High inter- and intraobserver inconsistencies in the assessment of MIB-1-based proliferative acitity of breast carcinomas: a survey of the Swiss working group of gynaeco- and breast pathologists

Adjuvant chemotherapy decisions in breast cancer are increasingy based on the pathologist's assessment of tumor proliferation. The Swiss working group of gyneco- and breast pathologists has surveyed the inter- and intraobserver consistencies of MIB-1 immunohistochemistry read-outs.

Methods: In a 1st round, 5 pathologists evaluated MIB-1-labeling index (LI) in 10 breast cancers by counting (among 500 tumor cells) and by eyeballing. In a 2nd round, 15 pathologists evaluated MIB-1-LI in three G2 breast cancers by counting and eyeballing, but in predefined fields of interest. In both rounds, MIB-1 was assessed in centrally immunostained slides and on slides immunostained in the participating laboratories. Results were discussed among the authors as well as ways to improve consistencey. In a 3rd round, four months later, the same three carcinomas were sent out once more for analysis by the same pathologists, along with three new carcinomas.

Results: We found a high degee of inter- and interobserver inconsistency in the read-outs of MIB-1-LI. Inconsistencies were not explained by any of the following factors: (i) pathologists' divergent definitons of what qualifies as a positive nucleus (ii) the mode of assessment (counting vs. eyeballing), (iii) immunostaining technique/protocol/antibody, and (iv) the selecetion of the area in which to count. Despite increasing awareness, consistency did not improve from round two to round three.

Conclusion: All authors of this survey were surprised how inconsistent we were in assessing MIB-1-LI. Oncologists should be aware of this caveat when considering MIB-1-LI as a basis for treatment decisions in moderately differentiated breast carcinomas.