P1539: infection-related mortality clinical-biological scoring system (irm-score) after allogeneic hematopoietic stem cell transplantation: a validation study in gitmo centres

Topic: 30. Infections in hematology (incl. supportive care/therapy) Background: Infection-related mortality (IRM) remains a major issue in the context of allogeneic hematopoietic stem cell transplantation (allo-HSCT). Recent improvement in supportive care have resulted in a considerable reduction of the treatment-related mortality, even though more effective prophylactic strategies and interventions are warranted. In a previous monocentric study a prognostic clinic-biological scoring system, based on four pre-transplantation factors (patient age, IgA and IgM levels, patient and donor CMV serostatus), was developed in order to predict IRM after allo-HSCT [Forcina et al, Biol Blood Marrow Transplant 23 (2017) 2151–2158]. This score identified 3 groups of patients with significant difference in term of IRM (low, intermediate and high-risk group). Aims: We aimed to promote IRM-score prognostic validation in a multicenter setting in centres of the Gruppo Italiano Trapianto Midollo Osseo (GITMO). Methods: Adult patients (18 -78 years), undergoing first allo-HSCT for both malignant and non-malignant hematologic diseases between 1st January 2012 and 31st May 2015, have been enrolled. A cohort of 1241 patients had completed data for being analyzed. Patients were clustered according to the 3-tiered prognostic model developed by Forcina et al. into 3 risk classes: low-risk <10.17 points; intermediate-risk 10.17 – 11.11 points, and high-risk >11.11 points. Primary endpoint was validation of this IRM scoring system in a multicenter setting. Results: A total of 236 patients, belonging from the monocentric original-cohort used for the derivation of the score, have been included with longer updated follow-up. This original-cohort has been compared with a new multicenter-cohort of 1005 patients, employed for externally validating the prognostic score. Overall, median follow up was 46.45 months. The original-cohort showed the following main differences with respect to the multicenter-cohort: older age (p=0.0003), higher comorbidity score (p=0.003), higher incidence of pre-transplant MDR colonization (p=0.0002) and less patients with CMV negative/negative serostatus (p=0.0155). Overall, IRM score was: low-risk in 550 (44.32%), intermediate-risk in 362 (29.17%), and high-risk in 329 (26.51%) patients. Notably, risk groups were differently distributed between cohorts, with a higher proportion of low-risk patients and a lower proportion of intermediate-risk patients in multicenter versus original-cohort (p=0.0389). Globally, the risk groups showed a significantly different IRM in both cohorts (p<0.0001 for original-cohort and p=0.0068 for multicenter-cohort) (Fig.1). In the original-cohort, IRM at 100-days was: 5% in low-risk, 9% in intermediate-risk, and 28% in high-risk group. In multicenter-cohort, IRM at 100-days was: 5% in low-risk, 8% in intermediate-risk, and 7% in high-risk group. In the original-cohort, IRM at 2-years was: 10% in low-risk, 23% in intermediate-risk, and 40% in high-risk group. In multicenter-cohort, IRM at 2-years was: 9% in low-risk, 12% in intermediate-risk, and 17% in high-risk group. Summary/Conclusion: Both the original-cohort with extended follow-up and the new multicenter cohort confirmed that the IRM score is reliable, cost-effective and applicable accross different transplant settings. IRM score may provide a useful clinical tool for the early identification of HSCT recipients at highest risk of infection mortality and application of tailored prophylactic/pre-emptive strategies. Fig.1Keywords: Bone marrow transplant, Infection