Validation of a reduction in time for avian influenza virus isolation using specific pathogen‐free embryonated chicken eggs

Abstract

Background

The international gold standard for avian influenza virus (AIV) diagnosis is virus isolation (VI) in specific pathogen‐free embryonated chicken eggs (ECEs). AIV isolation typically involves a 6‐day turnaround, during which premises under suspicion for notifiable AIV infection are held under restriction regardless of molecular diagnoses, often with significant welfare implications.

Methods

A reduction in time for negation by VI was investigated following experimental inoculation of AIV from known‐positive original clinical material into ECEs. VI data derived from more than 600 case investigations from epizootics of high‐pathogenicity AIV (HPAIV) in Great Britain since 2016 and from low‐pathogenicity AIV (LPAIV) cases in Great Britain since 2014 were examined to support a reduction in test timing using alternative regimens.

Results

HPAIVs were isolated during the first passage, and for LPAIV VI, the second passage could be reduced to 2 days. Power analysis showed that the benefit of reducing the number of days outweighed the risk of missing a positive isolate.

Limitations

Limited data were available from experimental inoculations.

Conclusion

This truncated methodology, which enables an earlier release of restrictions, may substantially ease the economic implications of restriction. It could also reduce bird welfare implications and improve international standards without loss of test performance.

INTRODUCTION

Avian influenza (AI) is a global threat to the poultry industry as well as to animal and human health. ¹ , ² , ³ Safe and accurate diagnosis is a cornerstone for effective disease control. The World Organisation for Animal Health (WOAH) historical ‘gold‐standard’ method for the diagnosis of AI is virus isolation (VI) in specific pathogen‐free (SPF) embryonated chicken eggs (ECEs). ⁴ Real‐time reverse transcription polymerase chain reaction (rRT‐PCR) has largely replaced VI in many settings, but there are some applications where robust and sensitive VI is still very important. The VI approach offers some advantages over rRT‐PCR, namely, it demonstrates the presence of an infectious virus, which can be important when defining the context and relevance of results, especially in investigations where all positive birds from a single case may have weak values in rRT‐PCR tests, which could be indicative of either early or late infection. ⁴ In the latter case, the data can be interpreted together with the clinical history and may fundamentally impact the decision on whether the case is proven to have actively infected birds at specific locations or premises with or without onward transmission risk. ⁴ Furthermore, while the fidelity of rRT‐PCR assays for highly sensitive and specific detection of AI virus (AIV) has been proven, there is always a theoretical possibility that nucleotide substitutions at key rRT‐PCR target sites in the viral genome may compromise the utility of the molecular test. ⁴ , ⁵ Therefore, in accordance with the WOAH Manual for Diagnostic Tests and Vaccines, ⁴ VI can be used to safely exclude AIV infection in kept bird case investigations, which acts as a trigger to lift disease control measures placed upon the premises while diagnostic tests are conducted. These measures are designed to prevent the spread of infection until the status of the case is known. This point is important as welfare considerations for premises under restriction while awaiting negation by day 6 VI are significant, and any reduction in the time to be able to release a premises from restriction would improve bird welfare options. ⁶

In Great Britain, our standard VI protocol requires two passages of varying duration. For notifiable avian disease (NAD; i.e., AI and Newcastle disease [ND]) investigations, quarantine/import testing and wild bird surveillance for AI, two rounds of virus propagation in ECEs are attempted, comprising an initial passage for 2 days (P1) followed by a second passage (P2) lasting up to 4 days (denoted as the ‘2 + 4’ control/standard passage model). ⁴ The full duration of the current VI procedure in Great Britain is therefore 6 days. During this period, the premises containing birds suspected of being infected with NAD agents remain under official disease control restrictions while diagnostic tests are completed.

The aim of this study was to determine from experimental inoculations of known‐positive clinical material into ECEs and analyses of VI data from high‐pathogenicity AIV (HPAIV) and low‐pathogenicity (LPAIV) case investigations in Great Britain whether the currently adopted two‐passage model for isolation of AIV could be reduced from 6 days (using the ‘2 + 4’ model) without any loss of test sensitivity, which would benefit Great Britain but also provide wider international improvement.

