Although pentavalent antimonial agents have been the preferred treatment for visceral leishmaniasis for decades, increasing numbers of treatment failures with antimony are being reported throughout the world [8-12]. For patients who do not initially respond to or have relapse after 20 to 30 days of parenteral antimony therapy, options include more lengthy re-treatment with antimony, prolonged treatment with conventional amphotericin B or pentamidine, or re-treatment with antimony in combination with a second agent (for example, aminosidine or interferon-) [8-13]. Other therapeutic alternatives for patients who have not previously received treatment or who have been unresponsive to antimony are also being actively tested; these alternatives include new lipid formulations of amphotericin B, which seem to be remarkably effective [4, 13-18].

In an initial study of Indian patients with visceral leishmaniasis, we showed that treatment with amphotericin B lipid complex was well tolerated and induced long-term responses in 100% of 25 patients who had not responded to or had had relapse after more than 30 days of conventional pentavalent antimony therapy [17]. Twenty-one of these patients received five infusions, 3 mg/kg of body weight, every second day over a 10-day period (total dose, 15 mg/kg); four patients received the same total dose, given as 3 mg/kg per day over a 5-day period [17]. To test the efficacy of the latter, particularly useful short-course regimen and to define the optimal dose of amphotericin B lipid complex in patients with visceral leishmaniasis that is unresponsive to antimony, we treated 60 patients with this condition for 5 consecutive days with total doses of 5, 10, or 15 mg/kg.

Methods

Top Methods Results Discussion Author & Article Info References

Patients and Entry Criteria

This open-label study, approved by the ethical committee of Banaras Hindu University, Varanasi, India, was conducted at the Kala-Azar Medical Research Center in Muzarffarpur, Bihar, India. Visceral leishmaniasis caused by L. donovani is endemic in this region. Informed consent was obtained from all patients or their parents or guardians. Each study patient had previously had visceral leishmaniasis treated with antimony and was evaluated at our center because of persistent or recurrent symptoms or signs suggestive of active visceral leishmaniasis. Typical symptoms of visceral leishmaniasis included fever, weight loss, weakness, and malaise; typical signs were splenomegaly, hepatomegaly, and an oral temperature greater than 38.5°C [1, 2]. Patients were eligible for the study if they had consistent symptoms or signs [1], had parasites shown on splenic aspirate smear (Figure 1), and had previously documented visceral leishmaniasis that was treated with full-dose pentavalent antimony (20 mg of sodium antimony gluconate per kg per day) for at least 30 consecutive days [12, 17]. Patients were excluded if they were younger than 5 or older than 65 years of age; had a history of cardiac disease; were pregnant; had a serious concurrent infection, such as tuberculosis or bacterial pneumonia; or had one of the following abnormal laboratory findings: granulocyte count less than 1.0 x 10⁹/L, hemoglobin level less than 60 g/L, platelet count less than 5.0 x 10⁹/L, prothrombin time greater than 4 seconds above control values, bilirubin level greater than 34.2 µmol/L, serum aminotransferase level greater than 200 IU, or creatinine level greater than 176 µmol/L.

View larger version (141K): [in this window] [in a new window]   Figure 1. Photoµgraph of a pretreatment splenic aspirate smear showing characteristic intracellular Leishmania amastigotes within splenic mononuclear phagocytes. Extracellular organisms were probably released during preparation of the smear (Giemsa stain; original magnification, x630).

Trial Procedures and Treatment

After microscopic documentation of characteristic amastigotes on splenic aspirate smear (Figure 1) [12, 17] and completion of baseline testing (standard hematologic and biochemistry profiles, urinalysis, chest radiography, electrocardiography, malarial smear, and examination of stool for ova and parasites [17]), patients were randomly assigned by sealed envelope to receive 5 days of treatment with 1 (group A), 2 (group B), or 3 mg (group C) of amphotericin B lipid complex per kg per day. The Liposome Co. (Princeton, New Jersey) provided the study drug (ABELCET) only and had no role in gathering, analyzing, or interpreting the data or in deciding whether the study findings were to be submitted for publication.

Treatment was initiated within 2 days after splenic aspiration was performed, and the study drug was given by 2-hour intravenous infusion [17], all patients received antipyretic treatment with 0.25 to 1.0 g of oral paracetamol (the only other treatment allowed) 1 hour before each infusion of amphotericin B lipid complex. Spleen size was measured below the left costal margin in the anterior axillary line.

