The Long and Short of Telomeres and Cancer Association Studies

A growing understanding of telomere biology coupled with advances in telomere length (TL) measurement methods has led to a deluge of association studies seeking to understand the relationship between germline TL and cancer. Telomeres consist of DNA nucleotide repeats and a protein complex that cap chromosome ends to preserve chromosomal stability. Telomere shortening is a consequence of normal cell division due to semiconservative replication of DNA at chromosome ends, known as the end-replication problem (1). In normal cells, when telomeres shorten to a critical length, cellular senescence or apoptosis is triggered. However, cancer cells can continue to divide, despite the apparent chromosomal instability, by upregulating key pathways related to telomerase, p53, or retinoblastoma protein. Telomeres of somatic cancer cells can be shorter than the corresponding normal tissue, but there is great heterogeneity in TL in cancer cells (2). Association studies of germline or surrogate tissue (eg, blood or buccal cells) and cancer have yielded variable results, particularly between cancer types but also within cancers (3,4).

Telomere maintenance is tightly regulated by a large, complex set of proteins (1). Cancer genome-wide association studies have mapped genetic variants to two notable genes, telomerase (TERT) (5) and regulator of telomere elongation helicase 1 (RTEL1) (6). TERT harbors susceptibility alleles for at least ten different cancers, suggesting that its subtle perturbation contributes to cancer risk. However, the functional consequence of the susceptibility alleles in or near TERT or RTEL1 appears to have minimal effect on TL. On the other hand, rare, highly penetrant mutations in key telomere biology genes, including TERT and RTEL1 (as well as CTC1, DKC1, NHP2, NOP10, TERC, TINF2, and WRAP53) cause dyskeratosis congenita, a cancer-prone inherited bone marrow failure syndrome in which telomere lengths are extremely short (eg, less than the first percentile for their age) resulting in an 11-fold increased risk of cancer (7–9). These findings have stimulated investigation of telomere length in surrogate tissues as a cancer risk factor in the general population.

In this issue of the Journal, Weischer and colleagues have conducted a large, prospective, population-based study to evaluate the association between peripheral leukocyte TL and either cancer risk or survival after cancer diagnosis (10). They measured TL by quantitative polymerase chain reaction (qPCR) in 47 102 individuals from two Danish cohorts with up to 20 years follow-up (median follow-up = 6 years); 3142 developed first cancers, and 1730 died. After adjusting for age and known cancer risk factors, short TL was associated with death after a diagnosis of cancer but notably not associated with the risk of developing cancer. A notable feature of this study is that the authors carefully adjusted the analysis for cancer risk factors (eg, smoking, obesity, and sedentary lifestyle), which have also been associated with TL.

Measurement of TL in large epidemiologic studies has become popular since 2002 when the qPCR method was first reported (11,12). qPCR is advantageous in such studies because it is amenable to high-throughput platforms and uses very small amounts of DNA. This is in contrast to the gold standard of terminal restriction fragment (TRF) Southern blots, which require larger amounts of DNA, are less practical for large studies, and are challenging to quantitate. Although Cawthon (12) showed adequate correlation between qPCR and terminal restriction fragment, others have suggested a nonlinear correlation (13). Moreover, most papers have not reported laboratory standards for qPCR/terminal restriction fragment correlations or important details concerning DNA extraction methods or handling. Hence, interlaboratory variability as well as measurement error may underlie some of the contradictory reports, which may be difficult to disentangle without better reporting standards.

Early association studies of TL and cancer suggested that shorter TL is associated with increased risk for certain cancers, such as renal cell and head and neck cancers (14,15). Two meta-analyses have been conducted to determine whether “short telomeres are associated with cancer risk” and found an increased risk overall (3,4). Specifically, short telomeres were associated with cancer risk (odds ratio [OR] = 1.35, 95% confidence interval [CI] = 1.14 to 1.60) in a meta-analysis of 21 published studies (3). In an analysis stratified by study design, Wentzensen and colleagues found that the observed high risk is mainly driven by the case–control studies (OR in pooled analysis = 1.96; OR in case–control studies = 2.9; OR in prospective studies = 1.16) (4). This suggests the possible effect of reverse causation bias in the case–control studies. Factors that likely contribute to this include unknown systemic effects (ie, inflammation) of cancer on surrogate tissue telomeres and the fact that most studies have not reported whether cases received chemotherapy or radiation therapy, both of which could damage telomeres, before DNA collection.

The Weischer et al. study is an important contribution to the literature, and its strengths include the population-based design to minimize selection bias, large sample size, long follow-up, and prospective study design. The latter is important for minimizing reverse causality. Moreover, the authors report replicate qPCR assays with tight coefficients of variation, indicating reliable results. Still, the lack of ethnic diversity in their study limits its applicability across populations, especially because preliminary studies suggest slightly different norms for TL values by populations. Future studies comparing ethnic groups are needed to understand if and why longer TLs are observed in African Americans (16). This could lead to insights into different cancer risk profiles between distinct ethnic groups.

Because telomere shortening is observed in the aging process, several studies have pursued a likely connection between TL and mortality. In a prospective, population-based study using qPCR, short telomere length at baseline was associated with higher cancer risk and mortality with a higher hazards ratio for the latter (3.1 and 11.1, respectively) (17). Weischer et al. found that individuals with shorter telomeres were more likely to die from cancer than those who had cancer and longer telomeres. In the Cardiovascular Health Study, shorter telomeres were associated with 60% risk of death, mainly from infectious or cardiovascular cause, but interestingly, not cancer (16). Although these prospective studies suggest a connection between short telomeres and mortality, there is more to investigate. It is possible that individuals with telomeres shorter than average could be at risk of dying sooner than those with longer telomeres, regardless of the cause of death, because of more limited cellular replicative capacity. Until even larger studies are done, preferably in prospective cohorts, it will be difficult to separate which causes of death are clearly linked to shorter TL.

In conclusion, association studies of TL and cancer have likely suffered from important methodological issues, including reverse causation bias, lack of interlaboratory comparisons of TL measurements, and limited reporting of DNA extraction and handling methods. Because therapeutic interventions can affect telomere lengths, it will be important to integrate clinical data in future studies. If telomere lengths are to be used as effective biomarkers for cancer prevention or intervention, it will be necessary to more thoroughly understand the connection between telomere biology, cancer, chronic disease, and mortality in the general population.

Funding

This work was support by the intramural research program of the Division of Cancer Epidemiology and Genetics , National Cancer Institute, National Institutes of Health.

The authors report no conflicts of interest. The study sponsor had no role in the writing of the editorial or the decision to submit it for publication.