Abstract
The purpose of this study was to determine the prevalence of p53 overexpression in feline oral squamous cell carcinomas (SCC) and to determine, if any, the association between p53 overexpression and lifestyle factors and environmental exposures, including exposure to environmental tobacco smoke (ETS). Questionnaires concerning exposure to ETS and other environmental factors were sent to owners of cats presenting to the Harrington Oncology Program with a diagnosis of oral SCC between 1991 and 2000. Additionally, 23 formalin-fixed biopsy samples from these cats, with information regarding ETS, were evaluated immunohistochemically for p53 expression using the CM-1 clone and the avidin-biotin-horseradish peroxidase method. Of the 23 samples evaluated, 15 (65%) showed positive nuclear staining for the CM-1 clone. Tumor biopsy samples from cats exposed to any ETS were 4.5 times more likely to overexpress p53 than were tumors from unexposed cats (P = 0.19). Among cats with any ETS exposure, those with 5 years or longer of exposure were 7.0 times more likely to overexpress p53 (P = 0.38). Longhaired cats (P = 0.18) and female cats (P = 0.35) were also more likely to show p53 expression in their tumors. These results provide additional support for a relationship between oral SCC development and exposure to household ETS and may implicate p53 as a potential site for carcinogen-related mutation in this tumor.
Keywords
Squamous cell carcinoma (SCC) is an aggressive and common oral neoplasm of the domestic cat. SCC is highly locally invasive and frequently ulcerative and may cause lytic change in underlying bony structures. Common presenting symptoms include ptyalism, halitosis, and mechanical interference with food prehesion. These tumors have a relatively low rate of metastasis but may spread to regional lymph nodes and, rarely, the lungs. 11
Effective therapies are few. Even with aggressive treatment including radical surgery, radiation therapy, and adjunctive chemotherapy, rates of survival for longer than 1 year are typically less than 10%. 11 Without treatment, affected cats are usually euthanatized within 4–6 weeks of diagnosis because of complications associated with local disease. 11
In both humans and domestic cats, SCC is the most commonly occurring oral neoplasia. 7 In humans there is a strong association between the development of SCC and the use of cigars, cigarettes, and smokeless tobacco products. 6,7,10 More recent studies have solidified this link by demonstrating the presence of adducts between DNA and tobacco–related carcinogens in these neoplasms and in oral squamous cells of individuals using tobacco products. 6,8,12 Hsu et al. 8 have demonstrated DNA adducts in oral and urothelial mucosa of smokers with both 4-aminobiphenyl (4-ABP) and polyaromatic hydrocarbons, two carcinogens found in cigarettes and cigarette smoke. Additional studies have suggested that oral SCC may be associated with occupational exposure to pesticides. 2,13 In contrast to human oral SCC, little is known about the etiology of feline oral SCC.
p53 is the most commonly disrupted gene in human cancer and is also the most frequently mutated gene in human oral cancer. 10 p53 mutations are seen in ocular SCC associated with exposure to ultraviolet light in both humans and domestic animals. 1,13–15 Neoplasia can be induced by inactivation of p53 (usually through mutation) or by inhibitory interaction with oncogene products. 10 p53 has also been found to be overexpressed in human tumors linked to carcinogens found in tobacco products, particularly oral SCC. 6,8–10 Furthermore, p53 mutations in conjunction with 4-ABP adducts have been demonstrated in human oral SCC and transitional cell carcinoma of the bladder, 6,8 implying a role for tobacco–mediated mutation of the p53 gene.
Tobacco smoke that is exhaled or is released into the air without being inhaled may affect the health of others. The so-called environmental tobacco smoke (ETS) contains carcinogens that may affect nonsmokers. High rates of p53 mutation in non–small cell lung tumors are seen in both smokers and patients who live with spouses who smoke, compared with those who neither smoke nor are exposed to ETS. 9 It has also been postulated that chronic exposure to smoke from tobacco products may put nonsmoking individuals at a greater risk of developing neoplasia. 9 In experimental studies, oral tumors have been induced in rodents through exposure of their oral mucosa to carcinogens found in smokeless tobacco. 7
Cats living with smokers may be exposed to the same environmental contaminants as their owners, both through inhalation and through oral ingestion during grooming of particulate matter deposited on the fur. 4 Like humans, cats exposed to household ETS metabolize nicotine into cotinine and demonstrate urinary cotinine levels that increase with exposure dose (Bertone, unpublished). Similarly, oral grooming may expose a cat's oral cavity to high levels of other chemicals, such as those contained in flea control products. If cats are exposed to a multitude of potentially toxic agents during grooming, the impact of these agents on their oral health must be considered. This includes their potential role in the development of oral tumors like SCC.
