Randomized Phase II Trial of Atovaquone with Pyrimethamine Or

MAJOR ARTICLE HIV/AIDS

Randomized Phase II Trial of Atovaquone with Pyrimethamine or Sulfadiazine for Treatment of Toxoplasmic Encephalitis in Patients with Acquired Immunodeﬁciency Syndrome: Downloaded from https://academic.oup.com/cid/article/34/9/1243/464079 by guest on 24 September 2021 ACTG 237/ANRS 039 Study

Keith Chirgwin,1 Richard Hafner,5 Catherine Leport,9 Jack Remington,6 Janet Andersen,7 Elizabeth M. Bosler,2 Clemente Roque,3 Natasa Rajicic,7 Vincent McAuliffe,4 Philippe Morlat,10 D. T. Jayaweera,8 Jean-Louis Vilde,9 Benjamin J. Luft,2 and the AIDS Clinical Trials Group 237/Agence Nationale de Recherche sur le SIDA, Essai 039 Study Teama 1Department of Medicine, State University of New York (SUNY) at Brooklyn, and Departments of 2Medicine and 3Radiology, SUNY at Stony Brook, and 4New York University, New York, New York; 5Division of AIDS, National Institute of Allergy and Infectious Diseases, Bethesda, Maryland; 6Stanford University School of Medicine and Palo Alto Medical Foundation, Palo Alto, California; 7Statistical and Data Analysis Center, Harvard School of Public Health, Boston, Massachusetts; 8University of Miami, Florida; 9Services des Maladies Infectieuses et Tropicales, Hoˆpital Bichat-Claude Bernard, Paris, and 10Service de Medecine Interne et Maladies Infectieuses, Hoˆpital Saint-Andre, Bordeaux, France

In this international, noncomparative, randomized phase II trial, we evaluated the effectiveness and tolerance of atovaquone suspension (1500 mg orally twice daily) plus either pyrimethamine (75 mg per day after a 200- mg loading dose) or sulfadiazine (1500 mg 4 times daily) as treatment for acute disease (for 6 weeks) and as maintenance therapy (for 42 weeks) for toxoplasmic encephalitis (TE) in patients infected with human immu- nodeﬁciency virus. Twenty-one (75%) of 28 patients receiving pyrimethamine (95% lower conﬁdence interval [CI], 58%) and 9 (82%) of 11 patients receiving sulfadiazine (95% lower CI, 53%) responded to treatment for acute disease. Of 20 patients in the maintenance phase, only 1 experienced relapse. Eleven (28%) of 40 eligible patients discontinued treatment as a result of adverse events, 9 because of nausea and vomiting or intolerance of the taste of the atovaquone suspension. Although gastrointestinal side effects were frequent, atovaquone- containing regimens are otherwise well tolerated and safe and may be useful for patients intolerant of standard regimens for toxoplasmic encephalitis.

Toxoplasmic encephalitis (TE) is the most frequent combination of pyrimethamine and either sulfadiazine cause of focal CNS infection in patients with AIDS. For or clindamycin [1–4]. However, the high frequency of primary treatment of TE, the standard treatment is a toxicity in HIV-infected patients (у40% of patients) often makes completion of a full course of this therapy impossible [1–3]. Experience with other regimens for treatment of acute disease, such as treatment with ma- Received 20 July 2001; revised 14 November 2001; electronically published 28 March 2002. crolides (azithromycin or clarithromycin) or tetracy- a Members of the study group are listed at the end of the text. clines (minocycline or doxycycline), alone or with pyr- Reprints or correspondence: Dr. Keith Chirgwin, Merck Research Laboratories, imethamine, is limited [5–9]. PO Box 4, UNB121, West Point, PA 19486 ([email protected]). Atovaquone, a hydroxynaphthoquinone, is indicated Clinical Infectious Diseases 2002;34:1243–50 for the treatment and prevention of Pneumocystis cari- 2002 by the Infectious Diseases Society of America. All rights reserved. 1058-4838/2002/3409-0012$03.00 nii pneumonia in people intolerant of trimethoprim-

