|
|
| |||||||||||||||||||||||||||||||||||||||||||||||
Background Progressive multifocal leukoencephalopathy affects about 4 percent of patients with the acquired immunodeficiency syndrome (AIDS), and survival after the diagnosis of leukoencephalopathy averages only about three months. There have been anecdotal reports of improvement but no controlled trials of therapy with antiretroviral treatment plus intravenous or intrathecal cytarabine.
Methods In this multicenter trial, 57 patients with human immunodeficiency virus (HIV) infection and biopsy-confirmed progressive multifocal leukoencephalopathy were randomly assigned to receive one of three treatments: antiretroviral therapy alone, antiretroviral therapy plus intravenous cytarabine, or antiretroviral therapy plus intrathecal cytarabine. After a lead-in period of 1 to 2 weeks, active treatment was given for 24 weeks. For most patients, antiretroviral therapy consisted of zidovudine plus either didanosine or stavudine.
Results At the time of the last analysis, 14 patients in each treatment group had died, and there were no significant differences in survival among the three groups (P = 0.85 by the log-rank test). The median survival times (11, 8, and 15 weeks) were similar to those in previous studies. Only seven patients completed the 24 weeks of treatment. Anemia and thrombocytopenia were more frequent in patients who received antiretroviral therapy in combination with intravenous cytarabine than in the other groups.
Conclusions Cytarabine administered either intravenously or intrathecally does not improve the prognosis of HIV-infected patients with progressive multifocal leukoencephalopathy who are treated with the antiretroviral agents we used, nor does high-dose antiretroviral therapy alone appear to improve survival over that reported in untreated patients.
Suspicion of progressive multifocal leukoencephalopathy is aroused by characteristic clinical and neuroradiologic abnormalities in an immunocompromised host. Other disorders, including cytomegalovirus infection, central nervous system lymphoma, and encephalitis caused by infection with the human immunodeficiency virus (HIV), may mimic progressive multifocal leukoencephalopathy, and definitive diagnosis requires the evaluation of tissue. Stereotactic brain biopsy has proved to be effective for this purpose.5,6 Recent studies indicate that the presence of JC virus in cerebrospinal fluid, as identified by the polymerase chain reaction (PCR), has high specificity for the diagnosis of active disease.7,8 Average survival after diagnosis in HIV-infected patients ranges from approximately 2.5 months9 to 4 months.4 Remission, prolonged survival, and even spontaneous recovery may occur, however.10,11,12 Berger et al. have estimated that in approximately 7 percent of patients, progressive multifocal leukoencephalopathy follows a more benign course, with survival of more than one year.13
Treatment with prednisone, acyclovir, vidarabine (given either intravenously or intrathecally), HLA-matched platelets, and interferon alfa have not resulted in consistent improvement in patients with progressive multifocal leukoencephalopathy.14,15,16,17,18,19 In patients with AIDS, both antiretroviral agents and cytarabine have been found to be efficacious in some small, uncontrolled studies but not in others. Such reports have led to the frequent use of cytarabine in patients with AIDS who have progressive multifocal leukoencephalopathy. Because cytarabine has severe side effects, including immune suppression, this unestablished therapy entails considerable risk.
AIDS Clinical Trials Group (ACTG) Study 243 was a multicenter study comparing antiretroviral medication alone with antiretroviral therapy plus cytarabine for the treatment of progressive multifocal leukoencephalopathy in HIV-infected subjects. The protocol ensured that the diagnosis of progressive multifocal leukoencephalopathy was based on tissue evaluation.
Methods
Study Design
This trial was a randomized, multicenter, open-label study, designed to enroll 90 patients of either sex. All subjects had HIV type 1 infection and clinical and radiologic findings indicative of progressive multifocal leukoencephalopathy. All subjects were required to have the diagnosis established within two months of study entry. Tissue obtained by stereotactic brain biopsy was evaluated both by standard neuropathological examination at the study center and by in situ hybridization for JC virus,20 conducted at the National Institute of Neurological Disorders and Stroke. Subjects were eligible only if at least one of these tests confirmed the diagnosis. Cerebrospinal fluid was evaluated by PCR for the presence of JC virus in a number of subjects. Other criteria for inclusion were an age of 18 to 65 years, an absolute neutrophil count of 750 cells per cubic millimeter or higher, a platelet count of 50,000 per cubic millimeter or higher, serum concentrations of alanine aminotransferase, aspartate aminotransferase, or both that were less than five times the upper limit of normal, the ability to provide informed consent or the assignment of a durable power of attorney, and, in the case of women, a negative serum pregnancy test and use of adequate contraception throughout the study.
