Biosafety risk assessment for production of candidate vaccine viruses to protect humans from zoonotic highly pathogenic avian influenza viruses

Abstract A major lesson learned from the public health response to the 2009 H1N1 pandemic was the need to shorten the vaccine delivery timeline to achieve the best pandemic mitigation results. A gap analysis of previous pre‐pandemic vaccine development activities identified possible changes in the Select Agent exclusion process that would maintain safety and shorten the timeline to develop candidate vaccine viruses (CVVs) for use in pandemic vaccine manufacture. Here, we review the biosafety characteristics of CVVs developed in the past 15 years to support a shortened preparedness timeline for A(H5) and A(H7) subtype highly pathogenic avian influenza (HPAI) CVVs. Extensive biosafety experimental evidence supported recent changes in the implementation of Select Agent regulations that eliminated the mandatory chicken pathotype testing requirements and expedited distribution of CVVs to shorten pre‐pandemic and pandemic vaccine manufacturing by up to 3 weeks.

of licensed influenza vaccines depends on a high level of structural similarity between the hemagglutinins (HA) of vaccine and circulating viruses. Therefore, pandemic vaccines with structurally well-matched HA antigens must be produced and administered as soon as possible after an emerging pandemic is detected. The National Pandemic Influenza Strategy calls for the United States (US) Department of Health and Human Services (DHHS) to maintain an updated library of CVVs and a strategic stockpile of vaccines to protect critical infrastructure in a pandemic emergency. 2 The Pandemic Influenza Plan 2017 Update includes expectations for DHHS and its partners to maintain a high level of readiness to start immunizing the US population with a well-matched pandemic vaccine within 4 months of a pandemic declaration. 3 Achieving this challenging pandemic vaccination goal requires aggressive time management in all vaccine development and manufacturing steps, including rapid development of a pandemic CVV and its immediate distribution to vaccine manufacturers by the World Health Organization and its (international) partners.

| PREPAR ATI ON AND D IS TRIBUTI ON OF C V VS AG AIN S T HPAI TO MAN UFAC TURER S
Most of the influenza vaccine supply for the United States is produced by growing viruses in embryonated chicken eggs. Pre-pandemic and pandemic vaccines for HPAI viruses developed using these technologies must be produced using attenuated CVV seeds that support worker safety during manufacturing. 4 CVVs derived from HPAI viruses for pandemic influenza preparedness (PIP) are generated using reverse genetic technology to remove the multibasic amino acid motif from the cleavage site of the HA, which is the major determinant of high pathogenicity in chickens; that is, HPAI virus. [5][6][7][8] Attenuated CVVs (with a monobasic amino acid HA cleavage site) are engineered by reverse genetics and characterized at public health laboratories under quality system regulations in compliance with Food and Drug Administration (FDA) and World Health Organization (WHO) guidance and subsequently transferred to vaccine manufacturers for development of vaccine virus seeds per current good manufacturing practice (cGMP) standards. [9][10][11][12] Currently, possession and transportation of wild-type HPAI viruses in the United States are regulated under Select Agent rules (CFR 9 part 121) by the United States Department of Agriculture (USDA) Agricultural Select Agent Program. 13 Furthermore, CVVs that are engineered with the attenuating monobasic HA cleavage site of an HPAI virus were considered Select Agents. However, CVVs with multibasic-deleted HA can be used at a lower Biosafety Level after exclusion from the Select Agent list per CFR9 121.3e guidance. 14 Exclusion from the Select Agent list was granted by USDA after review of the CVV information package with all the necessary experimental data supporting the loss of virulence for chickens and other phenotypic properties characteristic of low pathogenicity avian influenza (LPAI) viruses (Table 1, Figure 1A). This article describes the rationale and benefits of recent policy changes in the regulation of Select Agents in relation to development of CVVs for pandemic influenza preparedness and response purposes. 15

| B I OSAFE T Y RECORD OF C V V S SINCE 2 0 0 4
The regulatory policy framework for conducting biosafety risk as- In vitro characterization data inform three risk elements, as follows: (i) Genome composition. All CVVs against HPAI tested so far

| Mitigating residual virulence risk
The body of knowledge on the biological properties of PIP CVVs in chickens has increased by at least an order of magnitude since the original CVV regulatory framework was established in the United States more than a decade ago 13 ( These findings are consistent with the well-established importance of the multibasic and/or elongated cleavage site of the HA (molecular marker) for plaque formation in the absence of trypsin (in vitro marker) and the high virulence in chickens (in vivo IVPI). [34][35][36][37][38][39][40][41] The abundance of data indicates that H5/H7 CVV with monobasic cleavage site and trypsin-dependent plaque phenotype would have a negligible potential to cause severe disease in chickens.   26,46 All product-contact and raw materials are pre-qualified to be free of pathogenic infectious agents. It is worth noting that the use of synthetic DNA, which has been used increasingly in the generation of CVVs, results in the absence of wild-type HPAI HA genes in the production facility. Therefore, the risk of introducing a HPAI virus into the CVV production environment is extremely low to nearly zero. That said, a set of tests are performed after each CVV is produced to identify signals inconsistent with those that define the CVV as similar to LPAI viruses (   (Table A1). 51 To this end, groups of 8 to 10 birds were inocu-   5 These phenotypic differences are expected to result in rapid elimination of CVVs from exposed poultry because the reproductive number would be substantially lower than 1 (R 0 < 1). 53 These data indicate that PIP CVVs against HPAI with a PR8 backbone pose minimal risk to the health and well-being of poultry and other birds. Although avian (H1-H16) LPAI, including H7N9 LPAI of waves 1-4, and Testing by next-generation sequencing (NGS) has become standard practice for CVV characterization in 2016, adding sensitivity to the risk mitigation program (lower analytical limit of detection).

| REGULATORY POLICY REVIEW TO IMPROVE PANDEMIC PREPAREDNESS AND PUBLIC HEALTH RESPONSES
TA B L E 3 Biosafety risks and mitigation strategies for PIP CVV production and use swine (H1 and H3) influenza A viruses have zoonotic potential, they are not subject to Select Agent rules. However, the same scientific process to reduce virulence and transmissibility for agricultural animals by utilizing the PR8 backbone could be incorporated in public health and veterinary medical CVV risk assessment process to determine whether fewer enhancements are needed at BSL-2 than would be required for wild-type parent viruses, especially for parent viruses of foreign origin.
The recently updated policy governing PIP CVV development 15

CO N FLI C T O F I NTE R E S T
The authors declare no conflict of interest in this review.

D I SCL A I M ER
The findings and conclusions in this article are those of the authors and do not necessarily represent the views of the Centers for Disease  TA B L E A 1 Replication and shedding characteristics of H5 and H7 CVVs in chickens by intranasal inoculation a