Association of aplastic anaemia and lymphoma: a report from the severe aplastic anaemia working party of the European Society of Blood and Bone Marrow Transplantation

Abstract

Although anecdotal reports showing the uncommon association of aplastic anaemia (AA) and lymphoma have been published (Dorr et al, 1996; Koziner et al, 1975; Medinger et al, 2012; Suzuki et al, 2009; Veidt et al, 2005; Yoshioka et al, 1999; Zonder et al, 2002), data related to its incidence, diagnostic characteristics and centre management strategies are lacking. The Severe Aplastic Anaemia Working Party of the European Society of Blood and Bone Marrow Transplantation (SAAWP-EBMT) aimed to evaluate the diseases characteristics and treating centre attitudes regarding the treatment and outcome of patients suffering from both AA and lymphoma. Between 2013–2015, we collected data of patients with this combination. Congenital bone marrow failures, as well as AA or lymphoma that occurred after haematopoietic stem cell transplantation (HSCT) were excluded. Eighty-three (28%) of the 294 EBMT centres contacted participated in this study; 13 (16%) centres, from 7 different countries, reported the 26 cases included in the study. AA was diagnosed between 1983–2015; median age at AA diagnosis was 57 (10–78) years, 16 patients (62%) were male, 13 patients had severe AA (SAA) and 6 very severe AA (vSAA). Of the 18 cases investigated for paroxysmal nocturnal haemoglobinuria (PNH) clones by flow cytometry, two had a small clone, without clinical PNH signs or symptoms. Cytogenetics was available in 22/26(85%) cases; only one had an anomaly (trisomy 12). At AA diagnosis, 85% and 86% of patients required red blood cell and platelet transfusions, respectively. Lymphoma was diagnosed between 1958 and 2015, at a median age of 52 years (12–78). Five patients (19%) had Hodgkin lymphoma (HL) for which histopathology was available in 4 cases: 2 lymphocyte predominant type, 1 nodular sclerosis and 1 mixed cellularity. Non-Hodgkin lymphoma (NHL) was reported in 21 (81%) patients: 19 (73%) had a B-cell NHL and two patients an unspecified lymphoma. A subtype of B-cell NHL was reported in 18 cases: 4 diffuse large B-cell, 3 lymphoplasmacytic, 3 follicular, 3 nodal marginal zone, 2 chronic lymphocytic leukaemia, 1 splenic marginal zone, 1 mantle cell and 1 plasma cell neoplasm. The lymphoma was treated in 23 cases (91%). Lymphoma was detected at three different times with respect to AA (Fig 1): 11 patients presented with lymphoma before AA, 7 patients were simultaneously diagnosed with AA and lymphoma, and 8 patients presented with lymphoma after AA diagnosis . Patients presenting AA and lymphoma simultaneously were significantly older compared to the other times of presentation (Table I). HL was mainly observed in patients presenting lymphoma before AA (4 out of 5 HL). Various underlying pathogenic mechanisms may be involved in the different times of presentation. AA occurring after lymphoma and/or its treatment could have a different cause of marrow failure than typical AA. Alkylating agents included in lymphoma therapy, could lead to exhaustion of the stem cell pool, resulting in subsequent marrow failure (Gobbi et al, 2009). In line with this hypothesis, we found that more patients with AA occurring after lymphoma failed to respond to standard immunosuppressive therapy (IST) and needed a HSCT (Table I). Likewise, the use of purine nucleoside analogues have significant cytotoxic activity, resulting in prolonged lymphocyte depletion, especially in CD4 T-cells; this immune dysregulation might facilitate autoimmunity. In our series, only one patient was treated with fludarabine before the development of AA. When AA and lymphoma occur concomitantly, AA may emerge as a paraneoplastic autoimmune phenomenon of the lymphoma (Chandor, 1988), although our data could not fully confirm this hypothesis. In 3 of the 7 cases with concomitant presentation, the lymphoma was treated first, aiming to control both diseases; however, after chemotherapy, complete remission of the lymphoma was achieved in 2 of these cases but the AA did not respond and needed subsequently therapy. Furthermore, lymphoma has been reported following administration of anti-thymocyte globulin (ATG) (Calistri et al, 2006). In this series, lymphoma was diagnosed in 2 patients within one year after ATG therapy; in both cases, the AA remained active at lymphoma diagnosis (1 partial response, 1 non-responder). In this study, none of the reported lymphomas were Epstein–Barr virus associated. All these pathophysiological hypotheses concerning the association of AA and lymphoma are conceivable, but need to be proven. In patients where AA appeared first, the standard AA approach for therapy was observed, thus all patients received Correspondence