It Started Early - Now It is Speeding up...

H5N1 continues to cover the upper U.S.; is there any prospects for partial depopulation as demonstrated by the Germans in 2021?

The early fall H5N1 2.3.4.4b onslaught continues, with genotype still undetermined. We’ve seem to have “progressed” from about a case per day in September to 2.7 cases per day in the first 10 days of October. Anyone ready to admit that this is a wildlife reservoir endemic infection not suitably handled as a foreign animal disease or a select agent?? OK - we’ll keep pretending to please our trade partners’ imaginations.

First, here is the completed list from September, updated from last week’s column:

We are now 10 days into October reporting as of this weekend and HPAI reports have accelerated, if anything. We had 2 late week reports:

The Washington State Department of Agriculture (WSDA) confirms first detection of H5N1 in a large commercial poultry facility in Moses Lake. Another source confirmed that the site is a 1.97 million hen layer facility. Saturday I discovered that Indiana has reported 2 commercial duck infections: Commercial Duck Farm in Elkhart County Hit with Bird Flu | Hoosier Ag Today. Then I had to reconfigure this chart on Monday morning after discovering that 2 October 8th Utah turkey flocks had been added (out of order) over the weekend. I also added a Park County Montana backyard flock reported by FluTrackers. Finally, Wisconsin also reported a new backyard flock in Racine County this morning, marking at least 58 known infections to date. So, here is the updated current NVSL case list for the first 10 days of October only with the above cases added:

We have added detections at the rate of 2.7 per day since October 1st, after adding about 1 per day in September. 57 flocks have been depopulated in total, with 36 of those being commercial poultry operations. Turkeys have suffered by far the most, with 28 meat bird and 3 breeder flocks depopulated; however, we’ve also already lost 3 layer flocks with 5.5 million hens.

Here is a brief state-level breakdown:

Perhaps the most discouraging fact is that the outbreak hasn’t really yet moved into many of the poultry dense areas, being confined so far to the upper latitudes of the Midwest and Western states. If other poultry areas are hit with the same intensity that has struck the upper Midwest as the bird migration continues south, total losses will likely be catastrophic. We’ll see if weather changes or intensive biosecurity efforts can modify the outcomes. I’m not convinced that any of the areas struck to date were that negligent in biosecurity.

Some epidemiological feedback from state and federal responders, anonymized to protect confidentiality would certainly be helpful to all industry participants in better understanding likely routes of infection into affected flocks. Are multiple area infections being carried either by people or area/aerosol spread, as evidenced by nearly identical genotypes within areas? Or do most of these appear to be independent wild bird incursions with wider genotypic diversity? Epidemiology of infections continues to be a black box for all not directly involved.

Additionally, not to be a broken record, we still lack any evidence of H5N1 genotype(s) on the contemporary domestic avian outbreaks or wild bird isolates. Regardless, currently my heart truly goes out to colleagues at NVSL, the NAHLN Office, and within VS program and response units working under the barbaric conditions enforced by the current administration, OMB, and Congress through the shutdown, with all the associated uncertainty and flat-out cruelty towards Federal employees. Your continued service under such conditions is extraordinary and valued. Whatever disagreements I may have with you on data transparency pales in comparison with my appreciation for your professionalism, the work you do, and the dedication the current debacle requires of every Federal employee in this extraordinary crisis. I personally remember how frustrated I was in much less challenging political shutdowns. God be with you all and carry on…I remain confident the American people will eventually make this situation right; we really have no choice but to reclaim our excellence in regulatory governance and in leading the world in animal disease management.

Partial Culling - It Can Work, But…

The Germans just published a paper outlining a successful partial depopulation of an H5N1 HPAI outbreak in a broiler operation. This paper caught my eye, in part because USDA Secretary Rollins early on made a big point of implying “excessive” whole farm depopulations as a mistake in previous HPAI response activities. It’s useful examine this review of a partial broiler breeder farm depopulation in Germany from early 2021 in response to an H5N8 outbreak to better understand what factors can improve odds for successful partial depopulations: Strict Biosecurity and Epidemiological Segmentation Enable Partial Culling During a Highly Pathogenic Avian Influenza Outbreak

Background:

The affected poultry holding reared parent animals for broilers on the ground. It consisted of four identical, separate epidemiological units, which are colloquially called “farms”. Each farm consisted out of five units (barns) (Figure 2). Female birds were kept in barns 1–4 (10,000 hens per barn) and male birds in barn 5 (6,000 cocks). In total, there were approximately 180,000 animals on the holding at the time of the HPAI outbreak.

Here is the layout of the 4 closely spaced “farms” (or “clusters”) on this broiler breeding grow-out unit or premise:

Case Summary

Background: The mandatory procedures to be followed after official confirmation of an outbreak of category A animal infectious diseases, including highly pathogenic avian influenza (HPAI), is laid down in European and national legislation. Typically, an outbreak of HPAI results in the destruction of the entire poultry population on the affected holding.

