164 CHAPTER 6 In chapter 4, we performed a baseline (phospho)proteomics analysis of 26 patients with RCC, treated with sunitinib. We retrospectively compared samples of patients who were primary resistant to the treatment (PFS < 12 weeks) to patients who had clinical benefit (PFS ≥ 12 weeks), aiming to describe differences in biology between the two groups. We found a discriminatory 78-phosphosite signature and kinase activity associated with sensitive and resistant tumors. p-Tyr phosphoproteomics in resistant tumors showed upregulation of phosphosites that are associated with resistance to treatment in other tumor types, and with inflammatory processes. A comprehensive pathway analysis pointed towards VEGF-independent tumor angiogenesis as a possible contributor to sunitinib resistance. We reproducibly identified three differentially upregulated proteins in resistant patients that showed overlap with differential transcripts from an independent cohort39, one of them (EIF4A1/EIF4A2) was also exclusively phosphorylated in resistant patients. This is the first comprehensive phosphoproteomics analysis on clinical RCC samples in relation to the response to sunitinib. Other phosphoproteomics studies use in vitro or in vivo models40,41, or use clinical samples to characterize the disease, without correlation to treatment response42. Sunitinib resistance in RCC remains a hot topic, and many post hoc efforts to find molecular biomarkers for treatment outcome have been reported, assessing other layers of biology in clinical samples. Beuselinck et al. performed a transcriptomics analysis on 53 clinical baseline RCC samples and report four distinct molecular subtypes of ccRCC, associated with different responses to sunitinib39. Motzer et al recently published their integrated multi-omics analysis of 823 baseline tumor samples and found seven molecular subsets of RCC, that correlate with response to angiogenesis blockade and immune checkpoint inhibitors43. To date, only one prospective biomarker-driven trial in metastatic RCC has been published. The randomized phase II BIONIKK trial demonstrates feasibility of treatment allocation based on prospective molecular classification and suggested an improved sunitinib efficacy in one of the four molecular subgroups44. Although a reliable and practical predictive biomarker for sunitinib efficacy in RCC is not yet available, many important steps have been taken to improve our understanding of its biology and molecular features. It seems only a matter of time (and effort) before patients can actually profit from the upfront prediction of tumor response to systemic therapy. Our analyses on the role of phosphoproteomics is promising, as it clearly separates primary resistant tumors from sensitive ones based on kinase activity and protein expression. Ultimately, a targeted assay could be developed based on this and future work, computing a simple and practical result that can be interpreted by clinicians in all hospitals. There are various examples of mass spectrometry-based phosphoproteomics analyses that result in better understanding of cancer biology45 and report potential targets for treatment46,47 and prognostic biomarkers48. These analyses are most often performed on cell lines and patient-derived xenografts (PDX). Phosphoproteomics analyses on tumor tissue samples are scarce and often include low numbers of patients.49-51. To our knowledge, no reports are pub-
RkJQdWJsaXNoZXIy MTk4NDMw