Review Post-kala-azar dermal leishmaniasis Post-kala-azar dermal leishmaniasis E E Zijlstra, A M Musa, E A G Khalil, I M El Hassan, and A M El-Hassan Post-kala-azar dermal leishmaniasis (PKDL) is a complication of visceral leishmaniasis (VL); it is characterised by a macular, maculopapular, and nodular rash in a patient who has recovered from VL and who is otherwise well. The rash usually starts around the mouth from where it spreads to other parts of the body depending on severity. It is mainly seen in Sudan and India where it follows treated VL in 50% and 5–10% of cases, respectively. Thus, it is largely restricted to areas where Leishmania donovani is the causative parasite. The interval at which PKDL follows VL is 0–6 months in Sudan and 2–3 years in India. PKDL probably has an important role in interepidemic periods of VL, acting as a reservoir for parasites. There is increasing evidence that the pathogenesis is largely immunologically mediated; high concentrations of interleukin 10 in the peripheral blood of VL patients predict the development of PKDL. During VL, interferon ␥ is not produced by peripheral blood mononuclear cells (PBMC). After treatment of VL, PBMC start producing interferon ␥, which coincides with the appearance of PKDL lesions due to interferon-␥-producing cells causing skin inflammation as a reaction to persisting parasites in the skin. Diagnosis is mainly clinical, but parasites can be seen by microscopy in smears with limited sensitivity. PCR and monoclonal antibodies may detect parasites in more than 80% of cases. Serological tests and the leishmanin skin test are of limited value. Treatment is always needed in Indian PKDL; in Sudan most cases will self cure but severe and chronic cases are treated. Sodium stibogluconate is given at 20 mg/kg for 2 months in Sudan and for 4 months in India. Liposomal amphotericine B seems effective; newer compounds such as miltefosine that can be administered orally or topically are of major potential interest. Although research has brought many new insights in pathogenesis and management of PKDL, several issues in particular in relation to control remain unsolved and deserve urgent attention. Figure 1. PKDL papular lesions. and transmission may be anthroponotic as well as zoonotic, with rodents and canines as candidate reservoirs. Other VL endemic areas include countries in the Mediterranean basin, where Leishmania infantum is the species involved, and the New World, where the identical Leishmania chagasi circulates; in both areas canines are the reservoir hosts. Key features in the clinical presentation of VL are prolonged fever, hepatosplenomegaly, and weight loss. Dependent on the geographical region, in L donovani endemic areas between 5% and 60% of patients develop a dermatosis called post-kalaazar dermal leishmaniasis (PKDL) during or after treatment. This skin condition has a tendency to become chronic and is characterised by macular, papular, or nodular lesions in which leishmania parasites may be seen. PKDL is therefore considered a reservoir for leishmania parasites, especially during interepidemic periods of VL. Although the condition has been described for about 80 years, only recently has its relevance, in particular in Africa, been fully recognised. Recent studies have provided new insights into the pathogenesis of this condition that has important clinical and epidemiological implications. Lancet Infect Dis 2003; 3: 87–98 Clinical features Post-kala-azar dermal leishmaniasis (PKDL) is a complication of visceral leishmaniasis (VL). VL, also known as kala-azar, is caused by species of Leishmania that are transmitted by the bite of a female sandfly, and it is estimated that 200 million people are at risk for the yearly 500 000 cases.1 Most cases occur on the Indian subcontinent (India, Nepal, Bangladesh) and east Africa (Sudan, Ethiopia, Kenya), where Leishmania donovani is the causative parasite. Whereas VL is considered to be anthroponotic in India with people as the only known reservoir, in other areas the picture is less clear THE LANCET Infectious Diseases Vol 3 February 2003 The clinical features have been best described in reports from Sudan and India and are summarised in table 1. EEZ is at the Department of Medicine, College of Medicine, Malawi; AMM, EAGK, and AMEH are at the Department of Immunology and Clinical Pathology, and IMEH is at the Department of Parasitology and Medical Entomology, Institute of Endemic Diseases, Khartoum, Sudan. Correspondence: Dr EE Zijlstra, Department of Medicine, College of Medicine, Private Bag 360, Chichiri, Blantyre 3, Malawi. Tel +265 1 670202 or +265 8 844318; fax +265 1 673933 or 674700; email eezijlstra@malawi.net http://infection.thelancet.com 87 For personal use. Only reproduce with permission from The Lancet Publishing Group. Review Post-kala-azar dermal leishmaniasis Table 1. Comparison of most important features of Sudanese and Indian PKDL Clinical PKDL may occur in absence of previous VL PKDL may occur while still on treatment for VL PKDL may occur with evidence of visceralized disease (VL) Most commonly described predilection over body Type of rash Lymphadenopathy More severe disease found in Concomitant other post-kala-azar manifestations Sudan India Yes Yes Yes Face>trunk>arms>legs PN>MP>microP>M* Frequent Young children, short interval after VL Yes Yes No Yes Face>trunk>arms>legs Erythema, induration M,P,N Rare Not known Yes Epidemiology Frequency of PKDL following VL Highest prevalence reported in field study Interval between VL and PKDL most common range Age distribution 50–60% 4·8/100 5–10% 4·8/1000 0–6 months 0–13 months† Children, mean age 6 years 2–3 years 6 months–32 years Young adults Diagnosis Parasites may be demonstrated in smears Parasites may be demonstrated by PCR LST positive Most difficult differential diagnosis 20–30% 83% 16–65% Leprosy 20–40% 94% 0–67% Leprosy Treatment Treatment with stibogluconate Spontaneous cure 2–3 months The rule 4 months Not reported, all cases are treated *PN=papulo-nodular, MP=maculo-papular, MicroP=micropapular, M=macular, N=nodular. †In neighbouring Kenya, an interval of 30 years has been reported. Type of rash Distribution The rash has been best described from Sudan and India. In a recent cross-sectional descriptive study of 105 patients with PKDL in eastern Sudan, a papular or nodular rash was most frequently seen (51%) (figures 1 and 2); other types of rash were maculopapular (23%), micropapular (measles-like) (17%) (figure 3) and macular (9%) (figure 4).2 In a longitudinal study in the same area in which VL patients were monitored after treatment, the rash was almost always a mixture of a measles-like and maculopapular eruption.3 From India, three main presentations have been described, of which one or two may predominate: erythema and induration on the butterfly area of the face that shows photosensitivity; multiple symmetrical hypopigmented macules that may coalesce; and combinations of papules, nodules, and plaques.4–6 Other unusual manifestations include the annular, warty, papillomatous, fibroid, or xanthomatous types. From China and Nepal the same hypopigmented macular, maculopapular, and nodular types of rash or combinations have been described.7,8 Although there are differences in description of clinical findings from Sudan and other areas, in most reports macular (maculo) papular and nodular lesions are described as the hallmarks of PKDL, and are differentiated arbitrarily by size. Nodular lesions probably develop from papules over time, which seems also to be the case in PKDL reported from HIV-endemic areas. The erythema in the butterfly area is probably not appreciated in the darker African skin. Ulceration is not a feature of Indian or Sudanese PKDL. In Sudanese patients PKDL lesions typically appear around the mouth and spread to other parts of the face (figure 5); subsequent spread to upper arms and chest may follow. This pattern is most consistent in papular and nodular PKDL, but to a lesser extent in macular lesions as they may be more widely spread over the body. After treatment or during spontaneous regression, the lesions around the mouth remain longest and are the first to recur in case of relapse.9 In most severe cases the whole body may be affected sometimes with mucosal lesions on the lips or palate.9 Exposure to ultraviolet light may be an important factor in its pathogenesis and initially the distribution of the rash may mirror clothing habits.3 In some patients, the PKDL lesions