MATERIALS AND METHODS

SPF ECEs and sample inoculation

The 9‒11‐day‐old SPF ECEs (chicken anaemia virus negative) used in the study were supplied by VALO BioMedia. In accordance with the legal obligations under the Animals (Scientific Procedures) Act 1986, all egg inoculations were performed under Home Office project licences P5275AD31 (Diagnosis of statutory and endemic avian viral diseases) and PP7633638 (Investigation of influenza virus and avian avulavirus disease) in ACDP3/SAPO4 biocontainment facilities at the Animal and Plant Health Agency‐Weybridge.

Experimental inoculation of AIV from known‐positive original clinical material into ECEs and determination of haemagglutination (HA) activity for each shorter passage model

Original clinical material comprising brain or intestinal tissue pools infected with the HPAIV strain H5N8 A/chicken/England/035986/20 was inoculated undiluted into ECEs using each of the shortened and control/standard models as described in Table 1. The durations included 2, 3 or 4 days (referred to as the ‘1 + 1’, ‘1 + 2’, ‘1 + 3’ or ‘2 + 2’ passage models, with the numbers denoting the respective P1 and P2 lengths of time; Figure 1) following inoculation of AIV into ECEs.

TABLE 1.

Daily schedule for embryonated chicken egg (ECE) inoculations for each passage model per virus inoculum

Day	Minimum number of new ECEs used per model a	P1 or P2 stage of model	Total number of ECEs to undergo HA test on the day	Passage ‘model’ completed	Comments
0	5 ECEs	P1 started for all models	Not applicable	None	All passage models started
1	‘1 + 1’—3 ECEs ‘1 + 2’—3 ECEs ‘1 + 3’—3 ECEs	P1 for ‘1 + 1’ completed P1 for ‘1 + 2’ completed P1 for ‘1 + 3’ completed	1	No model completed	One ECE was passaged into three eggs for each model. No passage on this day for ‘2 + 2’ or ‘2 + 4’ models
2	‘1 + 3’—no ECEs ‘2 + 2’—3 ECEs ‘2 + 4’—3 ECEs	P2 for ‘1 + 3’ in process P1 for ‘2 + 2’ completed P1 for ‘2 + 4’ completed	4 (3 for ‘1 + 1’ and 1 for ‘2 + 2’ and ‘2 + 4’)	‘1 + 1’	‘2 + 2’ and ‘2 + 4’ model ECEs passaged and HA tested, ‘1 + 1’ model ECEs undergo HA test
3	0	P2 for ‘1 + 2’ completed P2 for ‘1 + 3’ in process P2 for ‘2 + 2’ in process P2 for ‘2 + 4’ in process	3 for ‘1 + 2’	‘1 + 2’	‘1 + 2’ model ECEs undergo HA test
4	0	P2 for ‘2 + 2’ completed P2 for ‘2 + 4’ in process	6 (3 for ‘1 + 3’ and 3 for ‘2 + 2’)	‘1 + 3’ and ‘2 + 2’	‘1 + 3’ and ‘2 + 2’ model ECEs undergo HA test
5	0	P2 for ‘2 + 4’ in process	0	0	ECEs candled
6	0	P2 for ‘2 + 4’ completed	6 (3 for ‘2 + 4’ and 3 remaining from day 0)	‘2 + 4’ (control/standard)	‘2 + 4’ ECEs undergo HA test

Abbreviation: HA, haemagglutination.

^a Total of at least 20 ECEs were required per virus inoculum for the evaluation of all passage models (including the ‘2 + 4’ control/standard model).

FIGURE 1.

