African Swine Fever Modified Live Vaccine Candidates: Transitioning from Discovery to Product Development through Harmonized Standards and Guidelines
Round 1
Reviewer 1 Report
This is concise, well written, and comprehensive summary of the status of vaccine development of different types of experimental vaccines for ASF. It is a good contribution to the field of vaccinology in ASF.
A few minor points:
Line 130: perhaps a statement should be added referring to the lack of knowledge of immune mechanisms mediating protection in ASF.
Lines 255-255: this statement is not very clear. It needs to be rephrased.
Lines 277-280: please, refer to the results obtained by the genetic modification introduced in the virus, as presented in reference 67.
Lines 500-504: Please, refer to results presented in reference 104 which in some way contradict results described in reference 97.
Author Response
This is concise, well written, and comprehensive summary of the status of vaccine development of different types of experimental vaccines for ASF. It is a good contribution to the field of vaccinology in ASF.
A few minor points:
COMMENT 1. Line 130: perhaps a statement should be added referring to the lack of knowledge of immune mechanisms mediating protection in ASF.
RESPONSE 1: Agree. Former lines 258-261 and former Ref. 63 speak to this suggestion, however this existing statement is better placed after Line 130. The existing statement from former lines 258-261 was moved to immediately after Line 130, modified and additional statements and references added to address a related comment from Reviewer 2.
Revised statement starting on Line 131 reads: "There is a general knowledge gap on the early protective immune mechanisms directly mediating resistance following ASF MLV vaccine immunization [37]. A seminal study by Oura et.al [38] provides direct evidence for the critical role of CD8+ cells in protection against ASFV. In historical studies sera obtained from protected animals provides passive immunity to naive pigs subsequently challenged with ASF, suggesting a protective role of B cells and antibodies [39-41]."
COMMENT 2. Lines 255-255: this statement is not very clear. It needs to be rephrased.
RESPONSE 2: Agree. The statement "The phenomenon of sterilizing immunity following MLV immunization with the ASFV-G--DI177L vaccine was recently reported [62]" was deleted, as this reference relates to sterilizing immunity following vaccination AND challenge, versus than just vaccination, which is what Malmquist reported. The preceding and subsequent statements were also slightly modified.
The revised statement now reads "Although the molecular tools to detect small amounts of modified live virus that may have persisted in the tissues of these immunized pigs were unavailable six decades ago, Malmquist's finding argues against the hypothesis that ASF MLVs utilize viral interference or long-term persistent infection mechanisms to establish host resistance."
COMMENT 3. Lines 277-280: please, refer to the results obtained by the genetic modification introduced in the virus, as presented in reference 67.
RESPONSE 3: Agree. The (former) reference 67 (new reference 72) was inserted to reference the results obtained by the genetic modification in the virus.
The revised statement reads " Due to likely regulatory safety concerns on NH/P68 residual virulence, and potential manufacturing issues associated with NH/P68 genome stability grown on the COS7 cell line [72], there is considerable development risk in advancing NH/P68 as a vaccine candidate.
COMMENT 4. Lines 500-504: Please, refer to results presented in reference 104 which in some way contradict results described in reference 97.
RESPONSE 4: Disagree. The results in (former) reference 104 do not directly contradict results described in (former) reference 97 since (former) reference 104 does not describe a backpassage/reversion to virulence study using ASFV-G-ΔMGF. That said, the author agrees that these two genetically highly similar viruses (ASFV-G-ΔMGF and HLJ/18-6GD) may be behaving differently in vivo with respect to genetic stability. A new concluding statement was added to emphasize this point.
The newly added concluding statement, Lines 555-557 reads "Further studies are necessary to better understand the in vivo genetic stability of pandemic genotype II recombinant virus containing multiple deletions in the MGF family."
