STK11 alterations as predictive biomarkers of resistance to immune checkpoint inhibitors in multiple cancers with mismatch repair gene alterations

2619 Background: STK11 is an important tumor suppressor gene that encodes for a serine-threonine kinase involved in a variety of metabolic functions. STK11 alterations are associated with an inert or cold tumor microenvironment in KRAS-mutant lung adenocarcinoma, and are a major driver of primary resistance to immune checkpoint inhibitors (ICIs). Here, we explored STK11 alterations as predictive biomarkers of resistance to ICIs in multiple solid tumors with mismatch repair (MMR) gene alterations. Methods: The cBioCancer Genomics Portal (cBioportal, http://www.cbioportal.org ) was used to obtain genomic data. The association between STK11 alterations and median overall survival (mOS) was explored in an ICI-treatment cohort comprising a total of 1572 evaluable patients. This cohort includes patients with cancers of the lung, colon, unknown primary, head and neck, gastroesophageal, bladder, skin (melanoma), brain, breast and kidney that had received at least one dose of ICI. These were then cross-referenced with patients who had MSK-IMPACT testing done in the context of routine clinical care. Kaplan-Meier survival curves were generated and compared using log-rank testing. Results: In the ICI-treated patients cohort (N = 1572), 96 patients (6.1%) had STK11 alterations. Patients with STK11 alterations had higher median tumor mutation burden (TMB) than patients without STK11 alterations (8.81 mut/Mb vs. 5.87 mut/Mb, p < 0.0001). Yet, they had worse mOS (7 m, 95% CI: 6-13 m vs. 20 m, 95% CI: 18-22 m, p < 0.001). MMR alterations (not including gene promoter methylation) were identified in 103 (6.6%) and co-occurrence of STK11 with MMR alterations was identified in 9 (0.6%) patients. Compared to patients without MMR or STK11 alterations (N = 1382), the mOS was better in patients with MMR alterations without STK11 alterations (N = 94) (59 m vs. 19 m, p = 0.0004) and trending for worse in patients with MMR and STK11 alterations (N = 9) (5 m vs. 19 m, p = 0.117). The mOS was better in patients with MMR alterations without STK11 alterations compared to patients with MMR and STK11 alterations (59 m vs 5 m, p = 0.016). Controlling for KRAS mutation status, a multivariable Cox regression model shows that besides MMR and STK11 alteration status (p = 0.0008), BRAF mutant status (p = 0.0001, HR 0.598; 95% CI, 0.459-0.779) was another biomarker associated with significantly better OS. Lung cancer diagnosis was associated with worse OS (p = 0.0016, HR 1.340; 95% CI, 1.117-1.607). Conclusions: In a large cohort of ICI-treated patients with multiple cancers, STK11 alterations were associated with worse mOS despite having higher TMB. MMR alterations were only predictive of response to ICIs in patients without STK11 alterations. Our findings suggest that STK11 alterations in patients with MMR alterations could potentially identify patients who are less likely to respond to ICIs but more data are needed.