occur preferentially in scars that become more prominent and regress again after treatment of PKDL (Köbner phenomenon).2 The typical pattern of distribution has resulted in the description of three clinical grades of severity. In grade one a scattered maculopapular or nodular rash occurs mainly in the face with or without some lesions on the upper chest and arms. Grade two is defined as a dense maculopapular or nodular rash covering most of the face and extending to the chest, back, upper arms, and legs, gradually becoming less distally, with only scattered lesions on the forearms and legs. Grade three is defined as a maculopapular or nodular rash covering most parts of the body, including hands and feet. In grade three crusting, ulceration, sloughing, scaling, and spreading to the mucosa of the lip (cheilitis) and the palate may occur (figure 6).2 88 THE LANCET Infectious Diseases Vol 3 February 2003 http://infection.thelancet.com For personal use. Only reproduce with permission from The Lancet Publishing Group. Review Post-kala-azar dermal leishmaniasis Figure 4. PKDL macular lesions. Other clinical findings Figure 2. PKDL nodular lesions that are confluent with plaque formation. The distribution and sequence of spread from the face to other parts of the body has not been described in Indian PKDL in longitudinal studies. However, especially for nodular lesions, similar predilection of the face with less frequent involvement of the arms and trunk further decreasing distally suggests a similar distribution pattern.5,10 As in Sudan, macular lesions do not seem to involve the face as frequently and may be more prominent on the trunk and limbs. In some reports it has also been suggested that the macular patches and the erythema precede the appearance of nodules; the influence of ultraviolet light has also been suggested.4,5,6 Patients are generally well and do not have fever.11 In the endemic area in eastern Sudan, where people are very much aware of the disease, which they call “kala-azar”, and where many have lost children because of this disease, the occurrence of a PKDL rash after treatment is regarded as a favourable sign (“the disease has come out”) meaning that the child is going to survive (EE Zijlstra and AM El Hassan, personal observation). In patients who develop PKDL early after VL there is a gradual transition from VL—during which the patient is ill, but without a skin rash, with demonstrable parasites in lymph node, bone marrow, or spleen—to PKDL—where the patient is well, the rash has developed, and parasites are no longer demonstrable except in the skin. Some patients present in the intermediate stage; one study showed that 15% of 416 PKDL patients had evidence of disseminated disease as shown by parasites in lymph node or bone-marrow aspirates. These patients with PKDL and concomitant VL may be more appropriately referred to as para-kala-azar dermal leishmaniasis cases and they take an intermediate position between VL and PKDL; they have larger spleen size (mean 4·1 cm [SD 3·6]) compared with PKDL patients who did not have demonstrable parasites in the bone marrow and lymph nodes (mean 2·8 cm [3·2]) but smaller than VL patients (mean 9·9 cm [5·4]).2 At least seven patients with similar concomitant VL and PKDL have been reported from India;12 another five patients with VL and presumably PKDL were seen in Iran,13 and one patient in France.14 Figure 7 shows the hypothetical association between the decreasing parasite load, the emerging immunological response, and the ensuing change in clinical syndrome. Lymphadenopathy may be seen in Sudanese VL and PKDL but it is rare in Indian VL and PKDL.15 Associated conditions Figure 3. PKDL micropapular lesions (measles-like). THE LANCET Infectious Diseases Vol 3 February 2003 PKDL may coexist with VL (see above); in Sudan, other post-kala-azar manifestations such as post-kala-azar mucosal leishmaniasis, uveitis, coniunctivitis, and blepharitis may be seen simultaneously in the same patient.16–18 Other complications after VL include post-kalaazar laryngitis and colitis.19 Similarly, in studies from India, http://infection.thelancet.com 89 For personal use. Only reproduce with permission from The Lancet Publishing Group. Review Post-kala-azar dermal leishmaniasis Age and sex distribution The age distribution of PKDL follows that of VL; in Sudan, the mean age of patients with PKDL and VL was identical (6 years) with equal numbers among boys and girls. Younger children have more severe PKDL;2,3,11 in India and Nepal, most cases are young adults with male predominance (1·3–3/1).4,8,23 There are, however, discrepancies between hospitalbased and community-based studies in India suggesting under-reporting of PKDL cases in children (age group 0–9 years) and females.23 Interval between VL and PKDL Figure 5. PKDL patient with typical perioral lesions. involvement of the mucosal surfaces, in particular in the mouth and the larynx, as well as eye involvement such as keratitis has also been described.4,20 Place of PKDL in the spectrum of manifestations of L donovani. Figure 8 shows the inter-relationship between PKDL and other clinical manifestations of leishmania infection in Sudan. PKDL may occur without a previous history of VL, or may come after or concomitant with a leishmanioma (figure 9). After PKDL subsides, immunity is the rule; in rare instances a patient with PKDL may have a relapse of VL.21 One study from India estimated this to occur in one of every 700 patients with PKDL. Reinfection, reinvasion from the skin, and renewed multiplication of latent parasites from the viscera have been suggested.21 A degree of immunosuppression induced by intercurrent diseases such as measles, malaria, and tuberculosis has been suggested by Nandy et al, 199822 as an explanation for relapse of VL in PKDL patients. In east Africa (Sudan, Kenya) the interval between VL and PKDL is short; all cases present 0–13 months after treatment of VL with most presenting within the first 6 months.2,11,24 Longer intervals of 3–30 years have also been reported.25 8% of cases have no previous history of VL;9 others present with concomitant VL and PKDL or may develop PKDL while still on VL treatment (up to 18%).2 A different situation exists in Asia (India, Nepal), where PKDL follows VL with an interval of 6 months to 6 years and most patients present after 2–3 years.4,10 In 15–20% of cases there is no previous history of VL.4 Longer intervals of up to 32 years have been reported.26 Epidemiology PKDL occurs mainly in L donovani-endemic areas and most studies reported are from Asia (mainly India) and east Africa, mainly Kenya and Sudan (table 1). Asia In India, PKDL follows VL in 5–10% of cases.4 The first cases of PKDL were reported from India by Brachmachari27 in 1922, who described the occurrence of eruptions and plaques in the skin containing leishmania parasites in patients previously treated for VL. After decades of intense transmission in India, mainly in Bihar, the incidence of both VL and PKDL declined in the early 1960s, probably the result of residual DDT spraying in the National Malaria Eradication Programme, which affected sandflies as well as malaria mosquitos. After discontinuation of insecticide treatment, a severe epidemic of VL occurred in Bihar and west-Bengal.28,29 In the interepidemic period, the number of PKDL cases outnumbered the VL cases, suggesting that the PKDL patients may have served as a reservoir. 