- Email: [email protected]
Tobacco smoking, alcohol drinking, and the risk of different histological types of nasopharyngeal cancer in a low-risk population
Oral Oncology 47 (2011) 541–545
Contents lists available at ScienceDirect
Oral Oncology journal homepage: www.elsevier.com/locate/oraloncology
Tobacco smoking, alcohol drinking, and the risk of different histological types of nasopharyngeal cancer in a low-risk population Jerry Polesel a,⇑, Silvia Franceschi b, Renato Talamini a, Eva Negri c, Luigi Barzan d, Maurizio Montella e, Massimo Libra f, Emanuela Vaccher g, Giovanni Franchin h, Carlo La Vecchia c,i, Diego Serraino a a
Unit of Epidemiology and Biostatistics, Centro di Riferimento Oncologico, IRCCS, Via F. Gallini 2, 33081 Aviano (PN), Italy International Agency for Research on Cancer, 150 Cours Albert Thomas, 69372 Lyon, France Departement of Epidemiology, The Mario Negri Institute for Pharmacological Research, Via G. La Masa 19, 20156 Milan, Italy d Department of Head and Neck Surgery, Azienda Ospedaliera ‘‘Santa Maria degli Angeli’’, Via Montereale 24, 33170 Pordenone, Italy e Unit of Epidemiology, Istituto Tumori ‘‘Fondazione Pascale’’, IRCCS, Cappella dei Cangiani, 80100 Naples, Italy f Departement of Biomedical Sciences, Università di Catania, Via Androne 83, 95124 Catania, Italy g Division of Medical Oncology A, Centro di Riferimento Oncologico, IRCCS, Via F. Gallini 2, 33081 Aviano, Italy h Division of Radiation Oncology, Centro di Riferimento Oncologico, IRCCS, Via F. Gallini 2, 33081 Aviano, Italy i Department of Occupational Medicine, Università degli Studi di Milano, Via Vanzetti 5, 20133 Milan, Italy b c
a r t i c l e
i n f o
Article history: Received 16 December 2010 Received in revised form 11 March 2011 Accepted 14 March 2011 Available online 7 April 2011 Keywords: Alcohol drinking Case-control study Nasopharyngeal carcinoma Tobacco smoking
s u m m a r y Nasopharyngeal carcinoma (NPC) is rare in western Countries. Tobacco smoking is a well-recognised risk factor, whereas the role of alcohol drinking is still in debate. We conducted a hospital-based case-control study in Italy on 150, histologically-confirmed, NPC cases of Caucasian ethnicity, aged 18–76 years, including 118 undifferentiated NPCs and 22 differentiated squamous-cell NPC. Controls were 450 Caucasian cancer-free patients admitted to general hospitals for acute conditions. Cases and controls were matched according to sex, age, and place of residence. Logistic regression was used to estimate odds ratios (OR) and the corresponding 95% confidence intervals (CI) while adjusting for known confounders. No significant association emerged between tobacco smoking and all NPCs (OR for current vs. never smokers = 1.52; 95% CI: 0.89–2.60). Conversely, for differentiated NPC only, statistically significant elevated OR were associated with increasing smoking intensity (OR for P15 cigarettes/day = 5.40; 95% CI: 1.34–21.76) and duration of the habit (OR for P32 years = 4.48; 95% CI: 1.11–18.04). Although alcohol drinking was not, per se, significantly associated to NPC risk, the combination of tobacco smoking and alcohol drinking accounted for 57% of differentiated NPCs, whereas it accounted for only 14% of undifferentiated carcinomas. Our findings suggest that, in western populations, NPC includes two separate entities: the differentiated NPC, associated with tobacco smoking like other cancers of head and neck, and the undifferentiated NPC, upon which tobacco smoking has little or no influence. Ó 2011 Elsevier Ltd. All rights reserved.
Introduction Nasopharyngeal carcinoma (NPC) is a rare cancer in Europe and North America, occurring in less than one case per 100,000 people each year. It is approximately 2-fold more frequent in men than in women.1 Very high incidence rates were reported in China, in South-eastern Asia, in the Arctic region,2 and in immigrants from these high-risk areas.2 The World Health Organisation has recognised three histological types of NPC.3,4 Type I, representing squamous cell carcinoma, is similar to carcinomas arising from other sites of head and neck. ⇑ Corresponding author. Tel.: +39 0434 659195; fax: +39 0434 659231. E-mail address: [email protected] (J. Polesel). 1368-8375/$ - see front matter Ó 2011 Elsevier Ltd. All rights reserved. doi:10.1016/j.oraloncology.2011.03.017
Type II is characterised as non-keratinizing carcinoma. Type III, representing undifferentiated carcinoma (also referred to as lymphoepithelioma or Schminke tumour), is characterised by the tumour infiltration of non-cancerous T-lymphocytes. In low-risk areas, undifferentiated carcinoma is the most frequent type, accounting for approximately 75% of new cases5; squamous cell carcinoma represents approximately 25% of cases, being nonkeratinizing carcinomas very rare.5 In high-incidence areas, infection with Epstein-Barr virus (EBV) is a well-established risk factor for NPC,6 particularly undifferentiated NPC7 and when the infection occurs early in life.2,4 Conversely, in low-incidence areas, the fraction of NPCs attributable to EBV is unclear, and NPC has been related to classic risk factors for head and neck cancer, notably tobacco smoking.4
542
J. Polesel et al. / Oral Oncology 47 (2011) 541–545
In 2004, the International Agency for Research of Cancer established a causal role for tobacco smoking in NPC onset,8 an evidence supported by investigations conducted in recent years.9–15 However, the majority of the researches investigating the aetiology of NPC were carried out in high-risk populations, with few investigations conducted in low-risk areas (four case-control studies16–19 and a cohort one20 in the United States (US)). These latter studies showed an excess of NPC among smokers, with up to 6-fold higher risks in heavy smokers than in never smokers. Only one19 of the US studies provided risk estimates according to histological types: the findings suggested that the association with smoking was much stronger with differentiated than with undifferentiated NPC. Studies evaluating the association between alcohol intake and NPC risk showed inconsistent results. A recent meta-analysis found a 33% increase in NPC risk among heavy drinkers,21 in the absence, however, of a clear dose–response relationship. Moreover, as for tobacco smoking, epidemiological evidence from low-risk areas is limited, and it is based exclusively on selected US populations.16,17,19,22 To investigate the impact of tobacco smoking and alcohol drinking on NPC risk in a low-risk European population, we conducted a case-control study in Italy taking into consideration the associations with different histological types.
