Long-Term Clinical, Virological and Immunological Outcomes Following Planned Treatment Interruption in HIV-Infected Children

R Freguja; A Bamford; M Zanchetta; P Del Bianco; C Giaquinto; L Harper; A Dalzini; TR Cressey; A Compagnucci; Y Saidi; Y Riault; D Ford; D Gibb; N Klein; A De Rossi

Disclosures

HIV Medicine. 2021;22(3):172-184. 

In This Article

Results

Main Study

A total of 101 (51 CT, 50 PTI) out of 109 children in the main trial participated in long-term follow-up, including 79 from Europe and 22 from Thailand.[10] Characteristics at trial enrolment were similar by arm (Table S1).

Five-year post-trial follow-up forms were completed for 47 (92%) of the children on CT and 46 (92%) of the children on PTI. Median (IQR) duration of follow-up from trial enrolment was 7.6 (6.9–8.2), a median of 5.1 (4.9–5.3) years after main trial end. Fifteen out of 50 PTI children had experienced two PTIs during the main trial. At the end of the main trial, one CT child was off ART, 43 on PTI re-initiated ART for a median (IQR) duration of 11.8 (7.1–22.8) months, six were undergoing a PTI and one child was off ART.

Clinical Outcomes During 5 Years of Follow-up After the Main Trial. There were no deaths. One PTI child in Thailand had pulmonary TB (this was diagnosed 4.4 years after the end of the trial); there were no other new CDC stage B or C events. There were 33 clinical events diagnosed in 18 (35%) CT children and 32 events in 13 (26%) PTI children. The rates [95% confidence interval (CI)] of events per 100 child-years were 13.1 (9.3–18.5) in the CT arm and 13.0 (9.2–18.4) in the PTI arm (rate ratio comparing PTI vs. CT of 0.91; 95% CI: 0.40–2.09; P = 0.83), with similar events in the two arms. Events included infections/infestations (n = 13 total), gastrointestinal disorders (n = 9), respiratory illness (n = 6), blood lipid abnormalities (n = 6), anaemia/thrombocytopenia (n = 5) and others (n = 26).

Laboratory Outcomes During 5 Years of Follow-up After the Main Trial. Immunology and virology results from the main study are summarized in Figure S1. At the end of the main trial, PTI children were less likely to have undetectable plasma viraemia (HIV-1 RNA < 50 copies/mL) than CT children (65% vs. 87%; P = 0.02; Table 1); by 2 years the proportions were similar (84% vs. 88%; P = 0.57). Based on a mixed model including all measures, proportions suppressed were predicted to be the same in the two arms by 4.9 years (95% CI: −0.7–∞) from the end of the main trial.

Mean CD4% was lower in PTI children at end of main trial [adjusted difference (PTI – CT) = −4.9%, 95% CI: −7.5 to −2.4; P < 0.001). By 2 years, the difference between arms was borderline (−1.8%, 95% CI: −4.6–1.0; P = 0.20), but the mixed model suggested the difference persisted beyond 2 years with mean values in the two arms predicted to be the same at 7.2 years (95% CI: 2.7–34.9). Results were similar for absolute CD4 count (Table 2) with the mean values in the two arms predicted to be the same at 5.0 years (95% CI: 2.2–14.0). Mean CD8% and CD8 were higher in PTI children at the end of the main trial and there was some suggestion that differences persisted longer for CD8% and absolute CD8 count (Table 1 and Table 2); although mean values in the two arms were predicted to be same at 11.0 and 18.3 years respectively, CIs did not preclude CD8% and CD8 count in the PTI arm not returning to the levels in the CT arm.

The CD4/CD8 ratio is an additional important marker of immune competence.[18–20] The mean CD4/CD8 ratio was lower at the end of the main trial: 64% of PTI children had CD4/CD8 ratio < 1 compared with 34% of the CT children (P = 0.003); by 5 years the corresponding proportions were 46% vs. 29% (P = 0.08). Mean CD4/CD8 ratios at 5 years post-trial were 1.22 and 1.08 in CT and PTI arms, respectively [difference (CT – PTI) = −0.15; 95% CI: −0.34–0.05; P = 0.14) (Table 1). In both arms the CD4/CD8 ratio at 5 years was strongly associated with the CD4/CD8 ratio at the end of the main trial: in the PTI arm, 12/20 participants with CD4/CD8 ratio < 1 at the end of the trial remained < 1 at 5 years, and only 3/14 participants with CD4/CD8 ratio ≥ 1 at the end of the trial were < 1 at 5 years (Fisher's exact for association, P = 0.03); corresponding proportions in the CT arm were 6/10 and 4/23 (P = 0.01). The mixed model predicted CD4/CD8 ratio to be the same in the two arms at 7.4 years (95% CI: 3.6–18.6) (Table 2).

Predictors of Earlier CD4/CD8 Recovery After Restarting Antiretroviral Therapy. Following re-initiation of ART after the last PTI, CD4/CD8 recovery was strongly associated with duration of ART since re-initiation and baseline CD4/CD8 ratio (Table 3). After adjustment for these predictors, the only factor that remained significantly associated with higher CD4/CD8 in the multivariable model was the minimum CD4/CD8 ratio before ART re-initiation. Although most recent HIV-1 RNA plasma viraemia (fitted as a surrogate for treatment adherence) was associated with CD4/CD8 recovery in the multivariable model, including it had a minimal effect on the impact of other predictors.

Models fitted for CD4% and CD8% recovery resulted in similar conclusions; in the corresponding multivariable model significant predictors (P < 0.05) of CD4% (CD8%) included duration of ART since re-initiation, baseline CD4% (CD8%) and lowest CD4% (highest CD8%) before ART re-initiation (data not shown).