Illustration of the different approaches (‘1 + 1’, ‘1 + 2’, ‘1 + 3’ and ‘2 + 2’) for each avian influenza virus passage model under investigation compared with the current (control/standard) ‘2 + 4’ model

Five eggs were inoculated on day 0 with 200 µL of homogenate to generate P1 eggs, which were then incubated at 37°C overnight. Each egg batch/passage required an extra non‐inoculated egg to act as a control for that specific egg batch/passage. At each passage stage, the egg embryos were decapitated as per the Home Office licensure, and the harvested egg fluid underwent HA testing. Shortened passage scenarios were carried out on day 1 post‐inoculation for the ‘1 + 1’, ‘1 + 2’ and ‘1 + 3’ models (Table 1). On this day, the HA test was performed on one day 1 P1 egg and 200 µL of supernatant fluid harvested from the egg was inoculated back into three eggs to generate P2 eggs for each of the ‘1 + 1’, ‘1 + 2’ and ‘1 + 3’ models, thereby using a total of nine eggs. Similarly, for the ‘2 + 2’ and ‘2 + 4’ (control/standard) models, an HA test was undertaken on a single day 2 P1 egg, and the harvested supernatant fluid was inoculated back into three eggs to generate P2 eggs for each model, which required a total of six eggs. ECEs were kept in an incubator/hot room at 37°C until they were ready for passage work and HA testing. The HA test was performed on the P2 ECE material for all models on the appropriate days according to the daily schedule (Table 1).

Isolation of virus in ECEs from samples submitted for virological investigation during HPAIV and LPAIV outbreaks occurring in Great Britain

VI data derived from more than 600 case investigations from the recent epizootics of HPAIV (clade 2.3.4.4b) and from LPAIV cases in Great Britain were reviewed to find a correlation with the evidence base from the experimental investigations detailed above in support of a reduction in test timing using alternative regimens. The assessment of samples from NAD responses covered the HPAIV epizootics occurring in Great Britain during 2016‒2017 (H5N8, 16 December 2016 to 31 December 2017), 2020‒2021 (H5N8, 1 October 2020 to 30 September 2021), 2021‒2022 (H5N1, 1 October 2021 to 30 September 2022) and 2022‒2023 (H5N1, 1 October 2022 to 30 September 2023). The clinical samples from LPAIV incursions into Great Britain since 2014 were similarly subjected to VI in 9‒11‐day‐old SPF ECEs according to the internationally recognised methods. ⁴

Typically, for NAD investigations, VI would be attempted on all sample sets comprising oropharyngeal and cloacal swabs and standard tissues (i.e., brain, lung, trachea, intestinal contents and mixed viscera homogenates) if carcasses were submitted. However, owing to the sheer scale of the 2021‒2022 and 2022‒2023 AI epizootics in terms of the number of suspect cases from which samples were submitted for laboratory virological investigation during the peak periods of these epizootics, and the neurotropism of the H5N1 2.3.4.4b HPAIV, a streamlined algorithm was implemented and VI was often attempted only on the brain samples (where carcasses were submitted) to protect staff resource and to help satisfy Home Office targets for reduction and refinement in the use of ECEs. ⁷ Brain tissue was preferred in these instances, as analysis of this sample type was useful in distinguishing HPAIV infection from LPAIV infection. For each NAD investigation, the previously described statutory AI testing algorithm with VI, rRT‐PCR and serology was employed. ⁸ For each suspected HPAI or LPAI case investigation, the VI methodology followed the control/standard ‘2 + 4’ model, whereby the official sample (200 µL) was inoculated into the allantoic cavity of each of three ECEs (P1 eggs), which were then incubated overnight at 37°C. In line with the Home Office project licence legislation, all ECEs were candled on day 1 post‐inoculation. From 2022, if the rRT‐PCR results were positive, the chorioallantoic fluid (CAF) from one of the three inoculated ECEs was typically taken off for HA testing per submission on day 1. Otherwise, on day 2 post‐inoculation, the CAF from one egg was taken off per sample and subjected to HA testing. In the event of haemagglutinating virus being detected on either day 1 or day 2 (P1), at least one isolate was harvested and placed into storage and the VI procedure was concluded by chilling all the remaining eggs inoculated for the submission at 4°C according to the Home Office regulations.

