False-Negative Interpretation of Breast Sentinel Lymph Node Touch Preps: Analysis of the Causes with Suggestions to Improve Diagnostic Accuracy

Abstract

Sentinel lymph node (SLN) biopsy has become widely accepted as an important procedure in staging breast cancer. False-negative results of touch prep (TP) examination at time of SLN biopsy requires additional surgery, delaying treatment and increasing cost. Therefore, we have analyzed our experience with false-negative interpretation on SLN TP’s. Eight-hundred and three consecutive SLN biopsies from 2003 to 2005 were obtained from the pathology archive of Roswell Park Cancer Institute. The intraoperative consultation results were correlated with the final diagnoses.  Twenty-five SLN intraoperative consultations had false-negative TP’s [false-negative rate = 3.1% (25/803), including 9 metastatic lobular carcinomas and 16 metastatic ductal carcinomas]. These cases were re-evaluated by 3 pathologists independently, and the metastases in the SLN sections were confirmed by positive cytokeratin staining.  Size of the metastatic focus, nuclear grade and the adequacy of TP’s were analyzed with regard to the cause of false-negative results. On re-screening of TP’s, we found that rare tumor cells of low nuclear grade were identified on 28% (7/25) of the TP’s (3 metastatic lobular carcinomas and 4 metastatic ductal carcinomas). In the remaining 72% (18/25) of TP’s, re-screening revealed no evidence of tumor.  Evaluation of these TP’s demonstrated that 50% (9/18) were unsatisfactory for evaluation or limited by scant cellularity. While cases that remained negative on re-screening tended to have smaller measured foci of tumor in the SLN (Average 0.65 mm vs. 0.94 mm from cases that were positive on re-screening), there was considerable overlap between these two groups. In conclusion, TP’s with scant cellularity, unsatisfactory TP’s and failure to identify tumor cells with low nuclear grade were found to significantly contribute to false-negative interpretations. We suggest that an additional TP or frozen section may be necessary if the cellularity of the initial TP is limited.  Correlation with the original core biopsy may be of value to help in identifying cancer cells of low nuclear grade.