Sub-study

Fifty-four (23 CT, 31 PTI) children were included in the sub-study. There were few differences at trial baseline between the children in the sub-study and the remainder: ethnicity differed primarily because Thai children did not participate; baseline weight-for-age and years on ART were marginally higher in sub-study participants (Table S2). Of 31 PTI children, 20 underwent one interruption, and 11 underwent two interruptions. Median (IQR) follow-up from trial enrolment in sub-study participants was 7.4 (3.7–8.5) years.

Over 5 years of post-trial follow-up, the difference in HIV-1 plasma viraemia between PTI and CT was significant (P = 0.011). Sub-study PTI children were less likely to be suppressed at the end of the main trial (57%, of PTI vs. 90% of CT had undetectable plasma viraemia; P = 0.005). This persisted at 1 year (65% PTI vs. 90% CT; P = 0.027), but differences thereafter were non-significant (Figure 1a). HIV-1 DNA levels were higher in the PTI arm at end of main trial [mean (95% CI) of 269 (124–414) copies/106 PBMCs in PTI vs. 162 (0–334) copies/106 PBMCs in CT]; levels thereafter remained higher, although at non-significant level, in PTI children (Figure 1b). Intracellular HIV-1 RNA levels were higher in PTI than in CT children at the end of the main trial (P = 0.012), and thereafter (at 3 and 4 years), but non-significantly different at 5 years (Figure 1c).

Figure 1.

Trends in HIV-1 plasma viraemia, HIV-1 DNA and intracellular HIV-RNA during 5 years of long-term follow-up (LTFU) in children on continuous antiretroviral (ART) therapy (CT) and those undergoing planned treatment interruption (PTI) of ART. (a) Proportion of children with undetectable HIV-1 plasma viraemia; (b) levels of HIV-1 DNA; (c) levels of intracellular HIV-1 RNA. **P < 0.05; *P < 0.10; P-values in panel: overall difference between PTI and CT groups. Whiskers represent 95% confidence interval of mean values.

CD4 and CD8 Cell Phenotypes. The percentage of CD4 cells was significantly lower in PTI than in CT (P = 0.012) at the end of the main trial. This difference persisted during long-term follow-up (overall, P = 0.046), but by 5 years was not statistically significant (Figure 2a). This reduction in the proportion of CD4 cells was due to a reduction in the CD4 memory cells, which were significantly depleted in the PTI group at the end of the main trial (P = 0.027), persisting throughout the study period and at 5 years (overall, P < 0.001; Figure 2c). The percentages of CD4-naïve cells did not differ significantly between PTI and CT (Figure 2b). At the end of the main trial, the PTI group had a significantly higher percentage of CD8 cells than did the CT group, and percentages of CD8 cells remained higher in PTI than in CT throughout follow-up (overall, P = 0.031). Both naïve and memory CD8 cells contributed to higher CD8% in the PTI children (Figure 2e-f).

Figure 2.

Trends in percentage of CD4, CD8 and their respective memory and naïve subpopulations during 5 years of long-term follow-up (LTFU) in children on continuous antiretroviral (ART) therapy (CT) and those undergoing planned treatment interruption (PTI) of ART. (a–c) Percentage of CD4 (a), CD4-naïve (b) and CD4 memory (c) cells in CT and PTI children. (d–f) Percentage of CD8 (d), CD8-naïve (e) and CD8 memory (f) cells in in PTI and CT children. **P < 0.05; *P < 0.10; P-values in panel: overall difference between PTI and CT groups. Whiskers represent 95% confidence intervals of mean values.

Mean CD4/CD8 ratio was lower at the end of the main trial in PTI than in CT, and 64% of the PTI arm had CD4/CD8 ratio < 1 compared with 27% of the CT arm; by 5 years corresponding proportions were 36% vs. 11%. PTI children with CD4/CD8 ratio < 1 at each time point had a higher proportion of detectable plasma viraemia than did PTI children with CD4/CD8 ratio > 1 and CT children (Figure 3a); they also had higher levels of intracellular HIV-1 RNA compared with CT children (Figure 3b). Children with CD4/CD8 ratio < 1 had significantly lower CD4 memory cells and higher naïve and memory CD8 cells compared with CT children (Figure 3).

Figure 3.

Trends in children on continuous antiretroviral (ART) therapy (CT) and those undergoing planned treatment interruption (PTI) of ART subgrouped by CD4/CD8 ratio during 5 years of long-term follow-up (LTFU). (a) Proportion of CT (grey diamond) and PTI children with CD4/CD8 ratio < 1 (PTI < 1), and PTI children with CD4/CD8 ratio >1 (PTI > 1) with undetectable HIV-1 plasma viraemia. (b) Levels of intracellular HIV-1 RNA in CT, PTI < 1 and PTI > 1. (c–f) Percentage of CD4-naïve cells (c), CD4 memory cells (d), CD8-naïve cells (e) and CD8 memory cells (f) in CT, PTI < 1 and PTI > 1. P-values in panel: paired comparisons between groups.

Thymic Function and Immune Activation. Levels of TRECs did not differ between PTI and CT children either at the end of the main trial or during follow-up (Figure S2a). A trend towards increased expression of CD38 on CD8 cells, a marker of immune activation, was observed in PTI children during therapy interruption,[9] which persisted at 1 year but then decreased, with the two groups showing comparable levels during follow-up (Figure S2b).

Comments

3090D553-9492-4563-8681-AD288FA52ACE

processing....