The influence of age‐associated comorbidities on responses to combination antiretroviral therapy in older people living with HIV

Abstract Introduction Multiple comorbidities among HIV‐positive individuals may increase the potential for polypharmacy causing drug‐to‐drug interactions and older individuals with comorbidities, particularly those with cognitive impairment, may have difficulty in adhering to complex medications. However, the effects of age‐associated comorbidities on the treatment outcomes of combination antiretroviral therapy (cART) are not well known. In this study, we investigated the effects of age‐associated comorbidities on therapeutic outcomes of cART in HIV‐positive adults in Asian countries. Methods Patients enrolled in the TREAT Asia HIV Observational Database cohort and on cART for more than six months were analysed. Comorbidities included hypertension, diabetes, dyslipidaemia and impaired renal function. Treatment outcomes of patients ≥50 years of age with comorbidities were compared with those <50 years and those ≥50 years without comorbidities. We analysed 5411 patients with virological failure and 5621 with immunologic failure. Our failure outcomes were defined to be in‐line with the World Health Organization 2016 guidelines. Cox regression analysis was used to analyse time to first virological and immunological failure. Results The incidence of virologic failure was 7.72/100 person‐years. Virological failure was less likely in patients with better adherence and higher CD4 count at cART initiation. Those acquiring HIV through intravenous drug use were more likely to have virological failure compared to those infected through heterosexual contact. On univariate analysis, patients aged <50 years without comorbidities were more likely to experience virological failure than those aged ≥50 years with comorbidities (hazard ratio 1.75, 95% confidence interval (CI) 1.31 to 2.33, p < 0.001). However, the multivariate model showed that age‐related comorbidities were not significant factors for virological failure (hazard ratio 1.31, 95% CI 0.98 to 1.74, p = 0.07). There were 391 immunological failures, with an incidence of 2.75/100 person‐years. On multivariate analysis, those aged <50 years without comorbidities (p = 0.025) and age <50 years with comorbidities (p = 0.001) were less likely to develop immunological failure compared to those aged ≥50 years with comorbidities. Conclusions In our Asia regional cohort, age‐associated comorbidities did not affect virologic outcomes of cART. Among those with comorbidities, patients <50 years old showed a better CD4 response.

positive patients have developed age-associated comorbidities such as cardiovascular, metabolic, pulmonary, renal, bone and malignant diseases, and these are often more prevalent compared with HIV-negative individuals [7,8]. Risk and management of comorbidities in ageing adults with HIV will continue to evolve as treatment improves and life expectancy increases [5,6].
Polypharmacy is also common in the HIV-positive older adult population [9,10]. The Swiss HIV cohort study comparing HIV-positive adults aged ≥50 years with HIV-positive patients aged <50 years on cART found that older patients were more likely to receive one or more co-medications compared with younger patients [11]. This study also determined that older patients had more frequent potential for drug-to-drug interactions when compared to younger patients. The effects of polypharmacy may be more substantial in older HIV-positive persons because of the increased chance of drug-to-drug interactions [9,12]. It has been shown that older HIV-positive patients have better adherence to cART than younger patients [13,14], and this can increase the likelihood of potential drug interactions. Drug interactions might be associated with a substantial risk for toxicity, decreased efficacy and subsequent emergence of drug resistance.
Another paper with the Swiss HIV cohort study investigated the prevalence of comedications and potential drug-to-drug interactions within a large HIV cohort, and their effect on ART efficacy and tolerability [15]. They found potential drugto-drug interactions increase with complex ART and comorbidities, but no adverse effect was noted on ART efficacy or tolerability.
Previous studies showed older HIV-positive individuals have a less robust immune response but, likely due to better adherence, a better virologic response [16][17][18]. However, multiple comorbidities among HIV-positive individuals may increase the potential for polypharmacy and older individuals with comorbidities, particularly those with cognitive impairment, may have difficulty in adhering to complex medication regimens [13]. However, the effects of age-associated comorbidities on the treatment outcomes of cART are not well known. In this study, we investigated the effects of age-associated comorbidities on therapeutic outcomes of cART in HIV-positive adults in Asian countries.

