Tropical ulcer is common throughout the wet tropics and sub-tropics. It is responsible for significant acute and chronic morbidity, particularly in rural communities, and its complications are sometimes lethal.

Acute tropical ulcer is a specific disease entity, whereas chronic tropical ulcer is the end result of a number of disease processes, including acute tropical ulceration, trauma, and infection. Tropical ulcer is known by a variety of local names such as naga sore, Aden ulcer, rice pickers' ulcer, Rhodesia sore, or veldt ulcer. Phagedenic ulcer (from the Greek ‘phagein’ to eat) is a synonym used particularly by French workers to emphasize the progressive nature of the skin necrosis seen in acute tropical ulcer.

Tropical ulcer is endemic in village communities throughout the tropics and in some subtropical regions. It is most common in the wet tropics and its incidence rises sharply in the wet season. Wherever nutrition, living conditions, and hygiene are poor, a high incidence of acute tropical ulcer will be found, especially amongst children and young adults, who suffer repeated trauma to their legs during their daily work in the field or rainforest. Medical services are often deficient in these areas and patients are rarely able to leave their work for periods of rest or treatment. Repeated trauma and infection compound the problem and lead to a high incidence of chronic ulcer.

Reliable statistics are generally not available, but a high incidence of tropical ulcer has been reported from areas of South America, West Indies, Africa, India, Vietnam, the Philippines, Malaysia, Indonesia, Melanesia, and the West Pacific. Studies from different parts of Africa indicate that up to 30 per cent of presentations to hospital are for tropical ulcer. Presentations to local aid posts and traditional healers are probably even more common. Most studies show the disease to be more common in males. Epidemics of acute tropical ulcer have occurred in Africa, Melanesia, and the Pacific Islands. Tropical ulcer was common amongst jungle fighters and prisoners in the tropics during the Second World War. Squamous cell carcinoma arising in chronic tropical ulcer is the most common skin cancer of the tropics.

Acute tropical ulcer results from a full-thickness necrotizing bacterial infection of the skin. In most cases, a minor puncture wound or dirty laceration provides an entry portal and initiates a spreading bacterial cellulitis which rapidly produces dermal gangrene, probably by a direct cytotoxic effect.

Initially a trivial injury to the leg may be disregarded and the patient remain ambulant. Saliva may be used to dress the wound. The necrotizing infection begins in a few days with the development of a painful papule and foul-smelling watery discharge or local blistering. A central area of gangrene rapidly becomes evident and the process spreads at a variable rate until it is brought to a halt, presumably by the host's defence mechanisms, and demarcation occurs. The patient may be febrile at this stage and there is an acute inflammatory response in surrounding skin. A hard slough usually forms after about a week and separation begins (Fig. 2) 2695. Lymphadenitis is not prominent at any stage. After about 2 weeks, slough separation reveals a layer of foul yellow or grey-green pus adherent to friable, infected granulation tissue (Fig. 3) 2696. In most cases the infection slowly resolves, leaving a bed of healthy granulation tissue surrounded by normal skin which may show signs of early epidermal growth and migration from the ulcer margin; this constitutes a typical acute tropical ulcer (Fig. 4) 2697.

The ulcer is virtually always on the leg below the knee, common sites being the lower one-third just above the malleoli, on the shin or dorsum of the foot. Tropical ulcers rarely occur above the knee and necrotizing skin infections of the trunk or upper limbs are best regarded as separate clinical conditions. If the ulcer remains unhealed after 3 months, it is regarded as chronic.

Full-thickness skin necrosis is produced by a synergistic bacterial infection in a host who may be predisposed by poor nutrition or intercurrent illness.

Since 1899 researchers have observed large numbers of fusiform bacilli and spirochaetes in the early stages of the disease. These were thought to be the same organisms as those found in patients with Vincent's stomatitis, gingivitis, and cancrum oris, and known as Fusobacterium species and Treponema vincenti. Similar organisms have been isolated from stagnant water and mud. Acute ulcers have been produced experimentally in man and animals using material from acute tropical ulcers or patients with acute suppurative gingivitis. It is thought the pathogens may gain access to a skin breach by direct inoculation at the time of injury or with the help of flies feeding on minor leg wounds. Some infections may originate from the patient's own mouth, if saliva is used as a dressing for a leg wound. Coliforms and Gram-positive cocci have also frequently been isolated from early ulcers.

