Short telomere lengths in peripheral blood leukocytes are associated with an increased risk of oral premalignant lesion and oral squamous cell carcinoma

Authors


Abstract

BACKGROUND

Oral premalignant lesions (OPLs) are precursors of oral squamous cell carcinoma (OSCC). Short telomeres in peripheral blood leukocytes (PBLs) are associated with increased risks of several cancers. However, it is unclear whether short leukocyte telomere length (LTL) predisposes individuals to OPL and OSCC.

METHODS

LTL was measured in PBLs from 266 patients who had a diagnosis of either OPL (N = 174) or OSCC (N = 92) and from 394 age-matched and sex-matched controls. The association between LTL and the risk of OPL or OSCC, as well as the interaction of telomere length, cigarette smoking, and alcohol drinking on the risk of OPL or OSCC, were analyzed.

RESULTS

The age-adjusted relative LTL was shortest in the OSCC group (1.64 ± 0.29), intermediate in the OPL group (1.75 ± 0.43), and longest in the control group (1.82 ± 0.36; Ptrend < .001). When the analysis was dichotomized at the median value in controls, adjusting for age, sex, smoking status, and alcohol drinking status, the odds ratio for the risk of OPL and OSCC associated with short LTL was 2.03 (95% confidence interval, 1.29-3.21) and 3.47 (95% CI, 1.84-6.53), respectively, with significant dose-response effects for both associations. Among 174 patients with OPL, 23 progressed to OSCC, and the mean LTL was shorter in progressors than in nonprogressors (mean ± standard deviation: 1.66 ± 0.35 vs 1.77 ± 0.44, respectively), although the difference did not reach statistical significance (P = .258), probably because of the small number of progressors. An interaction analysis identified short LTL, smoking, and drinking alcohol as independent risk factors for OPL and OSCC.

CONCLUSIONS

Short LTL was associated with increased risks of developing OPL and OSCC. The current results also indicated that short LTL likely predisposes patients to the malignant progression of OPL. Cancer 2013;119:4277–4283. © 2013 American Cancer Society.

INTRODUCTION

Oral leukoplakia, oral submucous fibrosis, and erythroplakia are the 3 major forms of oral premalignant lesions (OPLs), and oral leukoplakia accounts for 85% of all OPLs.[1, 2] Individuals with these lesions are at an elevated risk for developing oral squamous cell carcinoma (OSCC).[3-6] Lee et al reported that 31.4% of patients with OPL developed cancers in their upper aerodigestive tracts during 7 years of follow-up.[6] The overall malignant transformation rate of dysplastic lesions depends on the length of follow-up and varies from 11% to 36%.[5] Tobacco chewing, tobacco smoking, and alcohol drinking have been identified as major epidemiologic risk factors for OPLs and OSCC.[7] Association studies on genetic risk factors have been emerging, and the results suggest that genetic variants play an important role in OPL etiology.[8-10]

The telomeres are the extreme ends of each chromosome and protect them from degradation and end-to-end fusion.[11] Human telomeres are progressively shortened with each cell division and, thus, vary with age and cell type, with length ranging from 6 to 12 kilobases (kb) in somatic cells.[12] Oxidative damage and loss of telomere-binding proteins can also contribute to telomere shortening.[13, 14] The shortening can be compensated by telomerase, which is constitutively expressed in germ-line cells and in most malignant cells.[15]

Critically short telomeres become dysfunctional, and previous studies indicated an association between telomere dysfunction and the initiation and progression of malignancies in knockout mouse models[16, 17] and in human cancers.[18, 19] Constitutive short telomere length in peripheral blood leukocytes also has been associated with several human cancers.[19-25] A recent study demonstrated that OPL tissues had significantly shorter telomeres than normal epithelium.[26] However, to our knowledge, no study to date has evaluated the association of leukocyte telomere length (LTL) with the risk of OPL or OSCC. In this study, our objective was to investigate the association between LTL and the risk of OPL and OSCC using a case-control study design. We also performed an exploratory analysis of the association of LTL with OPL progression using a prospective study design.