10,28 Entomological work showed that Phlebotomus argentipes, the proposed vectors of VL in India, become infected and develop promastigotes in their midgut when allowed to feed on PKDL patients29,30 Figure 6. The distribution of PKDL may be described in grades of severity (see text for further explanation). and, therefore, seem capable of 90 THE LANCET Infectious Diseases Vol 3 February 2003 http://infection.thelancet.com For personal use. Only reproduce with permission from The Lancet Publishing Group. Review Post-kala-azar dermal leishmaniasis transmitting leishmania. This finding is in agreement with earlier reports that suggest that transmission of VL is anthroponotic in this area and no other vertebrate host than people has been found.31 The importance of PKDL in the epidemiology of leishmaniasis in India was also shown by Dye and Wolpert,32 who showed that the presence of as few as 0·5% durably infectious (PKDL) patients during an epidemic may cause VL to become endemic. Although there are few studies on the epidemiology of PKDL in India, one study reported a prevalence rate of 4·8/1000 in two villages in Varanasi district.33 Other countries in the Indian subcontinent from where PKDL has been reported include Bangladesh and Nepal; both are within the same nosogeographical area of L donovani.8,34 In China, VL occurs in the anthroponotic form (probably caused by L donovani) in the eastern plains for which effective control measures have been implemented; after 1960 only sporadic cases of VL and PKDL have been reported from that area.35–37 There are no reports on PKDL from the zoo-anthroponotic form of VL that occurs in central and north-west China, from where L infantum has been isolated from patients and dogs.35 From Taiwan both imported cases from mainland China and autochtonous patients have been reported.38,39 East Africa In Sudan, reports on PKDL remained scanty after the first report in 1938 by Kirk and Drew.40 Kirk and Sati41 described cutaneous infection in 57·5% of VL cases, but no clear distinction was made between PKDL and leishmaniomas. Sporadic cases were reported from two hospital-based studies.42,43 Similar to what was suggested from India, PKDL cases may have served as a reservoir for parasites at times when numbers of VL cases are low.44 In the 1990s a severe VL outbreak occurred in the endemic area in eastern Sudan with incidence rates in one village of 20·4–38·4/1000 person-years.45 56% of VL cases developed PKDL with similar age and sex distribution as VL cases.9,45 Between 1991 and 1996 416 patients with PKDL were seen in a cluster of villages in the same area.2 As in most patients, the skin lesions will persist for months and in some for years and a considerable number of people in any village can be expected to have PKDL. Indeed, in another study the point prevalence of PKDL in one village in the Parasites Immunity VL PKDL PKDL/VL Treatment Figure 7. The hypothetical relationship between decreasing parasite load, increasing immunity, and ensuing clinical presentation during treatment of visceral leishmaniasis. THE LANCET Infectious Diseases Vol 3 February 2003 Subclinical infection Death Infection VL Relapse Cure/ immunity PKDL Leishmanioma Figure 8. Diagram showing the interrelationship of clinical manifestations that may follow after leishmania infection in Sudan; the thickness of the lines corresponds with the likelihood of the occurrence of the following event. With permission from: Eldryd Parry, Richard Godfrey, Geoffrey Gill, David Mabey, eds. Principles of Medicine in Africa. Cambridge: Cambridge University Press; in press. area was 4·8%.46 A recent study on the natural history of PKDL in the same endemic area showed that the mean duration of PKDL was 9·7 months (range 2–28 months) before clearing.3 It seems likely that PKDL could have a role in transmission in Sudan, but so far this has not been proved; it is not clear whether transmission of VL in Sudan is anthroponotic or zoonotic, or both. L donovani has been shown by culture and PCR in materials from patients with VL, PKDL, and from rats and dogs; cultures that were further analysed by enzyme electrophoresis showed that all zymodemes of L donovani found in PKDL patients have also been seen in patients with VL and in dogs.47,48 In Ethiopia, where the endemic area of eastern Sudan extends into the Metema-Humera focus, a recent study showed a PKDL rate of 14% in patients who were seen only once at 6 months after treatment. By contrast with the Sudanese focus HIV infection is spreading in this area. PKDL was more common in HIV-positive patients (27·3% and 13·3% in HIV-positive and HIV-negative patients, respectively).49 In Kenya, PKDL was first described by Manson-Bahr50 in 1959. Reported PKDL rates after VL show considerable variability in four studies of 0·05%,51 1%50, 6%,24 and 30%.52 From L infantum-endemic areas reports on PKDL are scarce. Along the Mediterranean basin, sporadic cases occurred in Spain, Italy, France, and Israel; cases have also been reported from Iran.53–65 In one study in Spain up to 15% of patients developed cutaneous lesions after VL, which could be referred to as PKDL.54 In another report a patient had a PKDL rash for 15 years caused by a not previously described zymodeme of L infantum (MON-253) without obvious immunosuppression or HIV infection.61 More recent reports are from HIV-VL coinfected patients, in whom the clinical presentation of VL can be unusual and cutaneous lesions have been described to precede, accompany, and follow VL after treatment. A case of a dermatofibroma parasitised by leishmania parasites has been described; another patient developed Kaposi’s sarcoma-like lesions in the course of treatment of a third relapse of VL. It is unclear whether these are non-typical forms of PKDL or parasitised skin lesions in the course of VL.62,64 http://infection.thelancet.com 91 For personal use. Only reproduce with permission from The Lancet Publishing Group. Review Post-kala-azar dermal leishmaniasis experiments.22 Other studies suggest that different strains are involved in VL and PKDL; differences in antibody responses with variable patterns of reactivity by VL and PKDL serum samples have been shown.78 One study showed the cloning of a kinetoplast DNA mini-fragment from leishmania strains that was specific for strains from PKDL patients but not for strains from VL.79 Bozza et al80 provided support for this and showed that a strain from a PKDL patient had close relationship with L tropica and not with L donovani or L infantum. Pathogenesis and immunology Figure 9. PKDL in combination with leishmanioma. The girl did not have a history of VL. One HIV-infected patient from Italy developed PKDL after successful treatment of VL followed by highly active antiretroviral therapy (HAART); L infantum was identified in the skin by PCR58 A similar patient was described from Israel.59 PKDL has also been reported in patients with other forms of immunosuppression—eg, kidney transplant recipients or patients with Hodgkin’s lymphoma.14,57,65 In South America, where VL is mainly caused by L chagasi, PKDL seems uncommon; in one HIV-infected patient with VL caused by disseminated Leishmania braziliensis (a parasite that normally is restricted to the skin or mucous membranes) cutaneous lesions developed while relapsing from a previous episode of VL.66 In another case PKDL developed after VL caused by Leishmania amazonensis.67 Parasites Several studies from Sudan showed that in cultures from bone marrow or lymph node aspirates from patients with VL analysed by isoenzyme electrophoresis, L donovani, L infantum, and Leishmania archibaldi, which takes an intermediate position in the cladogram, are seen.68–70 However, the three species were all seen to be L donovani sensulato by Southern blotting and fingerprinting and were clearly different from a L infantum reference strain,69 which supports association between VL and PKDL in this area with L donovani rather than with L infantum. Of the three species, seven zymodemes have been seen to be circulating in the VL endemic area in eastern Sudan. Four zymodemes have now been isolated from PKDL patients: (MON-18 L donovani; MON-30 and MON-267 L infantum; and MON-82 L archibaldi),71 which does not suggest an association between a particular parasite subspecies and risk of developing PKDL. Similarly, although polymorphism was shown among L donovani strains using PCR single-strand conformation polymorphism, no correlation was found with the clinical manifestations of VL and PKDL.72 In India, Bangladesh, and Nepal strains that were isolated from VL and PKDL patients were all typed by electrophoresis as zymodeme LON-41 or MON-2;73–76 earlier typing found serologic similarity between strains,77 and the strains isolated from PKDL could cause VL in animal 92 The exact mechanisms underlying the development of PKDL still remain to be elucidated. There is accumulating evidence, however, that (developing) immune responses have a major role. In VL a specific cell-mediated immune (CMI) response to the leishmania parasite is absent, and only develops after treatment. This can be measured in vitro in experiments in which peripheral blood mononuclear cells (PBMC) are stimulated or in vivo by the leishmanin skin test. Early studies from India showed CMI responses during VL while two-thirds of patients with PKDL had a positive response after stimulation of PBMC, with a more marked CMI response in newly acquired PKDL (duration a few months to 1 year) compared with chronic PKDL (duration 8–30 years).81 