JAC Advance Access originally published online on August 17, 2006
Journal of Antimicrobial Chemotherapy 2006 58(4):811-815; doi:10.1093/jac/dkl342
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
A comparison of liposomal amphotericin B with sodium stibogluconate for the treatment of visceral leishmaniasis in pregnancy in Sudan
1 Médecins sans Frontières, 67-74 Saffron Hill London EC1N 8QX, UK 2 Department of Infection Tropical Medicine, Lister Unit, Northwick Park Hospital, Harrow Middlesex HA1 3UJ, UK
*Corresponding author. Tel: +44-20-7067-4235; Fax: +44-20-7404-4466; E-mail: manica.balasegaram{at}london.msf.org
Received 31 October 2005; returned 27 March 2006; revised 4 May 2006; accepted 29 July 2006
 |
Abstract |
|---|
 Top Abstract IntroductionPatients and methodsResultsDiscussionTransparency declarationReferences |
|---|
Objectives: Little is known about the treatment of visceral leishmaniasis (VL) in pregnancy, especially in resource-poor settings. We present a series of pregnant women with VL treated with either sodium stibogluconate or liposomal amphotericin B (AmBisome), or both, in eastern Sudan over 16 months.
Methods: We did a retrospective analysis of all pregnant VL patients treated in the Médecins sans Frontières (MSF) Um el Kher centre between January 2004 and April 2005. We diagnosed VL with laboratory confirmation of clinical suspects, and recorded the outcomes of treatment for pregnant women and their foetuses. We carried out a manual search of relevant publications and a systematic search of the literature in the MEDLINE database.
Results: We treated 23 women with sodium stibogluconate, 4 with AmBisome and sodium stibogluconate and 12 with AmBisome alone. There were 13 (57%) spontaneous abortions in the sodium stibogluconate monotherapy group, and none in either of the other two groups. All spontaneous abortions occurred in the first two trimesters. All patients, except one in the sodium stibogluconate group who defaulted, were discharged as cured in good clinical condition.
Conclusions: AmBisome treatment for VL appears to be safe and effective for pregnant women and their foetuses. We recommend the use of AmBisome as first-line treatment for these patients.
Keywords: kala-azar , neglected disease , AmBisome
|
Introduction |
|---|
|
TopAbstractIntroductionPatients and methodsResultsDiscussionTransparency declarationReferences |
|---|
Visceral leishmaniasis (VL), caused by Leishmania donovani, is endemic in Gedaref state in eastern Sudan. The disease is characterized by chronic fever, weakness and weight loss associated with (hepato)-splenomegaly, small volume lymphadenopathy and anaemia. Co-infection with diseases such as malaria or pneumonia is common. Information about VL in pregnancy is limited, but it is known to be life-threatening to both the mother and the foetus.
Médecins sans Frontières (MSF) was based in the Um el Kher treatment centre for VL in Gedaref state, a remote centre with mainly temporary structures made from mud and grass, without good roads, and without electricity. Transport in the area is extremely limited and most patients take several hours or even days to reach the centre. The poor clinical condition of pregnant VL patients on admission necessitates immediate treatment.1 In Sudan, two drugs are available to treat VL: sodium stibogluconate (Albert David Ltd, Calcutta, supplied by International Dispensary Association, Amsterdam) and liposomal amphotericin B (AmBisome; Gilead, USA). Little has been published on the effects of these drugs in pregnant women with VL.
We aimed to evaluate the efficacy and safety of sodium stibogluconate and AmBisome alone, and in combination, for the mother and the foetus. We did a retrospective analysis of all pregnant VL patients treated in the Um el Kher centre over a 16 month period, supplemented with a literature review.
|
Patients and methods |
|---|
|
TopAbstractIntroductionPatients and methodsResultsDiscussionTransparency declarationReferences |
|---|
Diagnosis of VL was by clinical examination with laboratory confirmation of clinical suspects. Confirmed cases were admitted to the centre for treatment as in-patients. Clinical diagnosis included fever of >2 weeks duration and either splenomegaly or at least two of the following: inguinal lymphadenopathy, hepatomegaly, weight loss, recurrent epistaxis or anaemia.
Laboratory confirmation was by direct agglutination test (DAT; 
6400), rapid diagnostic dipstick test (rK39; DiaMed-IT-Leish®) or lymph-node aspiration.2 DAT was the principal test used, though use of the other two tests was dictated by availability and urgency of diagnosis. Splenic aspiration is currently not permitted by the Sudanese Ministry of Health in rural health centres. Other routine investigations included haemoglobin and malaria screening (rapid diagnostic test Paracheck® or blood film). Services for voluntary counselling and testing for HIV were available from April 2004, but testing was not done routinely. For reasons of confidentiality, test results were not linked with patients' notes.
