Long-term prognosis of gastric cancer in a European country: a population-based study in Florence (Italy). 10-year survival of cases diagnosed in 1985-87.

This paper analysed, in a population-based series of 1976 gastric cancers diagnosed in Florence (Italy), from 1985 to 1987, the relationship between prognostic variables (demographic, clinical and pathological) and 10-year survival rates. Gastric cancer was mostly detected in elderly patients (mean age: 70.5 years) and at advanced stages (i.e. approximately 50% of the patients could not undergo radical surgery). Ten-year observed survival was 12.1% (95% confidence interval (CI): 10.6-13.6%) for the whole series and 20.8% (95% CI: 18.3-23.3%) for resected cases; relative survival was, respectively, 20.9% (95% CI: 18.4-23.4%) and 32.0% (95% CI: 28.1-35.9%). Ten-year relative survival was 86% for stage IA (95% CI: 73-99%) and 67% for stage IB (95% CI: 52-82%). Multivariate analysis showed a significantly better prognosis in females and a significantly worse prognosis in patients aged 65 years or more (reference: ≤59 years). In addition, an independent prognostic effect was observed for pT in the resected cases (reference: pT3; pT1: RR = 0.47, 95% CI: 0.34-0.64; pT2 = 0.71, 95% CI: 0.58-0.87; pT4: RR = 2.02, 95% CI: 1.49-2.75), pN (reference: pN0; pN1: RR = 2.13, 95% CI: 1.70-2.68; pN2-3: RR = 3.14, 95% CI: 2.42-4.07; pN + no. nodes involved unspecified: RR = 4.26, 95% CI: 3.11-5.83) and surgical margin involvement (reference: not involved; involved: RR = 1.36, 95% CI: 1.08-1.72). In addition, the stage, after adjustment for age, gender and surgical margin involvement, showed a strong independent prognostic value (reference: stage II; IA: RR = 0.37, 95% CI: 0.25-0.57; IB: RR = 0.70, 95% CI: 0.50-0.98; IIIA: RR = 1.80, 95% CI: 1.40-2.33; IIIB: RR = 2.82, 95% CI: 2.14-3.72; IV: RR = 3.29, 95% CI: 2.36-4.59). In conclusion, on the basis of a large population-based series, our results confirm the prognostic effect on long-term gastric cancer survival of pathological and demographic variables. In addition, the study shows that Italy had a relatively good, long-term survival when diagnosis was performed at early stages. However, only a few cases were diagnosed at stages when cure by radical surgery is more likely (i.e. stage I accounted for approximately 20% of the resected cases and less than 10% of all incident cases).