AUTHOR=Yin Xuejiao , Xu Aoshuang , Fan Fengjuan , Huang Zhenli , Cheng Qianwen , Zhang Lu , Sun Chunyan , Hu Yu
TITLE=Incidence and Mortality Trends and Risk Prediction Nomogram for Extranodal Diffuse Large B-Cell Lymphoma: An Analysis of the Surveillance, Epidemiology, and End Results Database
JOURNAL=Frontiers in Oncology
VOLUME=9
YEAR=2019
URL=https://www.frontiersin.org/journals/oncology/articles/10.3389/fonc.2019.01198
DOI=10.3389/fonc.2019.01198
ISSN=2234-943X
ABSTRACT=Background: DLBCL is the most commonly occurring type of non-Hodgkin's lymphoma, which may be found at various extranodal sites. But little is known about the particular trends of extranodal DLBCL.
Methods: A total of 15,882 extranodal DLBCL patients were included in incidence analysis from the Surveillance, Epidemiology, and End Results (SEER) database (1973–2015). The joinpoint regression software was used to calculate the annual percent change (APC) in rates. Nomograms were established by R software to predict overall survival (OS).
Results: The extranodal DLBCL incidence continued to rise at a rate of 1.6% (95% CI, 0.4–2.8, p < 0.001) per year over the study period, until it declined around 2003. The incidence-based mortality trend of extranodal DLBCL had a similar pattern, with a decrease happening around 1993. Five-year survival rates improved dramatically from the 1970s to 2010s (44.15 vs. 63.7%), and the most obvious increase occurred in DLBCL patients with primary site in the head/neck. The C-index showed a value for OS of 0.708, which validated the nomograms performed well and were able to forecast the prognosis of patients with extranodal DLBCL. The calibration curves showed satisfactory consistency between true values and predicted values for 1-, 5-, and 10-year overall survival, respectively.
Conclusions: The incidence and incidence-based mortality of extranodal DLBCL had been increasing for decades, followed by a promising downward trend in recent years. These findings may help scientists identify disease-related risk factors and better manage the disease. The prediction signature cloud identifies high-risk patients who should receive effective therapies to prevent the fatal nature of this disease, and low-risk patients to reduce over-treatment.