Article Text
Abstract
Introduction/Background The quality of pathological assessment is of pivotal importance in the management of cervical cancer patients if pelvic lymph node dissection (PLND) is to be replaced by sentinel lymph node (SLN) biopsy only.
Methodology The aim was to report the results of the quality control of pathological SLN assessment in the group of 300 patients treated per protocol. All specimens and full pathology reports from at least two patients per site were reviewed centrally in Prague. SLN ultrastaging protocol included a complete processing of all SLN tissue in slices of 2 mm thickness, 2 sections in 150 µm from each block until no tissue left, one stained with H&E and second examined immunohistochemically. The findings were classified as: Minor (minor protocol deviations), Major (additional sections completed at central pathology), or Critical (deviations could not have been fixed at central pathology). In the case of Major or Critical findings, sites were requested to submit all samples from all additional patients for the second-round assessment.
Results Samples from 77 cases from 36 sites with at least 1 patient treated per protocol were submitted for the central reading. Minor findings were identified in 27% of cases, Major in 30%, and Critical in 4%. Samples from 18 patients were submitted for the second control round; 9 were without deviations 7 had Minor, 2 Major, and 1 Critical findings. One site was prematurely closed due to repeated Critical findings. The most frequent deviations included incompletely processed SLN (36/95) and a lower number of immunohistochemical staining (31/95).
Conclusion Major deviations from the SLN protocol were identified in 33% of sites and 31% of samples. These deviations can impact detection of small metastases and, consequently, the safety of the SLN concept. Central pathology reading allowed for substantial improvement of SLN pathological assessment during the trial.
Disclosure Acknowledgements: This work was supported by a grant from the Czech Research Council (No 16-31643A). None of the authors declare a conflict of interest.