Familial concordance in cancer survival:: a Swedish population-based study
Familial concordance in cancer survival:: a Swedish population-based study
复制标题
DOI：
10.1016/s1470-2045(07)70282-6
复制
发表时间：
2007-11-01
期刊：
影响因子：
51.1
通讯作者：
Czene, Kamila
中科院分区：
文献类型：
--
作者：
Lindstrom, Linda S.;Hall, Per;Czene, Kamila
背景如今，癌症可以在家庭中聚集的事实已被普遍接受。本研究的目的是完成一个全面的分析癌症的生存一致性在父母和他们的孩子诊断为相同cancer.Methods我们使用了一个基于人口的瑞典家庭数据库，其中包括约300万个家庭和数据超过一百万个人与癌症。我们使用Kaplan-Meier方法分析了儿童生存率与父母生存率的关系。然后，我们通过使用两个多变量比例风险（考克斯）模型调整可能的混杂因素，对儿童与父母生存率的风险进行建模。结果在我们的单变量Kaplan-Meier分析中，与父母患有相同癌症且父母在诊断后10年内死亡的儿童，乳腺癌的生存率明显较差。（log rank p=0.01）、结直肠癌（p=0.04）和前列腺癌（p=0.05）。通过使用考克斯模型，我们注意到，与那些父母患有结直肠癌的儿童相比，父母存活率低的儿童癌症死亡的风险比增加（风险比[HR] 1.44 [95% CI 1.01-2.01]），肺癌（1.39 [1.00-1.94]）、乳腺癌（1.75 [1.13-2.71]）、卵巢癌（2.23 [0.78-6.34]）和前列腺癌（2.07 [1.13-3.79]）。除卵巢癌外，所有风险比估计值均具有显著性，儿童死亡风险随父母生存结局恶化程度增加的趋势显著，父母生存结局恶化程度定义为生存四分位数（即，好[最佳四分位数]，预期[中间两个四分位数]或差[最差四分位数]）。因此，父母的存活数据可能有可能指导治疗决定和遗传咨询。最后，强调癌症生存的遗传决定因素的分子研究现在是必要的。
Background Nowadays, the fact that cancers can aggregate in families is generally accepted. The aim of this study was to complete a comprehensive analysis of cancer-survival concordance in parents and their children diagnosed with the same cancer.Methods We used a population-based Swedish family database, that included about three million families and data for more than a million individuals with cancer. We analysed survival in children in relation to parental survival by use of the Kaplan-Meier method. We then modelled the risk in children in relation to parental survival by use of two multivariate proportional hazard (Cox) models adjusting for possible confounders of survival.Findings In our univariate Kaplan-Meier analysis, children with the same cancer as their parent and whose parent had died within 10 years of diagnosis showed significantly worse survival for breast (log rank p=0.01), colorectal (p=0.04), and prostate cancer (p=0.05) than those whose parents were alive at 10 years from diagnosis. By use of Cox modelling, we noted an increased hazard ratio for death from cancer in children with poor parental survival compared with those with good parental survival for colorectal cancer (hazard ratio [HR] 1.44 [95% CI 1.01-2.01]), lung cancer (1.39 [1.00-1.94]), breast cancer (1.75 [1.13-2.71]), ovarian cancer (2.23 [0.78-6.34]), and prostate cancer (2.07 [1.13-3.79]). All hazard-ratio estimates, except for ovarian cancer, were significant, with significant trends of increasing risk of death in children by degree of worsening survival outcome in parents defined in quartiles of survival (ie, good [best quartile], expected [middle two quartiles], or poor [worst quartile]).Interpretation Our findings suggest that cancer-specific survival in parents predicts survival from the same cancer in their children. Consequently, data on survival in a parent might have the potential to guide treatment decisions and genetic counselling. Finally, molecular studies to highlight the genetic determinants of cancer survival are now warranted.