Skip to main content
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2013 Mar 5.
Published in final edited form as: Am J Gastroenterol. 2012 Feb 21;107(5):782–789. doi: 10.1038/ajg.2012.21

Cigarette Smoking and Colorectal Cancer Risk by KRAS Mutation Status Among Older Women

NJ Samadder 1, Robert A Vierkant 2, Lori S Tillmans 3, Alice H Wang 2, Charles F Lynch 4, Kristin E Anderson 5, Amy J French 3, Robert W Haile 6, Lisa J Harnack 7, John D Potter 8, Susan L Slager 2, Thomas C Smyrk 3, Stephen N Thibodeau 3, James R Cerhan 9, Paul J Limburg 1
PMCID: PMC3588167  NIHMSID: NIHMS440197  PMID: 22349355

Abstract

OBJECTIVES

Existing data support a modest association between cigarette smoking and incident colorectal cancer (CRC) overall. In this study, we evaluated associations between cigarette smoking and CRC risk stratified by KRAS mutation status, using data and tissue resources from the Iowa Women’s Health Study (IWHS).

METHODS

The IWHS is a population-based cohort study of cancer incidence among 41,836 randomly selected Iowa women, ages 55–69 years of age at enrollment (1986). Exposure data, including cigarette smoking, were obtained by self-report at baseline. Incident CRCs (n = 1,233) were ascertained by annual linkage with the Iowa Cancer Registry. Archived tissue specimens from CRC cases recorded through 2002 were recently requested for molecular epidemiology investigations. Tumor KRAS mutation status was determined by direct sequencing of exon 2, with informative results in 507/555 (91%) available CRC cases (342 mutation negative and 165 mutation positive). Multivariate Cox regression models were fit to estimate relative risks (RRs) and 95% confidence intervals (CIs) for associations between cigarette smoking variables and KRAS-defined CRC subtypes.

RESULTS

Multiple smoking variables were associated with increased risk for KRAS mutation-negative tumors, including age at initiation (P = 0.02), average number of cigarettes per day (P = 0.01), cumulative pack-years (P = 0.05), and induction period (P = 0.04), with the highest point estimate observed for women who smoked ≥ 40 cigarettes per day on average (RR = 2.38; 95% CI = 1.25–4.51; compared with never smokers). Further consideration of CRC subsite suggested that cigarette smoking may be a stronger risk factor for KRAS mutation-negative tumors located in the proximal colon than in the distal colorectum. None of the smoking variables were significantly associated with KRAS mutation-positive CRCs (overall or stratified by anatomic subsite).

CONCLUSIONS

Data from this prospective study of older women demonstrate differential associations between cigarette smoking and CRC subtypes defined by KRAS mutation status, and are consistent with the hypothesis that smoking adversely affects the serrated pathway of colorectal carcinogenesis.

INTRODUCTION

Colorectal cancer (CRC) is the fourth most common incident and second most common fatal malignancy in the United States, with ~140,000 new diagnoses recorded and nearly 50,000 deaths attributed to CRC each year (1). Given this burden of disease, further understanding of common, potentially modifiable CRC risk factors represents an important public health priority. Despite widespread tobacco control efforts, about one in five US adults still characterize themselves as active cigarette smokers (2). Pooled analyses of data from numerous observational studies demonstrate increased risks for both benign and malignant colorectal neoplasia among cigarette smokers, as compared with non-smokers (3,4). Interestingly, our group (5) and others (615) have reported that cigarette smoking appears to be associated with distinct, molecularly defined CRC subtypes, although the spectrum of genetic alterations and epigenetic modifications involved in smoking-related CRC risk remains incompletely described.

The KRAS oncogene has been commonly implicated in colorectal carcinogenesis, with somatic mutations identified in 30–40% of sporadic CRCs (16,17). Tumor testing for KRAS mutations has been endorsed as an adjunct to chemotherapy planning (specifically, to inform the addition of anti-epidermal growth factor receptor agents, such as cetuximab or panitumumab, among patients with metastatic CRC) (18,19), emphasizing the relevance of this molecular marker in clinical practice. Since KRAS mutations are thought to occur at the early adenoma stage (17), it seems biologically plausible that exposures associated with both invasive and preinvasive disease might differentially modulate CRC risks based on the KRAS mutation status. Laboratory experiments have also shown that carcinogens found in tobacco smoke can induce cancer-related base substitutions, such as G:C→A:T transitions, in ras oncogenes (16,20). However, to date, relatively few epidemiological studies have examined associations between cigarette smoking and KRAS-defined CRC risks (913), including only one previous report from a prospective, population-based cohort study (13).

For the current report, we utilized data and tissue resources from the Iowa Women’s Health Study (IWHS), a prospective cohort study of cancer end points among older women, to examine associations between smoking habits and incident CRC by KRAS mutation status (negative or positive). KRAS-defined CRC risks were further evaluated with respect to anatomic subsite (proximal colon and distal colorectum), as another potential indicator of subtype-specific associations. These data extend prior analyses of cigarette smoking and CRC risk within the IWHS cohort (5,21) by including additional follow-up time and novel molecular marker associations.

METHODS

Approvals for the present study were obtained from the Institutional Review Boards for Human Research at Mayo Clinic Rochester, the University of Minnesota, and the University of Iowa.

Subjects

Details regarding the methods used for recruitment and enrollment of IWHS participants have been previously reported (22). Briefly, a 16-page baseline questionnaire was mailed out in January 1986 to 99,826 randomly selected women, ages 55 –69 years, who resided in Iowa and held a valid driver’s license. A total of 41,836 women (42%) returned the baseline questionnaire and these subjects constitute the parent IWHS cohort. As reported by Bisgard et al. (23), demographic characteristics and CRC rates were similar for the initial survey responders and non-responders. Vital status and state of residence were determined by mailed follow-up questionnaires in 1987, 1989, 1992, 1997, and 2004, as well as through linkage to Iowa death certificate records. Non-respondents were checked via the National Death Index to identify descendents. For the current analyses, women with a history of malignancy other than skin cancer (n = 3,830), unable to be followed longitudinally for at least 1 day (n = 10), or incomplete characterization of cigarette smoking at baseline (n = 660) were excluded (not mutually exclusive), leaving 37,399 women in the final analytic cohort.

