There were six articles that focused on surgical management of FCGS (11, 36, 39, 41, 45, 47). Of these, two articles were graded as level V C case reports discussing single cats; one cat underwent premolar and molar tooth extractions and achieved clinical remission that was sustained for at least 2.2 months (45), while the other cat was treated adjunctively with a carbon dioxide laser after dental extractions and achieved clinical remission that was sustained for at least 36 months (47).

There were three articles specifically investigating the efficacy of dental extractions in groups of cats, each reporting grade B evidence (11, 39, 41). The earliest of the three articles (39) included a sample size of 30 cats, in which a 60% clinical remission rate was achieved (no visible lesions, no oral clinical signs), with an additional 20% of cats considered significantly improved without the need for ongoing therapy, 13.3% of cats showing little improvement, and 6.7% of cats making no improvement. The next article (11) assessed the results of dental extractions performed on 21 cats; of these, 57.1% achieved clinical remission, 23.8% improved, and 19.1% suffered relapses. The results of these two articles are strikingly similar, with ~80% of both groups demonstrating substantial improvement or resolution of FCGS lesions and the remaining 20.0% of cats benefiting minimally or not at all from surgical treatment. The most recent and comprehensive article assessing dental extractions was a retrospective case series, including 95 cats treated with either full-mouth extractions or premolar and molar tooth extractions (41). This study reports that significantly fewer cats (28.4%, P = 0.002) recovered completely compared to the previous two studies (11, 39). While a greater percentage (39.0%) of cats showed substantial improvement, the authors point out that this was not statistically significantly different from the first extraction article’s (39) finding of 20.0% (P = 0.077). At the last recheck examination, 26.3% of their patients showed little improvement and required ongoing medical management [again, not statically significantly different from the first extraction article’s (39) finding, P = 0.214]. In this study, 6.3% of cats remained refractory to extraction treatment [yet again, not statically significantly different from the first extraction article’s (39) finding, P-value not provided]. Approximately one-third of the 95 cats had been treated with premolar and molar tooth extractions while the remaining approximately two-thirds were treated with full-mouth extractions; there was no significant difference in response to treatment between these groups (P = 0.377). These authors also found that treatment with antimicrobials, anti-inflammatories, or analgesics prior to or at the time of extractions was not associated with a better outcome. Cats that were reported to have resolution of abnormal behaviors associated with FCGS at the time of their first postoperative recheck had odds of a positive outcome (clinical remission or substantial improvement) 7.2 times as great as in cats without resolution. Just over two-thirds of the cats that achieved clinical remission or substantially improved did require additional medical management for a finite period of time beyond the immediate postoperative period, whereas the first study (39) reported no need for medical management in the 80% of cats that went into clinical remission or significantly improved, and the second study (11) did not discuss if ongoing medical management was necessary.

In the sixth article discussing dental extractions (36), the primary focus was not on surgical management, but instead the intent was to investigate if diets with differing omega-3 to omega-6 polyunsaturated fatty acid ratios would affect inflammation and wound healing when fed to cats postoperatively after premolar and molar tooth extractions. Cats’ FCGS was scored utilizing a five-point semi-quantitative scale examining the degree of inflammation preoperatively and 4 weeks postoperatively. While there was no statistically significant difference between groups fed different diets (P = 0.366), overall, there was significant improvement in FCGS scores of both groups (P = 0.017, P = 0.042), presumably owing to the dental extractions. With their data pooled, the 14 cats’ FCGS scores improved on average by 52.1% (range 0–72.0%).

The remaining 10 articles explore the efficacy of various medical management therapies (10, 12, 28, 37, 38, 40, 42–44, 46).

There are two studies that test the efficacy of oral cyclosporine; in the first (40), cyclosporine was administered to a group of eight cats that were not previously treated for FCGS with dental extractions, while in the second (12), cyclosporine was administered to nine edentulous cats. In the former study, 50.0% of cats achieved clinical remission, while the remaining cats showed 40.0–70.0% improvement in their semi-quantitative lesion scores during the 6-month follow-up period. In the latter study, 45.5% of cats achieved clinical remission, while 77.8% of cats showed a >40.0% improvement in their semi-quantitative stomatitis scores over the 6-week study period; by contrast, only 1 of 7 (14.3%) of cats in the placebo group showed a >40% score improvement.

A relatively similar success rate was achieved in a pilot study exploring the efficacy of fresh, autologous, adipose-derived mesenchymal stem cells intravenously injected into seven cats (10). Over the 6- to 24-month follow-up period, 42.8% of cats (n = 3) went into clinical remission, 28.6% (n = 2) demonstrated substantial improvement, and 28.6% (n = 2) did not respond. This article was awarded a level A EG because it not only utilized a semi-quantitative scoring system but also included histology from cats that achieved either clinical remission, substantial improvement, or that failed to respond. Histology results paralleled clinical gross examination of the lesions. While harvesting mesenchymal stem cells from the patient’s fat may preclude this treatment from easily being adopted widely at this juncture, the authors state that additional investigations using fresh, allogeneic cells are ongoing. Another unique aspect of this article is that the authors identified a potential biomarker to predict therapeutic outcome of stem cell treatment.

