Accurate and rapid diagnosis of the influenza virus infection is very important for effective treatment with antiviral agents, and global prevention and control of influenza viruses (Choi et al. 2010; Jernigan et al. 2011; Cheng et al. 2012; Nitsch-Osuch et al. 2013; Jackowska et al. 2013). Presumptive diagnosis of influenza can be made based on the clinical signs taking into consideration the host species involved. However, during periods of epidemic prevalence clinical diagnosis becomes problematic (Cho et al. 2012). In humans, horses and pigs usually the respiratory system is involved. However, in poultry, besides the respiratory signs, there may be diarrhoea, paresis and decreased egg production. Clinical signs and pathological lesions are not reliable as many of the signs are also manifested in several other diseases. Definitive diagnosis of influenza requires isolation and identification of virus. The success of confirmatory laboratory diagnosis is to a great extent dependent on the choice and quality of specimens, and their transport and storage conditions prior to processing in the laboratory (Kumar and Henrickson 2012). During the presence of clinical signs, oropharyngeal swabs are superior for AI virus isolation; however, cloacal swabs were found to be more suitable for isolation of AI virus during recovery phase (de Jong et al. 2006). In humans, more amount of virus is present in the nasopharynx than throat due to the tropism of the H5N1 virus for the lower respiratory tract. The samples such as tracheal and cloacal swabs, faeces, tissue samples including trachea, lungs, serum, etc., should be submitted to the laboratory (Slemons and Brugh 1998; Swayne et al. 1998; Swayne and Halvorson 2003; Kataria et al. 2005; OIE 2005; Alexander 2008; To et al. 2012).

Influenza viruses can be isolated in embryonated chicken eggs via amniotic cavity or allantoic cavity route (Shahsavandi et al. 2012; Simon-Grife et al. 2012; Lombardo et al. 2012). However, often it becomes essential to first inoculate the samples in the amniotic cavity route followed by virus amplification in allantoic cavity (Monto et al. 1981). Human influenza viruses have been reported to grow better in cell cultures than in fertile eggs. Some of the recent human isolates have failed to grow in allantoic cavity. This may be due to the fact that allantoic cells contain only SA α-2, 3 sialyloligosaccharides, whereas both SA α-2, 3 sialyloligosaccharides and SA α-2, 6 sialyloligosaccharides are present on the amniotic cells. The replication of the viruses in eggs can be detected by performing haemagglutination (HA) test with the allantoic fluid. The haemagglutinating viruses are identified by haemagglutination inhibition (HI) assays using reference influenza virus antiserum. Primary cell culture of chicken embryo fibroblast (CEF) and Madin Darby Canine Kidney (MDCK), embryonic swine kidney (ESK), primary swine kidney and swine testicle or swine lung epithelial cell lines, chicken kidney (CK) cell lines and primary human respiratory epithelial cells can also be employed with different virus adaptation efficiency for isolation of the virus (Sugimura et al. 2000; Swayne and Halvorson 2003; Nakharuthai et al. 2008; Shahsavandi et al. 2012; Simon-Grife et al. 2012; Lombardo et al. 2012). The CEF or kidney cells are most commonly used for plaque assays and virus neutralisation tests. However, many of the isolates may require pre-treatment or addition of trypsin. Haemadsorption and haemagglutination tests are the tools for recognizing newly produced viral particles. Influenza B and rarely influenza A virus will produce a CPE in MDCK cells. Influenza viruses isolated from embryonated eggs/tissue culture can be diagnosed by applying serological or molecular methods (George 2012). Determination of the viral load is essential for conducting studies on viral replication, and evaluation of the efficacy of new vaccines and antivirals, both in cell culture models as well as in animal models (WHO 2005).