Splenic aspiration was repeated 2 weeks after treatment ended (day 19). Parasite density score for aspirates obtained before treatment and those obtained on day 19 was graded microscopically (x10 eyepiece, x100 objective) by using a conventional logarithmic scale in which 0 indicated no parasites per 1000 fields and +6 indicated more than 100 amastigotes per oil-immersion microscopic field [12, 17]. The microscopist who examined the splenic aspirates and bone marrow aspirates (see below) was blinded to the treatment dose; the clinicians caring for the patients were not blinded. The designation of apparent response on day 19 required the absence of fever, clinical improvement with a reduction in spleen size, and a splenic aspirate score of 0 [12, 17]. Definitive or complete response was assessed 6 months after treatment and required the absence of signs or symptoms of relapse and a parasite-free bone marrow aspirate [12, 17]. No facilities were available to culture the splenic aspirates. None of the 60 patients was lost to follow-up.

Statistical Analysis

Data are expressed as the mean ± SE. We used analysis of variance to detect differences among the three study groups in clinical and laboratory results; the one exception was differences in sex, which were analyzed by using the chi-square test. Exact binomial 95% CIs for the individual proportions of patients responding to each dose were computed. By using STATXACT software (Cytel Software Corp., Cambridge, Massachusetts) and the Jonckheere-Terpstra test, we compared the percentages of patients who responded to increasing drug doses. A P value less than 0.05 was considered statistically significant.

Results

Top Methods Results Discussion Author & Article Info References

Patient Characteristics

Pertinent clinical features and laboratory results for the three patient groups are shown in the (Table 1). No significant pretreatment differences were seen among the three groups for the variables shown. Before study entry, patients in groups A, B, and C had been treated with pentavalent antimony (20 mg/kg per day) for 40 ± 4 days, 44 ± 6 days, and 39 ± 2 days, respectively. Eleven of 19 patients in group A, 11 of 20 patients in group B, and 14 of 21 patients in group C had previously shown an apparent response to antimony and were therefore considered to have had relapse. The remaining patients had never responded to antimony and thus were judged to have had primary drug unresponsiveness [12, 17]. Nine of the 60 patients had also previously received unsuccessful treatment with either pentamidine (n = 7) or conventional amphotericin B (n = 2). Two of these 9 patients were in group A, 3 were in group B, and 4 were in group C.

Response to Treatment

Regardless of the dose of amphotericin B lipid complex used (1, 2, or 3 mg/kg per day), all 60 patients showed a prompt clinical response. Fifty-one (85%) became afebrile within 5 days after the fifth and final infusion. As expected [17], almost all patients in each group initially developed infusion-related chills and increases in baseline body temperature (up to 40.1 °C) during the first hour of infusion. These reactions are typical of therapy with conventional amphotericin B, were not dose related (they developed in 100% of patients in groups A and C and in 95% of patients in group B), and were tolerated clinically in each instance. Chills seldom persisted beyond the first 30 to 60 minutes of infusion; the increase in body temperature usually lasted for as long as 1 hour after the infusion was completed. Reactions to amphotericin B lipid complex decreased in both frequency and intensity with successive infusions (these changes were similar to those seen in our previous experience [17]); 25 patients (42%) had no reaction to the fifth treatment.

Two weeks after treatment (day 19), all 60 patients had splenic aspirate scores of 0 and all were considered to have an apparent clinical and parasitologic response. At this time, increases in mean Karnofsky score, leukocyte count, hemoglobin level, and platelet count and reduction in spleen size were seen in all three groups (Table 1). On day 19, no patient showed any change in either mean blood urea nitrogen level (81.3 ± 4.4 nmol/L) or serum creatinine level (6.8 ± 0.2 µmol/L) compared with pretreatment levels (80.4 ± 4.3 nmol/L and 6.6 ± 0.1 µmol/L, respectively).

Patient Outcome

Five patients (three in group A and two in group B), none of whom had previously received treatment with conventional amphotericin B, had relapse after 2 to 6 months. Each of these patients developed new fever and increased spleen size and showed amastigotes on repeated splenic aspirate smear. The patients in group A who had relapse responded to re-treatment with twice as much amphotericin B lipid complex (total dose, 10 mg/kg) given over 5 days; the two patients in group B who had relapse responded to 5 days of re-treatment with a total dose of 15 mg/kg. At follow-up 6 months later, each of these five patients was considered to have a complete response. The remaining 55 patients, including all patients in group C, were healthy 6 months after treatment with amphotericin B lipid complex and had parasite-free bone marrow aspirate smears; thus, each was considered to have shown a definitive response. These patients have now been followed for a mean of 11.2 ± 0.2 months (range, 10 to 13 months), and all have remained healthy.

