Clinical outcome of three fractionation schedules of preoperative radiotherapy for rectal cancer

Abstract

Aim

To evaluate the effectiveness and normal tissue reactions of three fractionation schedules of preoperative radiotherapy for locally advanced rectal cancer.

Materials and Methods

Between 1996 and 2002, 168 patients with locally advanced rectal cancer were treated as follows: 53 patients received 25 Gy in 5 Gy per fraction (group A), 45 received 30 Gy in 3.0 Gy per fraction (group B), and 70 were treated with accelerated hyperfractionation: 42 Gy, 1.5 Gy per fraction, given twice a day with an inter-fraction interval of 6 hours (group C). The clinical characteristics of the groups were comparable. The patients did not receive concurrent chemotherapy. A Cox proportional hazard regression model was used to analyze the factors which may influence loco-regional tumour control (LRC) and overall survival (OS).

Results

The following variables significantly influenced LRC: fractionation scheme (5-year LRC 80%, 69%, and 90% in groups A, B, C respectively, p=0.016), haemoglobin concentration before radiotherapy (p= 0.012) and postoperative chemotherapy (p=0.01). Age, sex, stage of disease, location of tumour (distance of the tumour from the anal verge) and performance status did not appear significant for LRC. The overall 5-year OS was 64%, 59% and 74% in groups A, B, C respectively (p=0.056). The OS was significantly influenced by postoperative pathological stage (p=0.006), tumour location (p= 0.015) and postoperative chemotherapy (p=0.047). The most frequent acute radiation reaction was mild/severe diarrhoea, which appeared in 5%, 21.6% and 65.5% of the patients from groups A-C respectively. The median wound healing time in those who underwent abdominoperineal resections was 6, 6 and 4 weeks. Other reactions appeared less relevant. There was no significant difference in the incidence of late effects among the three treatment groups.

Conclusion

While due to the non-randomized character of the study the conclusions should be regarded as hypothesis-generating only, the analysis has shown an acceptable local effectiveness and tolerance of schedules A and C, and disappointing effectiveness of schedule B. The present study thus supports the data which suggest that the clinical effect of preoperative radiotherapy for rectal cancer is influenced not only by total radiation dose but also by overall radiation treatment time and dose per fraction.