Frédéric P. Gaschen, Dr.med.vet., Dr.habil., DACVIM, DECVIM-CA
Adverse reactions to food are common in dogs and in cats. They can elicit cutaneous and/or gastrointestinal manifestations. As shown in Figure 1, adverse food reactions encompass disorders with an immunological basis (food allergy, also called dietary hypersensitivity or dietary sensitivity), non-immunologic reactions (food intolerance), and toxic reactions (intoxications). While intoxications are encountered frequently at least in dogs (e.g., garbage can gut gastroenteritis), they will not be discussed in further details. Food intolerance occurs probably relatively often in small animals, and is associated with a variety of gastrointestinal signs, most importantly diarrhea and/or vomiting. Food allergy is a common cause of cutaneous signs such as pruritus, and may occasionally be associated with gastrointestinal signs. However, in some instances, chronic vomiting and diarrhea may be the only clinical expression of food allergy. Generally, the importance of reactions to food in the etiology of canine chronic enteropathies should not be underestimated.
Proposed nomenclature for adverse reactions to food in dogs and cats (adopted from Sampson1).
Epidemiology and Clinical Presentation
In humans, on the basis of recent surveys, it is believed that 3.5 to 4% of the US population have IgE-mediated food allergy.1
There are no comparable data for canine or feline food allergies. However, in a recent study performed in the United Kingdom, 7.6% of dogs presented to a veterinary clinic specialized in dermatology were diagnosed with food allergy. These dogs represented approximately 1/3 of all those diagnosed with any type of allergy affecting the skin.2 In the numerous dermatological studies performed on this topic, the most frequent clinical signs were: generalized pruritus (in some cases with a preference for specific regions), recurring otitis externa, and secondary pyoderma. Even though the prevalence of signs affecting the digestive tract was not clearly specified, an estimated 10-15% of cases presented to dermatologists also showed vomiting and/or diarrhea. Food allergy also plays a non-negligible role as a differential diagnosis in dogs with chronic enteropathies. In a prospective study of 70 dogs referred for chronic diarrhea of more than 6 weeks duration, clinical signs resolved after 7-10 days in 39 (56%) of patients after they were fed with a novel protein diet.3 It is likely that many of these patients with diet-responsive disease suffered from food intolerance or mild forms of idiopathic inflammatory diseases. Thirty-one dogs could be returned to a commercial (non-veterinary) diet after 14 weeks on the novel protein diet. In 8 dogs, clinical signs relapsed when the elimination diet was discontinued, and dietary challenge with proteins from single sources lead to a diagnosis of food allergy in 2 dogs while the other 6 were diagnosed with food intolerance.3
Dogs presented to dermatologists with cutaneous expression of food allergy are often young (one third to half the cases were 1 year or younger in several studies). Dogs with chronic enteropathies that were diet-responsive were younger than dogs with inflammatory bowel disease that required immune-suppressive steroid treatment (mean 3.5 vs. 7.5 years old).3 There is little evidence for breed predisposition or for heritability of canine food allergy,2,4 with the exception of Soft Coated Wheaten Terriers affected with protein-losing enteropathy.5
Food allergy is also a common rule out for cats presented with non-seasonal, persistent pruritus, especially if localized to the head and neck region. Miliary dermatitis, facial and neck dermatitis and otitis externa are frequently associated clinical signs. Simultaneous occurrence of gastrointestinal signs such as vomiting and diarrhea are occasionally reported. In a study of 55 cats from New Zealand with chronic idiopathic gastrointestinal problems, 27 cats responded to a novel protein diet. In 11 cats, the clinical signs did not recur after challenge with the original diet. In the remaining 16 cats, dietary challenge resulted in a relapse, and they were diagnosed with food intolerance or food allergy.6
The pathogenesis of food allergy is complex. It includes the following mechanisms: leaky intestinal mucosal barrier, abnormal antigen presentation to the immune system, dysregulation of the immune system, and leads to the loss of oral tolerance. The major food allergens are water soluble glycoproteins that are 10-70 kD in size and fairly stable to heat, acid and proteases.1 The most common allergens identified in spontaneously food allergic dogs may differ according to geographical location. They include cow's milk, eggs, meat proteins from beef, and chicken, and plant proteins from corn, wheat and soybeans.4
In human beings, most food allergies are due to type I hypersensitivity reactions involving a Th2 immune response after exposure to the allergen, and the production of IgE.1 It appears that the situation is different in dogs, as serum allergen-specific IgE concentrations poorly correlate with results of elimination trial and dietary challenge tests (see below). Type I hypersensitivity caused by s.c. sensitization of canine "high IgE responders" with specific allergens may elicit food allergy in the dog. This laboratory model may be useful for studying food allergy in humans,4 however it does not appear to truthfully mimic the spontaneous canine disease. It is suspected that type IV hypersensitivity or possibly local type I hypersensitivity may be at the origin of canine dietary hypersensitivity.
