Prior Mycobacterium avium complex infection is linked to immunological nonresponsiveness in HIV‐infected patients on highly active antiretroviral therapy

Abstract

Department of Internal Medicine, School of Medicine, Faculty of Health Sciences,University of Botswana, Gaborone, BotswanaHighly active antiretroviral therapy (HAART) leads toimmune reconstitution, as demonstrated by a substantialincrease in CD4 T-lymphocyte count, which can happeneven in patients with advanced HIV disease and severeimmunodepression [1].However, up to 40% of HIV-infected patients are‘immunological nonresponders’; that is, they have dis-cordant responses to long-term HAART characterized bycomplete suppression of HIVreplication in the absence of asignificant increase in CD4 T-cell count [2,3]. An increasein CD4 T-cell count of o30% and an absolute CD4 count ofo200cells/mL during the first 6–12 months of HAARTidentify such nonresponders [3], who are at increased riskof HIV-related morbidity and mortality [4], and whoserelative risk of clinical progression to AIDS is approxi-mately double that of responders [5]. To date, the riskfactors linked to immunological nonresponsiveness are alower nadir CD4 cell count before therapy [6], lower pre-HAART HIV RNA levels, older age, male gender, hepatitis Cvirus (HCV) coinfection, injecting drug use (IDU), and ofcourse poor adherence to therapy [7,8]. In addition, onestudy from France showed that Mycobacterium aviumcomplex (MAC) infections also predicted immunologicalnonresponsiveness [9].We reviewed the records of all HAART-nai¨ve patientswith AIDS presenting with CD4 counts of o100cells/mLattwo Infectious Diseases Units in Italy (one located inVerona in the north-east of Italy and the other in Cosenzain the south) and investigated whether opportunisticinfections or cancers recorded at presentation had aneffect on subsequent immune reconstitution on HAART.Fifty-three patients with these characteristics wereidentified in Verona and 20 in Cosenza (73 in total).Fifty-one patients (69%) were men. Their median age was43 years. Thirty-two patients (43%) were men who havesex with men, 15 (20%) were injecting drug users, and theothers were heterosexual. All patients who were injectingdrug users were HCV-coinfected. Twenty patients (27%)had Pneumocystis jiroveci pneumonia, nine (12%) disse-minated MAC infections, eight (11%) cryptococcal menin-gitis, eight (11%) neurotoxoplasmosis, seven (10%) Candidaspp. oesophagitis, six (8%) tuberculosis, six (8%) dissemi-nated Cytomegalovirus infection, four (5%) non-Hodgkin’slymphomas, three (4%) Kaposi’s sarcoma, and two (3%)progressive multifocal leucoencephalopathy.The median CD4 T-cell count at the time of HAARTinitiation was 60.68cells/mL and the median HIV RNA viralload was 572,633HIV-1RNAcopies/mL. The median fol-low-up time was 6.5 years. Six patients were nonadherentand excluded from the analysis.After a median follow-up period of 3 years, all 67adherent patients included in the analysis had sustainedviral load suppression (HIV RNA o50copies/mL), and themedian CD4 T-cell count was 391.79cells/mL.In the analysis of relationships with presenting oppor-tunistic infections or cancers, a lower increase in CD4T-cell count (median 59.75cells/mL) and total lymphocytecount (median 74.21cells/mL) was found only in patientswho had experienced MAC infections. How can MACinfections at presentation negatively influence immunerecovery? A number of hypotheses can be put forward[defects in lymphocytopoiesis resulting from bone marrowinfection, induction of immunosuppressive cytokines,tumour necrosis factor (TNF)-a or apoptosis [10]; long-lasting effects of anti-MAC drugs; MAC-induced low-levelchronic immune activation with its effects on T-cell lifespan [11] and on diminished bone marrow functionthrough TNF-a and transforming growth factor (TGF)-breleased by activated dendritic cells, natural killer cells andT cells], but no definitive data are available and furtherstudies are clearly needed.