Discussion

Top Methods Results Discussion Author & Article Info References

Our results indicate that 5 days of treatment with amphotericin B lipid complex readily induces initial and long-term responses in Indian patients with visceral leishmaniasis who had not responded to or had had relapse after standard antimony treatment. This short-course regimen of amphotericin B lipid complex is therefore an important therapeutic alternative that has obvious clinical appeal for this expanding [8], often difficult-to-manage subset of patients. These patients typically require prolonged antimony re-treatment lasting more than 30 days; treatment with conventional amphotericin B or pentamidine; or combination therapy with antimony plus a second agent, such as interferon- [8-13].

In addition to amphotericin B lipid complex [17], two other lipid-associated formulations of amphotericin B have also been tested in patients with visceral leishmaniasis in other endemic regions and in India [4, 14-16, 18]. Some of the patients in these studies had not responded to pentavalent antimony [4, 14, 16]. For amphotericin B cholesterol dispersion (Amphocil, Liposome Technology, Menlo Park, California), 2 mg/kg given for 7 consecutive days (total dose, 14 mg/kg) seems to be optimal in Brazilian children [15]. For liposomal amphotericin B (AmBisome, Vestar, San Dimas, California), total doses of 18 to 24 mg/kg administered for more than 10 days and 24 mg/kg given over 14 days seem to be optimal in Europe [4, 14] and Sudan [16], respectively. In a recently reported preliminary study also from Bihar, India, total doses of 6 to 14 mg of liposomal amphotericin B per kg administered over a 10-day period induced long-term responses in all patients [18].

It should also be pointed out that visceral leishmaniasis has, predictably, joined the list of AIDS-related opportunistic infections [1, 7]. Although initial clinical responses can be induced by treatment in most patients with AIDS-associated visceral leishmaniasis, relapse is typical once any form of therapy (including therapy with liposomal amphotericin B [7]) is discontinued [7, 19]. Although the patients enrolled in our study were not tested for HIV antibody, studies done in 1994 [20] and 1995 [17] indicate that the prevalence of HIV infection in Bihar, India, currently seems low. Other potential limitations of our study include slightly different numbers of patients treated previously with agents other than antimony, the absence of culture data to support the microscopic results for parasitologic effects, and the fact that this trial was conducted in only one geographic location. In view of the reported efficacy of lipid formulations of amphotericin B in visceral leishmaniasis in other parts of the world [4, 14-16], however, it seems unlikely that our results with amphotericin B lipid complex would not be applicable in other areas where visceral leishmaniasis is endemic.

Because 39 of 41 patients (95%) receiving a total amphotericin B lipid complex dose of more than 10 mg/kg (groups B and C) achieved definitive responses whereas 3 of 19 (16%) patients who received a total of 5 mg/kg (group A) had relapse, 10 mg/kg seems to be a reasonable total dose for amphotericin B lipid complex in patients who are unresponsive to antimony. In addition, a significant dose effect was seen when 6-month response rates for groups A, B, and C (84%, 90%, and 100%, respectively) were compared (Jonckheere-Terpstra test; P < 0.034). It is clear that a relapse rate of 16% (which occurred in group A) cannot be dismissed. Nevertheless, the high cost of all formulations of lipid-associated amphotericin B [4, 13-16], including amphotericin B lipid complex, suggests that the 84% response rate induced by 5 mg/kg (total dose in group A) might be considered pharmacoeconomically acceptable when balanced against the savings achieved by administering half as much drug as used, for example, in group B. Therefore, initially administering amphotericin B lipid complex at 1 mg/kg per day for 5 days and being prepared to re-treat any relapses with twice this dose may be a strategy worth testing in this patient population in India and in other areas of the world where visceral leishmaniasis is endemic.

Dr. Horwith: Clinical Research, The Liposome Co., One Research Way, Princeton, NJ 08540.

Dr. Murray: Department of Medicine, The New York Hospital-Cornell Medical Center, Box 130, 525 East 68th Street, New York, NY 10021.