全球X线摄影筛查随机对照试验的乳腺癌疾病自然史研究

Natural history of breast cancer：a systematic review of worldwide randomized controlled trials of mammography screening

目的：汇总全球乳腺X线摄影筛查随机对照试验（ RCT），分析乳腺癌疾病进展的1年转化概率。方法基于Cochrane 2013年乳腺癌X线摄影筛查系统综述和国际癌症研究署界定的全球范围内已完成的乳腺癌筛查RCT，根据试验名称和团队主要研究人员，结合＂乳腺癌＂、＂X线摄影＂、＂随机对照试验＂、＂发病率＂等检索词，在PubMed中检索并进行筛选，主要检索多年随访累积发病率，使用寿命表法折算为1年转化概率。结果最终纳入RCT 9项，文献23篇。从健康状态发展为癌前病变的概率数据缺失。干预组和对照组从健康状态发展为原位癌的1年转化概率范围分别为17．78／10万～50．21／10万和9．16／10万～26．84／10万，从健康状态发展为乳腺癌（不区分原位癌和浸润癌）的1年转化概率范围分别为143．75／10万～316．97／10万和141．45／10万～288．84／10万。干预组和对照组从健康状态发展为浸润癌的1年转化概率分别为159．79／10万～264．60／10万和170．12／10万～255．33／10万。从原位癌进展为浸润癌的概率因病理类型不同有所差异。结论文献报道的状态主要有完全健康、原位癌和浸润癌，不同状态间转化概率可为乳腺癌自然史模型研究提供部分参考数据，癌前病变相关参数量化将是下一步二次文献研究重点。

Objective To parameterize the 1？year transition probabilities between different health status of the natural history of breast cancer based on the data of randomized controlled trial of X？ray mammography screening worldwide. Methods Based on the breast cancer screening randomized controlled trials defined by a mammography screening review from the Cochrane 2013 and the International Agency for Research on Cancer, a systematic review was initiated in PubMed by searching names of the key investigators of the trials, combined with the diseases, screening intervention and outcome indicators. If applicable, all the original cumulative incidence rates were converted into one？year transition rate, using the life？tablenbsp;approach considering time length of follow？up. Results A total of 23 reports from 9 RCTs were included. The data on transition rate between the healthy status to precancerous lesions was absent. The 1？year transition rate from health to carcinoma in situ was 17. 78 to 50. 21 per 100 000 persons in the intervention group and 9.16 to 26.84 per 100 000 persons in the control group. Correspondingly, the 1？year transition rate from health to breast cancer （including carcinoma in situ and invasive cancer） were estimated as 143.75 to 316.97 per 100 000 persons in the intervention group, and 141. 45 to 288. 84 per 100 000 persons in the control group. Furthermore, the transition rate from the healthy status to invasive breast cancer was 159.79 to 264.60 per 100 000 persons in intervention group and 170.12 to 255.33 per 100 000 persons in control group. The transition rate from carcinoma in situ to invasive breast cancer varied among different pathological types. Conclusions The most common natural history states of reported by the included trials involved the full healthy status, carcinoma in situ and invasive breast cancer. The findings of transition rates between different health statuses will be informative for future model development of natural history studies of breast cancer. Information in relation to breast precancerous lesions still limited and needs to be further addressed.