Definition and Independent Validation of a Proteomic-Classifier in Ovarian Cancer

Abstract

Mass-spectrometry-based analyses have identified a variety of candidate protein biomarkers that might be crucial for epithelial ovarian cancer (EOC) development and therapy response. Comprehensive validation studies of the biological and clinical implications of proteomics are needed to advance them toward clinical use. Using the Deep MALDI method of mass spectrometry, we developed and independently validated (development cohort: n = 199, validation cohort: n = 135) a blood-based proteomic classifier, stratifying EOC patients into good and poor survival groups. We also determined an age dependency of the prognostic performance of this classifier, and our protein set enrichment analysis showed that the good and poor proteomic phenotypes were associated with, respectively, lower and higher levels of complement activation, inflammatory response, and acute phase reactants. This work highlights that, just like molecular markers of the tumor itself, the systemic condition of a patient (partly reflected in proteomic patterns) also influences survival and therapy response in a subset of ovarian cancer patients and could therefore be integrated into future processes of therapy planning.

Semi-Quantitative MALDI Measurements of Blood-Based Samples for Molecular Diagnostics

“…As detailed below, we find better reproducibility when we include information from both the fine structure and the bumps when determining the feature values for each peak. To maintain a consistent naming convention used in previously published literature [6,7,[11][12][13][14][15], we will use the general term "feature" to refer to the peaks and "feature value" to be the semiquantitative numerical value we calculate to represent the relative abundance of that feature (protein or peptide) within the sample. High-resolution images of a representative unprocessed and processed MALDI-TOF spectrum across the entire acquisition range (m/z = 3 to 30 kDa) is shown in the Supplementary Materials Figure S1.…”

“…These highly multiplexed data can be combined into diagnostic tests using machine learning techniques designed to work well in the clinical setting where we generally have more attributes than samples, without overfitting [7,8]. Multiple tests in the area of oncology were developed using this approach [9][10][11][12][13][14][15][16][17][18][19][20][21].…”

Explaining multivariate molecular diagnostic tests via Shapley values

“…It has already been observed that certain ML architectures facilitate SV calculations, e.g., tree SHAP [ 14 ]. The additive axioms satisfied by SVs facilitate SV calculations for tests based on ensemble averages, and ML methods based on regularized combinations of small coalitions of features also present the possibilities of exact SV calculations for tests which include large numbers of features [ 36 – 39 ]. Systematic studies of the convergence of sampling-based approximations to exact SV calculations for models with these architectures are underway.…”