TEXT

Top Abstract Text References

In Leishmania-human immunodeficiency virus (HIV)-coinfected patients, the first course of antimonial pentavalent compound is unsuccessful in more than 50% of patients (1). Relapses after apparent clinical recovery occur after a mean time of 4.5 months (8). Amphotericin B (AmB), an antifungal agent which has proved to be active also against Leishmania spp., is currently proposed as an alternative first-line treatment (12). AmB was first used with the deoxycholate formulation (Fungizone), but acute and chronic toxic side effects following AmB deoxycholate administration have promoted the use of lipid formulations which are much better tolerated (16). The liposomal formulation of AmB (Ambisome) represented progress in the fight against visceral leishmaniasis (VL). However, most immunocompromised AmB-treated patients, initially considered cured, relapsed clinically and parasitologically after 3 to 22 months (3). Secondary cures by Ambisome were also followed by relapses (4). These relapses, probably related to the immune status of patients, might also reflect ineffective drug levels in contact with some Leishmania parasites which could be in unusual locations in HIV patients (6). In some cases, primary or secondary resistance to AmB could be involved in treatment failures as was described for antimonial compounds. A previous study showed that different Leishmania infantum strains may require different AmB levels to be cleared (7). Thus, the median effective dose (ED₅₀) determined in BALB/c mice was 0.17 mg/kg of body weight for an isolate obtained from an untreated patient, whereas the ED₅₀ was 0.41 mg/kg for an isolate obtained from a patient who had received 12.4 g of AmB over 3 years (7). A study conducted on isolates obtained from the same patient before and after treatment was essential in order to assess the occurrence of secondary resistance to AmB. The aim of the present work was to determine ED₅₀s on parasites obtained from a Leishmania-HIV-coinfected patient before and after AmB therapeutic exposure, in a model associating L. infantum and BALB/c mice. As the patient lived in an area of endemicity during the treatment period, we have ensured by isoenzymatic study that the isolate obtained after 4 years of therapeutic exposure had the same profile as that of the initial isolate.

The first isolate was obtained in 1992 from an HIV-positive drug addict patient. The strain was identified as L. infantum; its zymodeme was close to zymodeme MON-1 but with a difference on one electromorph. During November 1992, the patient presented the first episode of typical VL: febrile pancytopenia and hepatosplenomegaly. This first episode was treated successfully with two courses of Glucantime (antimoniate meglumine). The first relapse occurred 10 months later, and subsequent relapses were treated either with AmB (Fungizone or Ambisome) or pentavalent antimony (Sb^v). Itraconazole and allopurinol were introduced occasionally as secondary prophylaxis. By December 1996, the patient had received a total dosage of Sb^v in excess of 325 g and a total dosage of AmB (whatever the formulation) exceeding 14 g. During the 4-year course, 27 Leishmania isolates were obtained from this patient, either on Novy-MacNeal-Nicolle medium or after inoculation into a golden hamster. All the isolates were stocked in nitrogen liquid. The first isolate was obtained before any treatment. Five isolates, randomly selected among the 27 isolates, were characterized by isoenzyme analysis (three isolates obtained from bone marrow in November 1992, September 1993, and September 1994 and two isolates obtained from blood in September 1994 and November 1996).

The isolates were characterized by starch gel electrophoresis with 15 enzymatic systems (11). The same isoenzymatic profile was identified each time and was identical to that of the reference strain L. infantum MON-1, except for one electromorph (malic enzyme) which was absent in every characterization.

AmB ED₅₀s were determined in a murine model on the first naive stock (stock A, November 1992) and on the 19th stock (stock B, November 1996), as previously described (7). Experiments were conducted with BALB/c male mice (5 weeks old, 20 ± 2 g) purchased from IFFA CREDO (l'Arbresle, France). Cryopreserved promastigotes of the patient stocks were thawed and cultivated on Novy-MacNeal-Nicolle medium and then on RPMI medium. On day 0, mice were inoculated via the tail vein with 2 × 10⁷ infective Leishmania promastigotes in a 0.1-ml volume. This procedure induced a liver parasite burden of 10.4 × 10⁷ amastigotes per mg of liver weight for the first isolate (stock A) and 0.52 × 10⁷ for the second (stock B). The stocks required high inoculum and 3 weeks to visceralize, reflecting a relatively low virulence in BALB/c mice for the two stocks studied in this work, in comparison with previous studies using other strains (7).

Mice infected with stock A and mice infected with stock B were randomly divided into five groups. On days 21, 23, and 25, control groups (10 mice) received isotonic sodium chloride, and four groups of five mice received 0.05, 0.1, 0.5, and 0.8 mg of AmB deoxycholate, respectively, per kg. On day 28, the animals were killed by cervical dislocation and autopsied. The Guiding Principles for Biomedical Research involving animals, published in 1986 by the Council for International Organizations of Medical Sciences, were followed during all procedures. The drug efficacy was determined by evaluating the liver parasite burden and the measure of efficacy doses. The liver parasite burden was evaluated after Giemsa staining of the smears. The number of amastigotes per 500 hepatocytes was calculated and related to liver weight (in milligrams) according to the formula of Stauber et al. (13). The percentage of parasite suppression was calculated as [1  (mean Stauber count of the treated group/mean Stauber count of the control group)] × 100. ED₅₀s (doses of the drug calculated to eliminate 50% of parasites compared to controls) were determined by the Michaelis-Menten model.

Statistical analysis. Results were expressed as means ± standard errors of the means. A one-way analysis of variance or a U test was performed to compare the influences of various parameters. A P value lower than 0.05 was considered statistically significant.

Control experiments. In group A, mice showed a mean of 0.40 L. infantum amastigotes per liver cell nucleus at the end of the 28-day period of experimentation (range, 0.1 to 0.84; n = 10). In group B, mice showed a mean of 0.02 L. infantum amastigotes per liver cell nucleus at the end of the same time period (range, 0.004 to 0.028; n = 10). The average liver parasite burdens obtained for each group were significantly different (10.4 × 10⁷ in group A versus 0.52 × 10⁷ in group B).

Treated groups. Maximal parasite suppression obtained with AmB deoxycholate was 96.8% ± 0.4% (Table 1) in group A and 100% ± 0% in group B (Table 1). These maximal parasite suppressions obtained at the same dose in the two groups (0.8 mg/kg) were not significantly different. The ED₅₀s were 0.053 ± 0.010 mg/kg in group A and 0.067 ± 0.016 mg/kg in group B. These results were not significantly different (P = 0.26).