Detection of Newcastle disease virus and assessment of associated relative risk in backyard and commercial poultry in Kerala, India

Abstract Background Newcastle disease (ND) is an economically important viral disease affecting the poultry industry. In Kerala, a state in South India, incidences of ND in commercial and backyard poultry have been reported. But a systematic statewide study on the prevalence of the disease has not been carried out. Objectives A cross‐sectional survey was performed to detect the presence of Newcastle disease virus (NDV) in suspect cases and among apparently healthy commercial flocks and backyard poultry, in the state and to identify risk factors for NDV infection. Methods Real‐time reverse transcription‐PCR (RT‐PCR) was used to detect the M gene of NDV in choanal swabs and tissue samples collected from live and dead birds, respectively and the results were statistically analysed. Results The predominant clinical signs of the examined birds included mild respiratory signs, huddling together and greenish diarrhoea. Nervous signs in the form of torticollis were noticed in birds in some of the affected flocks. On necropsy, many birds had haemorrhages in the proventriculus and caecal tonsils which were suggestive of ND. Of the 2079 samples tested, 167 (8.0%) were positive for the NDV M‐gene by RT‐PCR. Among 893 samples collected from diseased flocks, 129 (14.5%), were positive for M gene with pairwise relative risk (RR) of 15.6 as compared to apparently healthy flocks where 6 out of 650 (0.9%) samples were positive. All positive samples were from poultry; none of the ducks, pigeons, turkey and wild birds were positive. Commercial broilers were at higher risk of infection than commercial layers (RR: 4.5) and backyard poultry (RR: 4.9). Similarly, birds reared under intensive housing conditions were at a higher risk of being infected as compared to those reared under semi‐intensive (RR: 6.7) or backyard housing (RR: 2.1). Multivariable analysis indicated that significantly higher risk of infection exists during migratory season and during ND outbreaks occurring nearby. Further, lower risk was observed with flock vaccination and backyard or semi‐intensive housing when compared to intensive housing. When the M gene positive samples were tested by RT‐PCR to determine whether the detected NDV were mesogenic/velogenic, 7 (4.2%) were positive. Conclusions In Kerala, NDV is endemic in poultry with birds reared commercially under intensive rearing systems being affected the most. The outcome of this study also provides a link between epidemiologic knowledge and the development of successful disease control measures. Statistical analysis suggests that wild bird migration season and presence of migratory birds influences the prevalence of the virus in the State. Further studies are needed to genotype and sub‐genotype the detected viruses and to generate baseline data on the prevalence of NDV strains, design better detection strategies, and determine patterns of NDV transmission across domestic poultry and wild bird populations in Kerala.

had haemorrhages in the proventriculus and caecal tonsils which were suggestive of ND. Of the 2079 samples tested, 167 (8.0%) were positive for the NDV M-gene by RT-PCR. Among 893 samples collected from diseased flocks, 129 (14.5%), were positive for M gene with pairwise relative risk (RR) of 15.6 as compared to apparently healthy flocks where 6 out of 650 (0.9%) samples were positive. All positive samples were from poultry; none of the ducks, pigeons, turkey and wild birds were positive. Commercial broilers were at higher risk of infection than commercial layers (RR: 4.5) and backyard poultry (RR: 4.9). Similarly, birds reared under intensive housing conditions were at a higher risk of being infected as compared to those reared under semi-intensive (RR: 6.7) or backyard housing (RR: 2.1). Multivariable analysis indicated that significantly higher risk of infection exists during migratory season and during ND outbreaks occurring nearby. Further, lower risk was observed with flock vaccination and backyard or semi-intensive housing when compared to intensive housing. When the M gene positive samples were tested by RT-PCR to determine whether the detected NDV were mesogenic/velogenic, 7 (4.2%) were positive.

Conclusions:
In Kerala, NDV is endemic in poultry with birds reared commercially under intensive rearing systems being affected the most. The outcome of this study also provides a link between epidemiologic knowledge and the development of successful disease control measures. Statistical analysis suggests that wild bird migration season and presence of migratory birds influences the prevalence of the virus in the State. Further studies are needed to genotype and sub-genotype the detected viruses and to generate baseline data on the prevalence of NDV strains, design better detection strategies, and determine patterns of NDV transmission across domestic poultry and wild bird populations in Kerala.

