Stage migration after minor changes in histologic estimation of tumor burden in sentinel lymph nodes

Abstract

Melanoma metastasis size estimates are of prognostic significance for groups of patients, but to the authors' knowledge, measurement consensus does not exist. Maximum metastasis diameter, maximum centripetal tumor depth, microanatomic location of metastases, and complete metastasis volume were measured in 156 positive sentinel lymph nodes (SLNs) from 99 melanoma patients. The number of SLN-positive patients was increased by up to 41% using complete step-sectioning compared with less extensive protocols. Assessing maximum metastasis diameters, up to 27% of patients positive by the less extensive protocols went from 1 metastasis diameter group to a larger one when complete step-sectioning was performed. No patients were down-staged. Apparently minor protocol changes (eg, adding an extra step) led to substantial changes in maximum metastasis diameter. Similar protocol-dependent results were noted measuring the maximum centripetal tumor depth and the microanatomical location of metastases. By using semiquantitative tumor burden estimates, stage migration was always unidirectional (ie, moving from a lower to higher stage). Stereologic tumor burden estimates in step-sectioned SLNs also varied according to the number of step sections assessed, but could increase, decrease, or remain constant, so that stage migration was multidirectional. Adding extra steps to pathology protocols when assessing semiquantitative parameters leads to unidirectional stage migration ("the protocol trap"). This systematical bias makes it difficult to base treatment decisions on semiquantitative metastasis size estimates. Although based on metastatic melanoma, the principles described herein will apply when measuring nodal tumor burden in other metastasizing cancers, including breast carcinoma.