Histological subtypes drive distinct prognostic immune signatures in classical Hodgkin lymphoma.

Classical Hodgkin lymphoma (cHL) is a highly curable disease, with about 80% of patients cured using standard first-line chemotherapy. However, outcomes for relapsed/refractory patients re-main unfavorable and there is a critical lack of predictive biomarkers for early identification of these patients who may benefit from new therapeutic strategies. Here we evaluated the dynamic expression of 571 immune-related genes in a cohort of 42 cHL using NanoString technology. We identified a 19-genes immune signature predictive of relapse at the time of diagnosis, which was found to be strongly dependent on histological subtype. Moreover, comparative analyses be-tween paired diagnostic/relapsed biopsies of nodular sclerosis cHL showed 134 differentially expressed genes, highlighting an immune contexture shift at relapse not found in the mixed-cellularity subtype. Overall, these results strongly suggest that the predictive value of immune signature in cHL is influenced by histological subtype, a criterion that should be con-sidered when assessing new immunotherapy strategies. Sixty-six formalin fixed paraffin embedded (FFPE) cHL biopsies at diagnosis and/or at relapse were obtained from a total of 42 patients, including 30 r/r cHL patients and 12 non-r/r cHL patients. Thirty-five FFPE samples were available at diagnosis (n= 12 non-r/r cHL and n=23 r/r cHL) and FFPE paired diagnosis and relapse biopsies were available for 22/30 r/r cHL patients. Nodular sclerosis subtype was the most represented histological subtype in our series (n=24/42, 57.1%) followed by the mixed-cellularity subtype (n=15/42, 35.7%). Only two patients presented a lymphocyte-rich pattern (4.8%), and one patient displayed a lymphocyte-depleted pattern (2.4%). The average age at diagnosis of r/r patients was 30 years (ranging from 15y to 83y) and that of non-r/r patients was 31 years (ranging from 23y to 58y). Less than 25% (n=9/39) of cases were Epstein-Barr virus (EBV) positive. EBV status was unknown for 3 cases. Thirty-two (78%) patients, including 20/30 r/r patients and 12/12 non-r/r patients, had extensive disease at diagnosis (Ann Arbor stage III-IV). Of the 40 assessable patients, 30.9% (n=13) were treated with bleomycin, etoposide, doxoru-bicin, cyclophosphamide, vincristine, procarbazine and prednisone (BEACOPP) and 57.1% (n=24) received doxorubicin, bleomycin, vinblastine and dacarbazine (ABVD). The 3 remaining patients received alternative chemotherapy