Of 466 individuals with a positive HBsAg and active follow-up, 224 switched from TDF to TAF. After excluding 8 individuals with a HIV viral load above 200 cp/mL and 110 individuals with an eGFR ≥90 mL/min/1.73 m2, 106 individuals formed the study population. Sixteen (15.1%) participants were women, 20 (18.9%) were of African origin, and the median age was 53 years (interquartile range [IQR] 48–58, Table 1). Baseline eGFR was between 60 and 89 mL/min/1.73 m2 in 84 (79.2%) and below 60 mL/min/1.73 m2 in 22 (20.8%) individuals. Patients with an eGFR below 60 mL/min/1.73 m2 were older, more likely to receive a treatment for arterial hypertension, and had lower CD4 cell counts and higher urinary protein-to-creatinine ratios compared with individuals with an eGFR of 60–89 mL/min/1.73 m2. The median follow-up time on TAF was similar between both groups [14.6 months (IQR 12.5–16.2) in individuals with an eGFR of 60–89 mL/min/1.73 m2, and 14.1 months (IQR 12.4–16.9) in those with an eGFR below 60 mL/min/1.73 m2]. Sixty-three patients (59.4%) had only TDF replaced by TAF, the rest of the ART regimen remaining unchanged. The most common reasons for stopping TDF were the prevention of toxicity (37.1%), established renal toxicity (17.1%), and treatment simplification (11.4%).
Changes in eGFR and Urine Protein-to-creatinine Ratio
Adjusted mean changes in eGFR are shown in the Figure 1A. Individuals with an eGFR between 60 and 89 mL/min/1.73 m2 experienced changes in eGFR of −1.9 mL/min/1.73 m2 in the last year on TDF [95% confidence interval (CI) −3.3 to −0.5], which increased to 3.2 mL/min/1.73 m2 1 year (95% CI 1.2 to 5.2) after switching to TAF (P value for slope difference 0.001). The use of TDF in individuals with an eGFR below 60 mL/min/1.73 m2 led to a decline in eGFR (−4.9 mL/min/1.73 m2 in the year before switch, 95% CI −7.5 to −2.2), and switching to TAF was associated with improvements in eGFR (6.2 mL/min/1.73 m2 1 year after the switch, 95% CI 2.4 to 10.0, P value for slope difference <0.001).
Adjusted changes in eGFR and ALT levels in the year before and the year after switching from TDF to TAF. (A) Changes in eGFR according to the multivariable model stratified by eGFR at baseline (60–89 mL/min/1.73 m2 vs. below 60 mL/min/1.73 m2). Adjusted for age, sex, ethnicity, diabetes, treatment for arterial hypertension, and the time-updated use of integrase inhibitors, atazanavir, lopinavir, rilpivirine, cobicistat, and cotrimoxazole. (B) Changes in ALT levels according to the multivariable model stratified by ALT levels at baseline (normal vs. elevated*). Adjusted for baseline age, sex, body mass index, transmission risk (persons who inject drugs, men who have sex with men, and others), HCV infection, HBV replication at time of switch, and time-updated for hazardous alcohol consumption and efavirenz use. *ALT >25 for women and >35 for men. The bold line represents predicted mean values, the shaded area its 95% confidence intervals, and thin lines represent individual patient profiles. The dashed line represents the time of switch from TDF to TAF.
One year after switching to TAF, 58 (54.7%) individuals had urine protein-to-creatinine ratio values available, and the main demographic and clinical characteristics, including eGFR at baseline, were similar between the group with and without follow-up data. After adjusting for covariates, the use of TDF was associated with a change in urine protein-to-creatinine ratio of 0.9 mg/mmol in the last year before switch (95% CI −2.5 to 4.3), and switching to TAF led to a change in urine protein-to-creatinine ratio of −6.3 mg/mmol 1 year after the switch (95% CI −10.0 to −2.7, P-value for slope difference 0.01). Changes in urine protein-to-creatinine ratio after 1 year did not correlate with changes in eGFR over the course of time (correlation coefficient −0.02, P = 0.91). Our findings on eGFR and urine protein-to-creatinine ratio remained similar after restricting the analyses to individuals who had the replacement of TDF by TAF as the only ART change (see Table S1, Supplimental Digital Content, https://links.lww.com/QAI/B499).
HIV and HBV Outcomes and Changes in ALT Levels
One year after switching from TDF to TAF, 97 of 98 (99.0%) individuals with available HIV RNA follow-up measurements had a suppressed viral load (<200 cp/mL). The HBV viral load was suppressed (<50 IU/mL) in 64 of 66 (97.0%) individuals with available follow-up measurements. One individual with detectable HIV and HBV viral loads reported suboptimal adherence, and the other individual with detectable HBV DNA did not have a suppressed HBV viral load at the time of switch. HBsAg loss was observed in 4 (8.7%) of the 42 individuals with HBsAg measurements available 1 year after switching to TAF.
Thirty-six individuals (34.0%) had elevated ALT levels at baseline, 14 (13.2%) had significant fibrosis, and 6 (5.7%) had liver cirrhosis. Among those with elevated ALT levels, one individual (4.2%) had a detectable HBV viral load at baseline, compared with 3 (4.2%) among those with normal ALT levels. In adjusted analyses individuals with normal ALT levels at baseline experienced ALT changes of −1.3 IU/L (95% CI −3.8 to 1.2) in the year before and of −0.1 IU/L (95% CI −3.6 to 3.5) in the year after switching to TAF. In individuals with elevated ALT levels, the use of TDF was associated with increases in ALT of 5.9 IU/L per year (95% CI 2.1 to 9.8), whereas they experienced marked decreases of −11.8 IU/L per year (95% CI −17.3 to −6.4) after the switch to TAF (Figure 1B). EFV was the third drug used in 5 individuals (7.1%) with normal ALT levels and in 6 individuals (16.6%) with elevated ALT levels at baseline. In addition to switching to TAF, all but one of them received a non-EFV containing regimen. Restricting the analyses to individuals without EFV use or to individuals who only had TDF replaced by TAF did not change our findings (see Table S2, Supplemental Digital Content, https://links.lww.com/QAI/B499).
J Acquir Immune Defic Syndr. 2020;85(2):227-232. © 2020 Lippincott Williams & Wilkins