Drug resistance in Indian visceral leishmaniasis

Tropical Medicine and International Health volume6no11pp849±854november2001 Drug resistance in Indian visceral leishmaniasis Shyam Sundar Kala-azar Medical Research Centre, Banaras Hindu University, Varanasi, India Summary Throughout the world, pentavalent antimonial compounds (Sb v ) have been the mainstay of antileishmanial therapy for more than 50 years. Sb v has been highly effective in the treatment of Indian visceral leishmaniasis (VL: kala-azar) at a low dose (10 mg/kg) for short durations (6±10 days). But in the early 1980s reports of its ineffectiveness emerged, and the dose of Sb v was eventually raised to 20 mg/kg for 30±40 days. This regimen cures most patients with VL except in India, where the proportion of patients unresponsive to Sb v has steadily increased. In hyperendemic districts of north Bihar, 50±65% patients fail treatment with Sb v. Important reasons are rampant use of subtherapeutic doses, incomplete duration of treatment and substandard drugs. In vitro experiments have established emergence of Sb v resistant strains of Leishmania donovani, as isolates from unresponsive patients require 3±5 times more Sb v to reach similarly effectiveness against the parasite as in Sb v responders. Anthroponotic transmission in India has been an important factor in rapid increase in the Sb v refractoriness. Pentamidine was the rst drug to be used and cured 99% of these refractory patients, but over time even with double the amount of initial doses, it cures only 69±78% patients now and its use has largely been abandoned in India. Despite several disadvantages, amphotericin B is the only drug available for use in these areas and should be used as rst-line drug instead of Sb v. The new oral antileishmanial drug miltefosine is likely to be the rst-line drug in future. Unfortunately, development of newer antileishmanial drugs is rare; two promising drugs, aminosidine and sitamaquine, may be developed for use in the treatment of VL. Lipid associated amphotericin B has an excellent safety and ef cacy pro le, but remains out of reach for most patients because of its high cost. keywords kala-azar, visceral leishmaniasis, antimony, drug resistance correspondence Dr Shyam Sundar, 6 SK Gupta Nagar, Lanka, Varanasi 221 005, India. E-mail: shyam_vns@satyam.net.in Introduction Of all forms of leishmaniasis, visceral leishmaniasis is the most severe form of the illness, and if untreated is almost always fatal. The global annual burden of VL is estimated to be about 500 000 and 90% of cases occur in India, Sudan, Nepal, Bangladesh and Brazil. More than 100 000 cases of VL occur in India alone every year, and state of Bihar accounts for more than 90% of these (Bora 1999; Sundar et al. 2000a,b). Ever since the discovery of pentavalent antimonials about 60 years ago as therapeutic agent for VL, they have remained the rst-line treatment of choice all over the world. Primary resistance or unresponsiveness occur in approximately 1% of patients; and some cases relapse after initial response (secondary resistance or unresponsiveness; relapse), this occurs occasionally with all antileishmanial drugs like sodium antimony gluconate (Sb v ), pentamidine or amphotericin B. In HIV/VL coinfection, although treatment with antileishmanial drug may result in good response and an initial cure, relapses occur as a rule in most patients irrespective of the drug or regimen used. In India too, Sb v is the rst-line drug to treat kala-azar. Response to relatively small daily doses (600 mg max) for short duration (6±10 days) of Sb v had been excellent until the early 1980 (Peter 1981) when reports of treatment failure appeared, and modi cations for Sb v treatment were suggested to overcome the drug failure (Anonymous 1977; Thakur et al. 1984). The World Health Organisation (WHO) revised its recommendations twice, resulting in an increase in the daily dose (from 10 to 20 mg/kg) and duration (from 6 to 10 days to 20±40 days) (WHO 1984, 1990). ã 2001 Blackwell Science Ltd 849