Exclusion criteria included the administration within the past 14 days of interferon, ganciclovir, foscarnet, antiretroviral therapy other than zidovudine, didanosine, or zalcitabine, or experimental drugs for the treatment of progressive multifocal leukoencephalopathy; systemic chemotherapy for cancer; prior treatment with cytarabine; an active opportunistic infection; conditions precluding the placement of an Ommaya reservoir; intolerance of all the antiretroviral medications used in the study; allergy to or intolerance of filgrastim (granulocyte colony-stimulating factor); and other life-threatening complications likely to cause death within three months.
An ACTG neurologist conducted all neurologic evaluations. Optimal antiretroviral therapy was determined during a lead-in period of one to two weeks; the preferred therapy was 300 mg of zidovudine three times a day and 200 mg of didanosine (125 mg if the patient weighed less than 60 kg) twice a day. Subjects already receiving zalcitabine and those with a history of intolerance of didanosine received 0.75 mg of zalcitabine three times a day in addition to zidovudine. Subjects who were unable to tolerate either didanosine or zalcitabine could enter the study and receive zidovudine alone. Before enrollment was completed, new antiretroviral agents were approved by the Food and Drug Administration and were permitted in the study.
After the lead-in period, subjects were randomly assigned to one of three treatments, each of which was given for 24 weeks. Group 1 continued to receive the antiretroviral regimen established during the lead-in period, with dose adjustments to reduce or eliminate toxic effects. Group 2 continued to receive the established antiretroviral regimen and also received 4 mg of cytarabine per kilogram of body weight daily for 5 days by intravenous infusion, followed by a 16-day period when antiretroviral therapy alone was given. This 21-day cycle was repeated throughout the study. Group 3 received antiretroviral therapy plus 50 mg of cytarabine, administered intrathecally with an Ommaya reservoir, once a week for four weeks, then once every two weeks for eight weeks, then once every four weeks for the remainder of the study. Filgrastim was administered to all subjects in group 2 after each five-day cycle of cytarabine; it was administered to patients in the other groups when required to counteract neutropenia.
An external performance and safety monitoring board closely monitored the study. A first interim review was conducted at 18 months; on the basis of the interim results, the board recommended that the study continue. A second interim review was conducted when 50 percent of the expected events had occurred. After the second review, the board concluded that no treatment was likely to show a survival benefit, even if the study were continued to completion. On the basis of that recommendation, the study was terminated at 24 months.
Statistical Analysis
An intention-to-treat analysis was performed that included all eligible subjects randomly assigned to the three treatment groups. Kruskal–Wallis and Fisher's exact tests were used to compare continuous and discrete measures, respectively, among the treatment groups. Differences in the length of time to the occurrence of a first adverse effect due to drug toxicity and the length of time to death were tested by the log-rank test. The Kaplan–Meier method was used to estimate survival in the three groups.
Analyses were based on stochastic curtailment methods for the primary end point (survival). These methods were used to estimate the conditional power of the study if it were to be completed, given the observed data up to each interim analysis. These estimates were derived by simulation and by analytic techniques, with use of the normal approximation. Reported P values are two-sided.
Results
Subjects
The study was open for enrollment from April 1994 to August 1996. Sixty-four patients were enrolled at 13 ACTG sites. The study was approved by the institutional review board at each site, and all subjects provided written informed consent. Data on 62 patients enrolled before May 1996 were available during the second (final) interim analysis. Three subjects were lost to follow-up or withdrew during the lead-in period and were therefore not randomly assigned to a treatment group. Brain biopsies were positive in 49 patients by both methods (Table 1); 3 subjects were positive on in situ hybridization but not on microscopical examination; 5 were positive on microscopical analysis but not on in situ hybridization. Two additional patients were considered ineligible, since neither the neuropathological nor the in situ evaluation confirmed the diagnosis. The reported results are thus based on 57 patients who could be evaluated. One of these did not receive any study medication.
|
|
The majority of subjects (44, or 77 percent) had received zidovudine before entering the study. Didanosine alone or in combination had been taken by 21 patients (37 percent), stavudine by 14 patients (25 percent), and zalcitabine by 8 patients (14 percent). A history of treatment with saquinavir in combination with other antiretroviral agents was reported by three patients (5 percent; one in the intravenous-cytarabine group and two in the intrathecal-cytarabine group), and a history of ritonavir treatment by one patient in the intrathecal-cytarabine group. Seven patients did not report any prior use of antiretroviral agents.
