The New Nation - Internet Edition

Visceral leishmaniasis, or kala-azar as it is commonly known, is a sandfly transmitted parasitic illness marked by prolonged fever, splenomegaly, anorexia, and wasting. Considered to be a major neglected tropical disease in South Asia, kala-azar has resurged in endemic regions of Bangladesh since the 1990s, with the highest rates in the districts of Mymensingh, Pabna, and Tangail (1). In Mymensingh specifically, the average annual incidence rate between 1994 and 2004 was 5.8/10,000, and currently is as high as 300/10,000 in the most affected communities (1,2).

Post kala-azar dermal leishmaniasis, or PKDL, was first clinically described in Bengal by Brahmachari in 1922 (3). Diverse dermal lesions - from hypopigmented pinpoint marks, to erythematous papules, nodules, and others - appear in individuals who are otherwise asymptomatic, usually months to years after the occurrence and apparent effective treatment of classic kala-azar (4). Affected persons, though clinically well, harbour the Leishmania parasite in their skin lesions; sandflies which take a bloodmeal may thus become infectious, making PKDL patients an important reservoir in anthroponotic leishmaniasis transmission.

Despite being long recognized clinically, research on the risk factors, epidemiology, and management of PKDL is quite limited, especially in South Asia. Anecdotally, the number of PKDL patients in certain kala-azar endemic pockets of Bangladesh is clearly on the rise, but actual statistics remain incomplete and are only just beginning to be officially collected. PKDL is essentially a clinical diagnosis, because dermatopathological sampling of patients is impractical, and there is a lack of definitive laboratory testing. Under current Bangladeshi national guidelines, patients with PKDL are treated with 120 intramuscular sodium antimony gluconate (SAG) injections, in 20/month cycles over 6 months.

An active surveillance study exploring kala-azar, kala-azar treatment failure/relapse, and PKDL cases was commenced in the most affected upazilla of the country, Fulbaria, Mymensingh, in June of this year (2007). A field team selected communities with the highest rates of disease based on recent governmental data and administered surveys to individuals at their homes. Patients - current and from the past 5 years - were identified by symptom and treatment history oriented questionnaires, appropriate examination by physician, and rk39 (rapid, fingerprick blood) testing in the field. The study aims include a complete survey of 3 villages and a target sample population of more than 20,000 respondents. The data obtained will be compared with governmental passive surveillance statistics to assess rates of under-reporting. In addition, risk factor, laboratory, and further clinical characterization of especially current patients is being undertaken.

By November of 2007, interviewers have surveyed approximately 8,400 respondents with the completion of the study's primary field site of Chouder Village. From this population, a total of 32 current and 11 recent (treated and resolved within the last 5 years; one death during treatment) PKDL patients have been identified, and preliminary review of this emerging cohort has revealed several notable findings. While the village point prevalence for PKDL is 3.8/1,000 residents, one para (neighborhood), Nodipar demonstrated a rate nearly twice this, at 7.3/1,000 (Table 1). Also, rates of PKDL amongst treated kala-azar patients have been higher than expected in the entire community at 13.9% and reached 16% in 4 of 9 paras (Table 2).

Closer exploration of individual cases reveal important patterns regarding timing of PKDL related to kala-azar and patients' response to treatment. Based on respondents dating in the detailed surveys, presentation of PKDL lesions occurred within 2 years after kala-azar treatment in the majority (24/40) of patients, with 20% of patients within 6 months, and 40% between 6-24 months (Table 3). Also, of the 10 patients who experienced resolution of PKDL after treatment and from whom complete information was obtainable, 7 actually received less than the standard 120 day SAG regimen, secondary to extreme pain/injection intolerance, early and complete disappearance of lesions, or contemporaneous death or severe illness of co-patients receiving SAG (Table 4). Of note, these patients demonstrated similar timeframes of resolution to the 3 patients who received the full 6 month treatment course, with their lesions generally beginning to regress within 3-5 months of initiating treatment, and completely disappearing within 6-9 months (Table 4).

Additionally, three of the youngest patients presented with symptoms consistent with treatment-relapse kala-azar along with their PKDL lesions. A brief clinical description of one may be most illustrative: A 3.5 year-old female, whose father had kala-azar at the time of her birth, was first diagnosed with kala-azar at age 6 months after ongoing fever and poor growth. She improved after 20 continuous days of SAG injections as an inpatient at the Upazilla Health Complex Hospital, but again at age 1.5 (1 year later) presented with prolonged fever and marked anorexia. At that point, physicians at governmental facilities (based apparently on both clinical judgment and positive aldehyde (AT) testing) again felt that she had active kala-azar disease, and she was treated with 60 injections of SAG, in 20 injection/month cycles over 3 months. Approximately 2 months after the completion of this second regimen (21 months prior to our initial evaluation), she broke out with hypopigmented papules diffusely over both legs, and again experienced abdominal pain and suppressed appetite. At the time of our field examination and diagnosis, she appeared stunted and very ill, had conspicuous hypopigmented papules - confluent in areas - densely spread over her legs, arms, and mainly the cheeks of her face; she continued to have abdominal fullness with mild splenomegaly, regular fevers and debilitating anorexia, and had a recent history consistent with repeat respiratory and skin infections.

