Characteristics of participants in this study
In total, 2155 participants were selected from a long-term, first-line ART regimen, local Chinese cohort (Figure 1). The patients in this study were followed up to 8348 person-years in total, and with the median follow up time of 3.4 years (IQR: 2.3–5.0). The majority of the participants were male (92.9%, 2001/2155), with a median age at HIV diagnosis of 37 years old (IQR: 31–49). They mainly comprised of the Han nationality (85.4%). Further, 79.2% (1706/2155) of the participants were infected with HIV-1 through homosexual transmission, 65.1% (1403/2155) were infected with CRF01_AE HIV-1, and 68.5% (1477/2155) of participants used TDF+3TC+EFV as the ART regimen. Most participants (50.9%, 1096/2155) showed the zenith baseline VL of 4–5 (log copies/ml) and 37.4% (805/2155) of participants displayed the nadir baseline CD4+T cell counts of 200–350 cells/mm3 (Supplement Table 1).
In this study, 38.7% (835/2155) of participants experienced at least one episode of LLV. In terms of the LLV level, participants belonging to LLLV, MLLV, HLLV, and HBL groups accounted for 52.1% (435/835), 18.0% (150/835), 14.3% (119/835), and 15.7% (131/835), respectively. In terms of the duration of LLV, participants with Blip, LLV lasting for 3–6 months, 6–12 months, and above 12 months accounted for 72.1% (602/835), 11.4% (95/835), 11.4% (95/835), and 5.1% (43/835), respectively (Supplement Table 1).
High risk LLV contributes to VF
The VF rate increased along with increasing level of LLV (50–200 copies/ml: 1.6%, 200–400 copies/ml: 4.7%, 400–1000 copies/ml: 13.4%, Figure 2A) and the duration of viremia (Blip: 3.3%, 3–6 months: 7.4%, 6–12 months: 6.3%, >12 months: 11.6%, Figure 2B). On the basis of the estimated Kaplan–Meier curves, there were significant differences in the risk of VF (p < 0.01) between LLV level and duration categories. Compared with the control group (incidence rate (IR): 0.52/100 person years), those with > 400 copies/ml (IR: 2.97/100 person years, p < 0.01) and LLV duration of > 12 months (IR: 3.68/100 person years, p < 0.01) had the least favorable Kaplan–Meier curves and higher risk of VF (Figure 2C, D). The Cox proportional hazards model showed that only HLLV (HR: 5.93, 95%CI: 3.13–11.23, p < 0.01) and HBL (HR: 2.84, 95%CI: 1.27–6.34, p = 0.011) were associated with significantly increased risk of subsequent VF compared with the control group (all VL below 50 copies/ml).
Based on LLV duration analyses, we found that the risk of VF increased in the HLLV group for any duration; moreover, the HRs increased with prolonged duration (Blip: HR = 4.41, 95%CI: 1.96–9.92, p < 0.01; 3–6 months: HR = 6.59, 95%CI: 1.94–22.34, p < 0.01; 6–12 months: HR = 5.19, 95%CI: 1.22–22.18, p < 0.05; and > 12 months: HR = 9.63, 95%CI: 2.82–32.95, p < 0.01). In addition, any level LLV (> 3 months) coupled with HBL also showed increased risk of VF (3–6 months: HR = 6.20, 95%CI: 1.46–26.27, p < 0.05; 6–12 months: HR = 9.74, 95%CI: 2.82–33.66, p < 0.01; > 12 months: HR = 7.52, 95%CI: 1.01–55.84, p < 0.05). In contrast, no significant increase of VF was observed among MLLV and LLLV groups, even in those that lasted for > 12 months (200–400 copies/ml: HR = 3.43, 95%CI: 0.46–25.54, p > 0.05; 50–200 copies/ml: HR = 1.02, 95%CI: 1.01–1.03, p > 0.05).
In this study, VL was determined at 3–6 months interval and acted as a time-varying covariate. We further used the linear mixed-effect model to validate the contribution of LLV on VF35,36. Our results supported the contribution of HLLV and HLB to VF, but the MLLV group showed weak association with increased risk of VF (estimated 0.03, 95%CI: 0.00–0.06, p = 0.045). LLLV showed no increase of VF in both models (Table 1).
Factors associated with high risk LLV
We further assessed the socio-demographic information, transmission route, clinical data, and laboratory test results to explore the factors associated with high risk LLV. Using chi-square tests, we found that the different groups of zenith baseline VL, nadir CD4+T cell counts, and other demographic information, clinical data, and laboratory tests had varied frequencies of high risk LLV (Table 2). Multiple logistic regression analysis further supported that participants with higher zenith baseline VL (> 6 log copies/ml vs. < 4 log copies/ml: aOR = 3.49, 95%CI: 1.55–7.83, p < 0.01) and lower nadir baseline CD4+T cell counts (< 200 cells/mm3 vs. > 350 cells/mm3: aOR = 1.78, 95%CI: 1.14–2.78, p < 0.05) showed increased risk of high-risk LLV. Other associated factors included Manchu (aOR: 2.03, 95%CI: 1.27–3.25, p < 0.01), ART for over 60 months (aOR: 1.81, 95%CI: 1.21–2.69, p < 0.01), receiving AZT+3TC+NVP regimen (aOR: 2.26, 95%CI: 1.48–3.45, p < 0.01) or DDI+NRTI+NNRTI regimen (aOR: 9.96, 95%CI: 2.33–42.49, p < 0.01) and subtype B’ infection (aOR: 8.22, 95%CI: 2.34–28.92, p < 0.01).