Abstract
Objectives
The aim of this study was to determine whether feline chronic gingivostomatitis (FCGS) is more prevalent in shared vs single-cat households, whether the number of cohabiting cats or outdoor access represent risk factors for FCGS and whether the number of cohabiting cats is a useful prognostic indicator for standard surgical treatment.
Methods
Cats diagnosed with FCGS (study group) in the past 5 years at a referral institution were identified. The number of cohabiting cats, outdoor access, number of other cohabiting cats diagnosed with FCGS, ⩾6 month surgical outcome, when applicable, and historical signs of upper respiratory disease among any of the cohabiting cats, as well as patient demographic information, were recorded. Data were collected from medical records and by means of a telephone interview with the owners. The same information was collected from a group of cats of similar demographic characteristics diagnosed with periodontal disease but free of FCGS (control group).
Results
Seventy-six cats were included, of which 36 (47%) had FCGS and 40 (53%) served as controls. Bivariate analysis showed that cats with FCGS were significantly more likely to come from shared households, and had significantly more total cats per household compared with controls. Multivariate analysis also showed that cats in shared households had a significantly increased odds of FCGS compared with those from single-cat households. Historical signs of upper respiratory disease and outdoor access among cats within the same household were not associated with FCGS. The number of cohabiting cats was not associated with surgical outcome.
Conclusions and relevance
Cats with FCGS are more likely to live in shared households. The risk of FCGS correlates with the number of cohabiting cats. The epidemiological features of FCGS may support an infectious etiology. The number of cohabiting cats within a household is not a useful prognostic indicator for standard surgical treatment of FCGS.
Keywords
Introduction
Feline chronic gingivostomatitis (FCGS) is a relatively common disease of major clinical significance.1,2 It is clinically characterized by severe chronic inflammation and ulceration of the caudal and buccal oral mucosa, and occasionally other oral mucosal surfaces.1–4 Typical clinical signs include severe oral pain with reduced food prehension ability, ptyalism, mandibular lymphadenopathy, reduced grooming behavior and halitosis. 4 Additionally, affected cats frequently exhibit severe and extensive forms of periodontitis and tooth resorption. 5 Several treatments have been described,6–12 but the long-term outcome may be unfavorable. 2 Moreover, no prognostic indicators have been identified, making outcomes unpredictable. Refractory animals may require long-term palliative care, and may be euthanized owing to quality of life concerns.2,13
Multiple studies on FCGS have been published in the past few decades, with the vast majority focusing on etiopathogenesis or therapeutic approaches. Despite the progress made, the etiology of the disease remains unknown and available treatments continue to lack a rational basis. In addition, epidemiological studies of the disease are scarce,1,14 and many features have yet to be documented. Understanding the epidemiological characteristics of the disease is essential to better understand the possible etiopathogenic mechanisms involved, to propose novel treatments or preventive approaches, and to validate or refine current approaches.
Currently available options for the treatment of FCGS include medical and surgical intervention. 2 Initial medical management consists of palliative measures, including systemic analgesics to treat associated pain, anti-inflammatories to treat the oral inflammation, and antibiotics to treat secondary infections. Unfortunately, the long-term prognosis of this approach is poor. 10 The most widely used surgical treatment consists of extracting all premolar and molar teeth, or performing full-mouth extractions.11,15 The reported partial or complete remission rate of surgical treatment ranges between 70% and 80%. 11 Other available treatments are described mostly for cases that fail to respond to surgical intervention and offer variable response rates, and include systemic ciclosporine, 12 topical or systemic feline recombinant interferon omega, 10 and the systemic administration of adipose-derived stem cells.6,7
Several viruses of worldwide distribution have been associated with FCGS, including feline calicivirus (FCV),16"/>–26 feline herpesvirus (FHV), 22 feline leukemia virus (FeLV) 27 and feline immunodeficiency virus (FIV).20,27 Many of the epidemiological and clinical features of these pathogens have been documented.28"/>–32 For example, it is well established that multi-cat environments such as catteries, shelters and shared households, as well as roaming behavior, represent risk factors depending on the pathogen. Therefore, if the etiology of FCGS does, in fact, involve infection or coinfection with any of these pathogens, it is possible that cohabitation with other cats and outdoor access represent risk factors for the disease. Moreover, given the potential of some individuals to undergo cyclic re-infection when chronically exposed to viruses,33,34 it is also possible that the number of animals sharing a household could impact the outcome of surgical intervention, and thus may represent a useful prognostic indicator.
