Fulltext

Interlaboratory variability of MIB1 staining in well-differentiated pancreatic neuroendocrine tumors

Blank, Annika ; Wehweck, Laura ; Marinoni, Ilaria ; Boos, Laura ; Bergmann, Frank ; Schmitt, Anja ; Perren, Aurel

In: Virchows Archiv, 2015, vol. 467, no. 5, p. 543-550

Zum persönliche Liste hinzufügen
    Titel
    • Interlaboratory variability of MIB1 staining in well-differentiated pancreatic neuroendocrine tumors
    Autor
    • Blank, Annika. University of Bern, Institute of Pathology, Murtenstr. 31, CH-3012, Bern, Switzerland
    • Wehweck, Laura. Department of Cardiology, Pneumology and Internal Intensive Care Medicine, Klinikum Neuperlach, Städtisches Klinikum München GmbH, Oskar-Maria-Graf-Ring 5, 83575, Munich, Germany
    • Marinoni, Ilaria. University of Bern, Institute of Pathology, Murtenstr. 31, CH-3012, Bern, Switzerland
    • Boos, Laura. University of Bern, Institute of Pathology, Murtenstr. 31, CH-3012, Bern, Switzerland
    • Bergmann, Frank. University of Heidelberg, Institute of Pathology, Im Neuenheimer Feld 224, 69120, Heidelberg, Germany
    • Schmitt, Anja. University of Bern, Institute of Pathology, Murtenstr. 31, CH-3012, Bern, Switzerland
    • Perren, Aurel. University of Bern, Institute of Pathology, Murtenstr. 31, CH-3012, Bern, Switzerland
    Dokumententyp
    • Postprint
    Sprache
    • Englisch
    Veröffentlicht in
    • Virchows Archiv, 2015, vol. 467, no. 5, p. 543-550. Springer Berlin Heidelberg
    Andere elektronische Ausgabe
    OAI-PMH-ID
    • oai:doc.rero.ch:332720
    Summary
    Neuroendocrine tumors (NET) are routinely graded and staged to judge prognosis. Proliferation index using MIB1 staining has been introduced to assess grading. There are vivid discussions on cutoff definitions, automated counting, and interobserver variability. However, no data exist regarding interlaboratory reproducibility for low proliferation indices which are of importance to discriminate between G1 and G2 NET. We performed MIB1 staining in three different university hospital-based pathology laboratories on a tissue micro array (TMA) of a well-characterized patient cohort, containing pancreatic NET of 61 patients. To calculate the proliferation index, number of positive tumor nuclei was divided by the total number of tumor nuclei. Labeling index was compared to mitotic counts in whole tissue sections and to clinical outcome. Linear regression analysis, intraclass comparison, and log-rank analysis were performed. Intraclass correlation showed moderate-to-fair agreement. Especially low proliferating tumors were affected by interlaboratory differences. Log-rank analysis was performed for each lab and resulted in three different cutoffs (5.0, 3.0, and 0.5%). Every calculated cutoff stratified the patient cohort to a significant extent for the underlying stain (p<0.001, <0.001, and <0.001) but showed no or lesser significance when applied to the other stains. Significant and relevant interlab differences for MIB1 exist. Since the MIB1 proliferation index influences grading, local cutoffs or external standardization should urgently be introduced to achieve reliability and reproducibility.