HIV/AIDS • CID 2002:34 (1 May) • 1243 sulfamethoxazole [10, 11]. Also, atovaquone has demonstrated (histologically proven) or presumptive TE (on the basis of com- activity in vitro and in vivo against Toxoplasma gondii [12]. patible clinical and radiologic findings), either acute or relapsed; Animal model data suggest that the combination of atovaquone serological tests with results positive for HIV or a diagnosis of with either pyrimethamine or sulfadiazine may be synergistic AIDS that met the 1993 case definition of the Centers for [13]. Clinical studies of atovaquone with the older tablet for- Disease Control and Prevention [19]; a Karnofsky Performance mulation for the treatment of TE have been limited to salvage Score of у30; and the following laboratory values: hemoglobin, treatment for patients whose infections failed to respond to or у7.0 g/dL; absolute neutrophils, у1000 cells/mm3; platelets, who were intolerant of pyrimethamine-sulfadiazine or clin- у50,000 platelets/mm3; serum creatinine, р1.5 mg/dL; AST damycin [14–16]. and ALT, р5 times the upper limit of normal; and serum In this international, noncomparative study, we evaluated bilirubin, р2 times the upper limit of normal. Patients were the efficacy, safety, and tolerability of combination therapy with allowed to enter the study with the results of serological tests atovaquone (administered orally as a suspension) and either for T. gondii pending from the reference laboratory (laboratory Downloaded from https://academic.oup.com/cid/article/34/9/1243/464079 by guest on 24 September 2021 pyrimethamine or sulfadiazine as treatment for acute disease of J.R., Palo Alto Research Institute, Palo Alto, CA) and were for TE. The bioavailability of the original tablet formulation of subsequently excluded if the results were negative for T. gondii. atovaquone was limited, although it was substantially enhanced Exclusion criteria included the following: receipt of 196 h if administered with food. Atovaquone tablets have been re- of treatment for the current episode of TE before study entry, placed by a suspension formulation that has ∼2-fold greater need for continued treatment with drugs with evidence of ac- oral bioavailability, whether administered with or without food. tivity against T. gondii (including clindamycin, azithromycin, In a trial of atovaquone (tablets) alone for treatment of TE clarithromycin, sulfonamides, dapsone, antifolates, and IFN- among 87 patients whose illnesses failed to respond to treat- g), and concurrent CNS infections or neoplasms. ment or who could not tolerate standard therapy, post hoc Study medications. All patients received atovaquone in a analysis of correlations between plasma atovaquone concentrations and treatment outcomes suggested that responses were suspension at a dosage of 1500 mg orally twice per day with correlated with higher atovaquone levels [15]. Fifty-six (84%) meals or with orally administered nutritional supplements. Af- of 67 of patients with median plasma concentrations of ato- ter a 200-mg loading dose was provided on study day 1, patients vaquone of 118.5 mg/mL had conditions that had either im- received pyrimethamine at a dosage of 75 mg per day (for ! proved or were stable at 6 weeks. On the basis of pharmaco- participants weighing 60 kg, 50 mg per day). Patients were kinetic data from other studies of atovaquone suspension, the also administered leucovorin calcium at a dosage of 10 mg dose chosen for this study (1500 mg twice daily) has been daily for as long as they were receiving pyrimethamine. Sul- estimated to produce average plasma concentrations in the fadiazine was also administered at a dosage of 1500 mg 4 times range of 20–50 mg/mL [17, 18]. per day (for participants weighing !60 kg, 1000 mg 4 times daily). The duration of the treatment for acute disease was 6 weeks. After receiving treatment for acute disease, patients who PATIENTS AND METHODS responded to treatment continued to receive the same treatment regimen, as maintenance therapy, for up to 42 additional Study design. The primary objectives of this study were to weeks. Patients whose illnesses failed to respond to therapy or evaluate the activity, safety, and tolerance of atovaquone ad- who were intolerant of the study therapy received a standard ministered with either pyrimethamine or sulfadiazine for the treatment regimen for TE. Potent antiretroviral therapy was treatment of TE. The study was not designed or powered for not generally available during the time that this study was im- comparisons between the different treatment groups. This was plemented, but participants were allowed to receive antiretro- an open-label, randomized, phase II study with 2 treatment viral therapy at any time during the study. arms: atovaquone-pyrimethamine and atovaquone-sulfadiazine. Participants with a history of intolerance to either pyri- Adverse event management and reporting. To standardize methamine or sulfadiazine were assigned to receive the appro- the approach to managing treatment toxicity across the differ- priate treatment regimen. Written informed consent was ent study sites, the study protocol included dose-modification obtained from the patients or their guardians, and the guide- algorithms for toxicity management. These algorithms provided lines for human experimentation of the US Department of guidance regarding the order in which study medications were Health and Human Services and the institutional review boards discontinued and specified the maximum durations that study of the clinical centers were followed in the conduct of this medications could be held before being discontinued for tox- clinical research study. icity. The National Institute of Allergy and Infectious Diseases, Eligibility. Patients were required to meet the following Division of AIDS Table for Grading the Severity of Adult Ad- criteria for inclusion in the study: a diagnosis of definitive TE verse Events was used for reporting adverse events. This table

