No evidence for changing the radiation (RT) prescription in trial-defined chemoradiation for limited small cell lung cancer (LSCLC) to “conventional-fraction” RT: 10-year review of outcomes

Abstract

Purpose: To review the impact on LSCLC outcomes of changing from hypofractionated RT with posterior cord shielding starting with chemotherapy cycle 2 (“best” arm, randomized NCI-Canada regimen) to a plan of standard fraction RT/off-cord fields, where the change occurred due to theoretical concerns about the trial’s technique (i.e., shielding increases local failure) and fraction size (i.e., larger fractions=higher toxicity rates). Materials and Methods: From 1989–1999, 215 LSCLC patients (pts) received 6 cylces of alternating cyclophosphamide/anthracycline/vincristine and etoposide/cisplatin (EP). Thoracic RT (TRT) started with EP (at cycle 2 or 3) and was either: [1] 40 Gy/15 fractions/3 weeks/APPA fields/cord shield at 35 Gy or [2] 50 Gy/25 fractions/5 weeks/multiple fields/no cord shield. RT fields encompassed gross and suspected microscopic disease with 2 cm margins. Propphylactic cranial irradiation (PCI) was variably given. Analysis looked at the impact of each RT schedule on survival and recurrence patterns. Results: Overall survival for all pts at 2 and 5 years was 22.7% and 7.2%, respectively, with median survival at 14.7 months. 122 pts (57%) received 40 Gy; 92 (43%), 50 Gy. PCI was given to 21 (44%) 40 Gy and 47 (56%) 50 Gy pts. Disease-specific survivals (PCI pts only) per TRT dose, given in the table, showed no statistical differences. Failures were recorded in 132 pts (61%). Comparing 40 Gy vs 50 Gy: rate of any first relapse was 40% vs 42%; chest failures only were 34% vs 45% (PCI pts only). Brain failure rate was related to PCI use and not TRT dose. No difference in treatment-related toxicity rates per TRT dose was seen (p=0.35). Pts in each dose group showed no differences over a range of prognostic variables tested. Conclusion: Changing from a randomized trial-defined TRT prescription to RT “correcting” for theoretical concerns over the trial’s fractionation and technique was not valid since survival, thoracic control and toxicity rates were not improved. PCI had the most significant impact on brain relapse patterns. Tabled 1SURVIVAL40 Gy with PCI50 Gy with PCIMedian (months)20.920.02- year (%)41.344.45-year (%)20.722.2 Open table in a new tab