Tumor markers and rectal cancer: support for an inflammation-related pathway

Abstract

Inflammation may be a key element in the etiology of colorectal cancer (CRC). In this study we examine associations between factors related to inflammation and specific rectal cancer mutations. A population-based study of 750 rectal cancer cases with interview and tumor DNA were compared to 1205 population-based controls. Study participants were from Utah and the Northern California Kaiser Permanente Medical Care Program. Tumor DNA was analyzed for TP53 and KRAS2 mutations and CpG Island methylator phenotype (CIMP). We assessed how these tumor markers were associated with use of anti-inflammatory drugs, polymorphisms in the IL6 genes (rs1800795 and rs1800796), and dietary antioxidants. Ibuprofen-type drugs, IL6 polymorphisms (rs1800796), and dietary alpha tocopherol and lycopene significantly altered likelihood of having a TP53 mutation. This was especially true for TP53 transversion mutations and dietary antioxidants (OR for beta carotene 0.51 95% CI 0.27,0.97, p trend 0.03; alpha tocopherol 0.41 95% CI 0.20,0.84, p trend 0.02) Beta carotene and ibuprofen significantly altered risk of KRAS2 tumors. The associations between lutein and tocopherol and TP53 and KRAS2 mutations were modified by IL6 genotype. These results suggest that inflammation-related factors may have unique associations with various rectal tumor markers. Many factors involved in an inflammation related pathway were associated with TP53 mutations and some dietary factors appeared to be modified by IL6 genotype.

Inflammation may be a key element in the etiology of colorectal cancer (CRC). People with idiopathic inflammatory bowel disease (ulcerative colitis and Crohn's disease), have a greater likelihood of developing colon cancer ^{1, 2}. People who use anti-inflammatory drugs (aspirin/NSAIDs) have consistently been shown to have lower risk of CRC than non-users ^3–5. Furthermore, aspirin/NSAIDs appear to modify the association between many diet, lifestyle, and genetic factors associated with CRC, suggesting that the underlying inflammatory state of the bowel may be a key determinant of other risk factors ^6–8. Our previous assessment of aspirin/NSAIDs with specific mutations in colon tumors showed similar associations for all genetic and epigenetic alterations ^9–12.

Cytokines such as IL-6 have been cited as key molecular factors that link inflammation to epigenetic and genetic changes in tissue ¹³. IL-6 is a pleiotropic cytokine, meaning that it has multiple influences on different tissues, although it is generally viewed as a pro-inflammatory cytokine. Data from a colon cancer study have shown that the IL6 rs18000795 polymorphism was associated with TP53-mutated tumors among those using aspirin/NSAID (Slattery paper under review). Genetic variants in the IL6 gene also have been related to levels of circulating C-reactive protein (CRP) ¹⁴ and with plasma IL-6 response to immunization ¹⁵.

Chemical factors such as reactive oxygen and nitrogen species may act directly or indirectly to sustain chronic inflammation that contributes to the carcinogenic process. Antioxidants may play a role in balancing the harmful effect of free radicals that leads to cancer. Oxidative stress occurs when free radical generation exceeds the systems ability to neutralize and eliminate them. Reactive oxygen and nitrogen species are produced under the stimulus of pro-inflammatory cytokines such as IL-6 ¹³ Thus, genetic polymorphisms of the IL6 gene that have been shown to be associated with IL-6 levels as stated above may importantly regulate pathways involved in inflammation. Although dietary antioxidants and carotenoids have not been associated consistently with colorectal cancer in general, it is possible that they are related to specific tumor mutations involved in oxidation-related mechanisms.

There is little information on the association between inflammation-related factors and specific acquired epigenetic and genetic changes in rectal tumors. Some studies suggest that TP53 mutations may be induced by inflammation-related processes ^{13, 16}. Other studies have implicated inflammation as a contributor to CpG Island Methylated (CIMP) tumors ¹⁷. In this study we examine the associations between inflammation-related factors including recent use of aspirin and non-steroidal anti-inflammatory drugs, dietary carotenoids and anti-oxidants, and IL6 rs1800795 and rs1800796 polymorphisms and CIMP and mutations in the TP53 and KRAS2 genes in rectal tumors. We also evaluate the impact of aspirin/NSAID use and IL6 polymorphisms on tumor associations with dietary carotenoids and antioxidants.

