Treating HCV in dual diagnosis acute psychiatric inpatients with substance use disorder

selected by DAA therapy in regions targeted or not targeted by DAAs in GT-1l infected patients who failed to achieve SVR.Method: Shotgun metagenomics were used to generate full-length HCV GT-1l genome sequences.Briefly, library preparation was performed using Total RNA and Nextera XT kit (Illumina).Deep sequencing was performed by means of NextSeq500 (Illumina).Fulllength HCV sequences were analyzed using our original in-house MetaMIC ® software (1% cutoff ).Results: Among 771 patients with HCV infection treated with DAAs who experienced a virological failure and were referred to our center between 2015 and 2019, 316 (41.0%) were infected with genotype 1, and 20 of them (6.3%) with GT-1l.All GT-1l-infected patients were of African origin.Deep sequencing of full-length HCV GT-1l genomes was performed in 10 patients at baseline and at time of relapse.Treatment regimens were SOF/LDV ± RBV (n = 8), SOF/VEL (n = 1) and SOF/SIM (n = 1).At baseline, all of the 10 GT-1l patients had ≥3 dominant NS5A resistance-associated substitutions (RASs), including K24G/S (>99%), L31M (>99%) and H58P (>99%).Additionally, baseline NS5A-Q30R was present in 3 patients (>99% in 2 patients; 25% in 1 patient).Sequential analysis of full-length HCV genome sequences showed enrichment of NS5A-Q30R in one patient at failure (>99%) and the selection of NS5A RASs in other patients: baseline K24S (8% and 40%, respectively) replaced by K24G (>99%) in 2 patients; selection of Y93F and Y93H RASs at failure in 2 other patients (20% and 40% respectively).No NS5B RASs were detected.NS3 R155 K was selected in one patient failing SOF/SIM (30%).No amino acid changes of interest were observed at treatment failure in genome regions other than those targeted by the HCV DAAs administered.Data on retreatment by triple-combination regimens are ongoing and will be presented. Conclusion:We report the largest cohort of GT-1l-infected African patients failing DAAs thus far.The large number of NS5A RASs present at baseline explains lower SVR rates with NS5A inhibitor-based therapies in these patients.Our results emphasize the need for identifying this subtype and other "unusual" subtypes (e.g. in Africa where they are common, and the urgent need to guarantee equal access to last-generation DAA therapies in Africa.