doi:10.1111/codi.12717
Original article
Incidence and risk factors for colorectal neoplasia in patients with oral squamous cell carcinoma H. Kishikawa*, K. Sato†, T. Yamauchi†, A. Katakura†, T. Shibahara†, N. Takano† and J. Nishida* *Department of Gastroenterology, Tokyo Dental College, Ichikawa General Hospital, Ichikawa, Japan and †Oral Cancer Centre, Tokyo Dental College, Ichikawa General Hospital, Ichikawa, Japan Received 27 December 2013; accepted 4 May 2014; Accepted Article online 9 July 2014
Abstract Aim Colorectal adenoma and cancer are not regarded as being associated with primary oral cancer. The aim of this study was to determine whether screening colonoscopy should be performed for patients with oral cancer in addition to the upper gastrointestinal endoscopic screening that is now routinely performed. Method Between 2007 and 2013, 162 patients with oral squamous cell carcinoma were enrolled at Tokyo Dental College, Ichikawa General Hospital, and 136 individuals were assigned to colonoscopic surveillance. Advanced neoplasia was defined as an adenoma ≥ 10 mm, adenoma with villous histology or high-grade dysplasia regardless of size and invasive cancer. Associations between advanced neoplasia and clinical factors, including age, sex, body mass index, physical activity, smoking, alcohol consumption and oral cancer site and staging were determined. Results Advanced neoplasia, including five invasive cancers, was identified in 32 (23.5%) patients. An age- and sex-adjusted multivariate analysis revealed that smoking (Brinkmann index > 400; OR = 3.24, 95% CI = 1.28– 8.18), alcohol consumption (lifetime pure ethanol con-
Introduction An increase in the incidence of oral cancer and oral cancer mortality has been recorded in several geographical areas, including the Indian subcontinent, South America, Oceania, central and eastern Europe, Australia and Japan [1,2]. The primary risk factors for oral cancer are well documented and include past and present tobacco chewing, smoking and consumption of alcohol [3]. Several cohort studies have demonstrated a high incidence Correspondence to: H. Kishikawa, MD, PhD, Department of Gastroenterology, Ichikawa General Hospital, Tokyo Dental College, 5-11-13 Sugano, Ichikawa, Chiba 272-8513, Japan. E-mail: [email protected]
888
sumption > 600 l; OR = 2.84, 95% CI = 1.18–6.79) and a diagnosis of cancer of the floor of the mouth (OR = 7.97, 95% CI = 2.49–25.46) were independent risk factors for advanced colorectal neoplasia. Conclusion The prevalence of advanced colorectal neoplasia is unexpectedly high in patients with oral cancer. It should be recognized as a second primary tumour of oral cancer. Screening of oral cancer patients by colonoscopy should be routine practice, particularly among smokers and patients with a high intake of alcohol and cancer of the floor of the mouth. Keywords Oral cancer, advanced adenoma, colorectal cancer, colonoscopy, second primary tumours What does this study add to the literature? This is the first study suggesting a significantly high prevalence of colorectal neoplasia in patients with oral cancer. Although colonoscopy is not usually recommended for routine evaluation of patients with oral cancer, we recommend that screening for large bowel neoplasia should be offered.
of second primary tumours (SPTs) in patients with oral squamous cell cancer, and the incidence ranges from 8% to 21% [4,5]. SPTs in patients with oral cancer have been reported to occur mainly in the oral cavity, pharyngeal and laryngeal mucosa, lung and oesophagus [5]. Colorectal cancer is the third most common malignancy worldwide and approximately one million new cases are reported annually [6]. The majority of colorectal cancers develop from benign precursor lesions through the so-called adenoma–carcinoma sequence after at least a 10-year interval [7]. Thus, the removal of an adenoma decreases the risk of developing cancer [8]. Several risk factors are positively associated with the prevalence of advanced colonic neoplasia, including
Colorectal Disease ª 2014 The Association of Coloproctology of Great Britain and Ireland. 16, 888–895
H. Kishikawa et al.
smoking, alcohol consumption, increased body mass index (BMI) and a family history of colorectal cancer. Both alcohol and smoking are risk factors that are shared with oral cancer [9,10]. Several investigators have reported that colorectal cancer is one of the major second primary tumours of oral cancer or head and neck squamous cell carcinoma [5,11,12]. Considering these findings together, it is reasonable to hypothesize that colorectal adenoma and cancer may be an important category of SPT in patients with oral cancer. Thus, the removal of these lesions may improve survival in oral cancer. However, little is known about the precise relationship between oral and colon cancers. Currently, colonoscopic screening is not routinely performed for patients with oral cancer. In this study we assessed the incidence of neoplasia in colorectal lesions among patients with oral squamous cell cancer. The aim was to determine whether screening colonoscopy should be performed for patients with oral cancer. An additional goal was the identification of a higher-risk subset of those patients for whom colonoscopic screening would provide the greatest benefit.
Method
Colorectal neoplasia in oral cancer patients
Brinkmann index was calculated by multiplying the number of cigarettes smoked daily by the number of years spent as a smoker. To evaluate the dose–response relationship, subjects were stratified into four subgroups by Brinkmann index (nonsmokers; mild smokers, < 400; moderate smokers, 400–800; heavy smokers, > 800). Subjects who had consumed alcoholic beverages at least once a week for a year were defined as drinkers. The total lifetime pure ethanol consumption was then calculated, and the subjects were stratified into four subgroups (nondrinkers; mild drinkers, < 300 l; moderate drinkers, 300–600 l; heavy drinkers, > 600 l). A family history of colon cancer in first-degree relatives was defined as any self-reported cancer in a subject’s parents, siblings or children. Patients who reported exercising for one or more hours weekly were regarded as ‘physically active’ [16]. Patients were excluded from the study for a past medical history of colon polyp removal or surgery for invasive colon cancer. The study was approved by the Tokyo Dental College, Ichikawa General Hospital Ethics Committee and was conducted according to the principles of the Second Declaration of Helsinki. All patients provided written informed consent prior to enrolment.
Study population
Endoscopy
This study prospectively enrolled 162 patients with tumours of the oral cavity treated at the Oral Cancer Centre, Tokyo Dental College, Ichikawa General Hospital, Japan. All patients were registered between 2007 and 2013 and met the following criteria: (1) over 20 years of age and (2) a histologically confirmed diagnosis of squamous cell carcinoma of the oral cavity according to the ICD-O 10th revision (ICD-C00, C0206) [13]. Clinical stage was defined according to the International Union Against Cancer (UICC) TNM classification [14]. One patient was examined by colonoscopy for abdominal pain and then diagnosed with advanced colon cancer. All the other patients were asymptomatic and did not complain of apparent abdominal symptoms suggesting colorectal cancer. The participants independently completed a structured questionnaire concerning their drinking and smoking habits, past medical history, family history of first-degree relatives with colorectal cancer and physical activity. Their BMI was calculated as weight in kilograms divided by height in metres squared (kg/m2). BMI was classified into two categories: normal (< 23.0 kg/m2) and overweight (≥ 23 kg/m2), which differs slightly from the standard WHO criteria [15]. Subjects who had smoked 20 or more cigarettes per day for a year or more were defined as ‘smokers’. The
Colonoscopy was offered to all patients. All colonoscopies were performed by one experienced gastroenterologist (H.K.) using an electrical panendoscope (type CF260AI; Olympus, Tokyo, Japan). The adenoma size was measured during endoscopy using calibrated open biopsy forceps with a span of 7 mm. Polyps that were removed were measured before fixation. Abnormal findings were classified by the number of polyps, polyp size (≤ 5, 6–9, ≥ 10 mm), histological characteristics (lowgrade or high-grade dysplasia) and location. ‘Proximal’ lesions were located from the caecum to the splenic flexure, and ‘distal’ lesions from the descending colon to the rectum. The location was classified into two groups including ‘proximal’ in which patients with proximal only and both proximal and distal lesions were placed and ‘distal only’, which included patients with distal lesions only. Histopathological examination was carried out by two experienced pathologists and lesions were classified as one of the following: carcinoma, tubular adenoma, tubulovillous adenoma, villous adenoma, serrated adenoma, hyperplastic polyp or miscellaneous. Hyperplastic polyps were not considered to be neoplastic and were therefore not included in the analysis. An advanced adenoma was defined as one that was ≥ 10 mm in diameter or one containing villous histology or high-grade
Colorectal Disease ª 2014 The Association of Coloproctology of Great Britain and Ireland. 16, 888–895
889
H. Kishikawa et al.
Colorectal neoplasia in oral cancer patients
dysplasia regardless of size. Advanced neoplasia was defined as invasive cancer or advanced adenoma [17]. Intramucosal carcinoma and carcinoma in situ were classified as adenoma with high-grade dysplasia. The criterion for diagnosing invasive cancer was invasion of malignant cells beyond the muscularis mucosa. Patients were excluded from the analysis if the endoscopic examination was incomplete (caecum not visualized), but if a colonoscopic examination was incomplete because of an obstructing tumour the patient was included. If a patient had more than one polyp, the most advanced pathological finding of the largest lesion was included in the analysis.