Methods

Participants in the study were from the Kaiser Permanente Medical Care Program of Northern California (KPMCP) and the state of Utah. All eligible cases within these defined areas were identified and recruited for the study. Cases with a first primary tumor in the recto-sigmoid junction or rectum were identified between May 1997 and May 2001. Case eligibility was determined by the Surveillance Epidemiology and End Results (SEER) Cancer Registries in Northern California and in Utah. To be eligible for the study, participants had to be between 30 and 79 years of age at time of diagnosis, English speaking, mentally competent to complete the interview, could not have had previous colorectal cancer¹⁸, and could not have known (as indicated on the pathology report) familial adenomatous polyposis, ulcerative colitis, or Crohn's disease.

A total of 1505 rectal cancer cases were identified; of these, 982 were interviewed; reasons for non-response have been detailed ¹⁹. Block retrieval involved obtaining pre-operative biopsy prior to treatment as well as paraffin embedded tissue from the resection. In some instances because of radiation prior to resection, tissue was limited from the resection and therefore, biopsy specimens were used for making tumor DNA. In Utah, blocks were requested for all cases except those who refused release of blocks. For those who were not interviewed and had not signed a medical record release, the Utah Cancer Registry retrieved the blocks and released them to the study without key identifiers of name, address, and complete date of birth (year and month of birth were released). At the KPMCP, samples were retrieved from persons who signed a consent form or who had died. For the 1495 eligible rectal cancer cases identified at both centers, 239 people identified with rectal cancer had not given consent to have the tissue released (15.9%), and for an additional 234 cases, either tumor tissue could not be obtained or DNA could not be extracted. Tumor DNA was extracted from 81.4% of all rectal cancer cases identified, of which 750 cases had interview data. Controls were randomly selected from membership lists at KPMCP, social security lists, and driver's license list (people under 65 years); 1205 controls (68.8% of those selected) participated and are included in these analyses.

Genetic Analysis

Tumor DNA obtained from paraffin-embedded tissue was characterized by their genetic profile that include sequence data for exons 5 through 8, or the hotspots of mutations of the TP53 gene; sequence data for KRAS2 codons 12 and 13; and five CpG Island (CIMP) markers MINT1, MINT2, MINT31, p16, and MLH1. At this time there is no “consensus” as to the appropriate CIMP panel or method of detection. However, we have used our panel to demonstrate significant relationships between CIMP and numerous variables, including cigarette smoking and the BRAF V600E mutation, which were independent of microsatellite instability ^{20, 21}. This work has helped to support the legitimacy of the CIMP concept ²². Germline DNA was available from blood drawn at the time of the interview. We assessed two IL6 makers as previously described ²³ given the role of IL-6 in the inflammation process ²⁴.

Diet and Lifestyle Data

Trained and certified interviewers collected diet and lifestyle data as previously outlined ^{25, 26}. The referent year for the study was the calendar year approximately two years prior to date of diagnosis (cases) or selection (controls). Information was collected on demographic factors such as age, sex, and study center; physical activity as determined by a detailed physical activity questionnaire that obtained information on activity patterns 10 and 20 years ago as well as activity during the referent year ^{27, 28}; body size, including usual adult height and weight two and five years prior to diagnosis; cigarette smoking history; family history of colorectal cancer in first degree relatives; medical and reproductive history including use of hormone replacement therapy (HRT).

Regular use of aspirin and non-steroidal anti-inflammatory drugs were obtained from the following question: Before the referent date, did you ever take aspirin, excluding Tylenol, regularly? Some brand names for aspirin include Anacin, Arthritis Pain Formula, Ascriptin Tablets, Bayer, Buffrin, Empirin, Excedrin, and Vanquish. Before the referent date did you ever take other non-steroidal anti-inflammatory drugs or arthritis medicines such as ibuprofen, Motrin, Clinoril, Naprosyn, or Feldene.” Regular was defined as at least 3 times a week for one month. For those people who reported regular use, additional questions were asked about how long they used these medications on a regular basis and when they were last used.