A recent study 3 used a case control design to evaluate how environmental factors may be associated with the development of feline oral SCC. The results of this study demonstrated an association between the development of feline oral SCC and flea collar use, dietary tuna intake, and high canned food intake. Any household ETS was associated with a marginally significant twofold increase in risk, with the suggestion of an even greater increase in risk in cats exposed for 5 years or longer and those living with two or more smokers. Although this study showed some interesting associations, implication of any of these substances in the etiology of oral SCC would be stronger if a biologic link could be established.
This study was designed to determine the prevalence and levels of p53 overexpression in feline oral SCC. This study also aimed to define the association between p53 overexpression and exposure to ETS, and other environmental carcinogens, as well as diet and various aspects of feline care. A survey–based case control study was used for this purpose.
Materials and Methods
Study population and sample collection
Eligible cases for this study included all cats presenting to the Harrington Oncology Program at the Tufts University School of Veterinary Medicine (TUSVM) between January 1991 and June 2000 with a biopsy-confirmed diagnosis of oral SCC and a paraffin–embedded biopsy sample available for immunostaining. These included cats with SCC in maxillary, mandibular, gingival, laryngeal, pharyngeal, and lingual locations as well as SCC classified only as oral (n = 32).
Questionnaires were mailed to the owners of cats with a diagnosis of oral SCC for whom current addresses were available. Owners who had not responded by mail within 2 months were sent a second copy of the letter and the questionnaire. Nonrespondents to the second mailing were telephoned and asked to complete the questionnaire over the phone.
Owners were asked to complete a two-page questionnaire inquiring about the home environment of their pet in a specific year, corresponding to 2 years before the diagnosis of oral SCC. Some of the questions were on characteristics of the cat including age, birth year, sex, breed, hair length, reproductive status, and general medical history. Other questions inquired about aspects of the cat's care and home environment, such as number of years owned; frequency of grooming, bathing, and teeth brushing; amount of time the cat spent outdoors during the day and the night; house size; home location; and home-heating sources. Questions pertaining to ETS exposure queried whether the cat had ever lived in the same household as a smoker, the types of tobacco products used (cigarettes, cigars, pipes), the number of years the cat had lived with a smoker, the total number of smokers in the household, and the average number of cigarettes smoked per day by all household members combined. To measure the use of flea control products, owners were asked whether they had ever used flea control products on their cat, what type of products had been used (collars, shampoos, sprays, powders, pills, drops), and how frequently. The usual diet of the subjects was measured by asking owners how much of their pet's total diet consisted of dry food, moist canned food, table scraps, canned tuna fish, and liver.
Immunohistochemical staining
Thirty-two formalin–fixed, paraffin–embedded tumor biopsy samples from 61 diagnosed cases of SCC were obtained from the TUSVM Section of Pathology and other independent diagnostic laboratories. Biopsy samples that had been subjected to ethylenediaminetetraacetic acid decalcification due to the presence of bone in the biopsy tissue (n = 4) were eliminated because of potential interference with antigen retrieval. 5 Also eliminated were samples from cats for which no ETS survey data were available. Hematoxylin and eosin (HE)–stained sections of each sample were reviewed by one pathologist to reconfirm the diagnosis of SCC.
The samples were subjected to immunohistochemical analysis using the p53 CM-1 clone (Biocare Medical). Immunohistochemical detection was performed according to the manufacturer's directions: 5-µm sections were cut and mounted on charged slides. One slide from all biopsy samples was routinely stained with HE and evaluated by light microscopy to confirm the diagnosis.