1244 • CID 2002:34 (1 May) • HIV/AIDS specifies 4 grades of severity: grade 1, mild; grade 2, moderate; treatment; otherwise, the participant was classified as having grade 3, serious; and grade 4, life threatening. had no response to treatment. Assessment of outcome. Response to treatment was eval- Statistical analysis. The study was designed as a noncom- uated by assessment of clinical and radiographic parameters, parative, randomized phase II trial. Because of the limited en- as described elsewhere [5]. Clinical response was assessed with rollment, most evaluations combined participants randomized a standardized and quantifiable neurologic examination per- to receive a given treatment with participants directly assigned formed on the day of study entry and thereafter on day 4 and to the same treatment because of prior drug intolerance. One- at weeks 1, 2, 3, 6, 12, 24, and 48. There were 5 categories of sided 95% exact CIs were calculated for primary dichotomous response. (1) Complete response (CR) was defined as complete end points. No adjustment was made to the a level of these resolution of previously abnormal TE-related clinical signs and 95% CIs to account for multiple end points in this exploratory symptoms. (2) Partial response (PR) was defined as у1 levels phase II study. The median time to clinical events was estimated

of improvement in at least one-half of the TE-related clinical by actuarial methods, and distributions were calculated by the Downloaded from https://academic.oup.com/cid/article/34/9/1243/464079 by guest on 24 September 2021 parameters that previously had abnormal levels and no dete- Kaplan-Meier method. rioration or new findings for 2 consecutive examinations after the first week. (3) Nonresponse was defined as lack of clinical improvement after the first week. (4) Progressive disease was RESULTS defined as increasing severity of any neurologic sign or new clinical symptoms that appeared after the first week and per- Patients and treatment assignment. From October 1994 sisted for у2 examinations. (5) Relapse was defined as neu- through October 1996, 49 participants (34 from the United rologic findings at an examination after CR or PR was achieved States and 15 from France) were enrolled in the study at 14 that were worsened or new and were attributable to TE. clinical sites. Nine participants with serologic tests performed Radiographic responses were evaluated on the basis of the at the central laboratory that were negative for T. gondii were findings of serial neuroradiographic imaging. Scans were per- excluded from the analyses. The baseline demographic and clin- formed at the following times: at study entry; at week 3, if ical characteristics of the 40 eligible participants are summa- neurologic signs and symptoms were not resolving; at week 6; rized in table 1. One of these participants discontinued study at week 12, unless the scan at week 6 indicated that the illness participation after a cryptococcal infection of the CNS was was fully resolved; and whenever a patient’s condition was not diagnosed during the first study week and was considered un- improving or was deteriorating. Study sites were encouraged assessable for treatment response. The treatment assignments to use the same type of scan (CT or MRI) throughout the of the 39 remaining participants were as follows: 28 received study. One neuroradiologist (C.R.), who was blinded to the atovaquone-pyrimethamine (17 were randomized and 11 were treatment group, read and classified the scans from all sites for assigned to receive this combination), and 11 received atova- response. The examination within the first 7 days of study quone-sulfadiazine (10 randomized and 1 assigned). treatment served as the baseline. Results of a scan performed Eighteen participants received antiretroviral therapy (13 at after у3 weeks of treatment were required for a patient to be entry) at some time during the study. Of these, 11 participants considered assessable for radiographic response. received 1 nucleoside analogue, 6 received 2 nucleoside ana- There were 5 categories of radiographic response. (1) CR logues, and 1 other patient initiated a 4-drug combination that was defined as complete resolution of all abnormal findings on included nevirapine after study entry. CT or MRI scan that were attributed to TE. (2) PR was defined Time receiving study treatment and duration of follow- as no new lesions and no increase in size (i.e., in cross-sectional up. During the 48 weeks of follow-up, 34 of the eligible area) of existing lesions and a 125% decrease in size of more participants discontinued study treatment for the following rea- than one-half of the lesions. (3) Nonresponse was defined as sons: intolerance or protocol-defined toxicity (n p 11 ); failure no new lesions and a !25% decrease in size of more than one- of the illness to respond to treatment, or relapse (n p 7 ); cli- half of the lesions. (4) Progressive disease was defined as the nician/participant request (n p 6 ); loss to follow-up (n p 3 ); appearance of new lesions, an increase in the size of any existing concomitant HIV-related illness (n p 3 ); death (n p 3 ); and lesion, or both. (5) Relapse was defined as the appearance of not assessable for response (n p 1 ). Treatment discontinuation new lesions or an increase in the size of any existing lesion within the first 6 weeks (n p 20 ) was chiefly the result of tox- after a CR or PR was achieved at 6 weeks. icity or intolerance (n p 7 ) and treatment failure (n p 6 ). To determine overall treatment response, patients who were Study treatment discontinuation after 6 weeks (n p 14 ) was identified as having clinical CR or PR and who had a complete, most frequently the result of intolerance or toxicity (n p 4 ). partial, or unassessable radiographic response within the first The median time receiving the study treatment was 10 weeks, 6 weeks of treatment were considered to have had response to and the median duration of study evaluations for treatment