| Study design and data collection
We analysed data from the TREAT Asia HIV Observational Database (TAHOD), a prospective, observational cohort study of HIV-positive adults enrolled from 21 clinical sites, which is a contributing cohort to IeDEA Asia-Pacific [19]. We selected eligible subjects for this analysis among patients who were enrolled in TAHOD from 2003 to 2015. The TAHOD database and methods have been previously described [20]. Due to the observational nature of the cohort, viral load (VL) and CD4 testing are not performed on a predefined basis but depend on the site's local practices and the patient's financial circumstances. Institutional review board approvals were obtained at all participating sites, the data management and analysis centre (Kirby Institute, University of New South Wales, Sydney, Australia), and the coordinating centre (TREAT Asia/amfAR, Bangkok, Thailand). Patients provided written informed consent to participate in the TAHOD where required by local institutional review boards.

| Definitions
Patients were included in the analysis if they had been on cART for more than 6 months. Our failure outcomes were defined to be in-line with the World Health Organization (WHO) 2016 guidelines [21] as follows: (i) virological failure was defined as a single VL >1000 copies/mL; (ii) immunological failure was defined as CD4 count falling below 250 cells/ lL after a clinical failure, or persistent CD4 levels below 100 cells/lL (two consecutive CD4 counts below 100 cells/lL within six months). We assumed no treatment failure had occurred if there was an absence of VL or CD4 count. We utilized a single VL measurement, rather than a second confirmatory testing, as the median VL testing frequency in our cohort was 1 (interquartile range (IQR) 1 to 2) per patient per year. Patients were included in the virological failure analysis if they had at least one VL measurement available after six months on cART. Immunological failure analysis included patients with pre-cART CD4 count available and at least one CD4 measurement after six months from cART initiation. Both analyses were censored at four years from cART initiation.
Comorbidities evaluated included hypertension, diabetes, dyslipidaemia and impaired renal function. Hypertension was defined as a diastolic blood pressure ≥90 mmHg and/or systolic blood pressure ≥140 mmHg [22]; diabetes was defined as a fasting blood glucose level ≥7.0 mmol/L or 126 mg/dL [23]; dyslipidaemia was defined using any one of the following four criteria: total cholesterol ≥240 mg/dL, triglyceride ≥200 mg/dL, high-density lipoprotein cholesterol <40 mg/dL, low-density lipoprotein cholesterol ≥160 mg/dL according to National Cholesterol Education Programme ATP-III guidelines; impaired renal function was defined as an estimated glomerular filtration rate (eGFR) <60 mL/minute by CKD EPI equation [24].
Patients were grouped into four categories according to their age and comorbidities: (i) age <50 years with no comorbidities, (ii) age <50 years with comorbidities, (iii) age ≥50 years without comorbidities, and (iv) age ≥50 years with comorbidities. Age-associated comorbidity and cART adherence were included as time-varying variables. Time-fixed covariates included in the analyses were sex, HIV-1 exposure risks, baseline CD4 cell count, baseline viral load, cART regimen, prior AIDS-defining illness, hepatitis co-infection and smoking history. Ethnicity was reported descriptively but not included in the regression analyses due to the inclusion of site as a stratification variable. Year of cART initiation was not included in the multivariate model selection due to collinearity with cART adherence, as our cohort began collecting adherence data from 2011 onwards. However, we assessed the direction of the hazard ratios (HRs) by adjusting with other significant covariates in the absence of the adherence variable.

| Statistical analysis
Cox regression analysis was used to analyse time to first virological and immunological failure, stratified by clinical site. Risk time started six months from cART initiation. Patients who did not fail in either category were censored on the last date of VL testing for the virological failure analysis, and of CD4 testing for immunological failure analysis, all within four years from cART initiation. Sensitivity analyses were performed disaggregating by sex. Data management and statistical analyses were conducted using SAS software version 9.4 (SAS Institute Inc., Cary, NC, USA) and STATA software version 14 (STATA Corp., College Station, TX, USA). Table 1 shows the baseline characteristics of patients included in both the virological and immunological analyses. In the virological failure analysis, a total of 5411 patients were included from Cambodia, China, Hong Kong SAR, India, Indonesia, Japan, Malaysia, the Philippines, Singapore, South Korea, Taiwan, Thailand and Vietnam. The median age at cART initiation was 35 years (IQR 29 to 41), with 66% being aged <50 years without the presence of co-morbidities prior to cART initiation. Most patients were male (71%) and the majority were Thai (33%) and Chinese (27%). Heterosexual mode of HIV exposure was predominant (62%) and the median CD4 cell count at cART initiation was 130 cells/lL (IQR 40 to 228). Of the 5411 patients, there were 912 (17%) with virological failure. The median age was slightly lower at 34 years (IQR 29 to 40) and the median CD4 cell count was 97 cells/lL (IQR 25 to 200).

