Prospective randomised study of split-course radiotherapy versus cisplatin plus split-course radiotherapy in inoperable squamous cell carcinoma of the oesophagus.

Between 1983 and 1989, 211 patients with inoperable squamous cell carcinoma of the oesophagus were randomised in a study comparing split-course irradiation (two courses of 20 Gy in five fractions of 4 Gy, separated by a rest of 2 weeks) (arm A) and the same split-course irradiation in combination with cisplatin (CDDP) (3-4 days before each of the two courses of radiotherapy, repeated every 3-4 weeks, for a total of six cycles) (arm B). The Cox's regression model with retrospective stratification was used to compare the two arms to correct for the imbalance at randomisation of the T classification. The median overall survival was 7.9 (95% confidence interval (CI) 7.3-9.4) months in arm A and 9.6 (95% CI 8-13.5) months in arm B. The difference in overall survival was only borderline significant (P=0.048) with a reduction of the instantaneous rate of death of 24%. The 1 and 2 year overall survival rate were respectively 29% (95% CI 21-37%) and 15% (95% CI 8-22%) in arm A and 45% (95% CI 36-54%) and 20% (95% CI 13-27%) in arm B; thereafter, the survival curves became similar. The median progression free survival (PFS) was 5.0 (95% CI 4.6-5.7) versus 6.9 (95% CI 5.3-8.7) months (P=0.028) and the median time to local progression was 6.2 (95% CI 5.1-7.6) months versus 10.9 (95% CI 8.1-15.5) months (P=0.018), respectively, in arms A and B. Haematological toxicities were slightly more commonly observed in the combined group (1% versus 6%). This study shows that split-course irradiation in combination with CDDP is very well tolerated and should be preferred to radiotherapy alone.

RCT Entities:   Population Interventions Outcomes