Overt visceral leishmaniasis, i.e., infection with Leishmania sp. associated with high fever, weight loss, massive splenomegaly and markedly altered laboratory parameters, is generally considered fatal if untreated [1, 2]. The possibility of transient spontaneous remission has been mentioned [3] but, to our knowledge, has never been precisely documented because patients with documented VL are treated. We provide the first description of a complete, prolonged remission of fully symptomatic VL in a patient who had not received any anti-leishmanial therapy.

In 2008, relapse occurred with weight loss (4 kg), fever (40 °C) and arthralgia (Table 1). Physical examination showed an enlarged spleen, measuring 20 cm on the subsequent CT scan. Antinuclear antibodies were positive (dilution of 1/1280) with a nonspecific pattern. The bone marrow (BM) aspirate showed hemophagocytosis. All microbiological tests, including Histoplasma capsulatum antibody were negative except anti-leishmanial antibody titers that were again positive (ELISA >1.7 for a threshold at 1). Quantitative polymerase chain reaction (PCR) on a L. infantum kinetoplast DNA target found 30 parasites per mL, i.e., above the threshold for active VL (1 parasite/ml) [4]. Blood and BM cultures for L. infantum were negative.

Fever lasted 4 weeks before the diagnosis of VL was confirmed. The patient was then treated with liposomal Amphotericin B (l-AmB) (3 mg/kg/day D1-5 and D10) [5]. Complete fever resolution was obtained in 3 days along with a dramatic improvement of the patient’s general well-being. Ultrasonography showed regression of splenomegaly from 20 to 14 cm at the end of the first week, followed by normalization to 12 cm 6 months later. At that time laboratory parameters had returned to normal levels (Table 1) and quantitative PCR for L. infantum was negative. When last seen, in September 2010, the patient was completely asymptomatic.

We reexamined a bone marrow smear performed in 2006 during the first hospitalization. Independent, complete examination of this smear by 2 observers (10 h of observation under the microscope) revealed the presence of a single typical amastigote form of Leishmania sp. (Fig. 1). Material on the smear was then thoroughly scarped from the slide with a blade and resuspended in 30 μl of sterile distilled water. Real time PCR amplification was performed in a thermal cycler TAQMAN® (Applied Biosystems) using Leishmania kDNA specific primers as described by Mary et al. [4]. One negative blood slide and one negative tissue slide were used as controls. Amplification of the material retrieved from the slide occurred at 28 CTs, corresponding to the presence of 1.8 parasites (Fig. 1).

This is the first documented report of a patient with overt, confirmed VL experiencing a complete remission in the absence of any anti-leishmanial therapy. The diagnosis of VL at the time of the self-resolving episode was strongly suspected based on clinical presentation and presence of antileishmanial antibody, then unequivocally confirmed years later by the presence of an amastigote on a stored smear. Quantitative PCR from the material scraped from this same slide using Leishmania-specific primers was positive. The excellent specificity of this PCR has been confirmed by several teams [4, 6–9]. The last episode, during which PCR was positive in a blood sample was followed by a positive outcome under anti-leishmanial therapy with liposomal amphotericin B. No relapse occurred during the 2 years following administration of anti-leishmanial therapy, which is consistent with the high efficacy of l-AmB in Mediterranean VL [10].

It has been known for decades that patients with mild symptoms of VL may recover spontaneously and that a minority eventually develop overt clinical VL [11]. In L. infantum transmission foci, asymptomatic infections and mild clinical forms of VL are indeed frequent in humans [12, 13]. This is different from the spontaneous remission of fully symptomatic severe VL, as described here. Uncontrolled studies that show improvement or apparent cure in patients receiving experimental anti-leishmanial interventions for overt VL should thus be interpreted cautiously [14, 15]. We cannot definitely exclude the hypothesis that the short course of intravenous antibiotic therapy (ceftriaxone and ofloxacin during 8 days) during the first episode may have contributed to the patient's improvement. A few reports in vitro or in animals suggest indeed that quinolones may have some activity against Leishmania [16–18] thus possibly contributing to the control of parasite loads. Thereafter, the patient was appropriately treated and cured with liposomal Amphotericin B.

For confirmation of VL, PCR in the blood is more sensitive than conventional search of the parasite even in bone marrow aspirates. Time spent for reading Giemsa-stained smears in the search for Leishmania amastigotes increases the test sensitivity that raises to 95.4 % after 60 min and 89.7 % when 1.200 fields are examined [19]. When a macrophage activation syndrome is present, repetition of bone marrow aspirates also increases diagnostic sensitivity [20]. Why patients would spontaneously recover from overt VL is not clear. Interestingly, our patient met five out of the eight criteria of the hemophagocytic lymphohistiocytosis (HLH-2004 revised diagnostic guidelines) [21]. Therefore, if HLH had also occurred during the initial episode, phagocytosis of Leishmania-infected cells by previously activated macrophages, or a general activation of macrophages as observed in HLH, may have contributed to the control of parasite loads. This hypothesis is consistent with the absence of Leishmania amastigotes in bone marrow aspirates in 36.3 % of patients with VL and HLH [22].

In summary, this report demonstrates that complete spontaneous remission may occur in patients with overt, fully symptomatic VL. VL should therefore be considered in cases of self-resolving or relapsing episodes of fever of unknown origin. Confirmation should be based on both serological tests and specific PCR on a blood sample.

Patient was informed of the process by this attending physician using a procedure common to all French National Reference Centers (NRC) (http://www.parasitologie.univ-montp1.fr/conseil.htm) and gave his oral consent for data collection and publication. Mention of this consent was written in the medical chart.