MATERIALS AND METHODS

Study Population and Epidemiologic Data

In total, 266 patients with OPL (N = 174) or OSCC (N = 92) were included in this study. All patients were aged ≥18 years and had histologically confirmed OPL (leukoplakia and/or erythroplakia) or OSCC, as described previously.[27] Patients with acute intercurrent illnesses or infections and patients with a previous history of cancer (except nonmelanoma skin cancer) were excluded. Detailed clinical and follow-up information was abstracted from medical charts. A self-administered questionnaire was used to collect epidemiologic data, including demographic characteristics and history of tobacco and alcohol use. All patients donated a blood sample in heparnized tubes for molecular analyses before any treatment.

Healthy controls were identified from a pool of healthy individuals who were recruited in an ongoing case-control study[27] from the Kelsey-Seybold Clinic, the largest private multispecialty group practice in the Houston metropolitan area, with 18 clinics and more than 325 physicians. The majority of control participants were healthy individuals who attended the clinic for annual physical examinations. On the day of the interview, the controls visited the clinic specifically for the purpose of participating in this study and not for treatment. Epidemiologic questionnaire data were collected, including demographic characteristics, tobacco use history, family history of cancer, occupational and environmental exposures, alcohol drinking habits, and medical history. A blood sample was collected from each participant into a heparinized tube and was sent to the laboratory for molecular analysis. Control participants had no prior history of cancer (except for nonmelanoma skin cancer) and were frequency matched to the OPL patients on age (±5 years), sex, and ethnicity. In total, 394 healthy controls were identified and were included in the current analysis.

For both cases and controls, written informed consent was obtained for participating in the study and donation of blood samples. The study was approved by the institutional review boards of The University of Texas MD Anderson Cancer Center and the Kelsey Seybold Clinics. An individual who had never smoked or who had smoked <100 cigarettes in his or her lifetime was defined as a never smoker. A former smoker was defined as an individual who had quit smoking at least 1 year before diagnosis (cases) or before the interview (controls). A current smoker was defined as someone who was currently smoking or who had stopped smoking less than 1 year before diagnosis (cases) or before the interview (controls). Individuals who had never consumed alcohol or who consumed no more than 1 drink per month were defined as never drinkers, and those who had consumed more than 1 drink per month were defined as ever drinkers (1 bottle or can of beer, 1 medium glass of wine, 1 straight shot, or 1 mixed drink was defined as 1 drink).

Overall Leukocyte Telomere Length Assessment by Real-Time Polymerase Chain Reaction Analysis

Genomic DNA was extracted from whole blood using the QIAamp Maxi DNA kit (Qiagen Inc., Valencia, Calif) according to the manufacturer's protocol. The relative overall LTL was measured using the modified version of real-time quantitative polymerase chain reaction (PCR) described by Cawthon.[28] The ratio of the telomere repeat copy number (T) to the single gene (human globulin) copy number (S) was determined for each sample. The derived T/S ratio was proportional to the overall telomere length.

The PCR (15 μL) for telomere amplification consisted of 1 × SYBR Green Mastermix (Applied Biosystems Inc., Foster City, Calif), 200 nmol/L Tel-1, 200 nmol/L Tel-2, and 5 ng of genomic DNA. The PCR for human globulin (Hgb) amplification consisted of 1 × SYBR Green Mastermix, 200 nmol/L Hgb-1, 200 nmol/L Hgb-2, and 5 ng of genomic DNA. The thermal cycling conditions were at 95°C for 10 minute followed by 40 cycles at 95°C for 15 seconds and at 56°C (for telomere amplification) or 58°C (for Hgb amplification) for 1 minute. The telomere and Hgb PCRs were done on separate 384-well plates with the same samples in the same well positions. In each run, negative and positive controls, a calibrator DNA, and a standard curve were included. The positive controls contained a telomere of 1.2 kb and a telomere of 3.9 kb from a commercial telomere length assay kit (Roche Applied Science, Pleasanton, Calif). For each standard curve, 1 reference DNA sample (the same DNA sample for all runs) was diluted 2-fold serially to produce a 6-point standard curve between 20 ng and 0.625 ng of DNA in each reaction. The coefficient of determination (R2) for each standard curve was ≥0.99, with acceptable standard deviations set at 0.25 (for the Ct values). If the result was within the acceptable range, then the sample was repeated. The intra-assay coefficient of variation was <3%, and the interassay coefficient of variation was <5% for this assay in our laboratory.[23-25]

