Abstract
BACKGROUND: With therapeutic advancements improving survival outcomes in myelodysplastic syndromes (MDS), an increasing proportion of patients are at risk of developing second primary malignancies (SPMs). Nevertheless, the clinical profile, prognostic determinants, and overall survival (OS) of SPMs in this population remain inadequately characterized. This study sought to delineate the clinical features of SPMs in MDS patients and construct a predictive nomogram for overall survival (OS) in MDS-SPM (MSP) patients. METHODS: Data were obtained from the Surveillance, Epidemiology, and End Results (SEER) database for the period 2000-2021. Cox regression and competing risk analyses were performed to identify prognostic factors. A nomogram was then constructed and validated to predict the OS of patients with SPMs after MDS. RESULTS: A total of 2,967 eligible patients were included. The three most prevalent SPM sites were the respiratory system (18.03%), the urinary system (9.67%) and the colonand rectum (9.34%). The most common histological types of SPMs were adenomas and adenocarcinomas (38.93%), squamous cell neoplasms (9.91%) and ductal and lobular neoplasms (7.75%). Univariate and multivariate Cox regression analyses revealed that International Classification of Diseases for Oncology, Third Edition (ICD-O-3) histology, chemotherapy, age, the time interval between MDS and SPMs, SPM type, and marital status were associated with OS. Based on these factors, a nomogram was developed to predict OS. The time-dependent area under the curve (AUC) performed well in assessing the 5-year (0.69 and 0.7) and 7-year (0.73 and 0.74) OS of SPMs in the training and validation cohorts, demonstrating the nomogram's satisfactory discriminative ability. CONCLUSIONS: A prognostic nomogram was developed and validated to help clinicians assess the prognosis of MSP patients. The establishment of prognostic models can facilitate more comprehensive and precise treatment planning.