History of drug resistance First indications of drug resistance came from uncon rmed reports from north Bihar of about 30% patients not responding to the prevailing regimen of Sb v (Peters 1981). These reports were from Muzaffarpur, Samastipur, Vaishali and Sitamarhi, the four districts most affected by the kala-azar epidemic. An expert committee (of the Government of Bihar) recommended that Sb v be used in two 10-day courses with a 10-day interval (Anonymous 1977). In Goraul block of Muzaffarpur district, Aikat et al. (1979) followed these new recommendations and found only 1% patients refractory to Sb v therapy. However, only a few years later, Thakur et al. (1984) randomized patients to receive Sb v 20 mg/kg (maximum 600 mg) either for 20 days or longer in case of partial or delayed response, and showed that 86% of patients were cured in the former group. Surprisingly, the cure rate with 10 mg/kg for 20 days was much lower compared with the earlier results. In the same year, the WHO (1984) expert committee recommended that pentavalent antimony be used in doses of 20 mg/kg up to a maximum of 850 mg for 20 days, and a repetition of similar regimen for 20 days in cases of treatment failures. Four years later Thakur et al. (1988) again reviewed the WHO recommendations, and published a report of a clinical trial in which Sb v at 20 mg/kg (max. 850 mg) for 40 days cured 97% of patients, while 20-day treatment at the same doses cured only 81% of patients. Three years later, the same group reported a further decline in cure rate to 71% after 20 days of treatment at the same doses (Thakur et al. 1991a). Jha et al. (1992) found that even after 30 days of Sb v therapy, only 64% of patients could be cured in the hyperendemic district of Bihar. During 1991±1992, we (Sundar et al. 1995) observed that Sb (20 mg/kg without any upper limit) cured only 60%. Five years later, in a bigger study, 156 patients were randomized in three arms for treatment either with (a) Sb v alone for 30 days, or (b) Sb v plus interferon-c (IFN-c) for 15 days or (c) Sb v plus IFN-c 30 for days. Only 36% patients were cured with Sb v alone, and addition of IFN-c could improve the cure rate to 42 and 49% in groups b and c, respectively (Sundar et al. 1997a). Pentamidine was the rst drug to be used in patients refractory to Sb v, and high cure rates were reported (Jha 1983). But its ef cacy has declined over the years, and now it cures only approximately 70% of patients (Jha et al. 1991; Thakur et al. 1991a,b). Reasons for drug resistance Sb v is freely available in India, and both quali ed medical practitioners and unquali ed quacks use the drug. Unrestricted availability has led to widespread misuse of these compounds. In one study we found that most patients (73%) consulted unquali ed practitioners rst, who might not use the drug appropriately (Sundar et al. 1994). It is a common practice to start with a small dose and gradually build up to the full dose over a week; it was also advocated to have drug free periods to minimize the toxicity, especially renal toxicity. Even now in Bihar, physicians split the daily dose in two injections to be given twice a day. These practices result in build-up of a subtherapeutic blood levels, and increased tolerance of parasites to Sb v. We surveyed 312 patients with Sb v who had received one or more courses of antimony but failed to recover. Only 81 (26%) were treated according to the WHO guidelines. Most patients (72%) received the drug for < 40 days; 87 (28%) for < 20 days and 147 (47%) for < 30 days; 42% did not take the drug regularly and 112 (36%) stopped the drug on their own initiative. Almost half of the patients, receiving pentamidine as a second-line drug, had not received adequate antimony treatment before being labelled as refractory to Sb v. These facts indicate largescale misuse of antileishmanial drugs in Bihar, contributing to development of drug resistance (Sundar et al. 1994). There are several manufacturers of Sb v in India, and not all produce consistent quality products, resulting in occasional batches being substandard and toxic, and adding to the problems associated with Sb v therapy and serious toxicity and deaths related to the drug (Sundar et al. 1998a). Is there true Sb v resistance? There have been speculations (i) whether Indian Leshmania donovani has become truly refractory to Sb v or (ii) whether resistance occurs because of the inadequate doses being used in Bihar, or (iii) whether there are unknown host factors which determine the response to treatment. We tried to address some of these issues in two studies. In the rst, we investigated L. donovani isolates from 15 non-responding and nine responding patients. In vitro macrophage infection were established and ED 50 for Sb v in killing of intramacrophageal amastigotes were determined for each isolate. The mean ED 50 of these nine strains was 2.4 2.6 mg Sb v /ml. The ef cacy pro les of the 15 strains obtained from nonresponsive cases were far more heterogeneous, and showed a mean ED 50 of 7.4 3.7 lg Sb v /ml; similarly the mean ED 90 for isolates from non-responsive cases was ve times higher than that from responsive patients. Thus, the amastigotes from the unresponsive strains were several times more resistant to killing by Sb v than the sensitive ones, and this difference was highly signi cant (Lira et al. 1999). 850 ã 2001 Blackwell Science Ltd