Case Presentation: The presented case reports a deviation from this approach, demonstrating the practicality of partial culling in a highly bio-secure, epidemiologically segmented holding. These on-site circumstances together with the specific risk assessment led to the elimination of only the affected unit, thereby inhibiting the further spread of the disease. After the destruction of the respective unit (farm), the other farms were closely monitored and tested continuously negative for HPAI virus (HPAIV) despite intensive systematic sampling. In the end, this procedure saved approximately 138,000 animals, i.e. 75% of the poultry population of the holding from destruction.

Conclusion: This case demonstrates the effectiveness of proper management and high-level biosecurity to avoid excessive destruction of animals in case of an infectious disease outbreak. It might be suitable as a best-practice example in similar situations.

Here is the timeline of the outbreak diagnosis:

On 3 February, the veterinarian monitoring the flock took samples from dead animals in the rear section of barn 1 of farm 4 for a bacteriological examination with antibiogram. An antibiotic treatment (amoxicillin) via the drinking water was started as coccidiosis and clostridiosis were suspected.

On 4 and 5 February, mortality increased in both sections of the infected barn, but no further action was taken as the antibiotic treatment had just started and it was too early to decide whether it was effective or not. As the mortality continued to increase, the veterinary practice was notified on 6 February (Saturday), and on 7 February (Sunday) the veterinarian in charge of monitoring the flock took samples from the dead animals again. These samples tested positive for avian influenza virus in the laboratory of the veterinary practice on 8 February.

The local veterinary authority was informed immediately, and official samples were taken. At this point, the most prominent clinical sign was the high number of fatalities. The state laboratory of Mecklenburg-Western Pomerania tested the official samples positive for HPAIV H5 by real-time PCR RNA. This was also confirmed by the national reference laboratory of the Friedrich-Loeffler-Institut on the same day. All samples tested positive for HPAIV of the H5N8 subtype using a real-time PCR RNA detection method specific to this subtype…

Culling and safe disposal of the animals on farm 4 took place on 10 February 2021 by a specialist company for disease control…

Please refer to the paper for complete discussions of the epidemiological investigation, with the most likely route of infection being a broken air exhaust fan leading to condensate build-up in drip pan in the affected barn, and the multiple steps taken to assure all parties that infection remained isolated in the single barn and removed with the depopulation and cleanup-disinfection processes. Interestingly, infection was not found in birds in the other 4 barns on the infected “farm” (site or cluster in U.S. terminology?) which was nonetheless depopulated. The other 3 “farms” within the “holding” (production premises) were spared depopulation through intense biosecurity and regular monitoring for signs of infection.

Why did partial depopulation succeed in this instance? Let’s consider several alternative possibilities, none of them mutually exclusive:

• H5N8 in 2021 was a less virulent virus, less prone to onward area spread

I don’t personally have any information on that possibility; however, the general consensus seems to be that the 2024-25 H5 2.3.4.4b strain is extremely contagious for poultry, perhaps making partial depopulation with current strains more challenging than it may have been with this 2021 H5N8 strain.

• Broilers are less susceptible to H5 infections than layers and turkeys

Anecdotally it appears that turkeys and layers are hit much harder than broilers by HPAI. Is this a difference in susceptibility and is there research to back that observation? Host sensitivities (resistance, minimum infectious dose, etc.) will be major factors to consider in considering the wisdom of attempting partial depopulation in any species of poultry.

• Level of Biosecurity, including “controlled” ventilation

First, fail-safe aggressive fomite control is a given in order to consider partial depopulation. Additionally, in this case air entered each building through “filtered” sidewall air inlets (filter specifications not provided) and exited the roof peak through fan ventilation. The outbreak occurred in the winter when ventilation requirements were minimal. No data was provided on climatic or prevailing wind conditions. More research is needed on optimization of ventilation to minimize the risk for viral spread between buildings.

• Size Size Size

I suspect that we have not paid nearly enough attention to animal body weight as a correlate for respiratory viral load in facilities when considering the feasibility of partial depopulations. Here is a simple illustration of this point:

In the case study above the infected barn housed 10,000 birds- let’s estimate them at 5 lb. each. Now compare them to a “typical” 50,000 bird turkey site or a 250,000 bird layer building. Total pounds of living infected and shedding birds are magnitudes different between the scenarios! Which barn is most likely to “leak” virus to other buildings nearby? Which site(s) are the biggest risk to area disease status?

This is one reason why “partial depopulations” of laying facilities are so likely to fail, especially if diagnosis is delayed until confirmatory diagnosis from sick bird necropsies. The viral load from the incubating infected barn is growing exponentially, even before clinical signs commence several days prior to molecular diagnosis.