| Patient characteristics
The immunological failure analysis included 5621 patients in total, with a similar distribution of characteristics to the virological failure analysis. The median CD4 testing was two per patient per year (IQR 1 to 2). There were 391 patients (7%) who had an immunological failure. For each of the comorbidity groups, the median CD4 cell count at cART initiation was 116 cells/lL IQR (40 to 209) for age <50 without comorbidities, 157 cells/lL IQR (47 to 245) for age <50 with comorbidities, 116 cells/lL IQR (37 to 217) for age ≥50 years without comorbidities, and 164 cells/lL IQR (61 to 239) for age ≥50 years with comorbidities. Of the total patients in each comorbidity group, the number of patients initiating cART at CD4 cell count ≤200 cells/lL were 2662/3653 (73%), 894/1449 (62%), 198/285 (69%) and 145/234 (62%) respectively.
To determine the effects of the year of cART initiation on virological failure and to avoid collinearity with the adherence variable, we included year of cART initiation in the multivariate model without adjusting for adherence. As expected, later years of cART initiation were associated with decreased hazard for failure (
In the multivariate analyses, those aged <50 years without comorbidities (HR = 0.66, 95% CI 0.46 to 0.95, p = 0.025) and aged <50 years with comorbidities (HR = 0.54, 95% CI 0.38 to 0.76, p = 0.001) were less likely to develop immunological failure compared to those patients aged ≥50 years with comorbidities. Other factors associated with a reduction in hazard for failure were cART adherence ≥95% (HR = 0.16, 95% CI 0.09 to 0.29, p < 0.001) compared to adherence <95%, female sex (HR = 0.60, 95% CI 0.46 to 0.79, p < 0.001), homosexual mode of exposure (HR = 0.52, 95% CI 0.34 to 0.79, p = 0.002) compared to heterosexual mode of exposure, and higher CD4 count (CD4 51 to 100 cells/lL: To examine patterns of CD4 changes in our patient group, we plotted the median change in CD4 cell count for each of our comorbidities group. Figure 1 shows median CD4 increases at each six-month interval, categorized by comorbidity and age at cART initiation. Patients aged ≥50 years were shown to have slower increases in CD4 cell counts compared

| Sensitivity analyses
Factors associated with virological failure in males and females are shown in Table S1. cART adherence <95% was associated with failure in both sexes; however, higher CD4 cell count at cART initiation was associated with reduced hazard for failure in males, but this was not statistically significant in females. Table S2 reports risk factors for immunological failure in males and females. Age <50 with or without comorbidities, cART adherence ≥95%, and higher pre-cART CD4 cell count were associated with reductions in HRs in both males and females. Females who have never smoked were less likely to develop immunological failure, however this association was not evident in males. Overall the effects of the age-related comorbidity variable in the main analyses and in the sensitivity analyses remained similar suggesting that regardless of sex, those aged ≥50 years with comorbidities had worse immunological outcomes than their younger counterpart either with or without comorbidities.

