|Home | About | Journals | Submit | Contact Us | Français|
Telomere length plays an important role in maintaining chromosomal stability and in tumorigenesis. We hypothesized that telomere length in peripheral white blood cell DNA obtained from healthy individuals would be a predictor of future risk of developing non-Hodgkin lymphoma (NHL).
Using a new assay to measure relative telomere length, monochrome multiplex quantitative PCR, which strongly correlates with telomere length measured by Southern blot (Spearman r = 0.91, p < 0.0001) and has high precision (coefficient of variation = 7%), we compared telomere length in peripheral white blood cell DNA in 107 incident male NHL cases and 107 matched controls within the prospective Alpha-Tocopherol, Beta-Carotene Cancer Prevention (ATBC) Study cohort.
Median (10th, 90th percentile) telomere length was 1.10 (0.79, 1.43) in cases and 1.02 (0.78, 1.26) in controls (p = 0.0017, Wilcoxon sign test). There was a strong dose-response relationship between quartiles of telomere length and risk of NHL overall [odds ratios (95% confidence intervals) by quartile: 1.0; 1.1 (0.4-2.7); 1.8 (0.7-4.9); and 3.6 (1.4-8.9); p trend = 0.003)], and this association was similar across the most common NHL subtypes present in this study.
These results suggest that longer telomere length may be a potential predictor for future risk of NHL.
Telomeres are complexes of tandem repeats of the sequence TTAGGG that cap chromosomes, are essential for protecting chromosomal integrity (1), and shorten after every cell division. Short telomere length can cause genomic instability, which is associated with the initiation and progression of human cancers (2). At the same time, many incipient tumors can terminate their own growth by shortening their telomeres sufficiently to trigger replicative senescence or apoptosis (3). However, if sufficient numbers of mutations that promote growth and block cell senescence and apoptotic pathways accumulate in a cell before its telomeres shorten enough to trigger senescence or apoptosis and protect it from cancer, then unlimited proliferation may ensue. It follows that in some cell types, under some circumstances, long telomeres may actually increase the risk of cancer, by allowing more time and more cell divisions during which the cell can accumulate oncogenic mutations.
Most epidemiological studies have reported that relatively shorter telomere length measured in peripheral white blood cells and in some instances buccal cells is associated with increased risk of cancer (4-10). In contrast, some recent reports have suggested that longer telomere length may be associated with increased risk for certain tumors, such as breast cancer and melanoma (11,12). Most studies have used a case-control design (4-8), with several more recent reports using a prospective cohort design (9,10,13).
Given that peripheral white blood cells contain lymphocyte subsets that derive from lymphocytic stem cells and immunologically active tissue, we hypothesized that telomere length in peripheral white blood cell DNA might be particularly informative with regard to risk of developing non-Hodgkin lymphoma (NHL). To avoid potential disease bias, we carried out our study within a prospective cohort, the Alpha-Tocopherol, Beta-Carotene Cancer Prevention (ATBC) Study (14). In addition, we applied a modification of the original PCR-based assay developed by Cawthon (15), referred to as monochrome multiplex quantitative PCR (16), which has high assay precision and is almost perfectly correlated (Spearman r = 0.91, p < 0.0001) with the Southern blot method of measuring telomere length.
Details of the ATBC cohort study design have been described previously (17). Briefly, 29,133 male smokers, aged 50-69 years, were recruited from southwest Finland and randomized from 1985 to 1988. Subjects were provided α-tocopherol, β-carotene, both, or placebo. The study was approved by Institutional Review Boards at the National Cancer Institute and the National Public Health Institute of Finland. Participants provided written informed consent.
Incident NHL cases were identified using the Finnish Cancer Registry, which provides nearly 100% of case ascertainment in Finland (18). The hospital records of the identified NHL cases were reviewed by an expert study oncologist for confirmation of lymphoma diagnosis and histology. Through April 30, 2002, incident cases of NHL (n = 107) diagnosed after providing a whole blood sample in 1992 or 1993 were identified based on the histology information coded in the International Classification of Disease-Oncology second edition (ICD-O-2: 9590-9595, 9670-9677, 9680-9688, 9690-9698, 9700-9717, 9760-9764, 9820-9828, 9940-9941) (19). Cases were further grouped into NHL subtypes according to the World Health Organization guidelines (19). Controls were selected from the ATBC study participants who were alive, free of cancer at the time of the case diagnosis, and were individually matched to cases on date of birth (±5 years). DNA was extracted from whole blood samples using the phenol-chloroform method, and a monochrome multiplex quantitative PCR assay was used to determine the telomere measurements (16).