Statistical Analysis

All statistical analyses were performed using STATA statistical software (version 10.0; STATA Corporation, College Station, Tex). The Pearson chi-square test was used to test differences in the distribution of host characteristics between cases and controls for categorical variables, and the Student t test was used to test differences for continuous variables. The association between OPL risk and telomere length was assessed using unconditional multivariate logistic regression to estimate the adjusted odds ratio (aOR) and 95% confidence interval (CI). Analyses were adjusted for age, sex, smoking status, and alcohol drinking status, as appropriate. We then added an interaction term to the logistic regression models to test the interaction between telomere length and smoking (or alcohol drinking) in elevating the risk of OPL (or OSCC). All statistical tests were 2-sided, and the level of statistical significance was set at .05.

RESULTS

Cases and controls were matched in terms of age (P = .373) and sex (P = .894). The vast majority of patients and controls were white. There were significant differences in the distribution of smoking status between the patients with OPL/OSCC (cases) and the control group, with higher percentages of current and former smokers among cases than among controls (P = .004). Also, more cases were alcohol drinkers than controls (P < .001). The age-adjusted relative LTL was shortest in patients with OSCC (1.64 ± 0.29), intermediate in patients with OPL (1.75 ± 0.43), and longest in controls (1.82 ± 0.36; Ptrend < .001) (Table 1).

Table 1. Distribution of Demographic Data in Controls, Patients With Oral Premalignant Lesions, and Patients with Oral Squamous Cell Carcinoma
 No. (%) 
VariableControl GroupOPL GroupOSCC GroupPa
  1. Abbreviations: OPL, oral premalignant lesions; OSCC, oral squamous cell carcinoma; SD, standard deviation.

  2. a

    P values were derived from the Pearson chi-square test for sex, ethnicity, smoking, and alcohol drinking status; the Student t test was used for age; and the trend test was used for telomere length. All P values were 2-sided.

Age: Mean ± SD, y58.30 ± 11.1357.38 ± 11.6157.22 ± 13.85.373
Sex    
Men226 (57.36)99 (56.90)55 (59.78) 
Women168 (42.64)75 (43.10)37 (40.22).894
Ethnicity    
White356 (90.36)152 (87.36)89 (96.74) 
Hispanic17 (4.31)7 (4.02)0 (0) 
Black14 (3.55)4 (2.30)1 (1.09) 
Others7 (1.78)11 (6.32)2 (2.17).022
Smoking status    
Never207 (52.54)61 (40.40)35 (42.68) 
Former148 (37.56)58 (38.41)36 (43.90) 
Current39 (9.90)32 (21.19)11 (13.41).004
Alcohol drinking status    
Never251 (69.72)41 (27.70)17 (22.67) 
Ever109 (30.28)107 (72.30)58 (77.33)< .001
Relative telomere length: Mean ± SD1.82 ± 0.361.75 ± 0.431.64 ± 0.30< .001

When telomere length was dichotomized at the median value in controls, the aOR for OPL associated with shorter telomere length was 2.03 (95% CI, 1.29-3.21) after adjusting for age, sex, and smoking and alcohol drinking status. In quartile analysis in which the cutoff points were set using the quartile values in controls, compared with the fourth quartile, which had the longest telomere length, the aORs were 0.99 (95% CI, 0.52-1.90) for the third quartile, 1.83 (95% CI, 0.95-3.53) for the second quartile, and 2.19 (95% CI, 1.18-4.06) for the first quartile (Ptrend = .004) (Table 2, top). For OSCC, short LTL was associated with a 3.47-fold increased risk of OSCC (95% CI, 1.84-6.53). In quartile analysis, compared with the fourth quartile, which had the longest telomere length, the aORs were 1.21 (95% CI, 0.45-3.25) for the third quartile, 2.69 (95% CI, 1.06-6.87) for the second quartile, and 4.92 (95% CI, 2.04-11.8) for the first quartile (Ptrend < .001) (Table 2, bottom).