Reviewer 2 Report
The article by David Brake presents a thorough review of the historical and current literature on African swine fever virus vaccine development focusing in detail on those modified live vaccine candidates that are closest to or have been recently licensed. A compelling case is made for the need of standardised methodologies to be developed and accepted by WOAH to facilitate the acceptance of these vaccines globally. The final parts of the review discuss the main aspects to be considered to achieve this goal. This is a very timely contribution given the large increase in research activity in development of ASFV vaccines.
Specific points for consideration:
1. Lines 150 to 160. Passive transfer of sera from protected to naïve animals was shown historically to induce some partial protection implying that antibodies can have a role. The section on subunit vaccines should record the challenges associated with expression in native conformation of some key ASFV antigens, for example the major capsid protein p72. This protein requires a virus encoded chaperone, B602L, for correct assembly as a trimer in the virus particle. Subunit approaches have generally failed to take these complexities into consideration.
2. Recombination potential in the field is not considered in safety issues and appears to be a knowledge gap infrequently mentioned. This may be particularly relevant in regions in Africa where several genotypes are circulating or even in Asia since genotype I has been described circulating in China in addition to genotype II. Recombinant viruses may potentially evade vaccine induced protection.
Lines 698- 700. The need of vaccination to prevent transmission of virulent virus (ie R0 less than 1) to control disease outbreaks should be mentioned.
Line 703- 705. Harmonised humane endpoints are unlikely to be achieved due to variation in animal welfare standards in different countries in addition to non-quantitative assessment criteria. Viremia and clear signs such as temperature are easier to standardise.
Standardised cell lines for titration are discussed. I suggest to explain the criteria that would be used to identify such a cell line. The MA104 cell line is suggested as a possibility. However, a recent publication (Do et al., 2022) indicates an adaptation of ASFV to this cell line over several passages is required which would preclude its use. Other transformed porcine macrophage cell lines may be better suited for this purpose?
Author Response
Specific points for consideration:
- Lines 150 to 160. Passive transfer of sera from protected to naïve animals was shown historically to induce some partial protection implying that antibodies can have a role. The section on subunit vaccines should record the challenges associated with expression in native conformation of some key ASFV antigens, for example the major capsid protein p72. This protein requires a virus encoded chaperone, B602L, for correct assembly as a trimer in the virus particle. Subunit approaches have generally failed to take these complexities into consideration.
RESPONSE 1: Strongly agree on both excellent points above. The manuscript was revised accordingly, lines 131- 136 and lines 174-183 to reflect these suggestions.
Revised lines 131-136: "There is a general knowledge gap on the early protective immune mechanisms directly mediating resistance following ASF MLV vaccine immunization [37]. A seminal study by Oura et.al [38] provides direct evidence for the critical role of CD8+ cells in protection against ASFV. In historical studies sera obtained from protected animals provides passive immunity to naive pigs subsequently challenged with ASF, suggesting a protective role of B cells and antibodies [39-41]."
Revised lines 174-183: "One or more of the sets of most common ASFV subunit proteins that have been selected for proof-of-concept efficacy studies, may require the presence of virally encoded molecular chaperone proteins for proper protein folding into physical, stable conformation(s) mimicking native structures (i.e., icosahedral capsids) present on ASFV virion surface. For example, the major capsid protein p72 requires a virus encoded chaperone (B602L) for trimer assembly in the native virus particle [50, 51]. To date, ASFV recombinant subunit protein designs have largely failed to take into consideration these biologically relevant details which may increase the future probability of success. ASFV recombinant subunit protein-based vaccine approaches require a more basic understanding of ASFV structural protein assembly and the roles of nonstructural viral proteins in this process."
- Recombination potential in the field is not considered in safety issues and appears to be a knowledge gap infrequently mentioned. This may be particularly relevant in regions in Africa where several genotypes are circulating or even in Asia since genotype I has been described circulating in China in addition to genotype II. Recombinant viruses may potentially evade vaccine induced protection.
RESPONSE 2: Strongly agree. Several new statements on ASF MLV recombination potential were added along with supporting references.
