Validation of the Transplant Conditioning Intensity (TCI) Score for Allogeneic Hematopoietic Cell Transplantation in a Spanish Multicenter Cohort on Behalf of Geth

Background: The introduction of new conditioning intensity regimens for patients undergoing allogeneic hematopoietic stem cell transplant (HCT) does not often align with the traditional reducing/myeloablative conditioning intensity (RIC/MAC) classifications. Therefore, the Transplant Conditioning Intensity (TCI) score has been recently established to stratify the intensity of conditioning regimen more effectively than the existing nomenclature. This score sums the doses (weight scores) of each component of the pre-HCT regimen. TCI has been tested and validated in two large retrospective cohorts (n=8255; n= 4060, including PTCY in 10%) and it was predictive of early non-relapse mortality (NRM) and 2-years NRM and relapse incidence (RI). Methods: We reported the impact of TCI in predicting 100-day NRM, 1-year and 2-year NRM and RI. TCI was applied to 757 patients with acute myeloid leukemia (AML) who underwent HCT between 2011 and 2022 and were reported to the Grupo Español de Trasplante Hematopoyetico. All patients received peripheral blood stem cells. Results: According to TCI score, 44% of the patients were assigned to the low (TCI 1-2), 43% to the intermediate (TCI 2.5-3.5) and 13% to the high (TCI 4-6) category. The majority of patients had de novo AML (83%). Median age was 56 (range 18-82) years. Median follow up was 55 (range 2-140) months. Disease status at HCT was complete remission in 79% and 53% of the patients were male. Cytogenetic was adverse for 27% of the patients, while 56% had an intermediate and 12% a favorable risk. Conditioning intensity regimen was classified as MAC in 51%. All patients included in the low TCI (n=335) underwent a RIC and all those in the high TCI group had a MAC regimen (n=98). The majority of patients categorized as intermediate TCI received a MAC regimen (n=286, 88%). Patients received 14 different types of chemotherapy conditioning, the most frequent being fludarabine + 2 days busulfan in 35% of them. HCT was performed from a matched sibling (MSD, 36%) or unrelated donor (UD, 32%), haploidentical donor (24%) and a mismatched UD (8%). HCT comorbidity index was 1 to 2 in 33% of the cases and 36% received PTCY. We observed a higher frequency of MSD in the low group (47%) and of HAPLO in the intermediate group (49%) p<0.0001. Patients in the low group had a higher age and received PTCY p<0.0001. Patients were balanced within TCI grouping according other characteristics. For the entire cohort RI was 9.6% [95%CI 7.7-11.9], 23.2% [95%CI 20.3-26.4] and 28.1% [95%CI 25-31.6] at 100 days, 1 year and 2 years, respectively. Overall, 100 days, 1-y and 2-y NRM was 4.7% [95%CI 3.4-6.4], 12.6% [95%CI 10.4-15.2], 14.3% [95%CI 12-17.1], respectively. In univariate analysis, we didn't observe a statically significant difference between 100-day, 1-y and 2-y NRM according to TCI groups. NRM at 100 days was 4.2% [95%CI 2.5-7]; 5.3% [95%CI 3.3-8.4]; 4.1% [95%CI 1.6-10.8] for low, intermediate and high TCI group, respectively (p=0.78). The 1-y and 2-y NRM were 12% [95%CI 9-16.1]; 12.6 [95%CI 9.4-16.8]; 14.6 [95%CI 8.9-23.7] (p=0.82) and 15.1% [95%CI 11.7-19.6]; 13.2 [95%CI 10-17.6]; 16.9 [95%CI 10.7-26.5] (p=0.71) according to the 3 TCI categories. RI was 12% [95%CI 9-16], 7.2% [95%CI 4.8-10.7] and 9.3% [95%CI 5-17.3] according to TCI at 100 days (p=0.11). RI was 26.8% [95%CI 22.4-32], 18.2% [95%CI 14.4-23] and 27% [95%CI 19.4-37.7] according to low, intermediate and high TCI at 1 and 2 years (p=0.02). In multivariate analysis, we observed an increase in risk of NRM with each TCI group from low to high reaching significance only when comparing the high TCI score with low TCI (reference). Hazard ratios [HR] for intermediate and high TCI were 1.26 (95% CI,0.81-1.98, p =0.31) and 2.23 (95% CI, 1.17-4.25, p =0.015), respectively. HRs for RI were 0.65 (95% CI, 0.46-0.91, p =0.01) and 1.02 (95% CI, 0.61-1.69, p =0.94) for the intermediate and high group, respectively. Patients >65 years old had an increased NRM (HR 2.32, 95% CI, 1.23-4.38, p =0.009). Adverse risk was associated with higher NRM (HR 3.0, 95% CI, 1.32-6.79, p =0.008) an RI (HR 2.57, 95% CI, 1.61-4.09, p <0.0001) at 2 years. Conclusion: TCI effectively performed stratification in a real-life cohort of patients predicting NRM and RI at 2-year, but did not accurately predict early mortality. Notably, 30% of patients received PTCY highlighting the need for further refinement of TCI to account for PTCY, a steadily used graft-versus-host disease prophylaxis.