26 excluded for the following criteria: · Declined to participate (n = 15) · Past history of colorectal cancer (n = 2) · Past history of polypectomy (n = 9) Colonoscopy results available (n = 136)
Targeted condition absent (n = 68)
Targeted condition present (all adenomas and invasive cancer) (n = 68)
†Non-advanced adenoma (n = 36)
Statistical analysis
Database management and all statistical analyses were performed with SPSS for Windows software, version 19.0 (SPSS Inc., Chicago, IL, USA). The data were expressed as the mean standard deviation (SD) for continuous variables and as a proportion (%) for categorical variables. The frequency of advanced neoplasia in association with the following clinical findings was evaluated by Fisher’s exact test or a chi-square test: age (< 65 or ≥ 65 years), sex (male or female), BMI (< 23 or ≥ 23 kg/m2), exercise (< 1 or ≥ 1 h weekly), family history of colorectal cancer (yes or no), smoking habits (Brinkmann index < 400 or ≥ 400), total lifetime pure ethanol consumption (< 600 or ≥ 600 l), UICC staging of oral cancer (I, II or III, IV) and site of oral cancer (tongue, gingiva or floor of the mouth). Univariate and age- and sex-adjusted multivariate logistic regression modelling yielding odds ratios (ORs) and 95% confidence intervals (CIs) were used to identify variables with significant independent effects on the prevalence of advanced neoplasia among oral cancer subjects. A two-sided P-value of < 0.05 was considered to be statistically significant.
Results Study participants
Of the 162 patients in this study, 26 were excluded for refusing endoscopic examination, nine were excluded owing to a past medical history of polypectomy and two for past history of advanced colorectal cancer. The caecal intubation rate was 100% in the remaining 136 patients. Sixty-eight colonoscopies showed a cancer or adenomatous polyp and the other 68 colonoscopies were normal (Fig. 1). The mean time from the diagnosis of oral cancer to the screening colonoscopy was 18.5 months [median 5.3 (25th percentile 1.4 months, 75th percentile 18.6 months)].
890
Subjects scheduled for colonoscopy (n = 162)
‡ Advanced adenoma (n = 27)
*Invasive cancer (n = 5)
Figure 1 Overview of the study. SCC, squamous cell carcinoma. †Tubular adenoma smaller than 10 mm without villous histology or high-grade dysplasia. ‡Tubular adenoma of 10 mm or more with villous histology or high-grade dysplasia. *Cancers with invasion of malignant cells beyond the muscularis mucosa.
Characteristics of the study population
There were 95 (69.9%) men and 41 (30.1%) women of mean age of 63.4 12.4 years (Table 1). Additionally, 29.4% of participants engaged in at least 1 h of physical activity per week and 7.4% had a positive family history of colorectal cancer in a first-degree relative. Nondrinkers and mild, moderate and heavy drinkers accounted for 30.1, 13.2, 16.2 and 40.4% of the study population, respectively. The group comprised nonsmokers, mild, moderate and heavy smokers at rates of 33.8, 11.8, 28.7 and 25.7%. The primary tumour sites included tongue (n = 69), gingiva (n = 42), floor of the mouth (n = 17), buccal mucosa (n = 3) and others (n = 5). The other locations included cancers of the alveolus, hard palate and lip. The TNM stages included Tis to T4, N0 to N2 and M0. A total of 85 of the 136 patients (62.5%) had Tis, T1 or T2, and 51 (37.5%) had T3 or T4 tumours. There were 100 (73.5%) N0, 15 (11%) N1 and 21 (15.4%) N2 tumours. Table 1 also lists the demographic and clinical characteristics of patients with and without advanced neoplasia. Colonoscopy results
Of the 136 patients, 68 (50%) had abnormal findings (Table 2). Some type of adenoma or invasive cancer was found in 68 (50%) patients. There were 49 (36%) patients with low-grade and 14 (10.3%) with high-grade
Colorectal Disease ª 2014 The Association of Coloproctology of Great Britain and Ireland. 16, 888–895
H. Kishikawa et al.
Colorectal neoplasia in oral cancer patients
Table 1 Patient and tumour characteristics of the 136 patients with oral squamous cell carcinoma.
All patients n 136 Age (years) Mean SD 63.4 12.4 Median 64.5 Range 26–86 Sex, n (%) Male 95 (69.9) Female 41 (30.1) Body mass index (kg/m2), n (%) Mean SD 23.0 3.0 < 23 70 (51.5) ≥ 23 66 (48.5) Lifetime pure ethanol consumption* (l), n (%) Mean SD 620.7 676.7 Nondrinkers 41 (30.1) Drinkers < 300 18 (13.2) 300–600 22 (16.2) ≥ 600 55 (40.4) Cigarette smoking (Brinkman index)†, n (%) Mean SD 534.8 583.5 Nonsmokers 46 (33.8) Smokers < 400 16 (11.8 400–800 39 (28.7) > 800 35 (25.7) Physical activity, n (%) < 1 h weekly 96 (70.6) ≥ 1 h weekly 40 (29.4) Family history of colorectal cancer, n (%) No 126 (92.6) Yes 10 (7.4) Site of primary tumour, n (%) Tongue 69 (50.7) Gingiva 42 (30.9) Floor of the mouth 17 (12.5) Buccal mucosa 3 (2.2) Other 5 (3.7) Tumour classification, n (%) Tis 9 (6.6) T1 25 (18.4) T2 51 (37.5) T3 17 (12.5) T4 34 (25) Node status, n (%) N0 100 (73.5) N1 15 (11) N2 21 (15.4) N3 0 (0)
Presence of advanced neoplasia 32 68.8 8.6 68 51–86 26 (81.3) 6 (18.8) 22.4 2.9 19 (59.4) 13 (40.6) 977.6 770.7 6 (18.8) 0 (0) 5 (15.6) 21 (65.6) 675.6 435 5 (15.6)
Absence of advanced neoplasia 104 62.0 12.8 64 26–86 69 (66.3) 35 (33.7) 23.2 3.2 50 (48.1) 54 (51.9) 526.7 630.0 37 (35.6) 16 (15.4) 16 (15.4) 35 (33.7) 488.1 617.8 41 (39.4)
2 (6.3) 12 (37.5) 13 (40.6)
15 (14.4) 26 (25) 22 (21.2)
23 (71.9) 9 (28.1)
74 (71.2) 30 (28.8)
30 (93.8) 2 (6.2)
96 (92.3) 8 (7.7)
16 (50) 6 (18.8) 10 (31.3) 0 (0) 0 (0)
53 36 7 3 5
(51.0) (34.6) (6.7) (2.9) (4.8)
3 (9.4) 8 (25) 12 (37.5) 5 (15.6) 4 (12.5)
6 17 39 12 30
(5.8) (16.3) (37.5) (11.5) (28.8)
27 (84.4) 2 (6.3) 3 (9.4) 0 (0)
73 13 18 0
(70.2) (12.5) (17.3) (0)
Colorectal Disease ª 2014 The Association of Coloproctology of Great Britain and Ireland. 16, 888–895
891
H. Kishikawa et al.
Colorectal neoplasia in oral cancer patients
Table 1 (Continued).