For p53 staining, slides were deparaffinized in xylene and rehydrated in graded ethanols. Slides were washed with 1× phosphate-buffered saline (PBS) and subjected to a methanolic block (peroxidazed, Biocare Medical) to block endogenous peroxidase activity. To access the nuclear antigen, slides were subjected to heat-induced epitope retrieval by boiling at 100 C for 10 minutes in a pH-buffered Tris solution (nuclear decloaker, Biocare Medical), followed by 20 minutes cool down. After washing twice with deionized water, nonspecific binding was blocked with a recombinant serum-free blocking agent (background sniper, Biocare Medical). To block endogenous biotin, slides were treated with an avidin–biotin block (Biocare Medical) according to the manufacturer's instructions and washed three times in 1× PBS. Slides were incubated with the polyclonal CM-1 antibody (1:100 dilution in DaVinci green diluent, Biocare Medical) for 30 minutes at room temperature, followed by 10 minutes incubation with biotinylated goat anti–rabbit antibody (universal link, Biocare Medical). After two washes in 1× PBS, slides were incubated for 5 minutes with a streptavidin–peroxidase conjugate (Biocare Medical). After washing in 1× PBS, slides were treated with diaminobenzidene chromagen substrate according to the manufacturer's instructions (Biocare Medical) and counterstained with hematoxylin.
Slides were dehydrated in serial ethanols and cleared in xylenes and coverslips mounted with permount (Fisher Scientific). As a positive control for immunolabeling, sections from a human colon adenocarcinoma, known to overexpress p53, were used. Normal feline lingual and sublingual epithelia, obtained from an owner-donated, 12-year-old (approximately the mean age of our case population) domestic shorthaired cat served as a negative control and were immunohistochemically examined to verify the absence of nonspecific p53 staining. An additional negative control included known positive tissues processed in an identical manner in the absence of p53 CM-1 antibody.
Relative staining intensity was evaluated by a single individual without knowledge of any survey responses. Intensity was categorized as follows: (−) none or little nuclear staining, (+) < 50% of nuclei staining intensely, (++) > 50% of nuclei staining intensely.
Statistical analysis
The results of p53 staining in tumor biopsy samples were correlated with the survey results for patients with information available. Several aspects of ETS exposure were evaluated, including ever versus never exposed and, among cats ever exposed to household ETS, duration of exposure, number of household smokers, and average total number of cigarettes smoked per day by all household members. Flea control product use was categorized based on use of any flea control product, then by use of specific types of products (collars and shampoos). Cats were also divided into groups based on the makeup of their usual diet (mostly dry food, mostly canned cat food) and whether or not they had ever consumed tuna fish.
Odds ratios (OR) were used to estimate the association between various risk factors and p53 overexpression, with an OR of 1.0 indicating no association between exposure and overexpression. P-values from Fisher's exact test were used to evaluate the role of chance. A probability of P < 0.05 was considered statistically significant.
Results
Survey data
The study population consisted of 28 cats with both a diagnosis of oral SCC and available tumor biopsy samples that had not been decalcified. Current mailing addresses were obtained for all the diagnosed cases. Completed responses were obtained from 23 (82%) cases, 2 (8%) refused to participate, and 3 (13%) could not be reached by telephone.
Immunohistochemical staining
Of the 23 biopsy samples examined, 15 (65%), demonstrated positive nuclear staining for the p53 protein with the CM-1 antibody clone. Ten (43%) of the biopsy samples demonstrated a staining score of ++ (Fig. 1). Five of the biopsy samples (22%) demonstrated a staining score of +. No immunostaining was observed in eight of the tumor biopsy samples examined. Tumors exposed to ETS and staining positive for p53 were of maxillary, mandibular, and sublingual origin. Results of immunohistochemical staining in tumors at different oral locations are shown in Table 1.
Location of 23 SCCs and degree of p53 expression in the study population.∗
∗ (−), negative; (+), < 50% nuclei staining; (++), > 50% nuclei staining.
† Location within oral cavity not clearly specified in medical records.
Oral mucosal SCC in cat No. 5. Overexpression of p53 protein is noted in nuclei of tumor cell clusters. CM-1 antibody, avidin–biotin–horseradish peroxidase method, hemotoxylin counterstain. Bar = 25 µm.
Environmental exposures and p53 overexpression
Environmental factors were then evaluated for association with the likelihood of p53 overexpression. Tumor biopsy samples from cats exposed to ETS were 4.5 times more likely to overexpress p53 than tumors from cats that were not exposed to ETS, although the results were not statistically significant (P = 0.19). Among the population ever exposed to ETS, tumor biopsy samples from cats exposed for 5 years or longer were 7.0 times (P = 0.17) more likely to overexpress p53 than tumor biopsy samples from cats exposed for less than 5 years. In addition, female cats were five times more likely to overexpress p53 than males (OR = 5.0; P = 0.19), as were cats with short hair (OR of shorthaired versus longhaired cats = 5.0; P = 0.18). Expression of p53 was not strongly associated with any other factor (Table 2).
