Telomere length in prospective and retrospective cancer case-control studies.
Telomere length in prospective and retrospective cancer case-control studies.
复制标题
DOI：
10.1158/0008-5472.can-09-4595
复制
发表时间：
2010-04-15
期刊：
影响因子：
11.2
通讯作者：
Dunning AM
中科院分区：
文献类型：
--
作者：
Pooley KA;Sandhu MS;Tyrer J;Shah M;Driver KE;Luben RN;Bingham SA;Ponder BA;Pharoah PD;Khaw KT;Easton DF;Dunning AM
先前的研究报告称，淋巴细胞中较短的平均端粒长度与对常见衰老疾病的易感性增加有关，并且可能预测癌症风险。然而，大多数分析都是回顾性收集病例对照研究。使用高通量定量真实的时间PCR测量平均端粒长度。在东安格利亚乳腺癌（2243例，2181例对照）和结直肠癌（2249例，2161例对照）研究中，在癌症诊断后收集血液用于DNA提取。对EPIC-Norfolk队列中的乳腺癌（199例）和结直肠癌（185例）进行了前瞻性病例对照研究。血液在诊断前至少6个月收集，并与每个病例的两个无癌症对照的DNA相匹配。在回顾性研究中，平均端粒长度的最短（Q4）与最长（Q1）四分位数的年龄调整比值比为15.5（95%CI 11.6-20.8），p-het=5.7×10−75;乳腺癌的“每四分位数”p趋势=2.1×10−80，（95%CI 1.77-2.59），p-het=7.3×10−15;结直肠癌的“每四分位数”p趋势=1.8×10−13。在前瞻性EPIC研究中，乳腺癌的可比比值比[Q4 vs. Q1]为1.58（95%CI 0.75-3.31），p-het=0.23，结直肠癌风险为1.13（95%CI 0.54-2.36），p-het=0.75。在回顾性收集的病例中，平均端粒长度比对照组短，但在前瞻性研究中，等效关联明显较弱。这表明端粒缩短主要发生在诊断后，因此可能对癌症预测没有价值。
Previous studies have reported that shorter mean telomere length in lymphocytes is associated with increased susceptibility to common diseases of aging, and may be predictive of cancer risk. However, most analyses have examined retrospectively-collected case-control studies. Mean telomere length was measured using high-throughput quantitative Real Time PCR. Blood for DNA extraction was collected after cancer diagnosis in the East Anglian SEARCH Breast (2243 cases, 2181 controls) and SEARCH Colorectal (2249 cases, 2161 controls) studies. Prospective case-control studies were conducted for breast cancer (199 cases) and colorectal cancer (185 cases), nested within the EPIC-Norfolk cohort. Blood has been collected at least 6 months prior to diagnosis, and was matched to DNA from two cancer-free controls per case. In the retrospective, SEARCH studies, the age-adjusted Odds Ratios for shortest (Q4) vs. longest (Q1) quartile of mean telomere length was 15.5 (95%CI 11.6–20.8), p-het=5.7×10−75; with a ‘per quartile’ p-trend=2.1×10−80 for breast cancer, and 2.14 (95%CI 1.77–2.59), p-het=7.3×10−15; with a ‘per quartile’ p-trend=1.8×10−13 for colorectal cancer. In the prospective, EPIC study, the comparable Odds Ratios [Q4 vs. Q1] were 1.58 (95%CI 0.75–3.31), p-het=0.23 for breast cancer, and 1.13 (95%CI 0.54–2.36), p-het=0.75 for colorectal cancer risk. Mean telomere length was shorter in retrospectively-collected cases than in controls but the equivalent association was markedly weaker in the prospective studies. This suggests that telomere shortening largely occurs after diagnosis, and may not, therefore, be of value in cancer prediction.