All patients UICC staging, n (%) 0 I II III IV
5 28 48 16 39
(3.7) (20.6) (35.3) (11.8) (28.7)
Presence of advanced neoplasia
2 (6.3) 8 (25) 12 (37.5) 4 (12.5) 6 (18.8)
Absence of advanced neoplasia
3 20 36 12 33
(2.9) (19.2) (34.6) (11.5) (31.7)
*Total lifetime ethanol consumption. †Brinkmann index calculated by daily cigarette numbers multiplied by smoking years. Table 2 Characteristics of lesions found at colonoscopy. Patients, n (%) No neoplasia Any adenoma or invasive cancer Adenoma Adenoma histology Low-grade dysplasia High-grade dysplasia Adenoma size Mean SD (mm) ≥ 10 mm 6–9 mm ≤ 5 mm Number of adenomas Mean SD < 3 adenomas ≥ 3 adenomas Adenoma location Proximal Distal only Nonadvanced adenoma Advanced adenoma* Histology Low-grade dysplasia High-grade dysplasia Location Proximal† Distal only‡ Invasive cancer UICC staging I and II III and IV Location of invasive cancer Proximal Distal only Advanced neoplasia (invasive cancer and advanced adenomas)
68 (50) 68 (50) 63 (46.3) 49 (36) 14 (10.3) 8.6 4.2 27 (19.9) 9 (6.6) 27 (19.9) 1.7 0.99 55 (40.4) 8 (5.9) 21 42 36 27
(15.4) (30.9) (26.5) (19.9)
13 (9.6) 14 (10.3) 10 (7.4) 17 (12.5) 5 (3.7) 1 (0.07) 4 (2.9) 2 (1.5) 3 (2.2) 32 (23.5%)
*Includes single adenomas of at least 10 mm in diameter, or any high-grade dysplasia. †Proximal: colon from caecum to and including splenic flexure. ‡Distal only: colon from splenic flexure to rectum.
892
dysplasia. No villous or serrated adenomas were found. The size distribution of the adenomas was 19.9% smaller than 5 mm, 6.6% ranging from 6 to 9 mm and 19.9% larger than 10 mm. There were 8 (5.9%) patients with more than three adenomas. There were 42 (30.9%) patients with adenomas in the distal large bowel only and 21 (15.4%) with proximal lesions. Advanced adenomas were found in 27 patients and all were 10 mm or more in diameter. High-grade dysplasia was found in 14 of the 27 lesions. The remaining 36 patients had nonadvanced adenomas. Five (3.7%) patients had a cancer which was Stage III or IV. Thus 33 (23.5%) patients either had an invasive cancer or an advanced adenoma. No serious complications occurred during or after colonoscopy or polypectomy. Association between the presence of advanced neoplasia and various clinical parameters
Fisher’s exact test or the chi-square test were used to compare risk factors between the advanced neoplasia (32 patients) and nonadvanced neoplasia (104 patients) groups, as shown in Table 3. Univariate analysis demonstrated that age over 65 years, smoking, heavy alcohol consumption and presence of cancer of the floor of the mouth were associated with advanced colorectal cancer. A logistic regression analysis was conducted to further assess risk factors for the coexistence of advanced colonic neoplasia among subjects with oral cancer (Table 4). An age- and sex-adjusted multivariate analysis also revealed that smokers with a Brinkmann index > 400 (OR = 3.25; 95% CI = 1.28–8.18) and subjects with a lifetime alcohol consumption exceeding 600 l (OR = 2.84; 95% CI = 1.18–6.79) were independent risk factors. Moreover, patients who were smokers with an alcohol consumption over 30 l and a Brinkmann index of over 50 had an increased prevalence of advanced neoplasia (OR = 5.01; 95% CI = 1.55– 16.24). Diagnosis of cancer of the floor of the mouth
Colorectal Disease ª 2014 The Association of Coloproctology of Great Britain and Ireland. 16, 888–895
H. Kishikawa et al.
Colorectal neoplasia in oral cancer patients
Table 3 Clinical parameters related to ‘advanced neoplasia’ as determined by Fisher’s extact or chi-square tests.
Principal parameters
Prevalence of advanced neoplasia, n (%)
Age (years) < 65 12/75 ≥ 65 20/61 Gender Female 6/41 Male 26/95 Body mass index ≥ 23 13/66 < 23 19/70 Exercise ≥ 1 h weekly 10/40 < 1 h weekly 22/96 Family history of colorectal cancer Yes 2/10 No 30/126 Smoking Brinkmann index < 400 10/72 Brinkmann index ≥ 400 22/64 Lifetime pure ethanol consumption < 600 l 12/81 ≥ 600 l 20/55 UICC staging of oral cancer I or II 24/81 III or IV 8/55 Site of primary cancer Tongue (+) 16/69 ( ) 16/67 Gingiva (+) 6/42 ( ) 26/94 Floor of mouth (+) 10/17 ( ) 22/119
(16) (32.8)
P-value
< 0.05
(14.6) (27.4)
0.127
(19.7) (27.1)
0.321
(25) (22.9)
0.826
(20) (23.8)
1
(13.9) (34.4)
< 0.01
(14.8) (36.4)
< 0.01
(29.6) (14.5)
0.063
(23.2) (28.4)
1
(14.3) (27.7)
0.125
(58.8) (18.5)
< 0.01
was also an independent predictive factor for advanced colonic neoplasia (OR = 7.97; 95% CI = 2.49–25.46).
Discussion This is the first colonoscopic surveillance study in patients with oral squamous cell carcinoma. Our results show that a high incidence of colorectal neoplasia occurs in oral cancer subjects, and 32 of 136 patients with oral cancer (23.5%) had advanced colonic neoplasias. We also found that smoking, drinking, both drinking and smoking, and the presence of cancer of the floor of the mouth were strongly associated with advanced colonic neoplasia in subjects with oral cancer.
The previously reported results of screening colonoscopy suggested that the prevalence of advanced adenoma and cancer is relatively high in the asymptomatic general risk population over the age of 50 years. The reported incidence rate is approximately 6% in women and 10% in men [18,19]. We therefore consider that the high incidence of colorectal neoplasia is understandable among the extremely high-risk subjects in this study (approximately 69.9% were men, 66.2% smoked and 69.9% consumed alcohol). The subclinical development of colorectal cancer occurs over a long period and advanced colorectal cancer may be asymptomatic. Thus, colorectal cancer may be diagnosed by a screening examination or during investigation of an unrelated illness. This may explain why such a high prevalence of advanced adenoma in oral cancer has not been reported. We speculated that identification of advanced large bowel neoplasia among patients with oral cancer could improve the overall survival rate. There is currently only one previous report describing a high prevalence of SPT in cancer of the floor of the mouth [20]. Its association with advanced adenoma is most likely to be due to the heavy smoking and drinking habits of patients with this type of cancer. Our results regarding the high prevalence of colorectal neoplasia among patients with oral cancer also suggest that we should change the ‘field cancerization’ hypothesis proposed by Slaughter et al. [21]. Based on their hypothesis, the entire upper respiratory–digestive tract is at high risk of malignancy due to prolonged exposure to carcinogens, especially tobacco and alcohol. Epidemiological studies have shown that alcohol and smoking are associated with cancer in many organs besides the head and neck [9,10,22]. Positron emission tomography–computed tomography has recently been used as a tool to detect SPT in oral cancer [23], but its specificity is not acceptable for the detection of advanced colorectal neoplasia [24] and it is not a substitute for colonoscopy. Although few studies have referred to the prevalence of head and neck cancers in patients with colorectal cancer, several studies have suggested possible associations between these diseases [25,26]. The incidence of SPTs occurring in 1304 patients with colorectal cancer was compared with the number of cancers expected to occur in Japan. The data indicate that second primary oral cavity/pharyngeal cancers had a higher incidence than the statistically expected value, which is consistent with our study [25]. A limitation of the present study is that the results are derived from a small population of only Japanese patients without a control group. Larger studies with control data on colonoscopic findings in age- and
Colorectal Disease ª 2014 The Association of Coloproctology of Great Britain and Ireland. 16, 888–895
893
H. Kishikawa et al.
Colorectal neoplasia in oral cancer patients
Table 4 Univariate and age- and gender-adjusted multivariate analyses of independent factors determining ‘advanced neoplasia’ among patients with oral cancer. Prevalence of advanced neoplasia, n (%) Lifetime pure ethanol consumption (l) < 600* 12/81 ≥ 600 20/55 Cigarette smoking (Brinkman index)† < 400 10/72 ≥ 400 22/64 Alcohol drinking and smoking habits‡ Smoking ( ) or alcohol ( ) 6/57 Smoking (+) and alcohol (+) 26/79 Floor of mouth cancer ( ) 22/119 (+) 10/17
OR (95% CI) univariate
P
OR (95% CI) multivariate
P
(14.8) (36.4)
1 3.29 (1.44–7.48)
< 0.01
1 2.84 (1.18–6.79)
< 0.05
(13.9) (34.4)
1 3.25 (1.40–7.55)
< 0.01
1 3.24 (1.28–8.18)
< 0.05
(10.5) (32.9)
1 4.17 (1.59–10.97)
< 0.01
1 5.01 (1.55–16.24)
< 0.01
(18.5) (58.8)
1 6.30 (2.16–18.38)
< 0.01
1 7.97 (2.49–25.46)
< 0.01
*Includes non-nondrinkers and drinkers with alcohol consumption of 600 L. †Brinkmann index calculated by daily cigarette numbers multiplied by smoking years. ‡Alcohol habit was defined as positive if the subject’s alcohol consumption exceeded 30 l. Smoking habit was defined as positive if the subjects had a Brinkmann index more than 50.
