Effectiveness and Safety of Amphotericin B Deoxycholate, Amphotericin B Colloidal Dispersion, and Liposomal Amphotericin B as Third-Line Treatments for Cutaneous and Mucocutaneous Leishmaniasis: A Retrospective Study

INTRODUCTION

Leishmaniasis is a disease caused by protozoa of the genus Leishmania that generates a wide spectrum of clinical manifestations that depend on the species of Leishmania involved and the immune factors of the host.^1–3

In the New World, leishmaniasis is found from south Texas to South America, with Colombia being the country with the second most cases of leishmaniasis (7,541 cases in 2015). In the country, cutaneous leishmaniasis (CL) is the most frequent clinical form—98.2%—, followed by the mucocutaneous (mucocutaneous leishmaniasis; MCL)—1.2%—and visceral—0.5%—forms.^1,4,5

Most cases of leishmaniasis in Colombia are caused by the species Leishmania panamensis and Leishmania braziliensis, which are associated with an increased risk of mucosal involvement, with the latter also being associated with an increased risk of therapeutic failure.^6,7 The recommended treatment in the country is systemic, with the first line being pentavalent antimonials, whose effectiveness is approximately 75%. For cases of therapeutic failure and for some special cases, pentamidine, miltefosine, or amphotericin B (AB) are recommended.^8,9 However, there is great concern about therapeutic failure with the pentavalent antimonials. The experience in the Old World has shown that this failure can increase in incidence over time, even to the point of the drugs becoming useless, whereas a decrease in sensitivity to antimonials has been described in Colombia, especially in repeated cycles of drug administration and for L. braziliensis and L. panamensis.^7,10,11 The second and third cycles of pentavalent antimonials, recommended in some guidelines, are associated with levels of effectiveness of only 59% and 50%, respectively.^11,12 The effectiveness of pentamidine can reach up to 87%, whereas the response to miltefosine depends on the species, being most frequently associated with incidences of failure in L. braziliensis between 12% and 47%.^7,13,14 The aforementioned circumstances give relevance to the study of treatment strategies for therapeutic failure, with a particular interest in AB.

AB, a macrolide polyene used in the New World for 50 years in cases of therapeutic failure, binds to the ergosterol of the membrane of the parasite, increasing its permeability, which leads to successful therapy in 95–100% of cases.^14,16 Nevertheless, it is a drug with two main mechanisms of significant toxicity. First, recognition by Toll-like receptor 2 (TLR 2) triggers the release of inflammatory molecules such as interleukin 1B, leading to fever, arthralgia, headache, and gastrointestinal symptoms associated with the infusion. Second, vasoconstriction of the afferent arteriole and uptake by renal tubular cells trigger hypokalemia, hypomagnesemia, renal tubular acidosis, polyuria, and hyperazotemia.¹⁶ However, in lipid-associated formulations, deoxycholate is replaced by lipids that surround the AB monomers, reducing their uptake by susceptible tissues, thus reducing toxicity, mainly renal.¹⁶

The CL and MCL studies that have evaluated the role of the lipid-associated formulations of AB are mostly case reports and case series that have been performed primarily on travelers as the first or second lines of management, where an effectiveness of between 50% and 93% has been described. In these studies, the dosage and administration protocols have been variable, with a total average dose ranging between 7.5 mg/kg and 36 mg/kg, administered on different schedules.^15,17,18

MATERIALS AND METHODS

A retrospective descriptive study (opportunity sampling) was conducted with patients diagnosed with CL and/or MCL who were treated at the Central Military Hospital (Hospital Militar Central [HMC]) in Bogota, Colombia, between January 2010 and March 2016. The patients included in the study were diagnosed by direct smear, polymerase chain reaction, and/or anatomopathological study and presented failure with schedules and doses appropriate to the first two lines of treatment (pentavalent antimonials and miltefosine and/or pentamidine, respectively). Following the first two lines of treatment failure, the patients received treatment according to the protocol established by the Infectious Diseases Department of the hospital (see Table 1), which comprises AB in any of the three chemical forms: AB deoxycholate (ABD)—Fungizone^®(ABD; Ben Venue Laboratories, Inc. for Bristol-Myers Squibb Company, Bedford, OH), AB colloidal dispersion—Amphocil^®(ABCD; InterMune Inc., San Francisco, CA), or liposomal AB—Ambisome^® (LAB; Gilead Sciences, Inc., Foster City, CA). The follow-up of the protocol execution lasted for at least 6 months after administration of the drug.