K E Y W O R D S
F gene, M gene, Newcastle disease, poultry, prevalence, relative risk, RT-PCR

INTRODUCTION
Poultry production is an important contributor to agricultural productivity in India. Domestic poultry is reared for meat and eggs in large commercial farms as well as in small-scale backyard units. As in any livestock industry, infectious diseases pose a major threat to poultry production. One such disease is Newcastle disease (ND), which is caused by virulent strains of Avian orthoavulavirus-1 (commonly known as Avian Paramyxovirus-1 (APMV-1) or Newcastle disease virus (NDV)), a negative-sense single-stranded RNA virus of the genus Orthoavulavirus, family Paramyxoviridae (Dimitrov et al., 2019;ICTV, 2019). Based on pathogenicity to chicken embryos, strains of NDV are designated as lentogenic, mesogenic or velogenic, ordered by increasing virulence. In the United States, virulent NDV is regarded as a pathogen of national concern and a significant threat to animal agriculture (Brown & Bevins, 2017). On the basis of disease produced in chickens under laboratory condition, NDV isolates have been placed in five pathotypes; viscerotropic-velogenic, neurotropic-velogenic, mesogenic, lentogenic and asymptomatic enteric (Alexander, 2011).
NDV has been classified into Class I and Class II on the basis of its genome sequence (Czeglédi et al., 2006 (Aldous et al., 2010).
Cormorants, pigeons and psittacine birds are commonly infected with NDV and are known to serve as a source of virulent virus to domestic poultry. Strains of low virulence (lentogenic strains) are also prevalent in poultry and wild birds including waterfowl (Miller, 2014;Brown & Bevins, 2017).
In India, the poultry sector is valued at about 800 billion rupees (2015-2016) (approximately 11 billion USD) and the highly organised commercial sector contributes to a majority (about 80%) of the total market share (DAHDF, 2017). ND was first reported in India in 1928 and is currently present in most of the country. The first outbreak of ND in India occurred in 1927 in Ranikhet, Uttarakhand in North India and there have been several incidences of the disease in the country since then. Several reports on the isolation and characterisation of NDV from different parts of the country have also been published (Ananth et al., 2008;Kumanan et al., 1992;Morla et al., 2016;Roy et al. 2000;Tirumurugaan et al., 2011). Control efforts include regular vaccination of commercial poultry; however, backyard poultry are usually not vaccinated, which often leads to free circulation of the virus in these birds. The estimated total economic losses from January 2013 to July 2014 among commercial layers in Gujarat state of India was 3.7 million rupees (approximately 50,500 USD) (Khorajiya et al., 2018).
Being in the tropics, Kerala has a humid wet climate with seasonal monsoons. The state has both commercial and backyard poultry units.
Common diseases present in Kerala poultry are ND, infectious bronchitis (Fathima et al., 2018;Ravishankar et al., 2015), infectious bursal disease (Nandhakumar et al., 2020), fowl cholera and salmonellosis (Ravishankar et al., 2008). Of these, ND is economically important because outbreaks of this disease occur more frequently. In Kerala, the virus has been isolated from Indian mynah (Acridotheres tristis) (Sulochana et al., 1982), pigeons (Sulochana & Mathew, 1991) and Japanese quails (Mini et al., 2001) and chicken (Arun, 2004). The virus has also been detected in pigeons along with other viral and bacterial pathogens (Dhivahar et al., 2018;Reji et al., 2017). However, no systematic statewide study has been carried out to map the prevalence of NDV. This cross-sectional survey was performed to better estimate NDV prevalence in Kerala and identify risk factors for NDV infection among domestic poultry under varying production systems.

Source of samples
The state of Kerala was arbitrarily divided into three zones based on area (km 2 ) and geographic location (  Wild birds 0 1 0.00 (0-97.5) † Estimate and 95% confidence interval calculated by prop.test for denominator >30, by exact binom.test for denominator < 30. ‡ The vaccination status of 2 birds could not be ascertained and no NDV was detected. § The sex of broilers (n = 363) was not determined with detection of NDV in 84 samples.
pigeons and turkeys. The samples were collected in brain-heart infusion broth followed by storage at -80 • C until testing.

RNA extraction and real-time reverse transcription-PCR (RT-PCR)
Total RNA was extracted from specimens using MagMAX™ 96 AI/ND Viral RNA isolation kit (Thermo Fischer Scientific, Vilnius, Lithuania). Real-time reverse transcription-PCR (RT-PCR) was performed to detect the M gene of NDV using OneStep RT-PCR kit (Qiagen, Hilden, Germany) and previously published primers (Wise et al., 2004). Based on standard controls, a C t value of 35 and below was regarded as positive. In order to assess whether the M gene-positive samples contained lentogenic or mesogenic/velogenic NDV, all these samples were tested by RT-PCR targeting the F gene of mesogenic/velogenic NDV APMV-1 using previously published primers (Farkas et al., 2009). Any sample with a C t value of 40 or below was regarded as positive.