Dietary intake was ascertained using an adaptation of the CARDIA diet history ^{26, 29, 30}. Participants were asked to recall foods eaten, the frequency at which they were eaten, serving size, and if fats were added in the preparation. Nutrient information was obtained by converting food intake data into nutrient data using the Minnesota Nutrition Coding Center (NCC) nutrient database. Data are directly available from the NCC nutrient database for alpha, beta, gamma, and delta-tocopherol as well as for total alpha tocopherol equivalents, an estimate of the total biologically activity from all forms of tocopherols in the diet ³¹. Carotenoid values were available for alpha-carotene, lutein, lycopene, and zeaxanthine from the NCC database and have been previously described ³².

Statistical analysis

We assessed factors thought to be associated with an inflammation-related pathway including recent use (within two years of diagnosis or selection) of aspirin, ibuprofen-type drugs, or the combination of these two anti-inflammatory drugs where the referent group was those who did not use either aspirin or ibuprofen-type drugs, IL6 rs1800795, also referred to as −174, and rs1800796, also referred to as −572, genotypes; and dietary antioxidants and carotenoids that may have specific role in an inflammation-related pathway. We assessed beta carotene, lycopene, and lutein and zeaxanthine as well as vitamin C, dietary selenium, and alpha, beta, and gamma tocopherol using sex-specific tertiles of intake, noted as T1, T2, and T3 in the tables. Dietary variables were assessed by tertile of intake, based on the distribution of the controls for men and women separately. We also assessed differences in association by sex, a combination of any recent use of aspirin or ibuprofen-type drugs, and by IL6 genotypes. Recent use, within the past two years, of aspirin or ibuprofen-type drugs was assessed since recent use was shown to be a better predictor of rectal cancer risk than any or lifetime use ³³.

All statistical analysis was done using SAS version 9.1 (SAS Institute, Cary, NC). Tumors were defined by specific mutations detected as any TP53 versus no TP53 mutation, any KRAS2 mutations versus no KRAS2 mutation, or CIMP positive versus CIMP negative. CIMP positive was defined as at least two methylated markers. For TP53 and KRAS2 mutations, we also examined transversion and transition mutations since specific types of mutations were assessed because other studies have shown specific mutations to have etiologic associations ^{9, 34}. Because of few cases with MSI and BRAF (22 and 27 respectively) we were unable to examine inflammation-related factors with these mutations. Population-based controls were used to assess associations for the population overall while examining multiple outcomes defined by tumor status. Multiple logistic regression models were used to compare all interviewed cases, regardless of whether or not tumor tissue was obtained, to controls. Multinomial logistic regression models were used to assess associations for the population overall comparing specific types of mutations to controls. Cases could contribute one to three observations in the multinomial logistic regression models depending how an individual's number of tumor mutations (CIMP, KRAS2, TP53). The multinomial logistic regression models were adjusted for non-independent observations by creating independent clusters by subject using PROC SURVEYLOGISTIC. All logistic regression models were adjusted for age at diagnosis or selection and sex along with other factors that have been shown to be related to colon cancer including body mass index (BMI) in kg/m², long-term vigorous physical activity, pack-years of cigarettes smoked, dietary calcium per 1000 calories, and total energy intake per 1000 calories. Additional adjustment for center and stage did not alter results. Interaction was assessed by determining if the interaction term significantly improved the overall fit of the model by comparing a model with the interaction term as an ordered categorical variable to a model without the interaction term using the likelihood ratio test with one degree of freedom. P for trend was assessed over categories of variables; in the instance of genotypes, the trend p value was based on a model that included variant, heterozygote, and wild type.

Results

The majority of rectal cases were men and over 60 years of age (Table 1). Eleven percent of rectal cases with completed test results had a CIMP positive tumor; 48.3% had a TP53 mutation and 28.9% had a KRAS2 mutation.

Table 1.

Description of the study population.

