Correlation between Intratumoral Cellular Microenvironment, Clinicopathological Parameters and Prognosis in Hodgkin Lymphoma.

Hodgkin lymphoma (HL) is characterized by a minority of neoplastic, Hodgkin, Reed-Sternberg (HRS) cells surrounded by a heterogeneous background of non-neoplastic cells. Previous studies have described a prognostic role of the cellular composition in HL. The aim of the present study was to correlate parameters related to the intratumoral cellular microenvironment with clinicopathological features and prognosis in a cohort of previously untreated HL patients. For this purpose a tissue microarray Array (TMA) consisting of pairs of representative cores (2mm) from each tissue specimen and from positive and negative controls was constructed. Immunohistochemical stains of cellular antigens were assessed and graded. Tissue samples and clinical record data from a total of 87 HL patients were analysed. Histological subtypes were distributed as follows: nodular sclerosis (NS)61 cases (71%), mixed cellularity (MC) 15 (17%), nodular lymphocyte- predominant (NLP) 9 (10%) and lymphocyte depleted 2 (2%). The median age was 36.7 yrs (range 5–83 yrs and the male to female ratio 1.6. 56 had localised (67%) 27 had advanced stage disease (32%). Anaemia and leukocytosis were present in 4 (5%) and 15 (17%) cases, respectively. All patients were treated according to standard guidelines (chemo- radiotherapy for localised disease and ABVD/COPP- based chemotherapy for advanced stage disease). Immunohistichemical stains were used to characterise and quantify the neoplastic and non-neoplastic cells within tumor lesions. Following antigens were studied: CD3, CD20, CD30, CD15, CD10, STAT-6, MUM-2, FAS, p53 and MIB-1. Among histological subtypes NLP HL was, as expected, the one most frequently displaying CD20 positivity, while CD15 and CD30 were more common in classical HL (cHL). A strong CD3 expression in the non-neoplastic by-stander population was inversely correlated with the presence of leukocytosis in the peripheral blood. An opposite pattern was found in cases, mostly cHL, with high expression of STAT-6 in the tumor cell population. In addition to leukocytosis STAT-6 expression also correlated to younger age (60 yrs or less) and presence of B-symptoms. Time-related endpoint analysis was restricted to NS, representing the largest subgroup (n=61). Interestingly, a significant adverse impact on overall survival (OS) was detected in cases with a strong tumor cell expression of STAT-6. Even more marked was the unfavorable prognostic value of low or lacking tumor cell expression of CD30 (p=0.0012) or by-stander cell expression of CD3 (p=0,0133). All these findings were far more evident in younger (≤ 60 yrs) NS-HL patients than in their elderly counterparts. These three parameters where not, or only minimally, influenced by classical clinical prognosticators such as clinical stage, B-symptoms or anemia. Moreover, the adverse prognostic influence of these factors was mutually increased in those cases, where they occurred in combination (high STAT6 + low CD30: p=0.0001; low CD30 + low CD3: p=0.0004; high STAT6 + low CD3: p=0.0017). The role of the STAT family of proteins and that of T-regulatory cells in the pathophysiology of HL should be further elucidated.