Bone Mineral Density, Kidney Function, Weight Gain and Insulin Resistance in Women Who Switch From TDF/FTC/NNRTI to ABC/3TC/DTG

F Ibrahim; A Samarawickrama; L Hamzah; R Vincent; Y Gilleece; L Waters; S Kegg; B Barbini; L Campbell; FA Post


HIV Medicine. 2021;22(2):83-91. 

In This Article


Between May 2016 and March 2018, 102 women were screened, and 91 were randomized and received at least one dose of study drug; 32 were randomly assigned to continue TDF/FTC/NNRTI and 59 were switched to ABC/3TC/DTG (Figure S1). The majority of women were of black ethnicity (86%), with a mean (SD) age of 50.4 (6.6) years, median (IQR) CD4 cell count of 600 (479–800) cells/mm3, a median (IQR) BMI of 29.5 (26.3–32.6) kg/m2, and menstrual history indicating that at least 40 (44%) were post-menopausal. A similar proportion of women in both arms reported use of vitamin D supplements. The NNRTI at baseline was EFV in 59%, NVP in 18% and RPV in 23% (Table 1).

Twelve participants discontinued the trial prior to week 48 (four in the TDF/FTC/NNRTI arm and eight in the ABC/3TC/DTG arm). Treatment-limiting adverse events were experienced by one participant (3.1%) in the TDF/FTC/NNRTI arm (acute kidney injury) and nine (15.3%) in the ABC/3TC/DTG arm [neuropsychiatric (n = 5, of which two with suicidality); hypersensitivity (n = 2); other (n = 2)]; three of these participants continued in the switch arm on a modified regimen (ABC/3TC plus raltegravir, RPV or NVP). In addition, one person in each arm discontinued due to osteoporosis requiring bisphosphonates. No participants developed virological failure (Table S1); CD4 counts increased in the ABC/3TC/DTG relative to the TDF/FTC/NNRTI arm [adjusted mean difference = 74 (95% CI: 7–141) cells/mm3, P = 0.03].

Effect of ART Switch on BMD and Bone Biomarkers

Mean BMD values at baseline at the non-dominant total hip, non-dominant femoral neck and lumbar spine were somewhat higher in participants in the TDF/FTC/NNRTI arm than those in ABC/3TC/DTG arm (Table 2). At week 48, there was a modest but statistically significant increase in mean BMD of the non-dominant total hip [adjusted mean difference = 0.01 (95% CI: 0.002–0.03) g/cm2, P = 0.027] and of the lumbar spine [adjusted mean difference = 0.03 (95% CI: 0.01–0.05) g/cm2, P = 0.002] in those switching to ABC/3TC/DTG as compared with those who remained on TDF/FTC/NNRTI (Table 2; Figure S2). There was no significant change in BMD of the non-dominant femoral neck. Further adjustment for weight gain from baseline to week 48 minimally affected these results (Table S2). Similar results were obtained in a longitudinal analysis that took account of changes in BMD at W24 (Table S3).

We observed no significant change in plasma 25(OH)D or PTH levels between the two study arms through week 48. Serum ALP, a non-specific marker of bone formation, decreased significantly in women who switched from TDF/FTC/NNRTI to ABC/3TC/DTG (Table 2). Reductions in specific markers of bone formation (P1NP) and resorption (CTX) were minimally affected by further adjustment for baseline vitamin D status [mean difference between study arms from baseline to week 48: −2.93 (95% CI: −11.93–6.07), P = 0.517 for P1NP and −0.06 (−0.15–0.03), P = 0.199 for CTX] and only reached statistical significance for CTX in the repeated measures model (Table S3).

Effect of ART Switch on Renal Function

Serum creatinine concentrations increased and eGFR-creatinine decreased in the switch arm, consistent with the known effect of DTG on tubular secretion of creatinine.[14] By contrast, there was no significant change in cystatin C (a protein not cleared by tubular secretion) and eGFR-cystatin C in the participants who switched to ABC/3TC/DTG vs. those who remained on TDF/FTC/NNRTI. Participants who switched from TDF/FTC/NNRTI to ABC/3TC/DTG experienced reductions in total proteinuria (PCR) but not in ACR, RBPCR or fractional excretion of phosphate (Table 2).

Effect of ART Switch on Weight and Insulin Resistance

The mean weight of the participants remained stable in the TDF/FTC/NNRTI arm (86.3 kg at baseline vs. 86.2 kg at week 48) and increased from 77.4 to 79.2 kg in the ABC/3TC/DTG arm (Figure S3), giving a mean adjusted difference of 1.81 (95% CI: 0.03–3.59) between the two study arms (Table 2). No participants in the TDF/FTC/NNRTI arm vs. 19 (35.9%) in the ABC/3TC/DTG arms experienced > 5% weight gain (P < 0.001).

We used stored samples to explore whether the ART switch affected insulin resistance and the prevalence of metabolic syndrome. Participants with diabetes mellitus and those who had non-fasting or missing study visits at baseline or week 48 were excluded (Figure S1). The baseline characteristics of the 53 participants included in the metabolic analyses are shown in Table S4; those in the TDF/FTC/NNRTI arm were slightly heavier (P = 0.056) and had slightly larger waist circumference (P = 0.045) with no between-arms difference for all other measures. At week 48, there was a smaller increase in fasting insulin in the ABC/3TC/DTG arm compared with the TDF/FTC/NNRTI arm (P = 0.033), and this difference persisted after adjustment for weight and ethnicity [mean difference between the two study arms: −3.23 (95% CI: −6.31–0.14), P = 0.041 and further adjustment for differences in baseline vitamin D: −3.51 (−6.82 to −0.21), P = 0.038]. By contrast, we observed no between-arm differences in insulin resistance by any of the three indices (Table 3), including those in the ABC/3TC/DTG arm who experienced > 5% weight gain (Table S5). There was also no significant difference in the proportion of participants with metabolic syndrome over 48 weeks in each arm, including those in the ABC/3TC/DTG arm who experienced > 5% weight gain (Table S6).