Assessing the potential role of Rictor expression as predictive factor of response to PI3K/mTOR pathway inhibitors in preclinical models of squamous cell lung cancer

Squamous cell lung cancer (SQLC) is the second most prevalent histologic type of lung cancer and accounting for approximately 30% of newly diagnosed non-small cell lung cancer (NSCLC) cases. Systemic treatments for SQLC patients include cytotoxic chemotherapy and immune-oncology approaches. In contrast with lung adenocarcinoma, which is the other main subtype of NSCLC, no patient-tailored treatments are available so far for SQLC. Accumulating evidence suggests that the PI3K/mTOR axis is one of the most frequently altered pathways in SQLC. However, despite a plethora of clinical trials with numerous PI3K/mTOR targeted inhibitors, no significant increase in patients’ survival has been observed as compared to standard treatment options. A possible explanation for the outcome of those clinical trials might be the lack of reliable predictive biomarkers for better patients’ stratification. We and others have reported Rictor copy number gain (CNG) in a set of SQLC patients by performing targeted DNA sequencing on archival tissues. Another group has suggested the existence of Rictor focal amplification in subsets of lung cancers, including SQLC, and further suggested Rictor as a potential predictive biomarker of response to targeted therapy. However, no conclusive data were presented to show that Rictor amplification is driving activation of the PI3K/mTOR pathway in SQLC cells or representing a valid biomarker predictive of response to targeted inhibition of the pathway. Here, we used three different SQLC cell lines and 60 tissue specimens to show that CNG of Rictor is a recurrent event in SQLC, yet this is due to the polysomy of the short arm of chromosome 5 rather than to focal amplification. All three cell lines tested showed different Rictor CNG and different levels of its transcript and protein. In particular, the SQLC cell line harboring the higher CNG (H-1869) accordingly displayed higher level of Rictor protein. Therefore, we sought to test the possibility that the dosage of Rictor might affect the activation of PI3K/mTOR pathway and sensitivity towards its targeting agents. Unexpectedly, we found that Rictor levels did not parallel the biochemical activation of the pathway nor the sensitivity to dual mTORC1/C2 or PI3K/mTOR inhibitions. These observations were confirmed by genetic perturbation analysis, as reduction of Rictor levels through RNA interference did not lead neither to reduced cell viability nor to significant changes in drug sensitivity in the two cell lines tested. Overall, our findings suggest that Rictor does not represent a predictive biomarker of response towards PI3K/mTOR directed therapy.