|
|
| |||||||||||||||||||||||||||||||||||||||||||||
Background Intestinal microsporidiosis due to Enterocytozoon bieneusi is a cause of chronic diarrhea, malabsorption, and wasting in immunocompromised patients. Currently, there is no effective treatment.
Methods We conducted a randomized, double-blind, placebo-controlled trial of fumagillin (60 mg per day orally for two weeks) in patients with chronic E. bieneusi infection. Efficacy was assessed primarily by the clearance of microsporidia, as evidenced by analysis of stool specimens. Patients in whom microsporidia were not cleared received treatment for two weeks with open-label fumagillin. After clearance of the parasite, follow-up stool examinations were performed monthly to detect relapses.
Results Twelve patients were enrolled in this study, 10 with the acquired immunodeficiency syndrome and 2 who had received organ transplants. Clearance of microsporidia occurred in all six of the patients in the fumagillin group, as compared with none of the six in the placebo group (P=0.002). Treatment with fumagillin was also associated with increases in absorption of D-xylose (P=0.003) and in Karnofsky performance scores (P<0.001) and with decreases in loperamide use (P=0.01) and total stool weight (P=0.04). There were serious adverse events (neutropenia and thrombocytopenia) in three patients in the fumagillin group; one patient in the placebo group had severe diarrhea. All six controls subsequently had clearance of microsporidia after open-label treatment with fumagillin. Relapses of the infection were identified in two patients during follow-up (median follow-up, 10 months).
Conclusions Fumagillin is an effective treatment for chronic E. bieneusi infection in immunocompromised patients.
Fumagillin, an antibiotic derived from the fungus Aspergillus fumigatus, has been used to treat microsporidiosis in honeybees,16 and it has been effective in vitro against microsporidia.17,18,19,20 In humans, fumagillin was used more than 40 years ago for the treatment of intestinal amebiasis,21,22 and it is effective when used topically in the treatment of microsporidial keratoconjunctivitis.23,24,25
In preliminary studies, oral fumagillin has had efficacy in HIV-infected patients with E. bieneusi infection.26,27 We undertook this study in order to assess the efficacy of fumagillin in the treatment of intestinal microsporidiosis in immunocompromised patients.
Methods
Study Population
The study was approved by the ethics committee of the Saint-Louis Hospital in Paris. Enrollment began in June 1999, and written informed consent was obtained from all patients.
All participants were 18 years old or older and were immunocompromised because of infection with HIV type 1 (HIV-1) or the transplantation of bone marrow or an organ. Patients had chronic intestinal microsporidiosis, defined by the presence of microsporidia on two consecutive stool examinations two weeks before randomization. Patients were excluded if they had begun receiving a new antiretroviral agent or if their immunosuppressive regimen had been changed less than two months before enrollment. Other criteria for exclusion were concomitant treatment with antiarrhythmic agents or drugs that might increase the risk of thrombocytopenia or bleeding (e.g., aspirin or chemotherapeutic regimens), might affect the metabolism or absorption of fumagillin (e.g., rifampin, rifabutin, phenobarbital, cimetidine, phenytoin, or antacids), or might have activity against microsporidia (e.g., azole compounds). Patients were also excluded if they had any of the following laboratory abnormalities: a platelet count of less than 150,000 per cubic millimeter, an absolute neutrophil count of less than 1000 per cubic millimeter, a hemoglobin level below 8.0 g per deciliter, a serum creatinine concentration of less than 170 µmol per liter, serum levels of aminotransferases and alkaline phosphatase more than three times the upper limit of normal, or a serum lipase level more than two times the upper limit of normal.
Diagnosis of E. bieneusi Infection
All stool samples were analyzed in a central laboratory. E. bieneusi infection was suggested by the identification of typical spores in stools with the use of two different staining methods, and the identification of the species was confirmed by PCR, as previously reported.9,10,11 For a semiquantitative assessment of the number of spores per stool sample, the scoring system was as follows: 0 indicated no spores; 1 indicated rare spores; 2 indicated some spores; and 3 indicated numerous spores.27
Study Design and Treatment
The first phase of the study was a double-blind, placebo-controlled, randomized trial. Patients were randomly assigned in a 1:1 ratio to receive either fumagillin (20 mg three times daily on an empty stomach) or placebo orally for two weeks. Fumagillin and placebo were provided by Sanofi-Synthelabo Laboratories. After this double-blind phase (two weeks after the end of treatment [week 4]), patients with microsporidial spores in their stools received open-label fumagillin (20 mg three times a day) for an additional two weeks.