Risk factor assessment

Cigarette smoking patterns among IWHS participants were ascertained at baseline in 1986, including smoking status (never, ever (former or current)), age at smoking initiation (years), smoking duration (years), average number of cigarettes smoked per day, cumulative pack-years, and induction period (difference between the baseline date and age at smoking initiation). Potential confounding factors were also derived from the baseline questionnaire, including body mass index; waist-to-hip ratio; physical activity level; exogenous estrogen use; and daily intake of total calories, total fat, red meat, calcium, folate, methionine, vitamin E, sucrose, and alcohol. Family history of CRC and non-steroidal anti-inflammatory drug use were not systematically recorded at baseline and therefore were not included in this study. However, since neither of these factors has been associated with smoking status (to our knowledge), the potential for confounding from these unmeasured variables seems remote.

Case ascertainment

Incident CRC cases were identified through the Iowa Cancer Registry, which participates in the National Cancer Institute’s SEER (Surveillance, Epidemiology, and End Results) program (24). Annual matching between a computer-generated list of all IWHS cohort members and SEER registry data was completed using combinations of first, last, and maiden names; zip code; birth date; and social security number. Data from follow-up surveys indicate that the migration rate out of the IWHS cohort is < 1% annually, allowing for near-complete follow-up of cancer-related end points (25). Incident CRC cases were identified by ICD-O codes, with cancers located in the cecum, ascending colon, hepatic flexure, transverse colon, and splenic flexure (ICD-O codes 18.0, 18.2–18.5) categorized as proximal colon and cancers located in the descending colon, sigmoid colon, rectosigmoid junction, and rectum (ICD-O codes 18.6, 18.7, 19.9, 20.9) categorized as distal colorectum (26,27). Beginning in 2006, archived, paraffin-embedded tissue specimens were requested from incident CRC cases diagnosed among IWHS participants through 31 December 2002. Tissue specimens were subsequently retrieved for 732/1,255 cases (58 %). For the present study, 22 incident CRC cases were excluded due to incomplete smoking data. To assess the possibility of selection bias introduced by tissue availability status, general demographics, smoking patterns, and tumor characteristics (size and stage) were compared between incident CRC cases with retrieved vs. non-retrieved tissue specimens; no statistically significant differences were observed (P > 0.05 for any comparison; data not shown). All incident CRC cases were histologically confirmed by a single gastrointestinal pathologist. Following tissue processing (including DNA extraction), high-quality, usable samples were obtained for 555 CRC cases.

Tissue selection and DNA extraction

Paraffin blocks were serially sectioned in 5 or 10 μm increments. One slide was stained with hematoxylin and eosin, and areas of normal and neoplastic (defined as ≥50% dysplastic cells in the field of view) tissue were identified. Tumor and normal tissue samples were scraped from unstained slides and placed into separate tubes for DNA extraction, according to manufacturer’s instructions (Qiagen, Valencia, CA).

Characterization of KRAS mutation status

The tumor DNA was PCR amplified with primers for exon 2 (codons 12 and 13). Thermocycler conditions were 95 ° C for 10 min, followed by 35 cycles of 95 °C for 30 s, 60 °C for 30 s, and 72 °C for 30s. There was a final extension for 10 min at 72 °C. The PCR product was cleaned using Shrimp Alkaline Phosphatase and Exonuclease I. Next, the product was sequenced using the Applied Biosystems PRISM BigDye Terminator v1.1 cycle sequencing kit per directions on an ABI PRISM 3730 DNA analyzer (ABI, Carlsbad, CA). Data analysis was performed using the Mutation Surveyor software (SoftGenetics, State College, PA). KRAS mutation status (wild-type vs. specific mutation) was determined for 507 (91.4 %) of the available 555 CRC cases. Positive results for 30 of the 555 cases were confirmed using the DxS KRAS Mutation Test Kit (Qiagen), which utilizes fluorescent, allele specific real-time quantitative PCR to detect seven point mutations in the KRAS oncogene on the LC 480 (Roche, Indianapolis, IN) instrument.

Statistical analysis

Data were descriptively summarized using frequencies and percentages for categorical variables and means and standard deviations for continuous variables. Measures of agreement across molecularly defined tumor subtypes of CRC were examined by use of kappa coefficients. Follow-up was calculated as age at completion of the baseline survey until age at first CRC diagnosis, age at move from Iowa, or age at death. If none of these events occurred, a woman was assumed to be alive, cancer free, and living in Iowa through 31 December 2002. Cox proportional hazard regression analysis was used to estimate relative risks (RRs) and 95% confidence intervals (CIs) for associations between cigarette smoking exposures and incident CRC outcomes.

All eligible IWHS participants were included in the Cox regression analyses, regardless of eventual cancer status. Incidence was modeled as a function of age because age is a better predictor of cancer risk in our cohort than follow-up time (28). We assessed the effects of smoking status (never, ever, former, or current), age at smoking initiation (>30 or ≤ 30 years), total smoking duration (1–19, 20–39, or ≥ 40 years), average number of cigarettes smoked per day (1–19, 20–39, or ≥ 40 cigarettes per day), cumulative cigarette pack-years (1–19, 20 –39, or ≥ 40 pack-years), and smoking induction period (<35, 35–39, 40–44, or ≥ 45 years). For all such analyses, never smokers were modeled as the reference group. Tests for trend were carried out for each smoking variable by ordering the categorized values from lowest to highest category and including the resulting variable as a linear term with 1 df in a Cox proportional hazards model. The Cox regression proportionality assumption was formally evaluated by fitting and testing a smoking-by-time interaction term.

