nomogram

nomogram predicts long-term cancer-specific survival for patients with primary gastric diffuse large B-cell lymphoma (GDLBCL) by integrating clinical variables identified from the SEER database (1998–2014) using Lasso Cox regression.

Key Features:

• Dataset: Developed from 2,647 GDLBCL patients extracted from the SEER database spanning 1998–2014.
• Variable selection: Independent risk factors were identified using the Lasso Cox regression model.
• Predictors included: Surgery, chemotherapy, tumor stage, and age were incorporated as model covariates.
• Model outputs: Produces individualized cancer-specific survival rate estimates over time, including 5-year conditional survival metrics.
• Validation: Calibration plots were used to assess agreement between predicted and observed outcomes.
• Follow-up and risk patterns: Follow-up times reported as median 30 months and mean 52.8 months, with analyses showing decreasing overall cancer-specific survival over time and higher mortality in the first two years followed by stabilization.

Scientific Applications:

• Personalized prognosis: Provides individualized cancer-specific survival estimates for patients with primary gastric DLBCL.
• Risk stratification and surveillance planning: Identifies elevated cancer-specific mortality within the first two years post-diagnosis to inform monitoring intensity.
• Clinical research: Enables analysis of the prognostic impact of surgery, chemotherapy, tumor stage, and age on survival in GDLBCL cohorts.

Methodology:

Model development used Lasso Cox regression on SEER (1998–2014) data; follow-up times were summarized (median 30 months, mean 52.8 months); survival trend and conditional survival analyses were performed and calibration plots assessed model agreement.