Odds Ratios (OR) for the association between various risk factors and the likelihood of p53 overexpression.
∗ Fisher's exact test.
† Analysis was limited to cats exposed to any household ETS.
Discussion
Our data indicate that abnormal p53 accumulation frequently occurs in oral SCCs of domestic cats and suggest that p53 dysregulation may be involved in the carcinogenesis of this tumor. These findings are in accordance with those of other investigators who reported p53 overexpression in seven of eight conjunctival SCCs in domestic animals 15 and aberrant p53 expression in 29–100% of various SCCs in other domestic animals, 14 suggesting that p53 mutations in these tumors may be induced by UV radiation. However, because the tumors examined in this study are located in the oral cavity, an area exposed to little or no UV radiation, a different carcinogenic insult is likely responsible for p53 overexpression.
Although not statistically significant, our findings show a strong association between p53 overexpression and exposure to ETS. Also, female cats and those having a shorthaired coat were found to have an increased risk of p53 expression, although this increase was not statistically significant.
It is difficult to explain why female cats and cats with a shorthaired coat might have an increased likelihood of p53 overexpression. One might speculate that the protective effect observed in the male sex might be due to hormonal influences on DNA repair and damage. This, however, seems unlikely because a high percentage (88%) of cats surveyed were neutered. If one presumes that longhaired cats groom more fastidiously than their shorthaired counterparts, then the observed protective effect undermines the supposition that coat-associated carcinogens are related to the development of oral SCC. However, it is possible that the particulate matter from coat-associated carcinogens such as ETS may migrate deeper than the cat can thoroughly groom, thereby resulting in lower exposures in longhaired than in shorthaired cats, which might groom more thoroughly. An alternative explanation is that there is differential migration of smoke-associated carcinogens, and those most likely to cause p53 mutations are able to migrate deeper and are therefore less accessible to longhaired cats during grooming. It is important to remember that the small sample size for these variables limits the statistical power to evaluate these associations. Also the lack of statistical significance does not eliminate the likelihood that chance played a role in these apparent associations.
Although canned food, and in particular tuna intake, as well as exposure to flea collars were shown to increase the risk of oral SCC in cats, 3 there was no associated increase in p53 expression in these tumors. In contrast, the association between ETS and oral SCC seen in a case control study 3 was even stronger here when evaluated using p53 expression and paralleled the dose response seen in that study. Unlike the protective effect seen with other variables, the association between p53 overexpression and ETS seems a logical one because human tumors associated with both direct and indirect exposure to tobacco carcinogens are frequently found to overexpress p53. 6,9
Studies in human populations have demonstrated the presence of DNA adducts with tobacco-related carcinogens in neoplasms associated with tobacco use, as well as in the oral mucosal cells of cancer-free tobacco users. 6,8,12 Similar studies would need to be conducted in feline populations to establish or deny a causative role for ETS. It must also be considered that the ETS may alone not be causal but a key exposure in a series of cumulative genomic insults. In such a scenario one may not see an accumulation of DNA adducts with tobacco-related carcinogens.
In order to maximize the number of eligible cases for inclusion in the study, many of our subjects were diagnosed up to 9 years before the time of mailing of the questionnaires. Such a significant time lapse may influence the accuracy of recall of certain factors, particularly those pertaining to use and quantity of tobacco products and frequency of flea product application. However, the misclassification caused by owners being asked to recall exposures over a long period of time is likely to be nondifferential with respect to case status and should therefore result in underestimation of the association between ETS exposure and p53 overexpression.
Many oral SCC in these cats were not associated with p53 overexpression. Rather than pointing to one etiologic factor, the findings of this study seem to suggest that the development of oral SCC may be associated with multiple pathways rather than exclusively with p53 mutation. In summary, the results of this study, although not statistically significant, suggest that p53 overexpression, and possibly ETS, may be involved in the development of feline oral SCC. Further investigation into the role of environmental mutagens in these tumors is warranted.
Footnotes
Acknowledgements
This research was supported by National Institutes of Health training grant T35DK 07635, The International Feline Foundation, and the Geraldine R. Dodge Foundation.