HIV/AIDS • CID 2002:34 (1 May) • 1245 Table 1. Baseline demographic and clinical characteristics of study treatment in the ﬁrst 2 weeks. One of these 2 participants patients with toxoplasmic encephalitis who were treated with could not continue to receive orally administered medications atovaquone and either pyrimethamine or sulfadiazine. after 14 days because of preexisting pancreatitis; this patient

Drug regimen received did not achieve an adequate response by week 6 and died shortly thereafter. The other participant did not receive pyrimethamine Atovaquone plus Atovaquone plus pyrimethamine sulfadiazine after day 4 because of neutropenia. This patient continued to Characteristic (n p 28) (n p 12) receive atovaquone alone for 8 more days, when clinical pro- Sex gression was noted. In addition to the 7 participants with no Male 22 (79) 10 (83) response to treatment, 1 participant who received atovaquone- Female 6 (21) 2 (7) pyrimethamine had an MRI ﬁnding at week 6 that site inves- Race/ethnicity tigators interpreted as indicating an inadequate response (“not

White 13 (46) 5 (42) all lesions had decreased by at least 25%”), and the investigators Downloaded from https://academic.oup.com/cid/article/34/9/1243/464079 by guest on 24 September 2021 Black 9 (32) 2 (17) discontinued treatment with the study drugs. However, this Hispanic 5 (18) 5 (42) patient had a complete clinical response at weeks 3 and 6, and Injection drug use at baseline the MRI scan obtained at week 6 was not submitted for ex- Never 20 (71) 11 (92) amination by the study neurologist. Therefore, the patient was Previously 8 (29) 0 (0) coded as having a response to treatment, in accordance with Unknown 0 (0) 1 (8) the protocol criteria. Age, median years 35 37 Maintenance therapy. Fifteen participants who received CD4 count, median cells/mm3 25 28 the pyrimethamine combination and 5 who received the sul- Karnofsky score, median 70 70 fadiazine combination entered the maintenance therapy phase Weight, median kg 63 59 after having treatment responses at week 6. During the main- Baseline TE diagnosis tenance phase, these 20 participants received study drugs for Definitive 2 (7) 0 (0) a median of 18 weeks and continued study evaluations for Presumptive 24 (86) 10 (83) possible relapse for a median of 30 weeks. One participant (who Relapse 2 (7) 2 (17) received atovaquone-pyrimethamine) experienced relapse dur- NOTE. Data are no. (%) of patients, except as noted. ing the maintenance phase, at study week 17. One other patient had an apparent neurologic deterioration at week 15, and treat- response was 16 weeks. Ninety percent of the patients were ment with study drugs was discontinued because of suspected observed for at least 6 weeks. relapse. However, radiologic examinations performed at weeks Response to treatment for acute disease. Twenty-one(75%) 12 and 16 demonstrated complete resolution of lesions. of 28 participants (95% lower CI, 58%) who received atova- Toxicity and tolerability. Adverse events requiring treat- quone-pyrimethamine and 9 (82%) of 11 participants (95% ment discontinuation occurred in 11 (28%) of 40 eligible par- lower CI, 53%) who received atovaquone-sulfadiazine had an ticipants who received study drugs, 9 (32%) of 28 participants overall treatment response during the first 6 weeks. For the 9 received pyrimethamine and 2 (17%) of 12 received sulfadia- participants classified as having no response to treatment, the zine. Seven patients experienced these events during the first 6 median time to study-treatment discontinuation was 2 weeks; weeks of study treatment. The most common event requiring details of their evaluations are listed in table 2. Of the 7 patients discontinuation of study treatment was intolerance, as a result with no response to treatment who received the pyrimethamine of nausea or vomiting, or inability to tolerate the taste of the combination, 2 did not have clinical or radiologic evaluations atovaquone suspension (9 participants). Two other participants performed after week 1. One of these patients discontinued treat- (1 who received pyrimethamine and 1 who received sulfadia- ment with study drugs after 7 days because of intolerance of the zine) had abnormal liver function test results, for which the taste of atovaquone. Site investigators determined serum ato- protocol required permanent discontinuation of treatment with vaquone levels for the other patient during week 2 of therapy the study drugs. However, these abnormalities persisted after and decided to discontinue treatment with study drugs after the discontinuation of study drugs for both participants, and the results indicated that adequate concentrations could not be site investigators finally judged these adverse events to be un- achieved. related to the study treatment. Twelve additional participants The 5 other participants with no response to treatment who had temporary treatment interruptions caused by adverse received atovaquone-pyrimethamine had inadequate treatment events: 9 participants who received pyrimethamine (4 patients response or TE that failed to respond to therapy by the end of with hematologic events, 2 with vomiting, and 1 each with rash 6 weeks. Two of these participants experienced interruption in and liver or pancreatic enzyme elevations) and 3 participants

1246 • CID 2002:34 (1 May) • HIV/AIDS Table 2. Clinical course of patients with toxoplasmic encephalitis who had no response during induction treatment or who had relapse during maintenance therapy with atovaquone and either pyrimethamine or sulfadiazine.