Materials and methods Between 1992 and 2008, we conducted a case-control study on NPC within an established network of collaborating centres, including Aviano and Pordenone in North-eastern Italy, and Naples and Catania in southern Italy. Cases were 150 Caucasian patients aged 18–76 years (median age: 52 years) with incident NPC, admitted to major general hospitals in the study areas. All 150 NPC cases were histologically confirmed, with 118 (78.7%) undifferentiated NPCs (WHO type 3),3 22 (14.7%) keratinizing squamous cell carcinomas (WHO type 1)3 (here referred to as differentiated NPCs), and 10 (6.7%) not otherwise specified NPCs. EBV status was available for 61 NPC cases based on the detection of EBV nuclear antigen (EBNA) in tissue samples. All 57 undifferentiated NPCs and two out of four differentiated NPCs were EBV-positive. Three controls were frequency-matched to each case, according to sex, age (<40, 40–44, 45–49, 50–54, 55–59, 60–64, 65–69 and P70 years), and place of residence (northern Italy, central Italy, southern Italy and Islands). The control group included 450 Caucasian patients (aged 19–76 years; median age: 52 years) admitted
for a wide spectrum of acute conditions to the same hospitals where cases had been interviewed. Subjects admitted for malignant neoplasms, conditions related to tobacco smoking or alcohol consumption, or any disorder that might have induced long-term modification of diet were excluded from the control group. Controls were admitted for traumatic orthopaedic disorders (37%), other orthopaedic disorders (35%), acute surgical conditions (18%), and miscellaneous other illnesses, including eye, nose, ear, skin, or dental disorders (10%). All study participants signed an informed consent, according to the recommendations of the internal Boards of Ethics of study hospitals. Trained interviewers administered a structured questionnaire to cases and controls during their hospital stays, thus keeping refusal below 3%. The questionnaire collected information on sociodemographic factors; lifestyle habits; diet; a problem-oriented medical history; and family history of cancer. Information on smoking included lifetime status (i.e., never, former, or current smoker), daily number of cigarettes/cigars and grams of tobacco pipe smoked, age at starting and duration of the habit. In our computations, one gram of pipe-smoked tobacco corresponded to one cigarette, and one cigar to three cigarettes. Smokers were subjects who had smoked at least one cigarette/day for at least 1 year. Former smokers were defined as smokers who had abstained from cigarette smoking for at least 12 months before interview. Lifetime alcohol drinking status (i.e., abstainer, former, or current drinker) and the weekly number of drinks for the five most common alcoholic beverages or groups of beverages (i.e., wine, beer, herb liquors, grappa, and spirits) were assessed. Taking into account the different ethanol concentration, one drink corresponded to approximately 125 ml of wine, 330 ml of beer, and 30 ml of hard liquor (i.e., about 12 g of ethanol). Ages at start drinking and at quitting, for ex-drinkers, were also recorded. Former drinkers had abstained from any type of drinking for at least 12 months. The validity and reproducibility of questions on self-reported drinking and smoking habits in our study population were satisfactory.23,24 Odds ratios (OR) and the corresponding 95% confidence intervals (CI) were calculated by means of multiple logistic regression models including terms for sex, age, place of residence, year of interview, education level (i.e.,<7, 7–11, and P12 years), smoking and drinking habits, as specified in table footnotes.25 The test for trend was based on the likelihood-ratio test between the models with and without linear terms for each variable of interest.25 Percent attributable risks were computed according to the method described by Bruzzi et al.26 which used the distribution of the risk factors in NPC cases. All analyses were performed on all NPCs
Table 1 Distribution of 150 cases of nasopharyngeal carcinoma (NPC) and 450 controls according to matching variables.
Sex Men Women
Controls (n = 450)
All NPCsa (n = 150)
Undifferentiated (n = 118)
Differentiated (n = 22)
n
(%)
n
(%)
n
n
(%)
357 93
(79.3) (20.7)
119 31
(79.3) (20.7)
93 25
(78.8) (21.2) v for heterogeneity: p = 0.53
16 6
(72.7) (27.3)
Histological type
(%)
2
Age (years) <44 45–54 55–64 P65
111 138 117 84
(24.7) (30.7) (26.0) (18.7)
37 46 39 28
(24.7) (30.7) (26.0) (18.7)
32 38 29 19
(27.1) (32.2) (24.6) (16.1) v2 for heterogeneity: p = 0.06
3 5 7 7
(13.6) (22.7) (31.8) (31.8)
Place of residence North-East of Italy Centre, South, Islands
270 180
(60.0) (40.0)
90 60
(60.0) (40.0)
73 45
13 9
(59.1) (40.9)
(61.9) (38.1) v for heterogeneity: p = 0.81 2
a
Including 10 not otherwise specified NPCs.
543
J. Polesel et al. / Oral Oncology 47 (2011) 541–545
and, separately, for differentiated and undifferentiated NPCs versus all controls. Results Table 1 shows the distribution of cases (overall and by histological type) and controls according to study design variables. The distribution of undifferentiated and differentiated NPC cases by sex and area of residence was similar. Undifferentiated NPCs tended, however, to be younger than differentiated NPC cases (p = 0.06). Current tobacco smoking was not significantly associated with NPC risk (OR = 1.52; 95% CI: 0.89–2.60), but the association differed according to histological type (Table 2). The risk of current smokers versus never smokers seemed to be lower for undifferentiated NPC (OR = 1.37; 95% CI: 0.74–2.52) than for differentiated NPC (OR = 3.17; 95% CI: 0.91–11.03). Contrary to undifferentiated NPC, differentiated NPC also showed a clear relationship of cancer risk with intensity (OR for P15 cigarettes/day = 5.40; 95% CI: 1.34– 21.76) and duration (OR for P30 years = 4.48; 95% CI: 1.11–18.04; Table 2) of the habit. No clear relationship emerged between drinking status and NPC risk (Table 3). Current alcohol drinking of 28 drinks/week or more was not significantly associated with NPC (OR = 1.91, 95% CI: 0.83– 4.41; p for trend = 0.06). Duration of alcohol drinking (p for trend = 0.74) and age at beginning (p for trend = 0.97) were also unrelated to NPC risk, and no differences emerged according to histological type. The interaction of heavy tobacco smoking with alcohol drinking was also evaluated. Compared to never smokers who drunk <14 drinks/week, the ORs among subjects who smoked P15 cigarettes/day and drunk P28 drinks/week were 1.98 (95% CI: 0.84– 4.68) for undifferentiated NPC and 8.86 (95% CI: 1.47–53.20) for differentiated NPC. The fraction of NPC, overall and by histological type, attributable to heavy tobacco smoking and heavy alcohol drinking is
shown in Table 4. Although the confidence intervals of attributable fractions were broad due to small numbers, the contribution of tobacco smoking in differentiated NPC (55%; 95% CI: 13–98%) was larger than that in undifferentiated NPC (10%; 95% CI: 0–40%). Altogether, heavy tobacco smoking and heavy alcohol drinking accounted for 57% (95% CI: 16–98%) of differentiated NPCs, but only for 14% (95% CI: 0–42%%) of undifferentiated NPCs. Discussion The present study investigated, for the first time in a European country, the role of tobacco smoking and alcohol drinking in the aetiology of NPC. The findings provided additional evidence that, in western populations, NPC includes two separate entities: one, differentiated NPCs, clearly associated to tobacco smoking like other cancers of head and neck, and the other, represented by undifferentiated carcinoma, on which tobacco smoking has little or no influence. Likewise, the association of cancer risk with intensity, duration, and young age at starting smoking was substantially stronger for differentiated than for undifferentiated NPC, in agreement with one previous study from the US.19 The difference in the strength of the association of NPC (any histological type) with smoking in previous studies8–15 may derive, therefore, from a different proportion of differentiated and undifferentiated NPCs in various study populations. The only report that provided information on risk factors for NPC separately by histological type in a western Country showed, as we did, a stronger influence of smoking on differentiated than on undifferentiated NPC.19 However, differentiated NPCs represented 66.6% of cancer cases in the study by Vaugham et al.19 whereas they accounted for only 14.7% in the present study. Lower proportion of undifferentiated NPC was reported in Italy and, therefore, the high proportion in our present study is likely to derive from the selective referral of this rare malignancy to the specialised Ear, Nose & Throat (ENT) Clinics where the present study was conducted.
