CROI 2020 Abstract eBook

Abstract eBook

Poster Abstracts

Background: Protease inhibitors (PIs) cause drug-drug interactions (DDIs) with statins due to inhibition of drug metabolizing enzymes and/or the hepatic uptake transporter OATP1B1, which may alter the pharmacodynamic (PD) response to statins. There is a lack of data on real-life management of DDIs between antiretrovirals (ARVs) and statins. Methods: Patients of the Swiss HIV Cohort Study followed-up in the centres of Lausanne and Basel were eligible if they received a statin concomitantly to ARVs. Low-density lipoprotein (LDL), total cholesterol (TC) and plasma concentration of the statin were measured during a follow-up visit. Individual LDL target values were set using the Framingham score whereas TC target values were set according to the 2018 European AIDS Clinical Society recommendations. Statins concentrations were interpreted using published plasma concentration time curves. DDIs management was evaluated based on the statin dose adjustment considering coadministered ARVs and the PD response on the lipid profile. Results: Data were collected for 99 rosuvastatin, 93 atorvastatin, 46 pravastatin and 21 pitavastatin. DDIs management and PD response varied according to the statin (figure 1). Statin underdosing leading to suboptimal PD response was frequent with rosuvastatin and atorvastatin. However, the lipid target values were not always achieved in presence of PIs despite using the maximal recommended rosuvastatin dose. Similarly, suboptimal lipid control was observed with PIs despite high atorvastatin concentrations likely explained by inhibition of OATP1B1 resulting in less statin uptake in the liver, the site of action. Target lipid values were more often achieved with unboosted integrase inhibitors due to both their favourable DDIs profiles and neutral effect on lipids. Underdosing was less frequent with pravastatin and pitavastatin, nevertheless suboptimal lipid control was common regardless of coadministered ARVs and despite using maximal recommended pravastatin and pitavastatin doses. This is likely due to their lower efficacy compared to rosuvastatin or atorvastatin. Conclusion: Suboptimal management of DDIs with statins underdosing was observed in overall 30% of cases. Management of dyslipidemia in patients on PIs is challenging due to this ARVs class negative impact on lipid profile and DDIs potentially impairing the effect of statins. Integrase inhibitors based regimens and/or treatment with rosuvastatin or atorvastatin should be favoured in patients with refractory dyslipidemia.