In the absence of haemagglutinating virus following HA testing for the submission on day 2, the CAF harvest from each sample (P1) was inoculated back into three fresh ECEs for incubation for up to four more days at 37°C (P2), along with the remaining one or two ECEs for each sample (Figure 1). All ECEs were subsequently candled on day 3, which corresponded to day 1 post‐inoculation of P2 eggs. Eggs were also candled on day 4 post‐inoculation if the P1 eggs reached 14 days of age. HA testing was then performed on those eggs showing signs of embryo death to determine whether this was due to haemagglutinating virus. Furthermore, in strict adherence to the Home Office licence regulations, ECEs had to be candled daily from 14 days of age to check for egg death. Therefore, in the absence of haemagglutinating virus in the ECEs up to that time point, all inoculated ECEs remaining alive had to be candled on day 5 and subjected to HA testing on day 6 (Figure 1).

RESULTS

Virus isolation in ECEs using known‐positive original clinical material with variable passage parameters

Following ECE inoculation with the dilution series of HPAIV H5N8 A/chicken/England/035986/20, re‐isolation of virus at P1 was successful with the ‘2 + 2’ and ‘2 + 4’ (control/standard) models, eliminating the need for an additional passage. Positive haemagglutinating virus was recovered with the reduced ‘1 + 2’, ‘1 + 3’ and ‘2 + 2’ passage models but not for the ‘1 + 1’ model. As all P1 eggs had died on day 2, no P2 egg harvests were set up. All P2 eggs set up on day 1 were dead in the ‘1 + 1’ or ‘1 + 2’ models (data not shown).

Statutory isolation of HPAIV and LPAIV in ECEs during AI outbreak events in Great Britain

2016‒2017 HPAIV H5N8 (clade 2.3.4.6) epizootic in Great Britain

Table 2 shows the time periods (number of days) post‐inoculation into ECEs for isolation of HPAIV H5N8 from each of the 13 infected premises (IPs) comprising the 2016‒2017 HPAIV H5N8 outbreak in Great Britain. The period of this epizootic was defined as being from 16 December 2016 to 31 December 2017. In all cases, haemagglutinating virus was isolated at P1 (without passage) by day 2 post‐inoculation of ECEs from at least one epidemiological unit investigated per case. This typically occurred on day 2 post‐inoculation of ECEs, but virus was also isolated on day 1 in two cases. The second passage and later time points were not investigated as the eggs were already positive for each submission.

TABLE 2.

Haemagglutinating virus isolation data from the 13 infected premises during the 2016‒2017 high‐pathogenicity avian influenza virus H5N8 outbreak in Great Britain

Infected premises number	Notifiable avian disease reference number	APHA submission reference(s) for each infected premises	Day of isolation of haemagglutinating virus at P1 post‐inoculation
1	DPR2016/31	AV001417‐16	1
2	DPR2016/43	AV000001‐17 and AV000002‐17	2
3	DPR2017/06	AV000039‐17	2
4	DPR2017/24	AV000143‐17	2
5	DPR2017/35	AV000222‐17 and AV000223‐17	2
6	DPR2017/37	AV000248‐17	2
7	DPR2017/39	AV000281‐17	2
8	DPR2017/41	AV000296‐17 and AV000297‐17	2
9	DPR2017/51	AV000402‐17 and AV000403‐17	2
10	DPR2017/58	AV000503‐17	2
11	DPR2017/85	AV000851‐17	1
12	DPR2017/88	AV000870‐17 and AV000871‐17	2
13	DPR2017/101	AV001009‐17	2

2020‒2021 (H5N8), 2021‒2022 (H5N1) and 2022‒2023 (H5N1) HPAIV epizootic seasons

The VI data from the 13 IPs from the 2016‒2017 HPAIV H5N8 outbreak are further summarised in Table 3 along with the summary of the time periods (days) for isolation of haemagglutinating virus post‐inoculation into ECEs from samples submitted from the IPs over the 2020‒2021 (H5N8), 2021‒2022 (H5N1) and 2022‒2023 (H5N1) HPAIV epizootic seasons. Table 3 shows the numbers of positive IPs and suspect‐infected premises (SPs) testing negative for the presence of NAD based on the combination of rRT‐PCR and VI results. In all cases where haemagglutinating HPAIV was successful (n = 394), an isolate was harvested at the P1 stage either on day 1 or 2 post‐inoculation of the ECEs. The positive VI result from all IPs was also supported by the positive results from the statutory rRT‐PCR testing of the official clinical samples submitted from the respective IP and from the gross pathological examination when carcasses were submitted for investigation along with swabs (data not shown). Where no virus was isolated (using the control/standard ‘2 + 4’ model), negative rRT‐PCR results were also reported.