Current Diagnostic Gold Standard
In human medicine, the medical history is a mainstay in the diagnostic procedure as the immediate nature of the allergic reactions often allows to determine which food to suspect. However the double-blind, placebo-controlled food challenge remains the gold standard for the diagnosis of food allergies.1
In dogs and cats, a diagnosis of food allergy is traditionally confirmed by a change to a new diet made from a protein source to which the dog has not been previously exposed (elimination diet). Many excellent dry or wet elimination diets are currently available from pet food manufacturers. However, some specialists still prefer to recommend a homemade diet tailored to the each patient's previous exposure to dietary antigens as the best choice for dietary elimination. It is essential to provide a balanced diet based on a novel protein source (or on hydrolyzed proteins) and to make sure that the dog is fed this diet exclusively. Any deviation from this protocol must be avoided at all costs (e.g., feeding treats, table scraps, drugs coated with appetizing substances). Lack of owner compliance is probably an important reason for misdiagnosis, especially in dogs and cats with skin problems. The gastrointestinal signs generally resolve rapidly after the dietary modification (often 1 to 2 weeks, even though some cases may require more time to show clinical improvement). However, cutaneous manifestations of food allergy typically resolve much slower, and dermatologists recommend using the elimination diet for a minimum of 6 to 8 weeks before ruling out food allergy.
The disappearance of (cutaneous or digestive) clinical signs when a dog is fed an elimination diet is not sufficient to confirm a diagnosis of food allergy. For such patients the designations "diet-responsive disease" or "adverse reaction to food" are preferred. After a successful elimination trial, dogs and cats relapsing when fed their original diet probably suffer either from food intolerance or food allergy. The diagnostic confirmation of food allergy is based solely on dietary challenge with suspected allergenic proteins (e.g., beef, lamb, pork, cow's milk, corn, etc.) followed by recurrence of the clinical signs. This usually occurs after a few days for digestive signs, and up to 2 weeks for cutaneous signs. For obvious reasons, dietary challenges are frequently not favorably looked upon by dog and cat owners, and are often not performed in clinical practice. This is why, after a successful elimination trial, many dogs are left on the "hypoallergenic diet" without further diagnostics and often remain in remission for prolonged periods of time, or until they are (often accidentally) exposed to regular dog food again.
The gold standard for diagnosis is less than optimal as owners are requested to elicit a relapse of clinical signs in order to reach the final diagnosis. This is why several other diagnostic options have been explored with more or less success.
Intradermal Allergy Testing
It has been shown in several studies that this method was not helpful in the clinical diagnosis of food allergy in the dog.7
Detection of Serum Immunoglobulins
The evaluation of serum IgE and IgG concentrations was the object of great expectations. In humans with type I hypersensitivity reactions, antigen-specific immunoglobulins may be detected in the blood and allow the identification of the allergenic proteins. However, in spontaneous canine food allergy the determination of serum antigen-specific IgE and IgG concentrations has lead to disappointing results, even though the tests are widely available to the veterinary community (numerous false positives and negatives).4
Detection of Fecal IgE
The presence of antigen-specific IgE can also be detected in the feces. In food allergic Soft Coated Wheaton Terriers peak levels of antigen-specific fecal IgE were detected when the dogs were fed the proteins they were allergic to.5 However, this test probably suffers from the same limitations as the detection of serum antigen-specific IgE as the hypersensitivity in dogs and cats may not be a type I reaction.
Provocation Tests on the Gastric / Colonic Mucosa
Deposition of allergens on the gastric mucosa or under the colonic mucosa followed by observation for inflammatory reaction (redness, swelling) are 2 methods that have been studied in dogs (gastric mucosa sensitivity testing or GFST, resp. colonic mucosa allergenic provocation or COLAP). For these tests, solutions containing potential food allergens are deposited on the gastric mucosa (GFST) or injected under the proximal colonic mucosa (COLAP), and the mucosa is observed for a wheal and flare response during a few minutes. Concordance of the 2 tests with positive dietary challenge results was moderate (GFST 50%, COLAP 73%) in a colony of food allergic Soft Coated Wheaton Terriers.5,8 Unfortunately, both tests can only be performed in anesthetized patients and require availability of endoscopic equipment. Additionally, dogs undergoing COLAP must have their colon thorough cleansed with oral laxative solutions and enemas. Finally, the antigen injection and evaluation of the inflammatory response may be challenging. Because the immediate mucosal reactions reflect type I hypersensitivity, the limitations mentioned above (serum IgE and IgG, intradermal allergy testing) may also apply to GFST and COLAP.
