Fig. 2

Staining reactions submitted to NordiQC based on optimal and insufficient protocols. HER2 stains of four breast carcinomas (a–d) in three laboratories (1–3). The stains from laboratory 1 (upper row) were assessed as optimal, while stains from laboratory 2 (middle row) were too weak, and from laboratory 3 (lower row) were over-stained; both sets were assessed as poor. 1a) In tumour a, >30 % of the neoplastic cells show an expected strong and complete membranous staining, corresponding to score 3+; FISH test showed amplification, HER2/CEN17 ratio >6. 1b) In tumour b, >10 % of neoplastic cells show an expected moderate complete membranous staining corresponding to score 2+; tumour b is amplified, HER2/CEN17 ratio 2.4–2.9. 1c) Tumour c stains as tumour b, but is unamplified, HER2/CEN17 ratio 1.3–1.6; 1d) In tumour d, no staining is identified; the tumour is unamplified, HER2/CEN17 ratio 1.1–1.4. 2a) Tumour a shows a 2+ reaction that would be reflexed to FISH test; 2b-d) Tumours b, c and d show a 0/1+ reaction. As the amplified tumour b would not be reflexed to FISH test, the patient would erroneously not be offered trastuzumab treatment. 3a-d) Widespread cytoplasmic reaction is seen in all four tumours. The stain of unamplified tumour c might be interpreted as a 3+; the tumour therefore would not be not reflexed to FISH test and the patient would erroneously be offered trastuzumab treatment. FISH, fluorescence in situ hybridisation