Activating mutations of KRAS and ALK occur in ~32% and ~8% of patients with non-small cell lung cancer (NSCLC). Breakthrough targeted therapies including the KRAS G12C selective inhibitor – sotorasib, and the 3rd generation ALK inhibitor – lorlatinib, have demonstrated effectiveness in clinical trials. Yet, relapse occurs in up to 50% of patients without new driver mutations. This highlights the critical importance of transcriptional (non-genetic) mechanisms of resistance, which remains poorly understood.
To address this, we utilise a lineage tracing technique called SPLINTR (Single Cell Profiling and LINeage TRacing). SPLINTR unites lentiviral cellular barcoding with single cell RNA-seq (scRNA-seq) to allow simultaneous lineage tracing and transcriptional read-out of malignant cells across a treatment trajectory at clonal resolution. Additionally, barcode sequencing from gDNA can identify the clonal structure of tumours pre- and post-therapy. We hypothesised that this approach would reveal transcriptional evolution as malignant cells respond to targeted therapy and establish minimal residual disease comprised of drug-tolerant persisters (DTPs).
By using an in vivo sub-cutaneous model of SPLINTR barcoded EML4::ALK (H3122) and KRASG12C/+ (H358) NSCLC cell lines, we examined treatment response to lorlatinib and sotorasib respectively. This approach revealed almost no clonal restriction in DTP tumours compared to our vehicle arm, suggesting that all clones adapt similarly to targeted therapies and enter a DTP state. Additionally, we reveal that the DTP state itself is not transcriptionally canalised, rather it is comprised of diverse transcriptional states, expressing basal-like, ciliate-like, and neuroendocrine-like cell states to evade therapeutic pressure. Together, these data demonstrate co-ordinated maintenance of clonal structure in response to treatment, but with multiple avenues for adaptive transcriptional evolution under targeted therapy. Finally, we conducted an in vivo sensitiser CRISPR screen targeting the ‘druggable proteome’ (~800 genes) to propose a new therapeutic strategy to overcome DTPs in ALK-driven NSCLC.
This study deepens our understanding of the biology of drug tolerant persisters (DTPs) in cancer and develops on concepts of non-genetic tumour evolution following targeted therapy.