Survival after surgery for oesophageal cancer: a population-based study

Lancet Oncol. 2005 Nov;6(11):864-70. doi: 10.1016/S1470-2045(05)70347-8.


Background: Oesophageal resection alone remains the only potentially curative treatment for oesophageal cancer. However, few population-based studies of survival after surgery have been done. We aimed to assess whether survival after surgery for oesophageal cancer has improved since 1987.

Methods: We identified from the nationwide cancer register 764 patients in Sweden who had undergone resection alone for oesophageal cancer between Jan 1, 1987 and Dec 31, 2000, and followed up these individuals until Oct 18, 2004, through linkage to the nationwide registers of death, emigration, and total population. Tumour stage, location, and histology were recorded by review of histopathological reports. Hazard ratios, with 95% CI, for death adjusted for age; sex; comorbidity; and tumour stage, location, and histology were calculated by Cox proportional-hazards regression model.

Findings: Overall, the proportion of deaths up to 30 days after surgery decreased from 10.1% in 1987-91 to 4.9% in 1997-2000. Increased long-term survival was associated with low tumour stage. Survival was 46.5% 1 year after surgery, 24.1% 3 years after surgery, and 19.7% 5 years after surgery in 1987-91; 58.4%, 31.7%, and 24.9%, respectively, in 1992-96; and 61.7%, 39.9%, and 30.7%, respectively, in 1997-2000. The adjusted hazard ratio for death was 0.76 (95% CI 0.62-0.92) in 1992-96 and 0.57 (0.45-0.71) in 1997-2000.

Interpretation: Survival after surgery for oesophageal cancer has improved substantially since 1987. Because this improvement cannot be explained by a change in patient characteristics or tumour characteristics, it is probably a result of better surgery.

Publication types

  • Research Support, Non-U.S. Gov't

MeSH terms

  • Aged
  • Aged, 80 and over
  • Comorbidity
  • Esophageal Neoplasms / mortality
  • Esophageal Neoplasms / pathology
  • Esophageal Neoplasms / surgery*
  • Female
  • Humans
  • Male
  • Middle Aged
  • Population Surveillance / methods*
  • Proportional Hazards Models
  • Registries
  • Survival Analysis
  • Sweden