Survival and Risk Prediction Model for Acute Leukemia Patientsundergoing Hematopoietic Stem Cell Transplantation

Background

Allogeneic hematopoietic stem cell transplantation (allo-HSCT) is the only potentially curative treatment for acute leukemia. However, disease cure is not guaranteed post-transplantation, as numerous complications post-transplant can affect patient survival outcomes and warrant attention. Research indicates that post-HSCT, patients are prone to infections, pneumonia, liver and kidney dysfunctions, and neurological injuries. These complications not only directly impact the efficacy of the transplantation but also significantly decrease the patient's quality of life and may even threaten their life. This study aims to analyze the factors influencing survival in patients with hematological diseases undergoing HSCT and to develop a risk prediction model to provide reference for clinical treatment.

Objective

To analyze the risk factors affecting the survival of acute leukemia patients undergoing hematopoietic stem cell transplantation (HSCT) and to construct a survival prediction risk model to provide reference for clinical diagnosis and treatment.

Methods

From January 2018 to December 2020, acute leukemia patients who underwent HSCT at the Hematology Center of a tertiary hospital in Chongqing were selected. Multivariate regression analysis was used to identify factors influencing survival post-HSCT. A prediction model was constructed using R software, and the accuracy of the model was evaluated using the area under the ROC curve (AUC) and calibration curves.

Results

A total of 650 patients with malignant leukemia were included, with a follow-up period of 3 years. Among them, 115 patients died, resulting in a mortality rate of 18.54%. Multivariate Cox regression analysis showed that patient age (OR=1.02, P=0.035), diagnosis of ALL (OR=1.64, P=0.021), Risk (OR=1.8, P=0.01), HGB (OR=0.98, P=0.002), PLT (OR=0.97, P=0.006), GVHD (OR=0.35, P<0.001), GIB (OR=2.65, P=0.002), donor age (OR=0.98, P=0.025), PA (OR=0.99, P<0.001), NLR (OR=0.97, P<0.001), and SII/per100 (OR=0.98, P=0.002) were independent risk factors affecting survival post-HSCT in acute leukemia patients. By combining stepwise regression and clinical significance, the final model included Risk, Cells.CD34+, Donor.age, GVHD, Age, Diagnosis, HGB, Tumor.history, NLR, SII, and PA. The results showed that the AUC of the OS curve was 0.799 (95% CI 0.759 to 0.838) for 1-year post-transplant, 0.831 (95% CI 0.796 to 0.866) for 2-year post-transplant, and 0.778 (95% CI 0.739 to 0.817) for 3-year post-transplant. The consistency index (C-index) was 0.701, with a corrected C-index of 0.672 validated by bootstrap sampling. The K-fold cross-validation yielded AUC=0.8243 and Brier=0.1567, indicating good consistency between our prediction model and actual survival outcomes post-transplant.

Conclusion

Disease type, risk stratification, patient age, pre-transplant serum albumin, CD34+ cell count, donor age, patient age, GVHD, tumor history, and PA are high-risk factors affecting the survival of acute leukemia patients post-HSCT. Clinicians should focus on these high-risk patients and provide appropriate interventions to improve post-HSCT survival outcomes.

No relevant conflicts of interest to declare.