Predicting non-sentinel lymph node status after positive sentinel biopsy in breast cancer: What model performs the best in a Czech population?

Oldřich Coufal, Tomáš Pavlík, Pavel Fabian, Rita Bori, Gábor Boross, István Sejben, Róbert Maráz, Jaroslav Koča, Eva Krejčí, Iva Horáková, Vendula Foltinová, Pavlína Vrtělová, Vojtech Chrenko, Wolde Eliza Tekle, Mária Rajtár, Mihály Svébis, Vuk Fait, Gábor Cserni

Research output: Article

51 Citations (Scopus)

Abstract

Several models have previously been proposed to predict the probability of non-sentinel lymph node (NSLN) metastases after a positive sentinel lymph node (SLN) biopsy in breast cancer. The aim of this study was to assess the accuracy of two previously published nomograms (MSKCC, Stanford) and to develop an alternative model with the best predictive accuracy in a Czech population. In the basic population of 330 SLN-positive patients from the Czech Republic, the accuracy of the MSKCC and the Stanford nomograms was tested by the area under the receiver operating characteristics curve (AUC). A new model (MOU nomogram) was proposed according to the results of multivariate analysis of relevant clinicopathologic variables. The new model was validated in an independent test population from Hungary (383 patients). In the basic population, six of 27 patients with isolated tumor cells (ITC) in the SLN harbored additional NSLN metastases. The AUCs of the MSKCC and Stanford nomograms were 0.68 and 0.66, respectively; for the MOU nomogram it reached 0.76. In the test population, the AUC of the MOU nomogram was similar to that of the basic population (0.74). The presence of only ITC in SLN does not preclude further nodal involvement. Additional variables are beneficial when considering the probability of NSLN metastases. In the basic population, the previously published nomograms (MSKCC and Stanford) showed only limited accuracy. The developed MOU nomogram proved more suitable for the basic population, such as for another independent population from a mid-European country.

Original languageEnglish
Pages (from-to)733-740
Number of pages8
JournalPathology and Oncology Research
Volume15
Issue number4
DOIs
Publication statusPublished - dec. 1 2009

    Fingerprint

ASJC Scopus subject areas

  • Pathology and Forensic Medicine
  • Oncology
  • Cancer Research

Cite this

Coufal, O., Pavlík, T., Fabian, P., Bori, R., Boross, G., Sejben, I., Maráz, R., Koča, J., Krejčí, E., Horáková, I., Foltinová, V., Vrtělová, P., Chrenko, V., Eliza Tekle, W., Rajtár, M., Svébis, M., Fait, V., & Cserni, G. (2009). Predicting non-sentinel lymph node status after positive sentinel biopsy in breast cancer: What model performs the best in a Czech population? Pathology and Oncology Research, 15(4), 733-740. https://doi.org/10.1007/s12253-009-9177-6