Purpose: To perform a retrospective review of a single institution's 10-year experience in treating limited-stage small-cell lung cancer (LS-SCLC) with a concurrent chemoradiation regimen modeled after the experimental arm of a randomized National Cancer Institute of Canada trial in which hypofractionated radiotherapy started with cycle 2 of chemotherapy. We then looked at the impact on patient outcomes of changing the RT during the course of the decade to a "conventionally" (2 Gy) fractionated regimen, with a focus on toxicity and survival rates.
Methods and materials: Between 1989 and 1999, 215 LS-SCLC patients received six cycles of chemotherapy consisting of cyclophosphamide, doxorubicin, and vincristine alternating with etoposide and cisplatin every 3 weeks. Thoracic RT was administered concurrently with etoposide and cisplatin (at cycle 2 or 3) only and consisted of either 40 Gy in 15 fractions for 3 weeks or 50 Gy in 25 fractions for 5 weeks. RT fields encompassed gross and suspected microscopic disease with 2-cm margins. Prophylactic cranial irradiation (PCI) was offered to complete responders according to clinician preference. RT interruption during concurrent chemoradiation was used as the "marker" for treatment toxicity. The analysis compared the RT schedules for differences in toxicity, survival, and recurrence patterns.
Results: The overall survival rate for 215 patients at 2 and 5 years was 22.7% and 7.2%, respectively, with a median survival of 14.7 months. Thoracic RT consisted of 40 Gy in 3 weeks for 122 patients (57%) and 50 Gy in 5 weeks for 92 patients (43%). PCI was administered to 21 (44%) and 47 (56%) patients receiving 40 Gy and 50 Gy, respectively. The patient- and treatment-related variables were comparable between the two cohorts treated with the different RT prescriptions. RT interruptions during concurrent chemoradiation were recorded in 56 cases (26%), with a median duration of 5 days (range 1-18). No differences in treatment-related toxicity rates were demonstrated between the two dose cohorts (p = 0.35). The overall and disease-free survival rates (patients stratified by PCI use) at 5 years for the 40- and 50-Gy schedules were 14.3% and 12.0% (p = 0.71) and 20.7% and 22.2% (p = 0.76), respectively. Sites of first failure were recorded in 132 patients (61%). Comparing the 40-Gy and 50-Gy cohorts, the rate of any first relapse was 40% vs. 42% and the chest as the first relapse site was 34% vs. 45% (patients stratified by PCI use), respectively. The brain failure rate reflected PCI use and was not related to the thoracic RT schedule.
Conclusion: Changing from a hypofractionated to a conventionally fractionated RT thoracic prescription did not alter outcomes because the survival, thoracic control, and toxicity rates were statistically similar. This suggests that the hypofractionated schedule remains practicable and should be considered in the setting of randomized clinical trials. In view of the benefits that accelerated schedules provide for both patients and cost containment, clinicians may opt to use this tolerable regimen in managing LS-SCLC. Regarding the future development of novel chemoradiation programs, the most critical factor in ensuring improved outcomes for LS-SCLC may be limiting the duration of RT and overall treatment time.