TABLE 3.

Summary of the haemagglutinating virus isolation (VI) data from the high‐pathogenicity avian influenza virus (HPAIV) H5 epizootics occurring in Great Britain since 2016

		Day haemagglutinating virus isolated post‐inoculation of embryonated chicken eggs
Outbreak denomination	Positive IPs or negative SPs following PCR and VI in embryonated chicken eggs	1	2	3	4	5	6	No virus isolated after 6 days using the ‘2 + 4’ model	Total
2016‒2017 HPAIV H5N8	Positive IP (PCR and VI)	2	11					0	13
	Negative SP (PCR and VI)							32	32
2020‒2021 HPAIV H5N8	Positive IP (PCR and VI)	4	18						22
	Negative SP (PCR and VI)							26	26
2021‒2022 HPAIV H5N1	Positive IP (PCR and VI)		152						152
	Negative SP (PCR and VI)							70	70
2022‒2023 HPAIV H5N1	Positive IP (PCR and VI)		207						207
	Negative SP (PCR and VI)							82	82

Abbreviations: IP, infected premise; PCR, polymerase chain reaction; SP, suspect‐infected premise.

LPAIV incursions into Great Britain from 2014 to 2023

The annual VI data from the IPs associated with the LPAIV incursions into Great Britain from 2014 to 2023 are summarised in Table 4, which shows either the day of isolation of haemagglutinating virus following initial inoculation of ECEs or whether no viruses were isolated using the control/standard ‘2 + 4’ model. LPAIV incursions included non‐H5 and non‐H7 AIV cases.

TABLE 4.

Summary of the annual virus isolation data from the low‐pathogenicity avian influenza virus incursions into Great Britain since 2014

	Day haemagglutinating virus was isolated post‐inoculation of embryonated chicken eggs
Incursion year	1	2	3	4	5	6	No virus isolated after 6 days	Total
2014		2					0	2
2015		1					0	1
2016			1				0	1
2017							1	1
2018							1	1
2019		2		1			0	3
2020		5		1	2		1	9
2021		1					0	1
2022		1					0	1
2023	1						0	1

On three occasions, the associated LPAIV (H5N3 in 2017, H6N5 in 2018 and H6Nx in 2020) was detected by molecular methods only. Out of 16 VI‐positive cases stemming from the LPAIV incursions, haemagglutinating virus was detected by day 2 (from P1 eggs) in all but three cases (Table 4). With two cases (both occurring in 2020), harvesting of virus occurred on day 5 as the inoculated eggs were not examined for haemagglutinating activity on days 3 and 4. The data from these two cases were therefore not included in the analyses. The virus isolates were obtained on either day 3 or 4 post‐inoculation on three occasions, which supported our hypothesis that the second passage can be reduced from 4 to 2 days. Overall, the data further support the use of shortened passage models covering up to 4 days in duration with passage.

