© anika stock.adobe.com
A recent appellate court ruling cancelling the US EPA’s TSCA methylene chloride (MC) rule is a major blow for the EPA, with potentially far-reaching consequences for chemical risk assessment at the agency.
In its 15 September decision, the US Court of Appeals for the Fifth Circuit found that the agency made errors of law and lacked substantial evidence to support its rule, while subjecting the technical underpinnings of the agency's approach to unusually close scrutiny.
The court's rejections of the 'whole chemical' approach to risk determinations and the default assumption that workers do not wear personal protective equipment (PPE), in particular, may radically expand the workload of the agency, if the EPA determines it must revise completed evaluations. The agency, already under-resourced and failing to meet statutory deadlines, would be ill-equipped to absorb that additional work.
However, it is the court's criticisms of the agency’s scientific methods that may prove more consequential in the long term. The court questioned practices that most people involved in chemical risk assessment would consider routine, such as deriving a toxicological point of departure (POD) from a no observed adverse effect level (NOAEL).
The problem facing the agency in this regard is not simply additional work. Any replacement methods would need to satisfy both the court's reading of the law and the standards of the scientific community, and it is not yet clear that such methods exist.
Systematic review
One area the EPA may need to address is how it weighs and integrates evidence, after the court found it failed to make sufficient use of relevant studies when deriving a chronic exposure limit.
In its 2020 risk evaluation for MC, the agency conducted a systematic review, identifying relevant studies and evaluating their quality.
The many studies relevant to chronic exposure included a 1988 study in rats by Nitschke et al, which the agency rated high for quality, and three epidemiological studies in humans rated medium. The Nitschke study showed liver effects not seen in the human studies, and the agency selected this as the data source for the chronic exposure limit.
This approach was justified, the EPA said, because the Nitschke study was rated high for data quality and had the lowest exposure concentrations.
The court emphatically disagreed.
The agency acted irrationally by ignoring "evidence contrary to its presuppositions", the court said. "Even an ideal rat study could hardly justify the EPA entirely discarding three ‘acceptable’ human epidemiological studies," it said.
The court did not reject the selection of the Nitschke study as the most suitable data source, but rather the strategy of taking the data exclusively from a single study when other relevant studies were available.
Replacing methods
Setting aside questions about whether the court fairly characterised the agency’s process or whether a federal appeals court is the best place for a discussion about the right way to carry out systematic reviews, the EPA now must grapple with how to move ahead.
There may be existing methods that involve combining data from multiple studies, and the agency might use these, or it might develop new methods.
Unfortunately for the agency, there is considerable uncertainty about the viability of any future methods.
The agency does not yet know whether alternative approaches can reliably perform the same function as the rejected method, which has long been a standard part of chemical risk assessments. Shifting to a little-used or entirely new method could have unforeseen limitations when deployed in evaluations, such as yielding spurious results outside a narrow range of conditions, or failing because of scarce data.
And even if an alternative can perform the same function, it may fail a future legal test. Whether a future systematic review methodology will hold up in court is hard to predict because the Fifth Circuit generally did not indicate what the agency should have done, only what the agency did wrong, and other appellate courts could offer differing views.
Scientific standards
As well as complying with the law, any methods the agency uses must align with the norms and expectations of the scientific community, embodied for the purposes of TSCA by the Science Advisory Committee on Chemicals (SACC).
A method may function adequately as part of the assessment and comply with the law, yet still be rejected by the SACC. What happens in this scenario is an open question. TSCA requires the EPA to use the "best available science", and it seems unlikely that using a method explicitly rejected by the SACC would be consistent with that requirement.
Yet the Fifth Circuit’s ruling on MC does clash with the corresponding SACC report on certain scientific issues, suggesting that the committee is not the sole arbiter of sound science.
Valuations, NOAELs
Further scientific issues raised by the court compound the challenge for the EPA.
In its ruling, the court suggested a temporary 7% decrease in peripheral vision is only a minor impairment and should not be considered the basis for an unreasonable risk finding in the context of the acute exposure limit.
It also faulted the agency for using a NOAEL as the basis for its risk calculation and applying an uncertainty factor of 10 to account for intraspecies variability.
These three actions, taken together, were incompatible with a previous court ruling stating the role of the EPA under TSCA is not to reduce the risk to zero, the court said.
But here again, the agency may struggle to find alternative methods for these practices, which could be considered ‘normal’ within the context of chemical risk assessment.
Assessments usually exclude value judgements about what adverse effects are minor, given that such judgements make more sense as part of an impact assessment to determine the effectiveness of measures to bring down the risk. However, the court ruling implies that the agency needs to determine whether an adverse effect is minor or not as part of the risk calculation.
Similarly, chemical risk assessments often use NOAELs as the basis for risk calculations. It may not be the preferred method if more advanced methods, such as benchmark dose modelling, are available. But when such methods are not available, as is often the case, using NOAELs is generally considered viable. The court’s ruling seems to disregard this option.
Path forward
Looking at the scientific issues raised in the decision, the court’s rejection of the whole chemical approach to risk determinations seems almost trivial. It implies extra work, yes, but there is not the same uncertainty about the way forward, not least because the agency already has experience of a use-by-use approach to risk determinations for TSCA evaluations under the first Trump administration.
Something similar can be said about the court’s rejection of the default assumption that workers do not wear PPE. The agency already knows how to factor PPE into exposure assessments and has collected information about PPE use for many chemicals it has reviewed.
The key difference when considering systematic review for TSCA is the location of this topic at the intersection of science and law.
Ultimately, the EPA can rewrite a rule with different policy choices. Replacing scientific methods that are deeply embedded in modern risk assessment is a far harder challenge.