Patients whose stools no longer contained spores were then followed with monthly stool examinations, until parasitologic relapse or death, for a maximum of 12 months. Concomitant medications, including prophylaxis against opportunistic infections, were allowed. Antiretroviral therapy and immunosuppressive regimens were also allowed, provided that they were not modified during the placebo-controlled phase of the trial. Patients were also allowed to use antidiarrheal medications and to adjust their own doses of such medications.
Treatment and Follow-up
Base-line investigations included a complete medical history and physical examination, a complete blood count, urinalysis, biochemical analysis of serum, bacterial stool cultures, examination of stool for ova and parasites (including cryptosporidia), and assays for the Clostridium difficile toxin.
Patients were also assessed one, two, and four weeks after the initiation of fumagillin or placebo. Clinical and biologic measurements were recorded at each visit. Patients were asked to use a diary to record the number of bowel movements, the characteristics of their stools, the use of antidiarrheal medications, and any adverse events. The number and characteristics of bowel movements and the total weight of stools during two consecutive days before the visits on day 1, at week 2, and at week 4 were recorded. In addition, complete blood counts were obtained on days 10, 12, 17, and 19. The results of platelet counts, however, were not given to the investigators until week 4, unless the count decreased below 75,000 per cubic millimeter, resulting in the interruption of the study treatment. This procedure was used in order to keep the investigators unaware of the treatment-group assignment, since fumagillin had induced frequent thrombocytopenia in previous trials.26,27 At base line and at week 4, the blood D-xylose concentration was measured in a central laboratory 90 minutes after the oral administration of 25 g of D-xylose in order to assess intestinal absorption.4 Examinations of stool specimens for microsporidia were performed on days 1, 8, 15, 17, and 19. Similar follow-up was performed during the open-label phase of the trial, except that the investigators were aware of the results of stool examinations and platelet counts.
Efficacy End Points
Clearance of parasites, which was the primary efficacy end point, was defined as the absence of microsporidial spores in stools 15 and 17 days after either double-blind or open-label treatment. If there were conflicting results, a third stool sample was obtained on day 19 and analyzed. Laboratory staff members who analyzed the stool samples were unaware of the treatment-group assignment.
Secondary end points included semiquantitative assessments of the number of spores in stools and changes in body weight, stool weight, stool frequency and consistency, blood D-xylose levels, and use of antidiarrheal medications during the study period. All assessments were performed by physicians who were unaware of the treatment-group assignment, the results of stool examinations, and the platelet counts until week 4 of the double-blind phase. Parasitologic relapse was defined by the reemergence of spores in follow-up stool samples.
Safety End Points
All adverse events were recorded, regardless of their relation to the study drugs. Toxic effects were graded according to the scoring system of the World Health Organization (WHO). A serious adverse event was defined as any event not related to HIV infection or organ transplantation that was fatal or life-threatening, resulted in hospital admission or prolonged hospitalization, or caused permanent disability, cancer, or congenital anomaly; any clinical or biologic event of WHO grade 3 or 4; or overdose of the study drug.
Statistical Analysis
All analyses were performed according to the intention-to-treat principle. Analysis of the primary efficacy end point involved the comparison by Fisher's exact test of the percentages of patients in the two treatment groups in whom clearance of the parasites occurred. In the analysis of secondary end points, Wilcoxon signed-rank tests for paired nonparametric data were used for numerical variables and Fisher's exact tests were used for categorical variables. Analysis of variance and logistic-regression models were also used to assess the effect of randomization on secondary end points, while incorporating the effects of time. We used Kaplan–Meier analysis to estimate the time to relapse on the basis of the results of the monthly stool examinations.