sex-matched asymptomatic subjects should be performed to clarify more accurately the factors that can identify patients with colorectal neoplasia. We plan to investigate these factors in the future. It is possible that there is a greater need for colorectal surveillance in patients with oral cancer in First World countries where a higher prevalence of colon cancer has been reported compared with developing countries [27]. In conclusion, there was a significantly high prevalence of advanced adenoma and colorectal cancer in patients with oral cancer. A higher-risk subset of patients with oral cancer can be identified by combining the clinical parameters of cigarette smoking and alcohol consumption with a diagnosis of cancer of the floor of the mouth. Screening of oral cancer patients by colonoscopy should be performed and may decrease mortality from SPTs, especially among high-risk patients.
Acknowledgements We thank Assistant Professor Shogo Kaida, Professor Takashi Matsukubo and Professor Soichiro Miura for editing and reviewing this document. We are also grateful to Professor Jun Miyauchi and Professor Yoichi Tanaka for the pathological evaluation of the biopsy specimens. We appreciate the support provided by the medical and secretarial staff of the Endoscopy Division of Tokyo Dental College, Ichikawa General Hospital for their cooperation. No external financial support was received for this study.
894
Author contributions H. Kishikawa: study conception and design; H. Kishikawa, J. Nishida, T. Yamauchi, K. Sato: development of methodology; H. Kishikawa, T. Yamauchi, K. Sato, S: acquisition of data; H. Kishikawa, K. Sato: analysis and interpretation of data; H. Kishikawa, J. Nishida: drafting the article or revising it critically for important intellectual content; A. Katakura, T. Shibahara, N. Takano: final approval of the version to be published.
Conflicts of interest All authors declare that there are no potential conflicts of interest with respect to the performance of this study or publication of this manuscript.
References 1 Sankaranarayanan R, Ramadas K, Thomas G et al. Trivandrum Oral Cancer Screening Study Group, Effect of screening on oral cancer mortality in Kerala, India: a cluster-randomised controlled trial. Lancet 2005; 365: 1927–33. 2 Matsuda T, Marugame T, Kamo K, Katanoda K, Ajiki W, Sobue T. Japan Cancer Surveillance Research Group (2012) Cancer incidence and incidence rates in Japan in 2006: based on data from 15 population-based cancer registries in the monitoring of cancer incidence in Japan (MCIJ) project. Jpn J Clin Oncol 2012; 42: 139–47. 3 Garrote LF, Herrero R, Reyes RM et al. Risk factors for cancer of the oral cavity and oro-pharynx in Cuba. Br J Cancer 2001; 85: 46–54.
Colorectal Disease ª 2014 The Association of Coloproctology of Great Britain and Ireland. 16, 888–895
H. Kishikawa et al.
4 Crosher R, McIIroy R. The incidence of other primary tumours in patients with oral cancer in Scotland. Br J Oral Maxillofac Surg 1998; 36: 58–62. 5 Day GL, Blot WJ. Second primary tumours in patients with oral cancer. Cancer 1992; 70: 14–9. 6 Parkin DM, Bray F, Ferlay J, Pisani P. Global cancer statistics, 2002. CA Cancer J Clin 2005; 55: 74–108. 7 Muto T, Bussey HJ, Morson BC. The evolution of cancer of the colon and rectum. Cancer 1975; 36: 2251–70. 8 Zauber AG, Winawer SJ, O’Brien MJ et al. Colonoscopic polypectomy and long-term prevention of colorectal-cancer deaths. N Engl J Med 2012; 366: 687–96. 9 Cho E, Smith-Warner SA, Ritz J et al. Alcohol intake and colorectal cancer: pooled analysis of 8 cohort studies. Ann Intern Med 2004; 140: 603–13. 10 Liang PS, Chen TY, Giovannucci E. Cigarette smoking and colorectal cancer incidence and mortality: systematic review and meta-analysis. Int J Cancer 2009; 124: 2406–15. 11 Warnakulasuriya KA, Robinson D, Evans H. Multiple primary tumours following head and neck cancer in southern England during 1961–98. J Oral Pathol Med 2003; 32: 443–9. 12 Hashibe M, Ritz B, Le AD, Li G, Sankaranarayanan R, Zhang ZF. Radiotherapy for oral cancer as a risk factor for second primary cancers. Cancer Lett 2005; 220: 185–95. 13 International Statistical Classification of Diseases and Related Health Problems 10th Revision (ICD-10) Version for 2010. Available at http://apps.who.int/classifications/ icd10/browse/2010/en#/II (accessed April 2014). 14 Sobin LH, Gospodarowicz MK, Wittekind CH TNM Classification of Malignant Tumours. 7th edition. New Jersey: Wiley-Blackwell, 2010. pp. 1–21. 15 World Health Organization. Physical status: the use and interpretation of anthropometry Report of a WHO Expert Committee. World Health Organ Tech Rep Ser 1995; 854: 1–452. 16 Sanchez NF, Stierman B, Saab S, Mahajan D, Yeung H, Francois F. Physical activity reduces risk for colon polyps in a multiethnic colorectal cancer screening population. BMC Res Notes 2012; 5: 312.
Colorectal neoplasia in oral cancer patients
17 Lieberman DA, Weiss DG, Harford WV et al. Five-year colon surveillance after screening colonoscopy. Gastroenterology 2007; 133: 1077–85. 18 Brenner H, Hoffmeister M, Stegmaier C, Brenner G, Altenhofen L, Haug U. Risk of progression of advanced adenomas to colorectal cancer by age and sex: estimates based on 840,149 screening colonoscopies. Gut 2007; 56: 1585–9. 19 Kolligs FT, Crispin A, Munte A, Wagner A, Mansmann U, Goke B. Risk of advanced colorectal neoplasia according to age and gender. PLoS One 2011; 6: e20076. 20 Kramer FJ, Janssen M, Eckardt A. Second primary tumours in oropharyngeal squamous cell carcinoma. Clin Oral Investig 2004; 8: 56–62. 21 Slaughter DP, Southwick HW, Smejkal W. Field cancerization in oral stratified squamous epithelium. Clinical implications of multicentric origin. Cancer 1953; 6: 963–8. 22 Duell EJ. Epidemiology and potential mechanisms of tobacco smoking and heavy alcohol consumption in pancreatic cancer. Mol Carcinog 2012; 51: 40–52. 23 Weston BR, Iyer RB, Qiao W, Lee JH, Bresalier RS, Ross WA. Ability of integrated positron emission and computed tomography to detect significant colonic pathology: the experience of a tertiary cancer center. Cancer 2010; 116: 1454–61. 24 Stokkel MP, Moons KG, ten Broek FW, van Rijk PP, Hordijk GJ. 18F-fluorodeoxyglucose dual-head positron emission tomography as a procedure for detecting simultaneous primary tumours in cases of head and neck cancer. Cancer 1999; 86: 2370–7. 25 Tamamoto S, Yoshimura K, Ri S, Fujita S, Akasu T, Moriya Y. The risk of multiple primary malignancies with colorectal carcinoma. Dis Colon Rectum 2006; 49: S30–6. 26 Chiang JM, Yeh CY, Changehien CR et al. Clinical features of second other-site primary cancers among sporadic colorectal cancer patients- a hospital based study of 3,722 cases. Hepatogastroenterology 2004; 51: 1341–4. 27 Center MM, Jemal A, Smith RA, Ward E. Worldwide variations in colorectal cancer. CA Cancer J Clin 2009; 59: 366–78.
Colorectal Disease ª 2014 The Association of Coloproctology of Great Britain and Ireland. 16, 888–895
895
Copyright of Colorectal Disease is the property of Wiley-Blackwell and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.
