Background
We investigated the risk of liver stiffness progression and advanced liver fibrosis (LF) by diabetes mellitus (DM) status among people with HIV (PWH) and metabolic-dysfunction associated stetotic liver disease (MASLD).
Methods
PWH who had MASLD at their first vibration-controlled transient elastography or VCTE with both liver stiffness measurement (LSM) and controlled attenuation parameter (CAP) available (defined as baseline) and had >1 LSM over follow-up were included. MASLD was defined as the presence of liver steatosis (CAP ≥248 dB/m) with at least one cardiometabolic condition. Change in LSM from baseline (ΔLSM) was evaluated using a linear mixed-effects model, adjusting for age, sex, duration since DM diagnosis, body mass index (BMI), ALT and AST all fixed at baseline, time since first VCTE and time-varying exposure to INSTIs. An interaction term between time and baseline DM was included to assess the effect of DM on ΔLSM. Multivariable Poisson regression was used to evaluate the association between baseline DM and the incidence of LF (defined as LSM ≥8kPa).
Results
Among 345 PWH with MASLD included (35% female), median age was 48 (IQR, 43-54) years, median BMI was 25 (23-28) kg/m2 and median CD4 was 662 (500-826) cells/µL at baseline. 94% had HIV RNA <50 copies/mL and median antiretroviral therapy (ART) duration was 15 (7-18) years. Ninety-seven (28%) had DM at baseline, with a median DM duration of 16 (13-18) years. PWH with DM at baseline had higher FIB-4 (0.92 [0.69-1.26] vs. 0.76 [0.59-1.01]) and LSM (6.6 [5.4-8.2] vs. 5.5 [4.6-7.2] kPa). The median duration between the first and last VCTE measurements was 47 (42-61) months. In adjusted analysis, LSM declined over time with an average of -0.15 kPa/year (95%CI -0.28, -0.01, p=0.035). The ΔLSM over time was not associated with baseline DM (P-interaction=0.40). However, among 260 PWH-MASLD without LF at baseline, the risk of LF (incidence rate 3.7 [2.6-5.2]/100 person-years) was more than three times higher in those with baseline DM (adjusted incidence risk ratio [aIRR]: 3.35, 95%CI 1.67-6.75, P=0.001). There was no strong evidence suggesting the association between baseline DM and advanced LF risk differed by INSTI exposure (P-interaction=0.54).
Conclusions
Despite small decreases in LSM during follow-up, PWH with both MASLD and DM have a significantly higher risk of liver fibrosis. Prioritizing this population for treatments interventions may help mitigate liver disease progression.