Table 2. Risk Estimates for Telomere Length and Oral Premalignant Lesions or Oral Squamous Cell Carcinoma
 No. (%)  
Variable StrataControl GroupOPL GroupAdjusted OR [95% CI]bP
  1. Abbreviations: CI, confidence interval; OPL, oral premalignant lesions; OR, odds ratio; OSCC, oral squamous cell carcinoma; Ref, referent category.

  2. a

    Long and short groups were dichotomized at the medium telomere length in the controls.

  3. b

    Analyses were adjusted by age, sex, ethnicity, smoking, and alcohol drinking status, as appropriate.

Overall telomere lengtha    
Long197 (75.19)65 (24.81)1.00 [Ref] 
Short197 (64.38)109 (35.62)2.03 [1.29–3.21].002
Quartile    
Fourth quartile98 (72.06)38 (27.94)1.00 [Ref] 
Third quartile99 (78.57)27 (21.43)0.99 [0.52–1.9].983
Second quartile98 (68.53)45 (31.47)1.83 [0.95–3.53].073
First quartile99 (60.74)64 (39.26)2.19 [1.18–4.06].013
Ptrend   < .001
     
 No. (%)  
Variable StrataControl GroupOSCC GroupAdjusted OR [95% CI]bP
Overall telomere lengtha    
Long197 (88.74)25 (11.26)1.00 [Ref] 
Short197 (74.62)67 (25.38)3.47 [1.84–6.53]< .001
Quartile    
Fourth quartile98 (90.74)10 (9.26)1.00 [Ref] 
Third quartile99 (86.84)15 (13.16)1.21 [0.45–3.25].705
Second quartile98 (79.67)25 (20.33)2.69 [1.06–6.87].038
First quartile99 (70.21)42 (29.79)4.92 [2.04–11.8]< .001
Ptrend   < .001

We further evaluated whether there were interactions between LTL and cigarette smoking or alcohol drinking by introducing an interaction term into the logistic regression models. Compared with never smokers, who had long telomeres, the risk of OPL was 2.13 (95% CI, 1.10-4.12) for never smokers with short LTL, 2.44 (95% CI, 1.25-4.80) for ever smokers with long LTL, and 4.91 (95% CI, 2.57-9.36) for ever smokers with short LTL (Pinteraction = .896). The risk of OSCC was 4.51 (95% CI, 1.84-11.08) for never smokers with short LTL, 3.67 (95% CI, 1.32-10.21) for ever smokers with long LTL, and 9.90 (95% CI, 3.82-25.66) for ever smokers with short LTL (Pinteraction = .406) (Table 3). There was no significant interaction between LTL and smoking in elevating the risk of OPL or OSCC or between LTL and alcohol drinking in elevating the risk of OPL or OSCC risk (data not shown). It appeared that LTL, smoking, and alcohol drinking were independent risk factors for OPL and OSCC.

Table 3. Interaction of Telomere Length and Cigarette Smoking on the Risk of Oral Premalignant Lesions and Oral Squamous Cell Carcinoma
  No. (%)   
Telomere LengthaSmoking StatusControl GroupOPL GroupAdjusted OR [95% CI]bPPInteraction
  1. Abbreviations: CI, confidence interval; OPL, oral premalignant lesions; OR, odds ratio; OSCC, oral squamous cell carcinoma; Ref, referent category.

  2. a

    Long and short groups were dichotomized at the median telomere length in the controls.

  3. b

    Analyses were adjusted by age, sex, ethnicity, and alcohol drinking status.