In another study, all ten PKDL patients showed absent specific CMI response before treatment with intact response to (phytohaemagglutinin) mitogen, whereas in VL patients CMI responses both specific and generalised were absent. After treatment with antimony, specific CMI responses were restored in both VL and PKDL patients but this response was slower and took a larger amount of drugs in PKDL.82 More recent studies showed that during VL the immune response of PBMC to leishmania antigens is absent or skewed towards a Th2-type of response.83 After treatment the response changes from a Th2 to Th1 or a mixed Th1/Th2 type.83 The PBMCs of all Sudanese PKDL patients proliferate in response to leishmania antigen and produce interferon ␥; in about 20% of patients the cells also produced interleukin 10.84 In a further longitudinal study in Sudan, Gasim et al85 monitored 29 VL patients for 6–24 months from the time of diagnosis to the time of cure or development of PKDL. In both groups parasites were detected in the skin during VL, PBMC did not show proliferation to leishmania antigen, and interferon-␥ production was absent. Patients with VL who went on to develop PKDL had higher interleukin 10 concentrations in the skin and peripheral blood than those who did not develop PKDL. At follow-up, both groups did not differ in the acquisition of a degree of immunity as evidenced by proliferation of PBMC to leishmanial antigen as well as production of interferon ␥ and interleukin 10. However, it was noted that at day 30 after VL treatment, those who developed PKDL had higher PBMC immunological responses compared with those who developed PKDL later, suggesting that there is an association between the occurrence of the PKDL rash and the THE LANCET Infectious Diseases Vol 3 February 2003 http://infection.thelancet.com For personal use. Only reproduce with permission from The Lancet Publishing Group. Review Post-kala-azar dermal leishmaniasis appearance of leishmania-specific lymphocyte reactivity. The development of PKDL apparently depends on capacity to mount an immune response since in patients co-infected with HIV and VL, PKDL may develop after the start of HAART suggesting that PKDL develops in the context of immune reconstitution. This finding was also supported by an increased CD4 count, restored interferon-␥ production, and decreased interleukin-10 concentrations during PKDL compared with measurements during the VL episode.58,59 There seem to be discrepancies in the immunohistopathological findings in PKDL lesions from Indian and Sudanese patients. In the early stages of Indian PKDL limited numbers of CD4+ and CD8+ lymphocytes were seen in hypopigmented lesions; as the disease progresses to the nodular type, in dermal lesions as well as in lymph nodes a preponderance of CD8+ cells was found.86 By contrast in all lesions of Sudanese patients with PKDL most cells were CD3+ T cells with preponderance of CD4+ over CD8+ cells.84 Macrophages were seen in variable numbers. Natural killer cells were scarce. Interleukin 10 was the most prominent cytokine in the lesions. However, interferon ␥ was seen in all and interleukin 4 in most lesions. It was suggested that balance of the cytokines in the lesions may determine the outcome of the disease. The leishmanin skin test (LST) measures delayed type hypersensitivity in the skin by injection of killed leishmania amastigotes intradermally into the forearm. After 48–72 h the induration is measured in mm; a reaction 5 mm or more is usually considered positive. The LST is typically negative in VL and positive in 80% of successfully treated patients after 6 months.87 PKDL patients take an intermediate position: those with PKDL and concomitant VL are LST positive in 11% and those without concomitant VL in 37%.2 Pathology Irrespective of the clinical forms the epidermis shows several changes in different combinations. These include hyperkeratosis, parakeratosis, focal acanthosis, or atrophy of the rete pegs, and liquifaction degeneration of the basal cells.88 The last is associated with focal infiltration of the basal layer by lymphocytes. Under electron microscopy the lymphocytes are in intimate contact with melanocytes and basal keratinocytes. The latter cells seem to be damaged by the infiltrating lymphocytes, which is the major cause for the depigmentation seen clinically. The dermis is infiltrated by a mixture of lymphocytes and macrophages. By contrast with lesions of cutaneous leishmaniasis caused by Leishmania major (oriental sore) plasma cells are virtually absent in PKDL lesions. Lymphocytes are the dominant cells in most biopsies. In about half the cases epithelioid cells, scattered individually or forming compact granulomas, are seen. Compact granulomas are seen more frequently in nodular than macular and papular lesions.88 A neuritis involving small cutaneous nerves in PKDL lesions has been shown that may cause confusion in the differentiation from leprosy.89 The reported presence of parasites in biopsies varies, probably because of differences in type of rash and duration of lesions. In one Sudanese study all 15 biopsies showed parasites.11 In another comparative study, parasites could be THE LANCET Infectious Diseases Vol 3 February 2003 seen in only 17% of haematoxylin and eosin-stained sections; using a monoclonal antibody specific for L donovani leishmania parasites improved sensitivity to 88% of the biopsies.90 In India parasites are reported to be seen in about 90% of biopsies.91,92 Parasites are more easily shown in nodules than in papules and macular lesions. Predictors of PKDL No convincing clinical predictors have been identified that are helpful to predict who will develop PKDL and who will not. One Sudanese study showed spleen size at time of VL to be correlated with development of PKDL,85 but another study did not confirm this.2 In a further study from Sudan it was suggested that inadequate treatment regimens may be important.2 This possibility was also suggested from India where all patients presenting with PKDL had short duration of treatment for VL.3 It is unclear whether more effective treatment of VL—eg, with liposomal amphotericin would prevent PKDL; one limited comparative study in Sudan between stibogluconate and liposomal amphotericin B (AmBisome) in the treatment of VL, showed less PKDL in the amphotericin B group than the stibogluconate group.93 Several studies on Sudanese patients examined factors relating to the parasite and various immune responses.2,3,85,94,95 Persistence of parasites after successful treatment of VL may play a part. Osman et al94 showed that leishmanial parasite DNA was present by PCR in microscopically negative inguinal lymph nodes taken after treatment of VL in 82% of cases. Interestingly, 36% of these developed PKDL and 23% developed relapse of VL, whereas none of the 18% who were PCR negative developed these complications. Gasim et al95 showed that patients who had high C-reactive protein levels (>40 µg/mL) at the time of VL had higher risk of developing PKDL compared with those with C-reactive protein below 30 µg/mL. In another study high concentrations of interleukin 10 in the blood, and the presence of this cytokine in normal-looking skin during VL, predicted the subsequent development of PKDL.85 PKDL is more severe at a younger age and the conversion rates in the LST are lower in more severe PKDL. This may be the result of generalised ultraviolet light exposure since very young children often walk about undressed; it may also indicate the immaturity of the immune system in the very young.2,3 Another factor associated with the severity of PKDL is the interval between end of VL treatment and occurrence of PKDL; significantly more severe PKDL occurs after a shorter interval, suggesting a continuing Th2 response.3 Diagnosis In most endemic areas diagnosis will be made clinically by a history of previous VL, the temporal association with VL, the distribution and appearance of the lesions, by ruling out other disorders, and by the response to treatment. Parasitological confirmation may be sought if in doubt. Studies from India showed that smears are more likely to show amastigotes if taken from a larger lesion or from nodular (67–100%) lesions compared with papular http://infection.thelancet.com 93 For personal use. Only reproduce with permission from The Lancet Publishing Group. Review Post-kala-azar dermal leishmaniasis Table 2. Overview of studies on treatment of PKDL; only studies with more than one