DISCUSSION

While PCR has largely replaced VI in many applications where definitive detection of AIV is required, this method is nevertheless still used as an international ‘gold’ standard. ⁴ There are circumstances where confirmation or exclusion of the presence of infectious AIV is important for case investigation, understanding the pathobiology of the virus strain, creation of AIV repositories for phylogenetic analyses and in virus/strain characterisation studies. Current legislature in Great Britain dictates that VI must be performed in all statutory NAD laboratory investigations. ⁹ Therefore, in addition to being a diagnostic tool, a critical application of VI in ECEs is the ability to reliably negate suspicions of disease due to AIV in kept birds that have already returned preliminary PCR results indicating the absence of the virus. ⁵ VI has some advantages over PCR when seeking to confirm active infection, but obtaining such results is laborious and time consuming using methodology that was developed many decades ago. Therefore, in this study, we aimed to review the methodology and investigate if alternative designs (models) could achieve faster results with the same level of test sensitivity. Consequently, a scientific evidence base has been generated to support a revision to the internationally defined diagnostic methods for isolation of AIV prescribed by WOAH. ⁴ We report that a negative VI result for AIV in ECEs can be confirmed by day 4 (with passage included within that timeframe) following inoculation of clinical samples into ECEs, rather than 6 days, as is currently prescribed using the control/standard ‘2 days + 4 days’ model.

Analysis of the VI passage data from statutory NAD investigations during the epizootics of HPAIV H5N1 and H5N8 (clade 2.3.4.4b) occurring in Great Britain since 2016 showed that, where VI was successful, the isolate was generated from the P1 ECEs (day 1 or 2 post‐inoculation) (Tables 2 and 3). Table 3 shows that all 394 of the HPAIV‐positive isolates generated were detected within day 1 or 2 of the first passage. Aside from NAD investigations, successful isolation of other HPAIV H5 lineages (clade 2.3.2.1a) has also been reported by day 2 post‐inoculation of ECEs (data not shown; Joshua Lynton‐Jenkins, personal communication, 2024). There were 210 examples of ‘no virus isolations’ in either the first passage (days 1‒2) or the second passage (days 3‒6). In Table 4, which contains isolation data for LPAIV cases, 13 positive isolations were made in the first passage and three more isolations were made in the first 2 days of the second passage. The two isolations on day 5 were discarded as these eggs were not checked on day 3 or 4. There were three examples of ‘no virus isolation’ after 6 days. The specificity of the HA assay, particularly problems arising from bacterially contaminated ECEs (and the necessity of sterility tests), was not perceived to be an issue due to the close monitoring and candling of the inoculated eggs in accordance with the Home Office licencing as described. While a generic influenza A virus rRT‐PCR screening assay could be used instead of the HA assay to confirm the presence of haemagglutinating virus, such an approach would be too laborious and cumbersome during prolonged NAD outbreaks. By considering only the positive isolations, a retrospective power analysis was carried out to determine the probability that a positive sample would be isolated on days 1‒4 (the 2 + 2 model) rather than on day 5 or 6 (the 2 + 4 control/standard model). By using the 410 cases of positive isolations that occurred on days 1‒4 compared with 0 positive isolations that occurred on days 5‒6, and a 95% confidence interval, we calculated this probability as 98%. This analysis showed that the benefit of reducing the number of days outweighed the risk of missing a positive isolate.

Experimental inoculation of ECEs with known positive original clinical material provided further evidence, albeit limited, to support the adoption of the shortened passage model from 6 to 4 days' duration. Importantly, the viruses from statutory NAD investigation activities in the field (i.e., original clinical samples) were not necessarily well adapted to eggs, as these were derived from multiple different avian species and, therefore, may not efficiently and productively replicate in ECEs. Hence, the observation that all fresh clinical samples from naturally infected birds that contained virus (determined from rRT‐PCR testing), submitted for investigation across multiple outbreak seasons in Great Britain were viable with respect to VI within the 4‐day timeframe, thereby supporting this proposed reduction in the passage protocol (Tables 3 and 4). Furthermore, the infectivity of material inoculated from field samples may vary according to multiple factors, such as the condition of the birds when the samples were taken and temperature during transport, which could influence the likely rate of initial rapid virus replication from the samples. ¹⁰ , ¹¹ Nevertheless, despite these obstacles to successful VI, the outbreak data fully support a reduction in protocol length to 4 days with passage without the loss of test sensitivity, including specific situations where weak PCR results adjacent to the test borderline are obtained from clinical samples. ⁵ While the ‘2 + 2 model’ has not been validated for isolation of LPAIV from wild bird samples, it has been shown that the performance of the initial screening rRT‐PCR assay on the samples is the key determinator of success in detecting virus, with strong signals correlating with an increased probability of obtaining an isolate due to higher quantities of infectious virus in the sample.

