Materials and Methods

Animals—Cats examined at the University of Minnesota Veterinary Medical Center were recruited for the study. Cats selected for the study were all deemed healthy on the basis of owner reports and results of physical examinations. Cats were included when the owner consented to inclusion of the cat in the study and the animal had no history or clinical signs of major systemic illness, such as cancer, renal disease, hepatic disease, or endocrinopathy. The study was approved by the University of Minnesota Institutional Animal Care and Use Committee (assurance of compliance No. A3456-01). The University of Minnesota Institutional Review Board Human Subjects Committee reviewed the protocol and determined the study to be exempt from full review under federal guidelines 45 CFR 46.101(b) category 2.

Information regarding age, breed, sex, and health history was obtained for each cat. In addition, a urine sample was collected from each cat. Urine samples were collected by use of midstream catch during natural voiding or by cystocentesis. Urine samples were stored at −80°C until analyzed.

Assessment of exposure to ETS—Owners who consented to the inclusion of their cats in the study were asked to complete a questionnaire regarding exposure of their cats to ETS. Information was collected regarding the number of smokers in the household, daily cigarette use by all smokers in the household, number of cigarettes smoked in the house, percentage of time the cat was in the same room while a person was smoking, use of other tobacco or nicotine products, and number of years that the cat had been exposed.

Measurement of urinary concentrations of nicotine, cotinine, and NNAL—Total nicotine (nicotine plus nicotine glucuronide) content and total cotinine (cotinine plus cotinine glucuronide) content were measured in urine samples by use of gas chromatography-mass spectrometry, as described elsewhere.⁷ Internal standards ([CD₃]cotinine and [CD₃]nicotine) were added to each sample. Samples were also assayed for total NNAL (NNAL plus NNAL glucuronides), which are major metabolites of the tobacco specific N-nitrosamine NNK, by gas chromatography with nitrosamine-selective detection.⁸ For that analysis, iso-NNAL was used as the internal standard. Positive control samples were included with each set of urine samples; the positive control samples consisted of the specific analyte added to water. Water was used as a negative control sample with each set of urine samples. For simplicity, total NNAL was referred to as NNAL for this report.

Statistical analysis—Descriptive characteristics of the population of cats were compared between cats exposed to ETS and unexposed cats. Categoric variables, including breed, length of coat, and sex, were compared by use of a Fisher exact test. Continuous variables, including age and body weight, were compared by use of a Wilcoxon 2-sample test.

Concentrations of nicotine, cotinine, and NNAL in urine samples were compared between groups. For cats with values below the limit of detection, one half the value for the limit of detection was used in the analyses. Wilcoxon 2-sample tests were also conducted to compare distributions of nicotine, cotinine, and NNAL between groups.

For all tests, values of P < 0.05 were considered significant.

Results

Sixty-one healthy cats were recruited for the study. Characteristics of this population of cats were summarized (Table 1). Nineteen cats were from households in which cigarette smoking was reported, whereas 42 cats were from households in which there were no smokers. There was no detectable difference in age, body weight, sex, and length of coat between cats from households with smokers and nonsmokers. With regard to breed distribution, domestic shorthair cats were more likely to be from households with nonsmokers than from households with smokers.

Cats from households with smokers had lived in these environments for a mean of 4.4 years (range, 1 to 14 years). Fourteen of 19 (74%) cats had lived in these households for their entire lives. Cats were exposed to ETS from a mean of 21.5 cigarettes/d (Table 2). Three owners reported that they did not smoke in the house and that their cats were not exposed to ETS. All owners were also queried about the use of tobacco products other than cigarettes, such as chewing tobacco, and the use of nicotine-replacement products, such as nicotine gum or patches. None of the owners reported use of other tobacco products or nicotine-replacement products.

Urine (mean ± SD, 11.4 ± 7.90 mL) was collected from each cat. Concentrations of total nicotine, total cotinine, and NNAL were significantly higher in the urine of cats with smokers in the household, compared with concentrations for cats from households with no smokers (Table 3). Total NNAL content was greater than the limit of detection in 11 of 19 (57.9%) cats that were from households with smokers, whereas total NNAL content was greater than the limit of detection in only 3 of 42 (7.1%) cats that were from households with no smokers.

Mean ± SD total nicotine concentrations were significantly (P < 0.001) higher in urine from cats that had detectable amounts of NNAL (84.90 ± 94.70 ng/mL), compared with concentrations in urine from cats that did not have detectable NNAL concentrations (7.53 ± 14.90 ng/mL). Similarly, mean concentrations of total cotinine were significantly (P < 0.001) higher in urine samples in which NNAL was detectable (10.80 ± 10.80 ng/mL), compared with cotinine concentrations when NNAL was not detected (0.90 ± 1.96 ng/mL). In addition, when the data were separated on the basis of ETS-exposure status, both total nicotine and cotinine concentrations were significantly higher in urine samples in which NNAL was detected.