For analyzing the data, the following descriptors were considered: location of the lesions, diagnosis, total AB dose in mg, total dose in mg/kg, dose in mg/kg/day, adverse effects by the system, pretreatment creatinine, highest creatinine during treatment and posttreatment, time to cure, and therapeutic failure. We reviewed 165 histories of patients diagnosed with CL or MCL who received AB, of which 90 patients were excluded (see Figure 1).

View Full Size

Flowchart of included and excluded patients. AB = amphotericin B; ABD = amphotericin B deoxycholate; ABCD = amphotericin B colloidal dispersion; LAB = liposomal amphotericin B.

Citation: The American Journal of Tropical Medicine and Hygiene 102, 2; 10.4269/ajtmh.18-0514

• Download Figure
• Download figure as PowerPoint slide

This study was funded and authorized by the Ethics Committee of the Central Military Hospital (HMC) in Bogota, Colombia.

RESULTS

DISCUSSION

The purpose of this study was to perform a retrospective descriptive study describing the effectiveness and adverse effects of ABD, ABCD, and LAB as third-line treatments for the New World CL and MCL, a context in which no previous study has been conducted. Our findings cannot be interpreted as it was a randomized, controlled trial; thus, comparison of different treatments must be made cautiously. We nonetheless tentatively conclude that LAB at a dose of 3–5 mg/kg/day administered in seven doses on days 1–5, 14, and 21 is an alternative treatment for patients with CL and/or MCL who had presented therapeutic failure to the first, second, and/or third lines of treatment (defined according to international and national management guidelines),^8,9 apparently with better effectiveness and safety profiles than ABD and ABCD; this was statistically evidenced in adverse effects.

In several countries, AB is used as a second- or third-line drug for the treatment of tegumentary or visceral leishmaniasis once the drugs considered to be first line (pentavalent antimonials, pentamidine, and miltefosine) have failed. The effectiveness levels of pentavalent antimonials, pentamidine, and miltefosine are 50–80%, 87%, and 53–88%, respectively, and depend on the species of the parasite, the geographical acquisition area, the management schedule, and the immunological condition of the host.^7–12 Therefore, the number of patients who require treatment with presentations of AB is considerable.

ABD was the first AB developed. It requires daily application and is associated with adverse renal, hepatic, and electrolyte effects, among others. AB in its lipid-associated formulation was developed to decrease toxicity, especially renal, and to increase effectiveness. These improvements are achieved because the lipid association decreases the amount of medication free in the plasma and, therefore, its uptake by susceptible renal cells. The formulation also favors phagocytosis by macrophages and their distribution in the reticuloendothelial system, which increases the concentration of the drug in the phagosome, where the parasite is located.^{14–16,19–22}

For the present study, identification of the Leishmania species for each patient was not performed. However, Pérez JE et al.⁷ published a series of patients who received medication in the HMC for CL. In the said publication, the species were identified by hsp70 PCR–restriction fragment length polymorphism, finding that 95.4% of the patients were infected by L. braziliensis and 2.3% by L. guyanensis even though the predominant parasite species in the country is L. panamensis. Therefore, it is most likely that most patients in our study were infected by L. braziliensis.⁷

The effectiveness of LAB was greater than those of ABCD and ABD, although statistical significance was not established possibly because of sample size. The greater effectiveness of LAB could be attributed to pharmacokinetic and pharmacodynamic advantages, as shown in murine models where the concentration of LAB has shown increasing intralesional concentration of the drug and subsequent reduction of the parasitic load.²³

Similarly to what has been reported by other authors, we observed no difference in the probability of cure among patients who presented CL and MCL, except for those who presented perforation of the nasal septum.