Smoking associations were examined with respect to CRC subsets defined by KRAS mutation status (mutation negative or mutation positive) and anatomic subsite (proximal colon or distal colorectum). The outcome variable was incident CRC with the KRAS mutation status of interest; all other CRC cases (including those with missing or unknown KRAS mutation status) were considered censored observations at the date of first diagnosis. We also examined associations between cigarette smoking and incident CRC, based on the tissue availability status (available or not available) by using the same multi-outcome analytic approach as described above to determine whether incomplete tissue collection introduced any possible association biases. Two sets of Cox regression models were fit, one accounting for age and one adjusting for age and other potential confounding factors were body mass index (quartiles); waist-to-hip ratio (quartiles); physical activity level (low, moderate, or high); exogenous estrogen use (never or ever); and daily intake (quartiles) of total calories (kcal/day), total fat (g/day), red meat (g/day), calcium (mg/day), folate (μg/day), methionine (g/day), vitamin E (mg/day), sucrose (g/day), and alcohol (0, 0–3.4, or >3.4g/day). We also formally determined whether risk estimates for the smoking-related variables differed across KRAS-defined CRC subtypes using a competing risk form of Cox proportional hazards regression (29). This approach allowed us to specifically model and test the interaction between smoking (modeled as a covariate) and CRC subtype (included as a Cox regression stratum variable). All statistical tests were two-sided, and all analyses were carried out with the SAS (SAS Institute, Cary, NC) and S-Plus (Insightful, Seattle, WA) software systems (SAS[r] proprietary software, release 8.2 [TS2MO]; Splus, version 8.0.1 for Sun SPARC; and Sun OS, version 5.8, 32-bit:2006).

RESULTS

Smoking status was characterized as never, former, and current for 24,638 (66 %), 7,208 (19 %), and 5,553 (15 %) IWHS subjects, respectively. At baseline, never smokers were slightly older than either former or current smokers (mean ages 62.4, 61.9, and 61.4 years; P<0.01). Current smokers had a lower body mass index (25.3kg/m2) compared with never (27.3 kg/m2) or former (27.2kg/m2) smokers. Other baseline characteristics also differed by smoking status including waist-to-hip ratio, physical activity level, exogenous estrogen use, and daily intakes of total energy, total fat, red meat, calcium, folate, methionine, vitamin E, sucrose, and alcohol (P<0.01 for each variable, except P=0.03 for vitamin E; Table 1).

Table 1.

Baseline characteristics of participants, by cigarette smoking status

Variable Cigarette smoking status
Never (N=24,638) Former (N=7,208) Current (N=5,553)
Age at enrollment, years 62.4 (4.24) 61.9 (4.2) 61.4 (4.2)
Body mass index, kg/m2 27.3 (5.02) 27.2 (5.4) 25.3 (4.7)
Waist-to-hip ratio 0.837 (0.1) 0.839 (0.1) 0.843 (0.1)
Physical activity, N (%)
 Low 10,874 (45.1) 3,204 (45.1) 3,302 (60.3)
 Medium 6,957 (28.9) 1,906 (26.8) 1,266 (23.1)
 High 6,279 (26.0) 2,000 (28.1) 910 (16.6)
Estrogen use, N (%)
 Never 15,521 (63.7) 4,111 (57.6) 3,281 (59.7)
 Ever 8,848 (36.3) 3,024 (42.4) 2,217 (40.3)
Alcohol consumption, g/day 2.1 (5.7) 5.5 (10.2) 8.3 (14.8)
Total energy, kcal/day 1,807.8 (712.4) 1,755.1 (764.6) 1,770.8 (760.0)
Total fat, g/day 68.4 (31.1) 66.3 (33.5) 69.3 (33.6)
Red meat, g/day 91.6 (73.8) 83.5 (77.7) 91 (71.6)
Calcium, mg/day 1,101.7 (561.6) 1,120.9 (590.7) 1,008.6 (582.8)
Folate, μg/day 432.7 (259.9) 438.3 (271.3) 398.5 (268.1)
Methionine, g/day 1.9 (0.9) 1.8 (0.9) 1.8 (0.9)
Vitamin E, mg/day 67.7 (149.7) 68.4 (150.6) 62.2 (148.1)
Sucrose, g/day 42.7 (23.5) 39 (26.1) 38.1 (26.2)
Age at smoking initiation, years N/A 21.7 (6.6) 22.2 (7.5)
Duration smoked, years N/A 25 (12.9) 39.2 (8.0)
Average number of cigarettes per day N/A 15.5 (10.4) 18.4 (8.8)
Cumulative pack-years N/A 21.4 (19.7) 36.1 (19.1)
Induction period, years N/A 40.3 (7.3) 39.2 (8.0)
Time since smoking cessation, years N/A 15.7 (12.1) N/A

Data presented as mean value (s.d.), unless otherwise indicated; calcium, folate, and vitamin E intake includes supplements.