Doppler Ultrasonography of Large Mesenteric Vessels
Vasodilation and vasoconstriction of capillary beds in the intestinal mucosa directly influence quantity and quality of the blood flow through large mesenteric arteries (celiac and cranial mesenteric arteries). This technique aims at evaluating the dynamics of blood flow in the postprandial phase, and made it possible to confirm the presence and cause of food allergies with excellent accuracy in comparison to the traditional dietary challenge test in a colony of food allergic Soft Coated Wheaten Terriers. A prolonged vasodilation of the intestinal mucosal capillary beds could be indirectly documented in food allergic dogs after they had been challenged with food they were allergic to.9 This procedure does not require any special patient preparation, is generally well tolerated, and is very promising. However, the presence of an experienced ultrasonographer with good ultrasound equipment is essential for accurate measurements and interpretation of the results.
Other Laboratory Analyses
An in vitro lymphocyte stimulation test was used in Japan to diagnose food allergy in a group of dogs with chronic enteropathies. During the test procedure, circulating lymphocytes (and memory T cells) are exposed to potential allergens in vitro. Using flow cytometry, the cell colonies are observed for any proliferation of CD4 lymphocytes. CD4 proliferation suggests that memory T cells recognize the specific antigen and induce an inflammatory response following exposure. A positive response indicates that the dog is allergic to that particular antigen. In a Japanese study, this test was performed in parallel with a dietary challenge test and concordance of results was very promising. However, food allergies could only be diagnosed in symptomatic animals.10 Several groups are currently evaluating modified versions of this test to investigate if the procedure could be useful as a reliable diagnostic test of food allergy. Potential pitfalls include discrepancies in the secondary and tertiary conformation of antigens used in vitro and those present in heavily processed commercial diets.
Antineutrophil cytoplasmic antibody with perinuclear staining (pANCA) have been used to differentiate between the 2 forms of IBD in human medicine (Crohn's disease vs. ulcerative colitis). Circulating antibodies to pANCA are present in approximately 50-80% of patients with ulcerative colitis. pANCA are also present in serum from dogs with chronic intestinal inflammation. The results of pANCA serology are highly specific for dogs with chronic enteropathies.11 Dogs with diet-responsive disease appear to be more frequently pANCA-positive (62%) than those with IBD (23%). Although the study did not differentiate dogs with food intolerance from those with food allergy, pANCA have the potential to be useful in the diagnosis of canine food allergy.12
Outlook for Simplified Diagnosis of Food Allergy in Dogs and Cats
At this time, Doppler evaluation of mesenteric blood flow is the most promising alternative method for diagnosing food allergy associated with gastrointestinal signs in dogs. It has not been evaluated in dogs with cutaneous expression of food allergy or in cats yet. Even though it does not obviate the need for challenging, the changes in blood flow through the celiac and cranial mesenteric arteries were detectable before the onset of diarrhea in food allergic Wheaton Terriers.9 The in vitro lymphocyte stimulation test would be useful if it could be performed in dogs without requiring in vivo challenge of the patient with the offending dietary antigens. This may be achieved if the technique can be refined. However, immediate processing of the samples is required. This implies a close proximity between the hospital and the immunology lab, and is reserved to academic and large referral institutions. Ultimately, development of a canine, resp. feline model of canine, resp. feline food allergy appears to be an essential step towards refining and validating better alternative diagnostic methods for food allergy in these 2 species.
Summary of Therapeutic Approach
Treatment of food allergy is relatively simple once the offending allergens have been identified. It consists of exclusive feeding with a novel protein source and avoiding all known allergens for the particular dog. Use of diets based on hydrolyzed peptides is another option, however, allergic reactions to soy peptides have been reported and may be a limiting factor for the use of hydrolysates.13 A theory has been proposed that, in some dogs, the protein source used in the initial elimination trial may end up eliciting a new allergic reaction and causing recurrence of clinical signs ("sacrificed protein"). In these dogs, further treatment requires the use of a second, different novel protein in order to keep the animal free of clinical signs. Finally, optimal treatment duration has not been clearly determined, but it seems logical to keep feeding the "hypoallergenic" diet as long as possible.
1. Sampson. J Allergy Clin Immunol 2004; 113: 805.
2. Chesney. J Small Anim Pract 2002; 43: 203.
3. Allenspach et al. J.Vet Intern.Med. 2007; 21: 700.
4. Day. Proc.Nutr.Soc. 2005; 64: 458.
5. Vaden, et al. J.Vet.Intern.Med. 2000; 14: 60.
6. Guilford, et al. J.Vet.Intern.Med. 2001; 15: 7.
7. Kunkle and Horner. J Am.Vet Med Assoc. 3-1-1992; 200: 677.
8. Allenspach, et al. J.Small Anim Pract. 2006; 47: 21.
9. Kircher, et al. J.Vet.Intern.Med. 2004; 18: 605.
10. Ishida, et al. J.Vet.Intern.Med. 2004; 18: 25.
11. Allenspach, et al. Am.J.Vet Res. 2004; 65: 1279.
12. Luckschander, et al. J.Vet.Intern.Med. 2006; 20: 221.
13. Puigdemont, et al. Am.J Vet Res. 2006; 67: 484.