Proteomic Analysis of Clinical Trial Samples Uncovers Mechanism of Drug Resistance

A proteomic analysis of tumor samples from a Phase I clinical trial (NCT01226316) has identified increased mTORC1-driven translation as a mechanism of resistance to the AKT inhibitor capivasertib. By combining discovery proteomics with targeted assays developed under National Cancer Institute’s Clinical Proteomic Tumor Analysis Consortium (CPTAC) guidelines, researchers have uncovered new insights into the mechanics of drug response in breast cancer. Their findings are published in Cancer Research Communications and the Journal of Proteome Research.

Proteomics Predicts Treatment Response

Capivasertib is a selective AKT inhibitor approved (in combination with fulvestrant) to treat a subset of breast cancer patients with specific genetic alterations (PIK3CA, AKT1, or PTEN), but research (NCT01226316, NCT01683694, NCT02047702) has shown objective response rates of less than 30%. To understand why, researchers analyzed tumor samples from patients with PIK3CA-altered tumors collected prior to capivasertib treatment.

The study found that total concentrations of capivasertib’s direct targets—AKT1 and AKT2—were not different between patients who demonstrated clinical benefit and those who did not. However, proteomic analysis identified that patients in the unresponsive group showed higher levels of proteins associated with translational activity, including structural subunits of ribosomes and mitochondrial proteins. Subsequent pathway analysis revealed that this resistance profile was driven by upregulation of mTORC1 and EIF2 signaling.

CPTAC Guidelines Enable "Fit-for-Purpose" Verification

To verify their findings and create a reliable tool for patient stratification, the team developed a multiplexed targeted proteomics panel. The development of these assays followed CPTAC guidelines for "fit-for-purpose" assay characterization as well as biomarker verification standards. The team created 53 multiple reaction monitoring assays targeting 50 proteins involved in the mTORC1 and EIF2 signaling pathways. Notably, 45 of the 53 peptide assays demonstrated high precision with low variability. All assay characterization data are available to the public via the CPTAC Assay Portal.

The Power of Translational Proteomics

Researchers leveraged their new assays to reproduce the capivasertib-resistant profile in breast cancer cell lines with a protein profile closely matching that of the resistant patient tumors. In the discussion, the authors emphasize that "the reproducibility of the profile associated with capivasertib resistance across two orthogonal proteomics approaches and two fully independent models lends credibility to the proteins’ ability to detect capivasertib sensitivity".

By moving beyond genomic mutations to measure proteins, researchers were able to identify a resistance mechanism that had previously gone undetected. Their analysis found that, while many cancer cells exploit the PI3K/AKT/mTOR pathway, proteomic profiles can help identify those primed to resist AKT inhibition.