| DISCUSSION
We hypothesized that age-associated comorbidities may worsen therapeutic outcomes of cART, because of the risk of  polypharmacy and additive negative effects of these health conditions. However, our results showed that presence of age-associated comorbidities did not affect virological outcomes of cART, and patients <50 years with comorbidities had better immunological outcomes compared with patients ≥50 years with comorbidities. The prevalence of age-related comorbidities in this study population was similar to the results from other studies. The prevalence of dyslipidaemia among HIV-positive populations differs depending on the methodology and patient population studied, ranging from 20% to 80% [25]. According to the Swiss HIV Cohort study, the prevalence of hypertension and diabetes mellitus were 56.3% and 4.1% respectively [1]. In that study, the eGFR (calculated by the Modification of Diet in Renal Disease Study equation) of older HIV-positive participants was lower than that of younger HIV-positive patients. HIV-positive patients may have greater risk of non-infectious comorbidities than the general population, because of the effects of HIV itself, prevalent risk factors, and antiretroviral medications [26]. The treatment of older HIV-positive patients is complicated by preexisting comorbid conditions, including cardiovascular, hepatic and metabolic complications that may be exacerbated by the effects of HIV infection per se, immunodeficiency and metabolic and other adverse effects of combination antiretroviral therapy [27,28]. Synergistic deleterious effects of chronic immune activation on the course of HIV infection with the immune senescence of ageing may promote this accelerated course [27].
A study from Italy showed that age-related non-infectious comorbidities were more common among HIV-positive patients than in the general population [26]. They performed a case-control study involving ART-experienced HIV-positive patients treated from 2002 through to 2009. These patients were compared with age-, sex-and race-matched adult controls from the general population. The prevalence of hypertension, renal failure and diabetes mellitus of the HIV group <50 years were 13.2%, 3.78% and 6.17% respectively. The rates were greater than the general population.
Multiple studies have demonstrated that, despite successful ART and viral suppression, immune recovery is less robust with increasing age, highlighting the importance of early diagnosis and treatment of HIV [16,[29][30][31][32]. Consistent with previous studies, patients aged ≥50 years with comorbidities in our study had a greater rate of immunological failure compared to patients <50 years with comorbidities. As shown in Figure 1, patients aged ≥50 years were shown to have slower increases in CD4 cell counts compared to patients <50 years. This is consistent with previous studies as well. The poorer immune recovery in older populations could be caused in part by decreased thymic function in these groups [31]. In addition, late diagnosis can be more frequent in older populations, and low baseline CD4 cells might affects the immunological responses. However, in our study cohort, the highest proportion of those who initiated ART late was in the age <50 years without co-morbidities group, and the median CD4 cell count at baseline was lowest for those age <50 years without comorbidities and those age ≥50 years without co-morbidities. Nevertheless, older patients derive substantial benefit from cART despite having a less robust immunological response than expected given their adherence to therapy and excellent virological responses [27]. cART provides substantial benefit for older and younger HIV-positive patients [33], and older patients are more likely to achieve virological control of HIV replication [34,35] and less likely to develop subsequent virological breakthrough [34], findings that correlate with better adherence to therapy by older patients [35]. Consistent with previous studies, our study showed that older patients had similar virological outcomes compared with younger patients.
Overall, the effects of the age-related comorbidity variables in the main analyses and in the sensitivity analyses remained similar, suggesting that regardless of sex, those aged ≥50 years with comorbidities had worse immunological outcomes than their younger counterparts either with or without comorbidities.
The limitations of the study included the presence missing data. As TAHOD is an observational cohort, data collection depends entirely on the standard of care at each individual site. Patients with good clinic attendance may have more frequent comorbidity testing which may lead to earlier or more frequent diagnosis of a comorbidity. Patients with poor clinic attendance may also have these comorbidities present but not detected. As the cohort does not impose specific study procedures or treatment interventions, the study results should be interpreted with this in mind. The cutoff points for virologic and immunologic failures may not necessarily be relevant for individual patient management, but the failure definition is in line with current WHO guidelines for general clinical practice. In addition, the comorbidity variable was defined according to the availability of our data. We were not able to assess the effects of other comorbidities, as we were limited to the data variables being captured in our cohort. Furthermore, our cohort sites are generally urban referral centres. Patients are selected for enrolment based on the likelihood of remaining in care. Therefore, the generalizability of the reported findings is limited.  25 TREAT Asia, amfAR -The Foundation for AIDS Research, Bangkok, Thailand

C O M P E T I N G I N T E R E S T S
The authors do not have any competing interest to declare. age-related comorbidity and cART adherence are time-updated variables. Missing values were included in the regression analyses, however global p-values were tested for heterogeneity excluding missing categories. Significant p-values are highlighted in bold. Variables not associated with significant p-values are presented in the final table adjusted for the variables with significant p-values. CI, confidence interval; HR, hazard ratio; NRTI, Nucleoside reverse transcriptase inhibitor; NNRTI, Non-nucleoside reverse-transcriptase inhibitor; PI, Protease inhibitor.

SUPPORTING INFORMATION
Additional Supporting Information may be found in the online version of this article: Table S1. Multivariate analyses for factors associated with virological failure in males and females Table S2. Multivariate analyses for factors associated with immunological failure in males and females