Concordance between results of inexpensive statistical models and multigene signatures in patients with ER+/HER2- early breast cancer

Mod Pathol. 2021 Jul;34(7):1297-1309. doi: 10.1038/s41379-021-00743-8. Epub 2021 Feb 8.

Abstract

Multigene signatures (MGS) are used to guide adjuvant chemotherapy (aCT) decisions in patients diagnosed with estrogen receptor (ER)-positive HER2-negative early breast cancer. We used results from three MGS (Oncotype DX® (ODX), MammaPrint® (MP) or Prosigna®) and assessed the concordance between high or low risk of recurrence and the predicted risk of recurrence based on statistical models. In addition, we looked at the impact of MGS results on final aCT administration during the multidisciplinary meeting (MDM). We retrospectively included 129 patients with ER-positive HER2-negative early breast cancer for which MGS testing was performed after MDM at University Hospitals Leuven between May 2013 and April 2019 in case there was doubt about aCT recommendation. Tumor tissue was analyzed either by ODX (N = 44), MP (N = 28), or Prosigna® (N = 57). Eight statistical models were computed: Magee equations (ME), Memorial Sloan Kettering simplified risk score (MSK-SRS), Breast Cancer Recurrence Score Estimator (BCRSE), OncotypeDXCalculator (ODXC), new Adjuvant! Online (nAOL), Mymammaprint.com (MyMP), PREDICT, and SiNK. Concordance, negative percent agreement, and positive percent agreement were calculated. Of 129 cases, 53% were MGS low and 47% MGS high risk. Concordances of 100.0% were observed between risk results obtained by ODX and ME. For MP, BCRSE demonstrated the best concordance, and for Prosigna® the average of ME. Concordances of <50.0% were observed between risk results obtained by ODX and nAOL, ODX and MyMP, ODX and SiNK, MP and MSK-SRS, MP and nAOL, MP and MyMP, MP and SiNK, and Prosigna® and ODXC. Integration of MGS results during MDM resulted in change of aCT recommendation in 47% of patients and a 15% relative and 9% absolute reduction. In conclusion, statistical models, especially ME and BCRSE, can be useful in selecting ER-positive HER2-negative early breast cancer patients who may need MGS testing resulting in enhanced cost-effectiveness and reduced delay in therapeutic decision-making.

Publication types

  • Research Support, Non-U.S. Gov't

MeSH terms

  • Adult
  • Aged
  • Biomarkers, Tumor / analysis*
  • Biomarkers, Tumor / genetics
  • Breast Neoplasms* / genetics
  • Breast Neoplasms* / pathology
  • Female
  • Humans
  • Middle Aged
  • Models, Statistical*
  • Receptor, ErbB-2 / genetics
  • Receptor, ErbB-2 / metabolism
  • Receptors, Estrogen / genetics
  • Receptors, Estrogen / metabolism
  • Retrospective Studies
  • Transcriptome*

Substances

  • Biomarkers, Tumor
  • Receptors, Estrogen
  • ERBB2 protein, human
  • Receptor, ErbB-2