Among the 507 CRC cases for which KRAS status could be classified, the molecular subtype distribution was 342 (67 %) mutation-negative and 165 (33 %) mutation-positive tumors. The agreement between KRAS mutation status and other molecular subtypes was MSH-H, κ=0.25; CpG island methylator phenotype positive, κ=0.27; and BRAF mutation positive, κ=0.31. Multivariate-adjusted risk estimates for associations between cigarette smoking and incident CRC, stratified by KRAS mutation status, are presented in Table 2 (age-adjusted risk estimates were generally similar; data not shown). In general, smoking-related risk estimates were higher for KRAS mutation-negative than for KRAS mutation-positive tumors. Tests for trend across age at initiation (P trend=0.02), average number of cigarettes per day (P trend=0.01), cumulative pack-years (P trend=0.05), and induction period (P trend=0.04) exposure levels were statistically significant for KRAS mutation-negative tumors. Positive associations were also observed for KRAS mutation-negative tumors with smoking status (P trend=0.08) and total smoking duration (P=0.06), but the trend tests were not statistically significant. Women who smoked an average of ≥40 cigarettes per day were at the highest risk for KRAS mutation-negative tumors (RR=2.38; 95% CI=1.25–4.51, compared with never smokers). In contrast, none of the smoking variables were significantly associated with KRAS mutation-positive CRCs. Of note, tests for heterogeneity in the smoking-related risk estimates across levels of KRAS mutation status were not statistically significant (P>0.05 for each comparison), likely due in part to power limitations imposed by the available sample size.

Table 2.

Associations between cigarette smoking and incident colorectal cancer (CRC), by KRAS mutation status

Smoking variable Person-yearsa KRAS mutation negative (n=342)
KRAS mutation positive (n=165)
No. of casesa Median years to diagnosis RR (95% CI)b No. of casesa Median years to diagnosis RR (95% CI)b
Never smokers 375,486 226 12.42 1.00 (ref.) 110 11.93 1.00 (ref.)

Ever smokers 180,409 116 10.90 1.23 (0.97–1.57) 55 10.69 1.05 (0.74–1.50)

 Former 104,111 68 10.75 1.20 (0.91–1.59) 34 10.77 1.06 (0.70–1.60)

 Current 76,297 48 11.69 1.29 (0.92–1.79) 21 10.69 1.04 (0.63–1.71)

P trend 0.083 0.81

Age at initiation

 > 30 years 17,795 4 9.32 0.38 (0.14–1.02) 7 12.71 1.09 (0.48–2.50)

 ≤30 years 161,711 111 10.90 1.35 (1.06–1.72) 46 10.05 1.01 (0.70–1.46)

P trend 0.024 0.95

Total duration

 1–19 years 40,381 23 11.65 1.15 (0.74–1.77) 14 11.09 1.14 (0.62–2.09)

 20–39 years 87,073 48 10.83 1.14 (0.82–1.58) 20 11.73 0.86 (0.52–1.42)

 ≥40 years 50,848 43 10.06 1.40 (0.99–1.97) 18 9.20 1.09 (0.65–1.83)

P trend 0.06 0.99

Average number of cigarettes

 1–19 per day 95,965 58 10.99 1.13 (0.84–1.52) 28 10.05 0.99 (0.64–1.53)

 20–39 per day 73,546 48 10.90 1.29 (0.93–1.80) 22 11.73 1.00 (0.61–1.65)

 ≥40 per day 9,022 10 10.96 2.38 (1.25–4.51) 3 12.45 1.37 (0.43–4.34)

P trend 0.01 0.83

Cumulative pack-years

 1–19 74,225 46 11.19 1.20 (0.87–1.67) 18 11.65 0.78 (0.46–1.34)

 20–39 59,187 32 10.29 1.03 (0.70–1.51) 24 10.25 1.45 (0.91–2.31)

 ≥40 42,566 35 11.71 1.55 (1.07–2.25) 9 12.13 0.72 (0.36–1.44)

P trend 0.05 0.98

Induction period

 < 35 years 34,086 8 11.39 0.51 (0.25–1.03) 11 12.22 1.21 (0.63–2.34)

 35–40 years 46,825 28 11.46 1.55 (1.03–2.35) 10 11.05 0.94 (0.47–1.90)

 40–44 years 52,266 41 11.81 1.64 (1.16–2.32) 14 13.42 0.95 (0.53–1.71)

 ≥45 years 46,331 38 10.09 1.12 (0.79–1.60) 18 8.58 1.02 (0.61–1.71)

P trend 0.04 0.95

CI, confidence interval; RR, relative risk.

a

Do not always sum to total due to missing smoking data.

b

Adjusted for age, body mass index, waist-to-hip ratio, physical activity level, alcohol consumption, exogenous estrogen use, and daily intake of total calories, fat, sucrose, red meat, calcium, folate, vitamin E, and methionine.

Further analyses were conducted to explore smoking-related associations with incident CRCs defined by both KRAS mutation status and anatomic subsite (Table 3). For KRAS mutation-negative tumors located in the proximal colon, statistically significant risks were observed with smoking status (P trend=0.04), age at initiation (P=0.03), and average number of cigarettes per day (P trend=0.01). Conversely, null associations were observed between the analyzed smoking variables and KRAS mutation-negative tumors located in the distal colorectum, as well as KRAS mutation-positive tumors located in either the proximal colon or distal colorectum. Of note, relatively small event rates minimized our ability to obtain robust risk estimates for some of the KRAS-defined, subsite-specific CRC associations. In analyses defining cases as only those with available tissue, results comparing ever smokers with never smokers (RR=1.19; 95% CI=1.00–1.42, P=0.05) were similar to those based on all incident cases (RR = 1.20, 95 % CI=1.07–1.35, P=0.003), supporting a low likelihood of selection bias introduced by tissue availability status.

Table 3.

Associations between cigarette smoking and incident CRC, by KRAS mutation status and anatomic subsite

Smoking variable Person- yearsa CRC subtype
KRAS mutation negative
KRAS mutation positive
Proximal colon
Distal colon/rectum
Proximal colon
Distal colon/rectum
No. of casesa RR (95% CI)b No. of casesa RR (95% CI)b No. of casesa RR (95% CI)b No. of casesa RR (95% CI)b
Never smokers 375,486 129 1.00 (ref.) 95 1.00 (ref.) 57 1.00 (ref.) 53 1.00 (ref.)