While infected turkey and layer barns are excellent viral factories in local areas, uninfected barns are also excellent H5N1 “viral receptors or magnets” in the same areas! 1.6 million pounds of laying hens in a barn take in a huge amount of air, even in the winter. It only takes a few viral particles delivered to one hen for a short period of one day to transmit HPAI to the next barn on a site or to a separate complex “X” miles down the road (“X” distance TBD with further research). “Size matters” for both the viral dose delivered and for the receiving population of susceptible animals. This was seen repeatedly in California dairy cattle herds and layer flocks in the fall of 2024 in that state.

I recently discovered interesting research in Frontiers-Veterinary Science authored by a distinguished group USDA FAD researchers, modelers and university collaborators entitled: Epidemiologic consequences of preclinical transmission of foot-and-mouth disease virus in cattle. The study modelling here is quite technical with the main question being the effects of the inclusion of preclinical transmission in simulated FMD outbreak outcomes under various scenarios.

This investigation addresses the critical relevance of preclinical (incubation) phase transmission in determining the outcomes of simulated FMD outbreaks in the U.S. cattle production system. Our results demonstrate that the inclusion of even a single day of preclinical infectiousness exacerbates outbreak extent and duration under both optimal and suboptimal detection scenarios. These findings align closely with previous studies that emphasize the epidemiological importance of the incubation period in pigs and cattle.

This is one of those common-sense conclusions that we all too often continue to ignore in assessing all highly infectious disease outbreaks, including HPAI. The responders in Germany were actually quite fortunate to succeed, given a delayed diagnosis. Here is their likely infection to depopulation timeline:

The biggest objective for any high-consequence highly infectious disease outbreak remains an early diagnosis and response to minimize chance for further spread. In this case the officials were fortunate that this infection remained confined to one barn for 11 days until they managed to get the flock safely depopulated.

However, I’d point out that in the U.S this is currently about the same timeline we could expect with our HPAI diagnostic protocol! We might gain a day or 2 on the front end with really prompt reporting of clinical illness (6 versus 8?); however, we must field a FADD team and get samples to a NAHLN lab for screening. Assuming a non-negative result there, the flock would be quarantined, and depopulation planning could commence; however, no action can actually take place until the infection is confirmed at NVSL at least a day or 2 later. This all assumes that VS and the state officials remain adequately staffed to keep up with centralized diagnostics and incident command. It would be interesting to study 10 recent commercial flock outbreaks to measure real world performance under current response protocols. This is even something industry groups could assess independently of government assistance.

Regardless, we are missing opportunities to look for several high consequence diseases with localized screening tests. I noted with interest that the Germans ran an H5N1 test locally and found the infection! Their downfall was that the veterinarian failed to use it earlier - perhaps a cost or an awareness issue? Screening needs to be cheap, pooled, highly specific (no one wants false positives) and reasonably sensitive. We don’t need reference lab sensitivity standards because these tests can be repeated daily on non-clinical samples as an adjunct to high-risk symptomatic testing. “Screening-positive” samples could then be reported and investigated much earlier in the herd disease process, hopefully prior to clinical signs and extended viral shedding.

For anyone interested, here is a poultry house for sale on Alibaba! I use the picture as an illustration of industrial poultry housing ventilation to avoid any proprietary issues with picturing existing buildings!

The point I want to make is that we don’t really know “jack” yet about managing H5 HPAI spread risk in large populations of birds housed in ventilated facilities like this. The large number of susceptible birds exposed 24 hours per day to incoming air/dust/ particulates from the surrounding environment seems to multiply incredibly small odds for infection many times over from viruses shed by wildlife, nearby poultry, dairy cattle, swine, or even possibly from human employees. We haven’t proven this phenomenon, but repeated experiences with highly contagious viral outbreaks in multiple species (swine and poultry) provide a lot of indirect evidence for this risk.

Some will argue that the solution is to return to small family poultry operations- an attractive memory, but a non-starter for meeting today’s protein needs. The fact that Alibaba offers industrial poultry facilities for sale world-wide on the internet says something about the prospects for a return to viable back yard chicken production. We need to determine how to safely mange viral challenges for large groups of animals instead; that is how the world will be raising poultry, pigs, and dairy, regardless of any U.S. designs for a return to Old MacDonald’s nesting boxes.

Finally, please don’t contact me directly about magic UVC, electrostatic, or air filtering “solutions” for viral loads. I’m now nearly 74 years old and was being sold versions of these products in my 20’s. The sheer volume of air required to ventilate these units makes treatment “at scale” extremely challenging. Solutions will come and I will endorse them as the market adopts them based on proven value.

I have some very interesting swine influenza information remaining to present from the Leman Swine Conference - next time! As a tease - USDA has fixed the Swine Influenza Surveillance Program testing algorithm to better cover H5N1. Also, swine vets are working hard to keep human flu out of influenza free swine herds, with a similar challenge faced in timely diagnosing of flu in poultry (preclinical shedding in workers).

Stay tuned!

John