P395: PREDICTING SURVIVAL AFTER HAEMATOPOIETIC STEM CELL TRANSPLANTATION IN ADULT T‐CELL ACUTE LYMPHOBLASTIC LEUKAEMIA: DEVELOPMENT AND VALIDATION OF THE SAVED MODEL

Meng-Yu Xiao,Jian-Ying Zhou,Chen-Hua Yan,Yuan Kong,Xiao-Lu Zhu,Hai-Xia Fu,Yun He,Feng-Rong Wang,Yuan-Yuan Zhang,Xiao-Dong Mo,Wei Han,Jing-Zhi Wang,Yu Wang,Huan Chen,Yu-Hong Chen,Xiang-Yu Zhao,Ying-Jun Chang,Lan-Ping Xu,Kai-Yan Liu,Xiao-Jun Huang,Xiao-Hui Zhang
DOI: https://doi.org/10.1097/01.hs9.0000968488.95751.ba
2023-01-01
HemaSphere
Abstract:Background: T-cell acute lymphoblastic leukaemia (T-ALL) is a malignant blood disease derived from early T-cell progenitor cells. Although intensive chemotherapy has significantly improved the cure rate in patients with T-ALL, the prognosis remains poor in some patients, especially adults with T-ALL (age ≥18 years). Haematopoietic stem cell transplantation (HSCT) is an option for these patients with poor prognoses. However, no prognostic model for adult T-ALL with HSCT is available yet. Aims: The objective of this study was to establish a prediction model for the prognosis of adult T-ALL patients receiving HSCT according to clinical features and laboratory indicators. Methods: A total of 199 adult T-ALL patients under HSCT from January 2007 to December 2021 were retrospectively identified and separated into a derivation cohort (135 patients) and a validation cohort (64 patients) according to the time of transplantation. Their OS was analysed using the Kaplan-Meier method. The prognostic factors were evaluated by univariate Cox regression, and P <0.1 was included in multivariate analysis. P<0.05 was considered statistically significant in multivariate analysis models. A complete statistical analysis was performed using SPSS version 23.0. Results: We developed a prognostic risk score system for adult T-ALL with HSCT based on multivariable Cox regression, which we termed SAVED on the basis of the 5 included predictors. Specifically, we assigned point values to each parameter based on the β coefficient: pre-transplant central nervous system leukaemia scored one point; grade III-IV acute graft-versus-host disease (aGVHD) scored 1 point; no chronic graft-versus-host disease (cGVHD) scored 1 point; pre-transplant haemoglobin<110 g/L scored one point, and 3 months post-transplant CD3< 540 cells/µL scored 2 points. Therefore, a simplified grading scale (SAVED score) was developed on a scale ranging from 0 to 6. Death was estimated to be divided into three groups: low risk (0-1), intermediate risk (2-3) and high risk (4-6). The 3-year mortality rates of each subgroup in the derivation cohort were 9.59%, 31.58% and 87.5%, respectively. The 3-year mortality rates in the validation cohort were 11.54%, 57.14% and 88.24%, respectively. Significant differences in overall survival were observed between the risk groups (P < 0.001 for each cohort). The area of the prognostic model under the subject operating characteristic curve in the derivation cohort was 0.841 (95% CI: 0.760-0.922) and that in the validation cohort was 0.830 (95% CI: 0.729-0.930). The calibration plots show that the predicted results are highly consistent with the observed results. An analysis of the decision curve indicated that adult T-ALL patients could benefit from the clinical application of the prognostic model. Summary/Conclusion: The SAVED score is the first model to predict the prognosis of adult T-ALL patients following HSCT. We anticipate that this approach will facilitate future studies to develop personalized and novel interventions for patients in different SAVED risk groups. Keywords: T cell acute lymphoblastic leukemia, Adult, Hematopoietic cell transplantation, Prognostic
What problem does this paper attempt to address?