The cohort of Spanish AIDS Research Network (CoRIS) is an open, multicenter and prospective cohort of HIV-positive adults, naïve to ART at study entry, seen for the first time from January 2004 in any of the 42 centers from 13 of 17 Autonomous Regions in Spain. Subjects agreed to participate in the study by signing an informed consent form. Ethical approval for CoRIS was granted. A complete description of CoRIS has been published elsewhere. Briefly, CoRIS collects a minimum dataset as provided for in the cohort protocol which includes baseline and follow-up of socio-demographic, immune-virological and clinical data including ART medication, with start and stop dates, as well as reasons for drug discontinuation. Furthermore, all centers are invited to provide data on incident non-AIDS events (NAEs), including non-AIDS–defining malignancies and cardiovascular, renal, liver, psychiatric, bone, and metabolic events. Patients are followed periodically in accordance with routine clinical practice. For this study we have chosen the patients who started treatment from 2012 to 2017.
All methods were carried out in accordance with relevant guidelines and regulations and all experimental protocols were approved by Reina Sofia Hospital Ethics Committee.
Eligible individuals were antiretroviral naïve patients who initiated ART and remained on it for at least 6 months, ≥18 years old and who achieved viral suppression, defined as plasma viral load below 50 copies/ml, within 3-9 months after ART initiation. Patients with no follow-up after viral suppression were excluded from the analyses. In addition, for analyses on association of VLE with first occurrence of any serious NAE, we also excluded individuals who were monitored in centers not providing data on NAEs.
We defined viral load elevations below 200 copies/ml (VLE 50-199) as at least one viral load between 50 and 199 copies/ml after virological suppression. Following this definition, for descriptive purposes, individuals were classified into three groups: uncommon VLE50-199 or not elevation of viral load (viral load elevation between 50-199 copies/ml in <25% of the measurements or viral load < 50 copies/ml), frequent VLE 50-199 (viral load elevation ranging from 50 to199 copies/ml between 25% and 50% of the measurements) and very frequent VLE 50-199 (viral load elevation between 50-199 copies/ml in more than 50% of the measurements).
We have also considered for the analysis those patients with non-consecutive elevations of the viral load higher than 200 copies and lower than 1000 copies. We denominate them VLE 200-999.
Outcomes of this study were experiencing first virological failure, defined as at least two consecutive viral loads more than 200 copies/ml or one more than 1.000 copies/ml, occurrence of clinical events (first AIDS events or death) and first serious NAE [non-AIDS–defining malignancies, cardiovascular, renal, and liver-related] after virological suppression.
A descriptive analysis of patients’ characteristics at ART initiation was carried out using frequency tables for categorical variables, and median and interquartile range (IQR) for continuous variables. Differences in socio-demographic and clinical characteristics between different groups of VLE (uncommon, frequent and very frequent) were assessed through the chi-squared test for independence for categorical variables, and the non-parametric Kruskal-Wallis tests for continuous variables.
A multivariable multinomial logistic regression model was fit to investigate sociodemographic and clinical factors associated with frequent VLE50-199 and very frequent VLE50-199 in comparison to uncommon VLE50-199.
For the analyses of association of VLE groups with first virological failure and clinical events after virological suppression, follow-up started at the date of the first viral load below 50 copies/ml within 3-9 months after ART initiation and ended at the date of the event of interest, date of last study contact or censoring date, whichever arose first. Analyses were right-censored 5 years after virological suppression due to low numbers of events observed thereafter.
For those three outcomes, VLE groups was analyzed as a time-varying covariate. We calculated cause-specific cumulative incidence curves and used Cox proportional hazards models to estimate crude and adjusted hazard ratios. The proportional hazards assumption was checked graphically and by tests based on Schoenfeld residuals.
Multivariable models were adjusted for sex (male, female), age (<50, ≥50), transmission group (homo/bisexual, injecting drug use, heterosexual, other/unknown), educational level (compulsory education or not, upper secondary or university, unknown), country of origin (Spain, not Spain, unknown), CD4+ cell count (<200, ≥200, unknown), HIV-1 viral load (<100000, ≥100000, unknown) and AIDS diagnosis (no, yes), hepatitis C virus antibodies (no, yes, unknown), hepatitis B surface antigen (no, yes, unknown), ART regimen (2 nucleoside reverse transcriptase inhibitor [NRTI] + 1 non-nucleoside reverse transcriptase inhibitor [NNRTI], 2 NRTI + 1 [boosted protease inhibitor[PI/b], 2 NRTI + 1 integrase strand transfer inhibitor [INSTIs], other/non-specified) and duration of ART. Wald tests were used to derive p-values. All statistical analyses were performed using SPSS Software (version 24) and r software.