2.2.1
Statistical analysis Collinearity in the model was ascertained using the vif() function. The log odds ratio output from the glm() was converted into a relative risk score and accompanying 95% confidence interval using the function odds_to_rr() from the sjstats package. This function uses the equation: Zhang and Yu (1998), Wang (2013) and Grant (2014).

Clinical signs
During the course of the study, samples were collected from cases of respiratory illness suggestive of ND. The clinical signs exhibited by the affected birds ranged from mild respiratory distress to difficulty in respiration ( Figure 2a) and huddling with other birds (Figure 2b).

Assessment of associated risk factors for NDV infection
In India, NDV control is primarily with vaccination and varies with bird age and rearing conditions. Hence, the relative risk (RR) was assessed for specific risk factors including: vaccination status, flock health status, bird health status, housing, sex, zone and bird type (Table 2; Fig-ure 3). In pairwise comparisons, no significant difference was observed between the infection risk associated with vaccinated and unvaccinated birds (RR: 1.1, 95% CI: 0.8-1.5) and 0.9 (95% CI: 0.7-1.3).
Notably, there was a higher risk for diseased flocks when compared to healthy flocks (RR: 15.6, 95% CI: 7.0-35.2). Further, the RR of flocks exhibiting mild respiratory signs was 6.5 (95% CI: 2.7-15.4) when compared to healthy flocks, which confirms respiratory symptoms to be a significant risk factor for NDV infection. Birds in diseased flocks were 2.4 times more likely to be infected with NDV as compared to those with mild respiratory signs (RR: 2.4, 95% CI: 1.7-3.5).
At an individual bird level, healthy birds were four times as likely to be M gene-positive than recovered birds (RR: 4.0, 95% CI: 1.0-16.1).
The risk of a positive test among sick birds was also twice the risk among dead birds (RR: 2.0, 95% CI: 1.0-4.1). Additionally, sick birds had a risk five times that of recovered birds (RR: 5.5, 95% CI: 1.4-22.3). Birds in intensive housing were twice as likely to be M genepositive compared to those reared under backyard conditions (RR: 2.1, 95% CI: 1.0-4.5) and nearly seven times as likely when compared to those in semi-intensive conditions (RR: 6.7, 95% CI: 3.8-11.7). There was no statistically significant risk associated with the sex of the birds although a vast majority of samples was taken from female birds; the RR of a positive test among female birds was 2.8 (95% CI: 0.9-8.7).
The results of the multivariable analysis are shown in Table 3 and Lastly, broilers were at a higher risk (OR 2.9; RR 2.7; RR 95% CI 1.8-3.9) of being infected as compared to layers, but backyard birds were not at significantly higher risk than layers (OR: 1.1; RR: 1.1; RR 95% CI: 0.3-2.8).