Ever smokers 180,409 72 1.37 (1.00–1.86) 40 0.98 (0.66–1.45) 30 1.19 (0.74–1.92) 25 0.91 (0.54–1.53)

 Former 104,111 41 1.30 (0.90–1.88) 24 0.98 (0.62–1.55) 19 1.23 (0.71–2.13) 15 0.88 (0.47–1.65)

 Current 76,297 31 1.47 (0.97–2.23) 16 0.98 (0.56–1.72) 11 1.13 (0.57–2.23) 10 0.95 (0.47–1.95)

P trend 0.04 0.93 0.57 0.80

Age at initiation

 > 30 years 17,795 4 0.66 (0.24–1.79) 0 4 1.40 (0.50–3.89) 3 0.77 (0.19–3.19)

 ≤30 years 161,711 67 1.45 (1.06–2.00) 40 1.11 (0.75–1.65) 25 1.13 (0.68–1.88) 21 0.88 (0.51–1.52)

P trend 0.03 0.71 0.61 0.63

Total duration

 1–19 years 40,381 17 1.56 (0.93–2.60) 6 0.66 (0.29–1.52) 10 1.89 (0.95–3.75) 4 0.38 (0.09–1.57)

 20–39 years 87,073 28 1.19 (0.78–1.83) 16 0.87 (0.51–1.50) 10 0.84 (0.41–1.73) 10 0.88 (0.44–1.77)

 ≥40 years 50,848 25 1.37 (0.88–2.16) 18 1.44 (0.85–2.44) 8 0.95 (0.44–2.05) 10 1.21 (0.60–2.47)

P trend 0.12 0.48 0.85 0.88

Average number of cigarettes

 1–19 per day 95,965 36 1.24 (0.84–1.81) 20 0.91 (0.55–1.49) 17 1.23 (0.70–2.16) 11 0.74 (0.37–1.48)

 20–39 per day 73,546 31 1.50 (0.99–2.28) 17 1.05 (0.61–1.79) 10 0.94 (0.45–1.95) 12 1.06 (0.54–2.10)

 ≥40 per day 9,022 5 2.29 (0.93–5.66) 3 1.46 (0.46–4.65) 2 2.00 (0.48–8.35) 1 0.84 (0.11–6.15

P trend 0.01 0.77 0.64 0.87

Cumulative pack-years

 1–19 74,225 30 1.41 (0.94–2.12) 14 0.84 (0.47–1.49) 10 0.97 (0.49–1.92) 8 0.59 (0.25–1.39)

 20–39 59,187 18 0.99 (0.59–1.67) 14 1.09 (0.61–1.93) 13 1.53 (0.80–2.92) 11 1.35 (0.69–2.66)

 ≥40 42,566 21 1.67 (1.03–2.70) 12 1.20 (0.65–2.24) 4 0.66 (0.23–1.86) 5 0.78 (0.30–2.00)

P trend 0.08 0.63 0.94 0.91

Induction period

 < 35 years 34,086 6 0.71 (0.31–1.61) 2 0.28 (0.07–1.13) 6 1.50 (0.64–3.52) 5 0.94 (0.34–2.64)

 35–40 years 46,825 16 1.69 (0.97–2.95) 10 1.17 (0.60–2.29) 6 1.38 (0.58–3.29) 4 0.56 (0.17–1.84)

 40–44 years 52,266 31 2.31 (1.53–3.48) 10 0.88 (0.45–1.70) 7 0.95 (0.40–2.24) 7 0.93 (0.41–2.09)

 ≥45 years 46,331 18 0.88 (0.53–1.46) 18 1.36 (0.81–2.29) 10 1.07 (0.53–2.14) 8 0.97 (0.45–2.09)

P trend 0.09 0.42 0.83 0.75

CI, confidence interval; CRC, colorectal cancer; RR, relative risk.

a

Do not always sum to total due to missing smoking and/or anatomic subsite data.

b

Adjusted for age, body mass index, waist-to-hip ratio, physical activity level, alcohol consumption, exogenous estrogen use, and daily intake of total calories, fat, sucrose, red meat, calcium, folate, vitamin E, and methionine.

DISCUSSION

In this large prospective cohort study of older women, cigarette smoking was more closely associated with incident CRCs characterized by KRAS mutation-negative, rather than KRAS mutation-positive, status. Consistent (though not always statistically significant) trends were observed across all categories of cigarette smoking exposure, with intensity (i.e., average number of cigarettes per day), duration, and induction period demonstrating the strongest associations with KRAS mutation-negative tumors. Findings from the current study complement our previous report of differential associations between cigarette smoking and CRC subtypes defined by microsatellite instability, CpG island methylator phenotype, or BRAF mutation status in the IWHS cohort (5). Together, these data support the hypothesis that cigarette smoking affects CRC risk through the serrated pathway of carcinogenesis (30).

Existing data on KRAS-defined CRC risks among cigarette smokers and non-smokers are limited and inconsistent across studies (913,15). Consistent with our findings, Slattery et al. (11) reported no statistically significant associations between cigarette smoking and KRAS mutation-positive colon cancers among men or women in a multicenter case–control study. However, smoking exposure of ≥ 20 cigarettes per day was reportedly associated with a 50 % increased risk for KRAS mutation-negative tumors among men (OR=1.5; 95% CI=1.2–1.9), although a null association was observed among women (OR=1.1; 95% CI= 0.8–1.4) (11). A subsequent report from the Slattery group (9) demonstrated no apparent association between active smoking and KRAS mutation-positive rectal cancer, although subjects who described long-term exposure to environmental tobacco smoke of >10 h per week were at increased risk for this KRAS-mutated rectal cancer (OR=1.50; 95% CI=1.04–2.20). In the only other prospective study reported to date, Weijenberg et al. (13) analyzed CRC specimens from a subset of Netherlands Cohort Study participants (n=648 cases and 4,083 subcohort controls). Ex-smokers were found to be at increased risk for KRAS mutation-negative tumors (RR=1.79; 95% CI=1.00–3.20), but not for KRAS mutation-positive tumors (RR=1.20; 95% CI=0.61–2.33). Current smokers were not at significantly increased risk for either of the KRAS-defined CRC subtypes.