In brief, the reagents in the 25 microliter PCR were 10 mM Tris–HCl pH 8.3, 50 mM KCl, 3 mM MgCl2, 0.2 mM each dNTP, 1 mM DTT, 1M betaine, 0.75× SYBR Green I, and AmpliTaq Gold DNA polymerase, 0.625 U. The four primers were (5′ to 3′): telg (at 900 nM), ACACTAAGGTTTGGGTTTGGGTTTGGGTTTGGGTTAGTGT; telc (at 900 nM), TGTTAGGTATCCCTATCCCTATCCCTATCCCTATCCCTAACA; hbgu (at 500 nM), CGGCGGCGGGCGGCGCGGGCTGGGCGGCTTCATCCACGTTCACCTTG; and hbgd (at 500 nM), GCCCGGCCCGCCGCGCCCGTCCCGCCGGAGGAGAAGTCTGCCGTT. From 5 to 70 ng of human genomic DNA were added per reaction well. Three-fold serial dilutions of a reference genomic DNA sample were used to generate two standard curves for each PCR plate (five concentrations with a high of 150 ng per reaction, and a low of 1.85 ng per reaction). Thermal cycling: 1 cycle of 15 min at 95°C; 2 cycles of 15 s at 94°C, 15 s at 49°C; 32 cycles of 15 s at 94°C, 10 s at 62°C, 15 s at 74°C with signal acquisition, 10 s at 84°C, and 15 s at 88°C with signal acquisition. The 74°C reads provided the Cts for telomeres; the 88°C reads provided the Cts for the single copy gene (beta-globin). After the run was complete, the MyiQ software (Bio-Rad iQ5 2.0 Standard Edition Optical System Software) was used to determine the T (telomere) and S (single copy gene) values for each experimental sample by the Standard Curve method.
Figure 1 presents the principle of telomere length measurement by quantitative polymerase chain reaction. The ratio of the telomere PCR signal to the single copy gene (in our case, beta-globin) PCR signal (i.e. the T/S ratio) is proportional to the average telomere length per cell. All T/S ratios of experimental DNA samples are expressed relative to the T/S ratio of the same reference DNA sample, which, by definition is assigned a T/S ratio of 1.0. For a given experimental sample, the T value is the number of ng of the reference DNA that matches the experimental sample for copy number of the telomere template, and the S value is the number of ng of the reference DNA that matches that experimental sample for copy number of the single copy gene template. T/S, therefore, is a relative and dimensionless value. Samples with a T/S > 1.0 have an average telomere length greater than that of the Standard DNA; samples with a T/S < 1.0 have an average telomere length shorter than that of the Standard DNA. Multiplex QPCR eliminates a major source of variation present in monoplex QPCR. In monoplex QPCR, variation in the amount of DNA pipetted into the T and S reaction wells results in variation in T/S; but in multiplex QPCR both T and S are measured in each reaction well, so the pipetting variation between wells does not affect T/S.
Cases and their matched controls were assayed consecutively within each batch. Blinded quality control samples were interspersed across batches to evaluate assay reproducibility, and samples were analyzed 1-2 times. The assay intra-class correlation coefficient was 80% and the coefficient of variation was 7%. Odds ratios (OR) and 95% confidence intervals (95% CI) were estimated using conditional logistic regression models. Telomere length was categorized into quartiles based on the distribution among controls. Tests for trend were calculated using the median value for each telomere length quartile. Variables that resulted in a ≥10% change in the beta-coefficient of telomere length in the base model that adjusted for matched age were considered confounders and included in final, multivariable models. Age at randomization, body mass index, smoking, physical activity, dietary intake of alcohol, blood pressure, and mitochondrial DNA copy number were not confounders in our sample. All p-values are two-sided.
Cases were comparable to control subjects (Table 1). Telomere length was very weakly and inversely correlated with age (Spearman correlation r = -0.07, p = 0.46), and weakly and positively correlated with pack-years of smoking (Spearman correlation, r = 0.12, p = 0.20) among controls, which would be expected given the relatively narrow age range of this older cohort of all tobacco smokers.
Telomere length was statistically significantly longer among cases than controls [median (10th, 90th percentile): 1.10 (0.79, 1.43) in cases and 1.02 (0.78, 1.26) in controls, Wilcoxon sign test p = 0.0017, Table 1]. The risk of NHL was significantly increased with longer telomere length, compared to the lowest quartile of telomere length (p trend = 0.003), and the association was consistent across the predominant NHL subtypes in this series of cases (Table 2). Adjustment for demographic factors shown in Table 1 had a negligible impact on the results (data not shown).
To determine if the association might be driven in part by longer telomere length among cases undiagnosed at the time of blood sample collection, we excluded cases diagnosed within the first year of follow-up (n = 17) after blood sample collection, and found that results were very similar [OR (95% CI): 0.8 (0.3-3.1), 1.5 (0.5-4.1), and 3.1 (1.2-8.2), for second, third and fourth quartiles of telomere length, compared to the lowest quartile of telomere length, p trend = 0.01].