Table 2 Distribution of 150 cases of nasopharyngeal carcinoma (NPC) and 450 matched controls, odds ratio (OR)a and corresponding 95% confidence interval (CI), according to smoking habits. Controls (n = 450)
All NPCsb (n = 150)
Undifferentiated (n = 118)
Differentiated (n = 22)
n
(%)
n
(%)
OR (95% CI)
n
(%)
OR (95% CI)
n
(%)
OR (95% CI)
(32.2) (29.8) (38.0)
39 48 63
(26.0) (32.0) (42.0)
1d 1.43 (0.81–2.52) 1.52 (0.89–2.60)
34 38 46
(28.8) (32.2) (39.0)
1d 1.43 (0.76–2.70) 1.37 (0.74–2.52)
4 7 11
(18.2) (31.8) (50.0)
1d 2.36 (0.60–9.36) 3.17 (0.91–11.03)
Intensity among current smokers (cigarettes/day)c <15 66 (14.7) 27 P15 105 (23.3) 34 v2 trend
(18.0) (22.7)
0.96 (0.45–2.05) 1.98 (1.05–3.75) 4.13; p = 0.04
20 24
(16.9) (20.3)
0.82 (0.33–2.03) 1.76 (0.83–3.73) 1.89; p = 0.17
5 6
(22.7) (27.2)
2.17 (0.43–10.92) 5.40 (1.34–21.76) 5.64; p = 0.02
Duration among current smokers (years)c <30 74 (19.1) 26 P30 97 (18.9) 37 2 v trend
(17.3) (24.7)
1.26 (0.60–2.64) 1.85 (0.94–3.64) 2.86; p = 0.09
21 25
(17.8) (21.2)
1.25 (0.53–2.96) 1.55 (0.71–3.39) 1.18; p = 0.28
3 8
(13.6) (36.4)
2.51 (0.42–15.17) 4.48 (1.11–18.04) 4.44; p = 0.04
Age at beginning among current smokers (years)c P18 92 (20.4) 25 <18 79 (17.6) 38 v2 trend
(16.7) (25.3)
1.23 (0.64–2.37) 2.09 (1.06–4.13) 4.33; p = 0.04
19 27
(16.1) (22.9)
1.16 (0.55–2.47) 1.81 (0.81–4.04) 1.98; p = 0.16
4 7
(18.2) (31.8)
2.20 (0.50–9.60) 8.81 (1.87–41.55) 7.30; p < 0.01
Year since quitting among former smokers (years) P15 60 (13.3) 25 <15 74 (16.4) 23 2 v trend
(16.7) (15.3)
1.81 (0.87–3.76) 1.40 (0.71–2.77)
19 19
(16.1) (16.1)
1.62 (0.72–3.67) 1.54 (0.72–3.23) 1.39; p = 0.24
4 3
(18.2) (13.6)
2.44 (0.40–15.09) 1.43 (0.23–8.91) 0.13; p = 0.72
Smoking status Never 145 Former 134 Current 171
a b c d
Histological type
Adjusted for sex, age, place of residence, year of interview, education, and total alcohol intake. Including 10 not otherwise specified NPCs. The sum does not add up to the total because of missing values. Reference category.
544
J. Polesel et al. / Oral Oncology 47 (2011) 541–545
Table 3 Distribution of 150 cases of nasopharyngeal carcinoma (NPC) and 450 matched controls, odds ratio (OR)a and corresponding 95% confidence interval (CI), according to drinking habits. Controls(n = 450)
All NPCsb (I = 150)
n
n
(%)
Drinking status Never 54 (12.0) 16 Former 25 (5.6) 5 Current 371 (82.4) 129 Intensity among current drinkers (drinks/week) <14 124 (27.6) 40 14–27 119 (26.4) 34 P28 128 (28.4) 55 2 v trend Duration among current drinkers (years)c <25 91 (20.2) 40 25–39 155 (34.4) 48 P40 124 (27.6) 40 v2 trend Age of beginning among current drinkers (years)c P21 105 (23.3) 41 18–20 152 (33.8) 57 <18 113 (25.1) 30 v2 trend a b c d
Histological type Undifferentiated (n = 118)
Differentiated (n = 22)
(%)
OR (95% CI)
n
(%)
OR (95% CI)
n
(%)
OR (95% CI)
(10.7) (3.3) (86.0)
1d 0.73 (0.21–2.50) 1.14 (0.57–2.28)
12 3 103
(10.2) (2.5) (87.3)
1d 0.62 (0.13–2.83) 1.35 (0.61–2.99)
2 1 19
(9.1) (4.6) (86.4)
1d 0.94 (0.07–12.37) 1.40 (0.29–6.71)
(26.7) (22.7) (36.7)
0.95 (0.44–2.04) 1.05 (0.47–2.33) 1.91 (0.83–4.41) 3.66; p = 0.06
34 27 42
(28.8) (22.9) (35.6)
1.11 (0.46–2.69) 1.32 (0.52–3.34) 2.23 (0.84–5.89) 3.58; p = 0.06
5 6 8
(22.7) (27.3) (36.4)
1.05 (0.18–5.96) 1.82 (0.30–11.05) 3.18 (0.46–21.79) 2.17; p = 0.14
(26.7) (32.0) (26.7)
1.36 (0.58–3.19) 1.05 (0.48–2.27) 0.87 (0.35–2.20) 0.11; p = 0.74
35 38 29
(29.7) (32.2) (24.6)
1.52 (0.58–3.96) 1.24 (0.50–3.05) 1.10 (0.37–3.25) 0.02; p = 0.88
3 7 9
(13.6) (31.8) (40.9)
1.38 (0.17–11.05) 1.53 (0.28–8.54) 1.43 (0.21–9.91) 0.15; p = 0.69
(27.3) (38.0) (20.0)
1.05 (0.48–2.29) 1.24 (0.59–2.62) 0.97 (0.43–2.20) 0.00; I = 0.97
35 46 21
(29.7) (39.0) (17.8)
1.35 (0.55–3.30) 1.53 (0.64–3.63) 0.95 (0.37–2.48) 0.04; p = 0.85
4 8 7
(18.2) (36.4) (31.8)
0.92 (0.15–5.65) 1.47 (0.27–7.88) 2.37 (0.42–13.44) 1.93; p = 0.16
Adjusted for sex, age, place of residence, year of interview, education, and smoking habits. Including 10 not otherwise specified NPCs. The sum does not add up to the total because of missing values. Reference category.
Table 4 Percent riska of nasopharyngeal carcinoma (NPC) and corresponding 95% confidence interval (CI) attributable to tobacco smoking and alcohol drinking.
Heavy tobacco smoking (P15 cigarettes/day) Heavy alcohol drinking (P28 drinks/week) Heavy tobacco smoking and heavy alcohol drinking a b
All NPCsb
Histological type Undifferentiated
Differentiated
(%)
(%)
(%)
19.4 (0.0–45.0) 9.9 (0.0–32.8) 23.6 (0.0–47.6)
9.8 (0.0–39.7) 7.2 (0.0–32.3) 13.7 (0.0–42.1)
55.4 (13.1–97.8) 27.8 (0.0–76.8) 57.2 (16.2–98.2)
Adjusted for sex, age, smoking habits, and drinking habits, when appropriate. Including 10 not otherwise specified NPCs.
In this study, alcohol drinking was not significantly associated with NPC risk. The absence of association with alcohol drinking may be explained by the lack of a direct contact of the nasopharynx with alcoholic beverages, although some exposure to ethanol or its metabolites (e.g., acetaldehyde) is possible through the breath. The small sample size should be considered a limitation of the present study, as well as the long time period it took to complete it (this investigation lasted 17 years). In particular, the analysis of interaction between smoking and alcohol was based on small numbers and the results must be interpreted with caution. Smoking habits and the histological classification of NPC have changed during study period. Although the prevalence of smoking habits declined in Italy over the study period, no substantial modification in average cigarette consumption was reported.27 Moreover, in the study population, the prevalence of drinking and smoking habits by period of accrual did not show substantial differences. In order to assure homogeneity, the same WHO classification3 was used for all cases. The lack of information on EBV status in the majority of NPC cases was an additional limitation of the present study. In the subset of cases with available EBV information, all undifferentiated NPCs were EBV-positive, suggesting an important role of EBV in the aetiology of this NPC histological type. Too few differentiated NPCs were evaluated to rule out the presence of EBV. The high prevalence of EBV in undifferentiated NPC suggests that EBV
should not be considered a confounder in the association between tobacco smoking and NPC risk, but rather, a largely alternative cause.19 This is somewhat similar to the findings on human papillomavirus (HPV) infection and cancer of the oropharynx: a higher prevalence of HPV infection in non-smokers and a stronger association with smoking among HPV-negative oropharyngeal cancers.28 Other potential weaknesses of case-control studies, such as the present one, comprise information and selection biases. The direct interview of cases and controls by the same trained interviewers, under similar conditions, in a hospital setting, minimised information bias. Careful attention was also paid to exclude from the control group subjects admitted for any condition related to the exposures under study, including tobacco smoking and alcohol drinking. In addition, the almost complete case ascertainment in the catchment areas, the nearly complete participation of identified cases and controls, and the use of a validated and reproducible questionnaire contributed to strengthen our findings.23,24 In conclusion, our findings from Italy showed that the fraction of undifferentiated NPCs attributable to tobacco smoking is probably small, whereas the majority of differentiated NPCs are associated to tobacco smoking as are the majority of other head and neck cancers. At a population level, the attributable fraction of tobacco smoking is therefore likely to vary according to the relative predominance of undifferentiated type in that population.