Follow-Up
All 57 patients, including those who successfully completed 24 weeks of therapy, were followed while receiving the assigned drug until the end of the study or until they died. The median follow-up was 8.7 weeks and did not differ significantly among the three treatment groups (P = 0.78 and P = 0.66 by the log-rank test for the three-way and two-way comparisons, respectively).
Seven patients completed the 24-week treatment. Three (one in each group) who were receiving active therapy at the time the study was terminated discontinued treatment at that time. Table 3 lists the reasons for permanent discontinuation of treatment. Twenty-two patients (39 percent) died during the study. One patient in the intrathecal-cytarabine group and one in the intravenous-cytarabine group discontinued treatment because of thrombocytopenia, as required by the protocol. Treatment was never dispensed to one patient in the antiretroviral-therapy-only group. Fourteen patients discontinued treatment at their own request, four at the request of the investigators, five because of drug-induced toxicity, and one at the start of other experimental treatment. The majority of the patients who discontinued treatment at their own or an investigator's request had been randomly assigned to receive intrathecal cytarabine (11 patients). The median length of treatment was longer for the antiretroviral-therapy-only group (8.9 weeks, vs. 6.4 weeks for the two cytarabine groups combined), but the difference was not significant (P = 0.56 and P = 0.29, respectively, by the log-rank test).
|
As antiretroviral medication, 29 patients (51 percent) received zidovudine combined with didanosine, 7 (12 percent) received zidovudine plus zalcitabine, and 7 (12 percent) received zidovudine alone. In addition to this therapy, 15 patients received stavudine, 5 lamivudine, 5 saquinavir, and 1 ritonavir. Saquinavir was given in combination with zidovudine and lamivudine to three patients and in combination with stavudine and lamivudine to one patient. Ritonavir was used in combination with zidovudine in one patient. Five subjects received other drug combinations containing zidovudine. Two patients received didanosine alone, and two zalcitabine alone. Compliance with antiretroviral therapy was assessed every 4 weeks and was consistently rated as good (more than 80 percent of medication taken) for the majority of the patients (90 percent) throughout the 24-week protocol.
Safety
The standardized ACTG scale for grading toxicity assigns a value from 0 to 5 to clinical and laboratory abnormalities, according to their severity. Twenty-three patients had drug-induced laboratory abnormalities rated grade 3 or higher. A larger number of patients in the intravenous-cytarabine group (11 of 20 patients [55 percent]) than in the other groups had laboratory evidence of drug toxicity, but the differences were not significant (P = 0.22 and P = 0.81 for the three-way and two-way comparisons, respectively, by the log-rank test).
No significant difference was detected among the groups with respect to blood chemical abnormalities (three patients in the antiretroviral-therapy-only group had such toxic effects, as did two in the intravenous-cytarabine group and one in the intrathecal-cytarabine group). Nineteen patients had evidence of hematologic toxicity rated grade 3 or higher during the 24-week treatment period; the majority (10 patients) were in the intravenous-cytarabine group (50 percent of that group had such effects) (Table 4). Toxic effects (grade 3 or higher) on hemoglobin and platelet counts were more common and occurred earlier in the intravenous-cytarabine group than in the other two groups (P = 0.05 and P = 0.01, respectively, for the three-way comparisons and P = 0.22 and P = 0.01, respectively, for the two-way comparisons, by the log-rank test). No statistically significant difference was detected in effects on the absolute neutrophil counts. Hematologic toxicity was the primary reason for dose modification in the intravenous-cytarabine group, but only one patient permanently discontinued treatment for this reason. Thus, the time to the first dose modification differed significantly among groups, whereas the time to permanent discontinuation of treatment did not.