In 1987 workers using anaerobic culture techniques identified and characterized a new species from acute ulcers known as Fusobacterium ulcerans. This bacterium is markedly cytotoxic to animal and human cell lines in tissue culture. A combination of aerobic and anaerobic culture techniques with light and electron microscopy has demonstrated anaerobes in combination with aerobes or facultative anaerobes, especially in early ulcers. The anaerobes are usually a combination of Fusobacteria, Bacterioides species, or anaerobic cocci; they are always found in the presence of aerobes. Spirochaetes are also identified by microscopy in early ulcers. It is thought that aerobes and anaerobes act synergistically to produce the dermal gangrene characteristic of acute tropical ulcer. The role of the spirochaetes in the pathogenesis of acute tropical ulcer is not known.

Streptococci, diphtheroids, and other bacteria are isolated with variable frequency from established ulcers; their significance is uncertain.

Repeated trauma to the legs provides numerous opportunities for organisms to penetrate the epidermis and may also reduce local defence mechanisms. Dirt, damp, and a high ambient temperature may provide an environment rich in the pathogens specific for acute tropical ulcer and suitable conditions for their rapid proliferation.

Population studies support an association with poor nutrition (especially protein and vitamin deficiency), and intercurrent illnesses such as malaria, respiratory infection, and anaemia; however, evidence of a causative effect is generally lacking, and certainly at a clinical level many patients with tropical ulcer are not overtly malnourished or debilitated.

The similarity between acute tropical ulcer and other acute necrotizing skin conditions such as cancrum oris, necrotizing fasciitis, Fournier's gangrene, and Meleney's synergistic gangrene is striking. All of these conditions are probably produced by a mixed infection with aerobic and anaerobic bacteria. It has been postulated that tissue oxygen tension is reduced by proliferating aerobic bacteria, thereby promoting rapid growth of the anaerobic species. In the case of acute tropical ulcer, toxins produced by the anaerobe F. ulcerans probably cause rapid tissue necrosis by a direct cytotoxic effect.

Histological examination of the acute lesion shows the features of gangrene with evidence of skin necrosis and infection. An acute inflammatory response is most evident at the periphery of the lesion, and is presumably important in halting the spreading infection. After the slough has separated, the ulcer is covered with a layer of fibrin, necrotic debris, and organisms. As the infection resolves, granulation tissue rapidly fills the base of the ulcer and epithelial proliferation and migration begins at the periphery. Untreated, the acute ulcer will evolve along one of the following paths. Firstly, healing by epithelial growth from the periphery. The new epithelium is thin, lacks adnexal structures, and is often completely devoid of pigment. It has little resistance to further trauma, infection and ultraviolet radiation. Secondly, initial healing with later episodes of trauma, infection and repeated ulceration, and thirdly chronic ulceration.

Most patients present soon after slough separation with an infected granulating ulcer of the lower leg. The history, site, and appearance of the ulcer are usually typical, and allow an accurate diagnosis on clinical grounds. Early presentation with an infected papule or minimal skin necrosis is not so common. A full history and clinical examination of the patient will reveal any predisposing factors such as malnutrition and intercurrent illness.

Swabs should be taken from the edge and floor of the ulcer for microscopy, culture (aeobic and anaerobic), and sensitivity. Special stains and dark-ground microscopy may be helpful to demonstrate spirochaetes and fusiform bacilli. Biopsy may be indicated if the clinical diagnosis is uncertain or neoplasm is a possibility. Haematological and other tests may be indicated to elucidate associated conditions.

Improved nutrition, hygiene, and socioeconomic conditions are suggested as factors required to reduce the incidence of acute tropical ulcer. Use of suitable footwear and long trousers with improved working conditions should reduce the incidence of trauma to the legs and subsequent wound contamination.

Prompt appropriate treatment of minor leg wounds should reduce the incidence of exogenous and endogenous infection which may progress to acute tropical ulcer.

Patients who present early with an infected papule should be treated promptly with systemic antibiotics. Crystalline penicillin, perhaps with the addition of metronidazole, will abort the infection and prevent skin necrosis and ulceration in most cases.

Most patients present later; once skin necrosis is established, surgical debridement with removal of slough and non-viable tissue is required to control infection. Patients who are septicaemic or who have large areas of skin necrosis require urgent admission for bed rest, elevation of the limb, and systemic administration of antibiotics prior to surgery.