patient are included unless no other data on a particular drug are available. Reference Year Country Type of study Drug Daily dose Yesudian113 1974 India Case series Amphotericin B Max 25 mg Variable alternate days 2 1 cured after 717·5 mg in 3 months; 1 cured after 375 mg in 6 weeks Thakur28 1984 India Case series SSG or SAG 10 mg/kg adult 20 days 20 mg/kg child 20 days 20 7 (36%) cured; 13 (64%) relapsed Saha114 1985 India Case study 600 mg bd 2 courses of 15 days 1 1 cured Thakur115 1987 India Randomised SSG study 10 mg/kg 15 mg/kg 20 mg/kg 120 days 120 days 120 days 36 36 36 18 cured 24 cured 33 cured Rai33 1989 India Case series SAG 10 mg/kg 40 days 13 13 cured Thakur10 1990 India Case series SSG 20 mg/kg* 120 days 53 47 (88%) cured after 120 days 2 cured after 180 days 2 cured after 200 days† Muigai24 1991 Kenya Case series SSG Nil 20 mg/kg 30 days 7 5 6 cured, 1 relapsed, cured after SSG+allopurinol 5 self cured Ramesh116 1992 India Case series Ketoconazole 800 mg Variable 4 1 cured after treatment for 9 months; 2 stopped because of side-effects; 1 defaulted El-Hassan11 1992 Sudan Case series SSG 20 mg/kg 10 mg/kg 10 mg/kg 30 days 15 days 30 days 8 7 2 4 cured, 4 relapsed 2 cured, 5 improved with residual lesions 2 not improved Ramesh91 1993 India Case series SAG 20 mg/kg‡ 105-150 days 14 9 cured after 120–130 days 3 cured after 105 days 2 cured after 150 days Khalil117 1996 Sudan Case series Terbinafine + itraconazole 250 mg 200 mg 4 weeks 4 weeks 9 1 cured after 2 weeks extension 8 initially improved but relapsed later Ramesh118 1996 India Case series Allopurinol 20 mg/kg variable 3 2 cured after 20 and 24 months; 1 improved but stopped after 6 months (side-effects) Hashim119 1995 Sudan Case series Liposomal 2 2 cured 11 11 cured 11 7 cured Rifampin Ketoconazole Duration Number Outcome studied amphotericin B Thakur120 1997 India Comparative Amphotericin B 1 mg/kg study SAG 20 mg/kg* 3x20 days§ 6–10 x 20 days§ Garg8 2001 Nepal Case series SAG 20 mg/kg 30–72 days 21 21 cured *Maximum daily dose 850 mg. †Not mentioned for two patients. ‡Maximum daily dose 1g. §Courses of 20 days with 20 days drug free interval. SSG=sodium stibogluconate. SAG=sodium antimony gluconate. (36–69%) and macular lesions (7–33%).4,96 Cultures may give higher yield than smears but are likely to be contaminated. Monoclonal antibodies and PCR increase the diagnostic yield considerably to 88% and 83–94%, respectively, but these techniques are restricted to well equipped laboratories.90,97–99 Serological tests such as the direct agglutination test (DAT) and ELISA are of limited value in endemic areas and so far no specific and practical serological test exists for PKDL since after VL leishmanial antibodies may persist for years and a positive antibody test in a patient with suspected PKDL may be the result of previous VL. Similar titres in the DAT and rK39 IgG ELISA are reported in Indian and Bangladesh PKDL compared with VL.100,101 In Sudan the positivity rate in the rK39 strip test was equal among patients cured of VL without PKDL and patients with PKDL.102 94 Adaptations of the antigen used may be helpful. One study showed antibody profiles to differ in VL and PKDL since only 10% of PKDL patients showed an antibody response to a 200 kDa axenic amastigote antigen compared with 97% of patients with VL; in both conditions antibodies to leishmania amastigote soluble antigen could be shown.103 In another study from India, western blot analysis also showed different humoral responses in serum samples of PKDL patients compared with VL patients; two antigens (110 and 65 kDa) elicited an antibody response in 97–100% of PKDL patients, compared with 51–71% of VL patients and none of the control patients including patients with leprosy.104 Given the often long interval between VL and PKDL in India, it was argued that this immune response was likely to be the result of the occurrence of PKDL rather than persistence of antibody of VL. A competitive ELISA using L donovani-specific monoclonal THE LANCET Infectious Diseases Vol 3 February 2003 http://infection.thelancet.com For personal use. Only reproduce with permission from The Lancet Publishing Group. Review Post-kala-azar dermal leishmaniasis antibody was positive in all seven Indian PKDL patients tested and may prove useful to monitor the success of treatment.105 Serological tests such as DAT and rK39 strip test may help exclude other conditions, in particular leprosy;101,104 a rK39 strip test detected 91% of PKDL cases with 100% specificity in India.106 After VL, 80% of patients are expected to develop a positive LST after 6 months. For PKDL patients in Sudan, varying LST positivity rates of 16%, 32%, and 65% have been reported which is probably a function of time after VL, severity of PKDL, and antigen used.2,3,9 There is no difference in conversion rates between those who develop PKDL and those who do not.2 In India conversion rates vary between 0% and 67%.81,82,107 Both in Indian and Sudanese PKDL the differential diagnosis includes a large number of other skin conditions of which leprosy is not uncommonly mistaken for PKDL; distinguishing between these two conditions may be difficult.108–112 An overview of differential diagnoses has been published.2 Treatment There are few controlled studies on the management of PKDL and most data come from small case series. In addition, there are differences in approach according to geographical area. An overview of studies available is given in table 2. Spontaneous healing In Sudan, spontaneous healing frequently occurs; in one study in non-severe PKDL this was found in all 134 patients, of whom 84% healed within 1 year.3 Patients with grade-one and mild grade-two PKDL may therefore be left untreated under careful follow-up. Those who will develop persistent lesions may be identified by significantly higher titres in the DAT and a more often negative LST.3 therefore it is probably better used in lower dose as a second drug in combination with SAG.116 In Sudan, a combination of terbinafine and itraconazole was not effective.117 In India, rifampicin was shown to be effective in one case report; in another case series allopurinol cured two patients after 20–24 months, while a third patient improved but developed side-effects after 6 months.114,118 Amphotericin B was shown to be effective with low toxicity in antimony-unresponsive Indian VL patients and seen to be superior to pentamidine.122 Experience with amphotericin B in PKDL is limited;113,120 in one study in India, amphotericin B appeared more effective (all 11 patients cured after 120 days) compared with SAG (seven of 11 cured after up to 400 days of up to ten courses of treatment), although the result was not significant. Amphotericin B is, however, more expensive and showed some nephrotoxicity.113 Similarly, liposomal amphotericin B, which is probably the most effective drug for VL, could be of potential benefit in PKDL; however, so far this has been reported effective in three case reports only.56,119 The new oral compound miltefosine that has been shown effective in the treatment of Indian VL is of considerable potential interest for treatment of PKDL since patients who are otherwise not ill may resent a repeated and lengthy course of daily treatment with intramuscular or intravenous stibogluconate.123 Topical application may be an alternative because miltefosine has good skin penetration and is used as topical treatment for skin metastasis of breast cancer; it has been shown to be effective in experimental cutaneous leishmaniasis.124,125 Unresolved issues and areas for research in PKDL Pathogenesis and epidemiology ● Conclusive identification of risk factors for PKDL ● Differences between Leishmaniasis donovani and Leishmaniasis infantum with regard to development of PKDL: involvement of certain subspecies, immune responses after infection, parasite load, response to treatment, genetic factors ● Explanation of differences in interval between VL and PKDL between African and Asian PKDL; need for longitudinal studies ● Confirmation of PKDL as a reservoir in transmission of VL ● Identification of treatment regimens and drugs (or combinations of drugs) for VL to prevent development of PKDL Treatment with pentavalent antimony In Sudan, patients with severe PKDL, those with lesions that have persisted for more than 12 months, and those with concomitant anterior uveitis or mucosal lesions are best treated from the start; treatment is with sodium stibogluconate.3 No firm data