Immunological and Virological Response to Initial Antiretroviral Therapy Among Older People Living With HIV in the Canadian Observational Cohort (CANOC)

A Zhabokritsky; L Szadkowski; AN Burchell; C Cooper; RS Hogg; M Hull; DV Kelly; M Klein; M Loutfy; A McClean; J Montaner; SL Walmsley

Disclosures

HIV Medicine. 2021;22(8):759-769. 

In This Article

Results

Of 13 040 CANOC participants initiating ART between 1 January 2000 and 31 December 2016, 12 783 (98%) initiated treatment with at least two NRTIs and an antiretroviral agent from another class. Of these, 12 489 (98%) had at least one post-treatment CD4 count or viral load and were included in this analysis. Demographic and clinical characteristics at treatment initiation are presented by age group in Table 1. Overall, median (IQR) age was 40 (32–47) years with 12.6% of participants in the 51–60 years age group and 3.4% in the over-60 group. Participants in age groups over 40 were more likely to be male (82–86% vs. 73–77%) and White (66–73% vs. 52–54%) than younger participants. Participants in the oldest age category (> 60 years) were less likely to have MSM (47% vs. 54–59%) and IDU (12% vs. 23–31%) as risk factors relative to younger groups. Participants ≤ 30 years had the highest pre-treatment CD4 count (median 310 vs. 224–250 cells/μL) and lowest viral load (median 4.69 vs. 4.81–5.00 log10copies/mL). Participants in the middle age category (41–50 years) had the greatest follow-up time [median (IQR) = 6 (3–9) years], whereas those ≤ 30 and > 60 had the least [4.4 (1.8–7.6) and 4.6 (1.7–7.7) years respectively].

Immunological Response

Overall, 12 278 (98%) participants had at least one follow-up CD4 count recorded in the database after treatment initiation. The number of CD4 count measurements per year increased with participant age at time of treatment initiation; participants ≤ 30 years old had 3.4 (2.5–4.5), those aged 41–50 had 3.7 (2.8–4.8), and those over 60 had 4.5 (3.3–5.8) measurements per year. In a univariable GEE model, age groups 41–50, 51–60 and > 60 years were all associated with lower absolute CD4 counts following ART initiation compared with the youngest age group (β = −51, −59 and −69 cells/μL, respectively) (Table 2). An age–time interaction was not significant and so both variables were included only as main effects in the multivariable model. After adjusting for covariates, including pre-treatment CD4 count, only the oldest age group (> 60) had lower absolute CD4 counts over time compared with the youngest group (β = −31 cells/μL). Injection drug use and a history of hepatitis C were associated with lower CD4 counts over time in a multivariable model, whereas MSM (compared with male, not MSM) and higher pre-treatment CD4 count were associated with higher CD4 counts following treatment initiation.

In an analysis of 4533 participants who had a pre-treatment CD4 count < 200 cells/μL, 3648 (80%) achieved a CD4 count ≥ 200 cells/μL during follow-up, 247 (5.4%) died, and 638 (14%) were censored. There was no difference in time to CD4 count ≥ 200 cells/μL by age group (Figure 1a). Overall, the probabilities of achieving a CD4 ≥ 200 cells/μL by 1 and 5 years were 0.61 [95% confidence interval (CI): 0.6–0.63] and 0.88 (95% CI: 0.87–0.89) respectively. After controlling for covariates, age > 60 years was associated with a lower likelihood of achieving a CD4 ≥ 200 cells/μL compared with age ≤ 30 years [adjusted hazard ratio (aHR) = 0.76, 95% CI: 0.6–0.96] (Table 3).

Figure 1.

Cumulative incidence functions of immunological response by age with a competing risk of death. (a) Time to CD4 count ≥ 200 cells/μL (n = 4533). (b) Time to CD4 count ≥ 500 cells/μL (n = 10 378).