To evaluate the potential effects of the trial vitamin supplementation on the relationship between telomere length and risk of NHL, we carried out further analyses stratified by α-tocopherol vs. no α-tocopherol supplementation, and β-carotene vs. no β-carotene supplementation. Risks were similar in each group, and tests for interactions were not statistically significant (data not shown).
To the best of our knowledge, this is the first prospective study addressing the relationship between telomere length and risk of NHL. Although the sample size is modest, we detected relatively strong effects, which were consistent across major NHL subtypes. One previous report of telomere length and NHL by Widmann et al. of 40 cases and 40 controls using a case-control design found that shorter telomere length was associated with increased risk of aggressive NHL (7). The reasons for this discrepant result with our study are not immediately apparent. Given that relatively shorter telomere length was present in DNA extracted from B cells, T cells, and granulocytes in cases vs. controls in the initial report (7), it would be expected that measuring telomere length in an aggregated “buffy coat” containing all peripheral leukocytes, as we did in our study, would give similar results. Widmann et al. measured telomere length by the Flow-Fish method whereas we used a new PCR-based method, which correlates almost perfectly with the Southern blot assay of telomere length. The correlation between the assay used in our report and the Flow-Fish method is not known. The prospective nature of our study is another potential difference between the two reports.
It is possible that tumor cells were present in the peripheral blood of undiagnosed patients in our study, particularly those with the more indolent NHL histologies. However, there is much evidence that telomere length is shorter in lymphoma cells (20-22). As a consequence, if tumor cells were circulating in the blood of undiagnosed patients, the average telomere length of their peripheral white blood cell DNA would tend to be shorter, not longer. This would have biased our results towards the null, rather than create the relatively strong association we report with longer telomere length. At the same time, given that we observed similar associations between longer telomere length and risk of NHL for a relatively indolent (i.e., CLL/SLL) and a more aggressive histology (i.e., DLBCL) (Table 2), it is unlikely that our results have been attenuated by this potential bias. Therefore, our overall interpretation of these results is that in a healthy population there is a normal range of variation in the average white blood cell telomere length between individuals, and those individuals with relatively long telomeres are at higher risk of later developing lymphoma.
Clearly, telomere shortening is an early and frequently observed finding in malignant transformation (23). The question is to what extent would the tendency to have relatively shorter or longer telomeres in normal tissue, which appears to reflect both genetic and environmental factors (24), be associated with future risk of developing cancer in general, and of developing specific tumor types in particular. From a theoretical perspective, one can argue that the tendency to have shorter or longer telomeres could each contribute to carcinogenesis (3). Optimal telomere length is a balance of cell proliferation, senescence, and control. Shorter telomeres may result in a greater tendency towards chromosomal instability, which could reflect a constitutional or acquired tendency towards carcinogenesis (5,6). Alternatively, given that telomeres become shorter with cell proliferation, ultimately triggering senescence or apoptosis, it is possible that the tendency to have cells with shorter telomeres may have an advantage for suppressing tumorigenesis when their telomeres reach a critical minimal length. It follows that cells with longer telomeres may favor a delayed senescence (3), and such cells could have more opportunity to acquire genetic abnormalities and be at higher risk of transformation.
In addition, the relationship between telomere length and risk for cancer may vary by cell type. Recently, Han et al., reported that shorter telomere length measured in buffy coats in the prospective Nurses' Health Study (25) by the original PCR-based assay developed by Cawthon (16), which strongly correlates with the new monochrome quantitative multiplex PCR used in this report (Spearman r = 0.83, p < 0.0001), was associated with a decreased number of moles and risk of melanoma and an increased risk of basal cell cancer (9).
In summary, in this prospective cohort, we found that longer telomere length was associated with increased risk of NHL. These findings require replication in larger studies that avoid or carefully address potential disease bias and that incorporate comparable methods to assess telomere length. Overall, our findings suggest that the relationship between telomere length measured in healthy tissue and risk of cancer may be complex and vary by disease and possibly study design.
This work was supported by intramural funds from the National Cancer Institute. We thank Jackie King and the other members of the BioReliance BioRepository (Rockville, MD) for blood sample handling, storage, and shipping, and for assisting with laboratory analysis monitoring.
Translational relevance: We report that longer telomere length measured in peripheral white blood cell DNA was associated with future risk of developing non-Hodgkin lymphoma. This work contributes to a growing number of findings that biomarker assays that measure genomic damage and stability in peripheral white blood cells of healthy people (e.g., telomere length, global methylation, alteration in mitochondrial DNA copy number) can identify individuals who are at higher risk of developing certain types of cancer in the future. The translational implications of our findings, and this overall area of research, is that the general population will eventually be screened for not only inherited variation in genes, but for acquired alterations in DNA as well, that together will be powerfully predictive of future cancer risk. This will identify individuals who could benefit from various preventive strategies (e.g., focused carcinogen avoidance, chemoprevention) as well as targeted early disease detection.