The Risk of Sinusoidal Obstruction Syndrome (SOS) in Allogeneic Hematopoietic Stem Cell Transplantation after Prior Gemtuzumab Ozogamicin Treatment: A Retrospective Study on Behalf of the Acute Leukemia Working Party of the European Group for Blood and Marrow Transplantation

Background. Gemtuzumab ozogamicin (GO) is a humanized anti-CD33 monoclonal antibody conjugated with calicheamicin, used in acute myeloid leukemia (AML) either as monotherapy or in association with chemotherapeutic agents (Hamann et al. Bioconjug. Chem 2002). It is known to induce a significant liver toxicity that might increase the risk of sinusoidal obstruction syndrome (SOS), especially in patients undergoing allogeneic haematopoietic stem cell transplantation (allo-HSCT). Risk is especially high in patients  transplanted within a short interval from last GO dose (≤ 3.5 months) (Wadleigh et al. Blood 2003). The current retrospective analysis aimed to assess the incidence of SOS and the outcome of AML patients receiving GO before allo-HSCT.

Patients and methods. The primary endpoint of the study was the assessment of SOS incidence in a series of 146 adult patients (71 women and 75 men) with AML undergoing allo-HSCT after prior exposure to GO. Relevant data were captured by a specific designed questionnaire focusing on SOS (Med C) that was distributed to all EBMT centers. Secondary endpoints were engraftment, non-relapse mortality (NRM), graft-versus host disease (GVHD), leukemia-free survival (LFS) and overall survival (OS). Median age was 50 years (range 19-70). In most cases (n=127 out of 137 for which the information was available, 93%) GO was administered in parallel with other chemotherapeutic agents. Median GO dose was 3 mg/m². Median number of cycles before allo-HSCT was 1 (range 1-6). The majority of the patients (n=79, 54%) received GO as induction treatment, while 51 (35%) and 16 (11%) received GO for relapsing or primary refractory disease, respectively. At time of allo-HSCT  97 (66%) patients were in complete remission. Of note, 11 patients underwent prior allo-HSCT and 8 patients prior auto-HSCT. Most patients received a reduced intensity conditioning (RIC) regimen pre allo-HSCT (n=84, 58%). SOS prophylaxis was given in a total of 69 patients (heparin n=57, ursodeoxycholic acid n=8, and defibrotide n=4). Cumulative incidence (CI) and Kaplan-Meier estimates were used when appropriate; Cox model was used for multivariate analysis.

Results. Overall, cumulative incidence of SOS was 7.75% (95% CI 4.1-12.9) (n=11). Among patients receiving RIC, 4 (4,9%) developed SOS as compared to 7 (11,5%) among patients receiving MAC regimen (p=NS). SOS was the main cause of death in 3 out of 11 cases. Median interval between last GO dose and allo-HSCT was 130 days (range 13-1126). Neither OS nor SOS incidence differed significantly for patients receiving GO shortly before allo-HSCT (≤ 3.5 months) as compared to the others. None of the analyzed risk factors had an impact on the risk of SOS in our study. With a median of 15 days, CI of neutrophil and platelet engraftment was 93.8% (95% CI 88.1-96,8) and 89% (95% CI 82.3-93.3), respectively. CI of acute GVHD at day 100 was 30.6% (95% CI 23-38.2), occurring in 84 patients (out of them 45 had grade II-IV GVHD), while CI of chronic GVHD at 5 years was 25.5% (95% CI 18.6-33).

With a median follow-up of 64 months the probability of OS and LFS at 5 years was 40% (95% CI: 31-48) and 37% (95% CI: 29-45), respectively. Cumulative RI and NRM at 5 years were 42% (95% CI: 34-50%) and 21% (95% CI: 14-28), respectively. In multivariate analysis patients transplanted with an active disease relapsed more often and had worse LFS and OS, while patients presenting with liver abnormalities before allo-HSCT had a worse OS (p<0.03).

Conclusion. Our results in a large cohort of AML patients show that GO administration prior to allo-HSCT does not result in an excessive incidence of SOS as compared to historical controls reported in the literature (Carreras et al. Biol. Blood Marrow Transplant 2011). Notably incidence is low even when allo-HSCT is performed within a short interval from the last dose of GO.