The viral antigen in the suspected samples can be demonstrated by agar gel immunodiffusion (AGID) (Pourbakhsh et al. 2000), immuno fluorescence test (IFT) (Skeeles et al. 1984; Yuen et al. 1998; Bano et al. 2003; Antarasena et al. 2006; Lee et al. 2012a; George 2012), immunoperoxidase test (IPT) (Shamseddini et al. 2002), immunohistochemical (Haines et al. 1993; Rimmelzwaan et al. 2001; Gu et al. 2007; Nakharuthai et al. 2008; Kalthoff et al. 2008; Chen et al. 2009; Chamnanpood et al. 2011; Bertran et al. 2012), immunochromatographic test (Welch and Ginocchio 2010; Sakurai et al. 2013) and enzyme-linked immunosorbent assays (ELISA) (Kodihalli et al. 1993; Hadjiev et al. 2000; Velumani et al. 2008; Luo et al. 2009; Shahsavandia et al. 2011; Ji et al. 2011; Chen et al. 2012; Kim et al. 2012; Read et al. 2012). Commercial diagnostic kits based on ELISA using monoclonal antibodies against nucleoprotein or matrix proteins are now available. Subtyping of influenza viruses is done by using monospecific antisera or monoclonal antibodies produced against the isolated antigens of each of the 18 HA and 11 NA types. The newly isolated virus can be subtyped with the haemagglutination inhibition and neuraminidase inhibition (HI/NI) tests using many polyclonal antisera encompassing all the subtypes (OIE 2005). Subtyping of AIVs can also be made by RT-PCR and PCR-ELISA (Dybkaer et al. 2004). The subtyping of the avian influenza viruses based on neuraminidase has also been reported (Huang et al. 2013).

The serological diagnosis is especially useful in documenting asymptomatic infections by demonstrating more than or equal to fourfold increase of antibody titre on paired sera (collected 2 weeks apart) to the specific antigen in case of an influenza infection (To et al. 2012). The non-specific inhibitors present in serum of various species of birds that may interfere with the specificity of the HI and other tests should be removed by treating the serum with receptor destroying enzymes (RDE) and potassium periodate (Eckroade et al. 1984). The haemagglutination inhibition (HI) (Simon-Grife et al. 2012; Skowronski et al. 2012), virus microneutralisation assay and double immunodiffusion (AGID) are employed to detect antibodies to nucleocapsid and matrix antigens for typing the influenza viruses. The sensitivity and precision of SRH to detect antibodies is more than CF or HI tests. The SRH also does not require sera to be pre-treated to remove non-specific inhibitors. Indirect and competitive ELISA assays have also been developed to detect antibodies to AIVs (Zhou et al. 1998; Shafer et al. 1998; Hadjiev et al. 2000; Skibbe et al. 2004; Simon-Grife et al. 2012). A latex agglutination test (LAT) has been reported (Xu et al. 2005) based on polystyrene beads coated with inactivated AIV H5N1 particles which can have potential application in the field for seromonitoring purposes.

