P013 (0088) DIGITAL IMAGE ANALYSIS OF THE MICROENVIRONMENT IN CLASSICAL HODGKIN LYMPHOMA COMPARING PRIMARY TUMOR AND RELAPSE

doi: 10.1097/01.HS9.0000547863.89384.e7
Biology and Microenvironment
Free

Franziska Jochims1, Sarah Kretschmer1, Ron Jachimowicz2,3, Falko Fend4, German Ott5, Martin-Leo Hansmann6, Christoph Thorns7, Annette Plütschow3, Andreas Engert2,3, Wolfram Klapper1

1Institut für Pathologie, Sektion Hämatopathologie, Universitätsklinikum Schleswig-Holstein, Campus Kiel,2Innere Medizin I, Universitätsklinikum Köln,3The German Hodgkin Study Group (GHSG),4Institut für Pathologie und Neuropathologie, Eberhard-Karls-Universität, Tübingen,5Abteilung für Klinische Pathologie, Robert-Bosch-Krankenhaus, und Dr. Margarete Fischer-Bosch Institut für Klinische Pharmakologie, Stuttgart,6Dr. Senckenberg Institut für Pathologie, Universitätskrankenhaus Frankfurt, Goethe-Universität, Frankfurt am Main,7Institut für Pathologie, Universitätsklinikum Schleswig-Holstein, Universität zu Lübeck

Classical Hodgkin lymphoma (cHL) is composed of an abundant inflammatory microenvironment whereas the neoplastic Hodgkin-Reed-Sternberg Cells (HRSC) represent less than 1% of the cellularity. The microenvironment has become a focus of research because of its association with patient outcome and as target for immune-therapy. So far only a limited number of studies compared the microenvironment at the time of first diagnosis (primary tumor) to the relapse. We aimed to understand the dynamics of the composition of the microenvironment from primary tumor to relapse. Since relapses of cHL are frequently more aggressive diseases than the primary tumor we hypothezised that this shift in aggressiveness might be accompanied by alterations of the microenvironment.

Paired tumor samples of 17 patients were identified in the files of the hematopathology department in Kiel (test cohort) and 21 in other centers in Germany (validation cohort). All patients had been treated before the era of immune checkpoint inhibitors. Tissue sections were stained immunohistochemically for CD3 (T-lymphocytes), CD20 (T-lymphocytes), CD30 (HRSC) and CD68 (macrophages), scanned with high resolution and whole slides were analyzed by image analysis (Tissue Studio, Definiens). CD3 and CD20 were evaluated as percentages of cells in relation to all cells in the tissue, CD30 and CD68 were analyzed as percent of stained area.

The average T-cell content in the primary tumor (P) was 49.55% (SD 17.35) compared to relapse (R) 37.74% (SD 16.21), B-cell content for P was 29.84% (SD 17.26) to R 32.65% (SD 18.86), macrophage areas for P was 3.89% (SD 4.19) to R 2.72% (SD 2.79) and HRSC area for P was 6.2% (SD 7.09) to R 3.89% (SD 3.45) in the test cohort. We did not observe a statistically significant change in B-cell, macrophage and HRSC content between primary diagnosis and relapse. However, we observed a tendency of T-cells to decrease from primary tumor to relapse (p = 0,0765; paired t-test). The decrease in T-cell content was confirmed in the validation cohort although statistical significance was not reached. We demonstrate for the first time, that the overall content of T-cells in cHL decreases from primary biopsy to relapse. The evolution of cHL under conventional chemotherapy therapy thus leads to a reduced T-cell content in the lymphoma. Future studies need to address, whether this finding reflects a reduction of T-cell dependence of the HRSC at relapse.

Copyright © 2018 The Authors. Published by Wolters Kluwer Health Inc., on behalf of the European Hematology Association.