A strong correlation was also observed between the clinical response to Sb v and the in vitro susceptibility of intracellular amastigotes comparing L. donovani isolates from these patients with kala-azar. The temporal as well as geographical concentration of Sb v treatment failures points strongly to the emergence of antimony resistant strains. The threefold increase in the average ED 50 of the clinically unresponsive strains seems relatively slight, and interpreting this modest shift in antimony sensitivity in the in vitro assays, one may conclude that clinical unresponsiveness could be overcome by modest increases in dose or duration of therapy; however, serious and fatal toxicity associated with the current regimen are at the limits of acceptability, and increasing the dose of Sb v any further would seriously jeopardise the safety of the patients. In India there are no zoonotic reservoirs, and only human to human (anthroponotic) transmission occurs; this means that once there is emergence of Sb v refractory parasites, they circulate in the community ef ciently as Sb v sensitive parasites get eliminated by the drug, and the proportion of patients with Sb v refractory parasites rises. In another study in 1994±1997, we formally enrolled patients treated with Sb v at our two treatment sites (i) at Muzaffarpur in the state of Bihar, where Sb v treatment failure had been reported, and (ii) at Varanasi in Uttar Pradesh (UP), where VL endemicity had started almost a decade later, to document the level of Sb v resistance in Bihar, and to determine whether it had spread to the neighbouring Eastern UP. Of 209 and 111 patients treated at Bihar and UP site, respectively, only 35% could be cured at Bihar. Only 184 of 209 could complete the treatment for the prescribed 30 days, and of these, primary unresponsiveness was seen in 52% patients whereas another 8% relapsed after an initial cure. In UP, on the other hand, 100/ 102 (98%) were cured initially and one (1%) relapsed (Sundar et al. 2000a). Thus it is apparent that Sb v continues to be effective in the state of UP, but in North Bihar, where most (approximately 90%) of the disease occurs, it is ineffective in most patients. The magnitude of resistance varies in different areas of Bihar, peaking in the 11 districts as shown in Fig. 1 (unpublished observations), most patients from southern Figure 1 Map of State of Bihar, India and Nepal showing the degree of antimony resistance in various kala-azar endemic areas. ã 2001 Blackwell Science Ltd 851

districts of Bihar (areas south of Ganges; Fig. 1) respond well to Sb v therapy, as do patients from West Bengal or UP. Therapeutic options Because of its low cost, especially in India, its ease of administration and its effectiveness until recently, Sb v has been the backbone of anti-kala-azar therapy. In areas outside north Bihar, it should continue to be the rst-line drug. However, with its declining ef cacy in VL hyperendemic regions of North Bihar, it should no longer be used as rst-line drug in areas with high resistance. What are the options available for these patients? Pentamidine and amphotericin B are the two alternative antileishmanial drugs. Pentamidine has been in use for last two decades, and as discussed above, its dwindling ef cacy coupled with serious toxicities associated with its use, render it unsuitable for use as a rst-line medication. Thus only amphotericin B remains, which induces high cure rates (>97%) at doses of 0.75±1.00 mg/kg for 15 infusions on alternate days. It has been used extensively in Bihar with uniformly good results. Infusion reactions and thrombophlebitis are almost universal with amphotericin B use, and occasionally hypokalaemia, thrombocytopenia, myocarditis, death, etc. might occur. Fortunately, amphotericin B induced nephrotoxicity is extremely rare in patients with VL. Occasional relapses (<1%) might occur with amphotericin B and these patients are successfully re-treated with the same drug (Giri 1994; Giri & Singh 1994; Mishra et al. 1994). However, for its use as a rst-line drug, there are several limitations. It has to be given in infusions and facilities for this are not available at most primary health centres in Bihar. Special amphotericin B treatment centres with trained personnel need to established in these areas; on the other hand continued use of Sb v, although convenient, cheaper and easily applicable, is associated with high incidence of cardiotoxicity and deaths (Sundar et al. 1997a, 1998a, 2000a). Lipid-associated amphotericin (L-AB) preparations (AmBisome and Abelcet) are as effective as conventional amphotericin B, and have negligible adverse reactions. It is possible to administer high doses of L-AB over a short period with high cure rates (Sundar & Murray 