PNPLA3 Status and Tirzepatide Liver Fat
In this exploratory post hoc analysis, PNPLA3 I148M allele presence did not appear to alter tirzepatide-associated improvements in MRI-assessed liver fat content and several cardiometabolic parameters in the studied type 2 diabetes population.
Evidence
This was a post hoc analysis of the phase 3, open-label SURPASS-3 MRI substudy. Insulin-naive adults with type 2 diabetes inadequately controlled on metformin, with or without a sodium-glucose co-transporter-2 inhibitor, had been randomized in the parent study to once-weekly tirzepatide or once-daily insulin degludec. The supplied packet does not locate a sample size for this genotype analysis. [pmid:42576670]
The analysis classified genotype CC as absence of the PNPLA3 I148M allele and CG or GG as its presence. Participant blood DNA was genotyped with a TaqMan allelic discrimination assay. Eligible analytic participants were enrolled in the MRI substudy, received at least one study-drug dose, and had a valid MRI scan at baseline or postbaseline. [pmid:42576670]
Investigators compared baseline-to-Week-52 changes in MRI-assessed liver fat content, body weight, waist circumference, HbA1c, fasting serum glucose, lipids, and liver enzymes across CC, CG, and GG genotype subgroups. Tirzepatide-treated participants had significant reductions in liver fat content, body weight, waist circumference, HbA1c, and fasting serum glucose at Week 52, with p < 0.001, regardless of genotype subgroup. [pmid:42576670]
No statistically significant differences were reported between genotype subgroups for the reported outcomes. The treatment-by-genotype interaction at Week 52 was not significant except for ALT and liver fat content; those interactions were significant at p < 0.05 because of statistical differences between GG and CG genotypes in the insulin degludec arm, also at p < 0.05. [pmid:42576670]
Analysis — Genotype consistency
The central finding is one of consistency rather than evidence that genotype changes treatment response. Across the evaluated PNPLA3 I148M subgroups, tirzepatide-associated reductions in liver fat content and the listed cardiometabolic measures were reported, and the study found no statistically significant outcome differences among genotype subgroups. This supports the authors’ limited conclusion that allele presence did not seem to affect tirzepatide-induced improvement in this study population. [pmid:42576670]
The comparator helps define the scope of that conclusion. The reported significant treatment-by-genotype interactions for ALT and liver fat content were attributed to differences between GG and CG participants receiving insulin degludec, not presented as evidence of differing tirzepatide benefit by PNPLA3 status. Thus, the results do not establish a genotype-directed treatment strategy; they describe an exploratory subgroup analysis within a randomized parent trial. [pmid:42576670]
The available findings report statistical testing but do not provide an exact effect size or confidence interval in the supplied packet for the genotype comparisons. Consequently, the evidence can characterize direction, consistency, and reported p-values, but it cannot quantify the magnitude or precision of any genotype-specific tirzepatide effect from the supplied material. Statistical significance in the reported analyses should not be interpreted as clinical significance. [pmid:42576670]
Limitations
The authors characterize the work as exploratory and state that interpretation is limited by small sample size. They also report that analyses were not adjusted for baseline body weight or ethnicity, which may confound observed outcomes. These features limit how confidently apparent consistency across genotype groups can be attributed specifically to PNPLA3 status. [pmid:42576670]
Generalizability is uncertain because the study had a small non-White population and an imbalance in Hispanic or Latino ethnicity across genotype subgroups. The authors note that genotype differences between ethnic groups have been described and call for further study to assess whether these findings generalize to different racial or ethnic groups. [pmid:42576670]
Some participants lacked postbaseline MRI measurements, and liver biopsy results were unavailable. The authors state that larger studies are needed to evaluate genotype-treatment interactions and the relative contribution of factors to liver fat-content changes by genotype. These limitations mean the analysis should be read as evidence about the studied population and measurements, not as a definitive account of PNPLA3-mediated treatment response. [pmid:42576670]