Pregnancy was confirmed by urine dipstick testing for HCG. Trimester was from the date of the last menstrual period and clinical examination for fundal height. No ultrasound was available.
Until September 2004, we treated pregnant VL patients with either sodium stibogluconate (20 mg/kg intramuscularly once daily for 30 days) or with a combination of AmBisome and sodium stibogluconate. The regimen was four doses of AmBisome 3–7 mg/kg daily on days 1, 6, 11 and 16, followed by 20 mg/kg sodium stibogluconate intramuscularly once daily for 30 days. We reserved the latter regimen for patients with clinically severe VL, namely prostration, anaemia and severe malnutrition. From October 2004, we changed our protocol and treated all pregnant patients with AmBisome 3–7 mg/kg on days 1, 2, 3, 5, 10 and 15. All patients received a standard supportive treatment of vitamin A, folic acid, ferrous sulphate, multivitamins and nutritional support. Antibiotics or antimalarials were added if indicated. During treatment and at discharge, patients were followed up clinically, focusing especially on weight gain and clearance of fever. A patient was considered cured if symptoms, in particular, fever, resolved and she did not re-present with VL to the centre in the 3 months following treatment. We recorded the outcome of treatment for both pregnant women and foetuses. Because of scarce resources and adverse road and weather conditions, we did not actively follow up patients who were discharged still pregnant or with newborn babies.
We analysed our data using Fisher's exact test, Student's paired t-test and the unpaired t-test. A manual search of relevant publications and a systematic search of the literature in the MEDLINE database using the string ‘kala azar, visceral leishmaniasis and pregnancy’ was carried out, and compared with our results.
|
Results |
|---|
|
TopAbstractIntroductionPatients and methodsResultsDiscussionTransparency declarationReferences |
|---|
Between January 2004 and April 2005, we treated 3267 patients for VL. The clinical condition of patients was generally poor on admission. Despite immediate treatment, 159 (5%) patients died.
One thousand two hundred and eighty-eight (39%) patients were female. Three hundred and eighty women were of child-bearing age (15–45 years); 39 (10%) were pregnant. All 39 pregnant VL patients were from Gedaref or neighbouring Sennar states. Table 1 shows their characteristics on admission. Thirteen (33%) women were in the first trimester, 16 (41%) in the second and 10 (26%) in the last. For 14 patients the history of previous pregnancies was recorded. Unfortunately the information was not documented in the case records for the remaining patients: three were in their second pregnancy, five in their fourth, one in her fifth, two in their sixth, one in her ninth, one in her eleventh and one in her fourteenth. One patient had been treated for VL in the centre in 2003 with a 30 day course of sodium stibogluconate.
|
On admission, 72% (28 of 39) of pregnant patients had a non-palpable spleen, compared with 43% (1401 of 3228) of patients who were not pregnant. Other clinical signs on admission did not differ between these groups (data available on request).
Table 2 shows the outcome of the patients and Table 3 the outcome of their pregnancies. In all three groups, patients gained weight significantly (Student's t-test: P 
0.01) by the end of treatment. Weight gain was lowest in the sodium stibogluconate group (Table 2). Weight gained on AmBisome alone was significantly greater than that on sodium stibogluconate (unpaired t-test: P < 0.05). This difference persisted when adjusted for patients who had a spontaneous abortion or who gave birth (Table 2; P = 0.03). Patients' temperatures in all treatment groups normalized between 24 and 96 h after starting treatment, confirming the diagnosis of VL. All patients, excepting one patient who defaulted, were discharged in a good clinical condition. One patient treated with AmBisome gave birth to a premature baby who died after 1 week of life. The mother was reported to have died 2 weeks after discharge in another hospital, because of an unknown cause.
|
|
There were 13 spontaneous abortions during treatment: all were among patients treated with sodium stibogluconate during the first and second trimesters (Table 3). Spontaneous abortions occurred between days 13 and 30 of treatment. Two were accompanied with acute onset of fever. None of the patients treated by AmBisome or combination therapy had an abortion during treatment. The risk of spontaneous abortion in the first two trimesters was significantly higher (P < 0.002; Fisher's exact test) for patients treated with sodium stibogluconate than with AmBisome. Malaria did not appear to increase the risk of spontaneous abortion. Of the four patients treated with sodium stibogluconate who tested positive for malaria, two were still pregnant on discharge, one had a healthy child and one spontaneously aborted before sodium stibogluconate treatment was started.