Duration of Induction response treatment with Duration of Combination study drug, follow-up, Status, patient drug Clinical Radiographic weeks weeks Comments No response during induction treatment 1 Sdz Progressive disease Progressive disease 5 48 Switched to standard therapy for clinical and radiologic progression 2 Sdz Progressive disease Progressive disease 2 8 Patient died at week 8; CNS lymphoma suspected based on the appearance of the lesion on MRI 3 Pyr Progressive disease NA 1 25 Pyr stopped on day 4 due to neutropenia; atovaquone continued until relapse on day 14 4 Pyr Progressive disease NA 2 22 Treatment switched to Pyr plus Sdz at week 2; patient died of “toxoplasmosis” (presumptive diagnosis) at week 22 5 Pyr None NA 2 6 Worsening of preexisting pancreatitis precluded the use of oral (study) drugs; patient died at week 6 6 Pyr Partial None 8 48 Patient poorly adherent to study medications; CT scan was performed at week 8 instead of at week 6 7 Pyr NA NA 1 1 Patient died 1 week after entering study 8 Pyr NA NA 1 48 Patient could not tolerate taste of atovaquone 9 Pyr NA NA 1a 1 Serum atovaquone levels achieved indicated that therapeutic concentrations were not obtainable Relapse during maintenance therapy 10 Pyr Complete Partial 17 17 Patient lost to follow-up after relapse was detected

NOTE. NA, not available; Pyr, pyrimethamine; Sdz, sulfadiazine.

a Duration was 13 days. Downloaded from https://academic.oup.com/cid/article/34/9/1243/464079 by guest on 24 September 2021 September 24 on guest by https://academic.oup.com/cid/article/34/9/1243/464079 from Downloaded who received sulfadiazine (2 patients with skin rash events and patients [15]. Clinical response was observed in 45 (52%) of 1 with neutropenia). Grade 3 and 4 (severe and potentially life these 87 patients and radiographic response in 32 (37%) of 87 threatening) adverse events that occurred are listed in table 3. after 6 weeks. At 18 weeks, 35 (40%) of these 87 patients had died, and 7 had clinical progression of disease or experienced relapse. Adverse experiences resulted in treatment discontin- DISCUSSION uation in 22 (25%) of the 87 patients. In contrast to the current In this study of atovaquone in combination with either pyri- study, gastrointestinal intolerance did not cause any of the treat- methamine or sulfadiazine for the treatment of TE, 30 (77%) ment discontinuations. of 39 patients (95% lower CI, 66%) had a response at 6 weeks. In another study of maintenance therapy for 65 patients who Only 1 (5%) of the 20 patients entering the 42-week mainte- had resolution of an acute episode of TE and intolerance to nance phase experienced a relapse. Eleven patients (28%) dis- у1 of the standard treatment drugs, 75% of the subjects received atovaquone alone, and the remainder of the patients also continued treatment with the study drugs because of intoler- Downloaded from https://academic.oup.com/cid/article/34/9/1243/464079 by guest on 24 September 2021 ance (primarily nausea and vomiting or unacceptable taste), received either pyrimethamine or clindamycin [16]. Seventeen and 12 others temporarily interrupted treatment with the study (26%) of the 65 patients had relapse after a mean of 8 months drug because of various adverse events. Nine of the 11 treat- of study therapy. Overall, there was a high frequency of illnesses ment-terminating adverse events resolved promptly after dis- that failed to respond to treatment or of relapse among patients continuation of study medications, and the other 2 events (el- who received atovaquone alone in these earlier studies [14–16], evations in transaminase levels) were considered unlikely to and 1 subject in our study experienced progression of TE after have been caused by the study drugs. However, some treatment receiving atovaquone alone for 9 days during study weeks 1 failures occurred during or soon after treatment interruptions, and 2. In view of the apparent high risk for a poor outcome, which may have contributed to the failures. atovaquone should not be used as a single agent for treatment Previous studies of atovaquone for treatment of TE have for acute disease or maintenance therapy for TE. used the tablet formulation at a dosage of 750 mg 4 times per The differences in study design and treatment regimens day for salvage therapy. In a pilot study, 8 patients who had among these previous studies and the current trial make com- illnesses that failed to respond to standard therapy or who were parison of results difﬁcult. The salvage treatment studies eval- intolerant of standard therapy were treated with this regimen uated patients whose illness failed to respond to or who were of atovaquone alone. Seven responded to acute-disease therapy, intolerant of standard therapy, and such patients may not be but 2 of these patients later experienced relapse [14]. A later representative of patients who present with untreated TE. In study evaluated the same regimen as salvage treatment in 87 addition, the salvage studies primarily evaluated atovaquone

Table 3. Incidence of grade 3 (severe) or 4 (potentially life-threatening) adverse events pos- sibly related to the study treatment in patients with toxoplasmic encephalitis.

