Background

Cancer cells evolve and can acquire the ability to evade death by therapies, resulting in patient relapse. Despite an initial response to therapy nearly all patients with small cell lung cancer (SCLC) relapse. Through advanced computational analyses we can decode patterns left behind on cancer cell DNA by therapeutic agents. We sought to study the genetic alterations induced by SCLC therapies to determine their contributions to resistance, while also targeting potential genomic vulnerabilities.

Methods

We performed whole genome sequencing on PDX-derived tumors and matched peripheral blood representing XX samples of SCLC.

Results

Analysis of genomic alterations detected therapeutic exposures including those of platinum therapies and temozolomide (TMZ). In TMZ treated patients, we identified a distinctive pattern of DNA damage with a striking increase in the single nucleotide variant burden of highly expressed genes, resembling transcription-coupled damage. We discovered a striking indifference to the standard replication timing mutational bias. We treated TMZ naïve PDX models with TMZ and established that TMZ-induced mutational biases can be replicated outside of patient biopsies.

Conclusions

We revealed a novel pattern of DNA damage in patients treated with TMZ which may impart a potential increase in neoantigens, suggesting that checkpoint blockade may be clinically relevant.