		Cases		Controls
		n	%	n	%
Center
	KPMCP	614	64.6	746	61.9
	Utah	337	35.4	459	38.1
Age
	<40	25	2.6	33	2.7
	40 to 49	122	12.8	145	12.0
	50 to 59	248	26.1	304	25.2
	60 to 69	318	33.4	392	32.5
	70 to 79	238	25.0	331	27.5
Sex
	Male	559	58.8	673	55.9
	Female	392	41.2	532	44.1
		951
AJCC Stage
	I	464	48.8	N/A
	II	157	16.5	N/A
	III	227	23.9	N/A
	IV	84	8.8	N/A
	Missing	19	2.0	N/A
Tumor Block Avalible
	Yes	750	78.9	N/A
	No	201	21.1	N/A
CIMP
	Low	596	79.5	N/A
	High	74	9.9	N/A
	Missing	80	10.7	N/A
TP53
	Wild Type	364	48.5	N/A
	Mutated	340	45.3	N/A
	Missing	46	6.1	N/A
KRAS2
	Wild Type	528	70.4	N/A
	Mutated	215	28.7	N/A
	Missing	7	0.9	N/A

Recent ibuprofen-related drug use appeared to be more strongly associated with rectal tumors overall and with specific mutations than was aspirin (Table 2). The strongest associations were observed for TP53 mutations with slightly weaker associations observed for KRAS2 mutations. CIMP positive tumors were not significantly associated with either drug. The IL6 rs180796 polymorphism was associated with TP53 mutated tumors; those with a C allele were significantly more likely to have a TP53 mutated tumor compared to those with the GG genotype (OR 1.70, 95% CI 1.18, 2.33). Examination of specific carotenoids and antioxidants showed that higher levels of beta carotene decreased the likelihood of a KRAS2 mutation; lycopene had a significant inverse association with TP53 mutations. Unlike the other dietary variables in Table 2, there was a significant interaction between sex and alpha tocopherol and rectal tumor mutations. Alpha tocopherol was associated with a strong protective effect for both CIMP positive and TP53 mutations among women, while among men alpha tocopherol was associated with a twofold increased risk of CIMP positive tumors. Higher levels of beta tocopherol also were significantly associated with reduced risk of TP53 mutations. Neither vitamin C nor selenium was associated with rectal cancer overall or with specific tumor mutations (data not shown in the table).

Table 2.

Associations between inflammation-related factors and rectal cancer overal and specific rectal tumor mutations