We determined that we needed a sample of 10 patients per group, assuming an estimated rate of clearance of parasites of 80 percent with fumagillin and 10 percent with placebo, with alpha and beta levels of 5 percent and two-tailed tests. We planned to conduct the first interim analysis of efficacy and safety after 10 patients had completed the trial. On October 16, 2000, the data and safety monitoring board reviewed the data and recommended that the enrollment of patients be discontinued. The decision was endorsed by the scientific committee for the trial.
All P values reported are two-sided. A P value of less than 0.02 was considered to indicate statistical significance for the primary end points, because of the one interim analysis that had been planned and conducted before stopping the study. Data are reported as medians and interquartile ranges, unless otherwise indicated. Statistical analysis was performed with the use of SAS software, version 8 (SAS Institute). The study was designed and completed by the investigators of the Agence Nationale de Recherches sur le SIDA 090 study, who held the data and performed all statistical analyses. The sole role of Sanofi-Synthelabo was to provide active drug and placebo for the trial.
Results
Study Patients
Between June 1999 and November 2000, 19 patients were screened and 12 underwent randomization. Reasons for not undergoing randomization included death before randomization (in the case of one patient), thrombocytopenia (three patients), alterations in the results of liver-function tests (two patients), and absence of microsporidia in stools (one patient).
The base-line characteristics of the 12 patients who underwent randomization are summarized in Table 1. Of these patients, 10 had HIV infection and had a median CD4 count of 16 per cubic millimeter (range, 4 to 99 per cubic millimeter). Three of these patients had not received antiretroviral medication. All were homosexual men with AIDS and had had multiple opportunistic diseases. The two remaining patients had undergone organ transplantation; one had received a kidney transplant 39 months before enrollment and was receiving 2 mg of tacrolimus and 10 mg of prednisone per day, and the other had received a liver transplant 21 months before enrollment and was receiving 2 mg of tacrolimus per day.
|
The treatment groups were well balanced in terms of the base-line characteristics of the patients. At enrollment, all patients reported diarrhea, bloating, and mild abdominal discomfort. All patients had passed at least one liquid stool in the two days before the visit on day 1, and the total weight of stools was more than 300 g per day in all patients (data not shown). Associated intestinal pathogens (entamoeba species) were found in only one patient. D-Xylose–absorption tests were abnormal at base line in all but one patient (who was assigned to the placebo group).
Primary End Point
The patients were followed as required by the study protocol. No patient was lost to follow-up, and none prematurely stopped the study treatment. The shedding of E. bieneusi spores in stools ceased in all of the six patients in the fumagillin group, as compared with none in the placebo group (100 percent vs. 0 percent, P=0.002 by Fisher's exact test) (Table 2).
|
Microsporidial clearance was associated with a significant increase in D-xylose absorption in the fumagillin group but not in the placebo group (P=0.003), with normal D-xylose–absorption tests in five of the six patients in the fumagillin group at week 4 but in none of the six patients in the placebo group (P<0.02) (Table 2).
Microsporidial clearance in the fumagillin group was also associated with clinical efficacy. Indeed, there was a trend toward lower median total weight of stools in the fumagillin group than in the placebo group (P=0.04). At week 4, no patient in the fumagillin group was using loperamide, whereas four of the six patients in the placebo group still required this antimotility drug (P=0.009). There was also a trend toward more solid stools in the fumagillin group than in the placebo group (P=0.06). However, there were no clear changes noted in the number of bowel movements per day or in body weight. There was a significant increase in Karnofsky scores in the fumagillin group, whereas there was a decrease in the placebo group (P<0.001). New or relapsing opportunistic infections were recorded in three patients during the double-blind phase (pulmonary aspergillosis in one patient in the fumagillin group, disseminated histoplasmosis in one patient in the placebo group, and cytomegalovirus disease in one patient in the placebo group).
Open-Label Fumagillin and Follow-up
After the double-blind phase of the study, open-label fumagillin was offered to the six patients in the placebo group, in whom clearance of microsporidia had not occurred. Three patients prematurely discontinued the study treatment (one each 11, 13, and 14 days after open-label treatment began) because of thrombocytopenia. In all patients, however, follow-up stool specimens did not contain E. bieneusi. The overall cure rate associated with the fumagillin regimen in our study was therefore 100 percent (95 percent confidence interval, 76 to 100 percent).
