Skip to main content

Canine Influenza Virus


Influenza virus was first identified as an infectious disease of dogs in 2004. The first isolation of the virus was done at the Animal Health Diagnostic Center at Cornell in conjunction with a study being conducted by the University of Florida on respiratory disease in racing greyhounds. The virus was sequenced at the CDC and determined to be related to the H3N8 equine virus that was circulating in horses in the US. Specific genetic differences between the equine viruses and the viruses from canines defined the virus as a unique canine influenza virus (CIV). Within a year of the discovery, CIV was found in pet dogs in Florida and in the New York City area. Since that time the virus has been found in different areas of the US but in most cases, the infection was contained and did not become enzootic. The exceptions to this pattern were Florida, New York City area, and Colorado.

Currently, there is an increase in awareness of CIV due to the release of a CIV vaccine, the detection of the virus in several new areas and more testing in support of the vaccine release. The Northeast region (New York City to Philadelphia) continues to see CIV activity with a new introduction into northern Virginia. Colorado also continues to have frequent CIV activity with new reports from Las Vegas and California (California has had positive dogs in the past). Virus has not been detected in Florida over the past year.

CIV continues to move slowly through the canine population. Risk factors for the infection are having dogs in closely confined conditions such as in boarding kennels, day care setting and animal rescue shelters. Animals being relocated from the rescue shelters seem to be a main source of the movement of the virus to new locations. The clinical signs associated with the infection are indistinguishable from the traditionally defined “kennel cough” now more appropriately referred to as acute respiratory disease in dogs. The morbidity rate in normal populations can be very high (60-80%) while the mortality rate due exclusively to CIV is very low. The significance of a CIV infection is that it compromises the normal defense mechanisms of the canine respiratory tract so that secondary bacterial infections are common sequelae. Dogs may cough for several weeks after infection, but they are not contagious at this time. Dogs are generally free of CIV by 7 day post onset of clinical signs.

The vaccine should be considered in those situations where one would have used the standard kennel cough vaccines and where there have been documented cases of canine influenza. Simply having a seropositive dog in a given area is not definitive proof of the existence of CIV unless one has proof that the dog originated in that area and never traveled to an area enzootic for CIV. Those involved in the breed rescue organizations should not be moving dogs from affected areas without vaccinating the animals that will be coming in contact with the rescued animal.

The diagnosis of canine influenza necessitates the identification of CIV in the acutely infected animal or demonstrating CIV antibodies in the later stages of the clinical event. Detection of CIV can be done by doing a PCR test for influenza virus or by isolating CIV from a clinical sample. The AHDC offers a PCR test that can detect any influenza virus in a specimen, not just CIV. We offer this test as other influenza viruses are capable of infecting dogs and it would be unfortunate to miss a new strain of influenza virus in dogs if one were only doing an H3 CIV PCR. Samples for PCR testing should be taken within 4 days of onset of clinical signs and the sample of choice is a nasal swab. Swabs should be placed into a tube with several drops of saline. DO NOT USE A BACTERIAL TRANSPORT MEDIUM FOR PCR TESTING OR FOR VIRUS ISOLATION. Swabs should be refrigerated and sent overnight on freezer packs. For antibody testing, an acute and convalescent sample (10-14days apart) are optimal, but in areas of low CIV prevalence, a single sample taken >7 days after acute onset may be sufficient.

Questions regarding CIV can be directed to Dr. Diego Diel, Director of the Virology Lab.

Appropriate Samples for Detection

As with most viral diseases, there are several ways to determine the role of a virus in a clinical event. While there is a tendency to want a single sample type that will work in all cases, the reality is that this is not possible. The list of possible tests is:

  1. At present the most reliable way to diagnose canine influenza virus infections is by serological tests. The direct link between canine influenza virus and a clinical event is through the collection of acute and convalescent serum samples. First sample is collected at the first presentation of the patient and then 2-3 weeks later. Serum can be separated from the clot and held in the refrigerator until collection of the second sample. Samples can be shipped overnight without cold packs or 2-day delivery with cold packs. For animals that have recovered from a case of “kennel cough”, a single serum sample can determine whether the animal has been infected with canine influenza virus at some point in the past. As this is a relatively new pathogen of dogs, we do not expect to find a high seroprevalence in unaffected dogs.

    Go to Canine Influenza Virus HI for test fee and submission-related information.
  2. At present, the most reliable way to directly detect the virus is through a PCR test on either swabs (nasal swabs preferred) or respiratory tissue. A generic PCR test that detects the highly conserved matrix gene is the test of choice. While we are currently looking for the H3N8 virus, a test unique to this virus would miss a potential influenza infection caused by another flu strain such as H5N1. Broad screening tests are to be used whenever possible in surveillance programs.

    Go to Influenza Virus Matrix PCR for test fee and submission-related information.
  3. Isolation of canine influenza virus is a relatively unreliable way to confirm the infection. The reason for this is not clear, but the timing of the sample collection may be the main culprit. Experimental infections have shown that nasal swabs are nearly twice as successful in isolating the virus as nasopharyngeal swabs taken at the same time on the same animal. We continue to encourage the submission of samples for virus isolation because it is essential to track the genetic changes that are occurring in the virus as it makes its way through the dog population. Samples that are submitted for PCR detection of the virus that are test positive, will be automatically set for virus isolation at no addition cost.

Note: All dogs that cough are not infected with canine influenza virus. The standard respiratory pathogens of dogs have not gone away. One should always consider a complete diagnostic work-up that would include cultures for bacteria and mycoplasma. Check our fee manual for sample submission information and fees.

Submission forms for the AHDC can be downloaded from the Sample Submission Requirements webpages.