A proposed model for prediction of survival based on a follow-up study in unresectable pancreatic cancer.
2013 (English)In: BMJ Open, ISSN 2044-6055, Vol. 3, no 12, 1-6 p.Article in journal (Refereed) Published
OBJECTIVES: To define an easy-to-use model for prediction of survival time in patients with unresectable pancreatic cancer in order to optimise patient' care. DESIGN: An observational retrospective study on patients with unresectable pancreatic cancer. The initial radiographs at presentation of symptoms were reviewed and the maximum diameter of the primary tumour was determined. The occurrence of liver metastases and performance status that determines initiation of chemotherapy was also used in the regression analysis to identify prognostic subgroups. SETTING: County hospital in south-east of Sweden. POPULATION: Consecutive patients with unresectable pancreatic cancer who were diagnosed between January 2003 and May 2010 (n=132). MAIN OUTCOME MEASURES: Statistical analyses were performed using Stata V.13. Survival time was assessed with Kaplan-Meier analysis, log-rank test for equality of survivor functions and Cox regression for calculation of individual hazard based on tumour diameter, presence of liver metastases and initiation of chemotherapy treatment according to patient performance status. RESULTS: The individual hazard was log h=0.357 tumour size+1.181 liver metastases-0.989 performance status/chemotherapy. Three prognostic groups could be defined: a low-risk group with a median survival time of 6.7 (IQR 9.7) months, a medium-risk group with a median survival time of 4.5 (IQR 4.5) months and a high-risk group with a median survival time of 1.2 (IQR 1.7) months. CONCLUSIONS: The maximum diameter of the primary tumour and the presence of liver metastases found at the X-ray examination of patients with pancreatic cancer, in conjunction with whether or not chemotherapy is initiated according to performance status, predict the survival time for patients who do not undergo surgical resection. The findings result in an easy-to-use model for predicting the survival time.
Place, publisher, year, edition, pages
BMJ Publishing Group , 2013. Vol. 3, no 12, 1-6 p.
Medical and Health Sciences
IdentifiersURN: urn:nbn:se:bth-6684DOI: 10.1136/bmjopen-2013-004064.ISI: 000330541400009Local ID: oai:bth.se:forskinfo65AB6E608DA72E3FC1257C5B00379D25OAI: oai:DiVA.org:bth-6684DiVA: diva2:834208
Open access article2014-07-172014-01-092015-06-30Bibliographically approved