Abstract:Background: Allogeneic hematopoietic stem cell transplantation (allo-HSCT) is the most important curative therapy for acute myeloid leukemia (AML), and haploidentical related donor (HID) has become the most important alternative donor with strong graft-versus-leukemia effect. However, relapse is still inevitable and is one of the most critical cause of transplant failure. Comprehensive prognostic models have been established to predict relapse in HID HSCT recipients. However, these studies included hematologic malignancies other than AML, and the primary endpoint was survival. Thus far, no comprehensive model could be used to predict the relapse of AML patients receiving HID HSCT in complete remission (CR). Aims: In the present study, we aimed to establish a comprehensive model with machine learning algorithm which could predict the risk of relapse in AML patients receiving HID HSCT in CR. Methods: A total of 951 patients were enrolled, then we randomly selected 70% of the entire population (n = 665) as the training cohort for developing machine learning model and the nomogram, and the remaining 30% (n = 286) as validation cohort. We used multivariate logistic regression analysis for feature selection. We included the variables whose coefficients with P values less than 0.1 as the input for the machine learning model. The machine learning model was evaluated among validation cohort. The accuracy, area under the curve (AUC) score, sensitivity, and specificity were computed in both data cohorts for model evaluation. In addition, the calibration curve and decision curve were plotted for identifying the usability of nomogram. Results: A total of five variables were included in the final model, and our equation was as follows: Probability (relapse) = $\ \frac{1}{1\+\exp(-Y)}$. Y = 0.5677 × (AML risk category at diagnosis) + 0.0690 × (induction courses for first CR) + 0.4583 × (disease status before HSCT) + 0.4061 × (multiparameter flow cytometry [MFC] before HSCT) – 0.1623 × (blood group disparity) – 2.9641. The threshold of probability was set as 0.1106, and the g-mean was 0.668. A nomogram was designed with a C-index of 0.707 (95% CI 0.645–0.770). The calibration plots revealed satisfactory agreement between nomogram prediction and actual observation for the probability of relapse at day. The optimal cutoff value of the total nomogram scores was determined to be 53 (Figure 1). The patients were separated into low- and high-risk groups. The 3-year cumulative incidence of relapse after HID HSCT was 19.0% (95% CI, 14.8%–23.2%) versus 8.2% (95% CI, 5.8%–10.6%) (Figure 2A, P < 0.001), 18.9% (95% CI, 13.8%–24.0%) versus 9.2% (95% CI, 6.1%–12.2%) (Figure 2B, P < 0.001), and 19.5% (95% CI, 11.9%–27.1%) versus 5.9% (95% CI, 2.1%–9.7%) (Figure 2C, P < 0.001), respectively, in the high- and low-risk groups in the total cohort, training cohort, and validation cohort. The 3-year probabilities of measurable residual disease occurrence, event-free survival, leukemia-free survival and overall survival in low-risk group were also significantly better than those in high-risk group. This model cod also predict the efficacy of preemptive immunotherapy after HID HSCT.Summary/Conclusion: In summary, we firstly established a comprehensive model to predict the relapse of AML patients receiving HID HSCT in CR, which can be popularized easily, can help to provide risk-stratification directed relapse prophylaxis, and may further decrease the risk of relapse. In future, prospective studies with independent cohort can further confirm the efficacy of our predicted model. Keywords: Haploidentical stem cell transplantation, AML, Relapse

A nomogram to predict the risk of relapse in patients with minimal change disease: a retrospective cohort study

LASSO-derived Nomogram Predicting New-Onset Diabetes Mellitus in Patients with Kidney Disease Receiving Immunosuppressive Drugs.

A preliminary nomogram model for predicting relapse of patients with primary membranous nephropathy

A Nomogram for Predicting T315i-Free Survival in Chronic Phase Chronic Myeloid Leukemia Patients: A Multicenter Retrospective Study

Development and Validation of a Risk Score in Chinese Patients with Chronic Heart Failure

Machine learning-based diagnostic prediction of minimal change disease: model development study

Development and External Validation of a Nomogram for Predicting Renal Function Based on Preoperative Data from In-Hospital Patients with Simple Renal Cysts.

Relationship between time to clinical remission and relapse in adults with steroid-sensitive minimal change disease: a retrospective cohort study

A Prediction Model for Acute Kidney Injury in Adult Patients With Minimal Change Disease

Development and Evaluation of a Nomogram for Predicting the Outcome of Immune Reconstitution among HIV/AIDS Patients Receiving Antiretroviral Therapy in China.

Validation of the Antineutrophil Cytoplasmic Antibody Renal Risk Score and Modification of the Score in a Chinese Cohort With a Majority of Myeloperoxidase-Positive Patients.

Validation of the Antineutrophil Cytoplasmic Antibody Renal sk Score and Modification of the Score in a Chinese Cohort Ri With a Majority of Myeloperoxidase-Positive Patients

A dynamic online nomogram for predicting renal outcomes of idiopathic membranous nephropathy

A Multifactorial Risk Scoring System for the Prediction of Early Relapse in CMML Patients with Allo-Hsct: a Nationwide Representative Multicenter Study

P573: novel prognostic model for relapse in patients with acute myeloid leukemia receiving haploidentical hematopoietic stem cell transplantation

A machine learning-based nomogram model for predicting the recurrence of cystitis glandularis

Machine Learning For Prediction And Risk Stratification Of Lupus Nephritis Renal Flare

Nomogram to predict risk of incident chronic kidney disease in high-risk population of cardiovascular disease in China: community-based cohort study

A nomogram prediction model for treatment failure in primary membranous nephropathy

Development and validation of a diagnostic nomogram model for predicting monoclonal gammopathy of renal significance

Nomogram for predicting risk of mild renal dysfunction among general residents from rural Northeast China