The purpose of this study was to determine whether FCGS is more prevalent in shared households than in single-cat households, whether outdoor access and number of cats within a household represent risk factors for FCGS, and whether the number of cats sharing a household can be used as a prognostic indicator for standard surgical treatment of FCGS.
Materials and methods
Study and control groups
The electronic medical records of the Cornell University Hospital for Animals were searched to identify client-owned cats with FCGS (study group) that were diagnosed by, or under the direct supervision of, a board-certified veterinary dentist at the Dentistry and Oral Surgery Service between 2013 and 2018. Patient demographic information collected from the medical records included sex (male or female), breed (domestic or purebred), age (years) and body weight (kg) at the time of diagnosis. Additional information was collected by means of a telephone interview with the owner of the cat using a standardized questionnaire. All phone calls were made by a veterinary student who was unaware of the aims of the study at the time of data collection.
The information collected included the number of cats cohabiting within the same household at the time of diagnosis (or lack thereof); whether the cat or any of the cohabiting cats (when applicable) were allowed to roam freely outdoors; and whether the cat or any of the cohabiting cats (when applicable) had historical signs of upper respiratory disease. Additionally, owners were asked whether any of the other cohabiting cats (when applicable) had been professionally diagnosed with FCGS while sharing the household with the affected animal, and how many (when applicable). The same information was collected from cats diagnosed with periodontal disease and free of FCGS (control group) at the Cornell University Hospital for Animals during the same time period.
Cats in the FCGS group that had been surgically treated (ie, premolar and molar teeth or full-mouth extractions) by, or under the direct supervision of, a board-certified veterinary dentist at Cornell University’s Dentistry and Oral Surgery Service 6 months or more prior to enrollment were identified. Additional information was collected during telephone interviews to determine surgical outcome. Outcome was classified as poor when the owner responded yes to any of the following questions: (1) does the cat have persistent/unresolved oral pain or difficulty eating, (2) is there persistent/unresolved severe bad breath, (3) is there persistent/unresolved drooling, (4) is there persistent/unresolved inflammation in the mouth. The outcome was considered good when the owner responded ‘no’ to all of the questions above.
The study was exempt from Institutional Animal Care and Use Committee approval.
Statistical analysis
Continuous patient demographic variables (age, body weight, total number of cats in the household) were assessed for normality via the Shapiro–Wilk test and by visual inspection of Q–Q plots. For those with approximately normal distributions, the sample means were compared between the FCGS and periodontal disease (PD) groups via an independent t-test. If the variables showed substantial deviations from normality, the two groups were compared by Wilcoxon’s rank sum test. For dichotomous variables (breed, single- vs multi-cat household, indoor vs indoor/outdoor, historical signs of upper respiratory disease, presence/absence of other cats with FCGS in the household, type of surgery and surgical outcome), the FCGS and PD groups were compared by the χ2 test, with associated odds ratios (ORs) and 95% confidence intervals (CIs).
To determine if living in a shared household is associated with the odds of FCGS, a multivariable logistic regression model was evaluated, with diagnosis (FCGS vs PD) as the outcome and with all candidate predictors (single- vs multi-cat household, age, body weight, breed, indoor vs indoor/outdoor, historical signs of upper respiratory disease, presence/absence of other cats with FCGS in the household) as independent variables. Backward selection with a retention P value of 0.2 was used to choose predictors for the final model.
To determine if the number of cats living in the household is predictive of the odds of FCGS, a similar approach of logistic regression with backward selection was employed, replacing the dichotomous single-/multi-cat household indicator with the total number of cats in the household as the predictor of interest. All other candidate predictors remained the same in this model.
To assess the relationship between the number of cats in the household and the odds of a positive surgical outcome, multivariable logistic regression was performed with surgical outcome as the dependent variable and total number of cats in the household as the predictor, along with other candidate predictors (age, body weight, breed, indoor vs indoor/outdoor, presence/absence of historical signs of upper respiratory disease, presence/absence of other cats with FCGS in the household). Backwards selection with a P value threshold of 0.2 was again employed to determine the final model.
Collinearity diagnostics (tolerance, variance inflation factor) were assessed in a linear regression model. All statistical analysis was performed using SAS version 9.4.
