Incidence and mortality from oral and pharyngeal cancer vary widely across the world.1, 2 This variation depends mostly on differences in exposure to tobacco and alcohol, the 2 major risk factors for oral and pharyngeal cancer. However, genetic factors may also contribute to explain the observed differences in rates.2
Only a few epidemiologic studies have assessed familial risks in oral and pharyngeal cancer. A case-control study from the United States on 487 cases found odds ratios (OR) of 1.2 (95% confidence interval (CI), 0.7–2.3) in subjects with a family history of oral and pharyngeal cancer, and of 1.6 (95% CI, 0.7–3.8) for a family history of cancers of the esophagus and larynx.3 Another case-control study from Puerto Rico on 342 oral and pharyngeal cancers showed ORs of 2.5 (95% CI, 0.8–8.0) in subjects reporting a first-degree relative with oral and pharyngeal cancer, and of 2.6 (95% CI, 1.4–4.8) in those reporting upper aerodigestive tract cancers.4 A record-linkage study on the Utah Population Database, which includes ∼1 million individuals, found a standardized incidence ratio of oral and pharyngeal cancer of 1.8 (95%, 0.5–4.0) for subjects with a first-degree relative with cancer at the same site (8 observed versus 4.4 expected cases).5
The familial aggregation of oral and pharyngeal cancers, however, may have different genetic and environmental correlates in different populations, including the prevalence of various susceptibility genes and the exposure to various environmental factors. Thus, to provide further data on the issue, we have examined the relation between family history of oral and pharyngeal cancer/laryngeal cancer and the risk of oral and pharyngeal cancer and other cancers using data from a large multicentric case-control study conducted in Italy and Switzerland.
Material and methods
Data were derived from a case-control study of oral and pharyngeal cancer conducted between 1992 and 2005 in the greater Milan area and the provinces of Pordenone in northern Italy, of Rome and Latina in central Italy, and in the Swiss Canton of Vaud.6, 7 Data collection started and ended later in Milan. However, the same questionnaires and selection criteria were used in all centers.
Cases consisted of 956 patients under age 79 years admitted to major teaching and general hospitals in the areas under study with incident, histologically-confirmed carcinoma of the oral cavity and pharynx (median age 58 years, range 22–78). Controls were 2362 subjects under age 79 years (median age 58 years, range 19–79), selected among patients admitted to the same hospitals as cases for a wide spectrum of acute, nonneoplastic conditions, unrelated to known risk factors for oral and pharyngeal diseases, tobacco and alcohol consumption, and long-term modifications of diet. Seventeen percent of controls were admitted for traumas, 33% for other orthopedic disorders, 25% for surgical conditions, and 25% for miscellaneous other illnesses, including eye, nose, ear, skin or dental disorders. Less than 5% of cases and controls contacted refused to participate.
Centrally trained interviewers interviewed cases and controls during their hospital stay, using a structured questionnaire that included information on socio-demographic characteristics, anthropometric measures, lifestyle habits (including tobacco smoking and alcohol drinking), a validated food frequency section, and personal medical history. Informed consent was obtained prior to the interview from the patients and from the hospitals, in respect to the legislation in force. The subjects were specifically asked how many sisters and brothers they had, and whether their parents, siblings, children, grandparents or spouse had ever had cancer (excluding nonmelanoma skin cancer). For each relative with a history of cancer, the subject was asked to report the vital status of the relative at the time of interview, his/her current age or the age at death, the site of the tumour and the age at cancer diagnosis. In the present analysis, we considered the history of cancer in first-degree relatives only, i.e., parents, siblings and children. On account of recall and classification difficulties, some sites were combined (i.e., the whole intestinal tract, Hodgkin and non-Hodgkin lymphomas, cervix and corpus uteri). No verification of the cancer diagnoses in the relatives was performed.
We estimated the OR and the corresponding 95% CIs of oral and pharyngeal cancer according to history of cancer at selected sites in first-degree relatives using conditional multiple logistic regression models.8 The models were conditioned on sex, age and study centre, and included terms for education, tobacco smoking, alcohol drinking, and number of brothers and sisters.
Table I gives the distribution of 956 cases and 2362 controls according to center, age, sex and selected covariates. By design, the proportion of women was higher in controls than in cases, and the age distribution was similar in cases and controls. Cases reported higher tobacco and alcohol consumption.
