Posted in Aesthetic Surgery, Oncosurgery, Trauma

Skin Grafts


Skin is the primary protective barrier of the body against external mechanical, chemical or biological threats. Raw areas created on the skin surface by a variety of causes exposes the body to the harmful effects of these threats. Thus it is important to make good any loss of skin or mucosal surface lost by burns, surgical procedures or other injuries.

In managing the defects both of skin and mucous membrane which follow excision of lesions, closure by direct suturing is used when the defect is small enough and is otherwise suitable. Likewise, small defects produced by burns or ulcerations are allowed to epithelise primarily. When the defect is too large, the potential methods of reconstruction are the use of a free skin graft, a local or distant skin flap, a composite flap or a free flap.

Skin grafting is the method commonly used to close superficial defects of skin and mucous membrane. Its main advantages are the technical ease of the procedure and the minimal donor site morbidity. During its transfer from donor to recipient site, a free skin graft is completely, even if only temporarily, detached from the body. While being so detached, such a graft remains viable for a limited period whose precise limit depends on the ambient temperature at which the graft is maintained. In order to survive permanently, it has to become re-attached, and obtain a fresh blood supply from its new habitat. The processes that result in its re-attachment and revascularisation are collectively referred to as ‘take’.

Anatomy of the human skin

The skin is a two-layered organ which forms the protective covering of the body. The outer layer is the epidermis, an avascular cellular structure, and the deeper layer is the dermis, which is essentially a meshwork of collagen and elastic fibres. The two are separated by a lamina, the basement membrane. Downward prolongations of the epithelium penetrate the dermis in the form of eccrine sweat glands and pilosebaceous units like apocrine glands. These specialised structures are collectively called the adnexa. The dermis is also penetrated from below by a network of blood vessels, lymphatics and nerves.

The epidermis

The epidermis has two main elements – the epithelial cells and the pigment system

The epithelial cells

The epithelial cells take the form of stratified squamous epithelium in which the cells, as they mature, move from the deepest layer, the basal layer, towards the surface. In the process of this, the cells become keratinised and flattened until at the surface, they finally desquamate. The thickness of the epidermis varies greatly in different body area, the thickest in the pressure zones and the thinnest in the eyelids.

The differences in the histologic appearance of different layers of the epidermis is caused by the progressive keratinisation of the cells. The structure of the cells alter as they move towards the surface, the nuclei being lost as the cells become part of the keratin layer. This regular orderly progression to maturation which takes place in normal skin from the basal layer of cuboidal shaped cells abutting on the dermis, to the final keratin layer of flattened cells on the surface, is called ‘orthokeratinisation’. The units of the cell line maturing in this way are termed ‘epidermal keratinocytes’, regardless of its stage in the maturation process.

The pigment system

The cells responsible for the production of pigment in the epidermis are the ‘melanocytes’. They are derived from the neural crest ectoderm, from where they migrate early in the foetal life to the epidermis. In the epidermis, they lie among the cells of the basal layer and the cells of the hair bulbs.

The melanocytes are triangular in section with numerous dentritic processes which run between the cells of the basal and supra-basal layers. The pigment which they produce (melanin) is synthesised as granules (melanosomes) in their cytoplasm. According to Wolff (1973), the granules move outwards along the dentritic processes from where they are transferred by phagocytosis into the adjacent keratinocytes. In these cells, the granules cluster together as a pigmented cap over the nucleus. An ‘epidermal melanin unit’ is formed by the melanocyte and the adjoining specific number (between 20 and 36) of keratinocytes subserved for it.

The melanin cuts down the transmission of ultraviolet light through the epidermis, thus reducing its damaging effect. Coloured skin has the same ratio of melanocytes to basal cells as white skin. Its darker colour is due to a greater amount of pigment concentrated in the keratinocytes.

The dermis

The structural basis of the dermis is provided by the interweaving network of collagen and elastic fibres. Its upper surface is irregular, with upward projections called dermal papillae which fit into corresponding irregularities made by rete ridge pattern on the deep surface of the epidermis.

The superficial layer, the papillary dermis has a fine, loose structure and contains fine collagen, elastic and reticular fibres which run vertical to the surface, supporting the capillary loops, lymphatics and terminal nerve fibres. The deeper layer is called reticular dermis in which the collagen network is coarser in structure and runs parallel to the surface of the skin. Its deep surface has a pattern of indentations into which the fat of the superficial fascia extends to surround the secretory coils of the sweat glands.

The basement membrane

This term refers to the lamina between the epidermis and dermis, though it is doubtful if it acts strictly as a membrane. It is not of much significance in skin grafting, as even the thinnest of split-skin grafts contain at least some amount of dermis.

The adnexa

This is the collective term applied to the specialised epithelial prolongations which pass from the epidermis deep into the dermis. They include the pilosebaceous units and the sweat gland apparatus.

The epithelial cells of the adnexal structures constitute a biological cell line distinct from the cells of the epidermis, called adnexal keratinocytes. The distinction is more by means of biological behaviour than histological. One example is their different reaction to solar radiation. Compared with epidermal keratinocytes, the adnexal keratinocytes are relatively resistant to the effects of such radiation. Notwithstanding the differences, there exists a congruence in the basic potential of these cells. Given an adequate stimulus like the loss of surface epithelium from burning or the cutting of a split-skin graft, the adnexal cells are capable of ‘dedifferentiating’ and reverting to their ‘original’ function, of becoming epidermal keratinocytes to resurface on the skin.

The pilosebaceous unit

The pilosebaceous unit is a complex structure incorporating the hair follicle and sebaceous gland. The hair follicle extends into the dermis to varying depths, even reaching the subcutis in the scalp, beard area, upper lip and eyebrows. Towards its deepest part, it expands to form the hair bulb. The hair is formed in the hair bulb, and grows upward through the epidermis to the surface. The hair is surrounded along its length by a multi-layered  sheath of follicular adnexal keratinocytes.

The sebaceous gland buds off from the side of the hair follicle. This has a lobulated structure and opens via a duct into the upper part of the follicle. The sebaceous glands are particularly numerous in the nasal skin and the naso-labial fold, and their prominence and activity in different sites vary greatly in different individuals. In the oral mucosa, they open directly into the mucosal surface. Similar direct openings of the glands into the surface is also seen in the tarsal plates of upper and lower eyelids.

The sweat gland apparatus

The eccrine glands are widely distributed over the entire skin, excluding some particular sites like the red margin of the lips. Structurally, each of these consists of a simple tube which passes downward from the skin surface deep into the dermis and ends in a coil. In some of the glands, the coil reaches a sufficient depth to lie in one of the projections of fat which extends up from the superficial fascia. More than half of the coiled element is secretory; the remainder forms part of the duct element. From the secretory portion, the duct initially continues as part of the coil, but later emerges from it to pursue a gently undulating course up through dermis until it reaches the epidermis. It spirals through the epidermis to reach the surface.

Indications for skin grafting

The common indications for skin grafting include

Large raw surfaces on skin /mucous membrane after excision of benign or malignant lesions.

  1. Gross loss of skin caused by burns and crush injuries.
  2. Non-malignant ulcers or granulating areas (e. g. diabetic ulcers and pressure sores), which have failed to epithelise after routine conservative treatment
  3. Raw skin /mucous membrane surfaces left at the flap donor site, not amenable to direct suturing.
  4. Raw mucous membrane surfaces produced by pre-prosthetic surgical procedures like vestibuloplasty.
  5. Raw bony surfaces produced by maxillectomy, eyeball exenteration, and excision of palatal lesions.
  6. Interpositional material in the treatment of temporomandibular joint ankylosis.

Types of skin grafts

The free skin grafts consist of the entire thickness of the epidermis and a variable amount of dermis. They are designated according to their dermal component as

  1. Full-thickness skin grafts, consisting of epidermis and the entire thickness of the dermis, and
  2. Split-skin grafts, containing epidermis and a variable proportion of the dermis. According to the relative thickness of dermis included, the split-skin grafts are further subdivided into thin, intermediate and thick grafts.

These various categories of grafts are not completely distinct from each other. They merely represent convenient reference points on a continuous scale of decreasing thickness from the full-thickness skin graft to the graft containing little more than epidermis. The real difference in practice is between the full-thickness skin graft and the split-skin graft.

The full-thickness skin graft, once cut, leaves behind no epidermal elements in the donor area from where resurfacing can take place; the split-skin graft leaves adnexal remnants, pilosebaceous follicles and/or sweat gland apparatus as foci from which the donor site can resurface. As a result, the donor site of the split skin graft heals spontaneously, and requires no care other than that usually accorded to any raw surface; the donor site of a full-thickness skin graft has to be closed by direct suture or, if it is too large for this, covered with a split-skin graft. This limits the size of the full-thickness skin graft which can usually be cut in practice. Extensive defects are split-skin grafted; the full-thickness skin graft is restricted to small defects.

The full-thickness skin graft takes less readily than the split skin-graft, and before it can be used, conditions have to be optimal. The full-thickness skin graft remains virtually at its original size; the split-skin graft tends to contract subsequently of circumstances permit, e.g. across a flexure. Within broad limits, the thinner the graft, the more it contracts secondarily. The stability of the graft depends on dermis, and the thicker graft stands late trauma better than the thin graft.

History of skin grafts in oral cavity

Moscovitz in 1916 treated sulcal scarring by the use of skin grafts placed over a mould which was inserted through a submental approach. 10 days later, an intra-oral incision opened the graft pocket into the mouth. Weiser (1918) and Pickerill (1919) used the intra-oral route for placing skin grafts. Gillies and Waldon (1920) and Kilner and Jackson (1921) sued he same technique for correction of post-traumatic scarring and pre-prosthetic problems. Pichler (1931) used grafts in maxillectomy operations to prevent cavity shrinkage and promote early healing.

Schuchardt (1952) suggested suturing the mucosal edge to periosteum and grafting only the periosteal surface to overcome the problems created by the soft tissue-based graft. Obwegesser (1964) and Rehrmann and Pelser (1965) combined a buccal vestibuloplasty using a skin graft with the surgical lowering of the whole floor of the mouth. Hopkins et al (1974, 1980) described the use of a mucosal transposition flap from the lower lip with a skin graft on the donor site and resection of the mylohyoid ridges.

Donor sites

The thickness, appearance, texture and vascularity of skin vary greatly in different parts of the body, and have a strong influence in the selection of donor site appropriate to a particular surgical situation.

Full-thickness grafts

The full thickness skin graft may be harvested from a wide selection of donor sites, the main criteria for selection being colour matching, vascularity and the available area of the donor skin.

Post-auricular skin

The posterior surface of the ear extending on to the adjoining post-auricular hairless mastoid skin makes the best donor skin when the face is being grafted. It gives a most excellent skin colour and texture match, and when replacing eyelid skin, is virtually undetectable. The vascularity both of the graft and the usual sites to which it is usually applied make it the easiest of full-thickness skin grafts to get to take. It is the smallness of area of skin available which limits the size of the defect which can be used to cover. The donor site is closed by direct suture.

Upper eyelid skin

In the adult, skin is nearly always available on the upper eyelid, and this can be useful particularly when the defect is of another eyelid. The match of colour and texture is outstandingly good. The area available is obviously limited, though the redundancy of the upper eyelid skin usually present in the older age-group, the group in which such grafts are more often needed, allows more skin to be harvested than one might expect.

Supraclavicular skin

The skin of the lower posterior triangle of the neck gives a reasonable colour and texture match used on the face although it is distinctly inferior to post-auricular skin. The larger area of skin is available but unless the neck defect is grafted with a split-skin graft, the increase is not sufficient to make it obviously useful. Grafting the neck creates a cosmetic defect of its own and this is likely to be particularly undesirable in the female where the donor area is often exposed. These adverse factors restrict its usefulness considerably and it is not often used.

Flexural skin

The antecubital fossa and the groin are both described as possible donor sites. The dermis is thinner than average, and the skin is mobile on the deeper tissues, but only a limited width is available unless a graft is used to cover the donor site. On the face, the cosmetic result is comparable to that of the supraclavicular skin.

In the antecubital fossa, even if the donor site defect can be closed directly, the resulting scar is very obvious, and if the closure is under much tension, hypertrophy of the scar is a hazard. Its use as a donor site is, therefore, not recommended now. The groin area is useful if a long narrow graft is needed, closure in such circumstances being relatively simple. Its main use is in hand surgery, particularly in managing flexion contractures.

Thigh and abdominal skin

The texture and colour match of thigh and abdominal skin grafted to the face is usually poor. The skin either stays extremely pale or becomes hyperpigmented relative to the rest of the face. An added deficiency is a loss of constantly varying fine play of normal facial expression, the grafted area taking on a rather mask-like appearance, due possibly to its thicker dermis.

Both these sites provide a source of skin for the palm of the hand. The thickness of the dermis in both sites, where there is no ageing skin atrophy, provides a good pad to take the necessary pressure when used on the sole of the foot. If a graft of any size is used, the donor site in its turn must be grafted and even when the donor site can be directly sutured, the scar usually stretches badly.

Split skin grafts

This graft has a much wider usage than the full thickness graft, and within limits, the surgeon is able to control its thickness and make use of that variable in its characteristics and clinical behaviour. The donor site is chosen in any particular instance by taking into account such factors as the amount of skin required, whether a good colour and texture match is possible, local convenience (as in grafting from forearm to hand with need for only one dressing), the necessity of having no hair on the graft, the cutting instrument available, the desirability of avoiding the leg in the aged or out-patient.

The sites usually used as donor sites are:

  1. the thigh,
  2. the upper arm,
  3. the flexor aspect of forearm and
  4. virtually the whole of the reasonably plane surface of the torso.

When these sites are not available or when all possible sites are needed, skin can be taken from:

  1. the other aspects of forearm and
  2. the lower leg.

Harvesting of graft

The full-thickness skin graft is cut with a scalpel while the split-skin graft, of whatever thickness, is usually cut with a special instrument.

Local anaesthesia for graft cutting

Local anaesthesia can be used for harvesting skin grafts by injection of the local anaesthetic agent or by the topical application of the local anaesthetic into the skin area being harvested.

The topical anaesthetic agent commonly used is a mixture of lignocaine and prilocaine. Both agents are very slowly absorbed into the superficial layers of the skin, with negligible absorption into the blood stream, and this allows larger areas to be anaesthetised, though the anaesthesia may not extent to the deeper part of dermis. The area of anaesthesia required is marked out on the skin, and is covered liberally with the anaesthetic cream. As a rule, at least half an hour should be allowed for sufficient absorption to permit the cutting of the graft. Although pallor of the skin area is often seen, it is not a reliable indicator either of the presence of anaesthesia or its surface extent. The patient has to be tested for both.

When the anaesthetic agent is being injected, the addition of hyaluronidase to the solution makes it possible to cut a reasonable size of graft readily. The exact amount of hyaluronidase which has to be used is not critical; 1500 IU added to 100 ml of anaesthetic solution works satisfactorily. The mixture diffuses rapidly and leaves a uniformly flat skin surface. The diffuse increase in tissue turgor also increase the dermal thickness and makes the area slightly more rigid, which is helpful while harvesting thin grafts.

Full-thickness graft

As the full-thickness graft is to be accurately fitted into the defect, a pattern of the defect to be grafted is made to ensure that the graft is at normal skin tension at its new site. Aluminium foil and polythene sheet are the materials most often used for making patterns. The pattern may be made before or after excision. If the defect is irregular, matching points may be tattooed with methylene blue on the defect, and on the graft before it is cut.

While cutting the full-thickness graft with scalpel, it should be carefully cleared of fat on its deeper surface. Time and care should be spent in the cutting of the graft so that no fat is left on its deep surface. This procedure requires both skill and care. Alternatively, the graft may be cut without special regard to the inclusion of fat, the fat being subsequently removed with scissors. This removal of the subcutaneous fat is not considered necessary in case the donor site is upper eyelid or post-auricular region.

The graft is easier to cut if the area is ballooned with fluid, usually I:200,000 noradrenalin. Using the pattern already made, the outline is marked on the skin with methylene blue, and then it is incised and undercut. It often helps to pull the skin of the graft taut over the knife with hooks so that the cutting is done blindly, largely by touch. Alternatively, the graft can be held turned back so that cutting is done under vision. This latter method is less precise and usually results in more fat being left on the graft.

Split skin grafts

Unlike full-thickness grafts, the cutting of split-skin grafts require specialised instruments. The instruments most commonly used are:

  1. The Humby knife and
  2. The power-driven dermatome.

The Humby knife

The instrument originally used to cut split-skin grafts was the Blair knife, which has a blade approximately 25 cm long. The difficulty to cut grafts of uniform thickness with this instrument prompted Humby to add a roller mechanism to it. Modified versions of the Humby knife, of which the Watson modification is the most popular, have since been produced. A scaled-down version called the Silver knife, which uses a razor blade to provide the cutting edge, is useful when only a small graft is required. The Humby knife can be used only on convex surfaces, but despite this disadvantage, its convenience makes it the most frequently used instrument for routine graft cutting.

The donor site most often used is the thigh. While positioning the patient, the leg is placed with the appropriate group of muscles relaxed so that by pressing the muscle group either medially or laterally, the maximum of plane surface is presented to the knife. The same principle may be applied to any donor site.

Graft thickness is controlled by adjusting the distance between the roller and the blade. Despite the presence of a gauge on the instrument, most surgeons assess the thickness by holding the knife up to the light to see the clearance between the blade and the roller. Clearance of a little less than 0.5 mm gives a graft of average thickness. This initial assessment is adjusted as necessary by watching both the graft as it is cut, and the bed from which it is being cut.

Ideally, the blade when cutting moves to and fro smoothly over the skin surface which does not move at all with the knife. Drag, resulting from friction between the blade and the skin, causes the skin to move with the blade, making the graft harvesting more difficult. It cannot be completely eliminated, but lubrication with liquid paraffin of the skin surface and the surface of the blade next to the skin reduces it considerably.

The direction and orientation of the cut depends on the convenience of the surgeon. A little in front of the knife and moving smoothly a fixed distance from it, a wooden board is held pressed down on the skin. This serves the dual purpose of steadying the skin and flattening it as the blade reaches it. The edge of this board is also lubricated with liquid paraffin so that the forward movement of the knife is in unison with the movement of the board.

Assessment of graft thickness

Although a setting of the roller is suggested, the surgeon must be prepared to modify it accordingly. The first few millimetres of the graft cut provides an initial indication of the thickness and the setting can be adjusted accordingly.

The translucency  of the graft is the main index of its thickness. A very thin graft is so translucent that the grey of the blade shows through. Opacity of the graft increases with increasing thickness until the full-thickness graft has the colour and appearance of cadaveric skin. A split-skin graft of intermediate thickness is of moderate translucency. The pattern of bleeding of the donor site gives a further indication of the graft thickness. The thin graft produces a high density of tiny bleeding points; the thicker graft gives a lower density of larger bleeding points.

The thickness of the skin from the point of view of clinical atrophy and graft cutting, and the presence of remnants of the adnexa from which the sites can heal, vary in different parts of the limb. In general, the skin of the lateral aspect is thicker than the medial, and distal is thicker than proximal. In the thigh when atrophy is clinically obvious, the lateral aspect should be chosen if at all possible.

The power-driven dermatome

The power-driven dermatome is a complex and fragile instrument. It consists of a rapidly oscillating cutting blade which is driven electrically or by compressed air. With the skin held steady, and lubricated with liquid paraffin, the instrument is able to move forward smoothly.

The main advantage of the power-driven dermatome is the ability to cut a graft of controlled width and accurately controllable thickness from almost any part of the trunk or limbs. It is also capable of cutting thin grafts much more consistently when compared to other instruments. The straight margin, and the uniform thickness of the graft which it cuts, means that one need not leave any quantity of skin between adjoining donor sites in the knowledge that the whole area will heal uniformly and quickly. This facility makes it practically possible to cut successive crops of skin from the same donor site, which is a valuable property when the skin is at a premium such as in extensive burns.

Storage of skin

By storage at a low temperature, skin cut in excess of current requirements can be preserved for later use as needed. The increase in the use of ‘delayed exposed grafting’ has greatly increased the need for storage. Within the temperature range 0 –37˚ C, the survival time of a stored graft is a function of its temperature; the lower the temperature, the longer the survival time.

The graft is wrapped in gauze moistened with saline and placed in a sterile, sealed container. Unless specially long survival (e.g. up to 21 days) is needed, the storage temperature is not of paramount importance, but it is generally considered that 4˚ C gives the best results.

Preparation of recipient site

Free skin grafts are applied either to raw surfaces surgically created, or at least surgically clean, or to granulating wounds. The practise of grafting and site preparation varies with the two types of surfaces.

The surgically clean surfaces

A completely dry field is essential before the graft is applied, since graft failure is most often due to the presence of haematoma. To achieve this, several measures are used. Time is the single most important factor in this regard. The steps of the operation should be planned to give the area to be grafted the longest possible time for the normal haemostatic mechanism to become effective. While waiting for bleeding to cease, the area should be left covered with gauze soaked in saline.

Only the actual bleeding point should be picked up by the mosquito forceps so that necrosis caused by the short fine catgut is minimal. Bipolar coagulation is a useful alternative. Once the defect is created, the continuous use of suction would keep the bleeding active. Even when a specific clot is to be sucked off, the suction nozzle should not actually touch the tissue.

After the graft is sutured in place, unless the graft bed is dry, it is a good practise to flush out under the graft with saline using a 20 ml syringe with blunt cannula, before the tie-over bolus is applied.

Granulating areas

Healthy granulations are flat, red and vascular, do not bleed unduly readily, are free from a covering surface film, and shows evidence of good marginal healing. Left ungrafted, granulations generally become more fibrous (less vascular) or oedematous. Infection tends to add to the difficulties of grafting.

Systemic antibiotics to which the colonising organisms are sensitive, are ineffective in eliminating them from a granulating surface. An antiseptic (such as chlorhexidene) applied locally is likely to be more effective. The presence of slough created a suitable environment  for continuing infection. Surgical excision is a rapid and highly effective method of eliminating it. In the process, excision of fascia is preferable to excision of fat. The Humby knife with the roller widely open may be used to excise both slough and heavily infected granulations.

Granulations, once clean and free of slough, should be grafted without delay. If this is not possible, the area should be kept adequately covered with a large and thick dressing. Crepe bandages may be used to exert pressure over the area. It is noticed that hydrocortisone ointment sometimes improves unhealthy granulations.

Application of the graft

The full-thickness skin graft, cut to its prescribed pattern, fits the defect accurately and is sutured edge to edge along its margin. Sufficient sutures are inserted to give as accurate edge apposition as possible, care being taken to avoid inversion of the edges. Sufficient sutures are left long to provide a snug tie-over, and the remainder are cut short. The spilt-skin graft is cut large enough to cover the defect with an overlap, and the sutures used to hold it in its overlapped position are left long to provide for the tie-over.

The application of a skin graft depend on whether the graft is being applied on the skin surface or inside the mouth and/or sino-nasal cavity, but in any site, two distinct techniques are available. In the first, pressure is applied to the graft; in the second, the graft is left exposed without pressure being applied to the graft.

The skin surface

Pressure methods

Pressure methods are preferable when the graft is small in area and are invariable when it is full-thickness in type. They are advisable even if the graft is split-skin when the defect is in an area which is inherently mobile (around eyelids and mouth), where the defect is markedly irregular in contour (the pinna), and when the defect is markedly concave (orbital cavity following exenteration). When grafting is carried out primarily (immediately following the creation of the defect) pressure methods are normally used since the pressure, apart from keeping the graft immobile, helps to achieve hemostasis.

The pressure is exerted on the graft surface by a bolus applied directly over the graft and further pressure may be added by the use of additional dressings and crepe bandage and/or Elastoplast. The pressure is not a necessary factor in the take of the graft, and is only a means of providing immobility of the graft and holding it in contact with the bed.

Various bolus materials are used – flavine wool (cotton wool prepared with flavine emulsion), cotton wool moistened with saline or liquid paraffin, cotton waste, and polyurethane foams are some examples. The bolus should be bulky and extent to the margins of the graft, the long tie-over sutures being then tied tightly over the bolus, anchoring the bolus and graft in a single mass. A layer of Sofra-tulle®  laid over the graft before the bolus is applied, might help to ease the first post-operative dressing.

In certain situations, the defect may have to be kept stretched while a split-skin graft is taking. This is to allow  as much skin as possible to be introduced into the defect to mitigate the effect of any subsequent graft contracture. The usual bolus materials are not rigid enough to keep the defect stretched and a bolus of dental impression compound may be used instead. Since it softens in hot water bath and hardens to rigidity on cooling, an accurate impression of the stretched defect may be made from it. The material with the graft draped over its surface is then inserted into the defect. The  sutures are placed in and along the margins of the defect and tied across the bolus drawing the defect over it and stretching it so that as much skin as possible is inserted. This technique has its main application in the reconstruction of upper eyelid defects.

The outer pressure dressing consists of the usual gauze, cotton wool and crepe bandage or Elastoplast. The bulk of dressing may be enough for immobilisation, but plaster of Paris should always be used if necessary to reinforce the dressings.

Exposed grafting

The exposed grafting was initially developed as a solution to the ineffectiveness of bolus grafting in areas which cannot be immobilised. In this technique, the graft is laid on the defect, without dressing of any kind, merely protected from being rubbed off, and allowed to attach by fibrin adhesion alone. While applying the graft, any air trapped under it should be pressed out and the skin allowed to overlap the defect margins. Fibrin adhesion occurs quickly and it helps the graft to tolerate minor movements without interfering with the process of graft uptake.

If exposed grafting is used primarily, control of all bleeding points is essential since pressure is not available to help hemostasis. Since this is practically difficult, a technique called ‘delayed exposed grafting’ is used, where the application of the graft is postponed until natural hemostasis has taken place, the skin being stored at a lower temperature in the interval. The waiting period (usually 2-5 days) is used to free the wound of all the residual blood clot, and the graft can be applied as soon as the surface has been cleared of clot. Late exposed grafting, allowing the wound to granulate, is another option, but it has little application in head and neck regions.

In using delayed exposed grafting, the most important factor to be taken care is that the surface of the defect should not be allowed to dry out. This is particularly important when the forehead and the scalp are the sites concerned.  For this reason, an occlusive dressing should be applied to the defect as soon as it is created.

Exposed grafting demands a degree of co-operation from the patient, and it has to be used with discretion in children. In the head and neck region, however, no elaborate instructions are usually needed; an explanation to the patient of the need for care is often sufficient.

Mesh grafting

The procedure of meshing the grafts considerably help to expand the area which an individual graft is able to cover. The graft, cut in the usual way is passed through an instrument from which it emerges shredded into a regular network of skin. An alternative is to manually create regular slits on the graft using a scalpel. Traction applied to the four corners of the graft expands the mesh, giving a considerable increase in area. Apart from the factor that it provides for a large surface area, meshed grafts also show a higher ease of graft uptake. The main disadvantage is the unpredictable cosmetic outcome, making it unpopular in cosmetically important sites. The chief indication of mesh grafting is to expand the extent of the area the graft is being used to cover.

Oro-nasal cavity

Pressure methods

In the mouth, the nasal cavity or the sinuses, pressure methods either involve the use of a bolus tie-over dressing or, when the bed has a firm bony base, the use of a dental appliance to exert pressure on the graft.

Bolus grafting

This method uses a tie-over bolus in the manner primarily designed for grafting on the skin surface. Unfortunately, with most materials, the bolus rapidly becomes soaked with saliva, food debris, bacteria etc. with resulting offensive smell. Polyurethane foam is considered to be the most suitable bolus material to be used in the mouth. After suturing the graft with sufficient overlap and leaving the sutures long, the surgeon himself compresses the bolus between his fingertips and holds it against the defect, while the assistant completes the tie-over sutures around the bolus. After the compressing finger is removed, the bolus expands to exert its pressure on the graft.

The overlap of the graft may become a difficult slough which may separate spontaneously, but more often it is trimmed off when the bolus is removed 7 days after its insertion. By the time, the part of the graft which is taken, is fixed to the bed. Any area of failure is left to heal spontaneously.

Bolus grafting is most suitable when used in obviously concave areas where mobility is minimal or where the graft bed can provide some stability, as in buccal mucosa or floor of the mouth. It is advisable to insert as much skin as the defect can accommodate to avoid subsequent tenting, and to compensate for graft contraction.

Dental appliance method

This method is used where treatment of the tumour involves resected part of hard palate and upper alveolus. The technique may be discussed under two headings – edentulous and dentulous patients.