J. Polesel et al. / Oral Oncology 47 (2011) 541–545
Funding This work was supported by the Italian Association for Research on Cancer (AIRC).
Conflict of interest statement None declared. Acknowledgments The authors wish to thank Mrs. O. Volpato and M. Grimaldi for study coordination and L. Mei for editorial assistance. We are also deeply grateful to Dr. Emilia De Santis for the revision of the histopathological diagnoses; Drs. G. Chiara (1st General Surgery Dept., General Hospital, Pordenone), G. Tosolini (2nd General Surgery Dept., General Hospital, Pordenone), L. Forner (Eye Diseases Dept., General Hospital, Pordenone), A. Mele (Hand Surgery and Microsurgery Dept., General Hospital, Pordenone), and E. Trevisanutto (Dermatology Dept., General Hospital, Pordenone) for helping in enrolment of subjects. References 1. Curado MP, Edwards B, Shin HR, Storm H, Ferlay J, Heanue M, editors. Cancer Incidence in Five Continents, vol. IX. IARC Sci Publ No. 160. Lyon: IARC; 2007. 2. Chang ET, Adami HO. The enigmatic epidemiology of nasopharygeal carcinoma. Cancer Epidemiol Biomarkers Prev 2006;15(10):1765–77. 3. Shanmugaratnam K, Sobin LH. Histological typing of tumours of the upper respiratory tract and ear. 2nd ed. Berlin, Germany: Springer-Verlag; 1991. 4. Vokes EE, Liebowitz DN, Weichselbaum RR. Nasopharyngeal carcinoma. Lancet 1997;350(9084):1087–91. 5. Barnes L, Eveson JW, Reichart P, Sidransky D, editors. World Helath Organization Classification of Tumours. Pathology and Genetics of Head and Neck Tumours. Lyon: IARC, IARC Press; 2005. 6. IARC. IARC Monographs on the evaluation of carcinogenic risks to Humans, In: Epstein-Barr virus and Kaposi’s Sarcoma Herpesvirus/Human Herpersvirus 8, vol. 70. IARC Sci Publ. Lyon: IARC; 1997. 7. Shah KM, Young LS. Epstein-Barr virus and carcinogenesis: beyond Burkitt’s lymphoma. Clin Microbiol Infect 2009;15(11):982–8. 8. IARC. IARC Monographs on the evaluation of carcinogenic risks to Humans, In: Tobacco smoke and involuntary smoking, vol 83. IARC Sci Publ. Lyon: IARC; 2004. 9. Yang X, Diehl S, Pfeiffer R, Chen CJ, Hsu WL, Dosemeci M, et al. Evaluation of risk factors for nasopharyngeal carcinoma in high-risk nasopharyngeal carcinoma families in Taiwan. Cancer Epidemiol Biomarkers Prev 2005;14(4):900–5.
545
10. Friborg JT, Yuan JM, Wang R, Koh WP, Lee HP, Yu MC. A prospective study of tobacco and alcohol use as risk factors for pharyngeal carcinomas in Singapore Chinese. Cancer 2007;109(6):1183–91. 11. Guo X, Johnson RC, Deng H, Liao J, Guan L, Nelson GW, et al. Evaluation of nonviral risk factors for nasopharyngeal carcinoma in a high-risk population of southern China. Int J cancer 2009;124(12):2942–7. 12. Hsu WL, Chen JY, Chien YC, Liu MY, You SL, Hsu MM, et al. Independent effect of EBV and cigarettes smoking on nasopharyngeal carcinoma: A 20-year followup study on 9, 622 males without family history in Taiwan. Cancer Epidemiol Biomarkers Prev 2009;18(4):1218–26. 13. Ekburanawat W, Ekpanyaskul C, Brennan P, kanka C, Tepsuwan K, Temiyastith S, et al. Evaluation of non-viral risk factors for nasopharyngeal carcinoma in Thailand: results from a case-control study. Asian Pac J Cancer Prev 2010;11(4):929–32. 14. Hsu WL, Yu KJ, Chien YC, Chiang CJ, Cheng YJ, Chen JY, et al. Familial tendency and risk of nasopharyngeal carcinoma in Taiwan: effects of covariates on risk. Am J Epidemiol 2011;173(3):292–9. 15. Ji X, Zhang W, Xie C, Wang B, Zhang G, Zhou F. Nasopharyngeal carcinoma risk by histological type in central China: impact of smoking, alcohol and family history. Int J Cancer 2010. doi:10.1002/ijc.25696. 16. Mabuchi K, Bross DS, Kessler II. Cigarette smoking and nasopharyngeal carcinoma. Cancer 1985;55(12):2874–6. 17. Nam J, McLaughlin JK, Blot WJ. Cigarette smoking, alcohol, and nasopharyngeal carcinoma: a case-control study among U.S. whites. J Natl cancer Inst 1992;84(8):619–22. 18. Zhu K, Levine RS, Brann EA, Gnepp DR, Baum MK. A population-based casecontrol study of the relationship between cigarette smoking and nasopharyngeal cancer (United States). Cancer Causes Control 1995;6(6):507–12. 19. Vaughan TL, Shapiro JA, Burt RD, Swanson GM, Berwick M, Lynch CF, et al. Nasopharyngeal cancer in a low-risk population: defining risk factors by histological type. Cancer Epidemiol Biomarkers Prev 1996;5(8):587–93. 20. Chow WH, McLaughlin JK, Hrubec Z, Nam JM, Blot WJ. Tobacco use and nasopharyngeal carcinoma in a color of US veterans. Int J Cancer 1993;55(4):538–40. 21. Chen L, Gallicchio L, Boyd-Lindsey K, Tao XG, Robinson KA, Lam TK, et al. Alcohol consumption and the risk of nasopharyngeal carcinoma: a systematic review. Nutr Cancer 2009;61(1):1–15. 22. Kasum CM, Jacobs DR, Nicodemus K, Folsom AR. Dietary risk factors for upper aerodigestive tract cancers. Int J Cancer 2002;99(2):267–72. 23. D’Avanzo B, La Vecchia C, Katsouyanni K, Negri E, Trichopoulos D. Reliability of information on cigarette smoking and beverage consumption provided by hospital controls. Epidemiology 1996;7(3):312–5. 24. Ferraroni M, Decarli A, Franceschi S, La Vecchia C, Enard L, Negri E, et al. Validity and reproducibility of alcohol consumption in Italy. Int J Epidemiol 1996;25(4):775–82. 25. Breslow NE, Day NE. Statistical methods in cancer research, vol. 1. In: The Analysis of Case-Control Studies. IARC Sci Publ No. 32. Lyon: IARC; 1980. 26. Bruzzi P, Green SB, Byar DP, Brinton LA, Schairer C. Estimating the population attributable risk for multiple risk factors using case-control data. Am J Epidemiol 1985;122(5):904–14. 27. Ferketich AK, Gallus S, Iacobelli N, Zuccaro P, Colombo P, La Vecchia C. Smoking in Italy 2007, with a focus on the young. Tumori 2008;94(6):793–7. 28. Herrero R, Castellsagué X, Pawlita M, Lissowska J, Kee F, Balaram P, et al. Human papillomavirus and oral cancer: the international agency for research on cancer multicenter study. J Natl Cancer Inst 2003;95(23):1772–83.