Original article
Incidence and risk factors for colorectal neoplasia in patients with oral squamous cell carcinoma H. Kishikawa*, K. Sato†, T. Yamauchi†, A. Katakura†, T. Shibahara†, N. Takano† and J. Nishida* *Department of Gastroenterology, Tokyo Dental College, Ichikawa General Hospital, Ichikawa, Japan and †Oral Cancer Centre, Tokyo Dental College, Ichikawa General Hospital, Ichikawa, Japan Received 27 December 2013; accepted 4 May 2014; Accepted Article online 9 July 2014
Abstract Aim Colorectal adenoma and cancer are not regarded as being associated with primary oral cancer. The aim of this study was to determine whether screening colonoscopy should be performed for patients with oral cancer in addition to the upper gastrointestinal endoscopic screening that is now routinely performed. Method Between 2007 and 2013, 162 patients with oral squamous cell carcinoma were enrolled at Tokyo Dental College, Ichikawa General Hospital, and 136 individuals were assigned to colonoscopic surveillance. Advanced neoplasia was defined as an adenoma ≥ 10 mm, adenoma with villous histology or high-grade dysplasia regardless of size and invasive cancer. Associations between advanced neoplasia and clinical factors, including age, sex, body mass index, physical activity, smoking, alcohol consumption and oral cancer site and staging were determined. Results Advanced neoplasia, including five invasive cancers, was identified in 32 (23.5%) patients. An age- and sex-adjusted multivariate analysis revealed that smoking (Brinkmann index > 400; OR = 3.24, 95% CI = 1.28– 8.18), alcohol consumption (lifetime pure ethanol con-
Introduction An increase in the incidence of oral cancer and oral cancer mortality has been recorded in several geographical areas, including the Indian subcontinent, South America, Oceania, central and eastern Europe, Australia and Japan [1,2]. The primary risk factors for oral cancer are well documented and include past and present tobacco chewing, smoking and consumption of alcohol [3]. Several cohort studies have demonstrated a high incidence Correspondence to: H. Kishikawa, MD, PhD, Department of Gastroenterology, Ichikawa General Hospital, Tokyo Dental College, 5-11-13 Sugano, Ichikawa, Chiba 272-8513, Japan. E-mail: [email protected]
888
sumption > 600 l; OR = 2.84, 95% CI = 1.18–6.79) and a diagnosis of cancer of the floor of the mouth (OR = 7.97, 95% CI = 2.49–25.46) were independent risk factors for advanced colorectal neoplasia. Conclusion The prevalence of advanced colorectal neoplasia is unexpectedly high in patients with oral cancer. It should be recognized as a second primary tumour of oral cancer. Screening of oral cancer patients by colonoscopy should be routine practice, particularly among smokers and patients with a high intake of alcohol and cancer of the floor of the mouth. Keywords Oral cancer, advanced adenoma, colorectal cancer, colonoscopy, second primary tumours What does this study add to the literature? This is the first study suggesting a significantly high prevalence of colorectal neoplasia in patients with oral cancer. Although colonoscopy is not usually recommended for routine evaluation of patients with oral cancer, we recommend that screening for large bowel neoplasia should be offered.
of second primary tumours (SPTs) in patients with oral squamous cell cancer, and the incidence ranges from 8% to 21% [4,5]. SPTs in patients with oral cancer have been reported to occur mainly in the oral cavity, pharyngeal and laryngeal mucosa, lung and oesophagus [5]. Colorectal cancer is the third most common malignancy worldwide and approximately one million new cases are reported annually [6]. The majority of colorectal cancers develop from benign precursor lesions through the so-called adenoma–carcinoma sequence after at least a 10-year interval [7]. Thus, the removal of an adenoma decreases the risk of developing cancer [8]. Several risk factors are positively associated with the prevalence of advanced colonic neoplasia, including
Colorectal Disease ª 2014 The Association of Coloproctology of Great Britain and Ireland. 16, 888–895
H. Kishikawa et al.
smoking, alcohol consumption, increased body mass index (BMI) and a family history of colorectal cancer. Both alcohol and smoking are risk factors that are shared with oral cancer [9,10]. Several investigators have reported that colorectal cancer is one of the major second primary tumours of oral cancer or head and neck squamous cell carcinoma [5,11,12]. Considering these findings together, it is reasonable to hypothesize that colorectal adenoma and cancer may be an important category of SPT in patients with oral cancer. Thus, the removal of these lesions may improve survival in oral cancer. However, little is known about the precise relationship between oral and colon cancers. Currently, colonoscopic screening is not routinely performed for patients with oral cancer. In this study we assessed the incidence of neoplasia in colorectal lesions among patients with oral squamous cell cancer. The aim was to determine whether screening colonoscopy should be performed for patients with oral cancer. An additional goal was the identification of a higher-risk subset of those patients for whom colonoscopic screening would provide the greatest benefit.
Method
Colorectal neoplasia in oral cancer patients
Brinkmann index was calculated by multiplying the number of cigarettes smoked daily by the number of years spent as a smoker. To evaluate the dose–response relationship, subjects were stratified into four subgroups by Brinkmann index (nonsmokers; mild smokers, < 400; moderate smokers, 400–800; heavy smokers, > 800). Subjects who had consumed alcoholic beverages at least once a week for a year were defined as drinkers. The total lifetime pure ethanol consumption was then calculated, and the subjects were stratified into four subgroups (nondrinkers; mild drinkers, < 300 l; moderate drinkers, 300–600 l; heavy drinkers, > 600 l). A family history of colon cancer in first-degree relatives was defined as any self-reported cancer in a subject’s parents, siblings or children. Patients who reported exercising for one or more hours weekly were regarded as ‘physically active’ [16]. Patients were excluded from the study for a past medical history of colon polyp removal or surgery for invasive colon cancer. The study was approved by the Tokyo Dental College, Ichikawa General Hospital Ethics Committee and was conducted according to the principles of the Second Declaration of Helsinki. All patients provided written informed consent prior to enrolment.
Study population
Endoscopy
This study prospectively enrolled 162 patients with tumours of the oral cavity treated at the Oral Cancer Centre, Tokyo Dental College, Ichikawa General Hospital, Japan. All patients were registered between 2007 and 2013 and met the following criteria: (1) over 20 years of age and (2) a histologically confirmed diagnosis of squamous cell carcinoma of the oral cavity according to the ICD-O 10th revision (ICD-C00, C0206) [13]. Clinical stage was defined according to the International Union Against Cancer (UICC) TNM classification [14]. One patient was examined by colonoscopy for abdominal pain and then diagnosed with advanced colon cancer. All the other patients were asymptomatic and did not complain of apparent abdominal symptoms suggesting colorectal cancer. The participants independently completed a structured questionnaire concerning their drinking and smoking habits, past medical history, family history of first-degree relatives with colorectal cancer and physical activity. Their BMI was calculated as weight in kilograms divided by height in metres squared (kg/m2). BMI was classified into two categories: normal (< 23.0 kg/m2) and overweight (≥ 23 kg/m2), which differs slightly from the standard WHO criteria [15]. Subjects who had smoked 20 or more cigarettes per day for a year or more were defined as ‘smokers’. The
Colonoscopy was offered to all patients. All colonoscopies were performed by one experienced gastroenterologist (H.K.) using an electrical panendoscope (type CF260AI; Olympus, Tokyo, Japan). The adenoma size was measured during endoscopy using calibrated open biopsy forceps with a span of 7 mm. Polyps that were removed were measured before fixation. Abnormal findings were classified by the number of polyps, polyp size (≤ 5, 6–9, ≥ 10 mm), histological characteristics (lowgrade or high-grade dysplasia) and location. ‘Proximal’ lesions were located from the caecum to the splenic flexure, and ‘distal’ lesions from the descending colon to the rectum. The location was classified into two groups including ‘proximal’ in which patients with proximal only and both proximal and distal lesions were placed and ‘distal only’, which included patients with distal lesions only. Histopathological examination was carried out by two experienced pathologists and lesions were classified as one of the following: carcinoma, tubular adenoma, tubulovillous adenoma, villous adenoma, serrated adenoma, hyperplastic polyp or miscellaneous. Hyperplastic polyps were not considered to be neoplastic and were therefore not included in the analysis. An advanced adenoma was defined as one that was ≥ 10 mm in diameter or one containing villous histology or high-grade
Colorectal Disease ª 2014 The Association of Coloproctology of Great Britain and Ireland. 16, 888–895
889
H. Kishikawa et al.
Colorectal neoplasia in oral cancer patients
dysplasia regardless of size. Advanced neoplasia was defined as invasive cancer or advanced adenoma [17]. Intramucosal carcinoma and carcinoma in situ were classified as adenoma with high-grade dysplasia. The criterion for diagnosing invasive cancer was invasion of malignant cells beyond the muscularis mucosa. Patients were excluded from the analysis if the endoscopic examination was incomplete (caecum not visualized), but if a colonoscopic examination was incomplete because of an obstructing tumour the patient was included. If a patient had more than one polyp, the most advanced pathological finding of the largest lesion was included in the analysis.
