Abstract Background/Aim: Correct assessments of prognostic and predictive markers (ER, PgR, HER2, histological grade) are essential in the choice of adjuvant therapy in primary breast cancer. Comparisons of department performances are necessary to ensure that daily diagnostic practice meets clinical requirements. Here we compared the distribution of biomarker evaluation scores among the pathology departments in Sweden. Methods: All breast cancer cases diagnosed 2013 to 2017 in Sweden were identified in the National Quality Register for Breast Cancer. All cases with results on ER, PgR, HER2, Ki67 and histological grade tested on surgical resection specimens were selected. Only pathology departments with more than 450 diagnosed breast cancer cases in the given period of time were included in the study, resulting in 35,066 cases from 29 investigated departments across Sweden. All pathology departments follow the national guidelines of biomarker evaluation. The distribution of breast cancer phenotypes across the regions of Sweden is considered homogenous. We used Kruskal-Wallis test with Dunn’s post-hoc analysis, and χ2 tests to compare the departments resulting in totally 406 comparisons per marker. We applied Bonferroni correction in multiple comparisons. The trend of biomarkers score distribution among departments was investigated using joinpoint analysis. Results: We found significant difference in biomarker evaluation scores among the pathology departments for all the investigated markers (p<0.001 for all markers). Regarding hormone receptors, only 3 comparisons were significantly different for ER status, while 56 out of 406 comparisons were significantly different for PgR status (p<0.05). The distribution of positive cases among the departments ranged from 85.2% to 91.6% for ER status and from 67.2% to 83.9% for PgR. Trend analysis showed that 21 departments were in the same trend for the distribution of hormone receptor positive cases (ER: 86.3%-89% PgR: 70%-77.1%, p<0.001 for both markers. Regarding HER2 status, 23 comparisons were significantly different. The median percentage of HER2 positive cases were 11.7%. The distribution varied between 7.2% and 16.2% and the joinpoint analysis showed 4 trends among the departments as follows: 2 departments with 7.1% and 8.2%; 10 departments between 9.2%-10.9%; 12 departments between 11.5%-13.3% and 5 departments between 14%-16.2% (p=0.01). Considering Ki67, 175 comparisons were significantly different (p<0.05). The distribution of Ki67 low cases ranged between 23.5%-67.4% showing 4 trends among the departments: 2 departments with 23.5% and 33.7%, 11 departments between 37.8%-48.2%, 14 departments between 49.1%-58.5 and 2 departments with 60.2% and 67.4% (p=0.008). We found 92 comparisons significantly different during the investigation of histological grade scores among the departments (p<0.05). The distribution of Grade 3 cases varied between 17.4%-37.9% and the joinpoint analysis showed 4 trends among the departments as follows: 3 departments between 17.4% and 22.5%; 11 departments between 24.8%-30.5%; 11 departments between 30.9%-32.5% and 4 departments between 33.7%-37.9% (p<0.001). Conclusions: Variability in biomarker evaluation exists among Swedish pathology departments despite using the same guidelines, internal and external quality assurance for several years. The lowest variability was found for ER and HER2 testing, while histological grade and Ki67 assessment showed the highest variability among departments. The discrepancy was likely due to pre-analytical, analytical as well as scoring factors. Citation Format: Balazs Acs, Irma Fredriksson, Ulla Wilking, Stephanie Robertson, Johan Hartman. Distribution of biomarker evaluation scores among the pathology departments of Sweden: Comparison study of 35,066 breast cancer cases [abstract]. In: Proceedings of the 2019 San Antonio Breast Cancer Symposium; 2019 Dec 10-14; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2020;80(4 Suppl):Abstract nr P5-02-07.
Read full abstract