Prediction Model Establishment of Prognosis Factors for Acute Myeloid Leukemia Based on the SEER Database
Authors
Affiliations
Acute myeloid leukemia (AML) with t (9;11) (p22; q23) presents as a varied hematological malignancy. The t (9;11) (p22; q23) translocation is the most common among 11q23/KMT2A rearrangements in AML. This research aimed to develop a nomogram for precise prediction of overall survival (OS) and cancer-specific survival (CSS) in AML with the t (9;11) (p22; q23) translocation. We utilized the Surveillance, Epidemiology, and End Results (SEER) database to identify patients diagnosed with t (9;11) (p22; q23) AML from 2000 to 2021. Prognostic factors for this AML subtype were determined using least absolute shrinkage and selection operator (LASSO) regression, which guided the creation of prognostic nomograms. To evaluate the model's discrimination, accuracy, and effectiveness, we employed the concordance index (C-index), calibration charts, receiver operating characteristic curves (ROC), area under the curve (AUC), and decision-curve analysis (DCA). The research was meticulously planned, executed, and documented in full adherence to the TRIPOD guidelines. The nomogram was developed using key variables including age, race, first primary tumor, and chemotherapy. The concordance indices (C-indices) were 0.704 for OS and for 0.686 for CSS. Patients were classified into high-risk and low-risk groups based on nomogram scores, with significant differences in OS and CSS between these groups (P < 0.001). This study developed innovative nomograms that combine clinical and treatment factors to predict 1-, 3-, and 5-year survival rates for patients with t (9;11) (p22; q23) AML.