In the edentulous patient, an acrylic dental plate is prepared pre-operatively. If the patient already has a denture which fits well, it can be used as well. This provides the basis of the splint which ultimately presses the graft against the bed. The plate fits the area untouched by resection, but a fresh mould of the post-resection defect is required. Dental impression compound and gutta-percha are the available moulding materials for this purpose. Multiple holes are bored in the part of the splint which correspond to the site of the resection at the time of making the splint. Then, a bolus of softened bolus material pressed against it extrudes through the holes, making the two into a single structure once the material cools.

After the defect is created, the splint, with its bolus heated and made malleable once again, is pressed hard into the defect. This gives a composite denture-splint which matches accurately the irregular contours of the post-resection surface. Fixed in this position, it holds the graft firmly against the defect. The methods used to fix the splint can be extra-oral or intra-oral.

Intra-oral fixation may be provided either by direct wiring to the upper alveolus or by wire suspension to the zygomatic arch. Direct wiring to the upper alveolus can be used when there is sufficient alveolus left after resection. Holes are drilled on the splint/ denture as planned and then, with the splint held in position, a curved bone awl is thrust through the upper alveolar bone corresponding to the site of the hole in the splint. A 0.4mm stainless steel wire is used to fix the splint/denture to the alveolar ridge. Zygomatic suspension is required when there is insufficient upper alveolus left after resection. For this method, cleats should be provided on the sides of the splint. With the splint in position, a wire is looped over each zygomatic arch and its ends brought into the oral cavity and fixed firmly to the splint. A splint wired intra-orally is tolerated well and is removed 7–10 days to allow inspection of the graft. Intra-oral fixation should be preferred whenever it is possible, as it is comfortable to the patient and convenient to the surgeon.

Extra-oral fixation is provided by attaching the splint to the skull through a system of universal rods or joints. For this, the splint/denture should be constructed with a metal plate inset in the midline on its anterior surface. Into this, the fixing rod may be screwed. The skull fixation is best provided by supra-orbital pins – rods with self-tapping screws attached to supra-orbital ridges on either side and attached to the intra-oral splint through a rigid connecting rod. Other alternatives, mostly of historical importance include the halo frame and plaster of Paris headcap.

When the patient has teeth, the making of dental splints is more complicated, and cap-splints are used. The teeth outside the line of the mucosal resection are cap-splinted. A screw attachment welded to the cap-splint carries an acrylic plate shaped to correspond roughly to the shape of the alveolar segment to be excised. A bolus of dental compound welded to this plate gives the final accurate splint needed to hold the graft in place. When cap-splints cannot be made or enough teeth are not present, a denture is made with holes to accommodate the remaining teeth. This can then be used as described for the edentulous patient.

The graft can be applied to the defect in two ways

  1. If the defect is markedly concave, the skin can be draped over the moulded dental compound so that when the splint is wired in position, it carries the graft with it. When this method is used, it is advisable to glue the graft (using skin glue) to the bolus to prevent it from slipping. The adhesion is lost over a few days, and so when the splint is removed at the first dressing, the two surfaces separate easily.
  2. When the defect is shallow, the graft can be sutured to the margins of the defect with the usual overlap. The dental splint with the bolus is then inserted and fixed in position.

Exposed grafting

The technical problem posed by exposed grafting inside the mouth is one of providing continuing contact and effective immobility for sufficiently long to allow vascularisation of the graft. This has been largely solved by the use of quilted grafting.

Quilted grafting

This method was described by McGregor in 1975. Quilted grafting is used in sites which are impossible to immobilise, and it finds its main application in defects of mobile parts of oral cavity, most frequently the side of the tongue. Its successful use has prompted many surgeons to make use of the technique in other intra-oral sites as well.

The split-skin graft is sutured with catgut to the margins of the defect with an overlap in the usual way, and multiple additional sutures are inserted through the graft and the underlying tongue muscle, anchoring them together, and giving the overall appearance of a ‘quilt’. Each quilting suture creates a point of immobile contact between the graft and the bed with a ‘mosaic of squares’, each sufficiently immobile to allow the graft to become vascularised.

As the blind insertion of the quilting sutures inevitably causes some bleeding beneath the graft, multiple slits should be made on its surface to allow for the escape of blood and oedema fluid.

The process of graft take

The graft initially adheres to its new bed by fibrin, and its immediate nutritional requirements are met by diffusion from plasma which exudes from the bed providing the so-called ‘plasmatic circulation’. This is quickly reinforced by the outgrowth of capillary buds from the recipient area to unite with those on the deep surface of the graft and re-establish a circulation of blood in the graft. This link-up is usually well advanced by the 3rd day.

Coinciding with the vascular link-up, the fibrin is infiltrated by fibroblasts which gradually convert the initial tenuous fibrin clot adhesion into a definitive attachment by fibrous tissue. The strength of this attachment increases quickly, providing an anchorage which allows the graft to be handled safely within 4 days. More slowly, a lymphatic link-up is added, and even more slowly, nerve supply is established although imperfectly and variably.

The processes most critical in graft take are revascularisation and fibrous tissue fixation. The speed and effectiveness of these processes are determined by the characteristics of the graft bed, the graft itself, and the conditions under which the graft is applied.

The graft bed

The bed on which the graft is laid must have a rich enough blood supply to vascularise the graft and also be capable of providing  the necessary initial fibrin anchorage.

Surfaces which show rapid and profuse outgrowth of capillary buds takes a graft readily. The capability of the surface to produce granulations is a good indicator of graft survival on that surface. The soft tissues of the face, muscle, fascia and fat are so vascular that they all accept grafts readily. Cartilage covered with perichondrium, bone covered with periosteum and tendon covered with paratendon takes skin grafts without difficulty. Bare cartilage and bare tendon cannot be relied upon to take a graft although if the area is too small, the vascularity of the surrounding tissue may be sufficiently profuse to allow the graft to bridge the area and cover it successfully. The bare cortical bone on the outer table of the skull and mandible lack sufficient vascularity to take a graft successfully. The hard palate, the surrounding maxillary bone, the walls of the orbit, the circum-orbital buttresses, and the bone of the diploë (after the outer table of skull is removed) all take up grafts readily. The dura mater, the mucoperiosteum and the mucoperichondrium are other surfaces which could be expected to take a skin graft successfully.

The influence of vascularity on graft take is best illustrated by the effect of radiation. A site with radiation injury is rarely capable of being successfully grafted, despite the fact that in the absence of such injury, it is routinely grafted without difficulty. Ideally, the excision should extent into the normal tissue beyond the radiation damage before the grafting is attempted but this is not always possible. A useful guide is provided by the amount of fibrosis and induration of the bed, and the amount of small vessel bleeding, compared at he time of excision with the amount expected if radiotherapy had not been given.

Any surface with sufficient vascular supply to support a graft has fibrinogen and the enzymes which convert it into fibrin in adequate quantities to provide the necessary adhesion, unless the surface is harbouring pathogens which destroy fibrin (e.g.- Strep. progenies and Ps. aeruginosa).

The graft

Skin grafts can vary both in their thickness and vascularity. These variables affect the revascularisation and consequently the ease of take of the graft. The number of cut capillary ends exposed when a thick skin graft is cut, is smaller than with a thin graft. Thus revascularisation is faster with thin grafts, and they tend to be taken easily. Nevertheless, the common head and neck donor sites have a rich blood supply, and even full thickness grafts from these sites compare favourably in their vascular characteristics with thin split skin grafts taken from elsewhere.

Conditions for take

Rapid vascularisation is the most important factor, and the distance to be travelled by the capillaries for the link-up needs to be as short as possible. The graft therefore has to be in the closest possible contact with the recipient bed. The most frequent causes of separation are bleeding from the bed resulting in haematoma, and tenting of the graft when used in concave sites.

The graft has also to lie immobile on the bed until it is firmly attached by fibrous tissue anchorage. The shearing strains which tend to make the graft slide to and fro and prevent capillary link-up are to be avoided.

The phenomenon of bridging

A graft may survive over bare cortical bone, tendon or cartilage, and even if separated from the bed by a clot, provided the area is small enough. In such circumstances, the graft survives solely by bridging, a phenomenon in which vascularisation takes place solely by capillary invasion from the graft bed. In most cases, bridging is strictly limited in area, and beyond this, the graft will not survive.

Healing of donor site

The cutting of a split skin graft leaves variable portions of the pilosebaceous apparatus and the sweat glands in the donor area and from these multiple foci, epithelium spreads until the area is resurfaced with skin. The pilosebaceous remnants are much more active as foci of epithelial regeneration than the sweat gland remnants. The donor site of a thin graft, with its full complement of cut pilosebaceous follicles, heals in approximately 7-9 days, while that of a thick graft, dependent virtually entirely on sweat gland remnants, may take 14 days or more. The quality of healed donor site skin derived solely from the sweat gland remnants is also poorer. Most grafts are of intermediate thickness and leave a percentage of pilosebaceous apparatus, and takes 9-14 days for donor site healing.

If the graft is so thick that no adnexal structures are left in the donor area, or infection in the area destroys any remnants which are left, the surface will granulate, and healing takes place by epithelial migration from the margins. This takes place very slowly and it is always advisable to split-skin graft such sites, particularly if fat is showing to any extent.

The skin of recently healed donor site looks more deeply coloured than normal, the colour slowly fading in time to leave the area paler than normal, often with areas of variation in pigmentation, indicative of the local minor variations in the thickness of the graft cut.

Donor site management

The management of the donor site is a very important, but mostly overlooked aspect of skin grafting. The problems are pain, the provision for an optimal local environment for the healing process, and removal of the dressings.

Pain usually settles in 3-4 days, and is often followed by itching. Although itching is a clinical indicator of satisfactory progress in healing, it causes more discomfort to the patient than pain and is more difficult to treat. Pain can be reduced by peri-operative application of topical local anaesthesia (in the form of jelly), or by impregnating the dressings with some liquid form of anaesthetic agent. Use of a long acting agent often gets the patient over the most painful period without the need to use potent analgesics.

The ideal donor site dressing would remain non-adherent during the healing phase, be absorbent, maintain a moist environment, and minimise the potential for bacterial contamination. Such an ideal dressing does not exist at present. Most cases are managed by dressing with Sofra tulle® (gauze impregnated with liquid paraffin and antibiotic), over which is laden absorbent gauze, the whole held in position with a crepe bandage. The dressing is removed after 10-14 days with care being taken not to disturb the graft.

Newly healed donor sites are often covered with a flaky keratinised layer, and the use of an emollient, non-irritant cream for 3-4 weeks is usually effective.

Mucosal grafts

Skin grafts have several disadvantages when used in oral cavity. They are

  1. Its colour and texture never match that of normal oral mucosa, although after several years the difference becomes less obvious.
  2. Unpleasant taste and smell in the absence of adequate hygiene, especially if adnexal structures are included in the graft.
  3. Poor adhesion of complete dentures when used in the maxilla.
  4. Scarring and discomfort of donor site.

In order to circumvent these minor drawbacks, mucosal grafts have been introduced to graft intra-oral sites. Mucosal grafts may be full-thickness grafts or split-mucosal grafts.

Full-thickness mucosal grafts

The concept of mucosal grafts was introduced by Peer (1955) who transplanted small areas of oral mucosa on to the conjunctiva. Lewis in 1963 deepened the anterior sublingual area using cheek mucosa, and Propper (1964) reported its application in periodontal surgery. Obwegesser (1965) and Steinhauser (1969) used cheek mucosa in maxillary vestibuloplasty. Robinson (1967) and Hall (1971) recommended keratinised masticatory palatal mucosa as the ideal tissue for denture support because of its similarity to attached gingiva, but the latter reported a troublesome ulceration beneath the denture in the healed donor site in the midline and recommended retaining the central palatine mucosa. Hall and O’Steen (1970) concluded that full-thickness palatal mucosa fulfilled the basic requirements of a skin graft by covering denuded soft tissue while the thick underlying connective tissue layer reduced contraction. The dissection of palatal mucosa was made by free hand between the lamina propria and submucous tissues, leaving the minor salivary glands, fat and neurovascular bundles intact. Guernsey (1973) also preferred palatal mucosa because of its resilience and toughness.

Dekker and Tideman (1973) showed that transplanted cheek mucosa tends to assume the appearance of normal mucosa of the edentulous alveolus. Tideman (1972) noted that taking full-thickness mucosa from the cheek caused post-operative trismus.

Split mucosal grafts

Steinhauser (1969) obtained a split mucosal graft using a mucotome developed from the dermatome (originally devised by Castroveijo in 1959). The advantages claimed for split mucosa included the rapid epithelialisation and scar-free healing of the donor area, the uniform thickness of the graft and reduced contraction at the recipient site. However, he concluded that because of its better stress-bearing capacity and because stability is more important than adhesion in lower jaw, skin is preferable to mucosa in lower labial vestibuloplasty.

Advances in skin grafting

Tissue-cultured skin graft

The development of epidermal culture systems has allowed skin grafting with sheets of cultured keratinocytes. This technique has recently been reviewed by Nanchahal and Ward (1992). It has been found to provide a high expansion factor in the management of burns (O’Connor et al –1981) and chronic ulcers (Leigh et al –1987). Allogenic cultured keratinocytes are rejected, and so autologous cells are necessary to provide permanent epidermal cover.

Application of cultured keratinocytes alone resulted in sloughing, blistering, scarring and wound contraction due to lack of dermal appendages. This led to the development of collagen substrate to support the keratinocytes for grafting. Now autologous keratinocytes are cultured with wide variety of dermal appendages like

  • Extruded collagen sheets.
  • Reconstituted collagen lattice
  • Fibroblast postulated collagen lattices
  • Collagen glycosaminoglycan substrates
  • Cadeveric dermis

In 1993, Kangesu et al used ‘kerato-dermal grafts’, prepared by combining autologous dermis with cultured keratinocytes, and reported significant improvement in the in vivo growth of the cells.

Donor site dressings

Though split skin graft donor sites have been traditionally dressed with non-occlusive dressings, recent evidence suggest that a moist wound provides a better healing environment. Dressings that provide wounds with a moist environment include semi-permeable films, semi-occlusive hydrogels and occlusive hydrocolloids.

Semi-permeable films are permeable to water vapour and gases including oxygen, but impermeable to water and bacteria. Semi-occlusive hydrogels, while having similar properties, possess an absorbent mechanism. The occlusive hydrocolloids are impermeable to gases, moisture and bacteria. According to Hutchinson (1989), the moist environment beneath these dressings do not encourage wound infection.


Skin graft as interposition material in TMJ ankylosis surgery

The use of skin grafts for arthroplasty dates back to Gluck in 1902. Skin grafts were first used in TMJ ankylosis surgery by Georgiade and Altany in 1957. In 1961, Franchebois and Souyris used a strip of de-epithelialised skin obtained from a full-thickness graft to cover the mandibular stump and obtained good results in 7 patients. In 1977, Popescu and Vasiliu described a full-thickness skin graft technique and reported a low rate of recurrence. Most failures were due either to insufficient availability of skin to cover the tip of the condyle or to the displacement of graft due to poor suturing. Recent studies by Meyer (1988), Kaban et al (1990), Clauser et al (1995) and Chossegros et al (1999) have reported around 90% success rate with inter-incisal width of more than 30 mm after one year follow-up. All these studies reported a low incidence of complications related to infection and inflammation.


Skin grafts have been used in a wide variety of clinical applications for a long time. The main indication is to provide a natural coverage to the raw areas left behind by surgical excisions, burns, ulcers etc.  A good clinical assessment and meticulous technique can provide an adequate coverage in such cases. Of late, the applications of skin grafting have grown into new arenas like pre-prosthetic surgery and TMJ ankylosis surgery. With the emergence of recent advances, one can now hope to provide the patient with a near-perfect natural wound cover.


  1. Fundamental techniques of Plastic Surgery and their surgical applications. 9th I.A. McGregor & A.D. McGregor. Churchill Livingstone 1995.
  2. Basic Principles of Oral and Maxillofacial Surgery. Vol. I. Peterson, Marciani, Indresano (eds.). Lippincott-Raven. 1997.
  3. Cancer of the Face and the Mouth: Pathology and Management for Surgeons. IA McGregor & FM McGregor. Churchill Livingstone 1986.
  4. Grabb and Smith’s Plastic Surgery. 5th S. J. Aston, R. W. Beasley, C. H. M. Thorne. Lippincott-Raven. 1991.
  5. Surgery of the Mouth and Jaws. J. R. Moore. Blackwell. 1985.
  6. Skin grafts. G. H. Branham, J. R. Thomas. In Facial Plastic Surgery. The Otolaryngologic Clinics of North America. Oct 1990. 23:5.


Posted in Aesthetic Surgery, Maxillofacial Trauma, Oncosurgery

Flaps For Reconstruction


Surgical practice routinely involves excision of body parts for treatment of pathologic lesions, producing defects of varying sizes. Defects may also be caused by other factors such as trauma, burns etc. Reconstruction of the lost body part is important in many respects. They include provision for cover, restoration of function and aesthetic rehabilitation.

A flap is defined as a tissue that is either transferred or transplanted with intact circulation. When vital structures are exposed in a complex wound or when reconstruction has significant functional or aesthetic consequences, a flap is generally required.

Reconstructive ladder

A defect may be managed by a wide variety of methods. The first objective in analysing a reconstructive problem is a correct diagnosis. The extent and type of missing tissue are assessed in order to formulate a plan for correction and reconstruction. In planning for management of defects of skin, mucous membrane and underlying structures, it would be prudent to follow what is known as the ‘reconstructive ladder’.

Small defects produced by burns or ulcerations may be allowed to epithelise primarily. Another option is closure by direct suturing when the defect is small enough and is otherwise suitable. When the defect is too large, the potential methods of reconstruction are the use of a free skin graft, a local skin flap, a distant skin flap, a composite flap or a free flap. Planning involves consideration of the simplest alternative followed progressively by the more complex, advancing up the ladder to the most complex. Progression from primary closure to skin grafts to local flaps to distant flaps and finally to microvascular free tissue transfers provides a set of options that can be applied to any reconstructive situation.



Allow wound to heal by secondary intention



Direct tissue closure


Skin graft



Local tissue transfers


Distant tissue transfers



Free tissue transfers.

General considerations

A skin flap in its basic form is a tongue of tissue consisting of the entire thickness of the skin plus a viable amount of the underlying subcutaneous tissue. It is transferred in order to reconstruct a primary defect and is inset into this defect. The transfer usually leaves a secondary defect, which is either closed by direct suture or covered by a skin graft.

An ideal flap should satisfy the following goals

  1. provisions of a suitable colour match to the surrounding skin of the recipient bed.
  2. assurance of a compatible thickness
  3. retention, or provision of recovery, of clinically perceptible sensory innervation
  4. attainment of sufficient laxity and tissue ablation such that mobile margins, as in an eyelid or lip, are spared retraction and deranged function
  5. assurance that the resultant suture lines of either primary or secondary defects are restricted to anatomic units and fall within natural skin lines.
  6. assurance that the reservoir from which the flap is mobilized is sufficiently lax to allow closure of the donor site resultant defect.

The flap may be raised from the tissues immediately adjoining, or very close to, the primary defect, in which case it is called a local/regional flap; alternatively it may involve the movement of tissue at a distance from the primary defect and is then called a distant flap.

When a local flap is transferred, movement takes place in the form of advancement, rotation, transposition or interpolation. Some flaps, at the time of transfer, are reattached to the body over their entire area, and the proximal end of such a flap, where it remains continuous with the adjacent skin, is referred to as its base.

With other flaps, the distal segment alone of the flap is inserted into the defect, its central segment and base remaining unattached. The base is then called the pedicle and the central segment is referred to as the bridge segment. These two, pedicle and bridge segment, act as the carrier and provide the channel for the blood supply of the distal segment. Once the distal segment establishes at the new site, which usually takes three weeks, the bridge segment is divided and either returned to its original site or discarded depending on the local situation. Insetting of the distal segment is then completed. In order to inset the distal segment, it is necessary to undercut its margin for a short distance.

While the pedicle of a flap usually consists of skin and subcutaneous tissue, it is occasionally reduced to its subcutaneous component, the distal segment alone, as an island flap, retaining skin as well as subcutaneous tissue.

When a distant flap is transferred, it is raised prior to transfer either single pedicled as a relatively long narrow tongue of tissue, or bipedicled as a strap of tissue with a pedicle at each end. Transfer to its destination is carried out in one of two ways –

  1. If the flap is relatively near its ultimate destination, it may be swung on its pedicle, following division of one pedicle if it was bipedicled, and waltzed to its destination.
  2. If it is at a greater distance from its destination, the flap is attached instead to a carrier, usually the wrist, on which it is conveyed to its destination.

Most local flaps have an axis around which they rotate or are transposed in the process of being transferred, and this is called the ‘pivot point’ of the flap. When it is at all possible, the bridge segment of any flap is ‘tubed’ in order to eliminate unnecessary raw surface and to reduce sepsis.

If it is felt that the blood supply of a flap would not be adequate for its survival if it were transferred straightaway, the circulation can be rendered more efficient by surgically outlining the flap. Such outlining is called ‘delay’. The term is also used for the procedure of surgically augmenting the blood supply of the flap.

According to McGregor and Morgan (1973), two distinct types of flaps can be distinguished according to their vascular characteristics and behaviour, each with a distinct geometry imposed by its vascular anatomy.

Axial pattern flaps

This type of flaps is constructed around a pre-existing anatomically recognized arteriovenous system. The vascular system running along its length makes it possible to construct a flap at least as long as the territory of its axial artery with minimal regard for consideration of its breadth. This factor also makes the flap more robust and better able to cope with any adverse circumstances that may arise.

Random pattern flaps

A random pattern flap has no pre-existing bias in its vascular pattern, and this lack places stringent limits on its dimensions, particularly in the ratio between its length and breadth. The degree of stringency placed on the dimensions of a random flap depends to a considerable extent on the richness of its subdermal vascular pattern.


There are a number of methods of classifying cutaneous flaps. Flaps may be classified by the arrangement of their blood supply (random vs. axial), by configuration (rhomboid, bilobed, elliptical etc.), location (local, regional and distant) ands by the method of transferring the flap (rotation, advancement, transposition etc.)

Local flaps

The local flap consists of tissue immediately adjoining the defect to be closed. It consists of skin and subcutaneous tissue. The local tissue may be dissected into a flap and mobilised to the defect by advancement, rotation, transposition or interpolation.

Advancement flaps

The advancement principle makes use of a linear configuration flap, raised and advanced to cover a rectangular primary defect which adjoins its distal end. Tissue transfer is achieved by moving the flap and its pedicle in a single vector. Advancement flaps may be categorized as single-pedicled, bipedicled or V-Y.

Single-pedicle flaps

Single-pedicle flaps are created by parallel incisions, which allow a sliding movement of the tissue in a single vector towards a defect. The movement is in one direction, and the flap advances directly on the defect. The flap is developed adjacent to the defect and one border of the defect becomes a border of the flap. Repair with an advancement flap involves both primary and secondary tissue movement. Complete undermining of the advancement flaps as well as of skin and soft tissue around the pedicle is important to enhance tissue movement. Bilateral advancement flaps are frequently combined to close various defects, resulting in ‘H’ or ‘T’ shaped repairs. It may also be used as an ‘island’ advancement flap.

It is virtually only in the face that the necessary skin laxity exists to allow advancement flaps to be used successfully. It works well in the repair of the defects of foreheads, helical rims, upper and lower lips, and medial cheek. Mucosal advancement flaps are also used for vermilion reconstruction.

The main drawback of this type of flaps is that a standing cutaneous deformity is created.

Bipedicle flaps

Bipedicle advancement flaps are used primarily for repair of large defects of the scalp. The flap is designed adjacent to the defect, and advanced to the defect at right angles to the linear axis of the flap. This leaves, apart from the dog-ear deformity, a secondary defect which must be repaired with a split skin graft. So, these flaps are seldom used now-a-days in the reconstruction of head and neck.

V-Y advancement flaps

In the case of V-Y advancement flaps, a V-shaped flap is pushed toward the defect. Thus the flap is moved into the recipient site without any wound closure tension. The secondary triangular donor defect is then repaired with wound closure tension by advancing the two edges of the remaining wound towards each other. Thus the wound closure suture line assumes a ‘Y’ shape, with the common limb of the ‘Y’ representing the suture line resulting from closure of the secondary defect.

V-Y advancement is useful when a structure or a region requires lengthening or release from a contracted state. It is particularly effective in lengthening the columella in the repair of cleft lip nasal deformities. It is also helpful in releasing contracted scars that are distorting adjacent structures as the eyelid or vermilion.

Pivotal flaps

There are three types of pivotal flaps – rotation, transposition and interpolation flaps. All pivotal flaps are moved towards the defect by rotating the base of the flap around a pivotal joint. The greater the degree of pivot, the shorter the effective length of the flap. This is because the pivotal point is fixed in position, and the base of the flap is restricted in pivoting around this point because of the development of redundant tissue at the base known as ‘standing cutaneous deformity (dog ear). Pivotal flaps must be designed to account for this reduction in effective length as they move around their point of pivot.

Rotation flaps

Rotation flaps are pivotal flaps that have a curvilinear configuration. Since the flap is being rotated to its destination, its ideal form is as the large arc of a circle of which the triangular primary defect is a small arc, with the flap and defect together making a half circle. They are designed immediately adjacent to the defect and are best used to close triangular defects. They are usually random in their vascularity but depending on the position of the base of the flap, they may be axial. Because the flap has a broad base, it vascularity tends to be reliable. Rotation flaps are useful in repairing medial cheek defects located near the naso-facial sulcus.

Large rotational flaps are particularly useful for reconstruction of sizable posterior cheek and upper neck defects. Incisions for the flaps are placed in a pre-auricular crease, and can extend for some distance along the anterior border of the trapezius muscle to facilitate rotation of upper cervical skin toward the area of the posterior cheek. A z-plasty at the base of the flap facilitates closure of the secondary defect.  Chin reconstruction often can be readily accomplished with rotational flaps. Smaller rotational flaps may also be used for repair of defects located in the glabellar area. Scalp defects may be reconstructed with one or more flaps.

Transposition flaps

In contrast to rotation flaps, which have a curvilinear configuration, the transposition flaps have a straight linear axis. It can be designed in such a way that one border of the defect is also a border of the flap, or alternatively, with borders that are removed from the defect with only the base of the flap contiguous with the defect. In its ‘classic’ form, this flap is a rectangle, usually near square, which is raised and moved laterally into the primary defect, previously triangulated in preparation for it. Such a transfer leaves a secondary triangular defect which is at least equal in area to the primary defect.

Transposition is the most common method of moving local flaps into skin defects of the head and neck. They can be elevated in a multitude of sizes, shapes and orientation and usually are random cutaneous flaps, but may occasionally be compound. It is a reconstruction option for small and medium-sized defects. Its main advantage is that the flap can be designed some distance away from the defect, and the surgeon can select areas of skin elasticity and redundancy. The chief drawback is the secondary cosmetic deformity, but this can be hidden in hear-bearing areas.

Interpolation flaps

The interpolation flap, though it is a pivotal flap with a linear configuration, differs from transposition flaps in that its base is located at some distance from the defect. Thus the pedicle must pass over or under intervening tissue, and must subsequently be detached in a second procedure. This is the greatest disadvantage of this flap. On occasion, the pedicle can be de-epithelised or reduced to subcutaneous tissue only and brought under the intervening skin, to allow a single-stage repair.

A common interpolation flap is the mid-forehead flap, which includes the median and paramedian flap. They are highly effective in midface reconstruction because of their excellent vascularity and superb skin colour and texture match. Apart from nasal defects, defects of the medial canthal region, upper and lower eyelids, medial cheek, melolabial region and upper lip may be repaired with mid-forehead flaps. They are thin and viable flaps and can be easily contoured.

Rhomboid flaps

First described by Limberg (1966) and named after him (Limberg flap), the distinctive feature of the rhomboid flap is the precision of its design, which involved both the shape of the defect and the size and shape of the flap used to fill it. The defect is made in the form of a rhombus (       ) in such a way that its shorter diagonal equals the four sides in length.

The flap is outlined by extending the line of the shorter diagonal, for a distance equal to its length. From the extremity of this line, again equal in length to the shorter diagonal, is drawn at an angle 60˚ to it. In this way, a flap adjoining the defect is enclosed, with a size and shape similar to the rhomboid of the defect. Both the lines extended from the rhombus for flap design can be directed in either directions so that the flap can be placed in one or either side of the defect, with a further possibility of designing two flaps on each side, making four potential flaps in all. In practice, the side and direction of the flap are determined by the tissue available to be included in the flap.

The flap is raised and swung into the defect. This leaves a secondary defect, which is then closed by direct suture. The area where the rhomboid flap has its greatest value is the temple, between the eyebrows and the anterior hairline, and the appropriate direction is along the ‘crow’s foot’ wrinkle lines lateral to the outer canthus.