Of 10 378 participants with a pre-treatment CD4 count ≤ 500 cells/μL, 6088 (59%) achieved a CD4 count ≥ 500 cells/μL during follow-up, 600 (6%) died, and 3690 (35%) were censored. Figure 1(b) demonstrates the difference in time to CD4 recovery to 500 cells/μL by age group and a stratified analysis (according to pre-treatment CD4 count) is shown in Figure S1. Participants ≤ 30 were most likely to recover with a probability of doing so by 5 years of 0.7 (95% CI: 0.68–0.73) compared with 0.53 (95% CI: 0.48–0.59) for participants > 60 years. After adjusting for the covariates including pre-treatment CD4 count, all older age groups were less likely to achieve a CD4 count ≥ 500 cells/μL compared with those ≤ 30 years old (Table 3); the adjusted hazard ratios for the middle age groups ranged from 0.87 (95% CI: 0.79–0.97) to 0.91 (95% CI: 0.83–0.98) while the over-60 group had a hazard ratio of 0.69 (95% CI: 0.57–0.84).

In multivariable PH models, IDU and a history of hepatitis C were both independently associated with a lower likelihood of achieving each CD4 outcome, and MSM was associated with a higher likelihood of CD4 recovery (Table 3). Being female was associated with a greater likelihood of achieving a CD4 count ≥ 500 cells/μL (vs. non-MSM men), but was not associated with achieving a CD4 count ≥ 200 cells/μL after adjusting for other factors. Participants starting on NNRTI-based regimens were less likely to achieve CD4 recovery to ≥ 500 cells/μL than those starting on INSTI-based regimens (aHR = 0.87, 95% CI: 0.77–0.98), but otherwise there were no associations between regimen and CD4 recovery outcomes.

Virological Response

Ninety-eight per cent (n = 12 262) of participants had at least one follow-up viral load and were included in the analysis of time to virological suppression. Participants > 60 years old had more viral load measurements per year [median (IQR) = 4.5 (3.5–5.7)] compared with younger participants [4.1 (3.3–5.2) for 51–60 age group, 3.8 (3.0–4.8) for 41–50 age group, 3.7 (2.9– 4.8) for 31–40 age group, and 3.7 (2.8–4.9) for those ≤ 30 years old; P < 0.001). Virological suppression was achieved by 10 770 (88%) participants during follow-up. A total of 1160 (9%) participants were censored and 332 (3%) died before achieving virological suppression. While there were some differences in time to suppression by age group (Figure 2a), these differences were not statistically significant (P = 0.08). Overall, the probability of achieving suppression by 6 months was 0.62 (95% CI: 0.61–0.63) and increased to 0.8 (95% CI: 0.79–0.81) by 1 year. After adjusting for covariates in a multivariable model, increasing age was associated with a greater likelihood of achieving suppression (aHR values ranged from 1.09 to 1.21 with P < 0.01 for older age groups compared with those ≤ 30 years) (Table 4). This effect was also seen when considering age as a continuous variable (aHR = 1.07, 95% CI: 1.05–1.09, P < 0.0001, per 10-year increase in age). MSM and starting on an INSTI-based regimen were also associated with a higher likelihood of suppression, whereas being female (vs. non-MSM men), IDU, having a higher pre-treatment viral load, and having a history of hepatitis C were associated with a lower likelihood of suppression.

Figure 2.

Cumulative incidence functions of virological response by age with a competing risk of death. (a) Time to virological suppression (n = 12 262). (b) Time to virological failure (n = 10 770).

Of the 10 770 participants who achieved virological suppression, 1632 (15%) had virological failure to > 200 copies/mL, 431 (4%) died during follow-up, and 8707 (81%) were censored. Participants > 60 years old were least likely to fail after suppression (Figure 2b) with a probability of doing so by 5 years of 0.08 (95% CI: 0.05–0.12) compared with 0.1 (0.09–0.12) for 51–60 years, 0.15 (0.14–0.16) for 41–50 years, 0.18 (0.16–0.19) for 31–40 years, and 0.21 (0.19–0.23) for ≤ 30 years (P < 0.0001). Increasing age was also associated with a lower likelihood of failure in PH models (Table 4) with the greatest difference between those aged 51–60 years (aHR = 0.45, 95% CI: 0.36–0.56) and > 60 years (aHR = 0.46, 95% CI: 0.29–0.71) compared with ≤ 30 years. When age was included as a continuous variable, the aHR (95% CI) was 0.78 (0.73–0.82) per 10-year increase in age (P < 0.0001). In a multivariable model, being female (compared with non-MSM men), IDU and having a history of hepatitis C were associated with greater likelihood of virological failure; MSM was associated with a lower likelihood of failure.

Comments

3090D553-9492-4563-8681-AD288FA52ACE

processing....