A Survival Prognostic Model for Gastro-Intestinal Non-Hodgkin Lymphoma

• Published in: Blood Vol. 144; no. Supplement 1; p. 3087
• Main Authors: Wu, Shaojie, Du, Junru, Ma, Jie, Huang, Li, Peng, Zhiqiang, Ling, Yiwen, Deng, Xiaofang, Zhu, Weijian, Li, Hongliang, Wang, Hua, Li, Yuhua
• Format: Journal Article
• Language: English
• Published: Elsevier Inc 05.11.2024
• ISSN: 0006-4971; 1528-0020
• DOI: 10.1182/blood-2024-200667

Primary gastro-intestinal lymphoma is the most common extra-nodal lymphoma. There is no widely-accepted survival prognostic model. We interrogated data from 1, 023 consecutive newly-diagnosed subjects. The most common site was stomach (n = 560, 54.74%) followed by the large intestine (excluding the ileocecal region) (n = 177, 17.30%), small intestine (n = 134, 13.10%), ileocecal region (n = 85, 8.31%). Most were of B-cell origin (n = 973, 95.1%). Diffuse large B-cell lymphoma (DLBCL) was the most common diagnosis (n = 668, 65.30%). Subjects with B-cell non-Hodgkin lymphoma had better survival compared with those with T-cell non-Hodgkin lymphoma (3.99 [2.55-6.25], P<0.001). According to the principle of non-randomized, non-intervention analysis, Survival of subjects with stomach or large intestine sites was better compared with other single (HR 1.73,95%CI,1.25-2.41,P<0.001). Subjects receiving chemotherapy with surgery had better survival compared with surgery only(HR 2.22,95%CI,1.09-4.52,P=0.023). Receiving R-CHOP (rituximab, cyclophosphamide, doxorubicin, vincristine, and prednisone) was associated with better survival compared with CHOP. A total of 668 patients with PGI-DLBCL were divided into the training(465 patients from 6 centers in Guangdong) and validation (203 patients from the other 3 non-Guangdong centers) sets. In multi-variable Cox regression analyses: age (HR 2.60,95%CI,1.50-4.51, P<0.001), albumin concentration (HR 0.94,95%CI,0.89-0.99, P=0.047), Lugano stage (HR 1.35,95%CI,1.03-1.77, P=0.030), and lactic dehydrogenase (LDH) (HR 2.36,95%CI,1.30-4.30, P=0.005) were independently correlated with survival. These co-variates were used to develop a survival prognosis nomogram model with C-statistics of 0.82 (95%CI,0.77-0.86) and 0.80 (95%CI,0.74 - 0.85) in the training and validation cohortst. Our model, if validated, may be useful to predict survival of patients with primary gastro-intestinal lymphomas. No relevant conflicts of interest to declare.