The optimisation of this methodology provides an evidence base to modify international standard approaches for VI of AIV. If international agreement is achieved, this will have a profound impact on the bird industry in instances where premises are negative using molecular tools but international standards require VI to be completed. While it is extremely difficult to align costs against the benefit of a reduced period assigned to VI testing for AIV, the impacts are likely to be substantial for both avian and human welfare. ¹² From an avian perspective, a reduction in time where disease control restrictions are in place will have an impact on flocks that are negative for NAD. Welfare implications in these situations can be considerable, especially where flocks reach age and size capacity within housing. Furthermore, being under disease control restrictions can have a knock‐on effect on downstream processes. For example, a reduction in the positive/negative virus detection reporting time length also benefits chicks in hatcheries with no ‘home’ to go to, as the sites of destination may still contain birds, or have not been able to carry out effective cleansing and disinfection ready to receive new birds. The financial cost and emotional burden felt by owners of poultry in this situation are considerable. The costs incurred when the affected premises are placed under disease control restriction include losses from untreated sick birds, losses due to birds not being able to be processed appropriately according to age and size and losses from birds not being available for the correct product/season, including birds growing too large to fulfil order requirements and birds missing key market dates, for example. Therefore, despite the unavailability of cost‐benefit financial data, these factors mean that the costs are already significant for a business and exacerbate the emotional stresses placed on farm staff. The evidence gathered in this study substantially impacts upon all features of the sector and will potentially have global outreach if it is agreed that it provides a robust evidence base to safely change method design. Data regarding negation for ND are being gathered in parallel to further support a change in this methodology that encompasses all NADs, since by definition, disease suspicions cannot safely exclude one disease from the other and both are relevant for exclusion of exotic avian viral disease. ⁴ , ⁹

The successful outcomes from these analyses collectively supported a reduction in the length of time businesses/premises remain under official restrictions when investigating suspicions of AI, thereby substantially reducing costs both to the government and industry through earlier detection or negation of infection, with concomitant reduction in impact on animal welfare with the shortened VI test turnaround times. In addition to the financial savings, of particular significance in reducing the total time length of the shorter passage model is the potential to lift disease control restrictions on the affected premises following a negative (i.e., ‘no virus isolated’ from a negative HA test result) 2‐day passage result at P1 following inoculation of the original clinical samples or on collected allantoic fluids at ECE passage in tandem with negative rRT‐PCR results.

AUTHOR CONTRIBUTIONS

Conceptualisation; data analysis; resources; writing—original draft and writing—review and editing: Scott M. Reid. Laboratory testing; data analysis and writing—original draft: Vivien J. Coward. Data analysis; resources and consultancy: Joe James. Resources and consultancy: Rowena D. E. Hansen. Data analysis: Colin Birch. Data analysis: Mayur Bakrania. Conceptualisation; data analysis; resources; consultancy and writing—original draft: Sharon M. Brookes. Conceptualisation; resources; consultancy and writing—review and editing: Ian H. Brown. Resources; consultancy and writing—review and editing: Ashley C. Banyard. All authors have read and approved the manuscript.

CONFLICT OF INTEREST STATEMENT

The authors declare no conflicts of interest.

ETHICS STATEMENT

The described virus isolation using SPF ECEs was performed according to the internationally recognised method described by WOAH under Home Office Project Licences P5275AD31 (Diagnosis of statutory and endemic avian viral diseases) and PP7633638 (Investigation of influenza virus and avian avulavirus disease).

Reid SM, Coward VJ, James J, Hansen RDE, Birch C, Bakrania M, et al. Validation of a reduction in time for avian influenza virus isolation using specific pathogen‐free embryonated chicken eggs. Vet Rec. 2024;e4842. 10.1002/vetr.4842

DATA AVAILABILITY STATEMENT

The data that support the findings of this study are available from the corresponding author upon reasonable request.