An evaluation of the impact of technical bias on the concordance rate between primary and recurrent tumors in breast cancer

Abstract

Purpose: Whether or not to biopsy the metastasis in recurrent breast cancer has become mired in controversy. Several studies have shown an important discordance of the immunohistochemical (IHC) determinations for ER, PR and HER2 between primary (PT) and recurrent tumors (RT). Yet it remains unknown within this what impact technical issues have. The aim of our study was to assess whether technical variability might have an impact on the concordance between PT and RT.

Methods: IHC determinations in paired biopsies from PT and RT were compared under routine vs study conditions. In the former, pathological analysis reproduced the conditions used in the routine of a University Pathology Department. In the latter, in a technical bias-minimizing manner, samples were re-assessed at the same timing and by two independent observers.

Results: 128 paired biopsies from 64 patients were analyzed under both conditions. Concordance under routine vs study conditions for ER was 66% vs 93.4% (p = 0.001), for PR 58.7% vs 80.3% (p = 0.064) and for HER2 86.8% vs 96.8% (p = 0.25). Kappa index under routine versus study conditions for ER was 0.27 vs 0.79 (p = 0.002), for PR 0.26 vs 0.39 (p = 0.47) and for HER2 0.67 vs 0.9 (p = 0.14).

Conclusions: Although discordance rate between PT and RT decreased under conditions minimizing technical issues, some discordant cases appeared not to be subjected to this confounding factor. Either for clinical practice or for future studies reassessment of PT in recurrent breast cancer should be encouraged.

Keywords: Estrogen receptor; HER2; Metastatic breast cancer; Progesterone receptor; Technical bias.