are available on the optimal treatment regimen; clinical experience has shown that since 20 mg/kg per day for 30 days was not satisfactory, treatment may need to be prolonged to 2–3 months or alternative treatment may be necessary.2,11 In India treatment is the rule and cure rates are 64–92% with sodium antimony gluconate (SAG) 20 mg/kg per day for 120 days.4,111,115 Other treatment Especially in India, where resistance to SAG is a serious problem, alternative treatment options have been explored. There is one case of successful surgical removal of a localised lesion by shave excision.121 Pentamidine has been shown to be effective (cure rate 93%) but with serious toxicity.28 Ketoconazole in high dose (800 mg/day) has to be given for up to 9 months to achieve cure; THE LANCET Infectious Diseases Vol 3 February 2003 Management ● Firm clinical and/or laboratory markers of PKDL that predict rapid self-cure or need for treatment ● Identification of effective and non-toxic treatment regimens for PKDL ● Evaluation of miltefosine as a potential candidate for treatment (oral or topical administration) ● Clinical and/or laboratory markers that predict parasitological cure after treatment Control ● Use of bednets as a tool to prevent sandflies from feeding on PKDL patients ● Education and training of health workers to diagnose and manage patients with PKDL in areas of high incidence ● Integration of diagnosis and management of PKDL in control strategies in VL endemic areas http://infection.thelancet.com 95 For personal use. Only reproduce with permission from The Lancet Publishing Group. Review Post-kala-azar dermal leishmaniasis Search strategy and selection criteria The data in this review were from papers identified from PubMed searches using the terms, “post-kala-azar dermal leishmaniasis”, “post-kala-azar”, “dermal leishmaniasis”, “PKDL”, and “PKADL”. Additional data originated from papers in reference lists of reviewed articles and from the authors’ personal archives. Reference were selected for their scientific contribution to various aspects of PKDL. Case reports were used for areas in which not other studies were available. English and French papers were reviewed. and repeated demonstration of leishmanial DNA may not reflect the presence of live parasites. Parasitological cure may well precede clinical cure leading to unnecessary long treatment regimens with considerable toxicity. In addition, clinical response may differ according to PKDL type; nodules and papules disappear in 120 days, macules in 200 days.10 PKDL usually heals completely, but in longstanding cases scarring may occur (figure 10). Conclusion Figure 10. This girl had PKDL for 3 years. After the first course of sodium stibogluconate (top) some lesions are healing but with scar formation. Multiple depressed scars after a second course of treatment (bottom). Treatment studies in PKDL are hampered by the absence of a clear and early marker of cure. Since parasites may be scarce or absent in smears taken at diagnosis, parasitological cure is difficult to assess in repeated skin smears. Taking repeated biopsies is unattractive and impractical. Similarly, molecular techniques are not helpful References 1 2 3 4 5 6 7 8 9 96 Desjeux P. Leishmaniasis. Public health aspects and control. Clin Dermatol 1996; 14: 417–23. Zijlstra EE, Khalil EAG, Kager PA, El-Hassan AM. Post-kala-azar dermal leishmaniasis in the Sudan: clinical presentation and differential diagnosis. Br J Derm 2000; 142: 136–43. Musa AM, Khalil EAG, Raheem MA, et al. The natural history of Sudanese post- kala-azar dermal leishmaniasis: clinical, immunological and prognostic features. Ann Trop Med Parasitol (in press). Ramesh V, Mukherjee A. Post-kala-azar dermal leishmaniasis. Int J Derm 1995; 34: 85–91. Napier LE, Das Gupta CR. A clinical study of postkala-azar dermal leishmaniasis. Indian Med Gaz 1930; 65: 249–57. Girgla HS, Marsden RA, Singh GM, Ryan TJ. Postkala-azar dermal leishmaniasis. Br J Derm 1977; 97: 307–11. Leng Y. A review of kala-azar in China from 1949 to 1959. Trans R Soc Trop Med Hyg 1982; 76: 531–37. Garg VK, Agrawal S, Rani S, et al. Post-kala-azar dermal leishmaniasis in Nepal. Int J Derm 2001; 40: 179–84. Zijlstra EE, El-Hassan AM, Ismael A. Endemic kalaazar in eastern Sudan. Post-kala-azar dermal leishmaniasis. Am J Trop Med Hyg 1995; 52: 299–305. PKDL is now recognised as a frequent complication of VL in most endemic areas with important clinical and epidemiological implications. Although in the past decade studies have considerably increased our understanding of PKDL, many issues remain unresolved and should be the subject of further research (panel). These might include studies that increase our basic understanding in pathogenesis and management, but should also be focused on the public-health aspects of PKDL, especially in relation to developing control strategies in VL endemic areas. Acknowledgments This project was supported in part by UNDP/World Bank/WHO Spcial Program for Research and Training in Tropical Diseases (project #980947). Conflict of interest None declared. 10 Thakur CP, Kumar K. Efficacy of prolonged therapy with stibogluconate in post kala-azar dermal leishmaniasis. Indian J Med Res 1990; 91: 144–48. 11 El-Hassan AM, Ghalib HW, Zijlstra EE, et al. Post kala-azar dermal leishmaniasis in the Sudan: clinical features, pathology and treatment. Trans R Soc Trop Med Hyg 1992; 86: 245–48. 12 Singh NKP, Agrawal SK, Jha TK. Visceral leishmaniasis associated with post-kala-azar dermal leishmaniasis. J Assoc Phys Ind 1989; 37: 191. 13 Kumar PV, Sadeghi E, Torabi S. Kala-azar with disseminated dermal leishmaniasis. Am J Trop Med Hyg 1989; 40: 150–53. 14 Frances C, Merle Beral H, Franceschini P, LessanaLeibowitch M, Escande JP. Kala-azar chez l’immunodéprimé. A propos d’un cas révélé par des signes cutanés. Presse Med 1984; 13: 2433–36. 15 Sen Gupta PC, Mukherjee AM. Lymphadenopathy in Indian kala-azar and dermal leishmanoid. Indian J Pathol Bacteriol 1968; 11: 172–78. 16 El-Hassan AM, El-Sheikhi EA, Eltoum IA, et al. Post-kala-azar anterior uveitis: demonstration of Leishmania parasites in the lesion. Trans R Soc Trop Med Hyg 1991; 85: 471–73. 17 El Hassan AM, Khalil EAG, El Sheikh EA, et al. Postkala-azar ocular leishmaniasis. Trans R Soc Trop Med Hyg 1998; 92: 177–79. 18 El-Hassan AM, Meredith SEO, Yagi HI, et al. Sudan mucosal leishmaniasis: epidemiology, clinical features, diagnosis, immune responses and treatment. Trans R Soc Trop Med Hyg 1995; 89: 647–52. 19 Fahal AH, El Hag IA, El-Hassan AM, Hashim FA. Leishmanial cholecystitis and colitis in a patient with visceral leishmaniasis. Trans R Soc Trop Med Hyg 1995; 89: 284. 20 Nandy A, Addy M, Banerjee D, et al. Laryngeal involvement during post kala-azar dermal leishmaniaisis in India. Trop Med Int Health 1997; 2: 371–73. 21 Sen Gupta PC, Mukherjee AM. Recurrence of kalaazar associated with post- kala-azar dermal leishmaniasis. J Indian Med Assoc 1968; 50: 1–7. 22 Nandy A, Addy M, Maji AK, et al. Recurrence of kala-azar after PKDL: role of co-factors. Trop Med Int Health 1998; 3: 76–78. 23 Bora D. Epidemiology of visceral leishmaniasis in India. Nat Med J India 1999; 12: 62–68. 24 Muigai R, Gachihi GS, Oster CN, et al. Post kala-azar dermal leishmaniasis: the Kenyan experience. East Afr Med J 1991; 68: 801–06. 25 Rashid JR, Chunge CN, Oster CN, et al. Post-kalaazar dermal leishmaniasis occurring long after cure of visceral leishmaniasis in Kenya. East Afr Med J 1986; 63: 365–71. THE LANCET Infectious Diseases Vol 3 February 2003 http://infection.thelancet.com For personal use. Only reproduce with permission from The Lancet Publishing Group. Post-kala-azar dermal leishmaniasis 26 Munro DD. du Vivier A, Jopling WH. Post kalaazar dermal leishmaniasis. Br J Derm 1972; 87: 374–78. 27 Brahmachari UN. A new form of cutaneous leishmaniasis-dermal leishmanoid. Indian Med Gaz 1922; 57: 125. 28 Thakur CP. Epidemiological, clinical and therapeutic features of Bihar kala-azar (including post kala-azar dermal leishmaniasis). Trans R Soc Trop Med 1984; 78: 391–98. 29 Addy M, Nandy A. Ten years of kala-azar in west Bengal, part 1. Did post kala azar dermal leishmaniasis initiate outbreak 24-Parganas? Bull World Health Org 1992; 70: 341–46. 30 Dinesh DS, Kar SK, Kishore K, et al. Screening sandflies for natural infection with Leishmania donovani, using a non-radioactive probe based on the total DNA of the parasite. Ann Trop Med Parasitol 2000; 94: 447–51. 