CROI 2019 Abstract eBook

Abstract eBook

Poster Abstracts

DDIs. CAB PBPK model accurately predicted CAB PK parameters (all within acceptable bioequivalence criteria (0.80-1.25) for single as well as repeat dose studies). DDI simulations predicted a mean change in systemic exposure for tested OAT1/OAT3 substrates of <25% after co-administration with CAB at steady state. Conclusion: A PBPK model of CAB was developed and validated that accurately predicted human pharmacokinetics observed in healthy volunteers. OAT1/OAT3 substrate drugs such as tenofovir, cidofovir, methotrexate were predicted to have a minimal risk of DDIs when administered with CAB. Similar CAB concentrations following oral and LA administration suggest that these results would apply to CAB LA. The predicted lack of interactions supports co- administration with OAT1/OAT3 substrates without dose adjustments.

failure was seen in 16%, 22% and 2.5% of patients with normal, reduced and low UGT1A1 activity, respectively. Failure rates were lower in patients with low activity UGT1A1 (p=0.012; Figure). The relationship remained significant when adjusted for baseline CD4 count (p=0.048) but not when adjusted for baseline VL (p=0.082) or both CD4 and VL [HR (95% CI): 0.18 (0.02-1.30); p=0.08]. Conclusion: The NEAT001/ANRS143 study analysed UGT1A1 genotypes with the largest sample size to date and suggested little impact on RAL PK. However, UGT1A1 genotype may be a better correlate of RAL pharmacodynamics because of the high intra-subject variability in RAL PK.

Poster Abstracts

472 PHARMACOGENETICS OF WEIGHT GAIN AFTER SWITCH FROM EFAVIRENZ TO INTEGRASE INHIBITORS Michael Leonard 1 , Kassem Bourgi 2 , John R. Koethe 1 , Megan M. Turner 1 , Jamison Norwood 1 , Beverly O. Woodward 1 , Husamettin Erdem 1 , Rebecca Basham 1 , Paxton Baker 1 , Peter F. Rebeiro 1 , Timothy R. Sterling 1 , Todd Hulgan 1 , David Haas 1 1 Vanderbilt University, Nashville, TN, USA, 2 Indiana University, Indianapolis, IN, USA Background: Weight gain has been reported in virologically suppressed HIV- positive patients who switch to integrase inhibitor (INI)-based antiretroviral therapy (ART). We studied pharmacogenetics of weight gain following switch from efavirenz (EFV)- to INI-based ART. Methods: Patients at an HIV clinic in the southeastern USA were on EFV-based ART for at least 2 years and with no viral load >1000 copies/mL within 6 months prior to switch. Weight gain from date of switch to weeks 24 and/or 48 (± 4 weeks) was assessed. We genotyped CYP2B6 and UGT1A1 polymorphisms that predict increased plasma EFV and INI exposure, respectively. Associations were tested with linear regression models. Results: The 101 evaluable participants (n=83 for week 24, n=66 for week 48) included 65 (64%) white, 27 (27%) black, 84 (83%) male, and 17 (17%) female participants. INIs were 58 (57%) dolutegravir, 34 (34%) elvitegravir, and 9 (9%) raltegravir. Median baseline weight was 81.7 kg (interquartile range: 69.7, 94.7). There were 30 (30%), 54 (55%), and 15 (15%) CYP2B6 normal, intermediate, and slowmetabolizers, respectively, and 38 (40%), 41 (43%), and 16 (17%) UGT1A1 normal, intermediate, and slowmetabolizers, respectively. CYP2B6 slowmetabolizer genotype was associated with weight gain at week 48 (β= 7.2, p=0.009). In CYP2B6 normal, intermediate, and slow metabolizers, at week 48, average weight gain was 0.2 kg, 2.8 kg and 2.0 kg, respectively. After controlling for sex, age, and weight at switch, associations persisted at week 48 (β=6.97, p=0.012). CYP2B6 genotype was associated with weight gain in whites at week 48 (β=11.25, p=0.003), but not in blacks (β= -0.58, p=0.090) (Figure). The above significant associations also tended to be present at week 24 (p=0.05 to p=0.09). UGT1A1 genotype was not associated with weight change at week 24 (β= -0.33, p=0.83) or week 48 (β=0.70, p=0.77). Conclusion: Among virologically suppressed patients who switch from EFV-based ART to INI-based ART, CYP2B6 genotype that is known to predict higher EFV plasma exposure pre-switch may be associated with greater weight gain after switch. These findings warrant replication in other cohorts. We hypothesize that patients with greater plasma EFV concentrations before switch

471 INFLUENCE OF UGT1A1*28 ON RALTEGRAVIR PK/PD IN THE NEAT001/ ANRS143 STUDY Rohan M. Gurjar 1 , Laura Dickinson 1 , Daniel F. Carr 1 , Wolfgang Stohr 2 , Stefano Bonora 3 , Andrew Owen 1 , Antonio D’Avolio 3 , Adam Cursley 2 , Jean-Michel Molina 4 , Gerd Fätkenheuer 5 , Giovanni Di Perri 3 , Anton Pozniak 6 , Laura Richert 7 , François Raffi 8 , Marta Boffito 9 1 University of Liverpool, Liverpool, UK, 2 MRC Clinical Trials Unit at UCL, London, UK, 3 University of Turin, Turin, Italy, 4 University Paris Diderot, Paris, France, 5 Cologne University Hospital, Cologne, Germany, 6 St. Stephen’s Center at Chelsea and Westminster Hospital, London, UK, 7 University of Bordeaux, Bordeaux, France, 8 CHU de Nantes, Nantes, France, 9 St. Stephen’s Center at Chelsea and Westminster Hospital, London, UK Background: Raltegravir (RAL) is metabolised by UGT1A1 and polymorphisms in the UGT1A1 gene have been associated with plasma concentrations in some but not all previous studies. This analysis represents the largest study to date for the effect of UGT1A1 polymorphisms on RAL PK/PD. Methods: NEAT001/ANRS143, a randomised study (n=805 participants), demonstrated non-inferiority of first-line darunavir/ritonavir (DRV/r; 800/100 mg o.d) plus RAL (400 mg b.d) compared with DRV/r plus tenofovir/ emtricitabine (245/200 mg o.d). Random, single samples were collected at weeks 4 and 24 post-therapy initiation for drug measurement. DNA was extracted and UGT1A1 polymorphisms genotyped using the Sequenom MassARRAY iPLEX. Nonlinear mixed effects modelling (NONMEM v. 7.3) was used to estimate PK parameters. Weight, age, sex, ethnicity and genotypes were investigated in the model for association with RAL apparent oral clearance (CL/F). Kaplan-Meier estimates and Cox regression were used to assess the relationship between virological failure by week 96 and UGT1A1 genotypes. Results: A total of 602 samples (n=313 week 4, n=289 week 24) from 349 patients were used in the model (n=264 with genotypes). UGT1A1 activity was defined as normal (*1/*1, *1/*36), reduced (*1/*6, *1/*28, *1/*37, *28/*36, *36/*37) or low (*28/*28, *28/*37, *37/*37). Although none of the covariates were statistically significant, RAL CL/F was reduced by 21% in patients with *28/*28 UGT1A1 . A post-hoc analysis assessed the impact of UGT1A1 *28 on predicted RAL AUC 0-12 and C 12 [low activity (n=40) vs normal/reduced activity (n=224)]. Geometric mean ratios (95% CI) were 1.35 (0.99-1.84; p=0.062) and 1.32 (0.99-1.77; p=0.062), respectively, suggesting minimal impact of UGT1A1 *28 on RAL PK. By week 96, virological

CROI 2019 174