For urine samples of cats from households with smokers, mean ± SD total nicotine concentration was significantly (P = 0.03) higher in samples that had detectable amounts of NNAL (100.00 ± 102.00 ng/mL), compared with concentrations in samples in which NNAL was undetectable (29.20 ± 28.00 ng/ mL). Mean total cotinine concentration in samples with detectable amounts of NNAL (12.80 ± 11.40 ng/ mL) was significantly (P = 0.01) higher than in samples with undetectable NNAL concentrations (68.00 ± 4.17 ng/mL).

For samples collected from cats that lived in households without smokers, mean ± SD total nicotine concentration differed significantly (P < 0.001) when NNAL was detectable and not detectable (28.40 ± 12.30 ng/mL and 3.09 ± 2.44 ng/mL, respectively). In those same urine samples, mean total cotinine content differed significantly (P = 0.01) when NNAL was detectable and not detectable (3.30 ± 2.36 ng/mL and 0.54 ± 0.81 ng/mL, respectively).

Discussion

Urinary biomarkers of tobacco, including nicotine, cotinine, and NNAL, have been used extensively in the evaluation of human populations, including, smokers, nonsmokers, and people exposed to ETS.^6,9–12 Objective measurements of exposure may facilitate study of the effects of ETS on health. Nicotine and its primary metabolite, cotinine, provide information about exposure but are not carcinogenic. Thus, NNAL, which is a major metabolite of the tobacco-specific carcinogen NNK, may be of use in the study of tobacco-associated carcinogenesis.⁶

Analysis of our results revealed that cats take up and metabolize tobacco-associated chemicals, including the carcinogen NNK, from their home environments. Although inhalation may be the major route of exposure to ETS in cats, it has been speculated that the meticulous grooming behavior of cats leads to increased carcinogen exposure through the oral route. Nicotine is found in hair and on household surfaces, both of which could serve as a source of exposure for pets.¹³

Although concentrations of all 3 tobacco biomarkers were increased in urine from cats that lived with smokers, compared with concentrations in urine from cats that lived with nonsmokers, 8 of 19 cats that reportedly lived in households with smokers had undetectable amounts of NNAL in their urine. Analysis of our results indicated that urinary concentrations of nicotine and cotinine are significantly higher in samples in which NNAL is detected. This suggests that cats with undetectable urinary concentrations of NNAL have had less exposure to ETS than the cats with measurable urinary concentrations of NNAL. Consistent with this observation is that cats from 3 households in which all smoking was conducted outside the house had undetectable concentrations of NNAL. Thus, the degree of exposure to ETS and tobacco-associated carcinogens may vary, even in a population supposedly exposed to ETS.

Because chemical mutagenesis and carcinogenesis are typically dose-related phenomena, it is probably not sufficient to simply classify populations as exposed or unexposed; instead, it is necessary to consider the amount of exposure and to have a method for quantifying exposure. With a more sensitive detection method, it is possible that more of the cats from households with smokers would have had evidence of exposure to NNK and thus evidence of urinary NNAL. Whether this would be biologically relevant can only be determined by studies that investigate the health effects of various amounts of exposure. Certain carcinogens may have a threshold effect whereby exposure below a certain value does not increase the risk of cancer.

Another point to consider is that a single urine sample was collected from these cats, which provides an estimate of exposure over a relatively short period. Collecting urine over a longer time period or collecting multiple samples to determine how reflective these spot samples were of overall exposure would have been useful.

It is interesting that 3 cats in the unexposed group had measurable urinary concentrations of NNAL. Compared with values for the remainder of the unexposed cats, these 3 cats had the highest values for urinary concentrations of nicotine, were in the top one third of urinary concentrations of cotinine, and had significantly higher mean urinary concentrations of nicotine and cotinine. We speculate that these cats may have been exposed to ETS, contrary to the beliefs of the owners. There are many reasons that an owner may not have provided accurate data concerning exposure of their pet to ETS. They may have been concealing their own smoking habit from others or they may have been unaware of the smoking habits of other family members, friends, and neighbors. Attitudes toward smoking and its social acceptability have changed a great deal in the past few decades. For many people, smoking may represent a source of guilt or embarrassment, and they may be unwilling to openly admit to it.^14,15 Identification of possible exposure to ETS among the reportedly unexposed group illustrates the importance of establishing biomarkers to better estimate exposure and avoid bias.