		Control	All Cases*			CIMP High			p53 Mutation			Ki-ras Mutation
		n	n	OR	(95% CI)	n	OR	(95% CI)	n	OR	(95% CI)	n	OR	(95% CI)
Recent aspirin use	No	870	726	1.00		51	1.00		260	1.00		165	1.00
	Yes	327	219	0.79	(0.64, 0.97)	23	1.10	(0.65, 1.86)	78	0.78	(0.58, 1.04)	47	0.74	(0.52, 1.05)
Recent ibuprofen use	No	935	802	1.00		62	1.00		292	1.00		181	1.00
	Yes	261	147	0.65	(0.52, 0.82)	12	0.78	(0.41, 1.50)	47	0.58	(0.41, 0.81)	34	0.67	(0.45, 0.99)
IL6
rs1800796	GG	863	647	1.00		55	1.00		229	1.00		155	1.00
	GC and CC	142	146	1.36	(1.05, 1.76)	7	0.72	(0.32, 1.62)	64	1.70	(1.21, 2.38)	26	1.02	(0.64, 1.61)
	P trend			0.05			0.41			0.01			0.77
Antioxidants and Carotenoids
Beta carotene	T1	398	356	1.00		31	1.00		123	1.00		91	1.00
	T2	410	325	0.93	(0.75, 1.14)	20	0.62	(0.35, 1.13)	109	0.91	(0.68, 1.23)	69	0.75	(0.53, 1.06)
	T3	397	270	0.87	(0.70, 1.08)	23	0.74	(0.42, 1.32)	108	1.02	(0.75, 1.38)	55	0.69	(0.47, 1.00)
	P trend			0.20			0.28			0.93			0.04
Lycopene	T1	398	340	1.00		31	1.00		132	1.00		78	1.00
	T2	410	330	0.93	(0.76, 1.15)	19	0.66	(0.36, 1.19)	126	0.93	(0.69, 1.23)	70	0.89	(0.63, 1.28)
	T3	397	281	0.82	(0.66, 1.02)	24	0.89	(0.50, 1.57)	82	0.62	(0.45, 0.86)	67	0.91	(0.63, 1.32)
	P trend			0.07			0.64			<.01			0.61
Lutein & Zeozanthine	T1	397	354	1.00		27	1.00		125	1.00		86	1.00
	T2	411	311	0.89	(0.72, 1.10)	19	0.69	(0.37, 1.27)	111	0.91	(0.68, 1.22)	72	0.85	(0.60, 1.21)
	T3	397	286	0.91	(0.73, 1.13)	28	1.07	(0.61, 1.87)	104	0.94	(0.69, 1.28)	57	0.75	(0.52, 1.10)
	P trend			0.37			0.82			0.69			0.13
Alpha Tocopherol Women	T1	175	159	1.00		18	1.00		60	1.00		36	1.00
	T2	181	135	0.81	(0.59, 1.11)	7	0.33	(0.13, 0.82)	51	0.81	(0.52, 1.25)	35	0.89	(0.53, 1.51)
	T3	176	98	0.60	(0.43, 0.84)	8	0.40	(0.17, 0.98)	30	0.47	(0.29, 0.77)	29	0.78	(0.45, 1.35)
	P trend			<.01			0.03			<.01			0.37
Men	T1	223	179	1.00		8	1.00		60	1.00		42	1.00
	T2	228	232	1.31	(1.00, 1.73)	17	2.39	(0.99, 5.80)	81	1.35	(0.91, 1.99)	44	1.04	(0.65, 1.66)
	T3	222	148	0.89	(0.66, 1.20)	16	2.25	(0.91, 5.57)	58	1.01	(0.67, 1.54)	29	0.69	(0.41, 1.15)
	P trend			0.48			0.09			0.94			0.17
Beta Tocopherol	T1	398	382	1.00		28	1.00		144	1.00		87	1.00
	T2	410	315	0.83	(0.68, 1.02)	23	0.88	(0.49, 1.57)	107	0.75	(0.56, 1.01)	64	0.71	(0.50, 1.02)
	T3	397	254	0.72	(0.58, 0.89)	23	0.91	(0.51, 1.63)	89	0.65	(0.48, 0.89)	64	0.80	(0.56, 1.15)
	P trend			<.01			0.75			<.01			0.20
Gamma	T1	399	331	1.00		26	1.00		121	1.00		71	1.00
	T2	408	278	0.74	(0.60, 0.92)	25	0.90	(0.50, 1.62)	101	0.74	(0.54, 1.00)	66	0.82	(0.56, 1.19)
	T3	398	342	0.91	(0.73, 1.14)	23	0.86	(0.47, 1.60)	118	0.86	(0.63, 1.17)	78	0.92	(0.64, 1.34)
	P trend			0.45			0.64			0.34			0.69

^*Includes cases without tumor marker data.

Adjusted for age, sex, BMI, long-term activity level, pack-years of cigarette smoking, dietary calcium, and energy intake.

Assessment of transition and transversion mutations for TP53 and KRAS2 showed some differences in association for these two types of mutations (Table 3). While recent aspirin use showed a stronger inverse association with TP53 transverions than TP53 transition mutations, recent ibuprofen use was more strongly associated with both TP53 and KRAS2 transitions than transversion mutations. Beta carotene and alpha tocopherol were more strongly associated with TP53 transversion than transition mutations, while lycopene and beta tocopherol were inversely associated with KRAS2 transition mutations but not KRAS2 transversion mutations. We did not obverse significant association for specific KRAS2 or TP53 mutations beyond those mentioned.

Table 3.

Associations between inflammation-related factors and p-53 and Ki-ras transitions and transversions.