Supported by: United States Agency for International Development and Centers for Disease Control, USA

Our growing experience with PKDL patients from Mymensingh challenges the existing understanding of the disease in Bangladesh, and suggests that PKDL is a more common and complex clinical phenomenon than currently assumed. South Asian PKDL is conventionally described as occurring in a limited proportion of treated kala-azar patients, who are otherwise asymptomatic, usually 2-3 years after classic kala-azar disease, and that spontaneous cure is never seen - making treatment mandatory (4). In con trast, initial evaluation of the study population in Fulbaria, Mymensingh reveals that a majority of patients' dermal lesions appeared within 2 years of resolution of KA, and in several individuals as early as within 6 months. Also, we have identified a few respondents whose PKDL lesions demonstrated rapid increase followed by steady improvement towards disappearance without any intervention, which raises the possibility that the disease may resolve on its own (without treatment). Considering the observations of the patients evaluated to date, PKDL appears to naturally evolve in distinct patterns and rates, varying significantly at the individual level.

In addition, a number of our PKDL patients described ongoing fevers, while at least three also had re-appearance of anorexia and abdominal fullness, all classic kala-azar symptoms. These observations indicate that the emergence of PKDL consistent lesions is not always independent of other visceral leishmaniasis features, and may occur in patients who had ineffective management of their original kala-azar and thus a persistence of systemic symptoms, often known as treatment failure or relapse cases.

In sum, the appearance of PKDL consistent skin lesions may be more closely associated with classic kala-azar than initially assumed in South Asia, representing a manifestation within a spectrum of Leishmania-host/ immunity interactions rather than a completely distinct clinical entity - a concept which has best been described in the framework of PKDL seen in Sudan (4).

These early observations also call into question the existing national PKDL treatment and disease control strategies. The current treatment regimen of 120 SAG injections is not only particularly lengthy and painful, but also risky: prolonged use of the drug increases the potential for serious cardiac side-effects as well as exposure to 'toxic' drug batches, which often appear on the market unexpectedly and may be discovered by the occurrence of multiple sudden fatalities within a short time (5). In fact, the unexplained and relatively sudden death of a 7-year old girl from Chouder in 2004, who was overall well and had completed 57 SAG injections in the midst of her PKDL treatment (as explained by several family members during a thorough verbal autopsy), raised strong concern for drug-related toxicity.

Additionally, this rigorous regimen is supported neither by comprehensive clinical trials in the literature, nor by the practices of other kala-azar endemic countries. For example, in neighbouring Bihar state in India where antimony-resistance is a major problem, SAG is rarely used for more than 3 months, and in East Africa, PKDL - when treated, is usually approached with short course, combination therapy, including more effective alternatives such as amphotericin or paromomycin. The experiences of our 7 patients who achieved resolution with shorter than standard treatment similarly suggest that the arduous 6 month SAG course is not always necessary for resolution.

Also, the experience from Chouder indicates that PKDL may occur in a notably higher percentage of treated kala-azar patients, and be more prevalent, then assumed. Though the determined point prevalence (of 3.8/1,000) for the entire village is slightly less than 4.8/1,000 - the figure reported from the only existing community based study, from an endemic area of Varanasi, India, published in 1989 - a third of Chouder's paras had rates higher than this of above 5.0/1,000 (5). Similarly, the proportion of our kala-azar patient cohort which has developed PKDL has been striking, with 14% in the entire community and 18% in two of the most affected para, significantly greater than the conventional 5-10% repeated throughout the literature for South Asian patients. Our current calculations likely underestimate the true rates of PKDL emergence in treated kala-azar patients, because identified kala-azar patients from the past 5 years - particularly those with more recent illness - may present with characteristic dermal lesions any time in the future.

Preliminary assessment of any ongoing study, particularly when sample sizes remain small, is vulnerable to mis- or over-interpretation, and analysis of eventually completed data may not support initial observations. Additionally, disease rates and patient characteristics from the highest-endemicity region of Bangladesh may not ultimately reflect and be generalizable to less affected areas of the country or elsewhere.

Nevertheless, these early community based data from Mymensingh urge not only further research into PKDL but also a timely re-evaluation of current national control measures. Ideally, kala-azar patients should be followed regularly for development of PKDL consistent skin lesions, with enhanced community based identification. Likewise, research on PKDL's epidemiology and clinic features would be best approached with close longitudinal observation (rather than cross-sectional or retrospective analysis). Finally, along with active and early case detection, disease control efforts should include an immediate exploration of alternative kala-azar and PKDL treatment strategies, and associated measures such as targeted vector control, particularly in areas with a high-density of affected individuals. -HSB