No. of patients, by drug regimen received Atovaquone plus Atovaquone plus pyrimethamine sulfadiazine Toxicity (n p 28) (n p 12) Chemical AST or ALT levels 15.0 times ULN 9 1 Alkaline phosphatase level 15.0 times ULN 3 2 Pancreatic enzyme levels (amylase or lipase) 12.0 times ULN 6 1 Hematologic Thrombocytopenia (!50,000 platelets/mm3)2 0 Neutropenia (!750 neutrophils/mm3)2 3 Signs and symptoms Nausea and vomiting 3 0 Rash 0 1 Dyspnea 0 1 Fever 3 0 Pain or discomfort 2 0

NOTE. ALT, alanine transaminase; AST, aspartate transaminase; ULN, upper limit of normal.

1248 • CID 2002:34 (1 May) • HIV/AIDS monotherapy. The high relapse rates observed during main- This trial was conducted before potent antiretroviral com- tenance therapy in these studies indicated that atovaquone may bination therapies were commonly used. It is likely that the be used most effectively in combination with another active immune reconstitution achieved with these treatments will im- drug. In addition, the bioavailability of atovaquone suspension prove the response to treatment for acute disease and reduce is signiﬁcantly (∼2-fold) greater than that of the tablet for- the risk of relapse with or without the use of maintenance mulation. Although drug levels were not routinely monitored, therapy. The results of a randomized trial indicated that patients plasma atovaquone concentrations were likely to have been with a sustained immunologic response to potent antiretroviral higher in the patients in this study [17]. therapy could safely discontinue primary prophylaxis or main- Clindamycin-pyrimethamine has been studied extensively as tenance therapy for toxoplasmosis [21]. When atovaquone is alternative treatment for TE. An earlier collaborative study un- used concomitantly with antiretroviral drugs, pharmacokinetic dertaken by the AIDS Clinical Trial Group and the Agence interactions with protease inhibitors and nonnucleoside re-

Nationale de Recherches sur le SIDA (ANRS) evaluated treat- verse-transcriptase inhibitors are unlikely, because atovaquone Downloaded from https://academic.oup.com/cid/article/34/9/1243/464079 by guest on 24 September 2021 ment response to pyrimethamine (using the same dosage as in does not appear to induce or inhibit cytochrome P-450 enzymes the present study—i.e., 75 mg per day after a 200-mg loading and is not significantly metabolized [22]. Atovaquone may dose) plus clindamycin (600 mg 4 times daily) administered cause a moderate increase of zidovudine concentrations (in- orally [4]. By means of a standardized approach-to-response crease in area under the concentration-time curve, 35%), ap- evaluation very similar to the one used in the present trial, 35 parently through inhibition of glucuronidation [23], but dose (71%) of 49 participants had a response at 6 weeks. Treatment- adjustment has not been recommended. terminating toxicity occurred in 8 participants (16%). In addition to the small sample size, short average duration Two randomized trials have studied patients with TE to com- of maintenance therapy, and the subsequent introduction of pare responses to treatment with pyrimethamine in combination potent antiretroviral combination therapies, other limitations with either sulfadiazine or clindamycin [2, 3]. In the first trial of this study need to be considered. Because of practical con- of 59 patients, 65% of those randomized to receive the clinda- straints common to all toxoplasmosis treatment trials—and mycin combination had a clinical response, and 23% experienced consistent with standard practice—the diagnosis of TE was toxicity that caused drug discontinuation by 6 weeks. Of those made presumptively for all but 2 patients. Some apparent treat- randomized to receive the sulfadiazine combination, 70% ex- ment failures may be due to lack of response of lesions not perienced a response, and 33% experienced treatment-limiting caused by T. gondii. In this study, 1 patient who received the toxicity [2]. In the second study of 292 patients, 68% of the sulfadiazine combination had illness that failed to respond to patients assigned to receive the clindamycin combination re- treatment at study week 2, although CNS lymphoma was highly sponded and 11% developed treatment-terminating toxicities, suspected at that time on the basis of the appearance of the whereas 76% of the patients assigned to receive the sulfadiazine lesions on an MRI scan. Another possible cause for treatment combination responded and 30% discontinued the drug as a failure is an inadequate serum concentration of atovaquone, a result of toxicities [3]. Again, differences in study designs limit parameter that was not systemically assessed in this study and assessment of similarity between the treatment responses in that likely would not be assessed in clinical practice. Because of study and those observed with the atovaquone combinations in these limitations, it is not possible to draw firm conclusions the present study. However, although the number of patients about the relative efficacy of atovaquone combined with pyr- enrolled in the present study was relatively small, the results imethamine or sulfadiazine. suggest that the atovaquone-containing combinations provide At this time, there are no well-studied alternative treatments response rates that are consistent with these 2 regimens (77% except clindamycin-pyrimethamine. The results of the current compared with 65%–76%). study suggest that atovaquone in combination with either pyr- There was a low relapse rate (5%) among patients who re- imethamine or sulfadiazine has sufficient activity to be consid- ceived maintenance therapy. A randomized study of 124 pa- ered for treatment of acute TE. The 2 atovaquone combination tients comparing daily versus three-times-weekly administra- regimens were safe, and the frequency of adverse events that tion of pyrimethamine-sulfadiazine as maintenance therapy for caused treatment discontinuation (28%) is consistent with the TE reported that relapse rates at 12 months were 17% and 19%, findings of other studies that used standard treatment regi- respectively [20]. Although the maintenance-phase results of mens (16%–33%) [2–4]. On the bases of the results of this trial the present study are promising, the limited time that partic- and previous studies of atovaquone for the treatment of TE ipants received the atovaquone combinations during the main- in patients intolerant of standard therapies, atovaquone- tenance phase (median, 18 weeks) restricts the assessment of pyrimethamine should be considered as an alternative treat- the effectiveness of these combinations for long-term main- ment for patients intolerant of sulfonamides, and atovaquone- tenance therapy. sulfadiazine may be considered for the treatment of patients