Contents lists available at ScienceDirect
Oral Oncology journal homepage: www.elsevier.com/locate/oraloncology
Tobacco smoking, alcohol drinking, and the risk of different histological types of nasopharyngeal cancer in a low-risk population Jerry Polesel a,⇑, Silvia Franceschi b, Renato Talamini a, Eva Negri c, Luigi Barzan d, Maurizio Montella e, Massimo Libra f, Emanuela Vaccher g, Giovanni Franchin h, Carlo La Vecchia c,i, Diego Serraino a a
Unit of Epidemiology and Biostatistics, Centro di Riferimento Oncologico, IRCCS, Via F. Gallini 2, 33081 Aviano (PN), Italy International Agency for Research on Cancer, 150 Cours Albert Thomas, 69372 Lyon, France Departement of Epidemiology, The Mario Negri Institute for Pharmacological Research, Via G. La Masa 19, 20156 Milan, Italy d Department of Head and Neck Surgery, Azienda Ospedaliera ‘‘Santa Maria degli Angeli’’, Via Montereale 24, 33170 Pordenone, Italy e Unit of Epidemiology, Istituto Tumori ‘‘Fondazione Pascale’’, IRCCS, Cappella dei Cangiani, 80100 Naples, Italy f Departement of Biomedical Sciences, Università di Catania, Via Androne 83, 95124 Catania, Italy g Division of Medical Oncology A, Centro di Riferimento Oncologico, IRCCS, Via F. Gallini 2, 33081 Aviano, Italy h Division of Radiation Oncology, Centro di Riferimento Oncologico, IRCCS, Via F. Gallini 2, 33081 Aviano, Italy i Department of Occupational Medicine, Università degli Studi di Milano, Via Vanzetti 5, 20133 Milan, Italy b c
a r t i c l e
i n f o
Article history: Received 16 December 2010 Received in revised form 11 March 2011 Accepted 14 March 2011 Available online 7 April 2011 Keywords: Alcohol drinking Case-control study Nasopharyngeal carcinoma Tobacco smoking
s u m m a r y Nasopharyngeal carcinoma (NPC) is rare in western Countries. Tobacco smoking is a well-recognised risk factor, whereas the role of alcohol drinking is still in debate. We conducted a hospital-based case-control study in Italy on 150, histologically-confirmed, NPC cases of Caucasian ethnicity, aged 18–76 years, including 118 undifferentiated NPCs and 22 differentiated squamous-cell NPC. Controls were 450 Caucasian cancer-free patients admitted to general hospitals for acute conditions. Cases and controls were matched according to sex, age, and place of residence. Logistic regression was used to estimate odds ratios (OR) and the corresponding 95% confidence intervals (CI) while adjusting for known confounders. No significant association emerged between tobacco smoking and all NPCs (OR for current vs. never smokers = 1.52; 95% CI: 0.89–2.60). Conversely, for differentiated NPC only, statistically significant elevated OR were associated with increasing smoking intensity (OR for P15 cigarettes/day = 5.40; 95% CI: 1.34–21.76) and duration of the habit (OR for P32 years = 4.48; 95% CI: 1.11–18.04). Although alcohol drinking was not, per se, significantly associated to NPC risk, the combination of tobacco smoking and alcohol drinking accounted for 57% of differentiated NPCs, whereas it accounted for only 14% of undifferentiated carcinomas. Our findings suggest that, in western populations, NPC includes two separate entities: the differentiated NPC, associated with tobacco smoking like other cancers of head and neck, and the undifferentiated NPC, upon which tobacco smoking has little or no influence. Ó 2011 Elsevier Ltd. All rights reserved.
Introduction Nasopharyngeal carcinoma (NPC) is a rare cancer in Europe and North America, occurring in less than one case per 100,000 people each year. It is approximately 2-fold more frequent in men than in women.1 Very high incidence rates were reported in China, in South-eastern Asia, in the Arctic region,2 and in immigrants from these high-risk areas.2 The World Health Organisation has recognised three histological types of NPC.3,4 Type I, representing squamous cell carcinoma, is similar to carcinomas arising from other sites of head and neck. ⇑ Corresponding author. Tel.: +39 0434 659195; fax: +39 0434 659231. E-mail address: [email protected] (J. Polesel). 1368-8375/$ - see front matter Ó 2011 Elsevier Ltd. All rights reserved. doi:10.1016/j.oraloncology.2011.03.017
Type II is characterised as non-keratinizing carcinoma. Type III, representing undifferentiated carcinoma (also referred to as lymphoepithelioma or Schminke tumour), is characterised by the tumour infiltration of non-cancerous T-lymphocytes. In low-risk areas, undifferentiated carcinoma is the most frequent type, accounting for approximately 75% of new cases5; squamous cell carcinoma represents approximately 25% of cases, being nonkeratinizing carcinomas very rare.5 In high-incidence areas, infection with Epstein-Barr virus (EBV) is a well-established risk factor for NPC,6 particularly undifferentiated NPC7 and when the infection occurs early in life.2,4 Conversely, in low-incidence areas, the fraction of NPCs attributable to EBV is unclear, and NPC has been related to classic risk factors for head and neck cancer, notably tobacco smoking.4
542
J. Polesel et al. / Oral Oncology 47 (2011) 541–545
In 2004, the International Agency for Research of Cancer established a causal role for tobacco smoking in NPC onset,8 an evidence supported by investigations conducted in recent years.9–15 However, the majority of the researches investigating the aetiology of NPC were carried out in high-risk populations, with few investigations conducted in low-risk areas (four case-control studies16–19 and a cohort one20 in the United States (US)). These latter studies showed an excess of NPC among smokers, with up to 6-fold higher risks in heavy smokers than in never smokers. Only one19 of the US studies provided risk estimates according to histological types: the findings suggested that the association with smoking was much stronger with differentiated than with undifferentiated NPC. Studies evaluating the association between alcohol intake and NPC risk showed inconsistent results. A recent meta-analysis found a 33% increase in NPC risk among heavy drinkers,21 in the absence, however, of a clear dose–response relationship. Moreover, as for tobacco smoking, epidemiological evidence from low-risk areas is limited, and it is based exclusively on selected US populations.16,17,19,22 To investigate the impact of tobacco smoking and alcohol drinking on NPC risk in a low-risk European population, we conducted a case-control study in Italy taking into consideration the associations with different histological types.
Materials and methods Between 1992 and 2008, we conducted a case-control study on NPC within an established network of collaborating centres, including Aviano and Pordenone in North-eastern Italy, and Naples and Catania in southern Italy. Cases were 150 Caucasian patients aged 18–76 years (median age: 52 years) with incident NPC, admitted to major general hospitals in the study areas. All 150 NPC cases were histologically confirmed, with 118 (78.7%) undifferentiated NPCs (WHO type 3),3 22 (14.7%) keratinizing squamous cell carcinomas (WHO type 1)3 (here referred to as differentiated NPCs), and 10 (6.7%) not otherwise specified NPCs. EBV status was available for 61 NPC cases based on the detection of EBV nuclear antigen (EBNA) in tissue samples. All 57 undifferentiated NPCs and two out of four differentiated NPCs were EBV-positive. Three controls were frequency-matched to each case, according to sex, age (<40, 40–44, 45–49, 50–54, 55–59, 60–64, 65–69 and P70 years), and place of residence (northern Italy, central Italy, southern Italy and Islands). The control group included 450 Caucasian patients (aged 19–76 years; median age: 52 years) admitted
for a wide spectrum of acute conditions to the same hospitals where cases had been interviewed. Subjects admitted for malignant neoplasms, conditions related to tobacco smoking or alcohol consumption, or any disorder that might have induced long-term modification of diet were excluded from the control group. Controls were admitted for traumatic orthopaedic disorders (37%), other orthopaedic disorders (35%), acute surgical conditions (18%), and miscellaneous other illnesses, including eye, nose, ear, skin, or dental disorders (10%). All study participants signed an informed consent, according to the recommendations of the internal Boards of Ethics of study hospitals. Trained interviewers administered a structured questionnaire to cases and controls during their hospital stays, thus keeping refusal below 3%. The questionnaire collected information on sociodemographic factors; lifestyle habits; diet; a problem-oriented medical history; and family history of cancer. Information on smoking included lifetime status (i.e., never, former, or current smoker), daily number of cigarettes/cigars and grams of tobacco pipe smoked, age at starting and duration of the habit. In our computations, one gram of pipe-smoked tobacco corresponded to one cigarette, and one cigar to three cigarettes. Smokers were subjects who had smoked at least one cigarette/day for at least 1 year. Former smokers were defined as smokers who had abstained from cigarette smoking for at least 12 months before interview. Lifetime alcohol drinking status (i.e., abstainer, former, or current drinker) and the weekly number of drinks for the five most common alcoholic beverages or groups of beverages (i.e., wine, beer, herb liquors, grappa, and spirits) were assessed. Taking into account the different ethanol concentration, one drink corresponded to approximately 125 ml of wine, 330 ml of beer, and 30 ml of hard liquor (i.e., about 12 g of ethanol). Ages at start drinking and at quitting, for ex-drinkers, were also recorded. Former drinkers had abstained from any type of drinking for at least 12 months. The validity and reproducibility of questions on self-reported drinking and smoking habits in our study population were satisfactory.23,24 Odds ratios (OR) and the corresponding 95% confidence intervals (CI) were calculated by means of multiple logistic regression models including terms for sex, age, place of residence, year of interview, education level (i.e.,<7, 7–11, and P12 years), smoking and drinking habits, as specified in table footnotes.25 The test for trend was based on the likelihood-ratio test between the models with and without linear terms for each variable of interest.25 Percent attributable risks were computed according to the method described by Bruzzi et al.26 which used the distribution of the risk factors in NPC cases. All analyses were performed on all NPCs
Table 1 Distribution of 150 cases of nasopharyngeal carcinoma (NPC) and 450 controls according to matching variables.
Sex Men Women
Controls (n = 450)
All NPCsa (n = 150)
Undifferentiated (n = 118)
Differentiated (n = 22)
n
(%)
n
(%)
n
n
(%)
357 93
(79.3) (20.7)
119 31
(79.3) (20.7)
93 25
(78.8) (21.2) v for heterogeneity: p = 0.53
16 6
(72.7) (27.3)
Histological type
(%)
2
Age (years) <44 45–54 55–64 P65
111 138 117 84
(24.7) (30.7) (26.0) (18.7)
37 46 39 28
(24.7) (30.7) (26.0) (18.7)
32 38 29 19
(27.1) (32.2) (24.6) (16.1) v2 for heterogeneity: p = 0.06
3 5 7 7
(13.6) (22.7) (31.8) (31.8)
Place of residence North-East of Italy Centre, South, Islands
270 180
(60.0) (40.0)
90 60
(60.0) (40.0)
73 45
13 9
(59.1) (40.9)
(61.9) (38.1) v for heterogeneity: p = 0.81 2
a
Including 10 not otherwise specified NPCs.
543
J. Polesel et al. / Oral Oncology 47 (2011) 541–545
and, separately, for differentiated and undifferentiated NPCs versus all controls. Results Table 1 shows the distribution of cases (overall and by histological type) and controls according to study design variables. The distribution of undifferentiated and differentiated NPC cases by sex and area of residence was similar. Undifferentiated NPCs tended, however, to be younger than differentiated NPC cases (p = 0.06). Current tobacco smoking was not significantly associated with NPC risk (OR = 1.52; 95% CI: 0.89–2.60), but the association differed according to histological type (Table 2). The risk of current smokers versus never smokers seemed to be lower for undifferentiated NPC (OR = 1.37; 95% CI: 0.74–2.52) than for differentiated NPC (OR = 3.17; 95% CI: 0.91–11.03). Contrary to undifferentiated NPC, differentiated NPC also showed a clear relationship of cancer risk with intensity (OR for P15 cigarettes/day = 5.40; 95% CI: 1.34– 21.76) and duration (OR for P30 years = 4.48; 95% CI: 1.11–18.04; Table 2) of the habit. No clear relationship emerged between drinking status and NPC risk (Table 3). Current alcohol drinking of 28 drinks/week or more was not significantly associated with NPC (OR = 1.91, 95% CI: 0.83– 4.41; p for trend = 0.06). Duration of alcohol drinking (p for trend = 0.74) and age at beginning (p for trend = 0.97) were also unrelated to NPC risk, and no differences emerged according to histological type. The interaction of heavy tobacco smoking with alcohol drinking was also evaluated. Compared to never smokers who drunk <14 drinks/week, the ORs among subjects who smoked P15 cigarettes/day and drunk P28 drinks/week were 1.98 (95% CI: 0.84– 4.68) for undifferentiated NPC and 8.86 (95% CI: 1.47–53.20) for differentiated NPC. The fraction of NPC, overall and by histological type, attributable to heavy tobacco smoking and heavy alcohol drinking is
shown in Table 4. Although the confidence intervals of attributable fractions were broad due to small numbers, the contribution of tobacco smoking in differentiated NPC (55%; 95% CI: 13–98%) was larger than that in undifferentiated NPC (10%; 95% CI: 0–40%). Altogether, heavy tobacco smoking and heavy alcohol drinking accounted for 57% (95% CI: 16–98%) of differentiated NPCs, but only for 14% (95% CI: 0–42%%) of undifferentiated NPCs. Discussion The present study investigated, for the first time in a European country, the role of tobacco smoking and alcohol drinking in the aetiology of NPC. The findings provided additional evidence that, in western populations, NPC includes two separate entities: one, differentiated NPCs, clearly associated to tobacco smoking like other cancers of head and neck, and the other, represented by undifferentiated carcinoma, on which tobacco smoking has little or no influence. Likewise, the association of cancer risk with intensity, duration, and young age at starting smoking was substantially stronger for differentiated than for undifferentiated NPC, in agreement with one previous study from the US.19 The difference in the strength of the association of NPC (any histological type) with smoking in previous studies8–15 may derive, therefore, from a different proportion of differentiated and undifferentiated NPCs in various study populations. The only report that provided information on risk factors for NPC separately by histological type in a western Country showed, as we did, a stronger influence of smoking on differentiated than on undifferentiated NPC.19 However, differentiated NPCs represented 66.6% of cancer cases in the study by Vaugham et al.19 whereas they accounted for only 14.7% in the present study. Lower proportion of undifferentiated NPC was reported in Italy and, therefore, the high proportion in our present study is likely to derive from the selective referral of this rare malignancy to the specialised Ear, Nose & Throat (ENT) Clinics where the present study was conducted.