26 excluded for the following criteria: · Declined to participate (n = 15) · Past history of colorectal cancer (n = 2) · Past history of polypectomy (n = 9) Colonoscopy results available (n = 136)
Targeted condition absent (n = 68)
Targeted condition present (all adenomas and invasive cancer) (n = 68)
†Non-advanced adenoma (n = 36)
Statistical analysis
Database management and all statistical analyses were performed with SPSS for Windows software, version 19.0 (SPSS Inc., Chicago, IL, USA). The data were expressed as the mean standard deviation (SD) for continuous variables and as a proportion (%) for categorical variables. The frequency of advanced neoplasia in association with the following clinical findings was evaluated by Fisher’s exact test or a chi-square test: age (< 65 or ≥ 65 years), sex (male or female), BMI (< 23 or ≥ 23 kg/m2), exercise (< 1 or ≥ 1 h weekly), family history of colorectal cancer (yes or no), smoking habits (Brinkmann index < 400 or ≥ 400), total lifetime pure ethanol consumption (< 600 or ≥ 600 l), UICC staging of oral cancer (I, II or III, IV) and site of oral cancer (tongue, gingiva or floor of the mouth). Univariate and age- and sex-adjusted multivariate logistic regression modelling yielding odds ratios (ORs) and 95% confidence intervals (CIs) were used to identify variables with significant independent effects on the prevalence of advanced neoplasia among oral cancer subjects. A two-sided P-value of < 0.05 was considered to be statistically significant.
Results Study participants
Of the 162 patients in this study, 26 were excluded for refusing endoscopic examination, nine were excluded owing to a past medical history of polypectomy and two for past history of advanced colorectal cancer. The caecal intubation rate was 100% in the remaining 136 patients. Sixty-eight colonoscopies showed a cancer or adenomatous polyp and the other 68 colonoscopies were normal (Fig. 1). The mean time from the diagnosis of oral cancer to the screening colonoscopy was 18.5 months [median 5.3 (25th percentile 1.4 months, 75th percentile 18.6 months)].
890
Subjects scheduled for colonoscopy (n = 162)
‡ Advanced adenoma (n = 27)
*Invasive cancer (n = 5)
Figure 1 Overview of the study. SCC, squamous cell carcinoma. †Tubular adenoma smaller than 10 mm without villous histology or high-grade dysplasia. ‡Tubular adenoma of 10 mm or more with villous histology or high-grade dysplasia. *Cancers with invasion of malignant cells beyond the muscularis mucosa.
Characteristics of the study population
There were 95 (69.9%) men and 41 (30.1%) women of mean age of 63.4 12.4 years (Table 1). Additionally, 29.4% of participants engaged in at least 1 h of physical activity per week and 7.4% had a positive family history of colorectal cancer in a first-degree relative. Nondrinkers and mild, moderate and heavy drinkers accounted for 30.1, 13.2, 16.2 and 40.4% of the study population, respectively. The group comprised nonsmokers, mild, moderate and heavy smokers at rates of 33.8, 11.8, 28.7 and 25.7%. The primary tumour sites included tongue (n = 69), gingiva (n = 42), floor of the mouth (n = 17), buccal mucosa (n = 3) and others (n = 5). The other locations included cancers of the alveolus, hard palate and lip. The TNM stages included Tis to T4, N0 to N2 and M0. A total of 85 of the 136 patients (62.5%) had Tis, T1 or T2, and 51 (37.5%) had T3 or T4 tumours. There were 100 (73.5%) N0, 15 (11%) N1 and 21 (15.4%) N2 tumours. Table 1 also lists the demographic and clinical characteristics of patients with and without advanced neoplasia. Colonoscopy results
Of the 136 patients, 68 (50%) had abnormal findings (Table 2). Some type of adenoma or invasive cancer was found in 68 (50%) patients. There were 49 (36%) patients with low-grade and 14 (10.3%) with high-grade
Colorectal Disease ª 2014 The Association of Coloproctology of Great Britain and Ireland. 16, 888–895
H. Kishikawa et al.
Colorectal neoplasia in oral cancer patients
Table 1 Patient and tumour characteristics of the 136 patients with oral squamous cell carcinoma.
All patients n 136 Age (years) Mean SD 63.4 12.4 Median 64.5 Range 26–86 Sex, n (%) Male 95 (69.9) Female 41 (30.1) Body mass index (kg/m2), n (%) Mean SD 23.0 3.0 < 23 70 (51.5) ≥ 23 66 (48.5) Lifetime pure ethanol consumption* (l), n (%) Mean SD 620.7 676.7 Nondrinkers 41 (30.1) Drinkers < 300 18 (13.2) 300–600 22 (16.2) ≥ 600 55 (40.4) Cigarette smoking (Brinkman index)†, n (%) Mean SD 534.8 583.5 Nonsmokers 46 (33.8) Smokers < 400 16 (11.8 400–800 39 (28.7) > 800 35 (25.7) Physical activity, n (%) < 1 h weekly 96 (70.6) ≥ 1 h weekly 40 (29.4) Family history of colorectal cancer, n (%) No 126 (92.6) Yes 10 (7.4) Site of primary tumour, n (%) Tongue 69 (50.7) Gingiva 42 (30.9) Floor of the mouth 17 (12.5) Buccal mucosa 3 (2.2) Other 5 (3.7) Tumour classification, n (%) Tis 9 (6.6) T1 25 (18.4) T2 51 (37.5) T3 17 (12.5) T4 34 (25) Node status, n (%) N0 100 (73.5) N1 15 (11) N2 21 (15.4) N3 0 (0)
Presence of advanced neoplasia 32 68.8 8.6 68 51–86 26 (81.3) 6 (18.8) 22.4 2.9 19 (59.4) 13 (40.6) 977.6 770.7 6 (18.8) 0 (0) 5 (15.6) 21 (65.6) 675.6 435 5 (15.6)
Absence of advanced neoplasia 104 62.0 12.8 64 26–86 69 (66.3) 35 (33.7) 23.2 3.2 50 (48.1) 54 (51.9) 526.7 630.0 37 (35.6) 16 (15.4) 16 (15.4) 35 (33.7) 488.1 617.8 41 (39.4)
2 (6.3) 12 (37.5) 13 (40.6)
15 (14.4) 26 (25) 22 (21.2)
23 (71.9) 9 (28.1)
74 (71.2) 30 (28.8)
30 (93.8) 2 (6.2)
96 (92.3) 8 (7.7)
16 (50) 6 (18.8) 10 (31.3) 0 (0) 0 (0)
53 36 7 3 5
(51.0) (34.6) (6.7) (2.9) (4.8)
3 (9.4) 8 (25) 12 (37.5) 5 (15.6) 4 (12.5)
6 17 39 12 30
(5.8) (16.3) (37.5) (11.5) (28.8)
27 (84.4) 2 (6.3) 3 (9.4) 0 (0)
73 13 18 0
(70.2) (12.5) (17.3) (0)
Colorectal Disease ª 2014 The Association of Coloproctology of Great Britain and Ireland. 16, 888–895
891
H. Kishikawa et al.
Colorectal neoplasia in oral cancer patients
Table 1 (Continued).