Table I. Distribution of 956 Oral and Pharyngeal Cancer Cases and 2362 Controls According to Center, Sex, Age and Selected Covariates (Italy and Switzerland, 1992–20051)
The sum may not add up to the total because of some missing values.
1 Cigar equals 3 cigarettes, 1 g of pipe tobacco equals 1 cigarette.
Ex-smokers were subjects who had stopped smoking for at least 1 year.
Table II presents the relation between oral and pharyngeal cancer risk and several aspects of family history of oral and pharyngeal cancer and/or laryngeal cancer. The ORs of oral and pharyngeal cancer according to family history of oral and pharyngeal cancer and laryngeal cancer were similar (OR = 3.0 and 3.2, respectively). The OR for oral and pharyngeal cancer and laryngeal cancer combined was 3.1. The ORs became 2.6, 3.8 and 3.1, respectively, for oral and pharyngeal cancer, laryngeal cancer and oral and pharyngeal/laryngeal cancers after adjustment for education, smoking, alcohol and number of brothers and sisters. No substantial differences emerged according to the type of relative and the age of the relative at diagnosis of oral and pharyngeal/laryngeal cancers, while the adjusted OR was 7.1 (95% CI, 1.3–37.2) for those with 2 or more first-degree relatives affected.
Table II. Odds Ratio of Oral and Pharyngeal Cancer According to Family History of Oral and Pharyngeal Cancer, Laryngeal and Oral and Pharyngeal/Laryngeal Cancers in First-Degree Relatives in a Multicenter Study From Italy and Switzerland, 1992–2005
Family history of oral and pharyngeal cancer
Family history of laryngeal cancer
Family history of oral and pharyngeal cancer /laryngeal cancer
Further adjusted for education, smoking, alcohol and number of brothers and sisters when appropriate.
Reference category: no family history.
Subjects with both a sibling and parent/child affected were 3 cases for a family history of oral and pharyngeal cancer, 2 cases and 2 controls for a family history of laryngeal cancer and 6 cases and 2 controls for a family history of oral and pharyngeal cancer/laryngeal cancer.
Table III gives the distribution of cases and controls according to family history of oral and pharyngeal cancer in separate strata of age, tobacco and alcohol consumption, and the corresponding ORs. No clear differences in risk emerged between strata of covariates. The tests for interaction were not significant for any of the strata considered.
Table III. Odds Ratio of Oral and Pharyngeal Cancer According to Family History of Oral and Pharyngeal Cancer/Laryngeal Cancers in Strata of Selected Covariates (Italy and Switzerland, 1992–2005)
Family history of oral and pharyngeal cancer/laryngeal cancer
The joint effect of smoking, drinking and family history of oral and pharyngeal cancer is shown in Figure 1. Compared to the lowest risk category, i.e., never/ex smokers (former smokers since less than 5 years were included in the “current” smokers category), drinkers of <21 drinks/week without family history, the risk was increased in those with one or more factors in the highest risk category. Moreover, in each combination of smoking and drinking habits, a family history of oral and pharyngeal/laryngeal cancers increased the risk of oral and pharyngeal cancer. The OR was 1.9 (95% CI, 0.8–4.5) for nonsmokers, moderate drinkers with family history of oral and pharyngeal/laryngeal cancers, 11.7 (95% CI 8.8–15.6) for current smokers and heavy drinkers without family history, and 42.6 (95% CI 18.2–99.8) for smokers and heavy drinkers who also reported a first degree relative with oral and pharyngeal /laryngeal cancers.
Table IV shows the number of cases and controls with history of other selected cancers in first-degree relatives, and the corresponding ORs. A significant increase in oral and pharyngeal cancer was observed with a family history of melanoma (OR = 5.8; 95% CI, 1.3–26.4) and lung cancer (OR = 1.4; 95% CI, 1.0–2.0). None of the other cancer sites showed a significant association. The OR was 1.3 (95% CI, 1.1–1.6) for all cancer sites combined, and 1.1 (95% CI, 0.9–1.4) for all sites except oral cavity, pharynx and larynx.
Table IV. Odds Ratio of Oral and Pharyngeal Cancer According to Family History of Selected Cancers in First-Degree Relatives (Italy and Switzerland, 1992–2005)
In this study, a family history of oral and pharyngeal cancer and/or, laryngeal cancer in first-degree relatives was directly and strongly associated with the risk of oral and pharyngeal cancer, and the risk was higher when 2 or more relatives were affected, and independent from alcohol and tobacco consumption. The risk of oral and pharyngeal cancer was also increased in subjects with a family history of cancers of the lung and skin melanoma.