A large number of molecular tests and techniques with different degree of success have been exploited for the diagnosis of influenza viruses. Various nucleic acid-based techniques (Pasick 2008) such as reverse transcription-polymerase chain reaction (RT-PCR) (Fouchier et al. 2000; Gupta et al. 2003; Starick and Werner 2003; Dybkaer et al. 2004; Heine et al. 2007), real time RT-PCR (Sugita and Matsumura 2003; Lee and Suarez 2004; Landolt et al. 2005; Hoffmann et al. 2007; Slomka et al. 2007; Lu et al. 2008; Kalthoff et al. 2008; Spackman and Suarez 2008; Wu et al. 2008; Choi et al. 2010; Shu et al. 2011; Jernigan et al. 2011; Nakauchi et al. 2011; Leijon et al. 2011; Chamnanpood et al. 2011; Takekawa et al. 2011; Romagosa et al. 2011; Fereidouni et al. 2012; Simon-Grife et al. 2012; Bertran et al. 2012; Read et al. 2012; Jackowska et al. 2013; Romanowska et al. 2013; Elizalde et al. 2014), duplex real time PCR (Kang et al. 2010; Lee et al. 2012b), multiplex RT-PCR (Pang et al. 2001; Poddar 2002; Malik et al. 2004; Chang et al. 2008; Lee et al. 2008; Hymas et al. 2010), multiplex real-time RT-PCR (Spackman et al. 2003; Payungporn et al. 2006; Ong et al. 2007; Chaharaeina et al. 2009; Liao et al. 2011), multiplex-microsphere-quantitative PCR (Liang et al. 2013), loop-mediated isothermal amplification (Poon et al. 2005; Imai et al. 2006, 2007; Jayawardena et al. 2007), reverse transcription loop-mediated isothermal amplification (Ge et al. 2013; Nie et al. 2013), in situ hybridisation (with labelled DNA probes) (Wu et al. 1999; Gupta et al. 2003; Chamnanpood et al. 2011), DNA micro-array (Wang et al. 2004; Lodes et al. 2006; Townsend et al. 2006; Sun et al. 2011), nucleic acid sequence based amplification (NASBA) (Collins et al. 2003; Lau et al. 2004) nucleic acid sequencing (Ghedin et al. 2005; Obenauer et al. 2006; Leijon et al. 2011; Tombari et al. 2013; He et al. 2012; Soltanialvara et al. 2012; Bertran et al. 2012; Jonges et al. 2013), and heteroduplex mobility assay (HMA) (Berinstein et al. 2002) have been deployed for rapid, sensitive and specific diagnosis of influenza viruses. The sensitivity of Real-Time Reverse Transcription-PCR was found to be adversely affected by newly occurring mutations in the Matrix genomic segment of A(H1N1)pdm09 and A (H3N2) influenza viruses (Yang et al. 2014). The sensitivity of PCR was reported to be 100-fold more than virus isolation procedures. The location of viral replication in tissues of infected birds was identified using a radio-labelled gene probe in in situ hybridisation (Van Campen et al. 1989). Monoclonal antibodies are also useful for localizing viral antigen in tissues by immunoperoxidase staining. Sequencing of HA gene for the specific characterisation of the virus is helpful. Detection of influenza A virus by a battery of broadly reactive anti-NS1 mAbs has recently been reported (Rahim et al. 2013). The RT-PCR has been found to be much more sensitive than many of the rapid tests, and virus isolation used for the initial detection of influenza viruses. It can also be applied for the retrospective diagnosis on preserved or fresh tissue specimens (Yuen et al. 1998; Ruest et al. 2003; Cattoli et al. 2004). Avian Influenza (AI) from reference viruses have been identified using an improved one-step reverse transcription real-time PCR (RRT-PCR) (Trani et al. 2006; Lu et al. 2007). Influenza A viruses have been genotyped using a web tool (Lu et al. 2007). The RT-PCR was found to be more sensitive than ELISA (Cattoli et al. 2004). A combination of PCR and ELISA known as PCR-ELISA has proven to be 100 times more sensitive than the detection by PCR alone (Dybkaer et al. 2004). A rapid and sensitive method for diagnosis of equine influenza (Ozaki et al. 2000) and H5N1 avian influenza (Deng et al. 2011) by antigen detection using immuno-PCR has been described. The DNA flow-thru chip, a three-dimensional biochip, was used for typing and subtyping of influenza viruses (Kessler et al. 2004). Rapid microchip-based electrophoretic immunoassays for the detection of swine influenza virus have been reported (Reichmuth et al. 2008; Kostina et al. 2011). Influenza viruses A and B have been detected simultaneously in a TaqMan-based real-time PCR using primers and probes constructed from highly conserved regions of the matrix protein and haemagglutinin gene of influenza virus A and B, respectively. The sensitivity of this technique was found to be much more than the virus isolations in cell culture (van Elden et al. 2001). Molecular assays also can be useful for subtyping, surveillance studies as well as selection of the candidate vaccine virus strains (Taubenberger and Layne 2001). Early detection of the disease can help in the timely implementation of prevention and control strategies against influenza (Jernigan et al. 2011; Nitsch-Osuch et al. 2013). However, the diagnostic techniques for influenza viruses should only be undertaken in referral laboratories having appropriate disease containment facilities by trained and skilled personnel.

Samples for diagnosis of Bird Flu in poultry

Diagnosis of avian flu requires referral laboratories equipped with trained scientific manpower and minimum level 3 biosecurity (BSL-3) measures. Precise and timely diagnosis needs appropriate samples to be sent to referral laboratories designated worldwide.

Portable real-time RT-PCR with lyophilised reagents may expedite surveillance results, and help better understand wild bird involvement in HPAI H5N1 transmission. Recently, a rapid H5N1 bird flu test kit (real-time RT-PCR assay based), detecting all known strains of H5N1 virus with a single test and with almost 100 % accuracy, has been reported to be developed for diagnosing bird flu within a few hours in humans.

Samples for Diagnosis of other flu viruses

It should be kept in mind that accurate diagnosis will require selecting the correct and high quality specimen, its transportation and storage conditions and finally processing. Normally, specimens should be collected within the first 3 days after onset of clinical symptoms of influenza (WHO 2005).

Live animal: Nasopharyngeal swabs and nasal or tracheal washings should be taken by endoscopy and immediately transferred to transport medium suitable for virus isolation. Serum samples should be sent for sero-surveillence.

Samples collected at postmortem: