AUTOMATED IMAGE ANALYSIS REVEALED A HIGH VARIABILITY OF CD30+HRS CELLS DENSITY IN CLASSICAL NODULAR SCLEROSIS HODGKIN LYMPHOMA

Background

Hodgkin lymphoma, classic type (cHL) tissue is characterized by an extreme disproportion between tumor (Hodgkin-Reed-Sternberg, HRS) cells ranging from 0.1% to 10% and surrounding non-malignant microenvironment cells. Last years brought an explosion of novel information about close HRS-microenvironment cross-talk, which promotes a tumor growth. Numerous studies tested the predictive value of microenvironment cells density using immunohistochemistry (IHC) methods and many of them were focused on lymphoma-associated macrophages (LAM). High LAM density at time of diagnosis was found to be an inferior prognostic factor. However, no studies were performed on HRS cells density and HRS:LAM ratio analysis. 

Aims

To assess an impact of HRS cells density and HRS:LAM relationship on disease development/prognosis in cHL, nodular sclerosis subtype using novel automated scanning system of the total tumor sample area. 

Methods
We analyzed cHL tissues, obtained at time of diagnosis, in fourteen patients with intermediate and advanced stages (KS IIB-IV). Six patients represented a relapsed/refractory (RR) cohort, eight patients were selected as age-sex matched relapse-free (RF) controls. Paraffin-embedded biopsies were prepared from pre-selected high-quality diagnostic samples stained with hematoxylin-eosin for morphology analyses and using CD30 (HRS) and CD68 (LAM) for cell-density analyses. Tissue array analyses were performed using the TissueFAXS (TissueGnostics, Vienna, Austria) system that combines detailed morphologic information offered by microscopy with the scientific accuracy of multi-channel flow cytometry. Data were analyzed with TissueQuest software. Each sample was previewed by scanning software and only significant tumor tissue was manually gated for further analyses, excluding fibrotic (≥10 %), necrotic or residual lymphatic structures. This area was considered as a Total Tumor Sample (TTS) area.

Results
Mean TTS area covered 17.7±11.9 mm2 with mean total number of cells 457,257±397,599. The mean TTS was not different in RF (17.9 mm2) compared to RR cohort (19.2 mm2, p=0.86). There was a trend for higher cell density in RF compared to RR cases (27,679 vs 20,861 cells per mm2, p=0.09). Density of CD30+ HRS cells was 2.4 fold higher in RF than in RR patients (429 vs 180 cells per mm2, p=0.055). Although mean density of CD68+ LAM was comparable in RF and RR group (2,372 and 2,716 cells per mm2, respectively, p=0.66), a CD30:CD68 ratio was higher in RF (0.31) than in RR (0.07, p=0.044) patients. The number of LAM per one HRS cell was 33.1 in RR, whereas only 12.1 in RF group (p=0.10). Our data showed that automated analysis of a TTS area may overcome technical limitations caused by a tissue selection, heterogeneity in HRS and bystander cell distribution or tissue fibrosis/necrosis comparing to analysis of small tissue samples (2-3 mm2).

Summary

Cell distribution in cHL tissue is highly variable, automated analysis of may bring more accurate cell-density results with lower variance. We describe differences in HRS and total cell density in diagnostic biopsies of relapse/refractory and relapsed- free cHL patients. Higher HRS:LAM ratio in relapsed patient presume more potent HRS-microenvironment interaction with the possible prognostic role.

Acknowledgement: supported by grant from the Faculty of Medicine and Dentistry, Palacky University Olomouc (IGA-LF-2015-001).

 

Keyword(s): CD30, Hodgkin's disease, Macrophage, Microenvironment

www.eha.org