HIV status was known in only 17 of the 39 patients [7 (58%) in the AmBisome group, 1 (25%) in the combination group, 9 (40%) in the sodium stibogluconate group]. Of these, none were HIV positive.
|
Discussion |
|---|
|
TopAbstractIntroductionPatients and methodsResultsDiscussionTransparency declarationReferences |
|---|
This series of 39 cases of VL in pregnancy is the largest ever reported. In our literature search (English language only) we found 22 such cases reported over the last 50 years.1,3–13 Patients came from India (eight cases), Sudan (four cases), Italy (eight cases), France (one case) and Brazil (one case). The largest case series was five (two studies).3,8 Seven were diagnosed with VL in their first trimester, 12 in their second and 3 in their third trimester. Sixteen of the 22 mothers received a documented treatment during their pregnancies. Six (five in India and one in Brazil) were treated with amphotericin B, seven (from the Mediterranean area infected with Leishmania infantum) with liposomal amphotericin B (AmBisome), two (in Italy) with meglumine antimoniate and one (in Sudan) with pentavalant antimony.3–5,7–9 Fifteen mothers showed no signs of infection after treatment and delivered healthy babies. Only the Sudanese patient, treated with pentavalent antimony, had a reported relapse of VL.9 Her small-for-dates neonate was also diagnosed with VL.
Of the other six patients, two did not have their treatment documented, while four were treated after delivery.6,9–13 Most of the latter were case reports focusing on newborns with congenital VL. The outcomes of the mothers were not reported. The delay of therapy till after delivery resulted in two cases of congenital VL, one spontaneous abortion and one small-for-dates. These few cases demonstrate the risks of deferring treatment in VL.
Our results indicate that AmBisome is safer than sodium stibogluconate for treatment during pregnancy. All spontaneous abortions occurred during sodium stibogluconate treatment when the symptoms of VL were subsiding, and most were not accompanied by clinical manifestations of the infection. We suspect that the spontaneous abortions in the first and second trimesters might have been induced by sodium stibogluconate and not by VL or by co-morbidity such as malaria. The use of sodium stibogluconate in the third trimester did not cause spontaneous abortion. Amphotericin is a potent drug that rapidly kills the leishmania parasite, and this in itself may protect the pregnancy.14 Our findings support this presumption.
The response of pregnant patients in all treatment groups was excellent. However, patients treated with AmBisome alone gained significantly more weight than those on sodium stibogluconate. Patients in the combination group were treated for longer, and thus received the most nutritional support, than patients in the other two groups, which could explain why this group gained the most weight. Differences in stage of pregnancy may also have accounted for different rates of weight gain. Weight is an important indicator of response to treatment, and the disparity in weight gain between the sodium stibogluconate and AmBisome groups was even greater after exclusion of patients who delivered or spontaneously aborted; thus, we believe this difference and the poor weight gain in the sodium stibogluconate group still to be of importance.
All patients were discharged in a good condition and no relapses re-presented to the treatment centre in the 3 months after discharge. However, a major limitation in our analysis is that we did not follow up patients after discharge; hence we do not know the long-term effects of treatment.
Another limitation in our study is the comparability of the groups; 11 (48%) patients in the sodium stibogluconate received treatment in the first trimester as opposed to only 2 (17%) in the AmBisome group. As this is a time when women are likely to have a spontaneous abortion, this may be a confounding factor. For this reason we also analysed outcomes in the second trimester. The fact that many spontaneous abortions (five out of eight treated) also occurred in the second trimester in the sodium stibogluconate group, while none occurred in the six AmBisome patients treated, implies to us that spontaneous abortions in the first trimester may have been induced by sodium stibogluconate as well.
A further consideration was that HIV status was known in only 44% of the 39 patients, and hence its impact on our case series is unclear. This was due to the late date of introduction (April 2004) of services for voluntary counselling and testing for HIV (VCT), non-linkage of test results to the notes (consent was required from the patients) as well as the level of uptake of VCT services. Little epidemiological information is known about the HIV prevalence in Gedaref state. The Gedaref state Ministry of health report a 1.7% HIV-positive test rate among potential blood donors in Gedaref hospital. In our treatment centre in Um el Kher, only 1% of our blood donors (1 out of 105 results known to us) and 10% of our adult kala-azar patients from VCT (41 out of 418 tested between May 2004 and April 2005) were positive for HIV. The fact that none out of the 17 known results were HIV positive as well as the overall high cure rates in this case series also makes it unlikely that there was a high prevalence in this group. Nonetheless, we cannot exclude an influence of HIV on our results.