DDIs in young (20-50 years) adults and aging PLWH (55-80 years) was verified against clinical data for amlodipine (AML, 10mg QD) and rosuvastatin (ROS, 10mg QD) both being administered with darunavir/ritonavir (DRV/r, 800/100 mg QD). The clinical data were obtained in the framework of a Swiss HIV Cohort Study project enrolling PLWH older than 55 years or from publications. The verified PBPK model was used to conduct virtual clinical trials for 15 DDIs involving ARVs in virtual individuals aged 20 to 99 years. DDI magnitudes were normalized to the youngest investigated age group. Pearson’s correlation was performed to analyse age-related changes of DDI magnitudes. Results: Clinical data for AML and ROS in combination with DRV/r were within the 95% confidence interval (CI) of the predictions for young individuals (20-50 years) and aging PLWH (55-80 years). DDI magnitudes were always predicted within 1.25-fold of clinical data (Tab. 1). Predicted magnitudes of the 15 investigated DDIs (10 inhibitions and 5 inductions) using the verified PBPK model did not change with age. The calculated correlation coefficient of the AUC-ratio [95% CI] was -0.23 [-0.65 0.30] with a p-value of 0.40. Conclusion: PBPK modelling in combination with limited clinical data demonstrated that DDI magnitudes with ARVs appear not to be impacted by aging. Thus, in the absence of severe comorbidities, management of DDIs can be similar in elderly compared to young PLWH. 449 PLASMA & INTRACELLULAR PK AND RENAL SAFETY OF TAF 25MG WITH BOOSTED PI AND LDV/SOF Kristina M. Brooks 1 , Jose R. Castillo-Mancilla 1 , Mary Morrow 1 , Samantha MaWhinney 1 , Joshua Blum 2 , David L. Wyles 2 , Jia-Hua Zheng 1 , Bethany M. Johnson 1 , Joe Gomez 1 , Ye Ji Choi 1 , Francesca Cendali 1 , Hannah Haas 1 , Lane R. Bushman 1 , Peter L. Anderson 1 , Jennifer J. Kiser 1 1 University of Colorado Anschutz Medical Campus, Aurora, CO, USA, 2 Denver Health and Hospital Authority, Denver, CO, USA Background: Ledipasvir/sofosbuvir (LDV/SOF) is a recommended therapy for Hepatitis C virus (HCV). LDV/SOF increases tenofovir (TFV) exposures by 40-98% with TFV disoproxil fumarate (TDF) due in part to inhibition of TDF hydrolysis by SOF. This increase is greater with boosted HIV protease inhibitors (b/PI), resulting in renal toxicity concerns. There are no PK or renal safety data for TFV alafenamide (TAF) 25mg with b/PI and LDV/SOF. Our study objectives were to compare the plasma/intracellular PK and renal safety of b/PI with TDF, TAF, and TAF+LDV/SOF in persons living with HIV (PLWH). Methods: PLWH 18-70 yrs on TDF with ritonavir (RTV)- or cobicistat (COBI)-b/ PI were eligible. The study had 3 phases (Ph): (1) TDF 300mg + b/PI x 12 wks, (2) TAF 25mg + b/PI x 12 wks, and (3) TAF 25mg + b/PI + LDV/SOF x 4 wks. Adherence was electronically monitored using Wisepill®. Visits occurred at the end of each phase to collect PK (time 0 [pre-dose], 1 and 4 hrs post-dose) and renal biomarkers. PBMC were isolated pre-dose and plasma at every time point. TAF, TFV, and TFV-DP were quantified via LC-MS/MS. Plasma TFV exposure was calculated using non-compartmental methods. PK and renal biomarkers were log-transformed prior to analysis with mixed models. Results were back-transformed and Ph comparisons were reported as GMR (95% CI). P<0.05 was considered statistically significant with no adjustment for multiple comparisons. Results: Ten participants (1 black female; 9 males [5 Hispanic, 4 white]) were enrolled; 9 were on darunavir (5 RTV, 4 COBI) and 1 on atazanavir/RTV. Plasma TFV exposures were 76-79% lower for Ph 2 and 3 vs. 1 (Table 1). TFV-DP in PBMC were 11.1-fold higher for Ph 2 vs. 1 and 13.5-fold higher for Ph 3 vs. 1. Plasma TAF/TFV and TFV-DP in PBMC did not significantly differ for Ph 3 vs. 2, but TFV-DP trended towards a ~20% increase. PBMC findings were similar after controlling for adherence. eGFR did not differ between phases. Renal biomarkers either trended toward or showed improvements following TDF to TAF switch, and did not worsen with LDV/SOF. Conclusion: TFV-DP in PBMC increased 11-fold with TAF 25mg relative to TDF with b/PI. This increase is within the range of TFV-DP observed historically with higher TAF doses. Unlike prior findings with TDF, adding LDV/SOF with TAF did not significantly increase plasma TFV or TFV-DP in PBMC. This is likely due to

Poster Abstracts

448 AGING DOES NOT IMPACT DRUG-DRUG INTERACTION MAGNITUDES INVOLVING ANTIRETROVIRAL DRUGS Felix Stader 1 , Perrine Courlet 2 , Hannah Kinvig 3 , Manuel Battegay 1 , Laurent A. Decosterd 2 , Melissa A. Penny 4 , Marco Siccardi 3 , Catia Marzolini 1 1 University Hospital Basel, Basel, Switzerland, 2 Lausanne University Hospital, Lausanne, Switzerland, 3 University of Liverpool, Liverpool, UK, 4 Swiss Tropical and Public Health Institute, Basel, Switzerland Background: The risk of drug-drug interactions (DDIs) is elevated in aging people living with HIV (PLWH) because their increased prevalence of comorbidities leads to a higher use of comedications. Currently, the impact of aging on the magnitude and subsequently the management of DDIs in aging PLWH is unknown. As it is neither feasible nor ethically possible to study every drug combination in aging PLWH, we used physiologically based pharmacokinetic (PBPK) modelling in combination with limited clinical data to investigate the impact of aging on DDI magnitudes involving antiretrovirals (ARVs). Methods: A whole-body PBPK model was built in the mathematical programming language Matlab® including age-dependent physiological changes for the simulation of elderly subjects. The ability of the model to predict

CROI 2020 157