To the Editor: During 2000–2010, serial Leishmania isolates obtained from an HIV-infected patient who was not responding to treatment showed a gradual decrease in in vitro miltefosine susceptibility. We performed L. donovani miltefosine transporter (Ldmt) gene analysis to identify an association between miltefosine resistance of reference L. donovani lines and variability in miltefosine response of L. infantum isolates. A new single-nucleotide polymorphism (SNP), L832F, was identified, which might be a marker of miltefosine resistance in leishmaniasis.

The patient, a 46-year-old woman, had lived in France since 1994 but regularly returned to Algeria, her country of birth. HIV-1 infection was diagnosed in 1991. Antiretroviral therapy was initiated in 1993, leading to undetectable viral load and a CD4+ T-cell count of 185 cells/mm³ (reference >450/mm³). Concurrent conditions were thoracic herpes zoster in 1996, hairy leukoplakia of tongue, oropharyngeal candidiasis, and chronic renal failure of unknown cause since 2000.

Visceral leishmaniasis was diagnosed in 1998 by culture of a bone marrow smear, which showed intracellular amastigotes. Use of meglumine antimonate (Glucantime; Sanofi, Paris, France), a drug of choice for the treatment of leishmaniasis, was contraindicated because of pancreatitis in the patient and in vitro isolate susceptibility variation; therefore, induction therapy consisted of liposomal amphotericin B (AmpB [AmBisome; Astellas Pharma US, Deerfield, IL, USA]) at a dose of 3 mg/kg/d for 5 consecutive days, then 1× week for 5 weeks (total dose 30 mg/kg) during 1998–2000 (Table). The same medication was administered for relapses at 4 mg/kg/d for 5 days, then 4 mg/kg 1× week for 5 weeks (total dose 40 mg/kg) during 2001–2010. Given the adverse effects of AmpB and the availability of oral miltefosine (Impavido; AEterna Zentaris Inc., Quebec City, Quebec, Canada), the latter drug was used for maintenance treatment during 2001–2007 at 50 mg 2×/d. Leishmaniasis was monitored by leukocytoconcentration and culture of blood samples on Novy-Nicolle-McNeal medium.

When signs of biological and clinical relapse appeared, bone marrow was aspirated for parasite detection. After culture of the aspirate and isoenzyme determination, the strain was identified as L. infantum, zymodeme MON-24. Eleven relapses were documented; all were confirmed by positive direct examination of bone marrow or blood, but cultures of only 7 samples yielded positive results (Table).

The susceptibility of 4 cryopreserved isolates (S₁, S₃, S₄, and S₆; Table) to AmpB and to miltefosine was studied in the in vitro promastigote and axenic amastigote form by determining the concentrations inhibiting parasite growth by 50% (1,2). The 50% inhibitory concentration (IC₅₀) was determined in parallel for the following reference L. donovani lines: a wild-type L. donovani LV9 (MHOM/ET/67/HU3) line (LV9 WT), a wild-type L. donovani DD8 (MHOM/IN/80/DD8) line (DD8 WT), a laboratory miltefosine-resistant line obtained from LV9 WT (LV9 miltefosine-R, resistant to 90 μmol/L miltefosine), and the laboratory AmB-resistant line obtained from DD8 WT (DD8 AmB-R, resistant to 1.4 μmol/L AmB) on promastigote and axenic amastigote forms (3,4).

The AmB susceptibility of the isolates did not change notably over time; IC₅₀ values ranged from 0.09 µmol/L to 0.24 µmol/L, regardless of parasite form, similar to those of wild-type reference strains (Table). In contrast, the IC₅₀ values of miltefosine increased greatly over time, from 5.00 µmol/L to 50.10 μmol/L. During the 6 years of follow-up with miltefosine maintenance therapy, the susceptibility of the isolate (S₆) obtained 6 months after miltefosine treatment withdrawal in 2008 was 6-fold higher than that of the first isolate (S₁) obtained in 2000.

The L. donovani miltefosine transporter protein (LdMT) promotes miltefosine translocation (5), and LdMT inactivation in L. donovani promastigotes leads to miltefosine resistance at the promastigote and amastigote stages (6). In 2003 and 2006 studies, several mutations were linked to the inability of parasites to take up miltefosine and to miltefosine resistance (5,7). In a 2009 study, the weak expression of LdMT and its β subunit LdROS3 in L. braziliensis isolates was linked to diminished sensitivity (8). We sequenced the entire Ldmt gene (3,294 bp) in the reference strains and the clinical isolates for SNP analysis (5,7). Only 1 new SNP, L832F, was found in the miltefosine-resistant reference strain (LV9 miltefosine-R) and in clinical isolate S₆. The L832 wild-type allele was found in isolate S₁ and in the miltefosine-sensitive reference lines (LV9, DD8, and DD8 AmpB-R), whereas both alleles were found in isolates S₃ and S₄, with a decrease in the wild-type allele (Table). The last isolate, which was obtained 3 years after miltefosine withdrawal and could not be subcultured, had reverted to the wild-type allele (L832).

These results point to a relation between the 832F allele and diminished susceptibility to miltefosine. Analysis of this case of miltefosine resistance in a patient co-infected with Leishmania sp. and HIV strongly suggests that an SNP (L832F) in the Ldmt gene could represent a molecular marker of miltefosine resistance in L. infantum and L. donovani.