Scarce cases of AB–resistant Leishmania have been described, in which changes in the sterol content of the membrane of the parasite have been documented, probably because of variations in the sterol methyltransferase enzyme, losing the site of action of any AB.^24,25 It is probable that the cases of therapeutic failure in our study are rather due to immunological or idiosyncratic conditions of the patient or pharmacodynamic variables^23,26 and factors, which are also related to therapeutic failure of other antimicrobial agents.⁷

For treatment with the lipid-associated ABs, patients received 3 mg/kg/dose. However, a percentage of them presented therapeutic failure after which they received the drug at a dose of 5 mg/kg/dose, obtaining a cure rate of 75%. This is not surprising, considering the reports of other authors, where a successful treatment by repeating the dose of the drug was observed.¹⁵

Patients who presented more and larger lesions were more likely to be cured; these variables were more imperative in the LAB group. Nonetheless, these findings do not coincide with those described in other studies.^9,22 On the other hand, patients with a lower kg/day dose (< 1 mg/kg/day) in the ABD group had a higher cure rate and a lower prevalence of adverse effects; this could be due to an increase in the number of days of exposure with no need of suspension.

The dose corresponding to P10 of the total dose for each group was calculated to identify a minimum effective dose. In all chemical forms, patients treated with P10 of the total dose had a lower probability of cure (40–50%). By contrast, 100% of those patients who received P90 of the total dose of ABD and ABCD were cured.

Cure was diagnosed after up to 3 months of follow-up. Usually, after 45 days, failure is diagnosed, which implies the presence of late responders. The time to present therapeutic failure in the current study was up to 6 months of follow-up, emphasizing the importance of follow-up for at least 6 months given the presence of late failure.

There was a statistical difference between the groups regarding adverse effects. The prevalence of adverse effects in the LAB group was lesser than that of the ABCD and ABD groups. Although statistical significance was not established, the most frequent symptoms associated with infusion were fever, asthenia, chills, vomiting, and headache. Similar data have been reported in other pathologies, where the adverse effects of immunological origin associated with infusion are higher in ABCD and are similar to those evidenced in ABD infusions.²⁰ LAB, whose charge is negative and forms smaller liposomes, is recognized by TLR 4 and not by TLR 2. The difference in this recognition may attenuate the response of macrophages and, thus, reduce the effects associated with infusion.²⁷

Creatinine levels (including elevation ≥ 0.3 mg/dL) showed a greater difference between the baseline and highest levels during treatment in the ABD group; nephrotoxicity was shown in one patient in this group. It has been described that the frequency of nephrotoxicity in patients receiving ABD is at least 3-fold higher than when using LAB.¹⁶

Patients who received ABD presented electrolyte abnormalities and transaminase elevations more frequently. Most patients received the drug by central venous catheter in the ABD group, 30.7% of them presented catheter-associated bacteremia that might be explained by the long period of use of an invasive central access device (more than 2 weeks). The risk of presenting catheter-associated bacteremia was considered greater when ABD was used because the duration of use of this device is at least 20 days versus five for amphotericin with a lipid-associated presentation. The use of a PICC has been recommended, although it is not clear that PICC use is associated with a lower risk of infection.²⁸

Patients who had previously received a greater number of treatment schedules with N-methylglucamine, any treatment with AB, and a larger total dose (35 mg/kg) of lipid-associated formulations had higher percentages of adverse effects.

This research has some limitations, including that the Leishmania species were not identified; the optimal follow-up of ML after treatment is 12 months, but this information was not available; retrospective design, sample size, loss of population due to failure to meet the inclusion criteria, convenience sampling, and the inclusion of only male population, who were mostly healthy, young, and in the military; and limitations inherent these characteristics that did not allow other statistical analysis. There is little information in the literature regarding this topic; therefore, despite the limitations of our work, the results are useful for patient management and constitute a basis for future studies, whose design can confirm the proposed hypotheses and answer cost-effectiveness questions.