Once any spreading infection is controlled and necrotic skin removed, local measures are used to eradicate the infection. Debridement is followed by frequent dressings with rest and elevation of the limb. Antibiotics should be ceased. Many substances have been used to dress tropical ulcers including paw-paw, pineapple pulp, honey, various enzymes, and antibiotics. Dilute Eusol (Edinburgh University Solution of Lime) is an effective and cheap application for removing residual slough. Isotonic, stabilized sodium hyopochlorite solution is also effective in controlling local infection. The dressing of choice should be effective, sterile, cheap, non-toxic, non allergenic, painless, and easily applied. Frequent changes of moist dressings are the key to rapid control of local infection. Isotonic saline should be used once the ulcer is clean and granulating. Topical antibiotics are not effective and lead to colonization of the ulcer by resistant bacteria; their use is contraindicated.

Healing is completed by providing epithelial cover to the clean granulating ulcer. Spontaneous epithelialization will occur at about 1 mm/day from the ulcer edge. Small ulcers may be allowed to heal in this way, but the result is a thin, unstable, depigmented scar which is susceptible to repeated ulceration following mechanical trauma, infection and solar radiation. Autografts of split-thickness skin allow rapid closure of the defect and produce a more stable pigmented scar. It would be ideal to close these ulcers with normal full-thickness skin, however this is not currently feasible for most patients with tropical ulcer. Split-thickness skin grafting is relatively cheap, widely available, and has a high rate of successful take. The healed graft lacks hair follicles, sweat and sebaceous glands, but is pigmented and so has at least some protection from solar damage and carcinogenesis. This is the most widely used technique for achieving cover of acute ulcers.

As a general rule, ulcers 1 cm or more in diameter should be closed promptly by a split skin graft once infection has been controlled. Patients must be advised to protect the healed graft permanently from trauma, infection and solar exposure.

Many acute tropical ulcers fail to heal and others break down after repeated episodes of trauma and infection. After an arbitrary period of 3 months they may be defined as chronic. Trauma, non-specific infection, and various other disease processes also give rise to chronic leg ulcers which are clinically indistinguishable from chronic tropical ulcer, except perhaps on the basis of history. For practical reasons, all such non-specific ulcers are usually included in the definition of chronic tropical ulcer.

Chronic ulcers show macroscopic evidence of chronic inflammation, with repair proceeding in the presence of active infection and tissue damage. The granulating ulcer floor is often pale, indolent, and firm. There may be old scars from healed ulcers which have broken down on several occasions. Regenerated epithelium is often thin, atrophic, and devoid of all pigment. The ulcer base is usually attached to the tibial periosteum and bone changes due to periosteal reaction and osteitis may be marked. Dense fibrosis may produce crippling contractures, or may lead to lymphatic obstruction with lymphoedema and secondary skin changes.

After several years, regenerating epithelium at the ulcer edge may show evidence of poorly controlled growth with heaping-up of the ulcer margin or migration of abnormal epithelium across the ulcer floor. Histological studies at this stage will usually reveal pseudoepitheliomatous hyperplasia, which seems to grade imperceptibly into very well differentiated squamous cell carcinoma as time passes.

The acute ulcer of a child often becomes a life-time affliction for the adult. The ulcer may show little change over decades until an increase in size, odour, or pain signals overt malignant change. A latent period of less than 3 years is rare; periods of 20 to 40 years are more usual. Chronic or recurrent ulcers, and even healed ulcers have a high risk of malignant change in the long term. Virtually all malignant ulcers are squamous cell carcinomas. Verrucose, exophytic types are common (Fig. 9) 2702, while ulcerative, invasive carcinomas rapidly erode the underlying tibia or fibula.

Many acute ulcers never heal, but become chronic indolent lesions. Healing is thwarted by repeated injury due to infection, mechanical trauma, and solar radiation. Chronic inflammation, dense fibrosis, bone changes, dystrophic calcification, lymphatic obstruction, and epidermal dysplasia all become evident and render healing unlikely.

Alternatively, repeated episodes of ulceration may be followed by more prolonged periods of healing and eventually chronic ulceration. The fragile depigmented epithelium of a healed ulcer on the lower leg is easily breached by the combined effects of mechanical trauma, infection, and solar radiation. Each episode of ulceration leads to more marked fibrosis and epithelial dysplasia, with less orderly healing. Most ulcers overly the tibia and cause ‘reactive’ bone changes with osteitis and new bone formation.

The mechanism of malignant transformation in tropical ulcer is not known; however, the common antecedent to development of squamous cell carcinoma seems to be the depigmented, atrophic epithelium associated with both healed and chronic ulcers. Almost all chronic tropical ulcers show marked pseudoepitheliomatous hyperplasia which may evolve gradually into frank malignancy. An analogy with malignant change in burn scars, chronic osteomyelitis sinuses, fistulae, and other chronic ulcers is clear; malignant change in chronic tropical ulcers is a form of Marjolin's ulcer.

