799 Discontinuation of Maintenance Therapy for Cryptococcal Meningitis in Patients Treated with HAART: A Multi-center Observational Study C. Mussini*1, P. Pezzotti2, J. M. Meda3, E. Martinez3, J. C. L. Bernaldo de Quiros4, P. Cinque5, V. Borghi1, P. Domingo6, P. Cahn7, P. Bossi8, M. Nelson9, A. Cossarizza10, R. Esposito10, A. Antinori11, J. A. Aberg12 1Policlinico, Modena, Italy; 2Inst Superiore di Sanità, Rome, Italy; 3Inst d'Investigacions Biomèdiques August Pi i Sunyer, Hosp Clin, Univ of Barcelona, Spain; 4Infectious Diseases Unit, Hosp Gregorio Marañon, Madrid, Spain; 5Clin of Infectious Diseases, Univ Vita e Salute, Milan, Italy; 6Hosp de la Santa Creu i Sant Pau, Barcelona, Spain; 7Hosp J A Fernandez, Buenos Aires, Argentina; 8Groupe Hosp Pitié-Salpêtrière, Paris, France; 9Chelsea and Westminster Hosp, London, UK; 10Modena Univ, Italy; 11Natl Inst for Infectious Diseases L Spallanzani, Rome, Italy; and 12Washington Univ, St Louis, MO
Background: HAART has decreased the incidence of new episodes and recurrences of opportunistic infections in patients with AIDS. A few data are available concerning discontinuation of maintenance therapy for cryptococcal meningitis.
Methods: This is an observational study. Inclusion criteria were a previous definitive diagnosis of cryptococcal meningitis, an increase in CD4 count to more than 100 cells/mL as a result of HAART and the subsequent discontinuation of maintenance therapy for cryptococcal meningitis. The primary end-points were the development of a new meningeal or extra-meningeal cryptococcal localization. Incidence rates of cryptococcal meningitis and death related to this opportunistic infection were calculated for the period in which the patients (pts) were not taking antifungal maintenance therapy. Events were assumed to have a Poisson distribution, and exact 95% CI were calculated for the incidence of endpoints.
Results: As of April, 2002, 100 pts were enrolled. Pts were on maintenance therapy for a median of 33.0 months (mos) (range 2.0-90.7). At discontinuation, the median CD4 count was 259 cells/mL (range 4-1231) while the median plasma viral load was < 2.30 log10 copies/ml (range < 1.60-6.34) and the serum cryptococcal antigen was negative in 56 pts. During a median follow-up of 28.4 mos (6.7-64.5) (262 person-yrs [PY]) 4 events were observed (incidence rate = 1.53 per 100 PY; 95% CI, 0.42-3.92): 3 meningitis and one extrameningeal cryptococcal infection. All events occurred in pts who discontinued antifungal maintenance therapy after 6 mos of a CD4 count above 100 cells/mL. Three (3) of these pts developed a recurrent episode in presence of a CD4 count above 100 cells/mL and had a negative serum cryptococcal antigen, that became positive during the recurrent episode.
Conclusions: This study confirmed that the risk of developing a new episode of cryptococcal infection is low even if not absent in pts who discontinue maintenance therapy for cryptococcal meningitis when CD4 increases to more than 100 cells/mL after HAART. In presence of a serum cryptococcal antigen becoming positive after discontinuation, a recurrent episode of cryptococcal infection should be suspected. At discontinuation pts should receive clear advice concerning the risk of relapse and be asked to refer the onset of any new symptom.