HIV/AIDS • CID 2002:34 (1 May) • 1249 who are intolerant of pyrimethamine. Further study is war- acute toxoplasma encephalitis in patients with AIDS. Antimicrob Agents Chemother 1991; 35:2049–52. ranted to better define the role of these combinations for the 7. Lacassin F, Schaffo D, Perrone C, Longuet P, Leport C, Vilde JL. Clar- treatment of TE. ithromycin-minocycline combination as salvage therapy for toxoplasmosis in patients infected with immunodeficiency virus [letter]. Anti- microb Agents Chemother 1995; 39:276–7. ACTG 237/ANRS 039 STUDY TEAM MEMBERS 8. Hagberg L, Palmertz B, Lindberg J. Doxycycline and pyrimethamine for toxoplasmic encephalitis. Scand J Infect Dis 1993; 25:157–60. Members who participated in the development, implementa- 9. Jacobson J, Hafner R, Remington J, et al. ACTG 156 Study Team: tion, or analysis of the trial are as follows: Didier Sicard (Service Dose-escalation, phase I/II study of azithromycin and pyrimethamine for the treatment of toxoplasmic encephalitis in AIDS. AIDS 2001; 15: of Internal Medicine, Hopital Cochin, Paris); Vincent Jeantils 583–9. (Service of Internal Medicine, Hopital Jean Verdier, Bobigny); 10. Hughes W, Leoung G, Kramer F, et al. Comparison of atovaquone Bertrand Dupont (Service of Infectious Diseases, Hopital Pas- (566C80) with trimethoprim-sulfamethoxazole to treat Pneumocystis carinii pneumonia in patients with AIDS. N Engl J Med 1993; 328: teur, Paris); Francois Raffi (Service of Internal Medicine, Hoˆp- 1521–7. Downloaded from https://academic.oup.com/cid/article/34/9/1243/464079 by guest on 24 September 2021 ital Hoˆtel Dieu, Nantes); Zelina Eid, Jean Dormant, and Brigette 11. El-Sadr WM, Murphy, RL, McCabe-Yurick T, et al. Atovaquone com- Bazin (ANRS; Paris, France); Susan Owens and Ann Wala- pared with dapsone for the prevention of Pneumocystis carinii pneumonia in patients with HIV infection who cannot tolerate trimetho- wander (Frontier Science and Technology, Research Founda- prim, sulfonamides, or both. N Engl J Med 1998; 339:1889–95. tion, Buffalo, NY); Victoria Rosenwald and Fred Valentine 12. Araujo FG, Huskinson J, Remington JS. Remarkable in vivo and in (New York University, New York, NY); Ernesto Scerpella and vitro activities of the hydroxynaphthoquinone 566C80 against tachy- Allan Rodriguez (University of Miami, Miami, FL); Donald zoites and tissue cysts of Toxoplasma gondii. Antimicrob Agents Che- mother 1991; 35:293–9. Smith and Patricia Jordan (SUNY Health Science Center, 13. Araujo FG, Lin T, Remington JS. The activity of atovaquone (566C80) Brooklyn, NY) L. Joseph Wheat and Jean Decker (Indiana Uni- in murine toxoplasmosis is markedly augmented when used in com- versity School of Medicine, Indianapolis); Cecilia Shikuma bination with pyrimethamine or sulfadiazine. J Infect Dis 1993; 167: 494–7. (University of Hawaii, Honolulu); Jay F. Dobkin (Columbia 14. Kovacs J and the NIAID-Clinical Intramural AIDS Program. Efficacy Presbyterian Medical Center, New York, NY); Donna Mildvan of atovaquone in treatment of toxoplasmosis in patients with AIDS. (Beth Israel Medical Center, New York, NY); Kenneth Skahan Lancet 1992; 340:637–8. 