July 1999 - Questions for Consideration
Questions for Consideration by the FIFRA Science Advisory Panel on the Risk Assessment of Biopesticides Containing Burkholderia cepacia or Other Microbes Related to Opportunistic Human Pathogens
Current scientific knowledge pertaining to the taxonomy, genetics, and pathogenicity of Burkholderia cepacia (Bc) isolates, prevents EPA from concluding with certainty that isolates used for biological control would not cause disease in cystic fibrosis (CF) patients.
Because of this, EPA believes that the regulatory basis for risk assessment for Bc biopesticides, provided they do not possess CF pathogenicity indicators (such as placement in genomovar III or possession of epidemic strain markers), should depend primarily on the potential for increased exposure to CF patients due to the use of these biopesticides.
In addition, if biological control strains of Bc could be determined to be non-pathogenic to CF patients, greater flexibility of the use of these biopesticides may be feasible. Therefore, it is important to determine, to the extent possible, the potential pathogenicity of these strains.
- Does the panel agree that exposure determinations should be
the key component of current risk assessment for biopesticidal
use of Bc?
- If so, what parameters should be included in the study designs,
and what factors should be considered to conclude that there is
a reasonable certainty of no harm to CF patients from exposure
to Bc based biopesticides? How much baseline data are needed to
determine typical background populations of Bc? How should the
fate of Bc be monitored after test applications, to determine
whether undue exposure to CF patients will occur?
Are available assay and strain identification methods adequate for monitoring the fate of these strains?
Can data from studies addressed in the following questions (e.g. genomovar data) be used to set acceptable levels of Bc in the soil/rhizosphere or on the crop?
- How can current knowledge of Bc taxonomy, genetics, and pathogenicity
factors of clinical isolates be used in risk assessment of biological
a) Genomovar I strains appear to have a low propensity to cause infection in CF patients, as reflected in the low proportion of CF clinical strains identified as belonging to genomovar I. What additional data would allow EPA to use genomovar analysis for determining pathogenicity of biological control strains to CF patients? Could better characterization of the clinical outcomes from strains subjected to genomovar analysis allow this process to be used more fruitfully in risk assessment?
b) Are there identified virulence factors that may be used to determine the pathogenic potential of biocontrol strains of Bc? Is there a role for animal models in the identification of virulence factors? What is the best method or combination of methods to determine whether putative virulence or pathogenicity traits are important in the infection of CF patients?
c) Concern has been expressed that even if biological control strains of Burkholderia cepacia were found to be nonpathogenic, they might become CF pathogens or contribute to the pathogenicity of CF strains. Such concerns have been largely undefined, which makes them difficult to incorporate in the risk assessment process. What specific, currently available properties or traits can EPA use to evaluate the potential for biological control strains, even if nonpathogenic, to adversely influence pathogenicity to CF patients? [Traits that do not directly effect pathogenicity, but could allow higher exposure of CF patients, e.g. improved environmental fitness, may also be relevant].
d) Are any of the current animal models adequate and sufficiently validated (that is, not likely to give false positive or false negative results), for use in testing biological control strains of Bc for CF pathogenicity?
- There are currently no specific criteria for strain markers
for biopesticides. Such markers are useful or necessary to monitor
biopesticides after registration. What might be the role of specific
genetic markers, such as RAPDs or AFLPs, that can be used to quickly
identify RAL-3 or other biological control strains related to
opportunistic pathogens? What markers might be useful, and how
should they be validated for sufficient specificity (e.g. how
many strains should be used to develop confidence in their specificity)?
- Is the CF patient:Bc, host:parasite relationship typical of
what is commonly considered as infection by an opportunistic pathogen?
Should any of the criteria or methods used to assess risk from
Bc be applied to other biocontrol organisms which are related
to opportunistic pathogens, if they are submitted for registration?
- Has EPA overlooked any scientific issues used to assess risk from Bc, that may be applied to Bc or other biocontrol organisms related to opportunistic pathogens?