Aims Altering the length of time specimens are placed in fixative without compromising analytical testing accuracy is a continuous challenge in the anatomical pathology lab. The aim of this study was to determine under controlled conditions the effects of variable fixation time on breast biomarker expression in human breast cancer cell line-derived xenografted (CDX) tumours.

Methods CDX tumours using strong oestrogen receptor (ER)-positive, Her2-negative (MCF7) and weak ER-positive, Her2 equivocal (T47D) breast cancer cell lines were fixed for various times ranging from 1 to 336 hours in 10% neutral buffered formalin. CDX tumours were processed according to routine biomarker testing protocols and stained for ER and Her2 immunohistochemistry (IHC) and processed for HER2 fluorescence in situ hybridisation (FISH). The tumours were evaluated using Allred scoring for ER and current ASCO/CAP guidelines for Her2, and by objective cell counting methodology.

Results No differences were found in expression of ER in either MCF7 or T47D CDX tumours under variable fixation. T47D tumours displayed equivocal Her2 staining when fixed for 24 hours, but fixation for ≤8 hours resulted in consistently negative staining while tumours fixed for >72 hours demonstrated consistent equivocal staining (p<0.01). Cell counting assays revealed only a significant increase in sensitivity in tumours fixed for >72 hours (p<0.01). As expected, FISH results were unaffected by variable fixation.

Conclusions Neither shortened nor prolonged fixation affects ER expression, consistent with previous findings. In equivocal Her2-expressing tumours, however, increasing fixation increased the sensitivity of Her2 IHC reporting while not affecting FISH.