Chlebowski R T, Lillington L M
Division of Medical Oncology, Harbor-University of California at Los Angeles Medical Center, Torrance 90509.
J Clin Oncol. 1994 Sep;12(9):1789-95. doi: 10.1200/JCO.1994.12.9.1789.
To test the hypothesis that clinical research results have driven changes in recent breast cancer management recommendations.
All breast cancer abstracts in the Program/Proceedings of the American Society of Clinical Oncology (ASCO) from 1984 to 1993 were prospectively reviewed in 31 areas and categorized by study type, study question, whether statistical significance was claimed, and whether the abstract was presented.
Of 1,372 abstracts, 54% reported on prospective clinical trials (PCTs) and 17% on randomized clinical trials (RCTs). The total number of published abstracts progressively increased (from 87 in 1984 to 221 in 1993) and author citations nearly quadrupled (from 430 in 1984 to 1,642 in 1993, P < .01); however, RCTs have come to represent a smaller proportion of reports: 37% (33 of 89) in 1986 versus 10% (22 of 221) in 1993 (P < .001). The size of adjuvant-therapy RCTs has progressively increased (mean +/- SEM subjects/trial, 237 +/- 43 in 1984 to 874 +/- 374 in 1993), but has remained small in advanced-disease RCTs (mean +/- SEM subjects/trial, 145 +/- 25 in 1984 to 146 +/- 34 in 1993). For adjuvant therapy, 14 of 90 RCTs (with 51,207 patients) reported a significant (P < .05) survival benefit for investigational therapies (16%). For advanced-disease therapy, only three of 141 RCTs (with 26,281 patients) reported a significant (P < .05) survival benefit for investigational therapies (2%). Randomization was rarely used in trials of dose-intensity with blood-product support (zero of 86 trials) or locally advanced disease.
For breast cancer ASCO abstracts in the past decade, we determined the following: (1) adjuvant trials have not infrequently supported study hypotheses; and (2) advanced-disease trials have consistently failed to identify new approaches with a significant impact on survival. These results suggest that a critical process evaluation of current policy and procedures involved in directing breast cancer research is warranted, especially for strategies in advanced disease.
检验临床研究结果推动了近期乳腺癌管理建议变化这一假设。
对1984年至1993年美国临床肿瘤学会（ASCO）会议议程/论文集中的所有乳腺癌摘要进行前瞻性审查，涉及31个领域，并按研究类型、研究问题、是否声称具有统计学意义以及摘要是否被发表进行分类。
在1372篇摘要中，54%报告了前瞻性临床试验（PCT），17%报告了随机临床试验（RCT）。已发表摘要的总数逐渐增加（从1984年的87篇增至1993年的221篇），作者引用次数几乎翻了两番（从1984年的430次增至1993年的1642次，P <.01）；然而，RCT在报告中所占比例越来越小：1986年为37%（89篇中的33篇），1993年为10%（221篇中的22篇）（P <.001）。辅助治疗RCT的规模逐渐增大（平均±标准误，每个试验的受试者数量，1984年为237±43，1993年为874±374），但晚期疾病RCT的规模仍然较小（平均±标准误，每个试验的受试者数量，1984年为145±25，1993年为146±34）。对于辅助治疗，90项RCT中的14项（涉及51207名患者）报告研究治疗具有显著（P <.05）生存获益（16%）。对于晚期疾病治疗，141项RCT中的仅3项（涉及26281名患者）报告研究治疗具有显著（P <.05）生存获益（2%）。在有血液制品支持的剂量强度试验（86项试验中的0项）或局部晚期疾病试验中很少使用随机化。
对于过去十年ASCO的乳腺癌摘要，我们确定了以下几点：（1）辅助试验经常支持研究假设；（2）晚期疾病试验一直未能确定对生存有重大影响的新方法。这些结果表明，有必要对指导乳腺癌研究的当前政策和程序进行关键的过程评估，特别是对于晚期疾病的策略。