There were three articles, each testing a different therapy [Zincreo (43), thalidomide with lactoferrin (44), and recombinant feline interferon omega (46)], that were graded V C and each yielded a 100% remission rate in their single cat patient. In another article (42), graded IV C, recombinant feline interferon omega was administered to two cats, yielding a 100% remission rate.

The remaining three articles that discuss medical management included a control group (28, 37, 38). In one study (38), the success of treating with local paramunization using PIND-ORF (parapoxvirus ovis) versus “conventional treatment” (the authors did not define the specific control therapy) was compared. There were 33 cats treated with paramunization, resulting in 42.0% of them achieving clinical remission or substantially improving, while only 13.0% of the 39 cats receiving “conventional treatment” achieved the same positive results. Another study (37) tested the efficacy of bovine lactoferrin administered with piroxicam versus piroxicam alone as a control; 77.0% of cats receiving both medications achieved clinical remission or substantially improved after 12 weeks, whereas the authors did not state the success rate of the control group, as control cases were converted to treatment cases after the fourth week. Finally, a study (28) compared the effectiveness of recombinant feline interferon omega in 19 cats versus the control treatment of prednisolone in 11 cats; 45.0% of the treatment cats substantially improved (10.0% of which achieved clinical remission), while 23.0% of control cats substantially improved (7.0% of which achieved clinical remission). However, these differences between treatment and control group were not statistically significant.

Of the 16 articles included in this review, 6 were small retrospective case series or single patient case reports (EDG IV or V). Lacking any statistical power, these articles provide weak evidence. That being said, it is heartening that, in general, more recent studies have a higher EDG and EG than earlier studies. Ideally, studies would build off of the results of one another, expanding promising level V experiments to a level II or I clinical trial. However, in compiling FCGS articles, this is rarely the case. Instead, the literature is at times circuitous, expending efforts in studies that stand to change the overall conversation minimally. For example, after a level III B article (39) was published discussing efficacy of dental extractions in 30 cats, 8 years later a level V C case report (45) was published discussing success of premolar and molar tooth extractions in a single cat, which does not advance collective knowledge about FCGS treatment. Similarly, a level V C single patient case report (46) published on recombinant feline interferon omega was followed 7 years later by a level IV C case report (42) published testing the same drug in two cats with an almost identical follow-up time and result.

This review identified that 9 of the 16 articles assessed outcome of FCGS treatment via semi-quantitative scoring systems. A novel custom scoring system was developed and defined in four of these articles (28, 36, 39, 40), while two articles utilized the same scoring system (10, 12), and three articles each modeled their scoring system off of different previously utilized scoring systems (11, 37, 41). This represents a significant obstacle both to the comparison of existing results and for the planning of future studies. Recorded outcome measures in articles included in this study include gross examination of oral lesions by veterinarians, other physical exam findings (such as body weight or body condition score or prominence of mandibular lymph nodes), owner-reported clinical signs in the home setting, and/or owner-perceived quality of life for their cat. This review not only found little consistency between articles in the scoring system used to record outcome but also in the duration of follow-up. Follow-up periods ranged from 2 (43) to 354 weeks (41). The lack of consistency between studies in data reporting makes direct comparison of results problematic. In order to draw sound comparisons, ideally all studies would adopt the same validated outcome measure, with histology of oral lesions being the gold standard. Universal adoption of a standardized semi-quantitative scoring system validated with histopathology results would be ideal. A longer follow-up time is obviously superior to a shorter follow-up period, but at some point practicality limitations outweigh the desire to continue collecting data. No study has explored the time cutoff after which response to either medical or surgical therapy remains static; until such a study is performed, it is reasonable that prospective studies continue collecting data at least 6 months beyond plateaued response to treatment.

In analyzing articles whose focus is medical management, it is important to bear in mind if the subjects are refractory to historically performed dental extractions, if they have failed to respond to previous medical management attempts, or if their FCGS is naïve to any previous therapeutic intervention. For example, it is noteworthy that the clinical remission rate was strikingly similar for cats involved in the two cyclosporine studies (12, 40) despite the fact that one group of cats had been resistant to previous medical management (40), whereas the other group had been refractory to full-mouth extractions (12).

The grading systems utilized in this study provide a framework for analyzing experiment design and strength of the data reported. The grading process is not intended to criticize specific articles, and it is important to bear in mind that each article has its own strengths and weaknesses. The key components that comprise a strong FCGS study include the following: a large enough study population to be representative, a prospective experimental design, a quantitative or at least semi-quantitative scoring system for disease severity, and a long enough duration of follow-up to convince the reader that results will be sustained indefinitely. None of the articles included in this review meets all four of the above criteria.