Table 2 Distribution of 150 cases of nasopharyngeal carcinoma (NPC) and 450 matched controls, odds ratio (OR)a and corresponding 95% confidence interval (CI), according to smoking habits. Controls (n = 450)
All NPCsb (n = 150)
Undifferentiated (n = 118)
Differentiated (n = 22)
n
(%)
n
(%)
OR (95% CI)
n
(%)
OR (95% CI)
n
(%)
OR (95% CI)
(32.2) (29.8) (38.0)
39 48 63
(26.0) (32.0) (42.0)
1d 1.43 (0.81–2.52) 1.52 (0.89–2.60)
34 38 46
(28.8) (32.2) (39.0)
1d 1.43 (0.76–2.70) 1.37 (0.74–2.52)
4 7 11
(18.2) (31.8) (50.0)
1d 2.36 (0.60–9.36) 3.17 (0.91–11.03)
Intensity among current smokers (cigarettes/day)c <15 66 (14.7) 27 P15 105 (23.3) 34 v2 trend
(18.0) (22.7)
0.96 (0.45–2.05) 1.98 (1.05–3.75) 4.13; p = 0.04
20 24
(16.9) (20.3)
0.82 (0.33–2.03) 1.76 (0.83–3.73) 1.89; p = 0.17
5 6
(22.7) (27.2)
2.17 (0.43–10.92) 5.40 (1.34–21.76) 5.64; p = 0.02
Duration among current smokers (years)c <30 74 (19.1) 26 P30 97 (18.9) 37 2 v trend
(17.3) (24.7)
1.26 (0.60–2.64) 1.85 (0.94–3.64) 2.86; p = 0.09
21 25
(17.8) (21.2)
1.25 (0.53–2.96) 1.55 (0.71–3.39) 1.18; p = 0.28
3 8
(13.6) (36.4)
2.51 (0.42–15.17) 4.48 (1.11–18.04) 4.44; p = 0.04
Age at beginning among current smokers (years)c P18 92 (20.4) 25 <18 79 (17.6) 38 v2 trend
(16.7) (25.3)
1.23 (0.64–2.37) 2.09 (1.06–4.13) 4.33; p = 0.04
19 27
(16.1) (22.9)
1.16 (0.55–2.47) 1.81 (0.81–4.04) 1.98; p = 0.16
4 7
(18.2) (31.8)
2.20 (0.50–9.60) 8.81 (1.87–41.55) 7.30; p < 0.01
Year since quitting among former smokers (years) P15 60 (13.3) 25 <15 74 (16.4) 23 2 v trend
(16.7) (15.3)
1.81 (0.87–3.76) 1.40 (0.71–2.77)
19 19
(16.1) (16.1)
1.62 (0.72–3.67) 1.54 (0.72–3.23) 1.39; p = 0.24
4 3
(18.2) (13.6)
2.44 (0.40–15.09) 1.43 (0.23–8.91) 0.13; p = 0.72
Smoking status Never 145 Former 134 Current 171
a b c d
Histological type
Adjusted for sex, age, place of residence, year of interview, education, and total alcohol intake. Including 10 not otherwise specified NPCs. The sum does not add up to the total because of missing values. Reference category.
544
J. Polesel et al. / Oral Oncology 47 (2011) 541–545
Table 3 Distribution of 150 cases of nasopharyngeal carcinoma (NPC) and 450 matched controls, odds ratio (OR)a and corresponding 95% confidence interval (CI), according to drinking habits. Controls(n = 450)
All NPCsb (I = 150)
n
n
(%)
Drinking status Never 54 (12.0) 16 Former 25 (5.6) 5 Current 371 (82.4) 129 Intensity among current drinkers (drinks/week) <14 124 (27.6) 40 14–27 119 (26.4) 34 P28 128 (28.4) 55 2 v trend Duration among current drinkers (years)c <25 91 (20.2) 40 25–39 155 (34.4) 48 P40 124 (27.6) 40 v2 trend Age of beginning among current drinkers (years)c P21 105 (23.3) 41 18–20 152 (33.8) 57 <18 113 (25.1) 30 v2 trend a b c d
Histological type Undifferentiated (n = 118)
Differentiated (n = 22)
(%)
OR (95% CI)
n
(%)
OR (95% CI)
n
(%)
OR (95% CI)
(10.7) (3.3) (86.0)
1d 0.73 (0.21–2.50) 1.14 (0.57–2.28)
12 3 103
(10.2) (2.5) (87.3)
1d 0.62 (0.13–2.83) 1.35 (0.61–2.99)
2 1 19
(9.1) (4.6) (86.4)
1d 0.94 (0.07–12.37) 1.40 (0.29–6.71)
(26.7) (22.7) (36.7)
0.95 (0.44–2.04) 1.05 (0.47–2.33) 1.91 (0.83–4.41) 3.66; p = 0.06
34 27 42
(28.8) (22.9) (35.6)
1.11 (0.46–2.69) 1.32 (0.52–3.34) 2.23 (0.84–5.89) 3.58; p = 0.06
5 6 8
(22.7) (27.3) (36.4)
1.05 (0.18–5.96) 1.82 (0.30–11.05) 3.18 (0.46–21.79) 2.17; p = 0.14
(26.7) (32.0) (26.7)
1.36 (0.58–3.19) 1.05 (0.48–2.27) 0.87 (0.35–2.20) 0.11; p = 0.74
35 38 29
(29.7) (32.2) (24.6)
1.52 (0.58–3.96) 1.24 (0.50–3.05) 1.10 (0.37–3.25) 0.02; p = 0.88
3 7 9
(13.6) (31.8) (40.9)
1.38 (0.17–11.05) 1.53 (0.28–8.54) 1.43 (0.21–9.91) 0.15; p = 0.69
(27.3) (38.0) (20.0)
1.05 (0.48–2.29) 1.24 (0.59–2.62) 0.97 (0.43–2.20) 0.00; I = 0.97
35 46 21
(29.7) (39.0) (17.8)
1.35 (0.55–3.30) 1.53 (0.64–3.63) 0.95 (0.37–2.48) 0.04; p = 0.85
4 8 7
(18.2) (36.4) (31.8)
0.92 (0.15–5.65) 1.47 (0.27–7.88) 2.37 (0.42–13.44) 1.93; p = 0.16
Adjusted for sex, age, place of residence, year of interview, education, and smoking habits. Including 10 not otherwise specified NPCs. The sum does not add up to the total because of missing values. Reference category.
Table 4 Percent riska of nasopharyngeal carcinoma (NPC) and corresponding 95% confidence interval (CI) attributable to tobacco smoking and alcohol drinking.
Heavy tobacco smoking (P15 cigarettes/day) Heavy alcohol drinking (P28 drinks/week) Heavy tobacco smoking and heavy alcohol drinking a b
All NPCsb
Histological type Undifferentiated
Differentiated
(%)
(%)
(%)
19.4 (0.0–45.0) 9.9 (0.0–32.8) 23.6 (0.0–47.6)
9.8 (0.0–39.7) 7.2 (0.0–32.3) 13.7 (0.0–42.1)
55.4 (13.1–97.8) 27.8 (0.0–76.8) 57.2 (16.2–98.2)
Adjusted for sex, age, smoking habits, and drinking habits, when appropriate. Including 10 not otherwise specified NPCs.
In this study, alcohol drinking was not significantly associated with NPC risk. The absence of association with alcohol drinking may be explained by the lack of a direct contact of the nasopharynx with alcoholic beverages, although some exposure to ethanol or its metabolites (e.g., acetaldehyde) is possible through the breath. The small sample size should be considered a limitation of the present study, as well as the long time period it took to complete it (this investigation lasted 17 years). In particular, the analysis of interaction between smoking and alcohol was based on small numbers and the results must be interpreted with caution. Smoking habits and the histological classification of NPC have changed during study period. Although the prevalence of smoking habits declined in Italy over the study period, no substantial modification in average cigarette consumption was reported.27 Moreover, in the study population, the prevalence of drinking and smoking habits by period of accrual did not show substantial differences. In order to assure homogeneity, the same WHO classification3 was used for all cases. The lack of information on EBV status in the majority of NPC cases was an additional limitation of the present study. In the subset of cases with available EBV information, all undifferentiated NPCs were EBV-positive, suggesting an important role of EBV in the aetiology of this NPC histological type. Too few differentiated NPCs were evaluated to rule out the presence of EBV. The high prevalence of EBV in undifferentiated NPC suggests that EBV
should not be considered a confounder in the association between tobacco smoking and NPC risk, but rather, a largely alternative cause.19 This is somewhat similar to the findings on human papillomavirus (HPV) infection and cancer of the oropharynx: a higher prevalence of HPV infection in non-smokers and a stronger association with smoking among HPV-negative oropharyngeal cancers.28 Other potential weaknesses of case-control studies, such as the present one, comprise information and selection biases. The direct interview of cases and controls by the same trained interviewers, under similar conditions, in a hospital setting, minimised information bias. Careful attention was also paid to exclude from the control group subjects admitted for any condition related to the exposures under study, including tobacco smoking and alcohol drinking. In addition, the almost complete case ascertainment in the catchment areas, the nearly complete participation of identified cases and controls, and the use of a validated and reproducible questionnaire contributed to strengthen our findings.23,24 In conclusion, our findings from Italy showed that the fraction of undifferentiated NPCs attributable to tobacco smoking is probably small, whereas the majority of differentiated NPCs are associated to tobacco smoking as are the majority of other head and neck cancers. At a population level, the attributable fraction of tobacco smoking is therefore likely to vary according to the relative predominance of undifferentiated type in that population.