Bilobed flaps

In some situations, a defect which, in other respects is suitable for local flap reconstruction, is somewhat distant from an area of availability or lies at such an inappropriate angle that a suitable flap cannot be designed. Zimany in 1953 described a ‘bilobed’ flap for such situations.

The overall outline of the defect plus the flap is a ‘cloverleaf’ – one of the outer leaves being the defect, other two the bilobed flap, sharing the same pedicle. In the transfer, the central lobe of the leaf is rotated into the primary defect. This leaves a secondary defect which, in turn, is filled by the remaining cloverleaf. The tertiary defect so left is in an area of tissue availability, and so may be closed by direct sutures.

Hinged flaps

Cutaneous hinge flaps (trapdoor, turn-in, turn-out flaps) may be designed in a linear or curvilinear shape with the pedicle based on one border of the defect. The flap is dissected in the subcutaneous plane and turned over on to the defect like the page of a book. The epithelial surface of the flap is turned downward to provide internal lining of a facial defect that requires both external and internal lining surfaces. The exposed subcutaneous surface of the hinge flap is covered by a second flap.

The vascular supply of a hinge flap is derived from the soft-tissue border of the defect. And consequently they have limited and often restricted vascularity. The dissection should proceed in such a way that the base is thicker than the distal portion of the flap. Hinge flaps are commonly used for repair of full-thickness nasal defects and to close mature sinofacial and salivary fistulae.


Regional flaps

The regional flaps are local cutaneous flaps which involves transfer of tissue from areas adjacent to the defect. The commonly used regional flaps in head and neck include nasolabial flaps, forehead flaps, glabellar flap, tongue flaps etc.

Nasolabial flap

A nasolabial flap consists of a finger of tissue lying astride the nasolabial skin crease. It can be based above or below (superiorly or inferiorly based); and its main use is as a transposed flap to reconstruct defects of the side of the nose and the upper lip, occasionally the lower lip and oral cavity.

The nasolabial site is one of the most consistently available areas of lax skin. The area of skin which is used as the flap and its extension upwards and downwards varies considerably depending on the site, size and shape of the defect it is designed to cover. The flap is raised as a finger of tissue with its width made to correspond to that of the defect. Within limits, the length of the flap is not significant in vascular terms. The secondary defect is normally closed by direct suture.

The distribution of hair on the face and whether hair on the flap is desirable or undesirable will determine the ultimate the ultimate details of placing the flap. Used for defects of the nose, the superior pedicle is almost invariable. The lip and intra-oral defects may be repaired by either superiorly or inferiorly based flaps.

A nasolabial island flap may be designed, which bridges intact skin to reach the side of the nose. This is used to reconstruct the lower half of the side of the nose, occasionally the alar base and the margin, and rarely the adjoining upper lip.

Peri-alar rotation-advancement flap

This technique, described by Webster (1955), addresses the problem of asymmetry of the upper lip caused by direct closure of elliptical defects of alar bases. It involves excision of a crescent of the skin and subcutaneous tissue immediately lateral to the ala of the nose and mobilisation of the skin of the adjacent nasolabial area from the underlying cheek and lip musculature. The crescentic defect is then closed directly and the effect is to rotate and advance nasolabial skin into the lip defect to be closed without creating asymmetry.

Forehead flaps

The forehead flaps transfer the skin area from the forehead, most often to cover a defect of the middle third of the face. The tissue transferred can vary from a relatively small area to virtually the entire forehead. The pedicles generally make axial use of the superficial temporal, supraorbital and supratrochlear vessels, entering from the margins of the forehead. The secondary defect of a forehead flap is closed by direct suture when it is small enough; where it is too extensive, it is split-skin grafted.

Flaps raised from forehead are based either laterally on the temporal region, or inferiorly on the supraorbital region. The flap based laterally uses part or all of the breadth of skin between the ear and the outer end of the eyebrow. The flaps raised with an inferior pedicle vary considerably in design, but there are two basic designs – a straight relatively narrow finger like flap passing upwards from the glabellar region, and a ‘sickle’ shaped flap which rises in upward direction but curves down on the scalp to run vertically downwards on the opposite side of the forehead.

The temporal island flap, parented by the laterally based forehead flap, is occasionally used in the resurfacing of the malar region and also in reconstruction of the eyebrows.

Forehead flaps generally provide excellent contour and texture matches in the sites to which they are usually transferred.

Glabellar flap

The glabellar flap acts by transferring the skin from the glabellar region to cover defects of the side of the nose where it adjoins the medial canthus and the cheek immediately below the canthus. It is constructed on the hairless area between the eyebrows and the adjacent forehead, pivoting around the region of the superior orbital foramen on the opposite side from the defect, and incorporating a significant portion of the supraorbital and supratrochlear arteriovenous systems. The flap itself is triangular in outline with the apex pointing upwards on to the forehead. When it is rotated into the defect, it leaves a triangular secondary defect in the centre of the forehead, and this is closed by direct suture.

A glabellar island flap may be used to reconstruct the canthal area or the adjoining side of the nose and the cheek. Its parent flap is the finger variant of the supraorbital flap, rather than the classical glabellar flap. The subcutaneous pedicle pass from the island downwards towards the glabellar region.

Median glabellar flap

This rectangular flap (Rintala and Asko-Selvaara, 1969), long in relation to its width, is placed vertically in the glabellar region. With its base approximately 1 cm above the medial end of the eyebrows, its rectangle extends down over the bridge of the nose. It uses the principle of straight advancement to redeploy the glabellar area of skin availability, and can be used to cover square shaped defects of the upper third of the nasal skin in the midline.  This method can be used only if the gap between the eyebrows is wide enough to allow a reasonably broad flap to be constructed.

Tongue flaps

Eiselberg used tongue flaps to repair oral defects in 1901. However, he gave credit to Gersuny at the end of the 19th century for the first use of a tongue flap to restore oral defects. Klopp and Schurter (1956), Bakamjian (1963) and Conley (1982) repopularised its use in the modern era.

The tongue flaps offer the advantage of adjacent and similar tissue for repair of intra-oral defects. An excellent axial and collateral circulation provides for flap viability. The lingual artery is the main vessel supplying the tongue. Anastomotic connections between the terminal lingual artery, the facial artery and the tonsillar branches of palatine artery are present. The branches that directly supply the tongue are the suprahyoid artery, the dorsalis lingual artery, the sublingual artery and the deep lingual artery. There are vascular arcades, which often perforate the midline of the tongue.

Dorsal based tongue flaps get most of their blood supply from an intact lingual artery. The rich collateral circulation of the tongue prevents tongue flap death as long as the base of the flap and its design allow for collateral circulation to develop.

A wide variety of tongue flap designs are possible, the common ones being anterior or posterior dorsal based lateral or midline flaps, posterior based lateral flap (Vaughn, 1983), posterior based bilateral lateral flaps, anterior based ventral flaps etc.

The common indications for tongue flaps include

  1. reconstruction of intra-oral structures following cancer excision,
  2. resurfacing of oral defects and
  3. closure of palatal fistulae in cleft palate and acquired OAF.


Distant flaps

Deltopectoral flap

The deltopectoral flap is an axial pattern flap, first described by Bakamjian in 1965. It is composed of fascia, subcutaneous tissue and skin, and can be used for soft tissue reconstruction of mandible and maxilla.

The flap design is horizontal and rectangular with a rounded end over the shoulder. Based medially on the anterior chest wall, the flap on its upper border follows the clavicle and along the lower border, it runs just above the male nipple and not lower than the anterior axillary fold. In females, the base of the flap should include the first five ribs. The base of the medial portion of the flap must be 2 cm from the lateral sternal border to preserve the perforating vessels from the internal mammary artery, which supply blood to the flap. The perforators enter the flap through the first four intercostals spaces.

The fascia overlying the deltoid and pectoralis major muscles should be elevated with the flap by sharp dissection. Several branches of thoracoacromion artery will have to be ligated when elevating this flap. After appropriate planning and marking, the skin, subcutaneous tissue and the muscle fascia are sharply divided. The flap is elevated with skin hooks, and sharply dissected from the underlying muscles. It is then passed on to the defect and sutured in place.

The deltopectoral flap cannot be used for intra-oral reconstruction without creating an oral fistula. This is a controlled fistula and is necessary to preserve the blood supply to the flap. After the flap is sewn in place, it usually needs to be tubed.

The flap is divided 3 weeks later and the fistula closed. Waiting 6 weeks before dividing the pedicle ensures better survival of the flap. If the flap as designed is not long enough to fill the defect, it may be delayed to gain additional length.

The deltopectoral flap is useful for most reconstructive problems of head and neck, especially when a large amount of skin is required for coverage. It has been used in the repair of defects in the cervical oesophagus, hypopharynx, oropharynx, base of the tongue, mandible, maxilla and skin of cheek, chin and neck. In situations where both mucosal defects and skin defects are to be corrected, the combination of an appropriate myocutaneous flap for mucosal replacement, and the deltopectoral flap for external skin replacement works well. This flap may be used in most cases without having to reposition the patient.

The main disadvantage of this flap is the need to form an oral fistula and close it at a second operation. The donor defect is also a cosmetic problem. Another potential hazard is the rare occurrence of distal flap necrosis.

Cervical skin flaps

Flaps of varying size, shape, site and direction have been designed which make use if neck skin for reconstructive purposes. It is possible in theory to raise a flap from virtually any site on the neck and transpose it upwards to cover a defect of lower face or oral cavity. Being random in type, such flaps are restricted in their useful length.

Side of the neck

The flaps making use of the side of the neck are generally horizontally designed, based anteriorly or posteriorly. Because of their limited application, they are rarely used and have not been systematised in design.

Occipito-mastoid based flaps

From a base which extends form the midline and even beyond the occiput and event o the angle of the mandible, these flaps pass in a generally downward direction curving slightly laterally. Based on the downward extension, they may be ‘nape-of-the-neck’ flaps or sternocleidomastoid flaps.

‘Nape-of-the-neck’ flaps

This random pattern skin flap was first described by Mutter (1842). It makes use of the neck skin lying above the trapezius muscle. The flap can be raised pedicled on the occiput and extended downward as required to the approximate level of the spine of the scapula. Swung on the upper pedicle, such a flap can be used as a transposed flap to reconstruct the lower face and the submandibular region.

It is not generally considered safe to raise the ‘nape-of-the-neck’ flap and transfer it immediately because of the high incidence of necrosis. Prior delay is advisable and this makes it suitable for the already existing defect rather than the fresh post-excisional one.

Sternocleidomastoid flaps

This is a single-pedicled random pattern flap (Bakamjian and Littlewood, 1964), based on the skin overlying the upper insertion of the sternomastoid and running downwards along the line of the muscle.  Pedicled above, it can be swung forward to be used both intra-orally and extra-orally.

Used as a skin flap, it has a bad reputation for necrosis (about 20%). Because of its vascular vulnerability, it is sometimes delayed before transfer. The incidence of necrosis can be greatly reduced by incorporating the underlying muscle as part of the transfer, thus making it a myocutaneous flap.

Apron flap

Described by Zovickian in 1958, this is a superiorly based flap which consists of skin from the submental region and front of the neck between the line of the carotids on either side. It is hinged along a line parallel to and just below the lower border of the mandible and is designed to be turned upwards along the inner side of the mandible to replace the mucosa of the anterior floor of the mouth and anterior part of lower alveolus. Problems resulting from the presence of beard skin led to its use as an island flap. This, combined with the compromise on vascularity following neck dissection, has made this flap unpopular.


Composite flaps

The composite flaps contain, apart from skin and subcutaneous tissue, underlying supportive tissue such as muscle, fascia or bone. Myocutaneous flaps with their excellent blood supply have proved to be very reliable in reconstruction following ablative surgery. In some instances, the underlying bone is pedicled on the muscle, thus making it an osteomyocutaneous flap

Muscle / myocutaneous flaps

Pectoralis Major Myocutaneous flap

The pectoralis major myocutaneous flap is the most useful flap for soft tissue reconstruction of bone and soft tissue defects of the mandible and maxilla secondary to cancer surgery. This flap was first described by Heuston and McConchie, who used it to reconstruct a chest wall defect in 1968. Ariyan (1979) demonstrated the vascular arrangement of the flap.

The pectoralis major muscle is a fan-shaped muscle on the anterior chest wall. It is bounded by the clavicle superiorly and the sternum medially. The lateral border forms the axillary fold. The superior fibres run parallel to the clavicle, and the inferior fibres run from the inferior border of the deltoid to the 5th and 6th ribs and lower sternum. The clavicular head of the muscle originates from the upper three ribs and the clavicle and inserts into the inter-tubercular groove of the humerus.

The pectoralis major muscle has two separate blood supplies. The thoracoacromial artery arises from the second portion of the axillary artery, and passes along the medial border of the pectoralis minor muscle and penetrates the clavipectoral fascia. It has 4 branches at this level. The pectoral branch is the largest, supplying the standard pectoralis major myocutaneous flap. The other blood supply to the muscle comes from the perforating branches of the internal mammary artery, which penetrates the muscle next to the sternum. The motor nerve to the upper half of the muscle is the lateral pectoral nerve, and that to the lower half is the medial pectoral nerve. Sensation is by the intercostals nerves.

The technique for flap elevation depends on the location of the defect. Most commonly defects of the head and neck require skin for lining the oral and maxillary cavity and bulk for filling bone and soft tissue defects.  Currently two specific PMMC flaps are commonly used. These are the PMMC island flap and the PMMC paddle flap.

For raising an island flap, measurement is from clavicle to the inferior margin of the skin island, the measuring tape rotated to the defect to arrive at the appropriate length of the flap. The skin island may be placed on any part of the muscle as needed. The skin incision is carried down to the fascia of the pectoralis major muscle. Sutures are used to secure the skin island to the fascia. Then the entire muscle may be elevated, or it may be divided lateral to the island, leaving the lateral portion intact. The muscle is elevated primarily by blunt dissection off the ribs, care being taken not to enter the chest cavity. Medial to the island, the muscle is divided 2 cm lateral to the sternum to avoid injuring the internal mammary artery. The muscle is divided medially up to the clavicle. Laterally it is necessary to divide the humerus attachment. After the flap is elevated, the skin between the upper portion of the chest and neck is elevated and the flap passed under the skin to the defect. The neck incision is closed over the muscle, and the chest skin is mobilised and closed primarily in most of the cases.

In cases where the island flap is not long enough to cover the defect, a paddle flap may be useful. Another variant is the bilobular or ‘Gemini’ PMMC flap, in which two separate skin islands are raised with the muscle for simultaneous replacement of oral mucosa and overlying skin.

The pectoralis major myocutaneous flap and its variants are the flaps of choice for most defects in mandible and maxilla. It has also been used for the reconstruction of pharngoeophageal area, the base of the tongue, the anterior skull base, midface, total nose and orbital defects.

The PMMC flaps are very reliable in their survival chances. Very good functional and cosmetic results have been obtained using PMMC flaps along with dynamic fixation of the remaining mandible after partial resections. The disadvantages of this flap are also mainly cosmetic, the loss of muscle being very noticeable especially in thin patients. With 18 other muscles assisting shoulder motion, functional problems are rare.

Forehead flap

The forehead flap in its most reliable configuration is a myocutaneous flap, composed of the frontalis muscle extending from the hairline to the eyebrows with the overlying skin and subcutaneous tissue. It was introduced by Carpue in 1816. Blair (1941), Moore and Byars and McGregor (1963, 1964) have used this flap for lesions of the gingival mucosa, buccal mucosa and lateral pharyngeal wall. It may also be used to reconstruct defects of hard and soft palate.

The flap may be designed with a dual or single blood supply. It makes use of the superficial temporal artery and the posterior auricular artery, branches of external carotid artery. It may be used as a hemi-forehead flap or a total forehead flap depending on the size and location of the defect. The total forehead flap should include both the vessels.

The base of the total forehead flap extends from the lateral canthus of the eye to a point 2 cm posterior to the ear, to include the posterior auricular artery. The distal extremity is a transverse line at the level of the opposite lateral canthus. The base of the hemi-forehead flap extends from the lateral canthus at the eyebrow to the root of the superior helix of the ear and distally, it ends in the midline. The width of both types depends on the distance between the eyebrows and hairline.

Both the hemi-forehead and the forehead flaps are elevated at the level of the pericranium. The donor defect is covered with a split skin graft.

Various methods have been used for placing the flap into the oropharynx.

  1. McGregor (1963) – the flap is rotated 180˚ laterally and entrance is gained into the oral cavity through a separate transverse incision.
  2. Hoopes and Edgerton (1966) – the flap is rotated 180˚ medially and placed through a tunnel created at the base of the forehead pedicle that lies between the flap and the zygomatic arch.
  3. Terz and Lawrence (1969) – the temporal fascia is incised and the flap folded medial to the zygoma to be transferred into the oral cavity.
  4. The flap is hinged downwards and inserted into the oral cavity through the posterior part of the submandibular component of the neck dissection incision.

The advantages of forehead flaps are its close proximity to the orofacial regions, excellent blood supply and firmness of tissue. The main disadvantages are noticeable the donor site defect, the need to divide the pedicle to close the oral fistula art a second operation, and the frequent complication of bleeding. Flap necrosis, both major and minor, is an occasional problem.

Temporalis flap

The temporalis flap was introduced in 1898 by Golovine. The flap is useful

  1. to obliterate skull base, maxillofacial and orbital defects.
  2. to close cerebrospinal fluid leaks, to cover dural tears secondary to trauma or cancer operations
  3. to reconstruct patients requiring midface augmentation for hypoplasia secondary to trauma, operation or congenital anomalies.
  4. to reanimate the face after injury or resection of the facial nerve.
  5. to reconstruct small intra-oral defects. The flap extends across the midline of the soft palate for repair of velum defects.

The temporalis muscle is a broad, fan-shaped bipennate muscle with two origins – deep one from the temporal fossa extending from the superior temporal line to infratemporal crest, and the superficial one from the deep temporal fascia. The insertion is onto the coronoid process and anterior ramus of the mandible. The deep temporal fascia invests the outer aspect of the mandible. The superficial temporal fascia lies on top of the deep fascia.

The blood supply of the deep temporal fascia is by the middle temporal vessel, a branch of superficial temporal artery. The temporalis muscle has a dual blood supply from the anterior and posterior deep temporal arteries. These vessels arise from the second portion of the internal maxillary artery.

The available length of the arc of rotation is estimated pre-operatively by palpating the superior extent of the muscle while the patient clenches his teeth. The incision is started in a skin crease anterior to the ear, and is extended superiorly toward the vertex, ending above the superior temporal line. The incision is carried down to the deep temporal fascia, and anterior and posterior flaps are developed above the deep fascia, until the entire muscle is exposed. Then the fascia is incised around the border of the muscle down to the calvarium. Elevation of the muscle is done in a subperiosteal plane.

The zygomatic arch may have to be divided to facilitate placement of the flap into the mouth. Additional mobility is gained by sectioning the coronoid process. A tunnel is created into the oral cavity by blunt dissection, and the flap is passed to the defect with the aid of traction sutures. Split skin grafts may be applied, but is not necessary.

The main advantage of the temporalis flap is its proximity to defects high in the oral cavity or on the face. Problem from loss of muscle function are minimal. It can support skin grafts, has a good arc of rotation, may be turned in different directions and is thin, providing for less bulky reconstruction.

The main disadvantage is the cosmetic deformity, though minimal, caused at the donor site. This may be corrected with autogenous or alloplastic materials, or camouflaged by hairstyle.

Platysma flap

The platysma flap was first used by Gersuny (1887) for reconstruction of a through-and-through cheek defect. In 1951, Edgerton described a lateral cervical island flap based on the platysma muscle for reconstruction of intra-oral defects. DesPrez and Kiehn (1959) reported the modified apron flap, which included the platysma muscle. In 1978, Futrell and colleagues reported the use of the platysma muscle as a true myocutaneous flap.

The platysma muscle lies deep to the subcutaneous tissue overlying the anterior and lateral aspect of the neck. Superficial cervical fascia separates it from sternocleidomastoid muscle, the great vessels of the neck and other underlying structures. The origin of the muscle is in the subcutaneous tissue just caudal to the clavicle and the acromion. Its insertion is just cephalad to the inferior border of the mandible. Laterally, it extends over part of the posterior triangle and sternocleidomastoid muscle. In the midline, it may merge at any point from the chin to the thyroid cartilage. Its function is to depress the lower lip.

The blood supply of the muscle and overlying skin was described by Hurwitz et al (1983) and Rabson et al (1985). The cervical skin is supplied by a random anastomosing network located superficial to the platysma. The principal vascular supply to the muscle is from branches of the facial artery. But it also receives rich blood supply from other vessels such as occipital, posterior auricular and superior thyroid arteries.

The flap may be raised on either a superiorly or inferiorly based pedicle. For use in the facial region, it must be raised as a superiorly based flap. If t is to be used in conjunction with neck dissection, it is elevated before the neck dissection is done. A skin island is designed on the inferior aspect of the muscle. Following an incision outlined to the platysma muscle, a supra-platysmal dissection is carried superiorly to the point of rotation. The incision through the inferior base of the skin island is carried deep to the platysma muscle and including the superficial cervical fascia. It is preferable to preserve as many blood vessels in the vicinity as possible.

The muscle is flipped 180˚ and brought through a tunnel into the mouth or is rotated and taken through a subcutaneous tunnel for coverage of extra-oral defects.  The neck is closed primarily.

The primary use of this flap is in the reconstruction of intra-oral defects of the palate, buccal mucosa, tongue, floor of the mouth and pharynx; and extra-oral defects in the cheek and lower lip region. Advantages of this flap include its close proximity, and a minimal donor site defect which can be closed primarily. This thin, pliable flap causes negligible impairment of functions like deglutition, speech or prosthetic appliance use. It can be used for mild facial augmentation and reamination following facial nerve injury.

The greatest disadvantage of the platysma flap is that it is not reliable because of its unpredictable blood supply, which can cause flap loss. Another problem is the folding of the muscle which causes a bulge in the neck. Also, it cannot be used in regions where tissue bulk is required.

Sternocleidomastoid flaps

The sternocleidomastoid flap was introduced in 1909 by Jinau for facial reanimation. In 1949, Owens described a compound neck flap that included the muscle. This flap can be used as a muscle, myocutaneous or a myo-osseous flap.

The sternocleidomastoid muscle has two heads which at the origin attach to the manubrium sterni and the medial third of the clavicle. The insertion is to the mastoid process and lateral third of the superior nuchal line of the occipital bone. It divided the neck into anatomical anterior and posterior triangles.

The blood supply to the muscle is through three arteries. Superiorly a branch of the occipital artery enters the muscle below the mastoid tip. The middle branch is from the superior thyroid artery. The inferior third is supplied by a branch from the thyrocervical trunk. The dominant vessel is the occipital artery.

The incision is determined by the proposed use of the muscle. If a muscle flap is to be used alone, a vertical incision over the midportion of the muscle extending from its origin to insertion may be used. As an alternative, two horizontal incisions can be placed. The flap can then be elevated by tunneling through these incisions. A myocutaneous flap can be developed by basing it either superiorly or inferiorly with a skin island or skin pedicle attached to the muscle. In cases of simultaneous neck dissection, the McFee or hockey stick incisions can be modified to accommodate its elevation.

For superiorly based flaps, the flap is developed by elevating the skin and platysma from the underlying sternocleidomastoid muscle. The sternal and clavicular head are transected and the muscle is elevated by dissection between its deep surface and the deep cervical fascia. The lower branches of the vascular supply may be ligated without affecting the survival of the flap, according to Ariyan (1979). But Sasaki (1980) and Marx & McDonald (1985) recommend maintaining two of the supplying branches, by dissecting the middle branch (superior thyroid) back to its parent vessel.

The muscle is separated from the fascia, taking care to preserve the spinal accessory nerve. Arc of rotation may be restricted by the superior thyroid artery or the spinal accessory nerve. The muscle is then transposed and sutured in position, and the incision closed in layers over suction drains.

When a skin island is being taken, the incisions are carried down to the muscle fascia, the skin sutured to the underlying fascia, and the dissection proceeded as for the standard muscle flap. After the flap is rotated into position, the muscle is sutured to the subcutaneous tissues, and the skin island to the mucosa or skin as required. Because of variability in the axial blood flow, fluorescein may be used intra-operatively to determine the amount of viable skin. 1000 to 2000 mg of fluorescein is injected intravenously and the flaps are observed.

Extra-orally, the sternocleidomastoid flap can be used form the cheek to the neck area, and intra-orally, from the palatal region to the larynx. Other uses of the flap are to provide soft tissue augmentation after parotidectomy, for facial reanimation in 7th nerve injuries, to obliterate the dead space around a bone graft and to provide a vascularised muscle bed in patients with poor recipient tissues.

Advantages of the flap are the close proximity to recipient site, good colour match, adequate bulk and minimal donor site morbidity. The disadvantages include flat neck deformity, disruption of cutaneous nerve supply of the neck and unreliability of vascular supply of the flap. Potential limitations include the need to resect the muscle as part of neck dissection and obese people with short necks. Complications include muscle atrophy and flap necrosis.

Trapezius flaps

Flaps from the shoulder and back have a long history. Mutter described the trapezius flap in 1842, which was named after him. Zovickian popularised these flaps in 1957. Several designs of the trapezius myocutaneous flaps have been used because of its triple blood supply.

The trapezius is a flat and triangular muscle that covers the superior posterior part of the neck and shoulder. It originates from the nuchal line of the occipital bone and the spinous processes of C-7 through T-12. It courses laterally to insert on the lateral third of the clavicle, the acromion and the spine of the scapula. It overlies the semispinalis and splenius capitis in the neck, and both rhomboid muscles in the back.

The main blood supply of the trapezius muscle is by a branch of the thyrocervical trunk, the transverse cervical artery. At the border of the muscle, the vessel divides into an ascending and a descending branch, which permits separate flaps based on the lateral and vertical portions of the muscle. The upper portion of the muscle in the neck is supplied by the occipital artery. So this part of the muscle can be used as a separate myocutaneous flap. The muscle is also supplied by numerous deep perforating vessels from the intercostal system.

Three flap designs have been described – the upper trapezius flap, the lateral trapezius flap and the lower trapezius flap. The upper flap is the myocutaneous version of the standard nape-of-the-neck flap in which a strip of underlying trapezius is raised along with the skin (McCraw et al-1979). The flap in the form of an island can be rotated to allow a skin paddle to replace a mucosal defect.

The lateral trepezius flap (Bertotti, 1980, Guillamondegui & Larson, 1981) is based on the transverse cervical arterio-venous system and is raised from more or less the same skin area as the upper trepezius island myocutaneous flap. Its anterior border corresponds to approximately to the anterior margin of trepezius and from there it extends backwards and downwards in general the direction of the spine of the scapula. In order to be certain of including the vessels in the flap both the muscle element and the skin should extend above the point at which the transverse cervical vessels disappear deep to trepezius. The pivot point of the transfer is the medial end of its feeding arterio-venous system. The first step is to dissect the pedicle at its medial end. The island of skin with the underlying muscle is raised from levator scapulae. The flap can be used in conjunction either with a radical or a functional neck dissection.

The lower trapezius myocutaneous flap may be used as an island flap or as a solid flap. The vessels must be identified in the neck and traced to the anterior border of the muscle. The anterior incision for the skin island extends along the anterior border of the trapezius. The muscle and its blood supply are elevated from the underlying structures. An appropriate skin island is cut and the underlying muscle is incised with care. The pedicle is then isolated and traced to its origin at the thyrocervical trunk. The island, now attached only by its pedicle, is rotated into the defect and sutured into place.

The transverse cervical trapezius myocutaneous flap is useful for repairing defects within the oral cavity, hypopharynx or skin of the lower cheek and chin. For it to be considered, the blood supply must be preserved during the neck dissection. The horizontal (shoulder) fasciocutaneous flap is useful for coverage of neck.

The flap is a ready source of supple skin of uniform thickness without excessive muscle bulk, which lends itself to more contouring during reconstruction. The donor area is usually hairless and may be closed primarily in most cases. Scars in this area are not obvious.

The main disadvantages are the short pedicle available, limited arc of rotation and need for patient repositioning to use the flap. This flap cannot be used with a McFee incision. Injury or resection of the spinal accessory nerve causes significant and painful shoulder problems. Failure to heal donor defects is a dreaded complication.

Latissimus dorsi flap

The latissimus dorsi myocutaneous flap is one of the most useful and most reliable flaps that have been described. This flap was used by Tansini (1896) to cover a mastectomy defect. The use of this flap for reconstruction of head and neck defects was first described by Quillen et al in 1978.