Other previous observational studies have described slightly different KRAS-specific CRC risk associations than we observed in the IWHS cohort. Diergaarde et al. (15) conducted a Dutch population-based, case–control study, which showed no significant difference between risks for KRAS mutation-positive or KRAS mutation-negative colon cancers among ever vs. never smokers (OR=1.4; 95% CI=0.7–2.8 and OR=0.8; 95% CI=0.5– 1.4, respectively). Using a different study design, Miyaki et al. (14) compared the prevalence of KRAS mutations in CRCs analyzed from a small group of cigarette smokers (n=28) and non-smokers (n=33), with no statistically significant difference detected between groups (32 vs. 39 % ; P=0.38). In an attempt to clarify the relationship between cigarette smoking, KRAS mutation, and colorectal neoplasia risk, Porta et al. (16) performed a meta-analysis of available observational data (including two studies with adenoma rather than adenocarcinoma end points) (10,12). The summary risk estimate for the association between tobacco use and KRAS mutation-positive tumors was not statistically significant (RR=0.96; 95% CI=0.83–1.13). However, no risk estimate for KRAS mutation-negative tumors was reported. Although not conclusive, data from the Netherlands Cohort Study and the IWHS (at least) suggest that further evaluation of cigarette smoking effects on KRAS -independent pathways of colorectal carcinogenesis may be informative.

Data from our study revealed statistically significant associations between smoking indicators and KRAS-defined, subsite-specific CRC risks. Conversely, in the Netherlands Cohort Study, smoking-related risk estimates for KRAS mutation-positive and KRAS mutation-negative tumors were reportedly similar when analyzed by colon, rectosigmoid, and rectal subsites (13). However, as with our study, small case numbers limited accurate risk assessment for some of the subgroup analyses. Interestingly, neoplasms arising through the serrated pathway of carcinogenesis (30) are typically characterized by a proximal colonic location, micro-satellite instability-high, CpG island methylator phenotype-high and BRAF mutation-positive status, and absence of KRAS mutations. These clinical and molecular features are consistent with the smoking-related associations reported in the current study, as well as in previous IWHS reports (5,31). Of note, other studies have also described limitations in colonoscopy-based screening and surveillance algorithms for reducing CRC risk in the proximal colon (32,33) and among active smokers (34). Coupled with our observations, further consideration of modified CRC early detection strategies tailored specifically to cigarette smokers seems indicated.

Major strengths of our study include the prospective design, detailed exposure data, prolonged follow-up time, near-complete case ascertainment, CRC tissue availability, and high-quality KRAS mutation analyses. Potential limitations should also be acknowledged. First, the reported findings cannot be directly extrapolated to other demographic subgroups (e.g., younger women, men, and non-Caucasian subjects), which will require further investigation in more diverse subject populations. Second, we were unable to retrieve adequate tissue specimens from all IWHS subjects with incident CRC for the planned molecular analyses. However, as noted above, tissue availability biases did not appear to influence the smoking-related, molecularly defined CRC risk estimates. Third, our sample sizes were relatively low in some of the CRC subsets defined by combinations of KRAS mutation status and anatomic subsite. Although we did find a statistically significant association between cigarette smoking and KRAS mutation-negative tumors arising from the proximal colon, it remains possible that associations based on other subtype/subsite combinations went undetected.

In summary, data from this prospective cohort study of older women suggest that cigarette smoking is associated with molecularly distinct CRC subtypes, which can be defined, in part, by KRAS mutation-negative status. These findings support a possible causative effect from tobacco exposure on KRAS-independent mechanisms of colorectal carcinogenesis, perhaps through methylation-induced silencing of DNA mismatch repair genes (and/or other growth regulating genes), resulting in tumors with the serrated pathway phenotype. Further investigation of smoking-related CRC risks based on other molecular markers and integrated pathways is ongoing in the IWHS cohort, which should yield additional insights regarding the mechanisms of carcinogenesis induced by this common, potentially modifiable exposure.

Study Highlights.

WHAT IS CURRENT KNOWLEDGE

  • Cigarette smoking is associated with a moderately increased risk for incident colorectal cancer (CRC) overall.

  • Smoking appears to be linked to higher risk for select, molecularly defined CRC subtypes.

  • To date, relatively few studies have examined smoking-related CRC risks stratified by KRAS mutation status, with mixed results.

WHAT IS NEW HERE

  • In this prospective, population-based study of older women, cigarette smoking was more strongly associated with KRAS mutation-negative tumors, particularly in the proximal colon.

  • These findings suggest that smoking is a greater risk factor for specific colorectal cancer (CRC) subtypes, and are consistent with the hypothesis that smoking adversely affects the serrated pathway of colorectal carcinogenesis.

  • If confirmed in other subject populations, then these observations may have important clinical implications with respect to CRC early detection and perhaps chemoprevention/chemotherapy strategies for cigarette smokers.

Acknowledgments

Financial support: This study was funded in part by National Cancer Institute Grants R01 CA39742 and R01 CA107333. No editorial assistance was obtained in preparing the manuscript.

Footnotes

Guarantor of the article: Paul J. Limburg, MD, MPH.

Potential competing interests: Dr Limburg served as a consultant for Genomic Health from 12 September 2008 to 19 April 2010. Mayo Clinic has licensed Dr Limburg’s intellectual property to Exact Sciences and he and Mayo Clinic have contractual rights to receive royalties through this agreement.