All patients UICC staging, n (%) 0 I II III IV
5 28 48 16 39
(3.7) (20.6) (35.3) (11.8) (28.7)
Presence of advanced neoplasia
2 (6.3) 8 (25) 12 (37.5) 4 (12.5) 6 (18.8)
Absence of advanced neoplasia
3 20 36 12 33
(2.9) (19.2) (34.6) (11.5) (31.7)
*Total lifetime ethanol consumption. †Brinkmann index calculated by daily cigarette numbers multiplied by smoking years. Table 2 Characteristics of lesions found at colonoscopy. Patients, n (%) No neoplasia Any adenoma or invasive cancer Adenoma Adenoma histology Low-grade dysplasia High-grade dysplasia Adenoma size Mean SD (mm) ≥ 10 mm 6–9 mm ≤ 5 mm Number of adenomas Mean SD < 3 adenomas ≥ 3 adenomas Adenoma location Proximal Distal only Nonadvanced adenoma Advanced adenoma* Histology Low-grade dysplasia High-grade dysplasia Location Proximal† Distal only‡ Invasive cancer UICC staging I and II III and IV Location of invasive cancer Proximal Distal only Advanced neoplasia (invasive cancer and advanced adenomas)
68 (50) 68 (50) 63 (46.3) 49 (36) 14 (10.3) 8.6 4.2 27 (19.9) 9 (6.6) 27 (19.9) 1.7 0.99 55 (40.4) 8 (5.9) 21 42 36 27
(15.4) (30.9) (26.5) (19.9)
13 (9.6) 14 (10.3) 10 (7.4) 17 (12.5) 5 (3.7) 1 (0.07) 4 (2.9) 2 (1.5) 3 (2.2) 32 (23.5%)
*Includes single adenomas of at least 10 mm in diameter, or any high-grade dysplasia. †Proximal: colon from caecum to and including splenic flexure. ‡Distal only: colon from splenic flexure to rectum.
892
dysplasia. No villous or serrated adenomas were found. The size distribution of the adenomas was 19.9% smaller than 5 mm, 6.6% ranging from 6 to 9 mm and 19.9% larger than 10 mm. There were 8 (5.9%) patients with more than three adenomas. There were 42 (30.9%) patients with adenomas in the distal large bowel only and 21 (15.4%) with proximal lesions. Advanced adenomas were found in 27 patients and all were 10 mm or more in diameter. High-grade dysplasia was found in 14 of the 27 lesions. The remaining 36 patients had nonadvanced adenomas. Five (3.7%) patients had a cancer which was Stage III or IV. Thus 33 (23.5%) patients either had an invasive cancer or an advanced adenoma. No serious complications occurred during or after colonoscopy or polypectomy. Association between the presence of advanced neoplasia and various clinical parameters
Fisher’s exact test or the chi-square test were used to compare risk factors between the advanced neoplasia (32 patients) and nonadvanced neoplasia (104 patients) groups, as shown in Table 3. Univariate analysis demonstrated that age over 65 years, smoking, heavy alcohol consumption and presence of cancer of the floor of the mouth were associated with advanced colorectal cancer. A logistic regression analysis was conducted to further assess risk factors for the coexistence of advanced colonic neoplasia among subjects with oral cancer (Table 4). An age- and sex-adjusted multivariate analysis also revealed that smokers with a Brinkmann index > 400 (OR = 3.25; 95% CI = 1.28–8.18) and subjects with a lifetime alcohol consumption exceeding 600 l (OR = 2.84; 95% CI = 1.18–6.79) were independent risk factors. Moreover, patients who were smokers with an alcohol consumption over 30 l and a Brinkmann index of over 50 had an increased prevalence of advanced neoplasia (OR = 5.01; 95% CI = 1.55– 16.24). Diagnosis of cancer of the floor of the mouth
Colorectal Disease ª 2014 The Association of Coloproctology of Great Britain and Ireland. 16, 888–895
H. Kishikawa et al.
Colorectal neoplasia in oral cancer patients
Table 3 Clinical parameters related to ‘advanced neoplasia’ as determined by Fisher’s extact or chi-square tests.
Principal parameters
Prevalence of advanced neoplasia, n (%)
Age (years) < 65 12/75 ≥ 65 20/61 Gender Female 6/41 Male 26/95 Body mass index ≥ 23 13/66 < 23 19/70 Exercise ≥ 1 h weekly 10/40 < 1 h weekly 22/96 Family history of colorectal cancer Yes 2/10 No 30/126 Smoking Brinkmann index < 400 10/72 Brinkmann index ≥ 400 22/64 Lifetime pure ethanol consumption < 600 l 12/81 ≥ 600 l 20/55 UICC staging of oral cancer I or II 24/81 III or IV 8/55 Site of primary cancer Tongue (+) 16/69 ( ) 16/67 Gingiva (+) 6/42 ( ) 26/94 Floor of mouth (+) 10/17 ( ) 22/119
(16) (32.8)
P-value
< 0.05
(14.6) (27.4)
0.127
(19.7) (27.1)
0.321
(25) (22.9)
0.826
(20) (23.8)
1
(13.9) (34.4)
< 0.01
(14.8) (36.4)
< 0.01
(29.6) (14.5)
0.063
(23.2) (28.4)
1
(14.3) (27.7)
0.125
(58.8) (18.5)
< 0.01
was also an independent predictive factor for advanced colonic neoplasia (OR = 7.97; 95% CI = 2.49–25.46).
Discussion This is the first colonoscopic surveillance study in patients with oral squamous cell carcinoma. Our results show that a high incidence of colorectal neoplasia occurs in oral cancer subjects, and 32 of 136 patients with oral cancer (23.5%) had advanced colonic neoplasias. We also found that smoking, drinking, both drinking and smoking, and the presence of cancer of the floor of the mouth were strongly associated with advanced colonic neoplasia in subjects with oral cancer.
The previously reported results of screening colonoscopy suggested that the prevalence of advanced adenoma and cancer is relatively high in the asymptomatic general risk population over the age of 50 years. The reported incidence rate is approximately 6% in women and 10% in men [18,19]. We therefore consider that the high incidence of colorectal neoplasia is understandable among the extremely high-risk subjects in this study (approximately 69.9% were men, 66.2% smoked and 69.9% consumed alcohol). The subclinical development of colorectal cancer occurs over a long period and advanced colorectal cancer may be asymptomatic. Thus, colorectal cancer may be diagnosed by a screening examination or during investigation of an unrelated illness. This may explain why such a high prevalence of advanced adenoma in oral cancer has not been reported. We speculated that identification of advanced large bowel neoplasia among patients with oral cancer could improve the overall survival rate. There is currently only one previous report describing a high prevalence of SPT in cancer of the floor of the mouth [20]. Its association with advanced adenoma is most likely to be due to the heavy smoking and drinking habits of patients with this type of cancer. Our results regarding the high prevalence of colorectal neoplasia among patients with oral cancer also suggest that we should change the ‘field cancerization’ hypothesis proposed by Slaughter et al. [21]. Based on their hypothesis, the entire upper respiratory–digestive tract is at high risk of malignancy due to prolonged exposure to carcinogens, especially tobacco and alcohol. Epidemiological studies have shown that alcohol and smoking are associated with cancer in many organs besides the head and neck [9,10,22]. Positron emission tomography–computed tomography has recently been used as a tool to detect SPT in oral cancer [23], but its specificity is not acceptable for the detection of advanced colorectal neoplasia [24] and it is not a substitute for colonoscopy. Although few studies have referred to the prevalence of head and neck cancers in patients with colorectal cancer, several studies have suggested possible associations between these diseases [25,26]. The incidence of SPTs occurring in 1304 patients with colorectal cancer was compared with the number of cancers expected to occur in Japan. The data indicate that second primary oral cavity/pharyngeal cancers had a higher incidence than the statistically expected value, which is consistent with our study [25]. A limitation of the present study is that the results are derived from a small population of only Japanese patients without a control group. Larger studies with control data on colonoscopic findings in age- and
Colorectal Disease ª 2014 The Association of Coloproctology of Great Britain and Ireland. 16, 888–895
893
H. Kishikawa et al.
Colorectal neoplasia in oral cancer patients
Table 4 Univariate and age- and gender-adjusted multivariate analyses of independent factors determining ‘advanced neoplasia’ among patients with oral cancer. Prevalence of advanced neoplasia, n (%) Lifetime pure ethanol consumption (l) < 600* 12/81 ≥ 600 20/55 Cigarette smoking (Brinkman index)† < 400 10/72 ≥ 400 22/64 Alcohol drinking and smoking habits‡ Smoking ( ) or alcohol ( ) 6/57 Smoking (+) and alcohol (+) 26/79 Floor of mouth cancer ( ) 22/119 (+) 10/17
OR (95% CI) univariate
P
OR (95% CI) multivariate
P
(14.8) (36.4)
1 3.29 (1.44–7.48)
< 0.01
1 2.84 (1.18–6.79)
< 0.05
(13.9) (34.4)
1 3.25 (1.40–7.55)
< 0.01
1 3.24 (1.28–8.18)
< 0.05
(10.5) (32.9)
1 4.17 (1.59–10.97)
< 0.01
1 5.01 (1.55–16.24)
< 0.01
(18.5) (58.8)
1 6.30 (2.16–18.38)
< 0.01
1 7.97 (2.49–25.46)
< 0.01
*Includes non-nondrinkers and drinkers with alcohol consumption of 600 L. †Brinkmann index calculated by daily cigarette numbers multiplied by smoking years. ‡Alcohol habit was defined as positive if the subject’s alcohol consumption exceeded 30 l. Smoking habit was defined as positive if the subjects had a Brinkmann index more than 50.