The latissimus dorsi muscle extends from the tip of the scapula to the midline of the back posteriorly, and to the iliac crest inferiorly. The anterior border extends on an oblique line from the axilla to the midpoint of the iliac crest. The muscle is triangular and originates from the spinous processes of the lower six thoracic vertebrae, the spines of the lumbar and sacral vertebrae, posterior iliac crest, inferior angle of scapula and the last four ribs. The upper an anterolateral borders are free. The insertion is into the humerus in the intertubercular groove.

The thoracodorsal artery, the terminal branch of subscapular artery (which arises from the axillary artery), is the dominant vascular supply. Its vascular arrangement allows the muscle to be split into medial and lateral flaps. The secondary vascular source is from a few perforators from the posterior intercostals and lumbar vessels.

To reach defects in the head and neck region, the skin island must be placed on the anterolateral margin of the inferior half of the muscle. This area has the least number of perforating vessels. The size of the island conforms to the size of the defect, and the location of muscle paddle is distal enough to reach the defect by rotating the flap.

The initial incision is along the anterolateral border of the muscle. The skin island is outlined and the incision carried to the muscle. Then a skin incision is made connecting the island with the axilla along the proximal border of the muscle. Lateral and medial skin flaps are raised as needed to expose the muscle. The muscle is separated from the underlying bones. As the dissection proceeds upwards, the thoracodorsal pedicle is identified in the undersurface of the muscle.

Several different routes to the head and neck are available depending on the location of the defect. The flap may be tunneled over the pectoralis major and clavicle but under the overlying skin; or under the pectoralis major but over the clavicle. Care should be taken to protect the artery and the skin island. The flap is delivered to the defect and sutured in place. The skin defect on the back may be closed primarily or split-skin grafted. Drainage catheters are used under the back flap and also in the neck region.

The latissimus dorsi myocutaneous flap has been used to repair defects of the entire cheek, hemiface, oropharynx, mandible, segments of cervical oesophagus, posterior scalp, pharynx, etc. Its major advantage is the large amount of available skin and the wide arc of rotation. The blood supply is reliable, the muscle thick and the skin hairless. The main disadvantage is the need to reposition the patient during the operation. The donor site skin grafts occasionally fail to take.

Prior surgery in the axilla or radiation to the axilla makes the use of this flap unwise. Also, patients who have undergone thoracotomy with latissimus dorsi transection, are not candidates for this flap.

Osteomyocutaneous flaps

The osteomyocutaneous flaps consist of a segment of underlying bone pedicled on myocutaneous flaps, and are used for simultaneous soft and hard tissue reconstruction.

Sternocleidomastoid with clavicle/sternum

This flap was first described by Conley in 1972. He used all or part of clavicle and a part of sternum with sternocleidomastoid muscle. The main indications are for defects secondary to traumatic mandibular fractures, mandibular osteoradionecrosis and mandibular defects following cancer ablation.

The technique involves raising the contralateral flap with clavicle or sternum or both to reconstruct the mandible. The flap is raised preserving the occipital, posterior auricular and superior thyroid artery.

The main advantages are the provision for one stage reconstruction, and the rapid and technical ease of elevation of the flap. The disadvantages include loss of protection of great vessels, and the minimal donor site morbidity. The obvious limitation for its use is the presence of contralateral positive neck nodes in malignancy, in which case the sternocleidomastoid muscle will have to be excised.

Pectoralis major with rib

This flap was first developed by Ariyan et al in 1980. It is based on success with pectoralis myocutaneous flap and good periosteal supply to ribs form the pectoralis muscle.

The pectoralis muscle has three heads of origin the clavicular head from anteromedial clavicle, sternocostal head from anterior sternum and six upper ribs and abdominal head is variable from the external oblique muscle. It is inserted into the lateral lip of the humerus and deep fascia of the arm.

Vascularised upper six ribs based on the periosteal blood supply can be harvested along with pectoralis flap for maxillary and mandibular reconstruction. Usually 5th or 6th ribs are harvested with the muscle flap.

This flap combines the advantages of the versatile and durable myocutaneous flap and the presence of osseous tissue for bone reconstruction. Large skin island with soft tissue bulk is available for oral and facial reconstruction. Its long pedicle allows it to be mobilised to greater distance.

The disadvantages include poor bone stock for mandibular reconstruction, limited vascular supply to the bone, chance of pneumothorax, poor contour stability and limited dental restoration potential and the significant donor site cosmetic defect.

Trapezius with spine of scapula

Trapezius osteomyocutaneous flap for mandibular reconstruction was first described by Panje and Cutting in 1980. Here, the spine of scapula is harvested along with trapezius muscle for oromandibular reconstruction.

It has the advantages of a versatile and durable myocutaneous pedicle in the muscular part, with a dependable vascular supply. Other advantages are the location of the flap close to the operating site, low morbidity of donor site bone defects less than 12 cm, provision for large skin island and improved bone stack in comparison to rib with pectoralis flap.

Disadvantages include the requirement for repositioning during the procedure for flap harvest, inadequate bone stock for defects larger than 12 cm, limited manipulation of bone relative to soft tissues and the frequent complication of shoulder morbidity due to acromion damage and denervation.

Latissimus dorsi with rib / iliac crest

Maruyama et al (1985) described the use of this flap with rib for reconstruction of a complete hemimandibulactomy defect secondary to osteoradionecrosis. The vascular supply of the rib graft comprised of perforating branches from the posterior intercostal artery and the periosteal supply from the thoracodorsal artery. Follow-up bone graft biopsy at 3 months revealed viable bone.

An innovative latissimus dorsi myocutaneous–iliac crest bone flap was described by Magi et al (1986). A segment of iliac crest was transplanted beneath the muscle 2 months before raising the flap. The composite flap was then used for mandibular reconstruction. The results were disappointing in their two patients.

Temporalis with calvarium

As early as 1890, Konig and Muller reported the clinical applications of calvarial bone transfers with osteocutaneous flaps. Watson-Jones in 1933 transferred calvarial bone on periosteal pedicles to repair depressed skull fractures. Conley (1972) designed an osteomuscular flap consisting of temporalis muscle and clavarial bone. The bone-muscle-fascial-periosteal flap was developed by McCarthy and Zide in 1984.

The temporalis flap has an excellent arc of rotation about the coronoid process. It can be rotated to reconstruct the defects of orbit, oral cavity and also of face.

The advantages of this flap include minimal morbidity of the donor site and large volume of bone with a curvature. It is claimed that the viability of the membranous bone is superior to the endochondral bone.

The flap’s weaknesses include its bulkiness, anterior mobilisation requirements, donor site volume defect and possible limitation of jaw movement. Choung et al (1991) introduced a bone–fascial–periosteal flap to overcome these limitations.


Free flaps

Microvascular techniques for anastomosis of small vessels allow the transfer of skin flaps from distant sites. Free vascularised tissue transfer is a major advance in maxillofacial reconstruction by providing tissue with inherent ability to heal. These flaps are particularly useful in cases where adequate tissue for reconstruction is not available in the vicinity of the defect, and where the recipient bed is not vascular enough to facilitate the take of an ordinary flap.

An artery and vein of adequate size and in proximity to the defect should be chosen for microvascular anastomosis. It is usually possible to choose vessels with adequate pedicle length for donor vessels to reach. Delicate handling of the vessels is essential to prevent damage to the intima. The recipient vessels may be sutured end-to-end or end-to-side to the donor vessels. Branches of the external carotid artery and internal jugular vein may be used as recipient vessels.


The small intestine has been used for reconstruction of cervical esophagus to provide a reconstructive pharyngeal conduit following pharyngolaryngectomy and for resurfacing the oral cavity defects.

A segment of proximal jejunum is harvested along with mesenteric vasculature, which is capable of providing a single arterial and venous pedicle suitable for anastomosis. After the harvest of jejunum the continuity of the intestine is restored.

The advantages of this flap for intra-oral reconstruction is that the transferred tissue has a mucosal surface which secrete mucous and that the vascular pedicle is of adequate caliber for anastomosis.

The disadvantages of this flap are that it requires abdominal operation and the operating period is prolonged.

Groin flap

This was the first free tissue transfer performed with micro-vascular technique. First described by Wood in 1863 and reevaluated by McGregor and Morgan in 1973 as an axial pattern flap. The skin overlying the Iliac crest and ilium are perfused by arteries which anastomose in the vicinity of the anterior superior iliac spine, the superficial and deep circumflex arteries and the superior gluteal artery.

The standard groin flap is based on the superficial circumflex iliac artery, usually a branch of femoral artery. Venous drainage is by a superficial and deep system of veins. Large area of tissue meassauring up to 24 by 16 cm can be successfully transferred. The donor defect is minimal. It can also be used as a de-epithelised flat flap for repair of soft tissue defects.

The disadvantages are a variable pattern of vascularity, short vascular pedicle and the excess bulk of the groin.

Lateral arm

The lateral arm provides a free flap based on the posterior radial collateral artery, which is a direct continuation of the profunda branchii. The flap is thin and pliable and consists of fascia and skin. The posterior cutaneous nerve of the arm accompanies the artery and can be transferred with the flap.

Several branches from the artery provide a periosteal blood supply to the humerus and can allow harvesting of a small segment of vascularised bone.

Latissimus dorsi

Described by Maxwell (1978), the flap consists of latissimus dorsi muscle and its overlying skin paddle. The dominant vascular pedicle is the thoracodorsal artery, arising from the subscapular artery (branch of axillary artery). The subscapular artery and its branches offer a variety of flaps suited for free tissue transfer. Venous drainage is by venae comitantes, which accompany the thoracodorsal and axillary arteries.

The flap offers a large amount of tissue with a good quality skin element, thus making it useful to fill large and full thickness head and neck defects. It is very reliable and is easy to use. The disadvantages are the bulk, risk of seroma formation and functional incapacitation in certain occupation groups (athletes and tennis players). The muscle bulk settles slowly over time as the denervated muscle shrinks.

Rectus abdominis

The rectus abdominis muscle can be transferred either as a muscle flap or as a myocutaneous flap. The muscle is supplied by the superior and inferior epigastric vessels. The larger pedicle is the inferior epigastric, which forms the basis of the rectus abdominis free flap.

This flap is very useful in cases where an extensive area of soft tissue cover is required. Alternatively, it is possible to take a small amount of muscle that contains two or three perforators, which supply a large area of skin. It has a consistent, reliable, long vascular pedicle with vessels that are easily dissected.

The removal of the muscle causes some abdominal weakness, and ventral herniation is frequent. The flap is often too bulky.

Dorsalis pedis

First described by O’Brics and Shanmughan in 1973, this fasciocutaneous flap transfers skin, superficial fascia from the dorsum of the foot using as its vascular basis the dorsalis pedis artery and the superficial veins, which pass proximally into the long saphenous system. The vascular pedicle of this flap can be dissected from the leg to a length of at least 10 to 15 cm. If additional venous drainage is required, the saphenous vein may be included in the flap.

It has been used both to provide skin cover and an intra-oral lining. Second metatarsal bone can be harvested along with the flap for mandibular and temporomandibular reconstruction.

Iliac crest – composite groin flap

The iliac bone along with its overlying skin paddle can be harvested as a free flap based on either the superficial or deep circumflex iliac arteries. This flap was first described for mandibular reconstruction by Daniel (1978). Taylor et al (1983) showed the superiority of the deep circumflex iliac arteries. The segmental nature of the vessels supplying the iliac crest allows maintenance of viability with multiple osteotomies. The deep circumflex iliac artery arises from the external iliac artery just above the inguinal ligament. The venous drainage of the flap consists of two venae comitantes which may unite to form a single trunk.

The flap provides an abundance of well-vascularised iliac bone (corticocancellous), which can be used to adequately reconstruct large mandibular defects including hemimandibular or even total mandibular defects if bilateral flaps are used. The donor defect is minimal.

The chief disadvantages are the risks of necrosis of the skin segment and abdominal wall herniation. The dissection is tedious.

Radial fore-arm flap

The radial forearm free flap, described by Soutar (1983), is a fasciocutaneous flap based on the radial artery. The venous drainage is dual; the paired venae comitantes accompanying the radial artery and the subcutaneous veins. These two systems communicate and either can be used to provide adequate venous drainage.

The entire fascia and skin of the volar aspect of the forearm can be used for microvascular transfer. The periosteum of the radius is supplied by a rich network of vessels from the radial artery, and a vascularised segment of radius lying between the insertion of the pronator teres and the radial styloid may be transferred with the flap.

The flap is extremely reliable, has a constant anatomy, and has large diameter vessels. A large amount of thin, pliable skin may be harvested. The blood supply to the bone and skin arises from a different system of perforators, thus permitting multiple osteotomies without compromising skin viability. Furthermore, the segmental blood supply preserves vascularity of all segments. This flap is most useful in mucosal defects of the floor of the mouth and anterior mandibular defects in combination with floor of the mouth defects.

The main disadvantage is the need to interrupt the radial artery and the resultant compromise to vascularity of the hand. The amount of bone available for use is limited to 10 or 11 cm in length. This limits its use to small defects.

Scapular / parascapular flap

The scapular and parascapular free flaps are based on transverse and descending branches, respectively, of the circumflex scapular artery, which is the largest branch of axillary artery. Each artery is usually accompanied by two venae comitantes.

The flaps are of fasciocutaneous type, though variable amounts of bone or muscle (serratus anterior) can be included in selected cases. Large cutaneous paddles based on either of the arteries may be raised. The descending branch provides multiple segmental vessels to the periosteum of the lateral border of scapula, allowing a vascularised segment of bone to be harvested.

The flap is relatively thin, the anatomy reliable, and a large area is available for transfer. It also allows a second large myocutaneous area based on thoracodorsal vessels to be harvested. The donor site can be closed primarily and the flap hidden in clothing.

The chief disadvantage is the need to reposition the patient. The dermis of the back is thick, and the scar tends to widen in course of time.

Fibula free flap

Introduced in 1975 by Taylor et al, the free fibula flap was one of the earliest osseous free flaps with extensive application in long bone reconstruction. In 1983, Chen and Yan described the vascular supply of free fibular osteocutaneous flap. Hidalgo (1992) reported the first application in mandibular reconstruction. Hayden (1992) described the nerve supply to the cutaneous paddle and advocated a neurosensory potential. He also demonstrated successful primary osseointegration with titanium implants in the fibula.

A long segment of fibula can be harvested as a free flap based on the peroneal artery. The blood supply is rich, consisting of both medullary and periosteal vessels, which allows osteotomies without jeopardising viability. The branches of the common peroneal nerve, the lateral cutaneous nerve of the calf or the sural communicating nerve may be harvested with the flap to provide a neurosensitised skin paddle.

The dissection is straightforward and provides vessels of moderate size. The bone is strong and can even be folded on itself to provide a double strut. Donor site morbidity is unusual as long as the distal 5 to 6 cm of fibula is left for ankle support.


Reconstruction of specific regions

The different techniques used in reconstruction of head and neck are broadly classified as follows: –

Intra-oral reconstruction techniques

Reconstruction of tongue

1.      Superficial tumors of tongue

  1. Primary closure
  2. STSG
  3. Healing by secondary intention

2.      Partial glossectomy without mandibulectomy

  1. Primary closure
  2. Local flaps eg. Nasolabial flap
  3. Regional flaps eg. Masseter flap
  4. Distant flap eg. Pectoralis major flap
  5. Free flap eg. Radial fore arm flap, Dorsalis pedis flap.

3.      Partial glossectomy with anterior mandibulectomy

  1. Free Osseocutaneous flap eg. Osseocutaneous medial forearm flap

4.      Partial glossectomy with posterior mandibulectomy

  1. Regional flap and mandibular swing
  2. Distant flap and mandibular swing
  3. Distant flap with reconstruction plate

5.      Total glossectomy + laryngectomy or laryngoplasty

  1. Regional myocutaneous flap eg. Pectoralis major flap.
  2. Free myocutaneous flap eg. Rectus abdominis flap, Latissmus Dorsi flap

Reconstruction of the floor of the mouth

  1. Reconstruction of the anterior floor of the mouth
  2. Reconstruction of the posterior floor of the mouth

Reconstruction of the buccal cavity

  1. Using temporalis myofascial pedicled flap
  2. Using buccal fat pad.


Closure of oro-antral fistula

Management of acute oro-antral communication

  1. Socket edge reduction and suturing
  2. Use of supportive packs or protective plate

Management of chronic / established fistulas or Large OAC

  1. Local flaps
  2. Buccal flaps
  • Advancement flap (Welty, Von Rehrmann & Berger).
  • Modified advancement (Laskin & Robinson)
  • Sliding flap (Moczair)
  1. Palatal flaps
  • Straight advancement
  • Rotational advancement
  • Hinging and Island flaps
  • Palatal submucosal connective tissue flap.
  1. Bridge flap
  2. Combined local flaps
  • Double flap
  1. Distant flaps
  2. Tongue flaps
  • Anteriorly based
  • Posteriorly based
  • Laterally based
  1. Temporalis muscle flap
  2. Buccal fat pad flap
  3. Osteoperiosteal flap
  4. Grafts
  5. Autogenous bone grafts
  6. Allografts
    • Gold foil.
    • Tantulum foil
    • Gold plate
    • Poly methyl methacrylate
    • Hydroxy apatite blocks.
    • Fibrin glue.

Extra-oral reconstruction techniques

Reconstruction of midfacial defects

Classification of mid-facial defects

Type I              Loss of midfacial skin only; buttress of the maxilla, orbital floor and palate intact

Type II             Partial maxillectomy with intact palate and orbital floor

Type III            Partial maxillectomy with resection of a portion of palate; orbital floor and Lockwood’s ligament remain intact

Type IV            Total maxillectomy and palatectomy; orbital support remains intact

Type V             Total maxillectomy and palatectomy with loss of orbital support or eye.


Midface defects produce significant functional and aesthetic consequences. Feeding and speech can be a problem. Oro-antral fistula and velopharyngeal incompetence can develop. Resection of Lockwood’s ligament may result in enophthalmos and orbital dystopia.

Type I defects

Patients with type I defects suffer from variable loss of soft tissues of cheek and lips. Bony framework is not affected. Palate and orbital floor remain intact.

If the defect is small and the surrounding tissue lax, primary closure may be possible. Small rhomboid flaps or subcutaneous pedicled flaps are used for superficial midface defects. Tissue expansion has been used successfully for superficial defects of cheek. This technique affords excellent colour and texture match with least amount of scar tissue formed.

For larger defects, regional or distant flaps like pectoralis major, deltopectoral, latissimus dorsi, temporalis and forehead flaps have been used with success.

Type II & III defects

The traditional method of reconstructing these defects is by skin-grafting the internal cavity and the placement of a maxillofacial prosthesis. The prosthesis usually serves as a denture and a palatal obturator, closing the oro-antral fistula and providing projection of midface. An adequate residual palatal arch and surrounding soft tissues are required to support the prosthesis. As the size of the defect increases, problems with stability often are exacerbated especially if the prosthesis must function as a support for orbital structures or reconstruct a missing cutaneous segment.

Some limited maxillectomy defects may be reconstructed using autogenous tissue. Choung et al described the use of ipsilateral or bilateral temporalis muscle flaps. Calvarial bone may be included in the flap, making it a myo-osseous flap. The zygomatic arch may be cut anteriorly and posteriorly so that the muscle is mobilised to its insertion on the coronoid process. The composite flap is passed into the oral cavity and sutured to the remaining septum or palate.

Type IV defects

Individuals with extensive defects are best served by reconstruction with regional or distant flaps to obturate palatal defects, to provide complete soft tissue coverage and to aid in retention of a prosthesis. The reconstuctive goal is to provide a healed wound, separation of oral and nasal cavities, support for intracranial contents and obliteration of maxillectomy defects.

A variety of pedicled regional flaps have been advocated for midface resurfacing. They include the deltopectoral flap, pectoralis major myocutaneous flap, forehead flap etc.

Free tissue transfers advocated to repair midface defects include the free omental flap combined with non-vascularised bone grafts, the free latissimus dorsi flap, the rectus abdominis flap, the free scapular fasciocutaneous flap and the free fibular osseocutaneous flap.

Each donor site has its own advantages and disadvantages. The rectus abdominis flap allows a two-team approach, thus reducing operating time. Both the latissimus dorsi and scapular flaps require a change in the patient position, but they provide long vascular pedicles and large volumes of tissue. By de-epithelising intervening segments of dermis between cutaneous paddles, the palatal, maxillary and orbital components of the defects may be simultaneously reconstructed.

Isolated soft tissue repair without bony reconstruction tends to lose midfacial projection and result in sagging. Coleman and Sandham noted that by preserving the angular artery to the tip of scapula, vascularised bone could be harvested along with the scapular flaps. This helps in closure of massive midface defects. The muscular component helps in the closure of dead space of maxillary sinus. The cutaneous portion is used to resurface face and palate.

Type V defects

When orbital floor and Lockwood’s suspensory ligament are resected, reconstruction should obliterate the orbital cavity and restore facial contour.

Ilankovan and Jackson described the split thickness vascularised calvarial bone either pedicled on the temporalis muscle or with a free flap based on superficial temporal artery, to reconstruct the floor of the orbit.

The temporoparietal fascial flaps have been used for orbital and eyelid reconstruction. The free vascularised forearm flaps have been used to reconstruct the orbital floor and provide overlying soft tissue.

In extensive maxillectomy defects, soft and hard tissue requirements are massive, and free tissue transfer is preferred. It offers the advantage of one-stage reconstruction without the constraint of fixed point of rotation observed in regional flaps.


Reconstruction of the lip

The successful reconstruction of the lower lip must meet some criteria.

  1. The reconstructed lip should be sensate.
  2. Retain sphincter or muscle function.
  3. Oppose vermilion to vermilion of the upper lip in a watertight continent seal.
  4. Allow sufficient space for food, dentures and so on.
  5. Acceptable aesthetic appearance.

Defect of up to one third of the lip can be closed primarily. Larger defects require tissue transfer, and the preferred donor site is adjacent cheek or upper lip.

Defect of 30% to 50%: –

In moderate-sized defects, reconstructive techniques that use lip tissue will yield an excellent result. Transfer of upper lip tissue (lip switch) pedicled on the labial artery (Abbè technique), preserving the oral commissure, or rotated around the commissure by the Estlander method can be accomplished.

Another method is Karapandzic technique of advancement rotation of segments of skin, orbicularis and mucosa – after division of other supporting muscles. This procedure redistributes remaining lip, yet preserves the motor and sensory function. The principal disadvantage has been the relative microstomia and the necessity of extensive circumoral incision and dissection.

Defects of 65% to 80%: –

A widely used technique is the advancement of the cheek tissue by the Webster- Bernard approach. Initial results are good; the continued chronic tension of the closure has culminated in a tight lower lip that functioned poorly.

A more satisfactory procedure for defects of this magnitude has been the Karapandzic lip rotation, although a microstomia is inevitable. Denture construction should be modified here.

The Karapandzic approach essentially is dissection of the remaining lower lip segment, modiolus (bilateral) and lateral upper lip tissue in the neurovascular pedicles and advancement of these components to reconstruct the lower lip deficiency. The method requires incisions extended laterally and transversely in adjacent cheek tissue, curving upward into the nasolabial folds and into the lateral upper lip toward the alar base. Once skin incisions are made, the residual depressions and elevations are divided, beneath which lie both the facial nerve and artery branches to be preserved. Similar incisions need to be made in the mucosa at the labial sulcus, extending into the labial mucosa.

After mobilization of the segments, closure of donor site defect in the cheek is done in a V to Y fashion. Microstomia produced can be corrected later with a lip-switch procedure.

Total resection of the lip: – (Karapandzic approach)

A massive resection of the lip, chin and mandible must be reconstructed with distant flaps and requires reconstruction of the lower lip as a separate unit. The transfer of composite flaps of skin and bone revascularised and the result allows for institution of radiotherapy in the early post postoperative period.

Reconstruction of the lip and chin as a single unit can be accomplished with a radial forearm flap, incorporating a plantaris tendon as a vascularised unit to provide support. The tendon is sutured into either modiolus (if still present). The provision of sensation to the reconstruction by suturing the antebrachial cutaneous nerve of the flap into the stump of the mental nerve has been included.

To effectively reconstruct the lower lip, one must provide not only skin and mucosa, but also functioning mimetic muscle. Dissection of a platysma myocutaneous flap with an extended muscle pedicle to include the cervical branches of the facial nerve would empower the muscle component. Bilateral flaps would also provide sufficient tissue for mucosa reconstruction.

Another use of a full thickness inferiorly based nasolabial flap or bilateral flaps as needed for lower lip reconstruction.


Defects of the head and neck region resulting from severe trauma and ablative procedures for neoplasms, and those associated with congenital deformities, can be functionally and aesthetically debilitating. The primary objective of the reconstructive surgeon is to restore a level of form and function that provides the closest approximation of the patient’s pre-disease state.

Reconstructive options for patients with large defects have significantly progressed over the last century. The early primary closure techniques with their poor aesthetic results have now been replaced by improved surgical techniques, including composite tissue transfer. Free tissue transfer has now enabled the surgeon to reliably reconstruct the defects in areas of inadequate tissue availability and regions of vascular compromised. Different types of flaps, if used judiciously and carefully, allows reliable aesthetic and functional reconstruction in specific situations.



  1. Head and neck microsurgery. William M Swartz & Joseph C Banis. Williams & Wilkins. 1992.
  2. Fundamental techniques of Plastic Surgery and their surgical applications. 9th IA McGregor & AD McGregor. Churchill Livingstone 1995.
  3. Use of flaps in reconstructive surgery of the head and neck. DM Morris, G Unhold. In Basic Principles of Oral and Maxillofacial Surgery. (eds.) Peterson, Marciani, Indresano. 1997.
  4. Cancer of the Face and the Mouth: Pathology and Management for Surgeons. IA McGregor & FM McGregor. Churchill Livingstone 1986.
  5. Use of local flaps for intra-oral reconstructive surgery. MS Block. In Basic Principles of Oral and Maxillofacial Surgery. (eds.) Peterson, Marciani, Indresano. 1997.
  6. Grabb and Smith’s Plastic Surgery. 5th SJ Aston, RW Beasley, CHM Thorne. Lippincott-Raven. 1991.
  7. Design of local skin flaps. WF Larrabee Jr. In Facial Plastic Surgery. Otolaryngology Clinics of North America. Oct 1990. 23:5.
  8. Soft tissue augmentation and replacement in head and neck. Otolaryngology Clinics of North America. Feb 1994. 27:1.
  9. Reconstruction of the mandible and oropharynx. Otolaryngology Clinics of North America. Dec 1994. 27:6.
  • Pedicled Osseous Flaps. AG Lane, CS Johnson, PD Constantino. In Augmentation Craniofacial Skeletal Augmentation and Replacement. Otolaryngology Clinics of North America. Oct 1994. 27:5.



Posted in Non Odontogenic Tumors (Malignant), Odontogenic Tumors (Malignant), Oncosurgery




Surgery is the oldest and continues to be the most reliable form of treatment of malignancy. The treatment of the neck in patients with squamous cell carcinoma of the (upper aero-digestive tract and other neoplasms of the head and neck region continues to be one of the most controversial issues in head and neck oncology.

Neck dissection is the removal of lymph nodes and lymph node bearing tissues of neck from the inferior border of mandible to the clavicle, as a treatment for head and neck malignancy.

The rationale of treating squamous cell carcinoma which has metastasised to regional lymph nodes is based on the fact that spread by tumour emboli is the norm and that in transit metastasises.

The resection of lymph node as part of the surgical management of head and neck cancer was initiated by Crile ( 1906) and consolidated in a definite manner by Hayes Martin (Martin et al 1951).

The goal in the management of malignancy is tumour ablation either preservation of all vital and uninvolved structures. Neck dissection is either a component of a primary resection or a procedure unto itself. Ideally, radiation controls the primary disease, while the neck dissection is reserved for unresponsive cervical disease. Depending on the extent and location of the disease the appropriate procedure is chosen.

Radical neck dissection has proved to be an extremely effective therapeutic measure and at present is probably the surgical procedure most frequently used in managing metastatic tumour suspected or actual in neck nodes. Here there is clearance of nodes from below the mandible to the clavicle involving submental, submandibular, jugular group of nodes and also nodes in the posterior cervical triangle. This makes the procedure a more multipurpose one.

In recent years, however, a feeling has grown among many surgeons (Bocca, 1975 ) that in certain circumstances radical neck dissection sacrifices structures unnecessarily and in an unselective manner. The structures concerned are the spinal accessory nerve, internal jugular vein and the sternocleidomastoid muscle. Of all these, the spinal accessory is the most important functionally, its resection resulting in drooping of the shoulder from loss of upper trapezius function. The effect of loosing sternocleidomastoid muscle on one side is cosmetic rather than functional. Concern with the effect of loosing these structures, particularly the spinal accessory, has resulted in fresh evaluation of the whole surgical approach to lymph node resections, with the aim of preventing the sequelae without compromising the therapeutic effectiveness. These modified versions of radical neck dissection have been collectively referred to as ‘functional’ or ‘modified’ neck dissection.