Specific author contributions: Study concept and design: Robert A. Vierkant, Kristin E. Anderson, James R. Cerhan, and Paul J. Limburg; acquisition of data: Lori S. Tillmans, Charles F. Lynch, Kristin E. Anderson, Amy J. French, Thomas C. Smyrk, Stephen N. Thibodeau, James R. Cerhan, and Paul J. Limburg; analysis and interpretation of data: N.J. Samadder, Robert A. Vierkant, Alice H. Wang, Kristin E. Anderson, Robert W. Haile, Lisa J. Harnack, John D. Potter, Susan L. Slager, James R. Cerhan, and Paul J. Limburg; drafting of the manuscript: N.J. Samadder, Robert A. Vierkant, Alice H. Wang, and Paul J. Limburg; critical revision of the manuscript for important intellectual content: N.J. Samadder, Charles F. Lynch, Kristin E. Anderson, Robert W. Haile, Lisa J. Harnack, John D. Potter, Susan L. Slager, Thomas C. Smyrk, Stephen N. Thibodeau, James R. Cerhan, and Paul J. Limburg; statistical analysis: Robert A. Vierkant, Alice H. Wang, Susan L. Slager, and Paul J. Limburg; obtained funding: Kristin E. Anderson, James R. Cerhan, and Paul J. Limburg.

