Entry and exclusion criteria

This open label study was carried out at the Kala-Azar Medical Research Centre in Muzaffarpur, Bihar, where visceral leishmaniasis caused by Leishmania donovani is endemic. It was approved by the ethics committee of the Institute of Medical Sciences, Banaras Hindu University, Varanasi.

Patients of any age with visceral leishmaniasis caused by L donovani were eligible if they had symptoms or signs of the disease (fever, weight loss, splenomegaly) and parasites were found by microscopy in a splenic aspirate smear.9 Patients were excluded if they were pregnant or breast feeding, HIV positive, had a serious concurrent infection such as tuberculosis or bacterial pneumonia, or if they had a granulocyte count <1×10⁹/l, haemoglobin concentration <40 g/l, or platelet count <40×10⁹/l. Eight patients were excluded by these criteria.

Trial procedures and treatment

After completing baseline testing (standard haematological and biochemistry profiles, urine analysis, chest radiography, electrocardiography, testing for anti-HIV by ELISA, and malaria smear)9 and providing written informed consent (from parent or guardian for children), participants were randomised by sealed envelope to receive 5 mg/kg of liposomal amphotericin (AmBisome, NeXstar, Paris) as a single infusion or as once daily infusions of 1 mg/kg for five consecutive days. An independent statistician prepared the randomisation envelopes using a computer based random number generator. The study staff opened consecutively numbered envelopes containing the treatment assignment after eligible patients fulfilled the entry criteria.

Treatment was started within three days after splenic aspiration. The dose was given as a 60 minute intravenous infusion with no pretreatment with antipyretic drugs.13 Patients were examined daily and remained in the unit until 14 days after the last infusion, at which time splenic aspiration and laboratory evaluation were repeated. We graded parasite density scores for aspirates taken before and after treatment in a blinded fashion using a conventional logarithmic scale of 0 (no parasites per 1000 oil immersion fields) to >6 (>100 amastigotes per field).9 Since we could not rationalise giving the single dose group subjects four additional placebo infusions, clinicians caring for the patients were not blinded to the treatment given.

We defined apparent cure 14 days after treatment ended as absence of fever, clinical improvement, reduction in spleen size, and a splenic aspirate score of 0 (apparent parasitological cure). 9 13 Definitive cure, assessed after six months of follow up, was defined as being healthy with no signs or symptoms of relapse. 9 13

Statistical analysis

We compared response rates in the treatment groups by Fisher's exact test and calculated exact binomial 95% confidence intervals for the individual proportions responding to each treatment regimen. The objective of this study was to gather preliminary data on the relative efficacy of two treatments that would, if the results warranted it, inform the design of a more precise trial with appropriate statistical power.

Patient characteristics and initial responses

The figure shows the flow of participants through the trial, and table 1 summarises the clinical features at entry to the study. Thirty six subjects had not responded to pentavalent antimony (sodium antimony gluconate, 20 mg/kg per day for ⩾28 days 3 9); none had received amphotericin B. The groups were well matched except for haemoglobin concentration and platelet count (table 1).

Flow of participant through trial

• Download figure
• Open in new tab
• Download powerpoint

At entry, all 46 patients in the single dose group and 43 of 45 in the five dose group had fever. Seven days after treatment started, 45 of the single dose group and 43 of the five dose group had no fever, and mean duration of fever was identical in the two groups (2.6 (0.3) v 2.6 (0.4) days). At evaluation, 14 days after the last amphotericin infusion (table 2), none of the 91 participants had fever and all but one in each group had no parasites in splenic aspirate smears and fulfilled the criteria for apparent cure. One patient in each group had scanty parasites in the repeat aspirate smear and their treatment was designated as a failure. Compared with baseline, patients had reduced spleen size and increased white blood cell count, haemoglobin concentration, and platelet count (table 1).

Adverse reactions

Twenty four (52%) patients in the single dose group and 20 (44%) in the five dose group had no reaction during or after any infusion. In the remaining patients, reactions were mainly mild, related to the infusion (typical of reactions associated with amphotericin B), and seldom persisted beyond one or two hours after the end of the infusion. In the single dose group, three patients experienced increased temperature alone, one chills alone, and 18 both; two patients vomited once. In the five dose group, four patients had at least one episode of higher fever alone, one chills alone, and 18 both; two had brief vomiting and two had transient back pain. Blood urea nitrogen and creatinine concentrations did not change from day 0 in either group (table 1).

Outcome and retreatment

We followed all apparently cured patients for six months after treatment. 3 9 13 Five patients had asymptomatic relapses that were confirmed by microscopy (three in single dose group, two in five dose group, table 2). All other participants were asymptomatic, appeared healthy, and were judged to have a definitive cure. Thus, overall, liposomal amphotericin (5 mg/kg) cured 92% of patients (84/91), 91% (95% confidence interval 79% to 98%) of those treated with a single infusion and 93% (82% to 99%) of those who received five daily infusions of 1 mg/kg (table 2).

The five patients who relapsed and the two initial non-responders were retreated with amphotericin B deoxycholate (15 alternate day infusions of 1 mg/kg). All responded and were considered cured six months later.

Resistance

Another issue is whether patients who relapse or do not respond to low dose treatment could develop resistance to amphotericin B. Our approach is to retreat patients in whom treatment fails with conventional amphotericin B. 4 13 Although the results are limited, none of the 13 patients who failed to respond to low dose liposomal amphotericin B13 or the 32 who did not respond to short course, low dose amphotericin B lipid complex treatment4 have shown clinical resistance to amphotericin B.

Cost of treatment

The principal rationale for testing low or single dose regimens is to identify a treatment that is not only effective and convenient but affordable. Commercially available lipid formulations of amphotericin B are expensive, and the cost of even short course or low dose regimens 4 10 14 are an insurmountable obstacle in the developing world.1 For example, US average wholesale prices for the various lipid formulations of amphotericin B are $188 (£125) per 50 mg vial for AmBisome (liposomal), $93 per 50 mg for Amphotec (cholesterol dispersion), and $194 per 100 mg for Abelcet (lipid complex).15 The regimen for liposomal amphotericin B approved by the US Federal Drugs Administration for visceral leishmaniasis consists of 3 mg/kg given on days 1-5, 14, and 21 (total dose, 21 mg/kg).16 AmBisome currently costs about $173 per 50 mg in India, and in a 25 kg patient the regimen equates to about $1900. The regimen used in our study, which produced cure rates of 92% rather than 97%, would cost $519. This is more affordable but still too high.

Amphotericin B deoxycholate (Fungizone) is cheaper than the lipid formulations (US wholesale price $17 per 50 mg vial, retail price in India $6 per 50 mg vial). 4 14 It is now first line treatment for visceral leishmaniasis in Bihar because of antimony resistance and produces cure rates of >95% in patients who complete treatment. However, the regimens are prolonged and therefore expensive, require daily infusions for 20 days or (more usually) on alternate days over 30 days, and are not always well tolerated. 4 17 Short course regimens using drugs such as AmBisome or Abelcet, 9 13 which also seem to produce fewer adverse reactions than the deoxycholate form, 4 13 are therefore attractive.

Although hospital stay and costs can be appreciably reduced by short courses and especially single dose regimens, 4 14 these savings do not offset the cost of the drug in countries such as India, where hospital charges are low. If definitive trials establish the efficacy of our low dose regimen, drug companies will have to cut prices for the potential benefits to be realised.