Results
Seventy-six cases met the criteria for inclusion, of which 36 (47%) were cats with FCGS and 40 (53%) were PD controls. Median age was 6 years (range 1–19 years). Cats had a mean body weight of 4.8 kg. Sixty-six (87%) were domestic shorthair cats and 10 (13%) were purebred. With regard to sex, 32 (42%) were spayed females and 44 (58%) were castrated males. Sixty-two cats (82%) lived in multi-cat households, with the median number of cats being 3 (interquartile range 2, range 1–11). Fifty (66%) were indoor only, while 26 (34%) had historical signs of upper respiratory disease. Other cats diagnosed with FCGS were reported in 13 households (17%). Of the 36 FCGS cats, 29 (81%) had surgery to address the condition 6 months or more prior to inclusion; 23 (79%) of those had full-mouth extractions and six (21%) had premolar and molar teeth extractions performed. Twenty-two (76%) of the cats that underwent surgery had a good outcome, with the remaining seven (24%) reported as poor.
Table 1 shows the baseline characteristics of the FCGS cases and PD controls. FCGS cats were more likely to come from multi-cat households (OR 7.29, 95% CI 1.50–35.31; P = 0.0061), had lower mean body weights (4.43 kg vs 5.19 kg; P = 0.0226) and had significantly more total cats per household (median 3.5 for FCGS vs 2 for PD; P = 0.0025). Additionally, results showed that multi-cat vs single-cat household increased the odds of historical signs of upper respiratory disease (OR 9.56, 95% CI 1.17–77.9; P = 0.035) and that each additional cat in the household increased the odds of historical signs of upper respiratory disease by 30% (OR 1.30, 95% CI 1.05–1.60; P = 0.0164).
Baseline characteristics of the study population
IQR = interquartile range; FCGS = feline chronic gingivostomatitis; PD = periodontal disease
In multivariable logistic regression, cats cohabiting with other cats had a significantly increased odds of FCGS (OR 7.05, 95% CI 1.08–46.12; P = 0.0416) compared with cats with PD. Other variables retained at the P <0.2 level included single vs multi-cat household, breed, sex, indoor/outdoor status, presence of historical signs of upper respiratory disease, presence/absence of other cats with FCGS in the household, and body weight; ORs and associated CIs are shown in Table 2. The model using the total number of cats in the household as a predictor additionally retained the variable age, and found that each additional cat in the household increased the odds of FCGS by 72% (OR 1.72, 95% CI 1.14–2.58; P = 0.0094).
Multivariable model predicting odds of feline chronic gingivostomatitis (FCGS)
OR = odds ratio; CI = confidence interval; NA = not applicable
For surgical outcomes, neither the dichotomous single/multi-cat household indicator nor the total number of cats was predictive of outcome in either bivariate analysis or multivariable logistic regression (all P values >0.56). Multivariable logistic regression with all candidate predictors retained only sex and presence/absence of historical signs of upper respiratory disease at the <0.2 level, although neither achieved statistical significance (sex, odds of good surgical outcome in spayed females vs castrated males: OR 4.27, 95% CI 0.50–36.32 [P = 0.1836]; historical clinical signs of upper respiratory disease, odds of good surgical outcome in those without signs vs those with clinical signs: OR 4.27, 95% CI 0.63–29.14 [P = 0.1381]). The type of surgery performed (ie, full-mouth vs premolar and molar teeth extractions) was not associated with outcome in either bivariate or multivariate analyses.
Discussion
The epidemiological features of FCGS reported in this study complement current knowledge of this devastating and poorly understood disease. This study showed that FCGS is more prevalent in multi-cat than in single-cat environments and that the risk correlates with the number of cohabiting cats. These findings support the notion that the etiopathogenesis of FCGS may involve infectious mechanisms, as has been previously suggested.35,36
Some of the infectious agents that have been implicated with FCGS include FCV, FHV, FeLV and FIV.10,16–25 Consistent with our observations about FCGS, the prevalence of FCV, FeLV and FHV is higher in multi-cat environments.29–32 Moreover, as was the case with FCGS in this study, the prevalence of FCV infection is proportional to the number of cohabiting cats.30,37 Therefore, the results of this study show that it is possible that FCV, FHV and/or FeLV are involved with the pathogenesis of FCGS. It should be noted that conflicting evidence exists regarding FHV and FeLV and their association with FCGS: although some studies have shown an association,20,38 others have not.17,24 Conversely, there is consistent evidence that FCV is associated with the disease, and an etiologic role is suspected.16–18,20,21,23,24
Free-roaming behavior is a known risk factor for FeLV, FIV, FHV and FCV infection.28,29,39 Thus, the lack of association between outdoor access and FCGS observed in this study could suggest that infection alone is not sufficient to cause the disease and that additional conditions related to multi-cat environments are required. For example, multi-cat conditions allow permanent exposure of cohabiting cats to viruses shed by chronic carriers, favor high rates of viral evolution and facilitate cyclic reinfection of susceptible animals.30,34,37,40 Non-infectious mechanisms related to multi-cat environments could also play a role in the etiopathogenesis. One possible explanation would be that the stress of living in multi-cat environment may predispose certain individuals to adverse outcomes like development of FCGS.