The following statements were added, lines 720-742:
"Notably and infrequently considered, as evidenced by the absence of any published studies, is the important environmental safety issue for the potential of ASF MLV vaccine strain recombination in the field. This may be particularly pertinent in East and South Africa where numerous co-circulating genotypes have been reported [119], as well as in specific parts of Asia where genotype I viruses are now circulating [120]. Recombination frequencies vary extensively among virus families and is well documented for many RNA viruses [121]. MLV swine and poultry vaccine strain recombination with RNA field viruses such as Porcine Reproductive and Respiratory Syndrome Virus [122] and infectious bronchitis virus [123] have been reported. Double stranded DNA (dsDNA) viruses such as ASFV generally have larger genomes because of the higher fidelity of their replication enzymes. High frequency recombination can occur in some dsDNA viruses, such as the well-studied α-Herpesviruses in which homologous recombination is relatively frequent and associated with viral replication and DNA repair [124]. Vaccinia virus, another large dsDNA virus that shares some orthologous genes with ASFV, undergoes nonhomologous (end joining) recombination and new gene acquisition with relatively low frequencies to produce novel recombinant viruses [125]. The authors hypothesize that since poxvirus infection often results in high viremia levels (similar to ASFV), vaccinia virus recombination is sufficiently frequent to seed a small pool of novel recombinant viruses with potentially novel traits into larger populations of newly produced virus particles. The potential for ASF MLV vaccine strain recombination in the field is a research gap that may be addressed through in vitro co-infection studies using new techniques to accurately estimate recombination from NGS data [126, 127]".
- The need of vaccination to prevent transmission of virulent virus (ie R0 less than 1) to control disease outbreaks should be mentioned.
RESPONSE 3. Strong agree. Per this suggestion, a few new statements were added, see Lines 787-794.
"This third criterion is critical in the context of the need for ASF MLV vaccines to prevent transmission. The basic reproductive ratio (R0) is a predictive parameter associated with the average number of secondary infections produced from a single infectious event [132]. Generally speaking, a R0 value <1 is indicates an infection will not spread in a susceptible population. Future field studies will be necessary to determine the R0 value following ASF MLV vaccination in a defined population of pigs within a defined area known to have circulating ASFV".
- Line 703-705. Harmonised humane endpoints are unlikely to be achieved due to variation in animal welfare standards in different countries in addition to non-quantitative assessment criteria. Viremia and clear signs such as temperature are easier to standardise.
RESPONSE 4. Agree. Human endpoint considerations were deleted from the text and a comment on temperature and viremia as reasonable target readouts for standardization were added.
Revised sentence, Lines 797-799 reads " A standardized clinical scoring system based on the highest value acute disease objective parameter(s) such as temperature, along with a laboratory analytical readout such as viremia and/or virus shedding (as measured by RT-PCR and/or virus titration) should help strengthen future ASF vaccine development".
- Standardised cell lines for titration are discussed. I suggest to explain the criteria that would be used to identify such a cell line. The MA104 cell line is suggested as a possibility. However, a recent publication (Do et al., 2022) indicates an adaptation of ASFV to this cell line over several passages is required which would preclude its use. Other transformed porcine macrophage cell lines may be better suited for this purpose?
RESPONSE 5. Agree. New statements were added on the cons of such an approach and included the Kwon HI, Do Dt et al 2022 reference.
The following new statements was added to Lines 680-686:
"However, some ASF MLV vaccine candidates may need to be adapted to consistently replicate in certain continuous cell lines, which may preclude their use in a shared standard potency release assay. For example, wild-type ASFV requires several passages in MA-104 cells to consistently grow to high tires [118]. Other transformed porcine macrophage cell lines such as PIPEC, WSL, ZMAC-4 should be evaluated against the current panel of ASF MLV vaccine candidates to determine if MLV adaptation is required for stable growth and produces consistent and reproducible titers that can be measured (e.g., HAD50, TCID50, or qRT-PCR).
