Lustbader E D, Williams W R, Bondy M L, Strom S, Strong L C
Fox Chase Cancer Center, Philadelphia, PA 19111.
Am J Hum Genet. 1992 Aug;51(2):344-56.
This paper presents the analysis of familial cancer data collected in a hospital-based study of 159 childhood soft-tissue-sarcoma patients. Two different statistical models detected excess aggregation of cancer, which could be explained by a rare dominant gene. For each kindred, we estimated the probability of the observed cancer distribution under the dominant-gene model and identified 12 families that are the most likely to be segregating the gene. Two of those families have confirmed germ-line mutations in the p53 tumor-suppressor gene. The relative risk of affection for children who are gene carriers was estimated to be 100 times the background rate. Females were found to have a slightly higher age-specific penetrance, but maternal and paternal lineages made equal contributions to the evidence in favor of the dominant gene. The proband's histology, ethnicity, and age at diagnosis were evaluated to determine whether any of these altered the probability of affection in family members. Only embryonal rhabdomyosarcoma was found to be a significant covariate under the dominant-gene model. While molecular genetic studies of familial cancer will eventually provide answers to the questions of genetic heterogeneity, age- and site-specific penetrance, mutation rates, and gene frequency, information from statistical models is useful for setting priorities and defining hypotheses.
本文介绍了在一项基于医院的针对159例儿童软组织肉瘤患者的研究中收集的家族性癌症数据的分析情况。两种不同的统计模型检测到癌症存在过度聚集现象,这可以用一种罕见的显性基因来解释。对于每个家系,我们估计了在显性基因模型下观察到的癌症分布概率,并确定了12个最有可能携带该基因的家系。其中两个家系已证实p53肿瘤抑制基因存在种系突变。基因携带者子女受影响的相对风险估计是背景发病率的100倍。发现女性具有略高的年龄特异性外显率,但母系和父系血统对支持显性基因的证据贡献相同。对先证者的组织学、种族和诊断时的年龄进行了评估,以确定这些因素是否会改变家庭成员受影响的概率。在显性基因模型下,仅胚胎性横纹肌肉瘤被发现是一个显著的协变量。虽然家族性癌症的分子遗传学研究最终将为遗传异质性、年龄和部位特异性外显率、突变率和基因频率等问题提供答案,但统计模型提供的信息对于确定研究重点和定义假设很有用。