A Th-1 immune response induced by interleukin 2, interleukin 12 and interferon-gamma appears to be crucial in resisting infection with L. donovani.15 Discussion in the literature on immunological changes in pregnancy mention depression of the cell-mediated immunity and a switch to a Th-2-dominant response, and to more recently, a ‘paradigm of Th1–Th2 cooperation’.16–20 It is unclear whether there is a real increased risk of developing clinical VL in pregnancy after exposure to the parasite. Of the women of child-bearing age admitted to our treatment centre, 10% were pregnant. WHO estimates the total fertility rate for Sudan to be 4.3 (2002), meaning that around 11% of Sudanese women of child-bearing age will be pregnant at any given time.21 However, owing to the small size of the study and the fact that we did not determine differences in exposure to infection in pregnant and non-pregnant women, we still cannot comment on the risk of developing VL during pregnancy, nor can we comment on the effect of VL on fertility.
None of our 14 patients for whom the previous history of pregnancies was known was primiparous. Therefore, unlike in malaria, the risk of patients developing VL during first pregnancy does not seem to be higher than during subsequent pregnancies.22 This contrasts with the findings of others who note more primigravidae with VL.3 However, if primiparous patients spontaneously aborted early in pregnancy, or died, they would have been under-represented in our cases. From our study there are insufficient epidemiological data to clarify the risks of VL in the primi or multiparous.
We observed few pregnant VL patients with splenomegaly, a feature not previously reported. All patients had had a fever for longer than 1 week and had other signs of VL: low body mass index (BMI), a confirmatory laboratory test (usually serology) and favourable response to treatment. Because the spleen enlarges slowly during VL, this could be an indication that pregnant women attended our treatment centre earlier than non-pregnant patients. Other explanations might be that clinical palpation of the spleen in pregnant women may be difficult, that pregnancy alters the phenomenon of splenomegaly or that some women had false-positive VL test results. Serological diagnosis through DAT and rK39 has a lower specificity and positive predictive value than direct parasitological confirmation. Therefore, in spite of the fact that patients had other signs of VL and clinically responded to treatment, our inability to do a relatively sensitive and specific test like splenic aspiration remains a further limitation to our results.
Sodium stibogluconate, a pentavalent antimonial, has been recommended for pregnant patients in endemic areas and for Mediterranean VL.7,12 However, antimony-containing drugs have been shown to have reproductive toxicity and mutagenic and oncogenic potential, though others have assessed these risks to be low.14,23,24 Amphotericin B has also been safely used for the treatment of VL in a small cohort of patients in India.8
AmBisome is non-mutagenic in rats, but reproductive toxicity studies were not conducted in pregnant humans.25 AmBisome has been shown to be reasonably safe in the mother and the foetus in mycotic infections.26 The US Food and Drug Administration considers AmBisome to be in pregnancy category B: animal reproduction studies have failed to demonstrate a risk to the foetus but there are no adequate and well-controlled studies in pregnant women. No category has been assigned to sodium stibogluconate.
We conclude that the risk of spontaneous abortion with sodium stibogluconate treatment for VL during pregnancy is significantly greater than with AmBisome. We recommend the use of AmBisome in pregnant women with VL. Routine pregnancy screening (clinical and laboratory) of all women of child-bearing age should be undertaken on admission for treatment of VL. Special care and observation is warranted particularly in the first two trimesters. We would not recommend delaying treatment at any stage of pregnancy, owing to risks of the disease to both the mother and the foetus.
|
Transparency declaration |
|---|
|
TopAbstractIntroductionPatients and methodsResultsDiscussionTransparency declarationReferences |
|---|
The authors declare that they have no conflicts of interest.
|
Acknowledgements |
|---|
No specific funding was obtained for this study. We undertook the work described in the course of our roles in patients' care while working for Médecins sans Frontières.
|
References |
|---|
|
TopAbstractIntroductionPatients and methodsResultsDiscussionTransparency declarationReferences |
|---|
1 Figueiro-Filho EA, Duarte G, El-Beitune P, et al. (2004) Visceral leishmaniasis (kala-azar) and pregnancy. Infect Dis Obstet Gynecol 12:31–40.[CrossRef][Medline]
2
Ritmeijer K, Melaku Y, Mueller M, et al. (2006) Evaluation of a new recombinant K39 rapid diagnostic test in Sudanese visceral leishmaniasis. Am J Trop Med Hyg 74: pp. 76–80.
