Related Experiment Video
Updated: Aug 19, 2026

Competing-Risk Nomogram for Predicting Cancer-Specific Survival in Multiple Primary Colorectal Cancer Patients after Surgery
Published on: September 27, 2024
[Prognosis of lymphosarcomas. Multivariate analysis in a series of 334 patients]
Abstract:
In this retrospective study, we analysed prognostic factors of 334 previously untreated patients with non Hodgkin's malignant lymphoma. The importance of the different prognostic variables was first studied with survival curves. Next, using a "segmentation" method and a logistic regression analysis, we studied the factors influencing the complete remission rate which is the most important prognostic factor. For the "low-grade" lymphosarcomas, with the segmentation method, we obtained five patients groups for which the complete remission rates varied between 7 and 95.5 per cent according to clinical staging, age and mediastinal involvement. For the "high-grade" lymphosarcomas, we obtained four patients groups in which the complete remission rates varied between 8.7 and 84.7 per cent according to mediastinal and sub-diaphragmatic involvement and clinical staging. Lastly, the Cox model was used in order to determine the best variable combination capable of predicting the survival of patients with clinical stages I or II. These variables were pathological grading, complete initial remission, clinical staging and BCG immunotherapy. With scores deduced from the Cox model, we were able to determine three prognostic groups for which the survivals at 3 years were very significantly different (7 to 89%).
