Quality assurance review: Intra-operative evaluation of sentinel lymph nodes in breast cancer

Background Intraoperative consultation (IOC) of axillary sentinel lymph node (SLN) biopsy continues to play a role in selected breast cancer patients. The reported sensitivity rates for intraoperative SLN evaluation in breast cancer range from 47% to 80%. We study a center where the majority of SLN IOC is performed by imprint cytology, and a protocol was established to limit microscopic examination to three slides for a reporting TAT goal of 30 min. Methods Approval to conduct this study was obtained from the REB. A retrospective review was performed on all consecutive SLN cases sent for IOC. Reported IOC assessments of all cases were compared with the final pathology. Results Of 164 patients, there were 22 (13%) false negative IOC events, including 15 missed macro-metastasis and 7 missed micro-metastasis. The overall sensitivity for touch imprint in detecting SLNs macro-metastasis was 70.9%. Reporting total turnaround time was on average 3 min longer, whereas sensitivity and specificity were not significantly different in the two protocol periods. Conclusion Implementation of an IOC policy for a maximum of three slides for imprint cytology did not result in a significant impact on the sensitivity, specificity, or total turnaround time for SLN in breast cancer patients. False negative IOC events were mainly due to sampling error. Quality review was made difficult by limited documentation related to the gross handling of the specimens at IOC. System factors identified include insufficient space for the IOC report on the pathology requisition, and the lack of clearly communicated expectations for documentation.

Automated summary

The study investigated the results of IOC of axillary sentinel lymph node in breast cancer patients. In imprint cytology, microscopic examination was limited to a maximum of three slides, with an average turnaround time slightly longer than the goal time. False negative events during surgery were mainly attributed to sampling errors.