1996; Sundar et al. 1997b, 2001), however, their high cost makes these compounds unaffordable in VL-endemic countries. Miltefosine Miltefosine is an alkyl phospholipid also known as hexadecylphosphocholine; it can be used orally and is best suited for use in remote areas. This drug is still being developed, and phase III trials are under completion, and it is likely to be registered soon for the treatment of VL. The recommended duration of treatment is 4 weeks although 3-week treatment is equally effective, and its doses are: for patients weighing ³25 kg ± 100 mg daily and those weighing <25 kg ± 50 mg daily. In most of the studies (Sundar et al. 1998b, 2000b; Jha et al. 1999) at this dose, >95% of patients were cured consistently. Its half-life is approximately 150 h, and steady-state is reached in about 3±4 weeks. Because of its teratogenic potential, it cannot be given to pregnant females and patients should not conceive for at least 2 months after the end of treatment. In the Indian national programme for kala-azar, it has been recommended that after it is registered for the treatment of VL, miltefosine be used as rst line drug in areas with >10% Sb v unresponsiveness. Miltefosine, being an oral drug, can be an important tool in containing the epidemics with its ease of distribution and administration. Its good tolerance with mild side-effects obviates the need for close monitoring. However, its long half-life (approximately 150 h) and teratogenic potential are important drawbacks, as development of drug resistance and foetal malformations (if given to pregnant females) may eventually lead to a premature end of using this very important drug. Other drugs Aminosidine, an aminoglycoside, has been investigated extensively for its antileishmanial effect. In India, a dose of 16 mg/kg intramuscularly for 21 days cured 93% of patients (Jha et al. 1998). Aminosidine could replace Sb v as rst line drug, however, had its production not been stopped by the previous manufacturer. As a result of the efforts of TDR/WHO, a new manufacturer, IDA-Pharmamed, Malta is likely to produce it. But as new trials have to be conducted, its registration and consequent commercial availability are several years away. Sitamaquine (WR6026) is another orally administrable antileishmanial compound which has been on the horizon for several years, but its clinical development has been very slow, and only one phase I/II pilot study has been completed (Sherwood et al. 1994). Combination therapy in VL Growing resistance of the parasite to antileishmanial drugs suggests that the currently used monotherapy needs to be reviewed. Multidrug combination treatment of VL, as practiced in tuberculosis and leprosy, should be given serious thoughts to prevent/delay the appearance of drug resistance. Although at present not many effective antile- 852 ã 2001 Blackwell Science Ltd

ishmanial drugs are available, once oral miltefosine, sitamaquine and parenteral aminosidine become available, these drugs along with amphotericin B and Sb v should be used in combination not only to combat drug resistance but also to shorten the duration of treatment. Thus, the northern districts of Bihar have the distinction of being unique in terms of large-scale Sb v failure; resistance is likely to go up and spread into the areas where Sb v is still effective. Rampant misuse of Sb v (inadequate doses and insuf cient duration) has led to the development of refractory strains which tolerate several times more drug than those still responsive to it. More work is needed to identify the changes occurring in these strains and molecular tools for identi cation of these strains need to be developed. In areas with Sb v resistance, it needs to be replaced with safer and more effective drugs. Unfortunately at the moment despite its limitations, amphotericin B is the only option for these patients. Oral miltefosine is likely to be available in the near future and should then be used as rst line therapy. When more clinically effective antileishmanial drugs are developed, combination chemotherapy may hold the key to curbing the emergence of drug resistance. References Anonymous (1977) Proceedings of the Meeting of an Expert Group on Kala-azar held at Indian Council of Medical Research Headquarters on 9 September, 1977, New Delhi. Indian Council of Medical Research, New Delhi. Bora D (1999) Epidemiology of visceral leishmaniasis in India. National Medical Journal of India 12, 62±68. Giri OP (1994) Treatment of visceral leishmaniasis unresponsive to pentamidine with amphotericin B. Journal of the Association of Physicians of India 42, 988±989. Giri OP & Singh AN (1994) Experience with amphotericin B in sodium stibogluconate-unresponsive cases of