Chemotherapy is the first line therapy for numerous cancers, however it is often not curative. The mechanisms that drive an effective chemotherapeutic response are incompletely understood and there no robust predictive biomarkers for response. Along with the direct cytotoxic effects of chemotherapy, the composition of the tumour microenvironment (TME), in particular the immune system, have been shown to influence treatment.
Here, we developed a unique approach to determine the molecular characteristics of a chemotherapy responsive TME, by exploiting the fact that inbred mice bearing tumours derived from clonal cancer cell lines display dichotomous responses to treatment with cyclophosphamide (CY). We established that two bilateral AB1-HA murine mesothelioma tumours respond symmetrically, yet dichotomously to CY; both tumours either respond or they both do not. This allows for one tumour to be removed and analysed while its therapeutic ‘fate’ is inferred from the response of the remaining tumour. Using this model, we performed RNA sequencing on tumours prior to CY treatment to characterize the pre-treatment microenvironment of responding and non-responding tumours.
We found that before treatment, responding tumours were enriched for immune-related genes and had significantly upregulated CD4+ T helper associated pathways, compared to non-responding tumours. Similar pathways were also indicative of a response to cyclophosphamide containing chemotherapy regimes in breast cancer patient cohorts. When CD4+ T cells were depleted or inflammation inhibited using dexamethasone, the response to CY was completely abrogated. Using upstream regulator analysis, we identified key regulators that were predicted to induce a chemotherapy sensitive TME. Pre-treatment of tumours by specifically targeting the identified key regulators sensitised tumours to CY, increasing response rates and significantly improving survival in a CD4+ T cell-dependent manner.
These findings indicate that a CD4+ T cell driven pre-treatment gene expression signature can serve as a predictive biomarker for response to CY. pre-treating tumours with drugs that target the identified key regulators of response can improve treatment efficacy by inducing a chemo-sensitive TME.