LongNever108 (78.26)30 (21.74)1.00 [Ref]  
ShortNever99 (76.15)31 (23.85)2.13 [1.10–4.12].025 
LongEver89 (74.79)30 (25.21)2.44 [1.25–4.80].009 
ShortEver98 (62.03)60 (37.97)4.91 [2.57–9.36]< .001.896
       
  No. (%)   
Telomere LengthaSmoking StatusControl GroupOSCC GroupAdjusted OR [95% CI]bPPInteraction
LongNever108 (91.53)10 (8.47)1.00 [Ref]  
ShortNever99 (79.84)25 (20.16)4.51 [1.84–11.08].001 
LongEver89 (88.12)12 (11.88)3.67 [1.32–10.21].013 
ShortEver98 (73.68)35 (26.32)9.90 [3.82–25.66]< .001.406

Next, we examined the combined effects of having multiple risk factors. Compared with individuals who did not have any of the 3 risk factors, those who had all 3 risk factors (LTL, smoking, and alcohol drinking) exhibited a 27-fold increased risk of OPL (OR, 27.37; 95% CI, 10.12-73.98) and a 35-fold increased risk of OSCC (OR, 35.24; 95% CI, 10.16-122.24) (Table 4).

Table 4. Cumulative Effect of Telomere Length, Alcohol Drinking, and Cigarette Smoking on the Risk of Oral Premalignant Lesions and Oral Squamous Cell Carcinoma
 No. (%)  
No. of Risk FactorsaControl GroupOPL GroupAdjusted OR [95% CI]bP
  1. Abbreviations: CI, confidence interval; OPL, oral premalignant lesions; OR, odds ratio; OSCC, oral squamous cell carcinoma; Ref, referent category.

  2. a

    Risk factors were short telomeres, ever smoking, and ever drinking.

  3. b

    Analyses were adjusted by age, sex, and ethnicity.

048 (84.21)9 (15.79)1.00 [Ref] 
1179 (82.49)38 (17.51)1.21 [0.54–2.72].636
2121 (67.98)57 (32.02)3.08 [1.38–6.85].006
312 (21.43)44 (78.57)27.37 [10.12–73.98]< .001
Ptrend   < .001
     
 No. (%)  
No. of Risk FactorsaControl GroupOSCC GroupAdjusted OR [95% CI]bP
048 (90.57)5 (9.43)1.00 [Ref] 
1179 (94.21)11 (5.79)0.68 [0.22–2.12].504
2121 (78.57)33 (21.43)3.63 [1.29–10.26].015
312 (31.58)26 (68.42)35.24 [10.16–122.24]< .001
Ptrend   < .001

Finally, we explored whether short LTL at baseline could predict future OPL progression and death in patients with OSCC. Among 174 patients with OPL, 23 progressed to OSCC during follow-up. The mean LTL was shorter in progressors than in nonprogressors (1.66 ± 0.35 vs 1.77 ± 0.44), although the difference did not reach statistical significance (P = .258), probably because of the small number of progressors. The mean LTL in progressors (1.66 ± 0.35) at baseline was similar to that in the patients with OSCC (1.64 ± 0.29). Among 92 patients with OSCC, 20 died, and their baseline mean LTL was 1.59 ± 0.24 compared with 1.65 ± 0.31 in the 72 patients with OSCC who remained alive. Again, the difference did not reach statistical significance (P = .393) because of the limited number of deaths.

DISCUSSION

Our current data demonstrate that short LTL is associated with an increased risk of developing OPL and OSCC. Our data also suggest that patients with OPL who have short LTL are at an increased risk of progressing to OSCC and that patients who have OSCC with short LTL are at an increased risk of death.

