# Trends in Serum Amylase Levels in People Living with HIV: A Comparison Between INSTI and NNRTI/PI-Based Regimens

## TL;DR

This study found that INSTI-based HIV treatments lead to lower serum amylase levels compared to older regimens, suggesting better pancreatic safety.

## Contribution

The study is the first to compare INSTI-based and NNRTI/PI-based ART regimens specifically regarding serum amylase levels.

## Key findings

- INSTI-treated patients had significantly lower serum amylase levels than NNRTI/PI-treated patients at all timepoints.
- The difference in amylase levels between the two groups showed a significant linear time-by-group interaction.
- No clinical pancreatitis was observed, and all amylase variations were asymptomatic and biochemical.

## Abstract

Background: Antiretroviral therapy (ART) has transformed HIV into a chronic manageable condition, yet metabolic toxicities including pancreatic enzyme alterations remain concerns. While older ART regimens have been associated with hyperamylasemia, the impact of integrase strand transfer inhibitor (INSTI)-based therapies on serum amylase levels has not been specifically examined. Purpose: This study aimed to compare longitudinal patterns of serum amylase levels between people living with HIV receiving INSTI-based versus NNRTI/PI-based ART regimens. Methods: This retrospective observational study analyzed 99 HIV-positive patients at Kaplan Medical Centre, Israel (2002–2023). Participants received either INSTI-based (n = 49) or NNRTI/PI-based (n = 50) regimens for ≥24 months. Serum amylase, viral load, CD4 counts, and metabolic parameters were measured at baseline, one year, and two years. Repeated-measures ANOVA assessed longitudinal changes. Results: NNRTI/PI-treated patients maintained significantly higher mean amylase levels throughout follow-up (baseline: 122.9 ± 42.1 U/L; two years: 129.6 ± 38.0 U/L) compared to INSTI-treated patients (baseline: 78.7 ± 32.3 U/L; two years: 68.4 ± 23.4 U/L; p < 0.0001 at all timepoints). A significant linear time-by-group interaction (p = 0.037) demonstrated divergent trajectories. No clinical pancreatitis was observed in either treatment group during the follow-up period, and all observed variations in serum amylase were biochemical and asymptomatic. While these findings are reassuring regarding acute pancreatic toxicity, the clinical significance of chronic subclinical enzyme elevations remains uncertain. Conclusion: INSTI-based antiretroviral regimens suggest a favorable pancreatic and metabolic safety profile compared with NNRTI/PI-based therapies.