Co-expressed genes enhance precision of receptor status identification in breast cancer patients

Abstract

PurposeTherapeutic decisions in breast cancer patients crucially depend on the status of estrogen receptor, progesterone receptor and HER2, obtained by immunohistochemistry (IHC). These are known to be inaccurate sometimes, and we demonstrate how to use gene-expression to increase precision of receptor status.MethodsWe downloaded data from 3241 breast cancer patients out of 36 clinical studies. For each receptor, we modelled the mRNA expression of the receptor gene and a co-gene by logistic regression. For each patient, predictions from logistic regression were merged with information from IHC on a probabilistic basis to arrive at a fused prediction result.ResultsWe introduce Sankey diagrams to visualize the step by step increase of precision as information is added from gene expression: IHC-estimates are qualified as ‘confirmed’, ‘rejected’ or ‘corrected’. Additionally, we introduce the category ‘inconclusive’ to spot those patients in need for additional assessments so as to increase diagnostic precision and safety.ConclusionsWe demonstrate a sound mathematical basis for the fusion of information, even if partly contradictive. The concept is extendable to more than three sources of information, as particularly important for OMICS data. The overall number of undecidable cases is reduced as well as those assessed falsely. We outline how decision rules may be extended to also weigh consequences, being different in severity for false-positive and false-negative assessments, respectively. The possible benefit is demonstrated by comparing the disease free survival between patients whose IHC could be confirmed versus those for which it was corrected.Electronic supplementary materialThe online version of this article (10.1007/s10549-018-4920-x) contains supplementary material, which is available to authorized users.

Microarray Normalization Revisited for Reproducible Breast Cancer Biomarkers

“…Up to now, we have compared pairs of pipelines, now we evaluate single pipelines: Following normalization via each pipeline, logistic regression is trained as described in our previous work [ 50 ], using a single cut-point, rendering no cases undecided. Predictions from gene expression for receptor status are then checked against the golden standard, IHC.…”

“…Based on that, we applied fSVA and obtained batch-corrected corrected expression values for all 3753 samples. Our odds-prediction algorithm [ 17 , 50 ] was applied, based on gene and cogene of each receptor (ER, PGR, HER2). Unexpectedly, prediction quality worsened: percent misclassified receptors increased from 10.7% (average for RmaMSingle) to 21.3%, see Table 6 .…”

“…However, for precision medicine, it is even more relevant if biomarkers result stable and reproducible: Do biomarkers, as determined from a given array (i.e., sample), lead to unique consequences for a given patient, irrespective of the particular pipeline being used? To clarify, we compute two test sets of predictive biomarkers: Prediction models for 3 hormone receptor status (ER, PGR, HER2) in breast cancer patients, as published earlier by our group [ 17 , 50 ]. In the present work, we apply the very same prediction algorithm (“odds”-method) on expression data normalized in different ways 6 different breast cancer subtyping algorithms (“biomarkers”: smgene, scmod1, scmod2 pam50, ssp2006 and ssp2003)), rendered by the well-known, publicly available “genefu” software package [50, 51].…”

“…Prediction models for 3 hormone receptor status (ER, PGR, HER2) in breast cancer patients, as published earlier by our group [ 17 , 50 ]. In the present work, we apply the very same prediction algorithm (“odds”-method) on expression data normalized in different ways…”

Decision theory for precision therapy of breast cancer

“…In previous papers 57 , 58 we have worked on such approaches, applying standard statistical means (odds-products). In this work we expand the approach by drawing on Dempster Shafer decision Theory (DST) 59 .…”