15. Torres RA, Weinberg W, Stansell J, et al. Atovaquone for salvage treat- and Pam Daniel (University of Cincinnati, OH); Michael Rog- ment and suppression of toxoplasmic encephalitis in patients with ers (GlaxoSmithKline, Research Triangle Park, NC); Gerard No- AIDS. Clin Infect Dis 1997; 24:422–9. tario (Abbott Labs, Abbott Park, IL). 16. Katalama C, Mouthon B, Gourdon D, et al. Atovaquone as long-term suppressive therapy for toxoplasmic encephalitis with AIDS and multiple drug intolerance. AIDS 1996; 10:1107–12. Acknowledgment 17. Dixon R, Pozniak AL, Watt HM, Rolan P, Posner J. Single-dose and steady-state pharmacokinetics of a novel microfluidized suspension of Atovaquone oral suspension (Mepron), pyrimethamine (Da- atovaquone in human immunodeficiency virus–seropositive patients. Antimicrob Agents Chemother 1996; 40:556–60. raprim), and leucovorin calcium were provided by GlaxoSmith- 18. Falloon J, Sargent S, Piscitelli SC, et al. Atovaquone suspension in HIV- Kline. infected volunteers: pharmacokinetics, pharmacodynamics, and TMP- SMX interaction study. Pharmacotherapy 1999; 19:1050–6. 19. Centers for Disease Control and Prevention (CDC). 1993 Revised clas- References sification system for HIV infection and expanded surveillance case definition for AIDS among adolescents and adults. MMWR Morb Mor- 1. Luft BJ, Remington JS. Toxoplasmic encephalitis in AIDS. Clin Infect tal Wkly Rep 1999; 41(RR-17):1–19. Dis 1992; 15:211–22. 20. Podzamczer D, Miro JM, Ferrer E, et al. Thrice-weekly sulfadiazine- 2. Danneman B, McCutchan JA, Israelski D, et al. Treatment of toxo- pyrimethamine for maintenance therapy of toxoplasmic encephalitis plasmic encephalitis in patients with AIDS: a randomized clinical trial in HIV-infected patients. Eur J Clin Microbiol Infect Dis 2000; 19: comparing pyrimethamine plus clindamycin to pyrimethamine plus 89–95. sulfonamides. Ann Intern Med 1992; 116:33–43. 21. Miro JM, Lopez JC, Podzamczer D, et al. Discontinuation of toxo- 3. Katlama C, De Wit S, Doherty O, et al. Pyrimethamine-sulfadiazine plasmic encephalitis prophylaxis is safe in HIV-1– and T. gondii– versus pyrimethamine plus clindamycin for the treatment of toxo- co-infected patients after immunological recovery with HAART: pre- plasmic encephalitis in AIDS. Clin Infect Dis 1996; 22:268–75. liminary results of the GESIDA 04/98-B study [abstract 230]. In: Pro- 4. Luft BJ, Hafner R, Korzun A, et al. Toxoplasmic encephalitis in patients gram and abstracts of the 7th Conference on Retroviruses and Op- with the acquired immunodeficiency syndrome. N Engl J Med 1993; portunistic Infections (San Francisco). Alexandria, VA: Foundation for 14:995–1000. Retrovirology and Human Health, 2000:119. 5. Saba J, Morlat P, Raffi F, et al. Pyrimethamine plus azithromycin for 22. Rolan PE, Mercer AJ, Tate E, Benjamin I, Posner J. Disposition of ato- treatment of acute toxoplasmic encephalitis in patients with AIDS. Eur vaquone in humans. Antimicrob Agents Chemother 1997; 41:1319–21. J Clin Microbiol Infect Dis 1993; 12:853–6. 23. Lee BL, Tauber MG, Sadler B, Goldstein D, Chambers HF. Atovaquone 6. Fernandez-Martin J, Leport C, Morlat P, Meyohas MC, Chauvin JP, inhibits the glucuronidation and increases the plasma concentrations Vilde JL. Pyrimethamine-clarithromycin combination for therapy of of zidovudine. Clin Pharmacol Ther 1996; 59:14–21.

1250 • CID 2002:34 (1 May) • HIV/AIDS