Most malignant tropical ulcers are well-differentiated histologically. Progression is often slow but relentless, with invasion of underlying bone and eventual pathological fracture. Metastases to the inguinal lymph nodes is usually manifest only when the primary is advanced, and distant spread is rarely a clinical problem.

Some patients will give a clear history of preceding acute tropical ulcer. However, many will not be able to give a diagnostic history of the initial ulceration, which may have occurred up to 50 years previously. Specific causes should be considered in all chronic ulcers. Clinical evidence of long-term complications, particularly malignant change must be sought.

Swabs from the edge and centre of the ulcer should be taken for bacteriological studies including microscopy, cultures, and sensitivity, as well as studies for acid-fast bacilli. Biopsy is indicated where areas of depigmentation and tissue proliferation raise the suspicion of malignant change, or if diagnosis is uncertain. Radiography may be indicated to show bone changes, especially malignant erosion, pathological fracture, inflammatory changes, or dystrophic calcification. Haematological and other studies may be indicated to elucidate intercurrent illness and nutritional status.

Treatment of acute tropical ulcer that achieves good epithelial cover and the avoidance of subsequent trauma and infection to the leg will prevent the development of chronic or recurrent ulcer.

Treatment of uncomplicated chronic tropical ulcer comprises three main aspects; control of infection, control of any factors hindering healing, and provision of robust epithelial cover which will protect against further ulceration and malignant change.

Control of infection requires admission to hospital for rest in bed and elevation of the limb with frequent dressing of the ulcer. Some patients require systemic treatment with antibiotics, chosen on the basis of laboratory determined sensitivities.

Excision of indolent granulation tissue, areas of dense fibrosis, and severely scarred depigmented skin may be essential to achieve ultimate healing. Contractures and bone deformity will usually need to be corrected before the ulcer is covered. Bone infection must be controlled.

Epithelial cover is usually achieved by applying a split-thickness skin graft to the granulating bed of the ulcer or after excision of the ulcer. There is a higher rate of graft loss than in patients with acute ulcers, and the healed graft has all the disadvantages mentioned under the treatment of acute ulcers. Delayed graft breakdown and further ulceration are common, as split skin grafting does not provide adequate cover in many situations.

Free full-thickness skin grafts generally will not take on the ulcer bed, and local pedicle flaps in the lower leg are rarely adequate for the large defects present. Cross-leg flaps have been used in many tropical countries, but have several major disadvantages. Composite free tissue autograft with microvascular anastomosis (e.g. free rectus abdominis graft) is probably the most effective method of providing a robust repair for a large defect of the leg. However the technique is technically demanding and has not been widely used in this context; it is perhaps more appropriate in the treatment of malignant tropical ulcers.

Treatment of chronic tropical ulcer is demanding, particularly when surgical resources are limited. Permanent stable repair of the defect is essential to prevent long-term morbidity and to return the patient to normal activities. Recurrent and chronic ulceration are crippling and are associated with a high incidence of malignant change. There is potential for major improvement in the results currently achieved in treating most patients with chronic tropical ulcer.

All chronic tropical ulcers should be regarded as premalignant. The latent period from the first ulcer is rarely less than 5 years and is often about 30 or 40 years. Patients with white patches of atrophic epithelium seem particularly prone to malignant change.

Clinical diagnosis of the typical raised, everted edge or invasive ulceration of squamous cell carcinoma is not difficult, but the index of suspicion must be high and all suspect areas must be biopsied. Complete excision of the ulcer for histological examination may be necessary. Occasional confusion with chronic yaws, syphilitic gumma, or chronic osteomyelitis should be clarified by histology.

The clinical and histological differentiation between squamous cell carcinoma and pseudoepitheliomatous hyperplasia may be difficult. Sampling errors and difficulties with histological interpretation may lead to underdiagnosis of carcinoma: the clinical behaviour of the lesion is the final indicator of its true nature. Eight cases of sarcoma apparently arising in chronic tropical ulcers have been reported; however, virtually all malignant tropical ulcers are squamous cell carcinoma.

It is assumed that adequate treatment of acute and chronic ulcers prevents neoplastic change in the long term. However, there are no reliable studies to prove this, and squamous cell carcinoma is occasionally observed in patients who have undergone previous split skin grafting. Protection of chronic and healed ulcers (particularly white patches of skin) from the intense solar radiation of the tropics is thought to be very important in preventing malignant change.