31 Ashford RW, Bettini S. Ecology and epidemiology: Old World. In: Peters W, Killick-Kendrick R, eds. The leishmaniasis in biology and medicine. London: Academic Press, 1987: 365–424. 32 Dye C, Wolpert DM. Earthquakes, influenza and cycles of Indian kala-azar. Trans R Soc Trop Med Hyg 1988; 82: 843–50. 33 Rai RN, Kumar M, Srivastava A, et al. Clinicoepidemiological profiles of post-kala-azar dermal leishmaniasis in Varanasi. J Com Dis 1989; 21: 214–17. 34 Harith AE, Chowdhury S, Al-Masum A, et al. Reactivity of various leishmanial antigens in a direct agglutination test and their value in differentiating post-kala azar dermal leishmaniasis from leprosy and other skin conditions. J Med Microbiol 1996; 44: 141–46. 35 Desjeux P. Information on the epidemiology and control of the leishmaniases by country or territory. WHO/Leish91.3 Geneva: WHO, 1991. 36 Zhong H-L, Zhang N-Z. Studies on leishmaniasis in China. Historical background, epidemiology, clinical aspects, legislature and control program. Chin Med J 1986; 99: 281–300. 37 Leng Y. A review of kala-azar in China from 19491959. Trans R Soc Trop Med Hyg 1982; 76: 531–37. 38 Cross JH, Gunning J-J, Drutz DJ, Lien JC. Introduced leishmaniasis on Taiwan. Southeast Asian J Trop Med Pub Health 1985; 16: 381–86. 39 Cross JH, Gunning J-J, Drutz DJ, Lien JC. Autochthonous cutaneous-subcutaneous leishmaniasis on Taiwan. Am J Trop Med Hyg 1985; 34: 254–56. 40 Kirk R, Drew CB. Preliminary notes on dermal leishmaniasis in the Anglo-Egyptian Sudan. Trans R Soc Trop Med Hyg 1938; 32: 265–70. 41 Kirk R, Sati MH. Studies in leishmaniasis in the Anglo-Egyptian Sudan. IV. A punctate rash in treated cases. Trans R Soc Trop Med Hyg 1940; 34: 213–16. 42 El-Hassan AM, Ahmed MAM, Abdul Rahim AG, et al. Visceral leishmaniasis in the Sudan: clinical and hematological features. Ann Saudi Med 1990; 10: 51–56. 43 Zijlstra EE, Siddig Ali M, El-Hassan AM, et al. Kala-azar in displaced people from southern Sudan: epidemiological, clinical and therapeutic findings. Trans R Soc Trop Med Hyg 1991; 85: 365–69. 44 El-Hassan AM, Ghalib HW, Zijlstra EE, et al. Post-kala-azar dermal leishmaniasis in the absence of active visceral leishmaniasis. Lancet 1990; 336: 750. 45 Khalil EAG, Zijlstra EE, Kager PA, El-Hassan AM. Epidemiology and clinical manifestations of L donovani infection in two villages in an endemic area in eastern Sudan. Trop Med Int Health 2002; 7: 35–44. 46 El-Hassan AM, Khalil EAG. Post-kala-azar dermal leishmaniasis: does it play a role in the transmission of Leishmania donovani in the Sudan? Trop Med Int Health 2001; 6: 743–44. 47 El-Hassan AM, Zijlstra EE, Meredith SEO, Ismail A. Identification of Leishmania donovani using a polymerase chain reaction in patient and animal material obtained from an area of endemic kala-azar in the Sudan. Acta Trop 1993; 55: 87–90. 48 Dereure J, Boni M, Pratlong F, et al. Visceral leishmaniasis in Sudan: first identifications of Leishmania from dogs. Trans R Soc Trop Med Hyg 2000; 94: 154–55. 49 Ritmeijer K, Veeken H, Melaku Y, et al. Ethiopian visceral leishmaniasis: generic and proprietary sodium stibogluconate are equivalent; HIV coinfected patients have a poor outcome. Trans R Soc Trop Med Hyg 2001; 95: 668–72. 50 Manson-Bahr PEC. East African kala-azar with special reference to the pathology, prophylaxis and treatment. Trans R Soc Trop Med Hyg 1959; 53: 123–36. 51 Southgate BA, Oriedo BVE. Studies in the epidemiology of East African leishmaniasis. J Trop Med Hyg 1967; 70: 1–4. 52 Cole ACE. Kala-azar in East Africa. Trans R Soc Trop Med Hyg 1944; 37: 409–30. 53 Baghestani S, Sodeifi M, Kumar PV. Post-kala-azar dermal leishmaniasis. Eur J Derm 1998; 8: 277–79. 54 Alvar J, Canavate C, Gutiérrez-Solar B, et al. Leishmania and human immunodeficiency virus coinfection: the first 10 years. Clin Microbiol Rev 1997; 10: 298–319. 55 Alarcon Cabrera G, Garcia Garcia B, Duaden Tello E, et al. Leishmaniasis cutanea postkala-azar en un paciente con immunodeficiencia adquirada. Actas Dermo Sif 1987; 78: 475–77. 56 Rios-Buceta L, Buezo GF, Penas PF, et al. Post-kalaazar dermal leishmaniasis in an HIV-patient. Int J Derm 1996; 35: 303–04. 57 Roustan G, Jiménez JA, Gutiérrez-Solar B, et al. Post-kala-azar dermal leishmaniasis with mucosal involvement in a kidney transplant recipient: treatment with liposomal amphotericin B. Br J Derm 1998; 138: 526–28. 58 Ridolfo AL, Gervasoni C, Antinori S, et al. Post-kalaazar dermal leishmaniasis during highly active antiretroviral therapy in an AIDS patient infected with Leishmania infantum. J Infect 2000; 40: 199–202. 59 Gilad J, Borer A, Hallel-Halevy D, et al. Post-kalaazar dermal leishmaniasis manifesting after initiation of highly active anti-retroviral therapy in a patient with human immunodeficiency virus infection. Isr Med Assoc J 2001; 3: 451–52. 60 Moral L. Post-kala-azar dermal leishmaniasis and Leishmania infantum. Br J Derm 1999; 140: 760. 61 Montelius S, Maasho K, Pratlong F, et al. Skin rash for 15 years. Lancet 1998; 352: 1438. 62 Rodríguez-Peralto CVM, Alonso S, Gómez-De n. -De la Fuente, Ibarrola C. Dermatofibroma parasitized by Leishmania in HIV infection: a new morphologic expression of dermal kala azar in an immunodepressed patient. J Cutan Pathol 1999; 26: 516–19. 63 Juan Martín F De, Roldán MLJ, Adana Pérez ES de, et al. Leishmaniasis dérmica postkala-azar. Esp Pediatr 1997; 46: 63–64. 64 González-Beato MJ, Moyano B, Sánchez C, et al. Kaposi’s sarcoma-like lesions and other nodules as cutaneous involvement in AIDS-related visceral leishmaniasis. Br J Derm 2000; 143: 1316–18. 65 Hernández-Pérez J, Yebra-Bango M, JiménezMartínez E, et al. Visceral leishmaniasis (Kala-azar) in solid organ transplantation: report of five cases and review. Clin Infect Dis 1999; 29: 918–21. 66 Hernández DE, Oliver M, Martínez C, Planas G. Visceral leishmaniasis with cutaneous and rectal dissemination due to Leishmania braziliensis in acquired immunodeficiency syndrome (AIDS). Int J Derm 1995; 34: 114–15. 67 Barral A, Pedral-Sampaio D, Grimaldi G Jr, et al. Leishmaniasis in Bahia, Brazil: evidence that Leishmania amazonensis produces a wide spectrum of clinical disease. Am J Trop Med Hyg 1991; 44: 536–46. 68 Ashford RW, Seaman J, Schorscher J, Pratlong F. Epidemic visceral leishmaniasis in southern Sudan: identity and systematic position of the parasites from patients and vectors. Trans R Soc Trop Med Hyg 1992; 86: 379–80. 69 Oskam L, Pratlong F, Zijlstra EE, et al. Biochemical and molecular characterization of Leishmania parasites isolated from an endemic focus in eastern Sudan. Trans R Soc Trop Med Hyg 1998; 92: 120–22. 70 Rioux JA, Lanotte G, Serres E, et al. Taxonomy of Leishmania. Use of isoenzymes. Suggestions for a new classification. Ann Parasitol Hum Comp 1990; 65: 111–25. 71 Pratlong F, Dereure J, Bucheton B, et al. Dedet JP. Sudan: the possible original focus of visceral leishmaniasis. Parasitol 2001; 122: 599–605. 72 El Tai NO, El Fari M, Mauricio I, et al. Leishmania donovani: intraspecific polymorphisms of Sudanese isolates revealed by PCR-based analyses and DNA sequencing. Exp Parasitol 2001; 97: 35–44. 73 El-Masum MA, Evans DA. Characterization of Leishmania isolated from patients with kala-azar and post-kala-azar dermal leishmaniasis in Bangladesh. Trans R Soc Trop Med Hyg 1995; 89: 331–32. 74 Le Blancq SM, Peters W. Leishmania in the Old World: 4. The distribution of L donovani sensu lato zymodemes. Trans R Soc Trop Med Hyg 1986; 80: 367–77. THE LANCET Infectious Diseases Vol 3 February 2003 http://infection.thelancet.com Review 75 Chatterjee M, Manna M, Bhaduri AN, Sarkar D. Recent kala-azar cases in India: isoenzyme profiles of Leishmania parasites. Indian J Med Res 1995; 102: 165–72. 76 Thakur CP, Dedet JP, Narain S, Pratlong F. Leishmania species, drug unresponsiveness and visceral leishmaniasis in Bihar, India. Trans R Soc Trop Med Hyg 2001; 95: 187–89. 77 Bray RS, Ashford RW, Mukherjee AM, Sen Gupta PC. Studies on the immunology and serology of leishmaniasis IX. Serological investigation of the parasites of Indian kala-azar and Indian post-kala-azar dermal leishmaniasis. Trans R Soc Trop Med Hyg 1973; 67: 125–29. 78 Ghosh AK, Das Gupta S, Ghose AC. Immunoglobulin G subclass-specific antileishmanial antibody responses in Indian kala-azar and postkala-azar dermal leishmaniasis. Clin Diag Lab Immunol 1995; 2: 291–96. 79 Das Gupta S, Ghosh DK, Majumdar HK. A cloned kinetoplast mini-circle fragment from a Leishmania spp. specific for post-kala-azar dermal leishmaniasis strains. Parasitology 1991; 102: 187–91. 