Neck dissection is indicated for metastatic disease in the superficial or deep cervical fasciae of the neck. Indications include

  1. Cervical metastasis whose primary has recovered on treatment.
  2. Cervical metastasis unresponsive to radiotherapy or chemotherapy.
  3. Anticipated primary resection contiguous with occult or overt cervical metastasis.


Neck dissection is contraindicated in disease beyond the superficial or deep cervical fascia. These include

  1. Poor surgical candidate.
  2. Rampant distant metastasis.
  3. Base of skull disease.
  4. Mediastinal or infraclavicular disease.
  5. Unresectable or uncontrollable primary disease.
  6. Extension into deep vital structures of neck.


For doing neck dissection one should have thorough knowledge of the surgical anatomy of the neck.

Neck dissection is actually a surgical dissection of the anterior and lateral neck done for the purpose of removing tumour and cervical lymph bearing tissues. A rational plan surgery for cancer of the cervical lymphatics is based on a thorough understanding and application of the detailed anatomy of the cervical lymphatic system, the deep cervical fasciae and the relationship of these to the muscles and neurovascular structures of the anterior and lateral neck.

The neck can be divided into – Anterior and Posterior triangles about the sternocleidomastoid muscle. The anterior triangle is bordered by anterior border of the sternocleidomastoid muscle and by the inferior border of the mandible. The anterior neck is divided into two equal halves by the vertical midline from the mental symphysis to the suprasternal notch. The posterior triangle is bordered by the anterior border of the trapezius muscle, by the posterior margin of sternocleidomastoid muscle and by the middle third of the clavicle.

Anterior triangle

The anterior triangle contains the cervical part of the aerodigestive tract – larynx and trachea, hypopharynx and oesophagus, thyroid and parathyroid gland; large neurovascular structure of carotid sheath; supra and infra hyoid muscles; associated lymphatic and neurovascular structures. It extends from the inferior border of the mandible cranially to the thoracic outlet of the chest caudally. The anterior triangle can be divided into submandibular triangle, submental triangle, carotid triangle and muscular triangle.

Submandibular triangle

The submandibular triangle is bordered superiorly by inferior border of the mandible, the two bellies of the digastric and the stylohyoid and mylohyoid muscles. The roof is by skin and platysma. This contains the submandibular gland, associated fasciae, lymphatic structures ( submandibular group of lymph nodes ), part of anterior facial vein , facial artery and the marginal mandibular nerve branch of facial nerve. The deep part of the submandibular gland loops between mylohyoid and hyoglossus behind the  posterior surface of the mylohyoid muscle.

When the skin flap is elevated in this region, the skin flaps stops at the lower border of the mandible since the skin and the platysma are leaving the operating field to overlie the facial muscles. The dissection is carried along the lower border of the mandible up to the attachment of the mylohyoid medially so that contents of the triangle can be dissected off the muscle. Submandibular gland is resected because of its associated lymph nodes.

Facial vessels ( Artery & Vein ) and the marginal mandibular nerve run in close approximation to the gland. The facial artery and vein come to lie alongside one another at the lower border of mandible, nearer the angle and it is here the vessels are sectioned. The veins stays superficial to the gland, running downwards and backwards deep to platysma and the mandibular division of facial nerve. It unites with the anterior branch of the retromandibular vein to form the common facial vein which joins with internal jugular vein. The artery passes upwards towards the mandible from the external carotid approaching the triangle deep to the gland, loops over it to come alongside the vein at the lower border of the mandible.

Carotid and Muscular triangles.

Omohyoid muscle demarcates the carotid and the muscular triangle. In the carotid triangle the carotid vessel lie superficial.

The lower and anterior part of the neck is the muscular triangle which contains the infrahyoid strap muscles of neck, the aerodigestive tract and the thyroid complex.

Posterior triangle

It is bounded by the borders of the Sternocleidomastoid muscle, Trapezius muscle and the middle third of the clavicle. The roof is by the platysma and the skin. The platysma cover is present only in the anterior part of the triangle, so dissection of the skin flap in the posterior part of the triangle is bit difficult. Floor is formed by prevertebral muscles – the scalene, levator scapulae and splenius which are covered by prevertebral fascia.

The contents of this triangle include fibrofatty lymphatic containing tissue, cranial nerve XI, the superficial and cutaneous components of the cervical nerve plexus and a host of small vascular bundles. The presence of which at surgery has in the past led to the description of this surgical area as “bloody gulch”.

The omohyoid present just above the clavicle in the lower part of the triangle marks out a small subclavian triangle. It has important surgical implication as in its depth are found the cervical and thoracic outflow of nerves and vessels into the axilla  ( the brachial plexus from the inter-scalene interval ) and the subclavian vessels arching over the first rib from the thorax to the axilla.

The transverse cervical artery a branch of the thyrocervical trunk runs from medial to lateral over the floor of the triangle, reaching the anterior border of the trapezius at the same point as the spinal accessory nerve. Transverse cervical vein is less constant in its relation. Artery lies between the fat that fills the triangle and the prevertebral fascia covering the floor of the mouth. The vein runs through the fat itself. In relation to omohyoid usually transverse cervical vein lies deep but sometimes it is superficial and joins the external jugular vein. Parallel to the transverse cervical vessels at the lower ends are the subclavian vessels.

The anatomical structures encountered during routine neck dissection are the sternocleidomastoid muscle, spinal accessory nerve, trapezius, internal jugular vein, carotids, hypoglossal nerve, cervical fasciae and the lymphatic of cervical region.


Sternocleidomastoid and Spinal accessory nerve

The sternocleidomastoid muscle is attached cranially to the lateral surface of the mastoid process of the temporal bone and to the lateral part of the superior nuchal line of the occipital bone. Caudally it divides into the sternal and clavicular heads to attach on the sternum and the clavicle.

It is invested by superficial layer of the deep cervical fascia, which splits into two laminae to cover the superficial and deep surfaces of this muscle. At the anterior margin of the sternocleidomastoid muscle, thickened fascia attaches to the angle of the mandible forming an angular band “ CHARPY’S BAND”. This band is tough and it is difficult to open into the interval between the sternocleidomastoid muscle , angle of the mandible and the parotid gland. This band should be sectioned.

Crossing the outer surface of the sternocleidomastoid muscle in its upper half is the external jugular vein, marking out its course from the angle of the mandible to its midpoint of the posterior margin of the sternocleidomastoid muscle. Within one finger-breadth cranial and parallel to the vein, the great auricular nerve crosses the posterior surface of sternocleidomastoid muscle from Erb’s point ( the midpoint along the posterior margin of the sterrnocleidomastoid muscle). From this point more nerves derived from the cervical plexus emerge and distribute the cutaneous part of anterior neck.

The spinal accessory nerve can be identified entering the deep surface of the sternocleidomastoid muscle 4cm or more below the mastoid process. It can be located at the Erb’s point just superior to where the greater auricular nerve surfaces from the deep neck. It can also be identified deep to trapezius which is about 2 cm finger width above clavicle. Its identification is important when preservation is planned as in modified neck dissection.

Functionally sternocleidomastoid muscle aids in elevation of the thoracic cage and shoulder girdle, fixation of the limb, aid in lateral flexion of the head to the shoulder and rotate the head to direct the chin upward to the opposite side. Trapezius is one of the several muscles that elevate the shoulder girdle and retract the girdle medially. So preservation of the spinal accessory is vital for this function.

Internal Jugular vein

This vein trunk is part of the complex of veins, arteries, nerves, lymphatics and lymph nodes which run the length of the neck in a trough between the prevertebral muscles behind and the laryngopharynx medially. From the skull base to the root of the neck the internal jugular vein receives venous tributaries from the parotid, floor of the mouth, occipital region and from thyroid. It has no posterior tributaries. Each tributary has to be contained by ligature for removal of internal jugular vein in radical or modified neck dissection.

Nodal tissues are found on the anterior, lateral and posterior aspects of the connective tissue sheath along the internal jugular vein. These are present posterolateraly to carotid sheath and these are removed along with internal jugular vein in radical neck dissection.

Although covered by lymph nodes internal jugular vein lies within its own fascia and advential covering. This helps in preserving the vein as in modified or conservative neck dissection.

Internal jugular vein drain the brain and the deep structures of the head and neck. So bilateral ablation of internal jugular vein results in venous stasis and massive cyanotic swelling of the face. Morfit and Perzik 1952 demonstrated a case with simultaneous internal jugular vein ablation. They noted that swelling and venous engorgement subsided following the development of alternative venous channels.

Carotid artery and hypoglossal nerve

The common carotid branches of in the anterior triangle of the neck into the internal and external carotid arteries. Internal carotid proceeds straight and unbranched to skull base into the carotid canal through the carotid sheath. The external carotid gives branches in neck. The superior thyroid artery arises below the greater cornu, the lingual artery arises below the level of digastric and the ascending pharyngeal arises posteriorly. The facial and maxillary arteries arise anteriorly deep to digastric muscle while occipital artery arise posteriorly. These arteries might some times need to be secured in the course of neck dissection. The ascending pharyngeal artery which arises at the junction of common carotid artery runs straight posteriorly to occipital region is usually not encountered in neck dissection.

The hypoglossal nerve ( Cranial nerve XII) emerges between the internal jugular vein and carotid artery to descend in the lateral groove between them, to the point above carotid bifurcation. Here it gives off the superior root of the ansa cervicalis ( the descendans hypoglossali of the ansa hypoglossi) and curves forwards across the lateral surface of both internal and external carotid artery. Then the nerve runs forward deep to stylohyoid and posterior belly of digastric and enters the posterior part of the submandibular triangle, where it lies on the lateral aspect of the hyoglossus muscle. It passes forward on the hyoglossus and disappears behind the posterior border of the mylohyoid muscle. It enters geinoglossus and supplies muscles of tongue except palatoglossus ( via Vagus ).

Digastric, Stylohyoid and Mylohyoid.

These determine the neck dissection in the upper neck ( submandibular area).

The digastric muscle is attached to the deep surface of the mastoid in the digastric groove as a posterior muscle belly, to lesser cornu of the hyoid bone as a tendon slip and to the deep surface of the mandible in the mental region as an anterior belly.

Associated with tendinous attachment of digastric to hyoid is the stylohyoid which splits it into two muscle bands.

All major structures crossing between the head and neck pass deep to the posterior belly of digastric-stylohyoid complex. These include the hypoglossal nerve, vagus nerve, carotid artery and internal jugular vein. Only the facial vessels and marginal mandibular nerve lie superficial in the submandibular region. This nerve has to be protected during neck dissection.

The suprahyoid muscles lying between the hyoid bone and the body of the mandible constitute a muscular framework of the mouth (Oral diaphragm). These include anterior digastric, geniohyoid and mylohyoid. They depress the mandible. Mylohoid is a sheet of muscle contributing to the floor of the mouth which is attached to the mylohyoid line on the inner aspect of mandible, the hyoid bone and to the muscle of opposite side by way of a midline muscular raphe.

The deepest part of submandibular gland and the duct enters the floor of mouth by passing around the posterior or free margin of mylohyoid muscle. In the loose connective tissue above the mylohyoid there is the deep part of the submandibular gland, sublingual gland and ducts and hypoglossal nerve.

The nerve to the anterior belly of digastric and mylohyoid muscles the mylohyoid branch of mandibular nerve is quite superficially placed in the suprahyoid region between the two muscles and can be injured during neck dissection. Its injury results in loss of muscular support to the floor of mouth and reduced tongue thrust in the first phase of swallowing.

Cervical fasciae.

The fascia and fascial planes of the head and neck were described in detail in the 1930s by Coller & Yglesius (1937) and by Grodinsky & Holyoke (1938). Fascia and fascial planes are important in neck dissection as these help in recognising proper tissue planes.

The fascia of the head and neck include superficial fasciae and deep fasciae. The deep fasciae is condensed in many layers as investing layer, pretracheal layer, prevertebral layer and carotid sheath.

Superficial fascia

The superficial fascia of the anterior neck contains the platysma muscle. A plane can be developed superficial to platysma muscle or a skin flap can be raised to include the platysma muscle in it.

Investing layer of deep fascia.

Deep to superficial fascia lies a discrete fascial lamina, the investing layer of the deep cervical fascia which invests all around the neck in a collar like fashion. This fascia attaches dorsally to the cervical spine, cranially to the occiput, mastoid and inferior border of mandible. It invests around the parotid and submandibular gland. It is also attached to the hyoid bone, clavicle and sternum caudally. It invests around the sternocleidomastoid and trapezius muscle. In the anterior suprasternal region it attaches in the anterior and posterior side of sternum creating an inter-fascial space ” space of burn” through which anastomotic superficial vessels cross the midline of the neck.

Carotid sheath and Pretracheal fascia.

The investing layer of the deep cervical fascia has deeper extensions of connective tissue which invest around the carotid artery, internal jugular vein and vagus nerve known as carotid sheath.

Multiple laminae of deep cervical fascia invest the suprahyoid and infrahyoid muscles and can be identified as thin muscle fascial laminae. The deepest of these are associated with thyrohyoid muscle, thyroid gland and trachea. They invest the cervical part of airway and referred to as pretracheal fascia or middle layer of deep cervical fascia. Lateral extensions of this fascia gives rise to carotid sheath.

The carotid sheath is continuos with the laminae of the neighbouring fasciae: that of pharyngeal wall, prevertebral muscles and fascia, sternocleidomastoid muscle and investing layer of deep fascia. Areolar connective tissue external to carotid sheath contains lymphnodes on the anterior, lateral and posterior aspects.

Prevertebral fascia

The deepest cervical fascia invests prevertebral muscles. It extends from base of skull to the tubercles of the transverse process of all cervical vertebrae and is continuos along the vertebral surface of the vertebral column into the posterior mediastinum. It is recognised as the investment for splenius capitis; levator scapulae; anterior, middle and posterior scalene muscles; longus colli and capitis muscles and for nerve structures. The cervical and brachial plexus are invested by the prevertebral fascia which form the floor of the posterior cervical triangle of the neck. Neck dissection is done superficial to this plane to avoid damage to the cervical, brachial plexus and phrenic nerve.

Lymphnodes and lymphatics of head & neck.

These structures are not visualised as such in a neck dissection. The procedure of neck dissection is designed to remove them as part of the overall specimen encased in the surrounding tissues, be it fascial covering, muscle or salivary gland.

Occasionally larger lymphatics are recognised as distinct channels in the lower part of the neck but more often they are noticed when they are divided and clear fluid is seen to well up into the wound. The thoracic duct is at risk when dissection is carried low down on the left side of the neck. Greenfield and Gottleib (1956) showed variations in the termination of the thoracic duct as in internal jugular vein, subclavian vein, external jugular vein, innominate vein and right internal jugular vein.

The cervical lymphatic system is divided into superficial and deep chains. Superficial lymphatics perforate the first layer of cervical fascia to empty into deep cervical lymphatic chain. Their involvement requires ablation of large area of skin.

The deep cervical lymphatics are important as they receive lymph from mucous membranes lining the mouth, pharynx, major salivary glands, thyroid and as well as the skin of head and neck. The deep cervical lymphatics accompany the internal jugular vein and their branches or lie within the major salivary glands. The deep cervical lymphatics along the internal jugular vein are grouped into jugulodigastric (subdigastric), juguloomohyoid (middle jugular) and lower / inferior jugular nodes. Lindberg & Jene 1972 showed that the most common areas involvement of the cervical lymphatic chain are the jugulodigastric or the junctional nodes of fisch. Internal jugular vein is sectioned between the upper and lower limits of the jugular group of lymph nodes and are removed along with the nodes.

Lymph nodes of the posterior triangle are designated as upper, middle and inferior cervical nodes or spinal accessory group of cervical lymphatics. From metastatic point, this group of lymphnodes receive drainage from the nasopharynx, in the upper portion it receives drainage from the subdigastric group of deep cervical lymphnodes.

In the submandibular area there are three group of nodes preglandular, interglandular and pre & retro vascular nodes. They drain the lower lip, cheeks, alveolar region, floor of the mouth and anterior tongue. They empty into the deep jugular chain of lymphatics. Preglandular nodes are present at the anterior to submandibular gland, interglandular lymphnodes are present within the gland while pre and retro vascular lymphnodes are present just anterior and posterior to facial vessels at the lower edge of the mandible. These group of lymphnodes are involved in metastatic cancer of oral cavity.

The lower parotid nodes are present in the lower pole of the parotid gland and may be involved secondarily to the preauricular parotid nodes in cancer of parotid gland or in cases of the cancer of the upper lip. So in neck dissection this tail of parotid gland is removed along with the inferior parotid lymphnodes in this area.


Superior jugular or jugulodigastric drain the posterior facial region including the tonsil. Middle jugular nodes drain middle part of aerodigestive tract and thyroid. Inferior jugular nodes drain the thyroid and cervical part of aerodigestive tract (oesophagus and trachea).

From surgical point of view lymphnodes of head and neck can be divided into five regions

Region I – contains the lymphnodes of the submental and submandibular region.

Region II, III and IV are the nodes that are present along the internal jugular vein and those within the fibrofatty tissue medial to sternocleidomastoid muscle.

Region II – corresponds to the upper third of jugular group and includes the upper jugular (jugulodigastric) and the upper portion of cervical lymphnodes found in close proximity to the spinal accessory nerve.

Region III & IV are divided based on omohyoid muscle.

Region IV includes nodes in lower jugular as well as scalene and supraclavicular lymphnodes that are located deep to lower third of sternocleidomastoid muscle.

Region V includes in the posterior cervical triangle along the spinal accessory nerve.

Staging of neck nodes

The TNM staging for classification of malignant tumours, classifies nodes based on their size and side of involvement. The neck nodes are classified into

NX    Regional lymph node cannot be assessed.

N0     No regional lymph node metastasis.

N1     Metastasis in a single ipsilateral lymph node, 3cm or less in greatest dimension.

N2     Metastasis in a single ipsilateral lymph node, more than 3cm but not more than 6cm in greatest dimension; or in bilateral or contralateral lymph nodes, none more than 6 cm in greatest dimension.

N2a   Metastasis in a single ipsilateral lymph node, more than 3cm but less than 6 cm in greatest dimension.

N2b   Metastasis in multiple ipsilateral lymph nodes, none more than 6cm in greatest dimension.

N2c   Metastasis in bilateral or contralateral lymph nodes, none more than 6 cm in greatest dimension.

N3     Metastasis in a lymph node more than 6 cm in greatest dimension.

Note: Midline swellings are considered ipsilateral.


Classification of neck dissection

A rational classification of neck dissection must take into account primarily the lymph node groups of the neck that are removed and secondarily the anatomic structures that may be preserved such as the spinal accessory nerve and the internal jugular vein. Medina ( 1989 ) based on these two points classified neck dissection into comprehensive, selective and extended.



Comprehensive neck dissection consist of the removal of all lymph node region ( I to V ) of one side of the neck. Included here are radical neck dissection and those modifications of radical neck dissections that were developed with the intention of reducing morbidity by preserving structures such as : internal jugular vein, spinal accessory nerve and sternocledomastoid muscle.

In Type I modified neck dissection spinal accessory nerve is preserved.

In Type II modified neck dissection spinal accessory and internal jugular vein are preserved.

In Type III modified neck dissection all the three structures are preserved.


Selective neck dissections consist of the removal only the lymph node groups that are at highest risk of containing metastasis according to the location of the primary tumour, while preserving the spinal accessory nerve, internal jugular vein and sternocledomastoid muscle. There are three types.

Lateral neck dissection consists of the enbloc removal of nodal regions of II, III and IV.

Supraomohyoid neck dissection and the expanded supraomohyoid neck dissection consist of enbloc removal of nodal regions I, II and III. In the expanded supraomohyoid neck dissection the nodal region of IV is also included.

The posterolateral neck dissection consist of the removal of the suboccipital and retroauricular lymph node groups and nodal regions of II, III, IV and V.

The extended neck dissections are any of the above neck dissections described above “extended” to include either lymph node groups that are not routinely removed, such as retropharyngeal or paratracheal, or structures that are routinely removed, such as the carotid artery or the levator scapulae.


There are various incisions used for neck dissections. Selection of incision for neck dissection is important to avoid complications such as

  1. Wound break down
  2. Skin flap necrosis
  3. Exposure of the carotid artery following neck dissection.


So it is essential to place incision properly being mindful of the vascular anatomy of the skin flaps to be created. Following Crile’s first radical attempt to use the ‘Y’ incision for neck dissection other kinds of incision types have been described ( Crile 1906, Martin et al 1951, Schobinger 1957, McFee 1960, Stell & Brown 1970 ). As the understanding of the vascular supply of the neck skin and clinical practices improved some incisions started to be preferred, however no particular incision is used uniformly.

The choice of incision type used for neck dissection depends on a number of factors. There are some fundamental specifications which are to be considered before placing a neck incision for neck dissection.

Conditions that are to be provided for ideal incision

  1. Incision should provide a good exposure for surgery. Thus by exposing all the important structures of the neck, the block dissection of lymph nodes in the laterocervical region can be done. Incision should help in reaching the primary focus easier.
  2. It should not destroy the vascular supply of the skin flaps. It should protect the skin and also should prevent skin necrosis by supplying maximum blood flow to the flaps.
  3. It should look good cosmetically.
  4. It should not limit the movement of the head and neck by forming contractures related to the scar.
  5. It should not lead to complications during the post operative recovery period or these should be minimised.
  6. It should prevent the damage that can occur after the operation to the anatomical structures especially the carotid artery.
  7. It should be capable of flap transfer for such cases as primary reconstruction or defect covering.
  8. It should be capable of tracheostoma and even pharyngostoma formation.


It is necessary to know the vascularization of the neck skin to avoid complications such as carotid artery exposure, pharyngocutaneous fistulae and flap necrosis following neck dissection.

Ellis 1963 demonstrated that the recovery of the neck skin is in a medial direction from the common carotid artery.

Kambic and Sirca 1967 stated that the arterial supply is in a vertical direction. According to this recovery of the skin on the anterior trapezius muscle is by descending branches of the facial and occipital artery while transverse cervical and supraclavicular arterial branches form the ascending branches.

Rabson et al 1985 in their study on cadavers, established that the arterial supply of the skin of the neck is multifaceted and that there are four arterial branches which pass from the platysma muscle through to the top skin’s surface. These platysma cutaneous arteries while supplying a particular region of the skin are also in anastomosis with each other.

Ariyan 1986 states that these anastomosis remain intact during neck dissection while the platysma is dissected from the skin.

Hetter 1972, Freeland and Rogers 1975 reported that there is alternative development of arterial supply even if facial, occipital and transverse cervical are ligated.

The vasculature of the can be summarised into

  1. The upper neck region anterior to the angle of mandible are supplied by branches of facial and submental arteries.
  2. Occipital and external auricular branches of external carotid supply the upper lateral neck, the area between ramus of mandible and the sternocleidomastoid muscle.
  3. The transverse cervical artery and suprascapular artery provide vasculature to the lower half of neck. The transverse cervical artery should be preserved unless there is oncological indication for ablation.

In addition there are large platysma-cutaneous branches and branches of superior thyroid  supplying the front middle portion of the neck.


The incision used in the dissection of the neck are generally classified into – vertical and horizontal. The combined procedure is also performed but the use of incisions in various directions will prevent sufficient arterial recovery of the flap. This increases the risk of flap necrosis in the tips of the flap.

Transverse incisions have cosmetic advantages as they follow the natural skin folds of the skin. Recovery of the scar in these folds are rapid and successful. Attie 1957 supported this view and became the first to practice this type of incision. Transverse incision are also easy to modify. This type of incision destroys the neck’s venous drainage from top to bottom and is major cause for flap separation (Freeland and Rogers 1975 ).

Vertical skin incisions in comparison to transverse skin incisions are regarded to be disadvantages, because they intersect the natural skin folds of the skin and the vascular supply of the neck. ( Attie 1957; McFee 1960; Stella and Brown 1970; Futrell and Cheretien 1976 ). The vertical incisions tend to contract along their long axes.  This lead s to deformity and restricted action. This can be prevented by giving a posterior sigmoid curve to the vertical incision ( McFee 1960; Futrell and Cheretien 1976 ).

Choosing a particular incision is important especially in the post irradiated neck area. In necks that have received radiotherapy vascular recovery in narrow angle flap’s corners and sides may be reduced. The flap necrosis ensures and may expose the carotid artery. For this reason trifurcation and vertical incision should not be practised in necks which has received radiotherapy. Here transverse incisions are reported to yield good results ( Grillo and Edmunds 1965; Stella 1969; Stella and Brown 1970 ).

The incisions used for neck dissection are

  1. Tri-radiate incision and its modification.
  2. Hayes Martin double ‘Y’ incision.
  3. McFee incision.
  4. Apron flap incision.
  5. ‘J’ incision.
  6. Hockey stick incisions & its modification.

Tri-radiate incision & its modification.

Crile (1906) stated this type of incision which gave sufficiently wide exposure to the operating field. The disadvantages of the tri-radiate incisions are flap necrosis due to disruption of the blood supply.

This incision has got a submandibular transverse incision and a longitudinal incision through the neck. The submandibular incision starts little beyond the midline near the lower border of the mandibular symphysis and it ends well back on the mastoid area, the two lines meet at the submandibular region at 120 0. The vertical incision starts from where the two lines meet.

In cases where curved horizontal incisions are used the incision continues as a smooth curve between the two extremities and the vertical incision starts from the lowest point of the curve.

According to Freeland & Rogers (1976) there is a horizontal watershed midway between the mandible and the clavicle. It will be ideal to place the horizontal incision below the carotid bifurcation to avoid carotid exposure when carotid blow out occurs. The vertical incision is placed little behind the carotid and continues down over the clavicle 3-4cm. This aids in wide exposure in the posterior triangle at the posterior-inferior angle.  The skin flaps are raised subplatysmally.


  1. This incision provides good exposure to surgical site.


1.Flap necrosis is high due to disruption of vasculature of skin flaps.

2.Occurrence of flap separation at the trifurcation site.

Modification of Tri-radiate incision

Many modifications have been described. Schobinger (1957), Cramer & Culf (1969) and Conley (1970).

Schobinger suggested that the vertical limb instead of being straight should be curved posteriorly in order to avoid lying directly over the carotids.

Cramer and Culf suggested a “S” shaped vertical incision for the same.

Conley suggested a posteriorly curving vertical incision rather than a horizontal incision. Here the incision starts from the submental region and ending by running downwards along the anterior border of trapezius to the level of clavicle gently curving posteriorly. The posterior part of the submandibular incision meets it at right angle approximately below the lobule.

Hayes Martin Incision

It is a paired “Y” incision. Here the submandibular component is met by a vertical limb which below becomes continuos with an inverted “Y” in the suprascapular region. Four flaps are thus created, the base of each extending to the limit of the neck dissection on each side.

The posterior flap with no platysma at its base is liable to have less adequate blood supply than the other flaps. This flap most often gets cyanosed.

The vascularity of the flaps can be put under less strain if the posterior flap is shortened by placing the vertical limb of the incision a little further back. Flaps can be broadened by raising the submandibular incision a little.

The advantages and disadvantages are same as Crile’s incision but flap necrosis and carotid exposure is more in this type of incision.

McFee incision

This incision differs from virtually all other incision as it avoids a vertical limb. Here two horizontal incisions are used one in submandibular region and other in the suprascapular region. Between these two a bipedicled flap is raised based anteriorly on the midline and posteriorly on the anterior border of trapezius. The flap is raised upwards to expose the lower part of neck until the dissection has proceeded far enough upwards to allow the resected specimen to be pulled through into the submandibular region.


  1. Excellent cosmetic result ( McFee 1960, McNeil 1978).
  2. There is no lessening of vascularity in the centre of the flap (Ariyan 1986).
  3. There is no angle intersection in incision (McFee 1960 ).
  4. Post operative wound recovery is rapid (McFee).
  5. It is suitable in necks receiving radiotherapy and in peripheral vascular disease (Maran et al 1989).
  6. Recovery of the flap will not fail despite the cessation of the ascending and descending vascular recovery due to wide bipedicled flaps. ( Stella & Brown 1970, Daniel & McFee 1987 ).


  1. The exposure is not good. ( Hetter 1972).
  2. It is not suitable for bilateral simultaneous neck dissection ( Chandler and Ponzoli 1969).
  3. The operating period is long (McFee 1960).
  4. Posterior triangle dissection is difficult ( Maran et al 1989, White et al 1993).
  5. Difficulty may arise while working under the bridge flap.
  6. In short neck it might be difficult to distinguish between the front tip of the incision from that of the tracheostomy.