References

  • 1.Siegel R, Ward E, Brawley O, et al. Cancer statistics, 2011: the impact of eliminating socioeconomic and racial disparities on premature cancer deaths. CA Cancer J Clin. 2011;61:212–36. doi: 10.3322/caac.20121. [DOI] [PubMed] [Google Scholar]
  • 2.Centers for Disease Control and Prevention (CDC) Vital signs: current cigarette smoking among adults aged >or=18 years—United States, 2009. MMWR Morb Mortal Wkly Rep. 2010;59:1135–40. [PubMed] [Google Scholar]
  • 3.Botteri E, Iodice S, Bagnardi V, et al. Smoking and colorectal cancer: a meta-analysis. JAMA. 2008;300:2765–78. doi: 10.1001/jama.2008.839. [DOI] [PubMed] [Google Scholar]
  • 4.Botteri E, Iodice S, Raimondi S, et al. Cigarette smoking and adenomatous polyps: a meta-analysis. Gastroenterology. 2008;134:388–95. doi: 10.1053/j.gastro.2007.11.007. [DOI] [PubMed] [Google Scholar]
  • 5.Limsui D, Vierkant RA, Tillmans LS, et al. Cigarette smoking and colorectal cancer risk by molecularly defined subtypes. J Natl Cancer Inst. 2010;102:1012–22. doi: 10.1093/jnci/djq201. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 6.Slattery ML, Curtin K, Anderson K, et al. Associations between cigarette smoking, lifestyle factors, and microsatellite instability in colon tumors. J Natl Cancer Inst. 2000;92:1831–6. doi: 10.1093/jnci/92.22.1831. [DOI] [PubMed] [Google Scholar]
  • 7.Yang P, Cunningham JM, Halling KC, et al. Higher risk of mismatch repair-deficient colorectal cancer in alpha(1)-antitrypsin deficiency carriers and cigarette smokers. Mol Genet Metab. 2000;71:639–45. doi: 10.1006/mgme.2000.3089. [DOI] [PubMed] [Google Scholar]
  • 8.Samowitz WS, Albertsen H, Sweeney C, et al. Association of smoking, CpG island methylator phenotype, and V600E BRAF mutations in colon cancer. J Natl Cancer Inst. 2006;98:1731–8. doi: 10.1093/jnci/djj468. [DOI] [PubMed] [Google Scholar]
  • 9.Curtin K, Samowitz WS, Wolff RK, et al. Somatic alterations, metabolizing genes and smoking in rectal cancer. Int J Cancer. 2009;125:158–64. doi: 10.1002/ijc.24338. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10.Martinez ME, Maltzman T, Marshall JR, et al. Risk factors for Kiras protooncogene mutation in sporadic colorectal adenomas. Cancer Res. 1999;59:5181–5. [PubMed] [Google Scholar]
  • 11.Slattery ML, Anderson K, Curtin K, et al. Lifestyle factors and Kiras mutations in colon cancer tumors. Mutat Res. 2001;483:73–81. doi: 10.1016/s0027-5107(01)00228-7. [DOI] [PubMed] [Google Scholar]
  • 12.Wark PA, Van der Kuil W, Ploemacher J, et al. Diet, lifestyle and risk of K-ras mutation-positive and -negative colorectal adenomas. Int J Cancer. 2006;119:398–405. doi: 10.1002/ijc.21839. [DOI] [PubMed] [Google Scholar]
  • 13.Weijenberg MP, Aardening PW, de Kok TM, et al. Cigarette smoking and KRAS oncogene mutations in sporadic colorectal cancer: results from the Netherlands Cohort Study. Mutat Res. 2008;652:54–64. doi: 10.1016/j.mrgentox.2007.12.008. [DOI] [PubMed] [Google Scholar]
  • 14.Miyaki M, Iijima T, Ishii R, et al. Increased frequency of p53 mutation in sporadic colorectal cancer from cigarette smokers. Jpn J Clin Oncol. 2002;32:196–201. doi: 10.1093/jjco/hyf047. [DOI] [PubMed] [Google Scholar]
  • 15.Diergaarde B, Braam H, van Muijen GN, et al. Dietary factors and microsatellite instability in sporadic colon carcinomas. Cancer Epidemiol Biomarkers Prev. 2003;12:1130–6. [PubMed] [Google Scholar]
  • 16.Porta M, Crous-Bou M, Wark PA, et al. Cigarette smoking and K-ras mutations in pancreas, lung and colorectal adenocarcinomas: etiopathogenic similarities, differences and paradoxes. Mutat Res. 2009;682:83–93. doi: 10.1016/j.mrrev.2009.07.003. [DOI] [PubMed] [Google Scholar]
  • 17.Pritchard CC, Grady WM. Colorectal cancer molecular biology moves into clinical practice. Gut. 2011;60:116–29. doi: 10.1136/gut.2009.206250. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18.Allegra CJ, Jessup JM, Somerfield MR, et al. American Society of Clinical Oncology provisional clinical opinion: testing for KRAS gene mutations in patients with metastatic colorectal carcinoma to predict response to anti-epidermal growth factor receptor monoclonal antibody therapy. J Clin Oncol. 2009;27:2091–6. doi: 10.1200/JCO.2009.21.9170. [DOI] [PubMed] [Google Scholar]
  • 19.Van Cutsem E, Dicato M, Arber N, et al. Molecular markers and biological targeted therapies in metastatic colorectal cancer: expert opinion and recommendations derived from the 11th ESMO/World Congress on Gastrointestinal Cancer, Barcelona, 2009. Ann Oncol. 2010;21(Suppl 6):vi, 1–10. doi: 10.1093/annonc/mdq273. [DOI] [PubMed] [Google Scholar]
  • 20.Harris CC. Chemical and physical carcinogenesis: advances and perspectives for the 1990s. Cancer Res. 1991;51:5023s–44s. [PubMed] [Google Scholar]
  • 21.Limburg PJ, Vierkant RA, Cerhan JR, et al. Cigarette smoking and colorectal cancer: long-term, subsite-specific risks in a cohort study of postmenopausal women. Clin Gastroenterol Hepatol. 2003;1:202–10. doi: 10.1053/cgh.2003.50030. [DOI] [PubMed] [Google Scholar]
  • 22.Folsom AR, Kaye SA, Prineas RJ, et al. Increased incidence of carcinoma of the breast associated with abdominal adiposity in postmenopausal women. Am J Epidemiol. 1990;131:794–803. doi: 10.1093/oxfordjournals.aje.a115570. [DOI] [PubMed] [Google Scholar]
  • 23.Bisgard KM, Folsom AR, Hong CP, et al. Mortality and cancer rates in nonrespondents to a prospective study of older women: 5-year follow-up. Am J Epidemiol. 1994;139:990–1000. doi: 10.1093/oxfordjournals.aje.a116948. [DOI] [PubMed] [Google Scholar]
  • 24.Hankey BF, Ries LA, Edwards BK. The surveillance, epidemiology, and end results program: a national resource. Cancer Epidemiol Biomarkers Prev. 1999;8:1117–21. [PubMed] [Google Scholar]
  • 25.Zhang S, Folsom AR, Sellers TA, et al. Better breast cancer survival for postmenopausal women who are less overweight and eat less fat. The Iowa Women’s Health Study. Cancer. 1995;76:275–83. doi: 10.1002/1097-0142(19950715)76:2<275::aid-cncr2820760218>3.0.co;2-6. [DOI] [PubMed] [Google Scholar]
  • 26.Cheng X, Chen VW, Steele B, et al. Subsite-specific incidence rate and stage of disease in colorectal cancer by race, gender, and age group in the United States, 1992–1997. Cancer. 2001;92:2547–54. doi: 10.1002/1097-0142(20011115)92:10<2547::aid-cncr1606>3.0.co;2-k. [DOI] [PubMed] [Google Scholar]
  • 27.Troisi RJ, Freedman AN, Devesa SS. Incidence of colorectal carcinoma in the U.S. : an update of trends by gender, race, age, subsite, and stage 1975–1994. Cancer. 1999;85:1670–6. [PubMed] [Google Scholar]
  • 28.Korn EL, Graubard BI, Midthune D. Time-to-event analysis of longitudinal follow-up of a survey: choice of the time-scale. Am J Epidemiol. 1997;145:72–80. doi: 10.1093/oxfordjournals.aje.a009034. [DOI] [PubMed] [Google Scholar]
  • 29.Lunn M, McNeil D. Applying Cox regression to competing risks. Biometrics. 1995;51:524–32. [PubMed] [Google Scholar]
  • 30.Leggett B, Whitehall V. Role of the serrated pathway in colorectal cancer pathogenesis. Gastroenterology. 2010;138:2088–100. doi: 10.1053/j.gastro.2009.12.066. [DOI] [PubMed] [Google Scholar]
  • 31.Limburg P, Vierkant R, Cerhan J, et al. Cigarette smoking and colorectal cancer risk: a long-term, prospective study among postmenopausal women. Clin Gastroenterol Hepatol. 2003;3:202–10. doi: 10.1053/cgh.2003.50030. [DOI] [PubMed] [Google Scholar]
  • 32.Baxter NN, Goldwasser MA, Paszat LF, et al. Association of colonoscopy and death from colorectal cancer. Ann Intern Med. 2009;150:1–8. doi: 10.7326/0003-4819-150-1-200901060-00306. [DOI] [PubMed] [Google Scholar]
  • 33.Lakoff J, Paszat LF, Saskin R, et al. Risk of developing proximal versus distal colorectal cancer after a negative colonoscopy: a population-based study. Clin Gastroenterol Hepatol. 2008;6:1117–21. doi: 10.1016/j.cgh.2008.05.016. quiz 1064. [DOI] [PubMed] [Google Scholar]
  • 34.Brenner H, Chang-Claude J, Seiler CM, et al. Long-term risk of colorectal cancer after negative colonoscopy. J Clin Oncol. 2011;29:3761–7. doi: 10.1200/JCO.2011.35.9307. [DOI] [PubMed] [Google Scholar]

RESOURCES