		p53 Transition*	p53 transversions	p53 Transition vs. Control		p53 Transversion vs. Control		Ki-ras Transition*	Ki-ras Transversion*	Ki-ras Transition vs. Control		Ki-ras Transversion vs. Control
		n	n	OR	(95% CI)	OR	(95% CI)	n	n	OR	(95% CI)	OR	(95% CI)
Recent aspirin use	No	185	58	1.00		1.00		108	57	1.00		1.00
	Yes	61	11	0.88	(0.64, 1.22)	0.48	(0.24, 0.94)	31	16	0.72	(0.47, 1.11)	0.76	(0.43, 1.36)
Recent ibuprofen	No	214	56	1.00		1.00		122	59	1.00		1.00
	Yes	33	13	0.57	(0.38, 0.84)	0.77	(0.41, 1.45)	18	16	0.52	(0.31, 0.87)	0.97	(0.54, 1.72)
IL6
rs1800796	CG	171	44	1.00		1.00		106	49	1.00		1.00
	GC and CC	46	12	1.61	(1.10, 2.36)	1.75	(0.89, 3.45)	14	12	0.82	(0.45, 1.49)	1.48	(0.76, 2.89)
	P trend			0.03		0.38				0.53		0.67
Antioxidant and Carotenoids
Beta carotene	Q1	84	34	1.00		1.00		59	32	1.00		1.00
	Q2	77	20	0.95	(0.68, 1.35)	0.58	(0.33, 1.04)	45	24	0.75	(0.49, 1.14)	0.75	(0.43, 1.30)
	Q3	87	15	1.18	(0.84, 1.65)	0.51	(0.27, 0.97)	36	19	0.70	(0.45, 1.10)	0.67	(0.37, 1.22)
	P trend			0.28		0.03				0.23		0.17
Lycopene	Q1	98	23	1.00		1.00		55	23	1.00		1.00
	Q2	90	30	0.89	(0.64, 1.23)	1.27	(0.71, 2.25)	51	19	0.93	(0.62, 1.41)	0.80	(0.43, 1.51)
	Q3	60	16	0.60	(0.42, 0.87)	0.69	(0.35, 1.35)	34	33	0.68	(0.43, 1.08)	1.46	(0.83, 2.58)
	P trend			<0.01		0.30				0.11		0.16
Alpha Tocopherol	Q1	81	27	1.00		1.00		52	26	1.00		1.00
	Q2	99	30	1.22	(0.88, 1.69)	1.01	(0.59, 1.76)	52	27	0.96	(0.64, 1.46)	1.03	(0.59, 1.81)
	Q3	68	12	0.84	(0.59, 1.21)	0.41	(0.20, 0.84)	36	22	0.67	(0.43, 1.06)	0.85	(0.47, 1.56)
	P trend			0.38		0.02				0.10		0.61
Beta Tocopherol	Q1	108	28	1.00		1.00		61	26	1.00		1.00
	Q2	74	22	0.69	(0.49, 0.96)	0.78	(0.43, 1.41)	42	22	0.67	(0.44, 1.03)	0.84	(0.47, 1.53)
	Q3	66	19	0.63	(0.45, 0.89)	0.73	(0.39, 1.34)	37	27	0.66	(0.42, 1.02)	1.14	(0.65, 2.02)
	P trend			<.01		0.30				0.05		0.65
Gamma Tocophero	Q1	87	22	1.00		1.00		48	23	1.00		1.00
	Q2	77	21	0.80	(0.57, 1.13)	0.76	(0.40, 1.43)	37	29	0.66	(0.42, 1.05)	1.20	(0.67, 2.14)
	Q3	84	26	0.89	(0.63, 1.26)	0.88	(0.47, 1.65)	55	23	0.92	(0.60, 1.43)	0.91	(0.48, 1.71)
	P trend			0.52		0.96				0.78		0.75

Adjusted for age, sex, BMI, long-term activity level, pack-years of cigarette smoking, dietary calcium, and energy intake.

Assessment of interaction between the IL6 rs1800796 polymorphism and dietary factors showed significant interaction with lutein/zeozanthine and KRAS2 mutations. Those with the GG genotype were at reduced risk of a KRAS2 mutation when they consumed high levels of dietary lutein/zeozanthine (Table 4). Alpha, beta, and gamma tocopherol interacted with IL6 genotype to alter risk for TP53 and KRAS2 mutations. The GG genotype reduced risk of both TP53 and KRAS2 mutations when tocopherol levels were high which the GC and CC genotypes reduced risk when these tocopherols were low. Among those with a GC or CC genotype, there was a twofold increased risk of a TP53 mutation when alpha and gamma tocopherol intake was high. There was no significant interaction between aspirin and ibuprofen-type drug use and any dietary factors (data not shown in table).

Table 4.