On presentation, most patients have large squamous cell carcinomas with bone attachment or invasion. Despite advanced local disease, metastatic disease is not common and potentially curable primary lesions should be treated by surgery in otherwise well patients.

Small localized lesions may be treated by wide excision and split-thickness skin grafting; however, most lesions require more extensive surgery. Amputation has been the mainstay of treatment for large lesions showing extensive bone invasion. The associated morbidity is poorly tolerated in rural communities and alternatives have been sought for suitable patients. Wide excision of the carcinoma with surrounding skin, muscle, and tibial cortex leaves a large defect which is difficult to repair. One approach is to allow the defect to granulate and eventually achieve epithelial cover with a split skin graft. Although the limb is salvaged, it is fragile and prone to further ulceration and pathological fracture. In recent years large defects of the leg have been successfully repaired using free composite autografts with microvascular anastomosis. Techniques such as the free rectus abdominis flap have proven to be versatile, reliable, and technically feasible in a provincial hospital setting and a number of such repairs have been performed in suitable patients with good results. Widespread use of such limb-saving techniques has great potential for reducing the disability caused by chronic and malignant tropical ulcer.

Wide excision and free flap repair allows limb salvage in patients with moderately advanced carcinomas, but amputation remains indicated for many extensive lesions, pathological fractures, and most undifferentiated carcinomas.

Evidence of disseminated malignancy is usually a contraindication to radical local treatment of the primary lesion or of any inguinal lymph node metastases. Patients with clinically involved but operable nodes should be subjected to radical groin dissection for both therapeutic and prognostic reasons, once the possibility of infective adenopathy has been eliminated. Patients whose inguinal nodes are not clinically involved should be seen at regular intervals and advised to return promptly if they become aware of a groin mass. Prophylactic groin dissection is generally not indicated as the incidence of occult nodal metastases is low; it may be considered for patients with advanced or undifferentiated primary lesions or those who are not likely to return for review. However, the value of this procedure has not been proven in malignant tropical ulcer.

Treatment of patients with disseminated disease by palliative immuno-, radio-, or chemotherapy is rarely indicated or appropriate in most tropical countries; these techniques have little effect on survival, are expensive, often toxic, and may keep the dying patient away from his family. A generous supply of an effective oral analgesic is probably the most valuable treatment that can be offered.

Adelaye A, ed, Davey's Companion to Surgery in Africa. 2nd edn. London: Churchill Livingstone, 1987: 39 - 45.

Adriaans B. Tropical ulcer—a reappraisal based on recent work. Trans R Soc Trop Med Hyg 1988; 82: 185 - 9.

Adriaans B, Garelick H. Cytotoxicity of Fusobacterium ulcerans. J Med Microbiol 1989; 29: 177 - 80.

Basset A. Tropical phadenic ulcer. In Simons RDGP, Marshall J, eds. Essays on Tropical Dermatology. Amsterdam: Excerpta Medica; 1969: 25 - 33.

Clements FW. Tropical ulcer with special reference to its aetiology. Med J Aust 1936; 2: 615 - 43.

Connor DH, Neafie RC. Tropical phagedenic ulcer. In Binford CH, Connor DH, eds. Pathology of Tropical and Extraordinary Diseases. Washington DC: Armed Forces Institute of Pathology, 1976; Vol. 1: 199 - 201.

Foster HM, Webb SJ, Malignant change in chronic tropical ulcers. Asian J Surg 1984; 7: 232 - 6.

Loewenthal LJ. Tropical phagedenic ulcer—a review. Int Rev Trop Med 1963; 2: 267 - 91.

Morris GE, Hay RJ, Srinivasa A, Bunat A. The diagnosis and management of tropical ulcer in East Sepik Province of Papua New Guinea. J Trop Med Hyg 1989; 92: 215 - 20.

O'Brien HD. Treatment of tropical ulcers. Br Med J 1951; 2: 1544 - 51.

Robinson FC, Adriaans B, Hay RJ, Yesudian P. The clinical and epidemiological features of tropical ulcer (tropical phagedenic ulcer). Int J Dermatol 1988; 27: 49 - 53.

Taylor GI, Corlett RJ, Boyd JB. The versatile deep inferior epigastric (inferior rectus abdominis) flap. Br J Plast Reconstr Surg 1984; 37: 330 - 50.