Apron flaps

This flap was described by Latyschevsky and Freund 1960. Here only a horizontal incision from mastoid to mentum gently curving inferiorly up to upper border of the thyroid cartilage is used.

While in bilateral neck dissection Freund 1967 described an apron flap extending from one mastoid to other.

According to Freeland and Rogers1975 study on the vasculature of the head and neck showed better safeguard of the vessels of neck flap in the superiorly based apron flap.


  1. Carotid artery is well protected.
  2. Protects the descending arterial recovery.


  1. It will damage the ascending arterial and venous recovery.
  2. Venous congestion and oedema might develop at the bottom corner.

“J” incision

Eckert and Byars 1952 practised the first forms of the “J” incision. They extended the classical thyroid necklace incision laterally to the trapezius muscle and then slightly to the mastoid area. Grandon and Brintnall 1960 popularised this type of incision.

Acar et al 1999 reported excellent result in 320 cases.

“J” incision has a vertical incision extending from the mastoid apex to 2 cm above the clavicle along the back of the sternocleidomastoid muscle, then it is extended in horizontal plane and continued parallel to the bottom edge of the cricoid cartilage and it is stopped at the point where the sternocleidomastoid muscle is attached to the clavicle.

In situation that required bilateral neck dissection, the incision continued symmetrically and was extended across to the mastoid apex.

The skin flap is elevated at the subplatysmal plane.


  1. Good exposure to surgical site.
  2. Carotid artery is well protected.
  3. Suitable in bilateral neck dissections.
  4. It can be modified easily.
  5. Less chance for occurrence of pharyngocutaneous fistula due to coverage of the top sutures used for closing hypophyranx and oesophagus.
  6. Flap vascularisation is good due to its posterior base.
  7. Suitable in necks which receive radiotherapy.
  8. Cosmetic result are good.
  9. Tracheostomy can also be performed with this type of incision in a horizontal segment.


  1. Contracture in the vertical top section.
  2. Flap separation in the horizontal flap section

Hockey stick incision

Lahey et al 1940 first described a kind of hockey stick incision (HIS) for resection of thyroid gland carcinoma. It has been modified for radical neck dissection and functional neck dissection by Eckert & Byars 1952; Attie J. N. & Brooklyn N. Y. 1957; Bocca E, Pinataro O & Sasaki C T 1980 and Ariyan 1986. Rush 1965 and Gratz et al 1004 extended this incision to combine it with block resection of the oral cancer lesions.

Hockey stick incision has got a longitudinal and transverse incision. The longitudinal portion runs from the mastoid process downward, 1-2 cm behind the anterior border of trapezius muscle and curves gently at the acromioclavicular junction. The transverse incision runs medially towards the sternum approximately 1cm below the clavicular margin.

This incision permits the elevation of a medio superiorly based skin flap.

Bilaterally placed hockey stick incision permits the elevation of a superiorly based large skin flap and allows the deglovement of the whole neck. This is used when tracheostomy is planned along with bilateral neck dissection. In order to make incision for tracheostomy the transverse part of the incision is displaced superiorly and there by run across the anterior lower neck. This incision results in deglovement of the whole neck for providing access to the oral cavity.

  1. Omura et al 1999 reported good access and cosmetic result with reverse hockey stick incision for neck dissection and access to oral cavity while hockey stick incision provided good exposure and allowed tracheostomy to be performed through the same incision.


  1. It has good exposure.
  2. It has good cosmetic result as the incision gets covered in the hair line and clothing.
  3. Viability of the hockey stick incision flap is good.


  1. Access to submental area is not adequate. The flap has to be stretched and retracted in a upwardly direction.
  2. Access to oral cavity is also not adequate.

Reversed hockey stick incision

A reversed hockey stick incision was first described by Schobinger as a long anterior skin flap for radical neck dissection and was extended to block resection of oral cavity by Babcock & Conley and Disanayaka. There is no three point trifurcation in hockey stick incision and reverse hockey stick incision and also they provide good exposure. Here the flap is raised inferiorly.

Here the incision is placed in submandibular region. The transverse incision curves gently 2-3 cm below the tip of mastoid process and runs medially towards the submental region approximately 2cm below the lower margin of the mandible.

Modifications of reverse hockey stick incision

  1. Upwardly extended reverse hockey stick incision – here the transverse incision of the reverse hockey stick incision is extended upward to the median of the lower lip resulting in the splitting of the lower lip. This provides excellent exposure to the operating field for the management of oropharyngeal cancer.
  2. Contra laterally extended reverse hockey stick incision – here the transverse incision is extended towards the conta lateral mastoid process. This incision is preferred for bilateral neck dissection (ipsilateral radical and conta lateral supraomohyoid neck dissection ) and access to oral cavity.


  1. It provides good surgical exposure.
  2. Better cosmetic result.


  1. The descending arterial recovery may breakdown (Hetter 1972).
  2. The front and the tip of the medial flap might become necrotic however this weakest point is behind the carotid artery bifurcation.


Radical neck dissection

This operation is defined as the en bloc removal of the lymph node bearing tissues on one side of the neck from the inferior border of the mandible to the clavicle and from the lateral border of the strap muscles to the anterior border of the trapezius muscle, included in the resected specimen are the spinal accessory nerve, internal jugular vein and the sternocleidomastoid muscle.


The first description of this systematic en bloc removal of the lymphatics of the neck was published by Crile in 1906. He stated that the removal of the internal jugular was essential due to intimate relationship of lymph nodes with it.

The routine removal of the spinal accessory was advocated by Blair and Brown (1933), who believed that the nerve has to be removed to decrease the operating time and increase the certainty of total neck node removal.

Martin (1951) championed the concept that a cervical lymphadenectomy for cancer was not adequate unless all the lymph node bearing tissues of one side of the neck were removed and that this was not possible unless the spinal accessory nerve, the internal jugular vein and the sternocleidomastoid muscle were included in the resection.


  1. In multiple clinically obvious cervical lymph node metastases, particularly when they involve lymph nodes of the posterior triangle of the neck and these are found to be closely related to the spinal accessory nerve.
  2. In large metastatic tumour mass or there are multiple matted nodes in the upper portion of the neck

Surgical procedure.

The initial step is to raise the skin flaps to the margins of the dissection, that is the lower border of the mandible, the clavicle below the midline in front and the anterior border of trapezius behind.

The anterior margin and the clavicular margin are readily defined. The submandibular skin flap requires to be raised with care to avoid damage to the marginal mandibular branch of facial nerve. It runs in loose areolar layer deep to platysma and should be looked for, identified and traced out forward and backward out of the operating field. It is usual at this stage to divide the facial vessels as they cross lower border of the mandible. The submandibular prevascular and retrovascular lymph nodes are in close proximity to the nerve and they must be carefully dissected out from it.

The posterior margin is much more difficult to define because of the consistency of the tissue filling the posterior triangle and the laxity of the trapezius muscle in anaesthetised patient. it is ideal to visualise the anterior border of the trapezius from end to end for subsequent clearance.

Clearance of the neck is usually begun from the posterior triangle in the region of the posteroinferior angle. Here the floor is deeper than rest of the triangle and  hence makes it difficult to dissect. Once beyond 3cm above the angle on the posterior margin of the dissection, it is safe to deepen the plane to bare the prevertebral muscles.

At this stage the posterior triangle of neck can be cleared up to the meeting point of sternocleidomastoid and trapezius at the upper angle of the triangle. In the process of this dissection about 5cm above the clavicle the accessory nerve will be cut en passant.

As the floor of the posterior triangle is cleared towards the posteroinferior angle the transverse cervical vessels are met passing across into trapezius that can be resected as a part of the dissection.

Dissecting, along the lower border towards the posteroinferior angle, the tissue immediately above the clavicle is incised as far laterally at the trapezius exposing the sternocleidomastoid medially and passing more deeply laterally.

Lateral to the sternocleidomastoid muscle, external jugular vein is ligated. Dissection further exposes the inferior belly of omohyoid that once divided, the dissection is continued deeply as far as the prevertebral fascia overlying the brachial plexus and the scalene muscles.

Once the transverse cervical vessels are divided, the fat filling the posterior triangle is dissected towards off the prevertebral fascia. Now the floor of the triangle can be identified from occiput to the clavicle. The dissection can move forward on a broad front clearing each prevertebral muscles in turn. In order to keep the dissection moving uniformly forward along the entire length of the beck, it become necessary to section the sternocleidomastoid near its mastoid insertion. Dissection is continued to reach the anterior branch of cervical plexus as they emerge at segmental intervals just posterolateral to carotid sheath. Clearance of lower part of posterior triangle proceeds medially past the brachial plexus seen emerging between the two scalenes but safe behind the fascia until the dissections clearly impeded by sternocleidomastoid and this is then divided close to its sternal and clavicular insertions taking care to avoid damage to the internal jugular vein. Internal jugular vein is in close proximity to the sternocleidomastoid muscle at this point.

Once the sternocleidomastoid is divided dissection is proceeded medially over the scalenus anterior. Here it is important to watch carefully for the phrenic nerve which runs vertically across downward along the lateral line of muscle fibers. Once the nerve has been identified, dissection is carried on superficial to it, leaving it on the muscle, at the same time extending upwards so that its origin from the cervical plexus mainly C4 can be demonstrated.

With its source visualized the branches of the plexus C 2 ,C3 & C 4 passing forward into the resection mass can safely be divided.

Section of these branches coupled with forward traction brings the carotid sheath into view. Plane of dissection is along the prevertebral muscle behind carotid sheath.

Just lateral to the carotid sheath  over the scalenus anterior muscle thyrocervical trunk is seen giving thyrocervical artery which is ligated and dissected with the specimen. As the carotid sheath is approached the internal jugular vein with its bluish colour comes into view. The dissection plane here should stay close to the carotids. The vagus lying between artery and vein is dissected free and retained. The remaining tissue of the complex i.e. the internal jugular vein and deep jugular group is dissected forward. The absence of posterior branches of the internal jugular vein allows dissection to be carried up and down over much of its length without difficulty.

Then the internal jugular vein is divided and double ligation using non-absorbable suture with one ligature transfixing the vein is a wise precaution. Now the dissection proceeds forwards and upwards baring the infrahyoid muscles, the anterior jugular vein divided and the ansa cervicalis sacrificed. As the dissection proceeds, superior and middle thyroid veins are divided. The superior belly of Omohyoid is peeled off the infrahyoid group as part of the specimen and when dissection has reached the hyoid bone the muscle is divided at its origin.

In the upper neck as the internal jugular vein is stripped off from the internal carotid artery the hypoglossal nerve comes into view (between the two). It is traced to the superior root of ansa cervicalis that is divided.

Assuming ligation of internal jugular vein to precede, clearance of the submandibular triangle the dissection of vein free from hypoglossal and vagus nerve and the internal carotid artery is continued with forward and upward traction of the specimen until the transverse process of the atlas is reached. It is at this level that the internal jugular vein is divided.

With the internal jugular vein mobilized any remaining fibers of sternocleidomastoid still attached to the mastoid process are divided until the digastric becomes visible, and the same level the lower pole of the parotid is sectioned in a line across to the angle of the mandible. Here the retromandibular vein is ligated.

Downward retraction of the sternocleidomastoid and the parotid pole brings digastric into view and depending on the local condition it can either be retracted or divided watching for the occipital artery in so doing. Beneath the digastric the internal jugular vein is exposed, doubly ligated and divided in the same way as the lower end.

Clearance of submandibular triangle is now began. The upper border of the triangle is defined by deepening the dissection along the lower border of the mandible to the bone in front of the masseter. Here the facial vessels are ligated and divided as they cross the lower border just in front of the masseter if this has not already been done.

At the angle of the mandible this line of section becomes continuous with the section line of the parotid.

The inner surface of the mandible below the mylohyoid line has no attached structure in front of the medial pterygoid muscle and a finger inserted medial to the mandible can mobilize the soft tissues off the bone. Stripping is pursued upward to the insertion of the mylohyoid.

In the symphyseal area clearance of the digastric muscle should be extended across the midline as far as the contralateral digastric muscle, making it possible to clear the submental  nodes.

Tissue between the anterior bellies of digastric is dissected off the midline raphe of mylohyoid. If posterior belly of digastric has been sectioned the entire digastric is removed as a part of the resection specimen and the dissection plane continues on the mylohyoid. If digastric is retained dissection proceeds around it returning behind it to mylohyoid.

In this part of dissection mylohyoid is a key structure.

The tissue resected off mylohyoid is recognized as the superficial lobe of the submandibular salivary gland with its duct. Above and parallel to its duct lies and lingual nerve.

Below and parallel to the duct runs the hypoglossal nerve with hypoglossal venous plexus.

The duct is freed from its surroundings, ligated and divided and the gland is then dissected completely off hypoglossus. In so doing facial artery, looping over the gland is exposed for ligation and division. Final dissection cleans the specimen from any minor remaining attachments.


De Santo et al 1985; Leemans et al 1990 reported a 3 to 7 % recurrence rate in the neck following a radical neck dissection with histologically tumour free nodes.

Strong 1969 reported an ipsilateral recurrence rate of 20 to 71% following a therapeutic radical neck dissection for clinically and histologically nodes. This series included large number of cases of uncontrolled primaries. He also found a recurrence rate of 36.5 % in patients with histologically single node and 71 % in histologically multiple positive nodes.

Johnson et al 1981, Snow et al 1986 and Carter et al 1985 have shown that the recurrence rate in the neck after radical neck dissection is significantly higher when extracapsular spread of tumour is demonstrated.

Carter et al 1985 also found that macroscopic extracapsular extension is associated with a higher risk for recurrence (44%) whereas microscopic extracapsular spread has a lesser recurrence rate (25%) similar to intracapsular node metastases (32%).

Snow et al 1986 has shown that patients with four or more histologically positive nodes and extracapsular spread have a significantly greater likelihood for distant metastases( >60%) than patients with one histologically positive node with extracapsular spread ( < 30 %).

Efficacy of radical neck dissection is to be considered with adjuvant radiotherapy.

Leemans et al 1990, Strong 1969, Vikram et al 1984 reported a reduction in neck metastases by the addition of radiation therapy to radical neck dissection.

However De Santo Et al 1985 in a review of 1192 neck dissections reported no difference in the probability of recurrence in the neck patients with N2 neck disease treated by dissection alone, preoperative radiation, postoperative radiation or miscellaneous radiation.

Modified radical neck dissection

Modified radical neck dissection with preservation of spinal accessory nerve ( Type I ).

A modified radical neck dissection with spinal accessory nerve preservation is defined as the enbloc removal of the lymph node bearing tissues of one side of the neck from the inferior border of the mandible to the clavicle and from the lateral border of the strap muscles to the anterior border of trapezius preserving the spinal accessory nerve. The internal jugular vein and sternocleidomastoid muscle are resected with the specimen.


The following observation resulted in exploring alternative cervical neck dissection.

  1. The morbidity associated with the radical neck dissection, especially the shoulder disability that results from the resection of the spinal accessory nerve, and to a lesser extent the cosmetic deformity that results from this operation especially when done bilaterally.
  2. The realisation that the spinal accessory is not in close proximity to the nodes involved by tumour and that its preservation does not compromise the oncological soundness of the operation.

Indications :

When there is clearly identifiable plane of dissection between the tumour and the nerve.

Surgical technique

The incisions are the same as that of radical neck dissection. After the flaps are raised, the spinal accessory nerve is exposed as it crosses the anterior border of the trapezius at about 2 cm above the clavicle, then the nerve is exposed through the posterior triangle incising the fascia and fatty tissue over a hemostat. This is continued through the upper portion of the sternocleidomastoid muscle, exposing the nerve in its entire course through the neck. The nerve is freed from underlying tissue. Rest of dissection is carried out as routine radical neck dissection.

Alternatively, the nerve can be identified high in the neck, medially to the posterior belly of digastric where it lies either posterior or lateral to internal jugular vein. From here it is exposed and isolated. Isolation of the nerve is easy if bleeding is kept to a minimum.


Dargent & Papillon 1945 and Skolnik et al 1967 were among the first to advocate the spinal accessory nerve preservation.

Subsequently this was used by Roy & Beahrs 1969, Carenfelt & Eliasson 1980, Brandenburg & Lee 1981, Chu & Strawitz 1978, Pearlman et al 1982. The recurrence rate following modified radical neck dissection was 4 % to 7 % similar to radical neck dissection of 3 % to 7 %.

Functional, Conservative or conservation neck dissection ( Type III ).

This operation is defined as the en bloc removal of the lymph node bearing tissue of one side of the neck including the nodes in level I to V, preserving the spinal accessory, the internal jugular vein and the sternocleidomastoid muscle. The submandibular gland may not be removed.


Suarey 1963 discussed the rationale for this type of neck dissection. He stated that it is oncologically safe to remove lymph nodes of the neck without sacrificing the sternocleidomastoid muscle, the internal jugular vein and the spinal accessory nerve.

Since in necropsy of patients with head and neck cancer he noted that

  1. Lymph nodes were always found in the fibrofatty tissue either away or near blood vessels, particularly veins.
  2. Lymph nodes are not part of adventia of vessels.
  3. He also observed that lymph nodes are not located within the muscular aponeurosis of the salivary gland capsule. i.e. in sternocleidomastoid muscle and submandibular gland. But lymph nodes are found within parotid.
  4. There is no surrounding communication of lymphatics with the surrounding fascia or muscle.

Saurez demonstrated its feasibility in 275 cases. This operation was popularised by Bocca  who coined the term functional, conservative or conservation neck dissection. Bocca 1980 emphasised the presence of muscular and vascular aponeuroses of the neck, which define compartments filled with fibrofatty tissue. The lymphatic system of the neck, contained within the compartments can be excised in an anatomic block by stripping the fascia off the muscles and vessels.


  1. In cases of primary in larynx and hypopharynx where the submandibular triangle is at low risk of containing metastases.
  2. In differential carcinoma of the thyroid.

Contra indication

In cases of node fixation.

Surgical procedure

The procedure involves dissection in submandibular region. Then the fascia over sternocledomastoid is dissected beginning from posterior to anterior border. Muscle is retracted posteriorly and spinal accessory nerve is identified at the upper 1/3rd and lower 2/3rd of muscle. Fibrofatty tissue with lymphatic tissue are dissected away from spinal accessory nerve. The dissection is carried over the splenius capitis and levator scapulae muscles in the posterior triangle. The superficial layer of deep cervical fascia is divided at the junction of sternocledomastoid and anterior border of trapezius. Here the external jugular vein is clamped and divided. Also the omohyoid muscle is divided. Then the specimen is brought forward from posterior to sternocledomastoid muscle. The specimen is then dissected of carotid sheath and internal jugular vein while taking care not to injure the thoracic duct.

 Selective neck dissection

It involves selective en bloc removal of only lymph node groups of neck depending on the location of the primary tumour which are most likely to contain metastasis.

The supraomohyoid neck dissection consist of the removal of nodal regions I. II and III. In expanded supraomohyoid neck dissection region IV are also removed.

The lateral neck dissection consist of en bloc removal of nodal regions II, III and IV.

The ‘posterolateral neck dissection’ consist of removal of suboccipital and retroauricular lymphnodes and nodes from regions II, III, IV and V.


Medina & Byers 1989 stated that en bloc removal of the nodes at highest risk for metastasis is anatomically justified and it has the same therapeutic value as radical or modified neck dissection. It is associated with less post operative morbidity.

Rouviere 1938 and Fish & Siegel 1964 showed that lymphatic drainage of the mucosal surface of head and neck flow in a predictable routes.

Lindberg 1972 demonstrated the jugulodigastric and midjugular nodes are most frequently involved nodes in patients with carcinoma of the oral cavity. Further he noted submandibular nodes are involved in carcinoma floor of mouth, anterior oral cavity, tongue and buccal mucosa.

Skolnik 1976 in a study of radical neck dissection specimens noted no radical involvement of posterior cervical triangle.

Shah 1990 in a retrospective study of 119 radical neck dissection found that oral cavity tumours metastasised to regions I, II and III nodes, whereas carcinoma of oro, oropharynx and larynx involved mainly region II, III and IV nodes.


  1. Supraomohyoid neck dissection is indicated in the surgical management of patients with T2 to T4 N0 or Tx N1 oral cavity tumours and when palpable nodes less than 3 cm, clearly mobile and located in either level I or II. This procedure is performed bilaterally in primaries of floor of the mouth and in the anterior tongue. This type of neck dissection is done in performed when elective neck dissection is indicated in the management of patients with squamous cell carcinoma of the lip or skin of midline associated with discrete metastasis in submental or submandibular region. A supra omohyoid neck dissection is done in conjugation with parotidectomy for squamous cell carcinoma, merkel cell carcinoma and selected stage I melanomas of zygomas and cheek.
  2. Lateral neck dissection is indicated in patients with tumours of larynx, oropharynx and hypopharynx staged T2 to T4 N0 or N1 or T1 N1 when palpable node is located in level I or II. This procedure is usually done bilaterally as the lymphatics drainage metastases bilaterally.
  3. The posterolateral neck dissection is indicated in the treatment of melanomas, squamous cell carcinomas or other skin tumours with metastatic potential such as the merkel cell carcinomas that originate in the posterior and posterolateral aspects of the neck and occipital scalp. It is rarely indicated in the treatment of squamous cell carcinoma of the aerodigestive tract.

Surgical technique

A unilateral neck dissection is usually performed through a apron like incision that extends from the mastoid tip to mandibular symphysis. The lowest point of the incision is usually located at the level of the thyrohyoid membrane. The incision can be modified to include a lip splitting incision or a descending limb can be added for added exposure. To perform a bilateral dissection an apron like incision is made extending from one mastoid to other overlying the thyrohyoid membrane.

A superior flap is raised in a subplatysmal plane up to the inferior border of the mandible. The inferior flap is raised to about 1 inch above the clavicle , however the inferior flap can be raised up to the clavicle.

The dissection begins in the submental region. The lymphatic fibrofatty specimen is pulled inferiorly and laterally away from the digastric muscle and the mylohyoid. The submandibular gland with lymphnodes is dissected next away from the mandible. The fascia overlying the posterior belly of digastric and omohyoid is incised. The dissection is carried out in the area posterior to omohyoid preserving the hypoglossal and superior thyroid artery. Then the posterior dissection is began from the anterior border of sternocledomastoid muscle. The fascia is retracted anteriorly and dissection is carried around the muscle up to the point where the spinal accessory nerve enters the sternocledomastoid muscle. The nerve is carefully dissected free from the surrounding structures. Above the level of the nerve, splenius capitis and levator scapulae the dissection is completed. The dissected specimen in this region is brought forward. The dissection is carried along the carotid sheath. The inferior level of dissection is usually to the level where the omohyoid crosses the internal jugular vein.  If nodes are found in level III then the dissection is extended to level IV below the omohyoid sacrificing the omohyoid muscle. This operation is called “extended supraomohyoid neck dissection”. The neck dissection is done similarly on the opposite side. When this operation is complete only small amounts of lymph nodes are left behind in the posteroinferior aspect of the neck.

 Extended neck dissection

Any of the neck dissections described above can be extended to remove either lymphnode groups or vascular neural or muscular structures not routinely removed in a neck dissection.

Neck dissection can be extended to remove retropharyngeal lymphnodes in primaries of pharyngeal wall. These lymphnodes are involved in tumours of base of tongue, tonsil, soft palate and retromolar trigone.

Paratracheal and pretracheal nodes are removed in carcinoma of transglottis and subglottis, cervical oesophagus, trachea and thyroid carcinoma.

Adequate removal of a metastatic tumour in the neck may dictate the need to extend a neck dissection to resect structures such as hypoglossal nerve, levator scapulae muscle or carotid artery.

Controversies exist over the ligation of internal and common carotid artery.

Byers and Ballantyne 1985 noted that prognosis of a patient with neck disease which warrants resection of carotid arteries is dismal and it is not justified to remove carotid arteries. Moore and Baker 1955 observed a 30% mortality rate and 45% cerberal complication rate among patients who underwent carotid artery ligation.

Goffinet et al 1985 reported encouraging results with large cervical metastasis attached to carotid artery who were treated by resection of tumour and intraoperative Iodine 125 seed was placed using Vicryl tubes. He showed a 77 % tumour control but 1-yr. survival rate was 15%.

Bilateral Radical Neck Dissection:

This procedure is not often performed. The loss of both sternocleidomastoids makes it difficult for all patients, impossible for some to lift the head from a recumbent position without using the hands to support it. The loss of both accessory nerves doubles the usual disability from this source. Bilateral dissection is most often carried out as two separate operative procedures separated in time when metastatic nodes develop on the contra-lateral side of the neck after a node dissection on the ipsilateral side.

The modifications in bilateral dissections concern the skin incisions and the management of internal jugular vein. If there is large interval between the neck dissection the skin incisions can be the one used for a single neck dissection. When the procedures are simultaneous or the gap in time is very short, the problem changes, because the anterior flap raised in most dissection incisions ceases to have a base if it is raised on both sides at once.

Duplicated Conley variant of triradiate incision can be used in these circumstances. This has the most assured blood supply.

This can be modified from a curve into a series of straight line and in this form has been recommended (Lore, 1973). The internal jugular vein is the main venous connection between the brain and the heart and its simultaneous loss on both sides has dramatic and undesirable effects. These effects are mitigated if it is possible to ablate the veins at intervals rather than simultaneously. Compensatory veins appear to open up which are efficient enough to prevent the more undesirable changes and the process of compensation seems to occur with surprising speed. An interval of four weeks is usually adequate though naturally the safe time gap is likely to vary some what in different patients and one would make it as long as possible, as well as giving consideration to retaining one vein if the time interval was short.

With any second neck dissection preservation of the vein is worth attempting and for shorter time lag from first neck dissection.

In the extreme situation of a bilateral simultaneous neck dissection can be done with one vein retained if at all possible.

An attempt at preservation of vein made on the first side dissected leaves the second variable for another attempt, should the first be unsuccessful.

Morfit and Perzik 1952 reported a case of bilateral simultaneous neck dissection without vein preservation. The patient developed massive edema and cyanotic swelling over face which subsided in due course of time.

Assuming preservation of at least one internal jugular vein, this procedure would result in virtually no disability and used in the correct clinical situation is likely to be equally effective.

Complications of neck dissection

In addition to various medical complications that can occur following any surgical procedure, there are a number of surgical complications associated with neck dissection. The complications can be divided into intraoperative and postoperative complications. Complication of cerberovascular accident is also common following head and neck dissection. It is important to manage this properly otherwise it might result in neurological deficit or death.

Sobol et al 1982, have described five pathological mechanism for cerberovascular complications following head and neck surgeries.

  1. Embolism from ulcerated plaque.
  2. Intravascular thrombosis with occlusion.
  3. Unintentional surgical ligation, transection or laceration of the carotid artery.
  4. Ligation of the external carotid artery with previous internal artery occlusions.
  5. Transient reduction in cerberovascular profusion pressure.

Intraoperative complications

These are mainly vascular complications which include bleeding (arterial / venous), chyle leak and nerve injury. Inadvertent arterial injury might result in serious

Most serious complications are vascular in nature.

I)Bleeding from artery:

Most dramatic injury is to the common carotid artery and its branches which occurs rarely during an orderly neck dissection. If bleeding is from the external carotid artery or one of its branches the vessel it  may be ligated.

The carotid system contribute to 85% of cerebral blood flow, if occluded collateral flow must develop from the circle of Willis.

Moore & Baker 1955 found a difference in effects following elective and nonelective ligation of the common carotid artery. In 69 patients following elective ligation 40 % developed cerebral complication and 17% died while in 87 cases of nonelective ligation 88% developed cerebral complication and 38 % died.

Ledgerwood et al (1980) compared primary repair versus ligation of carotids and reported excellent results with primary repair.

Small lacerations are repaired. In moderately large loss of artery end to end anastamosis is planned. Lacerations of internal carotid artery close to the base of skull are best ligated. When tumour invades the internal carotids ,the artery can be repaired with saphenous vein graft with synthetic suture materials.

Fleming and Petrie (1968), Miller and Bergstorm (1974) reported cases of internal carotid thrombus. In suspected cases of thrombus thrombectomy is performed and if not viable ligation of internal carotid artery is considered. .


II) Bleeding from vein:

Bleeding from internal jugular vein may be potentially more dangerous than arterial bleeding as it is difficult to control. A most dangerous injury is at the junction of internal jugular and subclavian veins. To prevent air embolism patient may be brought down to level position from an elevated position.

Hemorrhage from the proximal aspects of internal jugular vein is less dangerous. If proximal control cannot be obtained with clamps and ligature, then the jugular foramen may be packed with oxidized cellulose. Alternatively mastoid tip may be removed in an effort to gain direct control.

