The EACR’s Top 10 Cancer Research Publications is a regular summary of the most interesting and impactful recent papers in cancer research. It is curated by the Board of the European Association for Cancer Research (EACR).
The list below appears in no particular order, and the summary information has been provided by the authors.
Use the dropdown menu or ‘Previous’ and ‘Next’ buttons to navigate the list.
Sammut, SJ., Crispin-Ortuzar, M., Chin, SF. et al. Nature 601, 623–629 (2022).
Summary of the findings
Despite great strides in understanding the biology of cancer, we still cannot reliably predict response to chemotherapy (plus HER2-targeted therapy for ERBB2 amplified tumours). In this study, we aimed to: (a) understand the biological processes associated with response to chemotherapy in breast cancer and (b) build a machine learning framework that combines pre-therapy molecular and digital pathology data to predict response.
We molecularly profiled breast tumour biopsies obtained at diagnosis from 168 women using exome and RNA sequencing and analysed their histological architecture using digital pathology. Women received 18 weeks of pre-operative chemotherapy and response was assessed at surgery using the Residual Cancer Burden score.
The baseline pre-therapy features were monotonically associated with response. DNA features associated with response related to genomic instability, including tumour mutation and neoantigen burden, homologous recombination deficiency and TP53 mutations. Response to therapy was also modulated by the tumour immune microenvironment. Specifically, high lymphocyte density, lack of T-cell dysfunction and exclusion, and absence of HLA class I loss of heterozygosity were associated with a better outcome.
Combining these molecular and digital pathology features into a multi-omic machine learning model predicted a pathological complete response in an external validation cohort with an area under curve of 0.87.
A tweetorial describing this study can be found at: https://twitter.com/stephensammut/status/1468251380586590216
Future impact of the findings
By integrating highly dimensional multi-omic data generated by profiling the tumour ecosystem, we have increased our understanding of the biology of breast cancer and developed a framework that will transform the way we treat it. In the future, this framework could be used to determine which patients can be treated with standard of care therapies (if they are predicted to respond) or treated using novel therapies in clinical trials (if predicted to be resistant to standard therapies). Perhaps more importantly, this approach has demonstrated a new way of thinking about forecasting therapy response which could be adapted for use in other cancer types.