80 Bozza M, Fernandes O, Degrave WM, Lopes UG. Characterization of Old World Leishmania species using amplified minicircle variable regions as molecular probes. Trans R Soc Trop Med Hyg 1995; 89: 333–34. 81 Haldar JP, Ghose S, Saha KC, et al. Cell-mediated immune response in Indian kala-azar and post-kala-azar dermal leishmaniasis. Infect Immun 1983; 42: 702–7. 82 Neogy AB, Nandy A, Ghosh Dastidar B, Chowdhury AB. Modulation of the cell-mediated immune response in kala-azar and post-kala-azar dermal leishmaniasis in relation to chemotherapy. Ann Trop Med Parasitol 1988; 82: 27–34. 83 Kemp M, Theander TG, Kharazmi A. The contrasting roles of CD4+ cells in intracellular infections in human: Leishmania as an example. Immunol Today 1996; 17: 13–17. 84 Ismail A, Hassan ELAM, Kemp K, et al. Immunopathology of post kala-azar dermal leishmaniasis (PKDL): T-cell phenotypes and cytokine profile. J Pathol 1999; 189: 615–22. 85 Gasim S, Elhassan AM, Khalil EAG, et al. High levels of plasma IL-10 and expression of IL-10 by keratinocytes during visceral leishmaniasis predict subsequent development of post-kala-azar dermal leishmaniasis. Clin Exp Immunol 1998; 111: 64–69. 86 Ghosh MK, Nandy A, Addy M, Maitra TK, Ghose AC. Subpopulations of T lymphocytes in the peripheral blood, dermal lesions and lymph nodes of post-kala-azar dermal leishmaniasis patients. Scand J Immunol 1995; 41: 11–17. 87 Zijlstra EE, El-Hassan AM. Leishmanin and tuberculin sensitivity in leishmaniasis in the Sudan, with special reference to kala-azar. Trans R Soc Trop Med Hyg 1993; 87: 425–27. 88 Ismail A. Immune responses and immunopathology of post kala-azar dermal leishmaniasis (PKDL) (PhD thesis). Copenhagen: University of Copenhagen; 1999. 89 El-Hassan AM, Ali MS, Zijlstra EE, et al. Post kalaazar dermal leishmaniasis in the sudan: peripheral neural involvement. Int J Derm 1992; 31: 400–3. 90 Ismael A, Kharazmi A, Permin H, El-Hassan AM. Detection and characterization of Leishmania in tissues of patients with post kala-azar dermal leishmaniasis using a specific monoclononal antibody. Trans R Soc Trop Med Hyg 1997; 91: 283–85. 91 Ramesh V, Misra RS, Saxena U, Mukherjee A. Postkala-azar dermal leishmaniasis: a clinical and therapeutical study. Int J Derm 1993; 32: 272–75. 92 Mukherjee A, Ramesh V, Misra RS. Post-kala-azar dermal leishmaniasis: a light and electron microscopic study of 18 cases. J Cutan Pathol 1993; 20: 320–25. 93 Khalil EAG, Hashim FA, Ghalib HW, et al. Comparative study of liposomal amphotericine B (AmBisome) and sodium stibogluconate in the treatment of kala-azar in the Sudan. East Afr Med J 1998; 75: 481–84. 94 Osman OF, Oskam L, Zijlstra EE, et al. Use of the polymerase chain reaction to assess the success of visceral leishmaniasis treatment. Trans R Soc Trop Med Hyg 1998; 92: 397–400. 95 Gasim S, Theander TG, Elhassan AM. High levels of C-reactive protein in the peripheral blood during visceral leishmaniasis predict subsequent development of post kala-azar dermal leishmaniasis. Acta Trop 2000; 75: 35–38. 96 Sharma MC, Gupta AK, Verma N, et al. Demonstration of Leishmania parasites in skin lesions of Indian post kala-azar dermal leishmaniasis 97 For personal use. Only reproduce with permission from The Lancet Publishing Group. Review (PKDL) cases. J Commun Dis 2000; 32: 67–8. 97 Meredith SEO, Zijlstra EE, Schoone GJ, et al. Development and application of the polymerase chain reaction for the detection and identification of Leishmania parasites in clinical material. Arch Inst Pasteur Tunis 1993; 70: 419–31. 98 Osman OF, Oskam L, Kroon NCM, et al. Use of PCR for diagnosis of post-kala-azar dermal leishmaniasis. J Clin Microbiol 1998; 36: 1621–24. 99 Salotra P, Sreenivas G, Pogue GP, et al. Development of a species-specific PCR assay for detection of Leishmania donovani in clinical samples from patients with kala-azar and post-kala-azar dermal leishmaniasis. J Clin Microbiol 2001; 39: 849–54. 100 Singh S, Kumari V, Singh N. Predicting kala-azar disease manifestations in asymptomatic patients with latent Leishmania donovani infection by detection of antibody against recombinant K39 antigen. Clin Diag Lab Immunol 2002; 9: 568–72. 101 Harith AE, Chowdhury S, Al-Masum A, et al. Reactivity of various leishmanial antigens in a direct agglutination test and their value in differentiating post-kala azar dermal leishmaniasis from leprosy and other skin conditions. J Med Microbiol 1996; 44: 141–46. 102 Zijlstra EE, Nur Y, Desjeux P, et al. Diagnosing visceral leishmaniasis with the recombinant k39 strip test: experience from the Sudan. Trop Med Int Health 2001; 6: 108–13. 103 Kual P, Malla N, Kaur S, Mahajan RC, Ganguly NK. Evaluation of a 200-kDa amastigote antigen of L. donovani by enzyme-linked immunosorbent assy (ELISA) for the diagnosis of visceral leishmaniasis. Trans R Soc Trop Med Hyg 2000; 94: 173–75. 104 Salotra P, Raina A, Ramesh V. Western blot analysis of humoral immune response to Leishmania donovani antigens in patients with post-kala-azar dermal leishmaniasis. Trans R Soc Trop Med Hyg 1999; 93: 98–101. 98 Post-kala-azar dermal leishmaniasis 105 Chatterjee M, Jaffe CL, Sundar S, et al. Diagnostic and prognostic potential of a competitive enzymelinked immunosorbent assy for leishmaniasis in India. Clin Diag Lab Immunol 1999; 6: 550–54. 106 Salotra P, Sreenivas G, Ramesh V, Sundar S. A simple and sensitive test for field diagnosis of post kala-azar dermal leishmaniasis. Br J Derm 2001; 145: 630–32. 107 Neogy AB, Nandy A, Chowdhury AB. Leishmanin test in post-kala-azar dermal leishmaniasis. Trans R Soc Trop Med Hyg 1990; 84: 58. 108 El-Hassan AM, Hashim FA, Abdalla M, Zijlstra EE, Ghalib HW. Distinguishing post-kala-azar dermal leishmaniasis from leprosy: experience in the Sudan. Lepr Rev 1993; 64: 53–59. 109 Lal S, Garg BR. Post-kala-azar dermal leishmaniasis simulating leprosy. Lepr India 1976; 48: 81–83. 110 Muir E. The differential diagnosis of leprosy and dermal leishmaniasis. Indian Med Gaz 1930; 65: 257–58. 111 Ramesh V, Saxena UM, Misra RS, Mukherjee A. Post-kala-azar dermal leishmaniasis: a case report strikingly resembling lepromatous leprosy. Lepr Rev 1991; 62: 217–21. 112 Ramesh V. On the differences between post-kalaazar dermal leishmaniasis and leprosy. Trop Doct 1994; 24: 120–1. 113 Yesuian P, Thambiah AS. Amphotericin B therapy in dermal leishmanoid. Arch Derm 1974; 109: 720–22. 114 Saha SK. Dermal leishmaniasis after kala-azar infection: successful treatment with rifampin. n: Cutis 1985; 36: 81–82. 115 Thakur CP, Kumar K, Sinha PK, Mishra BN, Pandey AK. Treatment of post-kala-azar dermal leishmaniasis with sodium stibogluconate. BMJ 1987; 295: 886–7. 116 Ramesh V, Saxena U, Misra RS. Efficacy of ketoconazole in post-kala-azar dermal leishmaniasis. Arch Derm 1992; 128: 411–12. 117 Khalil EAG, Nur NM, Zijlstra EE, El-Hassan AM, Davidson RN. Failure of a combination of two antifungal drugs, terbinafine and itraconazole, in Sudanese post kala-azar dermal leishmaniasis. Trans R Soc Trop Med Hyg 1996; 90: 187–88. 118 Ramesh V. Allopurinol therapy in post-kala-azar dermal leishmaniasis. Acta Derm Venereol 1996; 76: 328–29. 119 Hashim FA, Khalil EAG, Ismael A, El-Hassan AM. Apparently successful treatment of two cases of post kala-azar dermal leishmaniasis with liposomal amphotericine B. Trans R Soc Trop Med Hyg 1995; 89: 440. 120 Thakur CP, Narain S, Kumar N, et al. Amphotericin B is superior to sodium antimony gluconate in the treatment of Indian post-kala-azar dermal leishmaniasis. Ann Trop Med Parasitol 1997; 91: 611–6. 121 Ramesh V, Misra RS, Khunger N, et al. Shave excision as an adjunct to the therapy of a rhinophyma-like complication in post-kala-azar dermal leishmaniasis. Acta Derm Venereol 1999; 79: 330–31. 122 Mishra M, Biswas UK, Jha DN, Khan AB. Amphotericin versus pentamidine in antimony-unresponsive kala-azar. Lancet 1991; 337: 926. 123 Sundar S, Rosenkaimer F, Makharia MK, et al. Trial of oral miltefosine for visceral leishmaniasis. Lancet 1998; 352: 1821–23. 124 Murray HW. Treatment of visceral leishmaniasis (Kala-azar): A decade of progress and future approaches. Int J Infect Dis 2000; 4: 158–77. 125 Schmidt-Ott R, Klenner T, Overath P, Aebischer T. Topical treatment with R hexadecylphosphocholine (Miltex ) efficiently reduces parasite burden in experimental cutaneous leishmaniasis. Trans R Soc Trop Med Hyg 1999; 93: 85–90. THE LANCET Infectious Diseases Vol 3 February 2003 http://infection.thelancet.com For personal use. Only reproduce with permission from The Lancet Publishing Group.