III) Chylous Leak:

Usually prevented by isolating the lymphatic pedicle between the carotid artery and the phrenic nerve and clamping this pedicle prior to dividing and ligating it. Intraoperative chylous leak should be recognized as opalescent fluid in posterior and inferior aspect of the dissected neck. Once it has been isolated it should be clamped and secured with a tie.

IV) Nerve Injury:

The spinal accessory, cervical, cutaneous and great auricular nerves are intentionally sacrificed during standard radical neck dissection.

If spinal accessory nerve is accidentally injured as in modified neck dissection, an interpositional nerve graft of greater auricular nerve can be used for repair. This can decrease the shoulder disability.

Injury to motor branches of cervical plexus can also result in shoulder disability.

Injury to phrenic nerve can occur without its transection. If the phrenic nerve is transected it should be reapproximated.

Injury to the cervical contribution to the phrenic nerve may be avoided by dissecting the neck from the posterior triangle anteriorly and transecting the cervical cutaneous nerve, distal to their phrenic contribution.

Injury to brachial plexus is uncommon and may be avoided by undermining the posterior triangle of the neck and visualizing the entire brachial plexus prior to transection or to clamping of the posterior triangle fat. Injury to vagus nerve will result in ipsilateral hypopharyngeal and vocal cord anesthesia or vocal cord paralysis. No cardiac or GIT dysfunction is known to result from transection of the vagus.

Unilateral loss of hypoglossal nerve function results in little disabilities. Bilateral loss of function may occur in laryngectomy with resection of base of tongue cancer and would result in severe crippling of the swallowing function.

Injury to cervical sympathetic chain results in Horner’s syndrome.

Paralysis of facial nerve is an unusual complication. Care must be taken at the time of sacrifice of posterior belly of digastric since the stylomastoid foramen through which the facial nerve emerge is located at the posterior superior aspect of the posterior belly. The facial nerve may be identified and dissected free.

Unilateral sacrifice of marginal mandibular branch of facial nerve may be necessary when there is obvious nodal metastases in the upper neck. This results in little disability and minimal cosmetic deformity. However loss of function of both marginal mandibular nerves does result in incompetent oral commissure.

Postoperative Complication:


It is one of the more common complications that predispose to wound infection, flap necroses, carotid exposure and fistula formation.

To prevent haematomas meticulous hemostasis should be attained and prior to closure at least two medium drainage catheter should be placed in the supraclavicular fossa and posterior triangle.

Two other catheters should be placed anterior to the carotid artery and adjacent to the pharyngeal or oral cavity closure in composite resections. Constant suction should be maintained during the early  post operative and bulky external compression drainage applied.

Chylous Fistula:

Rarely a chyle leak will be noted as a bulge beneath the skin of the supraclavicular fossa.

In addition to elevation of skin flaps, the chylous leak produces induration edema and erythema of the overlying skin. A chylous fistula may lead to chylothorax that mimics pleural effusion. This may require chest tube drainage.

Bilateral neck dissection produces complications related to venous and lymphatic obstruction. Airway obstruction occurs frequently. Increased intracranial pressure may occur, that can lead on to cerebral edema, impaired neurologic function, and blindness stroke.

Sequelae of radical neck dissection:

  1. Shoulder dysfunction: This is due to loss of trapezius function when spinal accessory nerve is cut. Participating muscles that compensate for loss of trapezius function are scapulae, rhomboidus and levator.

Abduction of shoulder becomes limited. Patient try to compensate by contralateral flexion of the trunk.

  1. Enlargement of sternoclavicular joint: This is partly due to prominence of the head of the clavicle following removal of sternocleidomastoid.
  2. Stress fracture of the clavicle
  3. Gustatory sweating:

This is due to aberrant regeneration of autonomic nerve fibres following injury. The most typical area is submandibular area on the upper cervical flap.

  1. Sensory loss:

This is due to sectioning of branches of cervical plexus. The area of sensory loss can involve the entire neck, chest from midline to just below the insertion of the deltoid laterally and inferiorly to the nipple line, posteriorly across the scapular spine, superiorly to the occiput including the entire ear.

Patient should be advised not to expose this denervated skin to extremes of temperature.

  1. Neuromas:

Neuromas of the cutaneous roots of cervical nerves are common and may cause pain, tenderness, hyperesthesia. Graham in his review of 50 patients could find no correlation between the occurrence of neuroma and the use of preoperative or postoperative irradiation or a particular type of cervical flap. He found no recurrence of neuromas after treatment by resection high ligation and local instillation of triamcinole.

  1. Scar contracture

Significant scar contracture of the cervical flap incisions may occur if wound dehiscence has occurred or if the vertical limb of a cervical incision is placed more than 4cm anterior to the border of the trapezius muscle.

Multiple Z plasty will remove the tension of the scar contracture but will not improve the cosmetic appearance.


A precise knowledge of the surgical anatomy is essential for neck dissection. Neck dissection is an effective procedure for the management of diseases of neck. This is an ideal treatment for management of malignancies and metastasis to the head and neck region. For management of No neck refined surgical techniques like selective or functional neck dissection have been developed in order to reduce the morbidity associated with radical procedures.


  1. Charles W. Cummings, John M. Fredrickson, Lee A. Harker, Charles J. Krause, David E. Schurller. Neck Dissection. Otolaryngology- Head and neck surgery. Vol. II, 2nd 1993: 1649-1672.
  2. Ian A. McGregor, Frances M. McGregor. Neck dissection. Cancer of the face and mouth – Pathology and management for surgeons. Churchill Livingstone.1986: 282- 320.
  3. Ian T. Jackson. Inrtra oral tumour and cervical lymphadenectomy. Grabb & Smith’s Plastic Surgery. Sherrel J. Aston, Robert W. Beasley, Charles H. M. Thorne. 5th Lippincott- Raven . 1997 : 439 –452.
  4. H. Sobin & Ch Wittekind. TNM Classification of malignant tumours. 5th edition. UICC, A John Wiley & Sons Inc. Publication. 1997.
  5. Hermanek, R. V. P. Hutter, L. H. Sobin & Ch Wittekind. TNM atlas. Illustrated guide to the TNM / pTNM classification of malignant tumours. 4th edition. Springer. 1997.
  6. Aydin Acar, Gürsel Dursun, Ömer Aydin,Yücel Akbaş. J incision in neck dissections. The journal of Laryngology and otology. 1998: 112: 55 – 60.
  7. Susumu Omura, Hiroki Bukawa, Ryoichi Kawabe, Shinjiro Aoki, Kiyohide Fujita. Comparision between hockey stick and reverse hockey stick incision: gently curved single linear neck incisions for oral cancer. Int. J. Oral Maxillofac. Surg. 1999: 28 : 197 – 202.

h g h g

Posted in Maxfac Tutorial, Odontogenic Tumors (Benign)



The ameloblastoma is a benign tumour of jawbones with locally invasive capacity. It is a true neoplasm of enamel-organ type of tissue which does not undergo differentiation to the point of enamel formation. Ameloblastoma is the best-known odontogenic tumour, very often used as the norm by which other odontogenic tumours are judged. Its unique clinical and pathologic behaviour was described by Robinson (1937) as being “usually unicentric, non-functional, intermittent in growth, anatomically benign and clinically persistent”.



Guzack in 1826 reported a tumour of the jaw, which may be the first recorded instance of an ameloblastoma. But the credit of reporting the first neoplasm of this nature goes to Broca (1968). The first detailed description of ameloblastoma was given by Falkson in 1879.

The name ‘adamantine epithelioma’ was introduced by Malassez (1885) and subsequently many synonyms like adamantinoma, adamantinoblastoma, epithelial odontoma and multilocular cyst, were used to describe the lesion. The name ameloblastoma was suggested by Churchill (1920) and Ivy & Churchill (1930).

A full account of historical aspect of ameloblastoma was detailed by Baden in 1965.


The literature consists mainly of many single-case reports and small series. A few extensive series with statistics of age, sex, race, site of growth etc. include that of Robinson (1937) who reported 379 cases, and that of Small and Waldron (1955) who listed 1036 cases including those of Robinson. More recent series include those of Mehlisch et al (Mayo Clinic Review-1972  – 126 cases), Sehdev et al (1974  – 92 cases), Regezi et al (1978  – 78 cases)and Ademola et al (1993  – 315 cases). The last mentioned series include the cases of Mosadoni (1975 – 29 cases), Doranola et al (1975 – 16 cases) and Adekeye (1980 – 109 cases).


Ameloblastoma is said to account for about 1% of all oral tumours. Many tumour registries list it as the most prevalent odontogenic tumour, but because this data is based on biopsied lesions and because innocuous lesions like compound odontomas are often not removed, this data may not reflect the true frequency.


Small and Waldron (1955) found the average age of patients at the time of reporting to hospital to be 38.9 years, but the average age at the time of discovery of the lesion was 32.7 years. All series have reported that more than half of all the cases was reported in patients in the age-group of 20 to 50 years. But the tumour can occur in any age, and have been reported in young children (Young & Robinson – 1962 and Lewin – 1966). The oldest patients have been over 80 years of age.


The occurrence of ameloblastoma is approximately equally divided between the sexes. Some authors have reported a slight preponderance in males, but the difference is considered insignificant.


The tumour is thought to be more frequent in Africans than in white race (Kegel). In the U. S. and in Denmark, it accounts for 0.07% and 1% respectively of all oral tumours. In Africa, this figure is 3.6%. But this data is difficult to interpret because of the lack of standardised data collection procedures in third world countries and because many people do not report for treatment until the disease becomes troublesome. This has prompted many authors to believe that the concept of higher incidence of the tumour in black races cannot be supported. However, a more recent report by Shear and Singh (19780 from South Africa showed that the incidence of ameloblastoma is very much higher in blacks than in whites. Chung et al (1969) reported that tumours of jaws including ameloblastoma have a higher incidence in Korea.

Site distribution

About 80% of the tumours occur in the mandible. This figure was as high as 99.1% in one case (Adekeye – 1980). Nearly three-fourth of the mandibular tumours occur in the molar-ramus area in almost all the series. But in Nigerians (Akinosi and William – 1969) and in Indians (Potdar – 1969), a predeliction to mandibular symphysis was noted, but by contrast, Adekeye noted a preponderance in the horizontal ramus of the mandible. The less common lesions of the maxilla mainly occur in the molar area, antrum and the floor of the nose. Lesions have even been described to occur in the tuberosity area, zygomatic bone and base of skull.


Very little is known about the causative factors. Robinson in 1937 noted that in one-third of the cases, there was a history of trauma, oral infection or impacted third molars. But considering that these factors are common in normal individuals also, this data is not considered significant.

Dietary deficiency has been considered a possible factor. An irregularity in the ameloblastic layer of enamel was seen in pigs maintained on rachitogenic (Vit. D deficient) diet, but no unequivocal conclusion could be made.

Ameloblastoma-like tumours could be produced in mice by injection of polyoma virus, and particles with characteristics of paramyxovirus have been noticed in ameloblastoma. Similar tumours have also been produced in animals by the injection of nitrosureas.


The earlier workers themselves had noted the resemblance of the tumour epithelium to the normal odontogenic apparatus and had suggested that the neoplasm was derived from a portion of this apparatus or from cells potentially capable of forming dental tissue, but the precise point of origin is unknown. Malassez had noted small collection of epithelial cells near the roots of teeth and suggested that ‘adamanto-epithelioma’ arose from these cells.

At present, most authorities consider the tumour to be of varied origin. It conceivably may be derived from

1.   Enamel organ

Early histologists considered ameloblastoma to be of origin from enamel organ due to the obvious histologic similarities. The main contesting point (Bland & Sutton – 1922) is that the age of onset (approx. 32 years) belies this concept. But considering the early symptomless character, slow growth of the tumour and the fact that it is seen most often in the region of the mandible where supernumerary tooth germs are generally present, it is suggested that the cells of enamel organ give rise to this neoplasm.

2.   Cell rests of enamel organ

After the tooth buds have separated from dental lamina, that structure regresse4s and disappears, although isolated groups of epithelial cells remain in the connective tissue as rests of Serres. These cell rests in the periodontal membrane were first demonstrated by Malassez, who along with many others suggested that ameloblastoma originated from these rests.

3.   Basal cells of oral mucosa

Many histologists have noted the connection of ameloblastoma with oral mucosa but it is generally agreed that the tumour probably grew up from beneath and then established a connection with the surface epithelium. That the tumour is essentially intra-osseous in origin lends credence to this theory, but some authors do raise the question of mucosal origin, and in line with this concept, some have considered the tumour to be a type of basal cell carcinoma. There are striking histological similarities between the two. They also resemble each other in their clinical growth (slow growth and local invasiveness).

4.   Cysts of dental origin

A number of reports have showed lesions which appeared clinically and radiographically as ordinary odontogenic cysts but was proved in fact to be ameloblastoma. Cahn (1933) argued that dentigerous cyst should be considered a potential ameloblastoma. Shteyer et al has reviewed the concerned literature. In a review of 641 cases by Stanley and Diehl, 17% of the cases were definitely associated with an impacted tooth or a dentigerous cyst. So it is suggested that ameloblastoma arises not infrequently from dentigerous cyst.

5.   Heterotopic epithelium

Lesions histologically similar to ameloblastomas have been discovered in sites other than jaws like the pituitary and long bones. These lesions are discussed separately.

Clinical features

The typical ameloblastoma begins insidiously as a central lesion of bone, which is slowly destructive but tends to expand the bone rather than perforate it. The tumour is seldom painful unless secondarily infected and shows few or no symptoms in the early stages. Quite frequently, it is first noticed during routine dental examination.

Later there is a gradually increasing facial deformity. As the lesion grows in size, an ovoid or fusiform swelling is noted, which is hard but not tender. Because of the slow growth of the tumour, it is probable that even in apparently early cases, it has already been present for a significant time before its discovery. In the absence of treatment, the tumour continues to enlarge in size and the surrounding bone becomes thin and fluctuence or ‘egg-shell crackling’ may be elicited. Perforation of bone is a very late feature.

Teeth in the area may become loosened and usually many teeth are extracted before the tumour is discovered. Resorption of roots of teeth is an extremely common finding.

In the maxilla, the sinus becomes involved and the tumour may extend into the orbit or nasopharynx.

In lesser-developed countries, it has been common to see the tumours having grown to enormous size by the time treatment is sought. Another presentation is a history of a series of operations in the jaw for a ‘cyst’ or an ‘abscess’ carried out over a period of years with recurrence after each one.

Roentgenographic features

The ameloblastoma has been described classically as a multilocular cyst of the jaw. The tumour exhibits a compartmental appearance with septa of bone extending into the radiolucent tumour mass.

The radiograph may present either a polycystic (multilocular) or a monocystic (unilocular) appearance. In the polycystic type, the bone is replaced by a number of small well-defined radiolucent areas, giving the whole lesion a ‘honey-comb’ or ‘soap-bubble’ appearance and at the same time, the jaw is expanded. In the unilocular type, there is a well-defined area of radiolucency resembling a single cyst. It is not uncommon to misdiagnose an ameloblastoma as a dentigerous cyst or a radicular cyst because of its positional relationship with teeth.

Other characteristics like thinning of cortical plate and root resorption are also observed radiographically.


Biopsy is the removal of tissue from a living individual for diagnostic examination. It is the least equivalent (most diagnostic) of all diagnostic procedures and should be carried out whenever a definitive diagnosis cannot be obtained using less invasive methods. Aspiration or incisional/excisional biopsy may be performed for hard- tissue lesions like ameloblastoma.


Any radiolucent lesion that requires biopsy should undergo aspiration biopsy before surgical exploration, to rule out vascular lesions and to obtain a more confirmative clinical diagnosis.

Incisional biopsy

Most lesions of the hard tissue have to be approached through a mucoperiosteal flap. Lesions totally within the jaw require the use of a cortical window, which is cut using burs or chisels. A trephine bur may also be used. If the cortex is resorbed, the opening is made larger using rongeurs and burs. The specimen is removed as a V-shaped section and the remaining lesion is left undisturbed. Then the flap is closed.

The specimen is immediately placed in 10% formalin solution. The tissue must be totally immersed in the solution. The biopsy data sheet is filled, mentioning the clinical and radiological aspects in detail. The specimen and the data sheet are sent to a pathologist who has expertise in oral pathology.

Histopathologic features

Macroscopic features

The specimen as received by the pathologist consists of the tumour with a surrounding margin of normal bone or of the completely resected tumour-bearing area of the jaw. The tumour presents as a cylindrical or fusiform swelling which expands the bone. If perforation has occurred, the tumour extends to the soft tissues.

On section, the tumour appears as a greyish-yellow mass replacing the bone. There may be cystic spaces, particularly in advanced cases. The cysts have a more or less smooth epithelial lining and contents vary from straw-coloured fluid to semisolid materials. There may be minute nodules of growth protruding from an otherwise smooth lining. This is true in the case of unicystic ameloblastoma also. One or more teeth may be involved in the tumour.

Microscopic features

The ameloblastoma closely resembles the enamel organ. The typical multilocular ameloblastoma has two main histological patterns – follicular and plexiform. The unicystic variety shows a different histologic picture from these two.

Follicular (simple) ameloblastoma

This is composed of many small discrete islands of the tumour consisting of a peripheral layer of cuboidal or columnar cells whose nuclei are generally well polarised. The cells resemble ameloblasts or pre-ameloblasts and shows fine cytoplasm or fine granulation in some cases. These enclose a central mass of polyhedral, loosely arranged cells resembling the stellate reticulum. Thus each islet resembles an enamel organ.

The tumour islets are separated from each other by a variable amount of connective tissue stroma that caries the blood vessels. The general resemblance of ameloblastoma to the normal enamel organ has also been emphasised in a number of histochemical and ultrastructural studies. But some workers have found that the enzyme activity is more similar to that squamous cell carcinoma.

Microcyst formation within the tumour follicle is a common occurrence. In some cases, the central stellate cells undergo degeneration. In other cases, the cyst formation is accompanied by well-marked cellular changes like swelling of cells and homogenisation of cytoplasm. The coalescence of small cysts to form larger ones is responsible in course of time for macroscopic cyst formation. Apart from a rare case of ‘odonto-ameloblastoma’, enamel is not found in the tumour.

The follicular type shows a large number of histologic variants like

  1. Acanthomatous ameloblastoma (Pindborg – 1970)

In this case, the cells occupying the central region undergo squamous metaplasia, sometimes with keratin formation.

  1. Granular cell ameloblastoma (McCallum et al – 1957, Mallick – 1957)

The cytoplasm of the central cells takes on a very coarse, granular, eosinophilic appearance. This character often extends to the peripheral cells as well. The cytoplasmic granules are eosinophilic and PAS positive. The ultrastructural studies (Tandler & Rossi – 1977) have shown that the granules represent lysosomal aggregates.

  1. Basal cell ameloblastoma

This variant resembles basal cell carcinoma of skin. The epithelial tumour cells are more primitive and less columnar, and are generally arranged in sheets.

  1. Spindle cell ameloblastoma

The central cells undergo metaplasia to spindle cells.

  1. Clear cell ameloblastoma (Waldron et al – 1985)

This tumour has histopathological characteristics of ameloblastoma but also shows clear cells. It closely resembles ameloblastic carcinoma. The more recent view (Gardner – 1993) is that they are different lesions and that the presence of clear cells are rare occurrences in both lesions.

  1. Ghost cell ameloblastoma

Some workers have demonstrated ‘ghost cells’ (degenerated empty cells) which could be differentiated from the ‘clear cells’ of Waldron.

  1. Desmoplastic ameloblastoma

A newly described variant, occurring predominantly in anterior maxilla, it shows the radiographic picture of a benign fibro-osseous lesion.

There is considerable controversy regarding the relationship of these histological variations to the aggressiveness of the tumour. But most studies in this direction remain inconclusive.

Plexiform ameloblastoma

In this case, the tumour cells are arranged in irregular masses, or more frequently, as a network of interconnecting strands of cells. Each of these masses show a lining of columnar/cuboidal cells and a central mass of stellate reticulum like cells although the latter is much less prominent. Cyst formation is also seen in this type of growth.

Unicystic ameloblastoma

First described by Robinson and Martinez (1977), the unicystic ameloblastoma resembles dentigerous, radicular or residual cyst clinically and radiographically. It is differentiated by one or more of these features (Ackerman-1988).

  1. Lining epithelium exhibiting early ameloblastomatous changes of cyst lining, as described by Vickers and Gorlin (1970). This variety is called mural ameloblastoma.
  2. Nodules of tumour projecting intraluminally. This type is mentioned as luminal/ intraluminal ameloblastoma.
  3. Ameloblatomatous lining epithelium proliferating into the connective tissue wall.
  4. Islands of ameloblastoma occurring isolated in the connective tissue wall.

The latter two types are includes in the group ‘invasive unicystic ameloblastoma’.

‘Plexiform unicystic ameloblastoma’ is the term used by Gardner (1970) to designate a plexiform type of epithelial proliferation occurring in the dentigerous cyst, which otherwise exhibit the usual histologic features of a dentigerous cyst.

Local spread and metastases

Ameloblastoma is a tumour that causes expansion of bone rather than destruction. At the same time, there is a certain amount of local invasion of the surrounding bone, though it is generally limited in extent. Compact bone forms a much more efficient barrier to invasion by ameloblastoma than does cancellous bone and because of this, the extent of tumour encroachment upon it is reflected reasonably accurately in the corresponding radiograph. Extension of tumour into the cancellous bone, on the other hand, is much less readily gauged, since neither the radiograph nor naked eye assessment at the time of operation can reveal the extent to which cancellous spaces have been infiltrated, since this process occurs well in advance of actual bone destruction, to the magnitude of several millimetres. Because of its tendency to infiltrate cancellous bone, the lesions of posterior maxilla, and rarely large mandibular tumours may in fact kill the patient by direct extension into the cranium.

However it is found that in the unilocular type of ameloblastoma, the infiltration into cancellous bone is minimal, if ever present.

There appears to be little correlation between the histological pattern in ameloblastoma and its clinical course. Ultrastructural and histochemical studies have been conducted by many authors for this purpose, but most studies remain inconclusive.

Regarding distant spread, ameloblastoma is rather like basal cell carcinoma, which means that metastatic dissemination of ameloblastoma is rare, but has been reported to occur. This brings us to two new terms – malignant ameloblastoma and ameloblastic carcinoma. These terms are often used for each other, and they continue to confuse clinicians even now.

Malignant ameloblastoma

As used by most investigators, malignant ameloblastoma is defined as the ameloblastoma that has been shown to metastasise. To qualify for this designation, the metastatic lesion must also be cytologically benign and must closely resemble the original lesion in the jaws histologically.

In early reviews by Small and Waldron, over 30 cases were collected from the literature in which metastatic deposits were reported to have occurred. On further review of them (Carr and Halpenin – 1968), many of them were found to be either cases of misdiagnosis or ones where the metastatic lesion could not be undoubtedly proven to have disseminated from the ameloblastoma. In those few cases where a metastasis of ameloblastoma has been accepted, the feature is a long standing disease, operated upon a number of times over the course of years, and finally presenting with metastatic deposits in the lungs (Vorzimer and Perla – 1932; Schweiter and Banfield – 1943).

The literature dealing with metastatic ameloblastoma was reviewed by Lee et al (1959). It has been argued that the demonstrable lung lesions are a result of aspiration implantation, although this is difficult to prove and unlikely considering the routine precaution taken to avoid such and occurrence. What lends credence to this argument are the history of previous surgeries and the fact metastases are usually found in sites where aspirated foreign bodies are usually found.

Though pulmonary metastases seem to be the usual form of dissemination, metastases have also been noted in other sites, such as cervical and media stinal lymph nodes, bone and liver, and other viscera. These are considered by some as result of haematogenous spread.

Ameloblastic carcinoma

This is generally defined as that type of ameloblastoma in which there has been obvious histological malignant transformation of the epithelial component, and in which the tumour has behaved in malignant fashion so that the metastatic lesion do not have resemblance to the primary odontogenic tumour but to a less well-differentiated carcinoma

There seems to be difference of opinion in the relation of ameloblastic carcinoma to primary intra-osseous carcinoma of odontogenic origin. While the WHO publication (Pindborg –1971) classifies odontogenic carcinoma into malignant ameloblastoma, primary intra-osseous carcinoma and other carcinoma arising from odontogenic epithelium. But Elzay (1982) has liberalised this concept and classifies primary intra-osseous carcinoma into

  1. Those arising from odontogenic cysts
  2. Those arising from ameloblastoma (malignant ameloblastoma and ameloblastic carcinoma)
  3. Those arising de novo

Diagnosis of ameloblastoma

Non-neoplastic epithelial proliferation

While typical tumours present no problems in microscopic diagnosis, some difficulties may arise in cases of non-neoplastic epithelial proliferation, such as in the walls of odontogenic cysts.

Churchill (1938) has demonstrated that the walls of dentigerous cysts may show non-neoplastic proliferation comparable to the appearance of ameloblastoma and emphasised that the two lesions were separate entities. Again Vickers and Gorlin (1970) has shown that cells of mural ameloblastoma might flatten down by the pressure of cyst contents to look like a non-neoplastic proliferation. Close examination of the lesion considering the size of the cells, and shape and size of the nuclei, is essential to differentiate the two in certain cases

Relationship with dentigerous cyst

This is a controversial subject. The statement that ameloblastoma may arise in dentigerous cyst implies neoplastic change in an initially non-neoplastic lesion. A monocystic lesion with the clinical and radiographic characteristics of dentigerous cyst may prove histologically to be a simple cyst or an ameloblastoma. It is particularly necessary to make a thorough examination of the cyst wall with special attention to all mural nodules and thickenings.

Differential diagnosis

From the treatment point of view, the histologic variations of ameloblastoma are not very significant, and the basic diagnosis of the tumour whether it is the typical multilocular type, unicystic type or peripheral (extra-osseous) type, is considered adequate. In all cases, the typical palisading, hyperchromatism, reversed polarity and vacuolisation of basal cells are present.

One potential error is confusing the acanthomatous pattern of ameloblastoma with squamous cell carcinoma. Here the basal cells should be looked for typical features of ameloblastoma, and squamous cells looked for signs of dysplasia.

The basal cell ameloblastoma is confused with basal cell carcinoma of skin and adenoid cystic carcinoma. The diagnostic problem of the latter occurs primarily in maxilla. The differential diagnoses of ameloblastoma exhibiting mucous cells or clear cells include muco-epidermoid carcinoma, renal cell carcinoma and various salivary gland tumours.

Other tumours that should be included in differential diagnosis include odontogenic ones like ameloblastic fibroma, squamous odontogenic tumour, calcifying odontogenic cyst and metastatic carcinoma.

Another potential misdiagnosis is the failure to recognise the ameloblastic carcinoma. This has the same histologic features as ameloblastoma, except that it exhibits dysplasia. Since it behaves like a carcinoma, it requires different treatment.

Before discussing management of ameloblastoma, it is important to mention two different lesions – extra-osseous ameloblastoma and extra-oral similar tumours.

Extra-osseous ameloblastoma

Extra-osseous (peripheral ameloblastoma) is a tumour which histologically resembles the typical central or intra-osseous ameloblastoma, but which occurs in the soft tissue outside and overlying the alveolar bone. Although there were some earlier reports, Stanley and Krogh (1959) were the first to describe a peripheral ameloblastoma in the lingual surface of molar-premolar area of the mandible.

The peripheral ameloblastoma appears to have a predilection for the acanthomatous and basal cell patterns. The relationship of tumour cells to overlying epithelium is highly variable.

Greer and Hammond (1978) have shown that the ultrastructure of peripheral ameloblastoma is the same as that of intra-osseous ameloblastoma.    Most important factor in the diagnosis is a distinction between a peripheral ameloblastoma and extra-osseous component of an intra-osseous ameloblastoma. Another factor is its histological similarity to basal cell carcinoma and basal cell pattern of ameloblastoma. Other differential diagnoses include squamous cell carcinoma and other odontogenic tumours.

The peripheral ameloblastoma is considered a less invasive lesion than its intra-osseous counterpart, and they do not require excessively drastic treatment.

Extra-oral tumours resembling ameloblastoma

Tumours somewhat similar in structure and histological picture to ameloblastoma occurs in pituitary gland, the tibia and ulna, the ovary and elsewhere.


Zulch in 1963 reported after reviewing 6000 CNS tumours that the pituitary craniopharyngioma accounts for 2.5% of all CNS tumours. This tumour occurs in the anterior lobe which is derived from the Rathke’s pouch, an outgrowth of the oral ectoderm. Even after the degeneration of the craniopharyngeal duct, some squamous epithelial residues remain in the infundibulum, giving rise to ameloblastoma-like tumours. Its greatest incidence is in children and in young adults below 25 years of age. It grows as a pseudo-encapsulated mass, usually in suprasellar area, and often destroys the pituitary gland. Microscopically, the resemblance to ameloblastoma is very close.