Interaction between IL6 rs1800796 genotype and carotenoids and tocopherols associated with TP53 and KRAS2 mutations

IL6 rs1800796 genotype:		TP53 Mut vs. Control				KRAS2 Mut vs. Control
		GG		GC and CC		GG		GC and CC
		OR	(95% CI)	OR	(95% CI)	OR	(95% CI)	OR	(95% CI)
Lutein & Zeozanthine	T1	1.00		1.75	(0.97, 3.13)	1.00		0.62	(0.25, 1.54)
	T2	0.94	(0.66, 1.34)	1.00	(0.52, 1.92)	0.78	(0.52, 1.16)	0.62	(0.26, 1.45)
	T3	0.90	(0.62, 1.31)	2.13	(1.24, 3.65)	0.61	(0.39, 0.96)	1.18	(0.59, 2.34)
	P Interaction	0.64				0.02
Alpha Tocopherol	T1	1.00		1.11	(0.63, 1.94)	1.00		0.32	(0.12, 0.85)
	T2	0.92	(0.65, 1.31)	1.53	(0.85, 2.75)	0.66	(0.44, 1.01)	1.20	(0.59, 2.42)
	T3	0.62	(0.42, 0.91)	1.90	(1.04, 3.46)	0.60	(0.38, 0.92)	0.99	(0.45, 2.19)
	P Interaction	<.01				<.01
Beta Tocopherol	T1	1.00		1.00	(0.56, 1.76)	1.00		0.29	(0.10, 0.85)
	T2	0.58	(0.40, 0.83)	1.34	(0.76, 2.36)	0.52	(0.34, 0.79)	0.97	(0.48, 1.96)
	T3	0.57	(0.39, 0.82)	1.32	(0.74, 2.36)	0.61	(0.40, 0.94)	0.95	(0.45, 2.02)
	P Interaction	0.06				<.01
Gamma Tocopherol	T1	1.00		1.20	(0.69, 2.06)	1.00		0.46	(0.19, 1.13)
	T2	0.72	(0.50, 1.05)	1.27	(0.67, 2.41)	0.75	(0.49, 1.17)	1.28	(0.60, 2.71)
	T3	0.82	(0.56, 1.20)	2.00	(1.10, 3.66)	0.84	(0.54, 1.31)	1.09	(0.48, 2.47)
	P Interaction	0.05				0.04

Adjusted for age, sex, BMI, long-term activity level, pack-years of cigarette smoking, dietary calcium, and energy intake.

Discussion

These findings suggest that inflammation-related factors, including aspirin, ibuprofen-type drugs, IL6 polymorphisms, and some dietary carotenoids and antioxidants are associated with specific rectal tumor mutations. Associations were more consistent and stronger for TP53 mutations than either CIMP-positive or KRAS2 mutations, lending support to the existing literature that shows TP53 is a target of inflammation in that a mutated TP53 gene is associated with increased inflammatory process while wild-type TP53 demonstrates less inflammatory processes.

Antioxidants include carotenoids such as beta carotene, lutein and zeaxanthin, and lycopene; vitamins such as vitamin C and E; and minerals such as selenium. Although they all have antioxidant properties that include deactivation of free radical, they also have unique properties that may be important as chemopreventive agents. Beta-carotene has been shown to be effective in protecting lipid membranes from free radical damage; lycopene has been found to be the most efficient singlet oxygen quencher of the carotenoids; and lutein and zeaxanthin are generally more effective than beta-carotene as scavengers of oxygen radical species ^35–38. Observation by Enger and colleagues ³⁹ suggest that beta-carotene may work at early stages of tumor development while lutein may work at later stages in the disease process. Our data suggest that beta carotene is most consistently associated with TP53 mutations, especially mutations that were transversions rather than transitions. Lutein and lycopene were associated mainly with KRAS2 mutations and in the case of lutein, the association was modified by IL6 genotype. Although the majority of TP53 mutations are transitions, antioxidants appeared to have stronger associations with transversion mutations. Past studies have shown that TP53 transversion mutations are associated with cigarette smoking ^{40, 41}, which could increase free radical production. Thus, it is logical that antioxidants may have a greater impact on transversion mutations.

