Purpose: To assess the relationship between machine energy (4-8 MV) and treatment outcome in patients treated with conservative surgery and radiation therapy.
Methods and materials: Between 1968 and 1985, 1624 patients were treated for clinical Stage I or II invasive breast cancer. The study population was limited to 1380 patients who underwent complete gross excision and received greater than or equal to 60 Gy to the tumor bed. Of these, 1125 were treated on a 4 MV, 153 on a 6 MV, and 102 on an 8 MV linear accelerator. Patients were selected for treatment on the 8 MV machine based on chest wall separations greater than 24 cm. Of patients treated on the 8 MV, netting was used for 42% and bolus was used for 26%. The median dose with bolus was 14 Gy in seven fractions (range: 2-34.2 Gy). Patients treated on the 8 MV accelerator were older, had a higher percentage of clinical T2 tumors, a higher percentage of pathologically positive nodes, and a lower incidence of extensive intraductal component (EIC). Median follow-up times were 130, 153, and 102 months, respectively, for survivors treated on the 4, 6, and 8 MV machines.
Results: We analyzed the site and 5-year crude incidence of first failure by machine energy and found the pattern of first failure site (local, nodal, or distant) to be virtually identical for each energy group. Of the local failures, 12 were in the skin of the treated breast, and these failures were evenly distributed by machine energy. We performed a multivariate analysis to adjust for factors known to predict for treatment failure. When adjusted for these other variables, machine energy was not associated with an increased (or decreased) risk of recurrence (RR for 8 MV vs. 4 MV = 0.94, p = 0.7; RR for 6 MV vs. 4 MV = 1.0, p = 0.9). We also analyzed the nature and incidence of treatment complications (rib fracture, radiation pneumonitis, soft tissue necrosis, and brachial plexopathy) and found no significant differences among the three treatment groups when stratified by treatment technique (tangents only vs. three-field). There was also no significant difference in cosmetic outcome at 5 years among the three groups.
Conclusions: We conclude that machine energy over the range of 4 to 8 MV does not significantly affect treatment outcome. Specifically, it was feasible to treat patients with large chest wall separations using an 8 MV machine without an increase in skin recurrences and with the improved dose homogeneity afforded by 8 MV machines as compared with those of lower energies.