|
Survival
Forty-two patients had died by the time the last analysis was performed (14 in each treatment group). Thirty-seven (88 percent) died of progressive multifocal leukoencephalopathy, two of progressive HIV disease, one of Pneumocystis carinii pneumonia, one of sepsis, and one of an unknown cause. Figure 1 shows the Kaplan–Meier survival curves for each of the three treatment groups; no significant difference among the groups was detected (P = 0.85 by the log-rank test). The intravenous-cytarabine group had the lowest median survival (7.6 weeks), but there were also four long-term survivors in this group who were still alive beyond week 60; however, two of the long-term survivors discontinued treatment early (during the first and seventh weeks of the study). No significant difference was found when we compared the antiretroviral-therapy-only group with the two cytarabine groups combined (P = 0.85 by the log-rank test).
|
Discussion
Most of the literature on the effect of antiretroviral therapy alone and in combination with intravenous or intrathecal cytarabine for the treatment of progressive multifocal leukoencephalopathy in patients with AIDS has been anecdotal and conflicting. Some investigators have reported improvement in patients with AIDS and suspected or biopsy-proved progressive multifocal leukoencephalopathy after the administration of zidovudine or other antiretroviral agents,21,22,23 but not all investigators have had similar results.24 There have also been reports of improvement in non–HIV-related progressive multifocal leukoencephalopathy when cytarabine has been given intravenously, intrathecally, or both,25,26,27 and there are similar reports of improvement in patients with AIDS.21,28,29,30 Some patients appeared to respond to intravenous cytarabine, others to intrathecal cytarabine, and some to a combination. On the other hand, Urtizberea et al. found no benefit of intravenous and intrathecal cytarabine.31 Fong et al., in a prospective and retrospective study of 28 patients, found no benefit in the 9 who had been treated with cytarabine.32 In addition, although cytarabine has an antiviral effect in cell culture,33 a therapeutic effect in vivo has not been established.34,35
ACTG Study 243 was a prospective investigation of progressive multifocal leukoencephalopathy that was designed to address several important issues. To ensure that the approximately 7 percent of patients who appear to have a more benign course were not overrepresented, subjects were required to have progressive multifocal leukoencephalopathy diagnosed within two months of study entry. All diagnoses were confirmed in brain-biopsy specimens by either typical neuropathological findings or in situ hybridization for JC virus. The in situ hybridization was performed at a central site where the investigators had the requisite expertise (the laboratory of Dr. Major at the National Institute of Neurological Disorders and Stroke).
Enrollment of patients and prevention of withdrawals are difficult in clinical trials of rapidly progressive diseases with high fatality rates. Further difficulties associated with the current trial included the frequency of concomitant disabling HIV-related illnesses, the need for subjects to undergo brain biopsy, and the side effects of the prescribed medications, both the antiretroviral drugs and cytarabine. To make the evaluation of effects of medication more accurate, subjects likely to die from causes other than progressive multifocal leukoencephalopathy in less than three months were not enrolled, and most subjects did not have progressive multifocal leukoencephalopathy at a stage that was so advanced as to preclude a therapeutic effect if cytarabine had been effective.
We believe our negative results have direct relevance to clinical practice. Although there was no double-placebo group (i.e., no group that received neither antiretroviral drugs nor cytarabine), the median survival (1.75 to 3.5 months) was very close to that predicted from a review of the literature (2.5 to 4 months), indicating that antiretroviral therapy had no benefit. There were also no significant differences between the group treated with antiretroviral drugs alone and the groups that received intravenous or intrathecal cytarabine. This study was conducted before the advent of highly active antiretroviral therapy, which has been reported to be associated with regression of progressive multifocal leukoencephalopathy.36,37 Our results provide useful comparative data for future studies incorporating highly active antiretroviral therapy, as well as agents more specifically directed against JC virus.
Supported by grants (1 PO1 NS3228, A1-25868, RR00036-37, RR00046, RR00051, RR00722, NS26643, and AI25915) from the National Institutes of Health and by the AIDS Clinical Trials Group, National Institute of Allergy and Infectious Diseases.
