Sherman Simon, Rathnayake Nirosha, Mdzinarishvili Tengiz
Eppley Institute for Research in Cancer, University of Nebraska Medical Center, Omaha, Nebraska, United States of America; College of Public Health, University of Nebraska Medical Center, Omaha, Nebraska, United States of America.
School of Natural Sciences and Engineering, Ilia State University, Tbilisi, Georgia.
PLoS One. 2015 Oct 14;10(10):e0140405. doi: 10.1371/journal.pone.0140405. eCollection 2015.
Carcinogenic modeling is aimed at mathematical descriptions of cancer development in aging. In this work, we assumed that a small fraction of individuals in the population is susceptible to cancer, while the rest of the population is resistant to cancer. For individuals susceptible to cancer we adopted methods of conditional survival analyses. We performed computational experiments using data on pancreatic, stomach, gallbladder, colon and rectum, liver, and esophagus cancers from the gastrointestinal system collected for men and women in the SEER registries during 1975-2009. In these experiments, we estimated the time period effects, the birth cohort effects, the age effects and the population (unconditional) cancer hazard rates. We also estimated the individual cancer presentation rates and the individual cancer resistance rates, which are, correspondingly, the hazard and survival rates conditioned on the susceptibility to cancer. The performed experiments showed that for men and women, patterns of the age effects, the individual cancer presentation rates and the individual cancer resistance rates are: (i) intrinsic for each cancer subtype, (ii) invariant to the place of living of the individuals diagnosed with cancer, and (iii) well adjusted for the modifiable variables averaged at a given time period. Such specificity and invariability of the age effects, the individual cancer presentation rates and the individual cancer resistance rates suggest that these carcinogenic characteristics can be useful for predictive carcinogenic studies by methods of inferential statistics and for the development of novel strategies for cancer prevention.
致癌建模旨在对衰老过程中的癌症发展进行数学描述。在这项工作中，我们假设人群中有一小部分个体易患癌症，而其余人群对癌症具有抗性。对于易患癌症的个体，我们采用了条件生存分析方法。我们使用1975 - 2009年期间从监测、流行病学和最终结果（SEER）登记处收集的男性和女性胃肠道系统中胰腺癌、胃癌、胆囊癌、结肠癌和直肠癌、肝癌及食管癌的数据进行了计算实验。在这些实验中，我们估计了时间段效应、出生队列效应、年龄效应以及人群（无条件）癌症风险率。我们还估计了个体癌症发病率和个体癌症抗性率，它们分别是基于对癌症易感性的风险率和生存率。所进行的实验表明，对于男性和女性，年龄效应模式、个体癌症发病率和个体癌症抗性率具有以下特点：（i）每种癌症亚型所固有，（ii）与被诊断患有癌症个体的居住地点无关，并且（iii）针对给定时间段内平均的可改变变量进行了良好调整。年龄效应、个体癌症发病率和个体癌症抗性率的这种特异性和不变性表明，这些致癌特征可用于通过推断统计方法进行预测性致癌研究以及开发新的癌症预防策略。
