Discontinuation of secondary prophylaxis for cryptococcal meningitis in HIV-infected patients responding to highly active antiretroviral therapy

Abstract

Several studies have demonstrated that secondary prophylaxis for specific opportunistic infections can be safely discontinued in patients in whom CD4 cell counts have increased and remained stable above certain thresholds [1]. These data have been useful when considering the safe withdrawal of secondary Pneumocystis carinii pneumonia (PCP) and cytomegalovirus prophylaxis under selected conditions, although very few data exist concerning other less common major opportunistic diseases [2]. We present here the one year results of a prospective study of HIV-1-infected patients who discontinued secondary prophylaxis for cryptococcal meningitis after showing durable immunological response to highly active antiretroviral therapy (HAART). The ethics committee of the hospital approved the study and all the patients who participated gave informed consent. Patients on secondary prophylaxis for cryptococcal meningitis were identified from the patient database. Staining or culture from adequate clinical samples was required to confirm the diagnosis. Candidates had to have received adequate induction therapy followed by ongoing secondary prophylaxis. To be considered for the study, patients had to be clinically stable, have been treated with any HAART regimen for at least 6 months and have experienced an increase of CD4 cells above 100/mm3 and a viral load reduction below 200 copies of HIV-1 RNA/ml (Amplicor HIV Monitor, Roche Diagnostics Systems, Branchburg, NJ, USA). Cerebrospinal fluid (CSF) and blood were obtained at inclusion for fungal cultures and the detection of cryptococcal antigen (Pastorex Crypto Plus, Sanofi Diagnostics Pasteur, Marnes-la-Coquette, France). Blood for fungal cultures and the measurement of serum cryptococcal antigen were also performed approximately every 3 months at each routine clinical visit. If the patient developed symptoms suggestive of cryptococcal meningitis, a lumbar puncture was performed to obtain CSF for fungal culture and the detection of cryptococcal antigen. The incidence of recurrences and their confidence intervals were calculated based on Poisson regression models. Four men and two women met the inclusion criteria. Their median age was 38 years (interquartile range 34–45). One patient had previously had PCP. Five patients were taking HAART regimens containing HIV-1 protease inhibitors. All the patients were taking cotrimoxazole prophylaxis against PCP. The median CD4 cell count at the diagnosis of cryptococcal meningitis was 42/mm3 (interquartile range 26–58). Five patients experienced at least one recurrence after the initial episode of cryptococcal meningitis (one recurrence, one patient; two recurrences, three patients; and three recurrences, one patient). The median time from the first episode to the first recurrence was 4 months (range 3–5) and from the first recurrence to the second recurrence 4 months (range 4–6). All recurrences but one occurred before the introduction of HAART. The only recurrence on HAART occurred 3 months after HAART. The median duration of HAART before the withdrawal of secondary prophylaxis for cryptococcal meningitis was 11 months (range 9–15). All patients were taking fluconazole as secondary prophylaxis for cryptococcal meningitis at doses of 200 mg (n = 5) or 400 mg (n = 1) daily. Fig. 1 shows the evolution of CD4 cell counts in the six patients included from the first episode of cryptococcal meningitis until the end of follow-up. By the end of follow-up, three patients had less than 200 copies of HIV-1 RNA/ml, and the viral load in the other three patients showed low values (823, 2825, and 12300 copies/ml, respectively). CSF and blood fungal cultures at inclusion were negative for all patients. Baseline CSF cryptococcal antigen was also negative in all patients; however, three patients had baseline serum cryptococcal antigen titres higher than 1 : 8 (two patients, 1 : 16; one patient, 1 : 32). Blood cultures on routine visits were negative, although the serum cryptococcal antigen was detected occasionally during follow-up in all patients at values ranging from 1 : 8 to 1 : 256. The detection of serum cryptococcal antigen was not associated with the time from the withdrawal of prophylaxis or with the presence of clinical symptoms. There was no need to perform a lumbar puncture for diagnostic purposes in any patient. No patient has been diagnosed with a recurrence of cryptococcal meningitis after 102 patient-months of follow-up (95% confidence interval 0–3.6 episodes per 100 patient-months).Fig. 1.: Evolution of CD4 cell counts in the six patients included from the first episode of cryptococcal meningitis until the end of follow-up. The time-point ‘0 months’ corresponds to the moment of discontinuation of secondary prophylaxis. Patients are indicated by capital letters from A to E: ––▪–– A; ––□–– B; ––&●–– C; ––○–– D; ––&◆–– E; ––⋄–– F.The results of this study support the concept that life-long secondary prophylaxis for cryptococcal meningitis may be safely discontinued in HIV-infected patients showing successful immunological responses to HAART. Recent practice guidelines for the management of cryptococcal diseases emphasize that the use of potent HAART to control HIV replication is a key element in preventing a relapse of cryptococcal meningitis in HIV-infected patients [3]. The small number of patients included limited the results of this study. Although the results from a prospective, randomized study would be more welcome than those from an observational study, such controlled studies may be difficult to perform because the prevalence of cryptococcal meningitis was lower than that of other opportunistic diseases in the past and may be even lower in the HAART era. The discontinuation of secondary prophylaxis for cryptococcal meningitis after immunological response to HAART was first reported by Aberg et al. [4]. In contrast to our series, most of the patients in the study by Aberg et al. [4] had negative values of serum cryptococcal antigen at inclusion and follow-up. Serum cryptococcal antigen is a useful screening test for the presence of cryptococcosis [5], although its serial monitoring has played a limited role in the management of HIV-infected patients with secondary prophylaxis for cryptococcal meningitis [6]. Similarly, the results of our study suggest that the measurement of serum cryptococcal antigen is not useful either in the management of HIV-infected patients with previous cryptococcal meningitis who discontinue secondary prophylaxis after a durable immunological response to HAART. We did not perform any follow-up lumbar punctures and therefore do not know how the evolution of the CSF cryptococcal antigen was from its negative value at baseline. In patients on secondary prophylaxis for cryptococcal meningitis, an increase in CSF antigen titres has been associated with a relapse of cryptococcal meningitis [6]. However, lumbar puncture is an aggressive procedure that was not clinically justified in our patients, and may not be readily accepted in the absence of clinical symptoms, even in the context of an investigational study. In conclusion, the discontinuation of secondary cryptococcal prophylaxis is safe for HIV-infected patients showing immunological recovery after HAART, and the clinical evolution may be enough to follow these patients accurately. Esteban Martíneza Miguel A. García-Viejoa Maria A. Marcosb José B. Pérez-Cuevasa José L. Blancoa Josep Mallolasa José M. Miróa José M. Gatella