The median duration of follow-up after fumagillin therapy was 10 months (range, 5 to 12). One patient died from end-stage HIV disease five months after the end of treatment without having had a relapse. His CD4 count was only 10 cells per cubic millimeter. Two parasitologic and clinical relapses were recorded one month after treatment was completed in the HIV-infected patients. At day 40 of follow-up, the estimated rate of relapse was 16.7 percent (95 percent confidence interval, 0 to 38.3 percent). The two patients with relapses received a second course of fumagillin through an expanded-access program. Both had clearance of parasites, but one later had another relapse. No relapse was documented during follow-up in the other patients. The last available CD4 counts were only 6, 35, 72, and 95 cells per cubic millimeter in the HIV-infected patients who had received highly active antiretroviral therapy before the trial.
Adverse Events
Five serious adverse events were reported (in four patients) during the double-blind phase of the trial (Table 3). One patient in the placebo group was admitted to the hospital because of severe diarrhea and weight loss. Three patients in the fumagillin group had laboratory abnormalities that did not require hospital admission: one patient had grade 4 neutropenia and grade 4 thrombocytopenia, one had grade 4 neutropenia alone, and one patient had grade 3 thrombocytopenia. Both patients with neutropenia had already had grade 4 neutropenia at base line. Platelet counts, however, were above 150,000 per cubic millimeter in all patients at base line. Thrombocytopenia and neutropenia completely resolved within 8 to 14 days after the cessation of fumagillin and other hematotoxic agents. No bleeding or purpura was noted in patients with thrombocytopenia.
|
Seven serious adverse events were reported during the open-label phase of the study. One patient with severe diarrhea had grade 3 hypokalemia, and one patient who was receiving foscarnet had diabetes insipidus. In two patients, grade 4 thrombocytopenia developed; one of these patients also had mild epistaxis and anal bleeding, and the other had hemoptysis and received a transfusion of platelets. Grade 4 anemia developed in two patients, both of whom had had low hemoglobin levels at base line. One patient had grade 4 neutropenia.
Discussion
This double-blind, placebo-controlled trial shows that fumagillin is an effective treatment for E. bieneusi microsporidiosis in immunocompromised patients. Clearance of E. bieneusi, as assessed by repeated parasitologic examinations of stool samples, was achieved in all patients receiving fumagillin, as compared with none of the patients receiving placebo. Parasitologic clearance was also associated with normalization of the results of D-xylose–absorption tests and with clinical benefit, as indicated by decreased weight of stools, decreased use of loperamide, and an increase in Karnofsky scores. The efficacy of fumagillin against this infection is further supported by the fact that patients who were initially assigned to placebo had clearance of microsporidial infection after receiving open-label fumagillin. All 12 patients in our study had clearance of infection, resulting in a cure rate with fumagillin of 100 percent (95 percent confidence interval, 76 to 100 percent).
During a median follow-up of 10 months after the completion of fumagillin treatment, there were only two relapses, both in patients with AIDS. Although the three patients who had not received antiretroviral drugs at base line but received highly active antiretroviral therapy during follow-up had immune reconstitution that prevented assessment of the long-term efficacy of fumagillin, there were no relapses in the two transplant recipients or in the four HIV-infected patients whose CD4 counts remained low.
Toxic effects on bone marrow were the principal adverse effect of fumagillin therapy.26,27 Thrombocytopenia was the most frequent adverse event, occurring in more patients than any other adverse event, with grade 3 and 4 adverse events reported in 4 of the 12 patients (33 percent). Thrombocytopenia began after one week of treatment, with a nadir three days after the end of treatment. Spontaneous recovery of a normal platelet count was observed in all patients one to two weeks after the discontinuation of fumagillin therapy. The mechanisms of fumagillin-induced bone marrow toxicity might be related to a direct toxic effect of fumagillin on megakaryocytic and myeloid progenitors in vitro.27 Bone marrow toxicity had not been reported in patients with intestinal amebiasis21,22 but was reported in patients with cancer and Kaposi's sarcoma, in whom fumagillin was tested as an antiangiogenic agent.28,29 Close monitoring of blood counts is therefore mandatory during fumagillin therapy and should be performed every other day from day 8 to day 19, and treatment should be stopped if the platelet count falls below 75,000 per cubic millimeter.