Adamantinoma of long bones

This lesion was first reported by Fischer in 1913. Discussed by Baker, Dockerty & Coventry (1954), the tumour has a superficial microscopic resemblance to ameloblastoma of the jaws. Most cases have been reported in the tibia but rare cases have been reported to occur in ulna, fibula and femur. Chagus et al has suggested that the lesion is actually a malignant angioblastoma, but electron microscopic studies have pointed towards an epithelial origin.

Management of ameloblastoma

Over the years, a wide variety of treatment modalities have been advocated by different authors in the management of ameloblastoma. These range from a very conservative enucleation through an intra-oral approach to a radical hemimandibulectomy or maxillectomy with liberal removal of a good amount of uninvolved bone. A detailed discussion of the various treatment options is given below.


For decades, it has been almost universally accepted that radiotherapy is not an appropriate treatment modality in the treatment of ameloblastoma. Most series have reported poor result with this method.

It is not that the tumour is inherently radioresistant. It has been found that the extra-osseous component of large ameloblastomas can be markedly reduced by irradiation (Hair-1963; Singleton-1970). Also, the other similar tumours are relatively radiosensitive. The reason why ameloblastoma is not controlled by radiation is that they are primarily intra-osseous lesions and their location within the bone provides resistance to radiotherapy, just as squamous cell carcinoma turns radioresistant once the tumour has invaded bone.

The objections to the use of radiotherapy are mainly three-fold.

  1. It is ineffective in controlling the tumour
  2. There is danger of inducing osteoradionecrosis
  3. There is possibility of inducing malignancy, post-radiation carcinoma (in the form of ameloblastic carcinoma) or post-radiation sarcoma (Becker et al – 1967).

However, Atkinson, Harwood and Cummings, in a reputed journal (Cancer – 1984), have argued that there is, in fact, a role for megavoltage radiation in the treatment of ameloblastoma. Nevertheless, considered as a whole, radiotherapy is not an acceptable means of treating ameloblastoma, except in inoperable cases, primarily when it has invaded the cranium. In such cases, its use is one of last resort.


Curettage is the removal of the tumour by scrapping it from the surrounding normal tissue. Excision is the local surgical removal with an attempt to include a rim of uninvolved tissue. Currently most surgeons conclude that curettage is the least desirable form of therapy. Sehdev et al (1974) reported that repeated curettage for mandibular ameloblastoma has given a cure rate of only 10%. Taylor (1968) reported a 63% recurrence rate and Rankow and Hickey (1954) gave a 91% recurrence rate.

The results of curettage for maxillary ameloblastomas are even worse than those in the mandible because of structural difference in the bone. The maxilla lacks a thick compact cortical bone and has an intimate relation to the nasal cavity, paranasal sinuses, orbital contents, pharyngeal tissues and structures entering and leaving the base of skull.

In 11 cases of Sehdev et al, all of them had recurrences. 63% either died of the disease or had massive recurrences. It is possible that opening of maxillary antrum to a direct and clinically invisible tumour by curettage might have contributed to delay in early recognition and hence the inadequate treatment of a recurrence.

The failure of curettage is probably related to the fact that nests of tumour cells extend beyond the clinical and radiographic margins of the lesion and are therefore impossible to eradicate by a scrapping procedure. Many types of chemical and electrical cauterisation have been used by surgeons in conjunction with curettage but they have reported only a slight improvement in cure rate.

Operative procedure

Curettage generally implies the removal of pathological tissue by vigorous scrapping. It is primarily approached intra-orally. A mucoperiosteal flap is reflected to create adequate access. In the mandible, the approach is usually restricted to the buccal aspect. Hazards of lingual access include injury to lingual nerve & mandibular neurovascular bundle and exposure of facial spaces in the floor of the mouth. When the lesion occurs in the maxilla, either a palatal or buccal / labial approach may be used.

Depending on the thickness of the cortical bone, either a rongeur or surgical bur is used to remove sufficient bone to expose the underlying pathoses. Then the lesion is removed from the underlying bony cavity by the use of angular / straight curettes, using a teasing motion with the convex surface of the curette placed against the bony wall. Usually the vital structures like the contents of the mandibular canal are displaced to the distal wall of the surgical defect; sometimes these structures are sacrificed. After the gross lesion is removed using the largest curette that can be employed using the available access, a margin of apparently normal bone should be removed by aggressive scrapping. After thus removing 1 to 3 mm of surrounding bone, all margins are smoothened with a rongeur or a large round bur. Adjunctive treatment like cauterisation may be employed at this stage.

The bony cavity is then irrigated liberally with normal saline. Small wounds are then closed primarily. Large wounds are packed with gauze impregnated with compound tincture of benzoin, balsam of Peru or Whitehead’s varnish. The gauze is layered in tiers. In the case of maxillary procedures, if the procedure has violated the maxillary sinus, the end of the gauze packing may be exited through a nasal antrostomy. The mucoperiosteal flap is then reapproximated and sutured. A topical antibiotic may also be used with the gauze pack.

The pack is removed approximately 2 to 3 inches everyday until the surgical defect is filled with granulation tissue. The patient is put on an antibiotic regimen and scrupulous oral hygiene is maintained.


Numerous complications have been reported following conservative therapy, particularly extensions to vital structures. Other reported complications are seeding into the lungs, direct extension into the brain and malignant transformation (Tekeuchi et al – 1981). Moreover, the curettage procedure breaks the cortical barrier, thus paving the way for residual tumour to grow into the soft tissues, which then becomes more difficult to treat.

Cautery (desiccation)

Various types of cautery have been used in the treatment of ameloblastoma, primarily as an adjuvant to curettage, but in some cases as a primary mode of therapy. Chemical agents (most recently Carnoy’s solution), electrocautery and cryotherapy have all been used. Cauterisation is basically an attempt to eradicate the tumour that has infiltrated beyond the clinical and radiographic margins of the tumour. In general, the use of cautery is empirical because of our lack of knowledge as to

  • how far the tumour in each case has extended into the cancellous bone
  • how far the caustic agent (heat / chemicals) penetrates into the cancellous bone
  • how effective is the agent in eradicating the tumour cells and
  • the possible harmful effects to normal tissue

Electrocoagulation (thermal cautery)

Mehlisch et al (1972) states that cautery when used as primary treatment resulted in a 50% recurrence rate. Thus it is potentially a more effective therapy than curettage. The secondary ischaemia and necrosis that occurs for some distance from the margins of the tumour may destroy the invading tumour cells.

Cautery has frequently been employed as an adjuvant to other methods of therapy to give a better result (Gardner and Pecak – 1980). Mehlisch et al reported no recurrences in 2 patients treated with this method.

Chemical cauterisation

In recent years, Carnoy’s solution (a fixing agent consisting of a mixture of absolute alcohol, chloroform and glacial acetic acid, which is sometimes modified by the addition of ferric chloride) has been advocated as a treatment for odontogenic keratocysts. Stoelinga and Bronkhorst (1988) have used it after enucleation in the treatment of unicystic ameloblastoma and reported no recurrences.

The depth of penetration of Carnoy’s solution is known. It penetrates cancellous bone up to 1.5 mm after 5 minutes and up to 1.8 mm after 1 hour (Voorsmit et al – 1981). In 1982, Voorsmit further reported that the injurious effect of this agent on adjacent soft tissue is negligible.

Eventhough Carnoy’s solution has been used in the treatment of classical multilocular ameloblastoma also, there is no information in the literature concerning its efficacy. Nevertheless, the use of Carnoy’s solution appears to be harmless and has the potential of reducing recurrences after curettage.


In the past two decades, cryotherapy has also been advocated in the treatment of ameloblastoma (Bradley-1986, Holland and Mellor-1981). It is used mainly as an adjunct to curettage in the hope of reducing recurrences. The obvious advantage of cryotherapy is that it is possible to devitalise the tissue with liquid nitrogen to a depth of 1.5 cm, the margin that is frequently used for surgical resection. In fact, the jaw can be frozen through its entire thickness if necessary.

However, there are potential complications such as sequestration, pathological fracture, transient anaesthesia of mandibular nerve etc. In general, the more extensive the freezing, the greater the risk. This is a comparatively new treatment method, and has not been used extensively as yet in the treatment of ameloblastoma.

Another method which has been described (Weaver and Smith-1963, Bradley-1978) in which the affected segment of bone is excised, frozen in liquid nitrogen to devitalise the tissue, and then reimplanted as an autogenous graft. This is only in experimental stages at present.

En bloc resection

The en bloc or marginal mandibular resection is a surgical procedure in which the entire tumour is removed intact with a rim of uninvolved bone while maintaining the continuity of the jaw. This procedure is advocated for the treatment of ameloblastoma when the lesion does not extent closer than 1 cm to the inferior border of the mandible.

Kramer (1963) reported that although there is an invasion of cancellous spaces of the bone by finger-like projections, the tumour does not invade the haversian systems of compact bone. He concluded that in relation to cortical bone, the clinical and radiological margin may be regarded as the true margin. However, in the cancellous bone, a margin of 1 to 2 cm beyond the clinical and radiological limit is considered the minimum acceptable margin. Various authors have reported good results with en bloc resection (Mehlisch et al – 1972, Sehdev et al – 1974). But most surgeons are of the view that for lesions of the maxilla, en bloc resection is not as successful and recommend segmental resection (Björklund – 1979, Chaudhuri – 1975).

Operative procedure

En bloc resection can be done from either an intra-oral or an extra-oral approach. The former is used when there is a good access and when the lesion is anterior to third molar region. The extra-oral approach is used when the lesion involves the ramus of the mandible or when immediate reconstruction is planned.

Intra-oral approach

For large mandibular lesions requiring the removal of large sections of bone, a midline lip-splitting incision is often made to increase the access into the posterior region. This is not necessary for lesions requiring less bone removal.

Connecting vertical incisions are made on the buccal and lingual sides through the mucoperiosteum 2 cm anterior and posterior to the anticipated area of resection. These incisions should extend deep into buccal and lingual folds. The teeth bordering the surgical margin should be extracted. Next, horizontal incisions connecting the lower ends of vertical incisions are made. The buccal and lingual mucoperiosteal flaps are then developed, but not reflected superiorly over the region of bone to be removed.

On exposure of the mandible, the bony segment is sectioned with an air-driven saw or bur, at least 1 to 1.5 cm from the radiographic margin of the lesion. For segmental resection of the same region, a continuity defect is created by removing an entire segment of the mandible, including the lower border.

Haemorrhage arising from the bony segments can be controlled by crushing the bone over small blood vessels with a blunt instrument or by using bone wax. The mucoperiosteum is then undermined both lingually and facially to relieve tension. They are approximated with interrupted silk sutures.

Post-operatively, good oral hygiene should be maintained and antibiotic coverage is necessary. The use of intermaxillary fixation depends on the amount of remaining bone.

Extra-oral approach

This is similar to that used in segmental mandibular resection except that a margin of inferior border of mandible is left intact.

Segmental (partial) mandibular resection / hemimandibulectomy

Segmental resection of jaws, including maxillectomy and hemimandibulectomy has been the most commonly used treatment for ameloblastoma. Obviously, those who used this method have reported the least number of recurrences.

Operative procedure

Depending on the size of the mandibular segment to be removed, a lip-splitting incision may or may not be necessary. When the lower lip is being split, finger pressure should be applied on either side of the vertical incision to control bleeding from labial arteries. After the lip split, larger arteries may be ligated, and the smaller ones cauterised.

A submandibular incision about 1 to 2 cm below the inferior border of the mandible, is made from the angle of the mandible to join the vertical lip incision. This incision is made through the skin and subcutaneous tissue, exposing the platysma.

Next, intra-orally, a horizontal incision is made through the mucoperiosteum on the facial and lingual aspects of alveolar ridge. These are then curved on to the retromolar region. The facial and lingual flaps are advanced below the horizontal incision using a periosteal elevator. The lingual flap is raised as deep as to expose the mylohyoid attachment.

A vertical mucoperiosteal incision is made 0.5 cm proximal to the anticipated anterior bony cut. The buccal flap beneath the horizontal incision is reflected to expose the mental neurovascular bundle, which is ligated and sectioned. Preservation of the marginal mandibular branch of the facial nerve, which would be located in the lateral aspect of the facial flap, is important for the normal function and appearance of the lower lip post-operatively.

The masseter is resected at the lower border of the mandible, and reflected from the ascending ramus using periosteal elevators. The medial pterygoid is then reflected from the lingual surface. The temporalis is incised and reflected off the coronoid process. The lingual flap is retracted to expose the mylohyoid, which is then severed off its attachment. The facial flap should expose the entire lateral surface of the body and ramus of the mandible.

Using an air-driven saw, bur or a Gigli saw, a vertical cut is made through the mandible anterior to the lesion. Using bone forceps, the proximal part of the mandible is rotated laterally, exposing the inferior alveolar nerve and vessels, at the lingula of the mandible. They are ligated and cut adjacent to the mandibular foramen. The lateral pterygoid muscle is sectioned from the condyle and TMJ capsule. The capsule is cut with a scalpel and the segment of mandible is disarticulated and removed using bone-holding forceps. Bleeding ensuing from the severed muscles is controlled by digital pressure, coagulation or ligation, depending on the size of the bleeding vessel.

In order to obliterate the dead space and for haemostatic purposes, the opposing surfaces of medial pterygoid and masseter are sutured together, and a drain is placed in the region. When immediate reconstruction is planned, the graft is inserted at this stage. The submucosal surfaces are approximated and sutured with buried chromic gut or vicryl. Mucosal surfaces are then closed with interrupted sutures. The vertical and horizontal mucoperiosteal flaps are approximated and closed, covering all exposed bone. The vermilion border of the lip is re-approximated and sutured. The mucosal surface inside the lip is restored in layers. The deep tissues in the submandibular region are re-approximated to obliterate the dead space. The platysma is then re-approximated and sutured. The subcutaneous tissues are sutured with gut / vicryl. The skin is closed with silk or prolene sutures. The external portion of the drain is secured with a binding suture.

A firm pressure dressing is applied on the lateral surface of the resected area. Intermaxillary elastic traction using previously placed archbars will minimise mandibular deviation, and encourage a more ideal wound healing. The patient should be fed through a naso-gastric tube for a week and scrupulous oral hygiene should be maintained. Dressings should be changed daily. Removal of drain depends on the amount of drainage. Alternate skin sutures are removed after 4 days and the remaining ones, after 6 days. After that, the naso-gastric tube may be removed and oral feeding may be begun.

Marginal (partial) maxillectomy

The marginal maxillectomy is the surgical procedure most often used for ameloblastoma of maxilla when the maxillary sinus is not involved.

Operative procedure

Marginal maxillectomy is accomplished through an intra-oral approach. A mucoperiosteal incision is made approximately 1.5 to 2 cm in all directions from the underlying tumour. It may be necessary to extract one or more teeth to complete these incisions.  A small mucoperiosteal flap is reflected on the uninvolved bone. Using an air-driven saw or bur, the vertical bony cuts are made at the height of the alveolar ridge. These cuts extent into the maxillary sinus. If the tuberosity is to be removed, a separation of the pterygoid plates can serve as the distal cut. A horizontal buccal osteotomy is made to connect the buccal cuts. A similar osteotomy is performed on the palatal aspect also. The saw blade / bur is directed to enter the maxillary sinus or the nasal cavity as is necessary. Then the resected portion of maxilla is easily removed.

A gauze strip impregnated with a suitable medicament is firmly packed into the maxillary sinus and is held in position with a pre-fabricated obturator. The surgical defect is later repaired with a permanent obturator, which permits inspection for recurrence.


Maxillectomy is the surgical procedure of choice when an ameloblastoma extends into the maxillary sinus or when a maxillary lesion has recurred.

Operative procedure

A full-thickness midline incision is made in the upper lip from skin to the mucous membrane. The incision is continued along the inferior portion and lateral aspect of the nose, to the medial canthus of the eye, and is joined by a transverse infra-orbital incision close to the margin of lower eyelid (Weber-Ferguson incision). An intra-oral incision is made through the mucosa at the height of the labio-buccal sulcus from the anterior vertical incision to the tuberosity. This incision should be made deep to touch the lateral wall of the maxilla.

Using blunt and sharp dissection, a sub-periosteal flap is raised superiorly to the inferior rim of the orbit, exposing the whole lateral surface of the maxilla. The infra-orbital vein and nerve are identified and ligated. The periosteum of the floor of the orbit is elevated and the contents of the orbit are safely retracted, exposing the inferior orbital fissure.

The facial flap should be retracted posteriorly, exposing the zygoma, which is sectioned vertically using a saw / bur / chisel and mallet. The cut is directed medially at an oblique angle extending through the zygoma along the floor of the orbit to the inferior orbital fissure. The nasal soft tissues are reflected from the maxilla and the adjacent surface of the nasal bone.

The fronto-nasal process of the maxilla is sectioned vertically from the orbit to the nasal fossa. The cut is directed laterally and obliquely, extending along the floor of the orbit to the inferior orbital fissure.

A vertical incision is made over the midline of maxilla, and a gingival flap is reflected. The central incisor is extracted, and a bone cut is placed through the tooth socket and alveolar process to the floor of the nose. A midline incision on hard palate is made which is turned transversely at its junction with soft palate to join the buccal incision. The hard palate is sectioned along the floor of the nose just medial to the nasal septum. The muscles of soft palate and superior constrictor are divided. Retraction of the cut muscles expose the pterygoid plates of the sphenoid bone and the muscles attached to it. Pterygomaxillary disjunction is done with a chisel. The sectioned maxilla is made free of any remaining skeletal attachments.

Pressure / ligation is used to control bleeding from the severed muscles. Often the greater palatine vessels have to be ligated and sutured. The exposed surfaces of sphenoid and ethmoid are smoothened and a split-thickness skin graft is placed in the raw bony and soft tissue surfaces. The raw edges of facial surfaces are reapproximated in layers. A 1-inch wide gauze strip impregnated with a suitable medicament is packed firmly against the skin graft. This packing should be layered firmly and may be maintained with a pre-fabricated maxillary surgical splint.

A firmly placed pressure bandage is applied over the lateral facial area of the surgical defect. Post-operative feeding is through naso-gastric tube, inserted through the intact nostril and antibiotic therapy is instituted. Proper oral hygiene should be maintained. Extra-oral dressings are changed daily, and after six to eight days, the facial sutures and packing are removed.

Factors governing the choice of treatment method

Since the aggressiveness and recurrence of ameloblastoma are controversial issues, there is still no single accepted treatment method for ameloblastoma. Before arriving at a final decision as to which treatment is best suited for a particular patient, a number of actors need to be considered.

Age and health of the patient

These are of prime importance in selection of treatment method. A mutilating radical surgery would be unwarranted in a small child or a very elderly patient, but for different reasons. In the case of the former, it is a question of post-surgical morbidity i. e. leaving the patient with a crippling surgical defect for the rest of his/her life. Again, for the very old patients, conservative therapy may be considered as the recurrences are known to take several years to occur.

It needs to be assessed whether the general health of the patient permits any radical procedure. It would not be preferable to do a radical surgery in medically compromised patients and in those with advanced systemic disease.

Clinical type of ameloblastoma

The unicystic type of ameloblastoma is known to have a much better prognosis than the typical intra-osseous type, and a conservative therapy is considered adequate. Some workers consider certain variants like basal cell ameloblastoma and granular cell ameloblastoma as more aggressive than the other types. Even in the unicystic type, the ‘invasive’ type (involving the connective tissue) is thought to be a more aggressive variety.

Site of the lesion

The porous nature of the maxilla, along with the proximity to the maxillary sinuses and nasopharynx, means a poorer prognosis for maxillary lesions than mandibular ones. So the maxillary tumours always need a more radical therapy than mandibular tumours.

The adjacent vital structures like neurovascular bundles and maxillary sinus should be preserved when possible, but may be sacrificed if the tumour has grown very close.

A tumour which is confined to the bone without perforation of cortical plates offer a better prognosis than one that has invaded soft tissues. Tumours with extra-osseous component make complete removal more difficult and require sacrifice of more (apparently) normal tissue.

An ameloblastoma in an inaccessible site like the base of the skull or pterygo-maxillary fissure area present a difficult surgical problem. Other measures like megavoltage radiotherapy has a role in such cases.

Size of the lesion

A small tumour confined to the mandible may be treated by conservative measures at the first instance while the larger ones require more radical surgery.

Chances of recurrence

The chance of recurrence after the procedure should always be borne in mind. Obviously the conservative methods like curettage and cauterisation have a higher risk of recurrence than the radical methods.

Morbidity and complications

The expected post-surgical morbidity and the complications related to the particular method are very significant factors since resection of jaw bones has serious implications on the functional and psychological well-being of the patient.

Patient preference

The patient should be informed of all treatment options, complete with the advantages and risks of each treatment modality, and reconstruction options. He/she should then be given a role in the final decision making.

Availability for follow-up

This becomes an essential factor particularly if conservative treatment is planned. It is better to perform a radical operation if there is less chance of patient turning up for follow-ups. A follow-up of at least 10 years is essential following curettage.

Controversies concerning treatment

The controversies concerning treatment modalities lie partly in the philosophy of individual surgeons. They should fully understand the clinical behaviour of the tumour and the results to be expected of each treatment method.

Curettage vs. radical surgery

There has been a lot of argument over the years between the surgeons comparing the risk of recurrence of conservative methods to the morbidity following radical surgery.

The recurrence rate for typical intra-osseous ameloblastoma treated by curettage is high, varying from 55-90% in various series (Gardner- 1987, Gardner and Pecak-1980) and in the posterior maxilla, it happens to be100% (Sehdev et al-1974). Muller and Slootweg (1985) have concluded that it is 75% in both jaws put together. This is thought to be because of the infiltrative capacity of ameloblastoma into cancellous bone. In comparison, the recurrence rate is much less for marginal / segmental resection (15% in Muller and Slootweg series). It follows that curettage is not a reliable method in the treatment of multilocular ameloblastoma.

It is generally agreed that curettage should never be used in the treatment of lesions of posterior maxilla since this region lacks a dense cortical plate and there is risk of invasion of the cranium through the foramina leading from the pterygomaxillary fossa, eventually killing the patient. Ameloblastoma of this site should be treated by marginal or segmental resection, depending on the size. Even in the ascending ramus of the mandible, curettage appears to be inadvisable due to the proximity to vital structures.

In the case of the body of the mandible, small tumours may be treated by curettage, provided the surgeon is fully aware of the high risk of recurrence and will be able to follow the patient closely for the next 10 years or more. Moreover, the patient must be fully aware of the risk. In short, the conditions where curettage is thought to be justified are the following.

  1. A case of ‘non-invasive’ unicystic ameloblastoma in the mandible or anterior maxilla.
  2. A small tumour in the body of the mandible in a child or a young adult, provided the patient can be followed up for 10 years or more.
  3. A small tumour in the body of the mandible in an elderly patient, as ameloblastoma takes several years to recur.
  4. A small tumour in the body of the mandible in a medically compromised patient.

It is always advisable to use some kind of cauterisation as an adjunct to curettage. All teeth in the region should be extracted. These steps might reduce the risk of recurrence.

Feasibility of retaining the lower border of mandible

The decision whether to do a marginal resection or a segmental resection depends largely on the size of the lesion. Obviously a segmental resection needs to be done if the cortical plate has been thinned or expanded, as this situation increases the risk of pathologic fracture.

However, if the lower cortical plate has not been involved with the tumour, a sparing of the inferior border should be performed. This is because the lower border of the mandible consists of a thick layer of dense cortical bone, which is not invaded by ameloblastoma. Thus retaining the inferior border does not increase recurrence.

Timing of reconstruction

In the past, many surgeons were sceptical about doing a primary reconstruction following segmental resection because of the uncertainty regarding the risk of recurrence. The reconstruction was delayed for 6 to 12 months or even more, since it was feared that recurrence might occur in the bone graft. In such cases, a metallic (stainless steel) reconstruction plate is used to maintain the apparent continuity of the jaw and to make the healing occur in near normal positions (El Fattah-1999). Disadvantages of this approach are

  1. Need for a second surgical procedure
  2. Fibrosis that accompanies the healing makes dissection more difficult during the secondary procedure.
  3. The secondary graft is placed in a bed of compromised vascularity, increasing the risk of graft rejection.

Most recent series have reported very good results with primary reconstruction and cases of recurrences on the graft have been very few and far between, especially when segmental resection has included 1 to 2 cm of apparently normal bone tissue.

Reconstruction options

Radical surgeries like segmental resection, hemimandibulectomy and maxillectomy leave the patient with a thoroughly incapacitating aesthetic and functional deficit. The functions affected include mastication, swallowing, speech and respiration in the case of maxillary surgeries. It is the responsibility of the surgical team to use the best available technique to reconstruct the jaw so that the post-operative quality of life would remain on the high side.

Maxillary reconstruction

Maxillary reconstruction is a challenging procedure for a reconstructive surgeon because of the complex shape and multiple functions of the region. The reconstruction options include

  1. Prosthesis – Obturator and splints
  2. Local soft tissue flaps

Buccal and palatal advancement flaps

Cheek flaps

Buccal pad of fat

  1. Regional flaps

Temporalis – myofascial / myo-osseous

Trapezius – muscle / myo-cutaneous / osseo-myo-cutaneous

  1. Free flaps

Rectus abdominus

Radial forearm

Iliac crest


The osseous component of these flaps includes rib, scapula, calvarium and mandibular symphysis.

Mandibular reconstruction

Mandibular reconstruction can also be accomplished by a variety of means. They include

  1. autogenous vascularised bone by pedicled flaps
  • Clavicle pedicled on sternocleidomastoid
  • Rib pedicled on pectoralis major
  • Scapula pedicled on trapezius
  • Calvarium pedicled on temporalis
  • Rib pedicled on latissimus dorsi
  1. autogenous vascularised bone by free flaps
  • iliac crest based on deep circumflex iliac artery
  • fibula based on peroneal artery
  • scapula based on circumflex scapular artery
  • radial forearm based on radial artery
  • rib based on intercostal artery
  • second metatarsal
  • calvarium based on superficial temporal artery
  1. autogenous non-vascularised bone
  • calvarium
  • iliac crest
  • rib
  • fibula
  1. allografts
  2. xenografts
  3. alloplastic materials
  • stainless steel reconstruction plate
  • hydroxyapatite

Allografts and xenografts have fallen out of favour of most authorities. When they are used, it is as banked bone after lyophilisation to prevent antigenicity-related problems.

Autogenous vascularised bone gives the best result because of its reliable blood supply. Free flaps including iliac crest are considered the best choice because of their width which facilitated dental rehabilitation by implants or dentures. Disadvantages of free tissue transfer include longer duration of surgery, and the need for sophisticated and expensive equipment and trained personnel. In the absence of these, autogenous non-vascularised bone is a good compromise option, especially for reconstructing small defects.


Ameloblastoma of jaws is a benign neoplasm of odontogenic origin with local invasive capacity. It presents clinically as a painless swelling and as a multilocular radiolucency in radiographs. It has a number of histologic variants.

Radical resection is the generally accepted treatment modality except in certain specific instances. The defect created by resection may be reconstructed primarily using various means, autogenous vascularised bone being the most preferred choice.


  1. Burket’s Oral Medicine: Diagnosis and treatment. 9th MA Lynch, VJ Brightman, MS Greenberg. 1994
  2. A Textbook of Oral Pathology – 4th WG Shafer, MK Hine and BM Levy. 1993.
  3. Pathology of Tumours of Oral Tissues. 4th RB Lucas. 1984.
  4. Contemporary Oral and Maxillofacial Surgery. LJ Peterson, E Ellis, JR Hupp and MR Tucker. 1992.
  5. Surgery of the Mouth and the Jaws. JR. Moore. 1985.
  6. Oral and Maxillofacial Surgery (Vol. II). DM. Laskin. 1987.
  7. Controversies in the nomenclature, diagnosis and treatment of ameloblastoma. DG. Gardner – In Controversies in Oral and Maxillofacial Surgery. P Worthington, JR Evans (eds). 1994
  8. Grabb and Smith’s Plastic Surgery. 5th edition. S. J. Aston, R. W. Beasley, C. H. M. Thorne. 1991.
  9. Ameloblastoma –clinical features and management of 315 cases; Ademola A.O. –J Craniomaxillofac Surg 1993
  10. Muzzafar et al – Maxillary reconstruction – functional and aesthetic considerations; PSR 1999.
  11. Oral and Maxillofacial Surgery. WH. Archer. 1975.
  12. Basic Principles of Oral and Maxillofacial Surgery. Vol. I. Peterson, Marciani, Indresano (eds.). 1997.