AZD2693, a PNPLA3 antisense oligonucleotide, for the treatment of MASH in 148M homozygous participants: Two randomized phase I trials.

Objective: A common genetic variant (rs738409) encoding an isoleucine to methionine substitution at position 148 in the PNPLA3 protein is a determinant of hepatic steatosis, inflammation, fibrosis, cirrhosis, and liver-related mortality. AZD2693 is a liver-targeted antisense oligonucleotide against PNPLA3 mRNA. We evaluated the safety, tolerability, pharmacokinetics, and pharmacodynamics in single and multiple ascending dose studies. Methods: AZD2693 was assessed in 3D cultures of homozygous PNPLA3 148M primary human hepatocytes and mice expressing human PNPLA3. The single ascending dose study investigated 2-110 mg doses in overweight/mildly obese but otherwise healthy volunteers. The multiple ascending dose study investigated three monthly doses (25 mg, 50 mg and 80 mg) in participants with MRI-proton density fat fraction (MRI-PDFF) ≥7%. Changes in liver fat content were assessed at baseline, weeks 8 and 12 by MRI-PDFF. PNPLA3 mRNA and protein knockdown levels were evaluated for the 80 mg dose. Results: AZD2693 potently reduced PNPLA3 expression in human hepatocytes and livers of mice. Clinically, AZD2693 was generally well tolerated (no adverse events leading to discontinuation or treatment-related serious adverse events). Half-life was 14-33 days across investigated doses. A least-square mean liver PNPLA3 mRNA knockdown of 89% and reduction of protein levels demonstrated target engagement. Changes in hepatic steatosis at week 12 were -7.6% and -12.2% (placebo-corrected least-square means) for the 25 mg and 50 mg doses, respectively. There was a dose-dependent increase of polyunsaturated fatty acids in serum triglycerides and decreases vs. placebo in high-sensitivity C-reactive protein and interleukin 6. Conclusions: AZD2693 reduced liver PNPLA3 with an acceptable safety and tolerability profile. These findings support the continued development of AZD2693. Clinical treatment options for metabolic dysfunction-associated steatohepatitis (MASH) are limited. The genetic risk factor with the largest effect size for progressing to poor liver-related outcomes in MASH is a single-nucleotide polymorphism in the gene PNPLA3 (p.I148M). In phase I single and multiple ascending dose studies, AZD2693, a liver-targeted antisense oligonucleotide, was well tolerated, reduced liver PNPLA3 mRNA and protein levels, and dose-dependently reduced liver fat content in homozygous PNPLA3 148M risk allele carriers. These data support continued development of AZD2693 as a potential precision medicine treatment for MASH. The phase IIb FORTUNA study is now ongoing.

Background: NCT04142424, NCT04483947.