Unlike FCV, FHV and FeLV, the prevalence of FIV is not associated with multi-cat environments. 28 Instead, given that transmission mostly occurs via bite wounds, free-roaming cats are at higher risk of FIV infection than strictly indoor cats.28,41 Neither of these epidemiological characteristics are consistent with those observed in this study regarding FCGS, which suggests that FIV may not play a primary role in the etiopathogeneis of the disease.
There are other pathogens of domestic cats that are also common in multi-cat environments. Specifically, the prevalence of Bartonella species and Bordetella bronchiseptica has been shown to be higher in such housing conditions.42–44 However, studies have shown that neither is associated with FCGS.17,18,24,35 Therefore, any epidemiological similarities between FCGS and the prevalence patterns of these two pathogens are most likely due to their infection and transmission mechanisms instead of a true association with the disease. It should also be noted that other pathogens with prevalence patterns comparable to the ones reported here in regards to FCGS could exist that might be involved in the pathogenesis of the disease. Therefore, the information presented here should not be interpreted as conclusive evidence that FCV, FHV or FeLV are involved in the pathogenesis of FCGS, especially considering that no studies have demonstrated that any of these viral agents fulfil Koch’s postulates in regard to this disease.
In this study, historical signs of upper respiratory disease were significantly more common in cats from shared households compared with single-cat households, and their prevalence was correlated with the number of cohabiting cats. This was not surprising considering that upper respiratory signs in cats are often caused by several different infectious agents, several of which are more common in multi-cat environments, including FHV, FCV and B bronchiseptica.44–47 Interestingly, this study showed that historical signs of upper respiratory disease were not associated with FCGS. This was unexpected considering that some of the infectious agents that have been associated with FCGS are also known to cause upper respiratory disease, including FHV and FCV. Taken together, the results of this study suggest that FCGS and upper respiratory infections in cats occur as independent events, and would suggest that they are caused by different mechanisms or infectious agents.
The overall results of surgical treatment of FCGS observed in this study were consistent with previous reports that show that 20–30% of cases fail to resolve, and that the outcome is not associated with the type of surgery performed (ie, premolar and molar teeth extractions vs full-mouth extractions).11,15 Unexpectedly, this study did not show an association between surgical outcome and the number of cohabiting cats. An association between the number of cohabiting cats and a poor response to surgical treatment had been expected based on the premise that FCV might be involved with the pathogenesis of FCGS, and that cyclic re-infection among cohabiting cats is a common event. 34 However, the lack of association observed would suggest that other factors are involved in the outcome of surgical treatment of FCGS. Identifying such factors would be of clinical value but would require additional studies.
It is noteworthy that at least one additional cat with FCGS was reported in 17% of the households of the study group. Although the diagnosis of these additional cats was not confirmed by the authors, the question posed to the owners asked if the diagnosis was made by a professional, so the responses are credible. Such apparent high prevalence of FCGS in multi-cat environments contrasts with the prevalence in the general cat population (<1%) 1 and further suggests that multi-cat environments represent a risk factor for the disease.
The finding that FCGS is common in shared households is not new. At least three studies have reported that a large proportion of cats with FCGS live in shared households.11,14,18 However, these studies did not aim to compare the housing conditions of cats with and without the disease, and a control group was not used to demonstrate or quantify any such association. Client-owned cats with PD were selected as controls in this study in order to minimize potential confounding factors. That is, no known predispositions or differences between PD and FCGS have been reported in regard to sex, breed or age. Moreover, studies have shown that PD is not associated with FCV infection. 39 One interesting difference found between the two groups was the significantly lower body weight of cats with FCGS. This difference could suggest that a lower body weight represents a risk factor for FCGS. However, a much more likely explanation is that the cats in the FCGS group were suffering from significantly more oral pain and reduced prehension ability, which, in turn, resulted in weight loss. A longitudinal study based on body weights and body condition scores would be necessary to demonstrate whether either of these explanations are correct.
Conclusions
Cats with FCGS are more likely to live in shared than single households; the risk of FCGS correlates with the number of cohabiting cats; historical signs of upper respiratory disease and FGCS occur independently; and the number of cohabiting cats within a household is not a useful prognostic indicator for standard surgical treatment of FCGS.
Footnotes
Acknowledgements
We thank Ashley Nichter and Stephanie A Rey for conducting the telephone interviews used to collect data.
Conflict of interest
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The authors received no financial support for the research, authorship, and/or publication of this article.
