2012年10月28-31日，第54届美国放射肿瘤学会（ASTRO）年会在美国波士顿举行。来自世界各地的放射肿瘤学领域的医生及相关人士共计11000多人参会。本届年会的主题是“通过创新改善患者治疗（Advancing Patient Care through Innovation）”。 对本届年会进行了专题报道（http://zt.cmt.com.cn/zt/astro2012/index.html），敬请关注！

　　会上，纪念Sloan-Kettering肿瘤中心的B. McCormick报告了一项研究，题为《低危乳腺原位导管癌（DCIS） ：RTOG9804Ⅲ临床试验结果》，其摘要如下：

低危乳腺原位导管癌（DCIS） ：RTOG 9804 Ⅲ临床试验结果

Low-risk Breast Ductal Carcinoma In Situ (DCIS): Results From the Radiation Therapy Oncology Group 9804 Phase 3 Trial

Purpose/Objective(s)
The quest to identify women with DCIS, who receive no significant benefit from radiation (RT), continues despite several phase III trials demonstrating RT reduces the risk of local failure (LF). RTOG 9804 identified “good risk” patients based on the best pathology knowledge of the time (mammographic detection, <2.5 cm size, margins ≥3 mm and low-intermediate nuclear grade). Women were randomized to RT or observation (OBS). Tamoxifen (Tam) for 5 years was optional in both arms.

Materials/Methods
Ipsilateral local failure (LF) was the primary end point; LF and contralateral failure (CF) were estimated using cumulative incidence and overall (OS) and disease-free (DFS) survival were estimated using Kaplan-Meier method. Treatment arms were compared by the log-rank test. With 1,790 patients, 80% power and using a 2-sided log-rank test at 0.05, 9,804 was designed to detect a reduction in 5 year LR from 6.0% to 3.5% with RT. Dose in the RT arm was 50 Gy in 25-28 or 42 Gy in 16 fractions, with no boost allowed.

Results
Due to lower than projected accrual, the study closed early in 2006 with 636 patients, and 585 eligible patients are included in this analysis. Median follow-up time was 7.17 years, with Tam used in 62%; mean age was 59. At 7 years, LF in the RT arm was 0.9% (95% CI: 0.0-2.2) versus 6.4% in the OBS arm (95% CI: 3.2-9.6), p = 0.0005. In the OBS arm, 12 of 18 LF occurred in the same quadrant. The 2 LF in the RT arm were in distant quadrants. No failures involving skin were observed in either arm. With limited events: age, grade, margin status, and size did not correlate with LF. Rate of grade (G) 1-2 worst acute toxicity was 30% versus 76%; G3-4 toxicities were 4.0% and 4.2%, respectively in the OBS and RT arms. The rate of worst late RT toxicity was: G1 30%, G2 4.6%, and G3 0.7%. OS and DFS were excellent in both arms.

Conclusions
In this “good risk” subset of DCIS, the LF rate was decreased significantly with the addition of RT. Longer follow-up is planned, as late failures continue to occur.