The results of this trial confirm our previous observations26,27 and support the use of oral fumagillin as a treatment for E. bieneusi microsporidiosis in patients with AIDS and other types of immunodeficiency. The mechanisms by which fumagillin inhibits microsporidial replication are, however, poorly understood. Fumagillin may act through the inhibition of methionine aminopeptidase 2 by irreversibly blocking the active site.30,31,32 Further studies are needed to elucidate the mechanism of action of fumagillin and to develop compounds with lower bone marrow toxicity.
Supported by the Agence Nationale de Recherches sur le SIDA (ANRS) and by a grant (to Dr. Tourneur) from Sidaction (Ensemble contre le SIDA).
Presented in abstract form at the 41st Interscience Conference on Antimicrobial Agents and Chemotherapy, Chicago, December 16–19, 2001.
We are indebted to Professors J. Dormont, J.M. Alexandre, D. Vittecoq, and J. Modaï for helpful discussions and support; to Dr. L. Bertin and Dr. A. Struxiano from Sanofi-Synthelabo for providing purified fumagillin; to Dr. P. Faure and Dr. I. Madelaine from the pharmacy of the Hôpital Saint-Louis; to Dr. O. Liguory for performing the PCR assays; to Professor F. Calvo, from the Centre d'Investigations Cliniques, Hôpital Saint-Louis, for assistance in conducting the trial at this center; to Professor G. Chêne, Professor T. May, Dr. H. Pelloux, and Professor D. Salmon-Ceron, members of the data and safety monitoring board; and to the other members of the scientific committee — Professor F. Calvo, Dr. L. Cotte, Dr. I. Desportes, Dr. I. Madelaine, Dr. C. Maslo, Dr. A. Metro, Professor F. Raffi, and Dr. E. Thervet.
Source Information
From the Department of Infectious Diseases (J.-M.M., M.T., A.G., S.B.), the Laboratory of Parasitology (C.S., F.D.), and the Department of Biostatistics (S.C.), Hôpital Saint-Louis and University of Paris VII; and the Laboratory of Parasitology, Groupe Hospitalier Pitié–Salpétrière, Assistance Publique–Hôpitaux de Paris (J.-G.G.) — all in Paris.
Address reprint requests to Dr. Molina at the Department of Infectious Diseases, Hôpital Saint-Louis, 1 Ave. Claude Vellefaux, 75475 Paris CEDEX 10, France, or at jean-michel.molina{at}sls.ap-hop-paris.fr.
References
Other investigators in the ANRS 090 Study Group included the following: Hôpital Paul Brousse, Villejuif — M. Merad, D. Vittecoq, J.-C. Duclos-Vallée, F. Saliba, A. Charpentier, E. Rudant; Hôpital Edouard Herriot, Lyons — F. Jeanblanc, D. Treille-Ritouet, A. Nageotte, F. de Monbrison, M. Rabodonorina; Centre Hospitalier, Lyons-Sud — H. Maiza, C. Pouteil-Noble, G. Goedert, P. Felman; Centre Médical, Forcilles, Forcilles — J.-P. Lebrun, C. Perret, R. Cabrit; Hôpital Robert Debré, Reims — I. Béguinot, G. Rémy, S. Daliphard, B. Gourdier; Hôpital Saint-Louis, Paris — L. Gérard, J.P. Clauvel, M.-E. Noguera, F. Sigaux.
| |||||||||||||||||||||||||||||||||||||||||||||
Related Letters:
Fumagillin for Intestinal Microsporidiosis
Carr A., Cooper D. A., Molina J.-M., Tourneur M.
Extract |
Full Text |
PDF
N Engl J Med 2002;
347:1381, Oct 24, 2002.
Correspondence
This article has been cited by other articles:
HOME | SUBSCRIBE | SEARCH | CURRENT ISSUE | PAST ISSUES | COLLECTIONS | PRIVACY | TERMS OF USE | HELP | beta.nejm.org Comments and questions? Please contact us. The New England Journal of Medicine is owned, published, and copyrighted © 2009 Massachusetts Medical Society. All rights reserved. |
Previous