With reference to potential sources of bias, we excluded from the control group all diagnoses potentially associated with the increased risk of oral and pharyngeal cancer, such as tobacco and alcohol related conditions.9 Moreover, the almost complete participation has likely reduced selection bias. Information on family history was self reported, and it is possible that cases of oral and pharyngeal cancer may tend to recall a family history of oral and pharyngeal cancer or other cancers more accurately than controls. Several studies, however, have shown that the recall of cancer in first-degree relatives is satisfactory and comparable in cases and controls, while the recall of cancer in second-degree relatives is less reliable.10–12 Therefore, we only considered first-degree relatives in our analysis. In a reproducibility analysis in our population, the Kappa of agreement between 2 interviews was >0.8 for family history of digestive tract cancers in first degree relatives.13 Moreover, family history of most cancer sites was not associated with an increased risk of oral and pharyngeal cancer in this study, suggesting that no major recall bias of cancer in general has affected our results. With reference to confounding, we allowed for tobacco and alcohol, the major risk factors for oral and pharyngeal cancer, education as a social class indicator, and for number of brothers and sisters.
Our findings of an elevated risk of oral and pharyngeal cancer in subjects with family history of oral and pharyngeal/laryngeal cancers are in broad agreement with other reports, although our point estimate is somewhat higher than in other studies.3–5
An inherited component of susceptibility to oral and pharyngeal cancer has been suggested by case reports of families with multiple affected members,14–16 by epidemiologic studies indicating familial tendency to oral and pharyngeal cancer or other cancers of upper aerodigestive tract,1, 17–23 by segregation analysis in first-degree relatives,24 by elevated risks associated with polymorphic genes involved in the metabolism of tobacco and alcohol (i.e., ALDH2),25–29 and by elevated risks associated with genes involved in DNA repair maintenance of genetic stability.30, 31
Familial aggregation of oral and pharyngeal cancers may also be due to shared environmental exposure to the main risk factors, i.e., alcohol and tobacco. However, the risk of oral and pharyngeal cancer was only moderately increased in individuals with a family history of lung cancer, and not related to a family history of cancer at other sites related to alcohol or alcohol and tobacco (i.e., esophagus or liver),9, 32 suggesting that tobacco and alcohol cannot totally explain this association. Our finding for a family history of lung cancer is consistent with the results of case-control studies from the United States3 (OR = 1.2) and Puerto Rico4 (OR = 1.2).
We found an association with family history of melanoma. Although the point estimate of the OR was around 6, this estimate was based on 6 cases and 3 controls only, with a lower confidence limit of 1.3. This association has not to our knowledge been previously reported, and needs therefore independent confirmation. In fact, given the multiple tests performed, this result may be due to chance.
The analysis of the joint effect with alcohol and tobacco found that family history was also a risk factor in nonsmokers and non or moderate drinkers. The ORs of family history in the low exposure category of smoking and drinking was 1.9, while the OR for high exposure to tobacco and alcohol in subjects without family history was 11.7. The OR of those with family history and high exposure to alcohol and tobacco predicted by assuming an additive effect would be 13, and 22 by assuming a multiplicative model.33 Our point estimate of 42.6 suggests, if anything, a multiplicative or a supra-multiplicative effect. In agreement with a study from Puerto Rico,4 compared to the lowest risk category, i.e., light smokers and light drinkers without family history, the OR was 1.8 for light smokers and drinkers with family history of oral and pharyngeal cancer/laryngeal cancer, 12.2 for heavy smokers heavy drinkers without family history, and 60.4 for those who were heavy smokers, heavy drinkers and also reported a first degree relative with upper aerodigestive tract cancer.
Thus, our study adds further and more precise quantification on the association of family history of cancer and oral and pharyngeal cancer and is, to our knowledge, the first comprehensive epidemiologic information on family history of cancer at several sites and risk of oral and pharyngeal cancer. Our results show that the best way to prevent oral and pharyngeal cancer is to avoid alcohol and tobacco exposure, even in subjects with a family history of this neoplasm. There is ample scope for prevention of oral and pharyngeal cancer, a neoplasm that has shown little sign of improvement in survival over the last few years.34
The authors thank C. Pasche and F. Lucchini for the Swiss data collection and validation, and Ms. I. Garimoldi for editorial assistance.