Source Information
From the University of North Carolina at Chapel Hill School of Medicine, Chapel Hill (C.D.H.); Harvard School of Public Health, Boston (U.D., C.Y.); Mount Sinai School of Medicine, New York (D.S.); Washington University School of Medicine, St. Louis (D.C.); Yale University School of Medicine, New Haven, Conn. (P.E.W.); Northwestern University School of Medicine, Chicago (B.C.); the Departments of Neurology and Epidemiology, Johns Hopkins School of Medicine, Baltimore (J.M.); the University of California at San Francisco School of Medicine, San Francisco (H.H.); the National Institute of Neurological Disorders and Stroke, Bethesda, Md. (E.M.); Frontier Science and Technology Research Foundation, Amherst, N.Y. (L.M.); and Washington, D.C., General Hospital, Washington, D.C. (J.T.).
Address reprint requests to Dr. Hall at the Department of Neurology, CB7025, University of North Carolina School of Medicine, Chapel Hill, NC 27599-7025.
References
The following persons made a substantial contribution to the conduct, design, and analysis of the study: Z. Antonijevic, Harvard School of Public Health, Boston; J. Berger, University of Kentucky School of Medicine, Lexington; J. Booss, Veterans Affairs Medical Center, West Haven, Conn.; M. Chappell, Community Constituency Group, San Francisco; P. Clax, Division of AIDS, Bethesda, Md.; B. Dezube, Beth Israel Deaconess Medical Center, Boston; M. Donovan Post, University of Miami, Miami; C. Pettinelli and L. Purdue, National Institute of Allergy and Infectious Diseases, Bethesda, Md.; S. Shriver, Social and Scientific Systems, Rockville, Md.; R. Levy, Northwestern University, Evanston, Ill.; K. Robertson, University of North Carolina, Chapel Hill; C. Marra, University of Washington, Seattle; B. Navia, Massachusetts General Hospital and Harvard Medical School, Boston; P. Jatlow, Yale University School of Medicine, New Haven, Conn. The following pharmaceutical companies provided both expertise and study medications: Amgen (R. Wong); Bristol-Myers Squibb (C. McLaren); Burroughs Wellcome (J. Rooney); Hoffmann–LaRoche (M. Salgo); Upjohn (R. Earhart). The members of the performance and safety monitoring board were as follows: B. Jubelt (chair), State University of New York Health Science Center, Syracuse; B. Barton, Maryland Medical Research Institute, Baltimore; J. Noseworthy, Mayo Clinic, Rochester, Minn.; L. Sharer, New Jersey Medical School, Newark. The participating AIDS Clinical Trials Units and investigators were as follows: Albany Medical College, Albany, N.Y. (S. Remick); Columbia–Presbyterian Medical College, New York (M. Crawford, J. Dobkin, G. Dooneief, and K. Marder); Johns Hopkins University, Baltimore (R. Becker, K. Carter, A. Khan, and V. Rexrod); Massachusetts General Hospital and Harvard Medical School, Boston (T. Flynn, M. Hirsch, and E. McCarthy); Mount Sinai Medical Center, New York (E. Chusid, P. Gerits, H. Mendoza, and H. Sacks); Northwestern University, Evanston, Ill. (C. Cooper, R. Murphy, and J. Phair); University of California, San Francisco (S. Forstat, D. Gary, and D. McGuire); University of Cincinnati, Cincinnati; University of Colorado Health Sciences Center, Denver (S. Canmann and K. Tyler); University Hospitals of Cleveland, Cleveland (S. Weaver, S. Gordon); University of North Carolina, Chapel Hill (E. Atos Radzion and W. Robertson); University of Washington, Seattle (A. Collier, C. Cooper, and J. Lund); Washington University School of Medicine, St. Louis (M. Glicksman, J. Voorhees, and M. Royal).
| |||||||||||||||||||||||||||||||||||||||||||||||
Related Letters:
Progressive Multifocal Leukoencephalopathy, HIV, and Highly Active Antiretroviral Therapy
Cinque P., Casari S., Bertelli D., Hall C. D., Yiannoutsos C., Clifford D. B.
Extract |
Full Text
N Engl J Med 1998;
339:848-849, Sep 17, 1998.
Correspondence
This article has been cited by other articles:
HOME | SUBSCRIBE | SEARCH | CURRENT ISSUE | PAST ISSUES | COLLECTIONS | PRIVACY | 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