sex-matched asymptomatic subjects should be performed to clarify more accurately the factors that can identify patients with colorectal neoplasia. We plan to investigate these factors in the future. It is possible that there is a greater need for colorectal surveillance in patients with oral cancer in First World countries where a higher prevalence of colon cancer has been reported compared with developing countries [27]. In conclusion, there was a significantly high prevalence of advanced adenoma and colorectal cancer in patients with oral cancer. A higher-risk subset of patients with oral cancer can be identified by combining the clinical parameters of cigarette smoking and alcohol consumption with a diagnosis of cancer of the floor of the mouth. Screening of oral cancer patients by colonoscopy should be performed and may decrease mortality from SPTs, especially among high-risk patients.
Acknowledgements We thank Assistant Professor Shogo Kaida, Professor Takashi Matsukubo and Professor Soichiro Miura for editing and reviewing this document. We are also grateful to Professor Jun Miyauchi and Professor Yoichi Tanaka for the pathological evaluation of the biopsy specimens. We appreciate the support provided by the medical and secretarial staff of the Endoscopy Division of Tokyo Dental College, Ichikawa General Hospital for their cooperation. No external financial support was received for this study.
894
Author contributions H. Kishikawa: study conception and design; H. Kishikawa, J. Nishida, T. Yamauchi, K. Sato: development of methodology; H. Kishikawa, T. Yamauchi, K. Sato, S: acquisition of data; H. Kishikawa, K. Sato: analysis and interpretation of data; H. Kishikawa, J. Nishida: drafting the article or revising it critically for important intellectual content; A. Katakura, T. Shibahara, N. Takano: final approval of the version to be published.
Conflicts of interest All authors declare that there are no potential conflicts of interest with respect to the performance of this study or publication of this manuscript.
References 1 Sankaranarayanan R, Ramadas K, Thomas G et al. Trivandrum Oral Cancer Screening Study Group, Effect of screening on oral cancer mortality in Kerala, India: a cluster-randomised controlled trial. Lancet 2005; 365: 1927–33. 2 Matsuda T, Marugame T, Kamo K, Katanoda K, Ajiki W, Sobue T. Japan Cancer Surveillance Research Group (2012) Cancer incidence and incidence rates in Japan in 2006: based on data from 15 population-based cancer registries in the monitoring of cancer incidence in Japan (MCIJ) project. Jpn J Clin Oncol 2012; 42: 139–47. 3 Garrote LF, Herrero R, Reyes RM et al. Risk factors for cancer of the oral cavity and oro-pharynx in Cuba. Br J Cancer 2001; 85: 46–54.
Colorectal Disease ª 2014 The Association of Coloproctology of Great Britain and Ireland. 16, 888–895
H. Kishikawa et al.
4 Crosher R, McIIroy R. The incidence of other primary tumours in patients with oral cancer in Scotland. Br J Oral Maxillofac Surg 1998; 36: 58–62. 5 Day GL, Blot WJ. Second primary tumours in patients with oral cancer. Cancer 1992; 70: 14–9. 6 Parkin DM, Bray F, Ferlay J, Pisani P. Global cancer statistics, 2002. CA Cancer J Clin 2005; 55: 74–108. 7 Muto T, Bussey HJ, Morson BC. The evolution of cancer of the colon and rectum. Cancer 1975; 36: 2251–70. 8 Zauber AG, Winawer SJ, O’Brien MJ et al. Colonoscopic polypectomy and long-term prevention of colorectal-cancer deaths. N Engl J Med 2012; 366: 687–96. 9 Cho E, Smith-Warner SA, Ritz J et al. Alcohol intake and colorectal cancer: pooled analysis of 8 cohort studies. Ann Intern Med 2004; 140: 603–13. 10 Liang PS, Chen TY, Giovannucci E. Cigarette smoking and colorectal cancer incidence and mortality: systematic review and meta-analysis. Int J Cancer 2009; 124: 2406–15. 11 Warnakulasuriya KA, Robinson D, Evans H. Multiple primary tumours following head and neck cancer in southern England during 1961–98. J Oral Pathol Med 2003; 32: 443–9. 12 Hashibe M, Ritz B, Le AD, Li G, Sankaranarayanan R, Zhang ZF. Radiotherapy for oral cancer as a risk factor for second primary cancers. Cancer Lett 2005; 220: 185–95. 13 International Statistical Classification of Diseases and Related Health Problems 10th Revision (ICD-10) Version for 2010. Available at http://apps.who.int/classifications/ icd10/browse/2010/en#/II (accessed April 2014). 14 Sobin LH, Gospodarowicz MK, Wittekind CH TNM Classification of Malignant Tumours. 7th edition. New Jersey: Wiley-Blackwell, 2010. pp. 1–21. 15 World Health Organization. Physical status: the use and interpretation of anthropometry Report of a WHO Expert Committee. World Health Organ Tech Rep Ser 1995; 854: 1–452. 16 Sanchez NF, Stierman B, Saab S, Mahajan D, Yeung H, Francois F. Physical activity reduces risk for colon polyps in a multiethnic colorectal cancer screening population. BMC Res Notes 2012; 5: 312.
Colorectal neoplasia in oral cancer patients
17 Lieberman DA, Weiss DG, Harford WV et al. Five-year colon surveillance after screening colonoscopy. Gastroenterology 2007; 133: 1077–85. 18 Brenner H, Hoffmeister M, Stegmaier C, Brenner G, Altenhofen L, Haug U. Risk of progression of advanced adenomas to colorectal cancer by age and sex: estimates based on 840,149 screening colonoscopies. Gut 2007; 56: 1585–9. 19 Kolligs FT, Crispin A, Munte A, Wagner A, Mansmann U, Goke B. Risk of advanced colorectal neoplasia according to age and gender. PLoS One 2011; 6: e20076. 20 Kramer FJ, Janssen M, Eckardt A. Second primary tumours in oropharyngeal squamous cell carcinoma. Clin Oral Investig 2004; 8: 56–62. 21 Slaughter DP, Southwick HW, Smejkal W. Field cancerization in oral stratified squamous epithelium. Clinical implications of multicentric origin. Cancer 1953; 6: 963–8. 22 Duell EJ. Epidemiology and potential mechanisms of tobacco smoking and heavy alcohol consumption in pancreatic cancer. Mol Carcinog 2012; 51: 40–52. 23 Weston BR, Iyer RB, Qiao W, Lee JH, Bresalier RS, Ross WA. Ability of integrated positron emission and computed tomography to detect significant colonic pathology: the experience of a tertiary cancer center. Cancer 2010; 116: 1454–61. 24 Stokkel MP, Moons KG, ten Broek FW, van Rijk PP, Hordijk GJ. 18F-fluorodeoxyglucose dual-head positron emission tomography as a procedure for detecting simultaneous primary tumours in cases of head and neck cancer. Cancer 1999; 86: 2370–7. 25 Tamamoto S, Yoshimura K, Ri S, Fujita S, Akasu T, Moriya Y. The risk of multiple primary malignancies with colorectal carcinoma. Dis Colon Rectum 2006; 49: S30–6. 26 Chiang JM, Yeh CY, Changehien CR et al. Clinical features of second other-site primary cancers among sporadic colorectal cancer patients- a hospital based study of 3,722 cases. Hepatogastroenterology 2004; 51: 1341–4. 27 Center MM, Jemal A, Smith RA, Ward E. Worldwide variations in colorectal cancer. CA Cancer J Clin 2009; 59: 366–78.
Colorectal Disease ª 2014 The Association of Coloproctology of Great Britain and Ireland. 16, 888–895
895
Copyright of Colorectal Disease is the property of Wiley-Blackwell and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.