3
Pagliano P, Carannante N, Rossi M, et al. (2005) Visceral leishmaniasis in pregnancy: a case series and a systematic review of the literature. J Antimicrob Chemother 55:229–33.
4 Gradoni L, Gaeta GB, Pellizzer G, et al. (1994) Mediterranean visceral leishmaniasis in pregnancy. Scand J Infect Dis 26:627–9.[Web of Science][Medline]
5 Caldas AJ, Costa JM, Gama ME, et al. (2003) Visceral leishmaniasis in pregnancy: a case report. Acta Trop 88:39–43.[CrossRef][Web of Science][Medline]
6 Kumar A, Mittal M, Prasad S. (2001) Treatment of leishmaniasis in pregnancy. Int J Gynaecol Obstet 72:189–90.[CrossRef][Medline]
7 Utili R, Rambaldi A, Tripodi MF, et al. (1995) Visceral leishmaniasis during pregnancy treated with meglumine antimoniate. Infection 23:182–3.[CrossRef][Web of Science][Medline]
8 Thakur CP, Sinha GP, Sharma V, et al. (1993) The treatment of kala-azar during pregnancy. Natl Med J India 6:263–5.[Medline]
9
Eltoum IA, Zijlstra EE, Ali MS, et al. (1992) Congenital kala-azar and leishmaniasis in the placenta. Am J Trop Med Hyg 46:57–62.
10 Mittal V, Sehgal S, Yadav TP, et al. (1987) Congenital transmission of kala-azar. J Commun Dis 19:184–5.[Medline]
11 Elamin A and Omer MI. (1992) Visceral leishmaniasis in a 6-week-old infant: possible congenital transmission. Trop Doct 22:133–5.[Web of Science][Medline]
12 Nyakundi PM, Muigai R, Were JB, et al. (1988) Congenital visceral leishmaniasis: case report. Trans R Soc Trop Med Hyg 82:564.[CrossRef][Web of Science][Medline]
13 Yadav TP, Gupta H, Satteya U, et al. (1989) Congenital kala-azar. Ann Trop Med Parasitol 835:535–7.
14 Leonard A and Gerber GB. (1996) Mutagenicity, carcinogenicity and teratogenicity of antimony compounds. Mutat Res 366:1–8.[Web of Science][Medline]
15 Goto H and Lindoso JA. (2004) Immunity and immunosuppression in experimental visceral leishmaniasis. Braz J Med Biol Res 37:615–23.[Web of Science][Medline]
16 Hegde UC and Nainan R. (1998) T helper-2 type of interleukins with de novo germinal center formation in the spleen during murine pregnancy. Am J Reprod Immunol 40:424–30.
17 Zhang QH, Huang YH, Hu YZ, et al. (2004) Predominant Th2-type response during normal pregnancy of rats. Sheng Li Xue Bao 56:258–62.[Medline]
18 Wegman TG, Lin H, Guibert L, et al. (1993) Bidirectional cytokine interactions in the maternal-fetal relationship: is successful pregnancy a TH-2 phenomenon? Immunol Today 14:353–6.[CrossRef][Web of Science][Medline]
19 Mellor AL and Munn DH. (2000) Immunology at the maternal-fetal interface: lessons for T cell tolerance and suppression. Annu Rev Immunol 18:367–91.[CrossRef][Web of Science][Medline]
20 Wilczynski JR. (2005) Th1/Th2 cytokine balance—yin and yang of reproductive immunology. Eur J Obstet Gynaecol Reprod Biol 122:136–43.[CrossRef][Web of Science][Medline]
21 World Health Organization. (2004) The world health report 2004—changing history. (World Health Organization, Geneva).
22 Okoko BJ, Enwere G, Ota MO. (2003) The epidemiology and consequences of maternal malaria: a review of immunological basis. Acta Trop 87:193–205.[CrossRef][Web of Science][Medline]
23 Paumgartten FJ and Chahoud I. (2001) Embryotoxicity of meglumine antimoniate in the rat. Reprod Toxicol 15:327–31.[CrossRef][Web of Science][Medline]
24 Winship KA. (1987) Toxicity of antimony and its compounds. Adverse Drug React Acute Poisoning Rev 6:67–90.[Medline]
25 Gilead Sciences International Limited. (1998) Summary of product characteristics [AmBisome]. (Gilead Sciences International Limited, Cambridge) (revised 2005).
26 King CT, Rogers PD, Cleary JD, et al. (1998) Antifungal therapy during pregnancy. Clin Infect Dis 27:1151–60.[Web of Science][Medline]
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||