DISCUSSION
The present study was carried out to assess the prevalence of NDV analyses, higher risk of NDV detection was associated with the intensive system of housing.
In Kerala, the primary vaccination strategy for ND in commercial layers is the use of a lentogenic strain of NDV at the first week of age, followed by a booster with a mesogenic strain (e.g., R2B) at two months of age. However, it is worth mentioning that the currently used vaccine strains were isolated three to seven decades earlier and are regarded as genetically distinct from the currently circulating virulent NDV strains. The high genetic distance between the vaccine and the current NDV strains prevents effective reduction of shedding of virulent virus from vaccinated birds (Dimitrov et al., 2017). However, results of the multivariable analysis indicate that vaccination did have a protective effect.
The M gene primers pick up vaccine strains that are largely (but not exclusively) lentogenic in nature. Hence, F-gene primers and probes were used to determine the prevalence of mesogenic/velogenic (M/V) strains. The low prevalence of M/V pathotypes suggests that a vast majority of circulating strains are probably lentogenic in nature. Since lentogenic strains were equally prevalent in non-vaccinated flocks, it can be presumed that at least some of them were wild-type strains.
Another reason for reduced detection of mesogenic/velogenic NDV may be due to the lack of sensitivity of the primers and probes for virulent NDV prevalent in India. However, to preclude any issues with sensitivity and specificity of the PCR-based pathotyping approach, future molecular epidemiologic and phylogenomic analyses are planned.
The occurrence of NDV in vaccinated birds has been previously reported in several other countries in Asia, Africa and Central America where NDV is endemic (Dey et al., 2014). Although vaccination protects against clinical disease, it fails to protect against viral shedding when birds are challenged with a genotype different than that contained in the vaccine (Kapczynski & King, 2005;Miller et al., 2007).
For example, a new subtype of virulent genotype XIII has been shown to cause severe outbreaks in vaccinated commercial broiler farms in Tamil Nadu, Southern India (Gowthaman, Ganeshan et al., 2019). Additionally, mixed infection with both vaccine and field strains in a flock is F I G U R E 4 Relative risk of Newcastle disease virus infection (M-gene positive) by risk factor from multivariable logistic regression analysis also possible. The failure of vaccination to protect a flock can be due not only to antigenic variation or genotype mismatch with circulating strains but also to poor flock immunity because of inadequate vaccination that can be attributed to poor vaccine storage conditions or faulty vaccine administration (Dortmans et al., 2012;Liu et al., 2018).
In the present study, NDV was found to be more prevalent in broilers (23.1%, 95% CI: 19.0-27.9) than in layers (5.2%, 95% CI: 4.1-6.4) or backyard birds (4.7%, 95% CI: 1.7-11.2). The broilers also had a higher RR (univariable analysis, Figure 3). Higher prevalence of NDV in broilers as compared to layers has been reported previously (Rahman et al., 2017). The greater chance of birds reared under intensive systems contracting viral infections has been reported previously. However, this system allowed for better control of disease. On the other hand, the control of diseases in small, rural, extensive poultry flocks in developing countries was difficult and that the incidence of diseases, such as ND, in these birds may represent a threat to intensively managed systems (Biggs, 1982).
Most of the clinical signs observed during sample collection were typical of ND (e.g., respiratory distress, ruffled feathers, and high morbidity). During necropsy, haemorrhages in the proventriculus and caecal tonsils were often observed, which are suggestive of ND. Similar observations have been reported in NDV outbreaks in Gujarat, India (Khorajiya et al., 2015). In the present study, samples collected from birds with clinical signs of respiratory illness and high morbidity were positive for NDV. However, there are other diseases, which can present a clinical picture similar to that of ND, for example, infectious bronchitis, infectious laryngotracheitis and avian influenza. In a recent study, co-infection of flocks with low-pathogenic avian influenza virus (AIV) and virulent NDV was reported and it was concluded that AIV may help increase the severity of NDV in layers (Gowthaman, Singh et al., 2019). It is important, therefore, that these disease conditions are differentially diagnosed. As a part of differential diagnosis, we tested a few samples for avian infectious bronchitis virus (IBV) and NDV and found some of them to be positive for IBV and negative for NDV (data not shown). Hence, respiratory signs exhibited by the birds can also be due to IBV. But viral infectious diseases such as infectious laryngotracheitis and pneumovirus infection are not currently known to pose a significant threat to poultry industry in Kerala. However, to rule out the possible contributory role of these respiratory viruses, selected samples from positive flock are being analysed by Next Generation Sequencing.
This study has several limitations. First, although this analysis suggests a higher number of outbreaks in zone 1, this may not give an accurate picture of the NDV prevalence in zones 2 and 3 as they were far away from the sampling centre location and information on outbreak occurrence in these areas may be under-reported. This study considers that the primers for M and F gene are capable of detecting prevalent NDV strains. This might not be the case as variant strains may be present in this region. This seems to be especially true in the case of F gene testing which gave a very low positivity rate. Lastly, non-random sampling is a limitation of the study for generalisation of results but does not affect the relationships observed between infection and risk factors.
On the contrary, during the study we have also observed that cases that appeared to be ND were not detected by the M or F gene primers.
It has been reported that techniques based on probe/primer hybridisation to a specific site are very sensitive to mismatches that often produce false-negative results (Cattoli et al., 2009;Kim et al., 2006). The M and F gene primers used in this study were described in 2004 and 2009, respectively, and there has been significant additions in the genomic data of NDV in the databases and new genotypes have been described which might not be detected by the primer/probes used. In addition, the authors who have designed the M gene primer in the assay have reported that the oligos were not an exact match to all APMV-1 isolates whose sequence is available in the databases at that time (Wise et al., 2004).

ETHICS STATEMENT
The study was undertaken after obtaining permission from Insti-

PEER REVIEW
The peer review history for this article is available at https://publons. com/publon/10.1002/vms3.747