Vitamin E represents a group of tocol and tocotrienol derivatives, which may have specific biological mechanisms of action although as a group act as antioxidants ⁴². Although alpha-tocopherol has been the primary tocopherol examined in epidemiologic studies because of its higher biological activity than other dietary tocopherols, higher biological activity does not necessarily equate with higher antioxidant capabilities ^{42, 43}. The gamma form of tocopherol has been hypothesized as being an equal or better antioxidant than the alpha form of tocopherol ⁴³ and may play an important role in the etiology of colorectal cancer because it is preferentially secreted into the intestine and may impact fecal mutagens. We observed consistent associations for alpha and beta tocopherol and TP53, especially transversion mutations. In our previous work we observed stronger associations for tocopherols among women than men ⁴⁴. Our current work suggests that this may be most associated with alpha tocopherol and TP53 mutations where we observed a significant difference in effect by gender. Studies have shown that estrogen can act as an anti-inflammatory agent and may decrease IL-6 serum levels ^45–47, it is possible that the gender differences observed could operate through these mechanisms.

The −572 (G>C) rs1800796 IL6 polymorphisms has been associated with body size, with individuals with the G allele having smaller waist-to-hip ratios ⁴⁸. We have previously reported that having a C allele of the rs1800796 polymorphisms slightly increased risk of rectal cancer ²³. Our data suggest that this increased risk is associated specifically with TP53 mutations, where the associations becomes much stronger than observed in the population overall. Furthermore IL6 rs1800796 genotype also appears to modify the association between tocopherol intake and rectal cancer risk. Among those with the GG genotype high levels of tocoperol reduce risk of both TP53 and KRAS2 mutated rectal tumors risk.

Our data suggest that factors associated with inflammation relate most consistently with TP53 mutations. It has been suggested by others that TP53 acts as a key molecular element in the inflammatory stress response pathway ⁴⁹. During chronic inflammation, a factor that is an important underlying contributor to colorectal cancer, nitric oxide, induces TP53 accumulation and post-translational modification ^{50, 51} which can lead to the selective clonal expansion of mutated TP53 cells ⁵⁰. Studies have shown that nitric oxide, in an inflammatory condition such as ulcerative colitis, can increase frequency of TP53 mutations ¹⁶. Other studies have shown that the guanine base (G) is highly susceptible to oxidative stress because it has the lowest oxidation potential of the four DNA nucleotide bases, thus transversion mutations involving the G:C > T:A or C:G occur more frequently under oxidative stress conditions ^{52, 53}. Our findings of associations with antioxidants that would reduce oxidative stress would reduce the likelihood of a TP53 mutation, especially a transversion mutation, is consistent with our knowledge of TP53 and inflammation.

CIMP positive tumors also have been proposed as a pathway with an inflammation component ⁴⁹. Given the few number of CIMP positive tumors, measures of association were imprecise. However, we did note a pattern of associations consistent with CIMP positive rectal tumors, some of which were statistically significant despite limited power. However, the association between alpha tocopherol and CIMP positive rectal tumors was significantly different for men and women. It is not clear why this difference would occur and given the low prevalence of CIMP high rectal tumors, we had limited power to evaluate these associations. It is also of interest to note that some of the associations observed for KRAS2-mutated tumors, such as beta carotene, had similar ORs as seen for CIMP, but were only significant for KRAS2 mutations where power was greater.

Our study is one of the largest studies of tumor mutations in rectal cancers that we are aware of. It is a population-based study that includes over 80% of eligible rectal cancer cases. However there are limitations that include lack of APC data and additional indicators of inflammation and inflammation-related exposures. Most importantly a direct measure of individual long-term GI inflammatory state prior to diagnosis was not available, and would be difficult if not impossible to measure in epidemiological studies. Thus, our surrogate measures of inflammation-related factors were used to explore this disease pathway.

In summary, our data support an inflammation-related pathway to rectal cancer, especially for TP53 mutations. Our findings of antioxidants, especially tocopherols and IL6 genotypes specifically associated with a TP53 pathway are support by the literature on TP53, inflammation, and oxidative stress. Confirmation of these findings by other groups is needed. If confirmed, our data suggest that lifestyle factors that could influence inflammation-related stress could therefore be potential chemopreventive agents. Since TP53 is a common mutation in CRC, these avenues towards prevention could potentially reduce the incidence of the disease.