visceral leishmaniasis in North Bihar. Journal of Association of Physicians of India 42, 690±691. Jha TK (1983) Evaluation of diamidine compounds (pentamidine isethionate) in the treatment of resistant cases of kala-azar occurring in North Bihar, India. Transactions of the Royal Society of Tropical Medicine and Hygiene 77, 167±170. Jha SN, Singh NKP & Jha TK (1991) Changing response to diamidine compounds in cases of kala-azar unresponsive to antimonials. Journal of the Association of Physicians of India 39, 314±316. Jha TK, Singh NKP & Jha S (1992) Therapeutic use of sodium stibogluconate in kala-azar from some hyperendemic districts of N. Bihar, India. Journal of the Association of Physicians of India 40, 868 (Abstract). Jha TK, Olliaro P, Thakur CP et al. (1998) Randomised controlled trial of aminosidine (paromomycin) v sodium stibogluconate for treating visceral leishmaniasis in North Bihar, India. British Medical Journal 316, 1200±1205. Jha TK, Sundar S, Thakur CP et al. (1999) Miltefosine, an oral agent for the treatment of Indian visceral leishmaniasis. New England Journal of Medicine 341, 1795±1800. Lira R, Sundar S, Makharia A et al. (1999) Evidence that the high incidence of treatment failure in kala-azar is due to the emergence of antimony resistant strains of Leshmania donovani. Journal of Infectious Diseases 180, 564±567. Mishra M, Biswas UK, Jha AM & Khan AB (1994) Amphotericin vs. sodium stibogluconate in rst-line treatment of Indian kala-azar. Lancet 344, 1599±1600. Peters W (1981) The treatment of kala-azar. New approach to an old problem. Indian Journal of Medical Research 73 (Suppl.), 1±18. Sherwood JA, Gachihi GS, Muigai RK et al. (1994) Phase 2 ef cacy trial of an oral 8-aminoquinoline (WR6026) for treatment of visceral leishmaniasis. Clinical Infectious Diseases 19, 1034±1039. Sundar S & Murray HW (1996) Cure of antimony-unresponsive Indian visceral leishmaniasis with amphotericin B lipid complex. Journal of Infectious Diseases 173, 762±765. Sundar S, Thakur BB, Tandon AK et al. (1994) Clinico-epidemiological study of drug resistance in Indian kala-azar. British Medical Journal 308, 307. Sundar S, Rosenkaimer F & Murray HW (1995) Immunochemotherapy for a systemic intracellular infection: accelerated response using interferon-c in visceral leishmaniasis. Journal of Infectious Diseases 171, 992±996. Sundar S, Singh VP, Sharma S, Makharia MK & Murray HW (1997a) Response to interferon-c plus pentavalent antimony in Indian visceral leishmaniasis. Journal of Infectious Diseases 176, 1117±1119. Sundar S, Agrawal NK, Sinha PR, Horwith G & Murray HW (1997b) Short-course, low dose amphotericin B lipid complex therapy for visceral leishmaniasis unresponsive to antimony. Annals of Internal Medicine 127, 133±137. Sundar S, Sinha PR, Agrawal NK et al. (1998a) A cluster of severe antimony cardiotoxicity in kala-azar due to high osmolality lot. American Journal of Tropical Medicine and Hygiene 59, 139±143. Sundar S, Rosenkaimer F, Makharia MK et al. (1998b) Trial of oral miltefosine treatment for visceral leishmaniasis. Lancet 352, 1821±1824. Sundar S, More DK, Singh MK et al. (2000a) Failure of pentavalent antimony in visceral leishmaniasis in India: report from the center of the Indian epidemic. Clinical Infectious Diseases 31, 1104±1107. Sundar S, Makharia A, More DK et al. (2000b) Short-course oral miltefosine treatment for visceral leishmaniasis. Clinical Infectious Diseases 31, 1110±1113. Sundar S, Jha TK, Thakur CP et al. (2001) Low dose liposomal amphotericin B (AmBisome) in refractory Indian visceral leishmaniasis ± a multicenter study. American Journal of Tropical Medicine and Hygiene (in press). ã 2001 Blackwell Science Ltd 853

Thakur CP, Kumar M, Singh SK et al. (1984) Comparison of regimens of treatment with sodium stibogluconate in kala-azar. British Medical Journal 288, 895±897. Thakur CP, Kumar M, Kumar P, Mishra BN & Pandey AK (1988) Rationalisation of regimens of treatment of kala-azar with sodium stibogluconate in India: a randomized study. British Medical Journal 26, 21±25. Thakur CP, Kumar M & Pandey AK (1991a) Evaluation of ef cacy of longer duration of therapy of fresh cases of kala-azar with sodium stibogluconate. Indian Journal of Medical Research 93, 103±110. Thakur CP, Kumar M & Pandey AK (1991b) Comparison of regimens of treatment of antimony-resistant kala-azar patients: a randomized study. American Journal of Tropical Medicine and Hygiene 45, 435±441. World Health Organization (1984) The leishmaniases: report of a WHO Expert Committee. WHO Technical Report Series 701, 99±108. World Health Organization (1990) The control of leishmaniases: report of an expert committee. WHO Technical Report Series 793, 50±55. 854 ã 2001 Blackwell Science Ltd