Immune reconstitution consecutive to HAART initiation is often complicated by adverse clinical manifestations caused either by the unmasking of preexisting untreated opportunistic infections or the clinical deterioration of a known and treated opportunistic infection . This immune reconstitution disease (IRD) may be caused by bacteria, parasites, viruses, or fungi, which stimulate a dysregulated inflammatory response. This response is most acute during the initial rapid recovery of immune functions in severely immunodeficient patients having initiated HAART. Superficial and systemic mycoses are a major cause of morbidity and mortality among patients living with HIV. IRD has been described with fungal agents such as Cryptococcus neoformans , Histoplasma capsulatum , and Pneumocystis jiroveci .
We have recently observed that the incidence of cutaneous mycoses seemed to increase after HAART initiation, suggesting that a mild form of immune reconstitution disease might increase the patency of cutaneous mycoses . Mucosal candidiases are a very frequent clinical event as immunity declines during the course of HIV infection. There is, however, scant literature on IRD and candidiasis. Our objective in the present study was to determine whether the initiation of HAART was followed by an increase in the incidence of mucosal candidiases, a possible manifestation of unmasking IRD.
All HIV-positive patients followed in Cayenne, Kourou, and Saint Laurent du Maroni Hospitals between 1 January 1996 and 31 December 2006 were enrolled in the French Hospital Database for HIV (FHDH). Time-independent variables such as sex, nationality, mode of acquisition of HIV, and time-dependent variables such as age, CD4 cell counts, HIV-1 viral loads, treatments, and clinical events are routinely entered by trained clinical studies technicians. Diagnoses are coded according to the 10th International Classification of Diseases. Therefore, specific details such as the size and exact topography of the lesion could not be obtained from the database. Patients included in the FHDH give informed consent to the use of their data. Their identity is encrypted before the data are sent to the Ministry of Health and the Institut National de la Recherche Médicale, which centralize data from Centers for Information and Care of HIV throughout France. This data collection is approved by the Commission Nationale Informatique et Libertés. The data were analysed using STATA 8.0 (STATA Corp., College Station, Texas, USA).
In this retrospective cohort study, a single failure Cox proportional hazards model was used to evaluate the adjusted relationship between failure and explanatory variables. The failure event was the incidence of candidiasis. We studied several models in which failure was oral candidiasis, vaginal candidiasis, oesophageal candidiasis, and all mucosal candidiases pooled together. Candidiasis was diagnosed clinically for oral candidiases, endoscopically for oesophageal candidiases and mycologically for vaginal candidiases. The main explanatory variables were age, sex, transmission mode, nationality (French citizens versus non-French citizens), CD4 cell count at the time of the visit (categorized as < 50, 50–199, 200–499, and > 500 cells/μl), the presence or absence of HAART, the time since treatment started (< 2 months, 2–4 months, 4–6 months, 6–12 months, > 12 months). The proportionality of the hazard functions was determined using Schoenfeld and scaled Schoenfeld residuals and the global proportional hazards test. Age, nationality, CD4 cell count category at the time of HIV diagnosis, and follow-up duration were transformed into dummy variables to compare different groups with a reference group. Because the period that follows HAART initiation is scrutinized more carefully than routine follow-up, the time to a clinical event may have been biased by this change in the frequency of consultations. An adjustment variable was created that tried to control for this difference. The ratio between the number of consultations per time unit during each period was generated and added to the Cox model.
A total of 1647 subjects with 27 662 observations were included, representing a total of 5902 person-years of follow-up. A total of 547 clinical episodes of mucosal candidiasis were recorded in 474 patients. Among these, there were 365 episodes of oropharygeal candidiasis, 134 oesophageal candidiases, and 48 episodes of vaginal candidiasis.
Table 1 shows that after adjusting for potential confounders, there was an increased incidence of candidiases in the first 2 months after HAART initiation, with the exception of oesophageal candidiasis. Although the crude incidence rate of oesophageal candidiasis seemed higher in the 2 months after HAART initiation, the lack of statistical significance may have resulted from the lower number of observations or from underdiagnosis. Unsurprisingly, higher CD4 cell counts were associated with a lower incidence of candidiases. Patients older than 60 years of age were more at risk of developing candidiases [adjusted hazard ratio (AHR) 1.56, 95% confidence interval (CI) 1.2–2; P = 0.002]. As observed elsewhere , patients on cotrimoxazole were more likely to develop candidiasis (AHR 1.4, 95% CI 1.03–1.9; P = 0.03). Both HAART and non-HAART regimens were associated with protection from candidiasis. The use of protease inhibitors (AHR 0.44, 95% CI 0.35–0.56; P < 0.001), non-nucleoside reverse transcriptase inhibitors (AHR 0.64, 95% CI 0.48–0.88; P = 0.006), but not nucleoside reverse transcriptase inhibitors (AHR 1, 95% CI 0.86–1.2; P = 0.8) was associated with a lower incidence of mucous candidiasis. This might have reflected the difference in virological potency of these different classes of antiretroviral agents, with more potent antiretroviral regimens allowing the eradication of the asymptomatic carriage of Candida. Although adjustments were made to control for the increased frequency of consultations after HAART initiation, this may not have been sufficient to control for all the differences that may occur. Adherence issues, however, and the search for toxicities are usually more specifically targeted than the search for candidiasis. Therefore, there may be a real increase in the incidence of candidiases. This could represent a benign form of IRD in which the pathogen is unmasked by the restored immune system.
1. Shelburne SA III, Hamill RJ. The immune reconstituion inflammatory syndrome. AIDS Rev 2003; 133:447–454.
2. Shelburne SA III, Darcourt J, White AC Jr, Greenberg SB, Hamill RJ, Atmar RL, Visnegarwala F. The role of immune reconstitution inflammatory syndrome in AIDS-related Cryptococcus neoformans
disease in the era of highly active antiretroviral therapy. Clin Infect Dis 2005; 40:1049–1052.
3. Nacher M, Sarazin F, El Guedj M, Vaz T, Alvarez F, Nasser V, et al
. Increased incidence of disseminated histoplasmosis following highly active antiretroviral therapy initiation. J Acquir Immune Defic Syndr 2006; 41:468–470.
4. Barry SM, Lipman MC, Deery AR, Johnson MA, Janossy G. Immune reconstitution pneumonitis following Pneumocystis carinii
pneumonia in HIV-infected subjects. HIV Med 2002; 3:207–211.
5. Nacher M, Vantilcke V, Mahamat A, El Guedj M, Vaz T, Randrianjohany A, et al
. Increased incidence of cutaneous mycoses following HAART initiation: a benign form of Immune reconstitution disease? AIDS 2007; 21:2248–2250.
6. Abgrall S, Charreau I, Joly V, Bloch J, Reynes J, Delta Coordinating Committee. Risk factors for esophageal candidiasis in a large cohort of HIV-infected patients treated with nucleoside analogues. Eur J Clin Microbiol Infect Dis 2001; 20:346–349.