Huncharek Michael, Kupelnick Bruce
Division of Radiation Oncology, Department of Clinical Oncology, Marshfield Clinic Cancer Center, Wisconsin, USA.
Am J Clin Oncol. 2003 Aug;26(4):402-7. doi: 10.1097/01.COC.0000026911.98171.C6.
Bacille Calmette-Guérin (BCG) immunotherapy is currently considered the most effective agent in the management of superficial bladder cancer. Prior work suggests that the efficacy of intravesical chemotherapy in preventing tumor recurrence may be greater than previously suggested. This latter finding, therefore, brings into question the currently perceived superiority of BCG treatment for this disease. A metaanalysis was performed to rigorously examine existing data relevant to this relationship and to quantify the relative efficacy of both treatment modalities on tumor recurrence. A prospective protocol outlining the above-noted metaanalysis was initially developed followed by a thorough search of the existing published literature using strict eligibility criteria. Nine randomized trials were found that met protocol specifications. These reports contained data on 2,261 patients that were statistically combined using a fixed effects model (Peto). The outcome of interest was the proportion of patients with recurrence at 1, 2, and 3 years following intravesical therapy (i.e., a summary odds ratio, ORp). Combining all nine trials using 1-year recurrence as the endpoint demonstrated significant statistical heterogeneity, although the ORp favored BCG over intravesical chemotherapy (0.89 [0.74-1.07]). This precluded statistical pooling of the data and sensitivity analyses were performed to determine the source of heterogeneity. These tests showed that the prior chemotherapy treatment in a large number of the randomized trials biased study results in favor of the BCG arms. Once the data were stratified on presence or absence of prior drug treatment, intravesical chemotherapy reduced 1-, 2-, and 3-year recurrence by 21% to 82% versus BCG, depending on the endpoint of interest. The available data suggest that clinical trials directly comparing intravesical BCG to intravesical chemotherapy must stratify on the presence or absence of prior chemotherapy. Recurrences following prior intravesical chemotherapy appear less responsive to drug therapy than those in chemotherapy-naive patients. The currently perceived superiority of BCG therapy may therefore be an artifact of this phenomenon, since most randomized trials include chemotherapy failures in their chemotherapy treatment arms.
卡介苗（BCG）免疫疗法目前被认为是浅表性膀胱癌治疗中最有效的药物。先前的研究表明，膀胱内化疗在预防肿瘤复发方面的疗效可能比之前认为的更大。因此，后一项发现使目前认为BCG治疗该疾病具有优越性的观点受到质疑。进行了一项荟萃分析，以严格审查与这种关系相关的现有数据，并量化两种治疗方式对肿瘤复发的相对疗效。最初制定了一份概述上述荟萃分析的前瞻性方案，随后使用严格的纳入标准对现有已发表文献进行了全面检索。发现有9项随机试验符合方案规范。这些报告包含了2261名患者的数据，使用固定效应模型（Peto）进行了统计合并。感兴趣的结果是膀胱内治疗后1年、2年和3年复发患者的比例（即汇总比值比，ORp）。以1年复发作为终点对所有9项试验进行合并，显示出显著的统计学异质性，尽管ORp显示BCG优于膀胱内化疗（0.89 [0.74 - 1.07]）。这使得无法对数据进行统计合并，并进行了敏感性分析以确定异质性的来源。这些测试表明，大量随机试验中的先前化疗治疗使研究结果偏向于BCG组。一旦根据是否有先前的药物治疗对数据进行分层，根据感兴趣的终点，膀胱内化疗与BCG相比，可将1年、2年和3年复发率降低21%至82%。现有数据表明，直接比较膀胱内BCG与膀胱内化疗的临床试验必须根据是否有先前化疗进行分层。先前膀胱内化疗后的复发对药物治疗的反应似乎不如未接受过化疗的患者。因此，目前认为BCG疗法具有优越性可能是这一现象造成的假象，因为大多数随机试验在其化疗治疗组中纳入了化疗失败的患者。
