Factors determining outcome in patients treated with interstitial implantation as a radiation boost for breast conservation therapy.

PURPOSE: To evaluate the relative utility of interstitial implant as a technique for tumor bed dose escalation and assess technical factors related to outcome. METHODS AND MATERIALS: From 1982-1994, a prospectively applied institutional policy of margin-directed boost dose escalation to the tumor bed was followed whereby interstitial implantation was commonly employed for a final margin status (FMS) < or = 2 mm. There were 509 treated breasts, of which 127 received an implant boost. For purposes of comparison, cases were broadly classed as "implant" (all FMS < or = 2 mm) and "nonimplant" (FMS < or = 2 mm or FMS > 2 mm). The implant target volume was determined at completion of whole breast irradiation by clinical assessment. All implants were constructed in accordance with a preplanning algorithm designed to maximize dose homogeneity within a prescription isodose goal of 0.50 Gy/h for 40 h. Local control and cosmetic outcome were evaluated with respect to extent of tumor, histopathology, FMS, extent of surgery, and systemic adjuvant therapy. Implant quality was assessed using four calculated parameters: strand separation quotient (SSQ), planar separation quotient (PSQ), global separation quotient (GSQ), and dose homogeneity index (DHI). The mean implant volume was 48.3 +/- 20 cc, the mean prescribed dose rate was 0.46 +/- 0.08 Gy/h, and the mean total implant dose was 19.94 +/- 1.52 Gy.
RESULTS: Cosmetic outcome was good/excellent in 90% of implant and 83% of all nonimplant cases, which was not statistically different. Cosmesis was significantly superior with implant when compared to nonimplant cases receiving an external boost of 20 Gy. Logistic regression analyses of implant cases revealed that reexcision volume and decreased DHI were associated with adverse cosmesis. There were 10 local failures in the implanted patients (4 within the prescribed isodose volume, 5 at the periphery, and 1 elsewhere in the breast). The local failure rate at 5 and 7 years in the implanted group was 3.9 and 9.0%, respectively, compared to nonimplant cases with a margin < or = 2 mm of 3.2 and 3.2%, respectively. These differences were not significant. The crude local failure rate in patients with an associated DCIS component was 12% a compared to 3% in patients with pure invasive histology (p = 0.06). A proportional hazards survival model revealed a significant association of local failure with the performance of a reexcision and young age.
CONCLUSION: We conclude that interstitial implant boost for breast conserving irradiation is associated with cosmesis that is superior than the same nominal dose of external beam boost, although this is highly dependent upon the technical quality of the source position and the relative uniformity of dose deposition. Breast implantation results in a rate of local control no better than dose-matched external beam boost in patients with a final margin < or = 2 mm. Local control with implantation might be further enhanced by increasing implant volume and/or improved target localization.