This systematic review evaluated whether prior exposure to immune checkpoint inhibitors (ICI) is associated with an increased risk of grade 3 or higher hepatotoxicity when patients with KRAS G12C–mutant non–small-cell lung cancer receive sotorasib. The authors identified a single matched analysis (reported as a meeting abstract) that compared patients with and without prior ICI exposure. In that matched comparison (386 patients per arm), the proportion of patients with grade 3+ hepatotoxicity was 13.2% in the ICI-exposed group versus 4.4% in the unexposed group, yielding a matched risk ratio (RR) of 3.00 (95% CI 1.76–5.10). The review reports that prior ICI exposure was therefore associated with a higher risk in this matched analysis, but the overall certainty of this finding was rated as very low.
The review separately examined the interval between last ICI exposure and initiation of sotorasib as a potential risk factor. Matched comparisons at predefined interval cutoffs did not demonstrate statistically significant differences: at 6 weeks the matched RR was 1.11 (95% CI 0.61–2.03), and at 12 weeks the matched RR was 1.08 (95% CI 0.52–2.27). Although these point estimates did not show clear increases in risk, the confidence intervals were wide and could not exclude meaningful clinical effects.
All five unadjusted cohort comparisons included in the review reported higher rates of severe hepatotoxicity with shorter ICI-to-sotorasib intervals. The CodeBreaK 200 trial data cited an example of 21/74 versus 8/75 patients experiencing grade 3+ hepatotoxicity in the earlier versus later interval quartiles. However, the effect sizes across unadjusted cohorts were heterogeneous (exact homogeneity P = 0.006), limiting pooled interpretation.
The authors conducted comprehensive searches of PubMed, Europe PMC, Embase, Web of Science, Scopus, the Cochrane Central Register of Controlled Trials (CENTRAL), and meeting proceedings for comparative studies reporting grade 3 or higher hepatotoxicity after KRAS G12C inhibitor therapy. From 530 unique records screened, seven sources met inclusion criteria; all included analyses involved sotorasib. Sources comprised published reports, a meeting abstract that provided the only matched analysis, and unadjusted cohort data including trial supplementary appendices.
The review synthesized results from matched analyses and unadjusted cohort data separately. Matched risk ratios (RRs) were pooled where applicable; unadjusted cohort data were synthesized using an exact conditional odds ratio (OR) approach. Certainty of evidence for outcomes was assessed using the GRADE framework. The authors made the extracted dataset and analysis code publicly available on OSF.
Key findings reported by the review are:
A single matched analysis (meeting abstract) showed prior ICI exposure was associated with a higher rate of grade 3+ hepatotoxicity (13.2% vs 4.4%; matched RR 3.00, 95% CI 1.76–5.10).
Matched comparisons of ICI-to-sotorasib intervals at 6 weeks and 12 weeks did not show statistically significant differences (RR 1.11, 0.61–2.03 at 6 weeks; RR 1.08, 0.52–2.27 at 12 weeks), but confidence intervals were compatible with potentially important effects.
All five unadjusted cohorts reported higher rates of severe hepatotoxicity with shorter intervals between ICI and sotorasib initiation. One trial example (CodeBreaK 200) observed 21 events among 74 patients in earlier interval quartiles versus 8 events among 75 patients in later quartiles.
There was statistically significant heterogeneity among the unadjusted cohort effect sizes (exact homogeneity P = 0.006).
The authors rated the certainty of evidence as very low across comparisons using GRADE. Reasons included reliance on a single matched abstract for adjusted comparisons, multiple unadjusted cohort analyses with heterogeneous effects, and wide confidence intervals for interval-based matched comparisons. Heterogeneity among unadjusted results further limits confidence in pooled estimates and inferences about a causative interval effect.
Based on available data, the review concludes that prior ICI exposure was associated with a higher risk of severe sotorasib-associated hepatotoxicity in the single available matched analysis, but this finding carries very low certainty. Shorter ICI-to-sotorasib intervals were associated with higher risk in unadjusted data only; current evidence does not establish a mandatory washout period before starting sotorasib.
In practical terms, the authors emphasize that liver-test monitoring recommended in the sotorasib prescribing information should apply to all patients. Additional surveillance after prior ICI exposure is noted as a possible precaution but remains untested by higher-certainty data.
Limitations explicitly reported in the source include the small number of comparative sources, reliance on a single matched analysis presented as a meeting abstract, and heterogeneity among unadjusted cohorts. The review could not definitively establish causation or a safe washout interval due to very low certainty evidence.
All extracted data and analysis code supporting the review are available in an OSF project cited by the authors. The systematic review was registered in OSF Registries. The authors declared no competing interests and reported funding sources from the National Research Foundation of Korea and the Korea Health Industry Development Institute. The article is a preprint posted on medRxiv and has not been peer reviewed; the authors note that it should not be used to guide clinical practice without further confirmation.