There have been many reports suggesting that LTL is associated with cancer risk in a cancer type-dependent manner: both short and long telomeres can confer increased cancer risks.[20-22, 24, 25] A recent large cohort study of 47,102 individuals demonstrated that, when all cancer types were combined, LTL was not associated with the risk of cancer.[22] However, because different cancers have different etiology, pooling all cancer types together may mask the significant associations of LTL with individual cancer types. There have been consistent data indicating that, in some cancers (eg, melanoma[22, 29, 30] and sarcoma[22, 25]), longer LTLs conferred increased cancer risks; whereas, in many others, shorter LTLs conferred increased cancer risks.[20-22] When pooling data for all cancer sites, the opposite associations cancel each other out, resulting in a null finding for the overall cancer risk.[31]

To our knowledge, no previous studies have specifically examined the association of LTL with the risk of OSCC or OPL. However, in the aforementioned large cohort study,[22] cancers of the “oral cavity and pharynx” were included as a specific type. With 76 patients, the observed OR for short LTL with this cancer was 1.17 (95% CI, 0.90-1.53), consistent with our current study and supporting the finding that short telomere length in peripheral blood leukocytes is associated with increased risks of OPL and OSCC.

In humans, telomeres are maintained in germline cells but shorten as somatic cells divide because of the down-regulation of telomerase. Telomere shortening limits the replication of somatic cells. Cancer cells invariably maintain telomere length through the expression of telomerase. Excessive telomere shortening before the expression of telomerase can lead to chromosomal fusion, which has been proposed as a mechanism for chromosome instability.[32] It was reported previously that the mean cytoplasmic human telomerase reverse transcriptase increased from normal, through oral epithelial dysplasia, to oral squamous carcinoma.[33] Markedly higher expression levels of human telomerase reverse transcriptase were reported in oral dysplasia compared with normal tissues,[34] suggesting that telomere length shortening is an early event during oral carcinogenesis. With regard to the potential mechanisms underlying the association of short LTL with increased risks of OPL and OSCC, we hypothesize that shorter LTL indicates higher genetic instability, leading to elevated risks. We previously demonstrated that LTL was inversely correlated with baseline and mutagen-induced genetic instability in lymphocytes,[20] supporting this hypothesis. It should be noted that regulation of the telomere length in mammalian cells may be chromosome specific,[35, 36] and previous studies have identified telomere shortening in specific chromosomes associated with the risk of different cancers.[23, 37] Thus, further investigation into the role of chromosome-specific telomere length in the risk of OPL and OSCC is a promising future direction.

Cigarette smoking and alcohol drinking are 2 major risk factors of oral cancer.[38-40] We did not observe a significant interaction between short LTL and these 2 risk factors in elevating the risk of OPL and OSCC. It appears that LTL is an independent risk factor for OPL and OSCC. However, individuals in our study who had 3 risk factors (short LTL, smoking, and alcohol drinking) exhibited a 27-fold increased risk of OPL and a 35-fold increased risk of OSCC. These data suggest that adding LTL as a biomarker to an environmental risk factor profile can help identify high-risk individuals. The findings also point to the importance of preventing LTL attrition and stopping smoking and alcohol drinking in reducing the risks of OPL and OSCC.

In our exploratory, prospective analysis, we observed that patients with OPL who later progressed to OSCC had baseline LTL similar to that of patients with OSCC but shorter than that of those who did not progress. We also observed that patients with OSCC who died had shorter baseline LTL than those who remained alive. Because of the limited numbers of progression and death events, these 2 comparisons did not reach statistical significance. However, the data suggest that baseline LTL can be a valuable predictor of future malignant progression in patients with OPL and can predict a worse prognosis in patients with OSCC. Future large, prospective studies are warranted to confirm these observations.

In conclusion, to our knowledge, the current study is the first molecular epidemiologic study of telomere length in peripheral blood leukocytes and its association with OPL susceptibility and progression. Intervention strategies aimed at preventing or reversing telomere shortening, together with ceasing the behaviors of smoking and alcohol drinking, may be effective in lowering the incidence of oral cancer.

FUNDING SUPPORT

This work was supported by a grant from the National Cancer Institute (Specialized Center of Research Excellence [SPORE] grant P50 CA097007) to The University of Texas MD Anderson Cancer Center (Scott M. Lippman, principal investigator; Xifeng Wu, project leader), the MD Anderson Cancer Center Research Trust (to Xifeng Wu); and the MD Anderson Cancer Center Start-Up Fund (to Jian Gu).

CONFLICT OF INTEREST DISCLOSURES

The authors made no disclosures.

Ancillary