J. Polesel et al. / Oral Oncology 47 (2011) 541–545
Funding This work was supported by the Italian Association for Research on Cancer (AIRC).
Conflict of interest statement None declared. Acknowledgments The authors wish to thank Mrs. O. Volpato and M. Grimaldi for study coordination and L. Mei for editorial assistance. We are also deeply grateful to Dr. Emilia De Santis for the revision of the histopathological diagnoses; Drs. G. Chiara (1st General Surgery Dept., General Hospital, Pordenone), G. Tosolini (2nd General Surgery Dept., General Hospital, Pordenone), L. Forner (Eye Diseases Dept., General Hospital, Pordenone), A. Mele (Hand Surgery and Microsurgery Dept., General Hospital, Pordenone), and E. Trevisanutto (Dermatology Dept., General Hospital, Pordenone) for helping in enrolment of subjects. References 1. Curado MP, Edwards B, Shin HR, Storm H, Ferlay J, Heanue M, editors. Cancer Incidence in Five Continents, vol. IX. IARC Sci Publ No. 160. Lyon: IARC; 2007. 2. Chang ET, Adami HO. The enigmatic epidemiology of nasopharygeal carcinoma. Cancer Epidemiol Biomarkers Prev 2006;15(10):1765–77. 3. Shanmugaratnam K, Sobin LH. Histological typing of tumours of the upper respiratory tract and ear. 2nd ed. Berlin, Germany: Springer-Verlag; 1991. 4. Vokes EE, Liebowitz DN, Weichselbaum RR. Nasopharyngeal carcinoma. Lancet 1997;350(9084):1087–91. 5. Barnes L, Eveson JW, Reichart P, Sidransky D, editors. World Helath Organization Classification of Tumours. Pathology and Genetics of Head and Neck Tumours. Lyon: IARC, IARC Press; 2005. 6. IARC. IARC Monographs on the evaluation of carcinogenic risks to Humans, In: Epstein-Barr virus and Kaposi’s Sarcoma Herpesvirus/Human Herpersvirus 8, vol. 70. IARC Sci Publ. Lyon: IARC; 1997. 7. Shah KM, Young LS. Epstein-Barr virus and carcinogenesis: beyond Burkitt’s lymphoma. Clin Microbiol Infect 2009;15(11):982–8. 8. IARC. IARC Monographs on the evaluation of carcinogenic risks to Humans, In: Tobacco smoke and involuntary smoking, vol 83. IARC Sci Publ. Lyon: IARC; 2004. 9. Yang X, Diehl S, Pfeiffer R, Chen CJ, Hsu WL, Dosemeci M, et al. Evaluation of risk factors for nasopharyngeal carcinoma in high-risk nasopharyngeal carcinoma families in Taiwan. Cancer Epidemiol Biomarkers Prev 2005;14(4):900–5.
545
10. Friborg JT, Yuan JM, Wang R, Koh WP, Lee HP, Yu MC. A prospective study of tobacco and alcohol use as risk factors for pharyngeal carcinomas in Singapore Chinese. Cancer 2007;109(6):1183–91. 11. Guo X, Johnson RC, Deng H, Liao J, Guan L, Nelson GW, et al. Evaluation of nonviral risk factors for nasopharyngeal carcinoma in a high-risk population of southern China. Int J cancer 2009;124(12):2942–7. 12. Hsu WL, Chen JY, Chien YC, Liu MY, You SL, Hsu MM, et al. Independent effect of EBV and cigarettes smoking on nasopharyngeal carcinoma: A 20-year followup study on 9, 622 males without family history in Taiwan. Cancer Epidemiol Biomarkers Prev 2009;18(4):1218–26. 13. Ekburanawat W, Ekpanyaskul C, Brennan P, kanka C, Tepsuwan K, Temiyastith S, et al. Evaluation of non-viral risk factors for nasopharyngeal carcinoma in Thailand: results from a case-control study. Asian Pac J Cancer Prev 2010;11(4):929–32. 14. Hsu WL, Yu KJ, Chien YC, Chiang CJ, Cheng YJ, Chen JY, et al. Familial tendency and risk of nasopharyngeal carcinoma in Taiwan: effects of covariates on risk. Am J Epidemiol 2011;173(3):292–9. 15. Ji X, Zhang W, Xie C, Wang B, Zhang G, Zhou F. Nasopharyngeal carcinoma risk by histological type in central China: impact of smoking, alcohol and family history. Int J Cancer 2010. doi:10.1002/ijc.25696. 16. Mabuchi K, Bross DS, Kessler II. Cigarette smoking and nasopharyngeal carcinoma. Cancer 1985;55(12):2874–6. 17. Nam J, McLaughlin JK, Blot WJ. Cigarette smoking, alcohol, and nasopharyngeal carcinoma: a case-control study among U.S. whites. J Natl cancer Inst 1992;84(8):619–22. 18. Zhu K, Levine RS, Brann EA, Gnepp DR, Baum MK. A population-based casecontrol study of the relationship between cigarette smoking and nasopharyngeal cancer (United States). Cancer Causes Control 1995;6(6):507–12. 19. Vaughan TL, Shapiro JA, Burt RD, Swanson GM, Berwick M, Lynch CF, et al. Nasopharyngeal cancer in a low-risk population: defining risk factors by histological type. Cancer Epidemiol Biomarkers Prev 1996;5(8):587–93. 20. Chow WH, McLaughlin JK, Hrubec Z, Nam JM, Blot WJ. Tobacco use and nasopharyngeal carcinoma in a color of US veterans. Int J Cancer 1993;55(4):538–40. 21. Chen L, Gallicchio L, Boyd-Lindsey K, Tao XG, Robinson KA, Lam TK, et al. Alcohol consumption and the risk of nasopharyngeal carcinoma: a systematic review. Nutr Cancer 2009;61(1):1–15. 22. Kasum CM, Jacobs DR, Nicodemus K, Folsom AR. Dietary risk factors for upper aerodigestive tract cancers. Int J Cancer 2002;99(2):267–72. 23. D’Avanzo B, La Vecchia C, Katsouyanni K, Negri E, Trichopoulos D. Reliability of information on cigarette smoking and beverage consumption provided by hospital controls. Epidemiology 1996;7(3):312–5. 24. Ferraroni M, Decarli A, Franceschi S, La Vecchia C, Enard L, Negri E, et al. Validity and reproducibility of alcohol consumption in Italy. Int J Epidemiol 1996;25(4):775–82. 25. Breslow NE, Day NE. Statistical methods in cancer research, vol. 1. In: The Analysis of Case-Control Studies. IARC Sci Publ No. 32. Lyon: IARC; 1980. 26. Bruzzi P, Green SB, Byar DP, Brinton LA, Schairer C. Estimating the population attributable risk for multiple risk factors using case-control data. Am J Epidemiol 1985;122(5):904–14. 27. Ferketich AK, Gallus S, Iacobelli N, Zuccaro P, Colombo P, La Vecchia C. Smoking in Italy 2007, with a focus on the young. Tumori 2008;94(6):793–7. 28. Herrero R, Castellsagué X, Pawlita M, Lissowska J, Kee F, Balaram P, et al. Human papillomavirus and oral cancer: the international agency for research on cancer multicenter study. J Natl Cancer Inst 2003;95(23):1772–83.