It may be useful to establish cutoff values for biomarkers that could classify individual animals as exposed or unexposed and to establish sensitivity and specificity of these assays for determining exposure status. However, it is probably premature to attempt to establish cutoff values. First, our standard for comparison was owner information, which may not have been particularly reliable. As mentioned, a small number of cats in the unexposed group may have actually been exposed. Second, in future investigations that use these biomarkers, it may be more useful to compare concentrations of tobacco biomarkers for a disease outcome, rather than use a simple dichotomous classification of exposed or unexposed. This would allow for detection of concentrations that are biologically relevant and identification of a threshold below which exposure may pose minimal risk. Finally, additional studies to compare fluctuations in biomarker concentrations over time in specific animals should be performed before committing to a single cutoff value.

In the study reported here, urinary concentrations of NNAL measured in cats from households with smokers were similar in magnitude to concentrations measured in human populations exposed to ETS, including nonsmoking women with spouses who smoke, infants with mothers who smoke, and hospitality workers exposed to ETS in the workplace.^8,9,12 Mean ± SD urinary concentrations of cotinine in exposed cats (8.53 ± 10.20 ng/mL) were similar to concentrations reported for ETS-exposed women (7.92 ± 8.98 ng/mL) but were lower than concentrations for ETS-exposed infants (23.4 ± 33.4 ng/mL) and hospitality workers (15.4 ± 17.9 ng/mL). In contrast, mean urinary concentrations of total nicotine were 10, 6, and 5 times as high in exposed cats (70.40 ± 86.00 ng/mL), compared with concentrations in ETS-exposed women (7.45 ± 10.0 ng/mL), hospitality workers (11.6 ± 11.9 ng/mL), and infants (13.4 ± 32.4 ng/mL), respectively.^8,9,12 Whereas the ratio of urinary nicotine to urinary cotinine in those human populations ranged from 0.57 to 1.33, the ratio of urinary nicotine to urinary cotinine in ETS-exposed cats in our study was 8.3. Even in unexposed populations, the ratio of urinary nicotine to urinary cotinine was 6.6 in the cats of our study, compared with 1.3 in women in another study,⁹ which suggests that proportionately more nicotine is excreted unchanged or conjugated with glucuronide in cats than in humans.

Although the literature regarding nicotine metabolism in cats is limited, studies^16–18 of experimental administration of¹⁴ C-labeled nicotine highlight similarities as well as differences between cats and other species, such as humans. Features of nicotine metabolism that appear to be shared among species include rapid distribution of nicotine to a wide array of tissues and metabolism of nicotine to cotinine, which begins within minutes after nicotine administration.¹⁶ In 1 study¹⁸ in which investigators administered radioactive nicotine by various routes, blood concentration of cotinine in cats typically was lower than the concentration in some of the other species studied, including squirrel monkeys and rabbits, which led the author of that study to conclude that cats are less similar to humans with regard to metabolism of nicotine to cotinine. In humans, approximately 80% of nicotine is metabolized to cotinine, and blood concentrations of cotinine in smokers are considerably higher than blood concentrations of nicotine.⁵ These differences may account for higher urinary concentrations of nicotine in humans, compared with the urinary concentrations in cats.

Evidence of ETS uptake coupled with the epidemiologic observations linking ETS exposure to certain malignancies in cats supports the supposition that ETS is carcinogenic for cats, similar to the situation in humans^2–4; however, additional investigations are clearly needed. Although associations between ETS and cancer in humans have been fairly weak and the risk conferred is of low magnitude, the consistency of these findings coupled with biomarker data led the Environmental Protection Agency to conclude that ETS is a carcinogen in humans.¹ By comparison, evidence for an association between cancer and ETS in companion animals is weak.

The study reported here revealed the feasibility for the use of urinary biomarkers in the assessment of ETS exposure in cats and will hopefully give rise to additional studies in which investigators use these biomarkers to investigate associations with disease. The cats selected for this study were all deemed healthy on the basis of owner reports and results of physical examinations. Use of these urine-based markers in a population of animals with disease, in which the animals may be considerably older and may have substantial metabolic limitations, could affect the concentrations of tobacco metabolites in the urine. In this study, we did not adjust for renal clearance of creatinine, which may be important in an older population of cats with disease. Another potential pitfall in the application of these biomarkers in the investigation of carcinogenesis is that they reflect current but not necessarily past exposure and represent single time points rather than cumulative exposure. These limitations would need to be considered when designing and interpreting results of studies.