There were four articles awarded a level I EDG, in part because they included either a control group (28, 36, 37) or a placebo group (12). While it is generally accepted that studies are strengthened by the inclusion of a control or placebo group, this practice is potentially problematic when applied to FCGS treatment clinical trials. One such problem is deciding upon the treatment administered to the control group. Each of the controlled studies included in this systematic review tests a different control treatment, namely a corticosteroid (28), a non-steroidal anti-inflammatory (37), or a diet (36). While selection of these control group treatments was intended to provide internal validation via direct comparison within studies, it renders comparison across studies difficult, if not meaningless. Control therapies have not been standardized; it could be informative to define a standardized control therapy against which to compare novel therapies. A second problem is the ethical dilemma that arises when cats enrolled in control or placebo groups experience prolonged suffering by not receiving superior, or any, treatment. One approach could, thus, be to eliminate placebo and control groups altogether. A benefit of this action is that more cats would be available to receive and test a treatment. Furthermore, spontaneous recovery or significant clinical improvement has never been reported in cats with refractory FCGS, which may devalue control or placebo groups (10). However, best practice in study design strives for the inclusion of control and placebo groups. The ethical predicament that accompanies use of control and placebo groups could be mitigated by allowing for analgesic administration to all cats throughout the study period. Additionally, if the treatment being tested is found to be promising, the cats in the control or placebo groups could be converted to cats receiving treatment, as occurred in the single placebo-controlled study discussed in this systematic review (i.e., a randomized controlled crossover study with one of the two treatments being a placebo) (12).

There were two articles awarded a level A EG for including histology of lesions post-treatment (10, 37). While FCGS is frequently diagnosed on the basis of clinical appearance and clinical signs, it is still beneficial to perform histopathology, ideally before and after treatment. Histopathology is useful not only to confirm the diagnosis (i.e., differentiate it from other diseases, such as squamous cell carcinoma) but also to help further classify and understand the lesion. The current lack of knowledge regarding the etiology of FCGS and the extent of mucosal response to therapy can only definitively be assessed via histological and immunohistochemical means, which validate and compliment the subjective assessment of clinical appearance and clinical signs. The more histopathological knowledge we have of this disease, the more potential there is to discovering the underlying etiology(s) and the most efficacious treatments.

There are limitations that may potentially affect the results and conclusions produced by this study. Inclusion and exclusion criteria were pre-defined to minimize bias; however, these criteria resulted in the exclusion of potentially influential articles. For example, an article exploring the efficacy of bovine lactoferrin was excluded because the study lacked a standardized definition of FCGS, with one experimental cat’s disease described as “severe gingivitis with hemorrhage” (48). Another article reporting the outcome of various therapies, including corticosteroids, antibiotics, megestrol acetate, levamisole, azathioprine, cyclophosphamide, and injectable gold (aurothioglucose) was excluded due to lack of defined treatment protocol, with inconsistent drug doses and no reporting of which cats received which treatment(s) (16). Another article examining various medical therapies, including diet adjustment, megestrol acetate, antibiotics, corticosteroids, levamisole, and mouthwash, was excluded because of inconsistent treatment protocols as well as no defined follow-up timeframe (8).

A second potential limitation is that only treatments for which there have been peer-reviewed studies published in the English language could be included in this study. Although chlorambucil, vincristine, 5-fluorouracil, sulodexide, tacrolimus topical, colchicine, and lysine have all been suggested as possible therapies (23), to the authors’ knowledge, there are no published reports assessing their efficacy within the veterinary peer-reviewed literature.

Because of the heterogeneity in outcome measurements, study design, population sizes, and follow-up times, it was impossible to perform a meta-analysis of articles’ data, and, thus, results of each study have been reported individually.

In considering treatment options for FCGS, veterinarians should strive to adopt an evidence-based approach to their therapeutic recommendations. While studies have shown the percentage of cats that respond to various medical and surgical interventions, clinicians must sift through these data and apply it to their individual patients. Many questions still remain – for which cats should medical management be attempted initially versus recommending surgical management at the time of initial diagnosis? At what point should medical management be discontinued in favor of switching to surgical management if a cat is failing to respond to therapy, versus administering an alternate medical management? Can we predict which cats will respond to various treatment types?

Veterinarians should strive to improve the quality of FCGS research and, thus, the evidence base available to inform colleagues’ therapeutic recommendations in order to optimize patient care. In addition, we should bear in mind both the strengths and weaknesses of the current literature when drawing conclusions. Large prospective studies are needed to compare existing treatments and demonstrate the promise of new treatments. With the emergence of novel and innovative therapeutics, the field of FCGS research would benefit from standardizing studies by adopting use of the same quantitative or semi-quantitative scoring system and extending follow-up duration to at least 6 months beyond plateaued response to treatment.