Prospective Use of a Triple FISH Probe to Predict sMDS/AML Following High Dose Therapy with Autologous Stem Cell Rescue for Lymphoma.

Abstract 3459

Secondary myelodysplasia/acute myeloid leukaemia (sMDS/AML) is an almost universally fatal complication of high-dose chemotherapy with autologous stem cell rescue (HDT+ASCR) for lymphoma, occurring in 2–10% of recipients at five years (Pedersen-Bjergaard et al. Blood, 2000; 95: 3273-9). There are no reliable predictive factors except prior chemotherapy and radiotherapy. At St Bartholomew's Hospital, London (Barts) 33/230 (14%) patients (pts) having cyclophosphamide+total-body irradiation (Cy-TBI)-conditioned HDT+ASCR for lymphoma developed sMDS/AML after a median follow-up of 6 years, and a retrospective analysis in those in whom a pre-Cy-TBI sample was available showed cytogenetic abnormalities in all of them. Subsequently a triple fluorescent in-situ hybridisation cytogenetic probe (TF) for the commonest abnormalities seen in sMDS/AML (EGR1: 5q31, RB1: 13q14 and D7Z1: 7q31) was developed (Lillington et al. Ann Oncol, 2002; 13: 40-3). The aim of the current study was to analyse the prospective use of TF to identify those at greatest risk of MDS, by including the test as part of the standard pre-transplant workup for all patients receiving the successor to Cy-TBI at Barts: BEAM (BCNU, Etoposide, Ara-C, Melphalan) HDT+ASCR during a 7-year period 2002–2009. 133 pts (47% female, 53% male; median age: 47, range: 18–68) received BEAM HDT for lymphoid malignancies during this time (22% diffuse large B-cell lymphoma, 21% follicular lymphoma, 11% transformed follicular lymphoma, 33% Hodgkin lymphoma and 13% other non-Hodgkin lymphoma). Median time from diagnosis to HDT was 29 months (range: 7–275 months). TF was performed in 125 cases (94%), using a pre-transplant bone marrow sample in 95% and harvested stem cells in 5%. An abnormal result (5q31 del) was detected in one case (0.8%). After a median follow-up of 64 months (range: 24–103), 4/133 (3%) patients developed sMDS/AML at a median of 23 months (6-85) following treatment, including the single patient with an abnormal TF result before BEAM. The TF pre-BEAM showed no abnormalities in 2 cases developing sMDS/AML and was not performed before HDT in the fourth. The low incidence of pre-transplant TF abnormalities in BEAM candidates (0.8%) compared to historic Cy-TBI candidates, and the failure of the test to detect the remaining cases (2.4%) has demonstrated this not to be an effective risk-stratifying tool. Whether withholding a potentially beneficial treatment from patients in whom an abnormality is detected would prevent or delay sMDS/sAML is unknown. However identifying patients at greatest risk of this almost universally fatal complication remains a challenge and prospectively assessable risk factors should continue to be sought while outcomes remain so dire.