Science Corner: Climbing the Fruit Tree of Science

In other universe, the biggest science story of the past month would be the discovery of yet another hint at the possibility of life on Mars. Yes, we’ve been down that road (and parallel paths for other celestial bodies) before, without definitive results. And this latest find is still far from first contact; we have observed chemical patterns that are associated with life on Earth, but which technically could result from abiotic processes as well. It’s another bread crumb, an encouraging sign that we should keep looking–and keep thinking about what to look for, and what to do when we find it. But like I said, while we could be talking about one of life’s biggest open questions, instead we find ourselves repeating one of life’s most tedious: Seriously?

The actual biggest science story of the past month, at least based on discussion generated, is almost certainly the US Health & Human Services (HHS) announcement regarding autism and acetaminophen. You may recall that Robert Kennedy, Jr. declared in April that “By September, we will know what has caused the autism epidemic and we’ll be able to eliminate those exposures,” a goal later softened to “some of the answers.” It sounded at the time as if this would involve a large-scale research campaign, an ambitious plan to say the least. Five months is hardly enough time to recruit study participants in the necessary numbers, let alone actually conduct research and prepare findings. So it is not much of a surprise that when September rolled around, what we got was not new findings but commentary and new decisions based on existing literature.

And actually, what we got–and your perception of it–likely varies considerably depending on whether you go by the press conference or the written documents. Let’s just go ahead and stipulate that no one should ever take personal medical advice from this or any president of the United States. (Even in the unlikely scenario that a future president is a clinician who previously provided you medical care, I’ll go ahead and say they should turn over your care to other actively practicing clinicians so they can focus on leading the country.) I don’t blame anyone for paying attention to the press conference; we live in a video-dominated age, and historically it has been reasonable to treat such public statements as fairly reliable representations of the topic at hand. In this case, the gap is wide, and so it is unfortunate that so much attention was paid to the embellished version. Let’s instead focus on the written documentation.

Specifically, I am referring to the HHS press release and accompanying fact sheet. There are three main points; we’ll tackle them in reverse of the order they appear, so we can start with the easiest one. The pre-existing Autism Data Science Initiative, announced back in May, is highlighted. By now, the funded projects have been chosen and presumably are ramping up. Based on a scan through the available abstracts, this seems like NIH business as usual. To get this kind of research funding, you typically have to already have some relevant results or work already in progress, and it reads as though that is the case here as well. And the timeline suggests decisions were made prior to recent announced changes in how funding would be reviewed. I don’t know what projects were rejected; there is a strong theme of environmental exposures across many of the studies, which aligns with the current HHS secretary’s philosophy on health. At the same time, the context is often gene-environment interactions, which is consistent with already established results on the role of genetic factors. So while there funded projects may be been selected with a particular slant, they don’t appear to be outside of the mainstream of current research. And of course I don’t know what will be done with the eventual findings of this work. But for now I think it is reasonable to believe the grantees are working in good faith and that their results will generate positive contributions to our understanding of autism.

There is also discussion of the potential association between chronic acetaminophen use during pregnancy, particularly late, and autism in the child born from that pregnancy. Both documents mention the conflicting data; some primary studies and meta-analyses (pooling data from multiple published studies) find evidence of an association and others do not. Findings of association are different from proving causality, as noted. As many have pointed out, chronic acetaminophen use during pregnancy may be an indicator of other challenges for the pregnancy or the child that could have a causal role rather than the acetaminophen itself. This could also explain why the association isn’t always observed; in some cases, that other causal factor may not be treated with acetaminophen. Or the inconsistency may be due to the fact that over-the-counter medication use is typically self-reported long after the fact in these studies. These are not randomized trials where people are assigned to get acetaminophen or not and its administration handled in a controlled setting and recorded. It is plausible that not everyone remembers accurately what they took, at what dose, and how often. (Anecdotally, not all young people of child-bearing age pay full attention to the differences between over-the-counter fever reducers.) All of those complications are likely making it harder to identify whatever true signal may or not may be involved. Nevertheless, it seems clear that if there is any causal link at all–which again is still undecided–at most chronic acetaminophen is one of many risk factors and represents a small modification to the overall likelihood of autism. (If there were a strong link, it would be easier to detect.)

What isn’t present in either document is an explanation for why this particular possible association is being highlighted now. The weight of the evidence hasn’t changed; nothing new has been published to suggest the answer is any clearer than it was a year or ten ago. And even though the preamble highlights the intent to explain recent increases in autism diagnosis rates, no evidence is provided to suggest that chronic acetaminophen use during pregnancy has changed in that time period. It has been available over-the-counter since 1960, and the evidence I could find indicates its use has been declining. The bottom line of consulting a physician about acetaminophen use during pregnancy is reasonable enough, but it was already the recommended approach. It is not clear why anything additional needed to be said now in any capacity, let alone in the hyperbolic terms used in the press conference. And let’s not forget that there is already a history of unfairly and inaccurately blaming women for their children’s autism, which should prompt more care, not less, when discussing any potential maternal factors.

That leaves us with the topic of leucovorin. It is a little harder to understand what is happening there just from the HHS materials. Leucovorin is a drug that has been around for decades and is FDA-approved to treat the side effects of certain cancer drugs. It had been approved for that use under the brand name Wellcovorin, sold by GlaxoSmithKline (GSK). When generics became available, GSK stopped using that brand name and withdrew the approval application since it was no longer necessary. Nothing changed in our understanding of the safety of the drug, or its efficacy for mitigating chemotherapy side effects; there just was no longer any value in the brand name.

What has changed now is a little murkier. The press release mentions a label update for Wellcovorin, indicating approval for treatment of cerebral folate deficiency (CFD), a rare condition whose symptoms can overlap those of autism spectrum disorder (ASD); further, some people can have both diagnoses. The press release goes on to suggest that Wellcovorin will also be approved for treating ASD, and indicates new clinical trials will be initiated by the NIH. The fact sheet also mentions these clinical trials, indicating they will investigate the potential of leucovorin for treating ASD whether or not there is a CFD diagnosis, but only regarding a label change only indicates that an update will be sought from GSK regarding CFD. Looking at the Federal Register on the subject, it reads as though so far all that has happened is that the FDA is (re)approving the withdrawn application for approval of Wellcovorin to treat chemotherapy side effects. They are also requesting GSK to submit a supplement to that application with the necessary materials to go through the approval process for labeling it for CFD. That would include available safety and efficacy data, and presumably also details on appropriate doses for prescribing.

While it is also unclear what happened to prompt these steps (whichever ones are actually happening), it is not entirely out of the blue either. Leucovorin is a form of folic acid which can cross the blood-brain barrier, so there is a plausible mechanism for how it could address a cerebral folate deficiency. There has been research on its use as a CFD treatment and an ASD treatment; the HHS materials reference this recent review of that literature. And some physicians have already been prescribing it off-label for both in some cases. So it may be that these developments would have happened anyway but are now being accelerated.

Or then again, maybe not, depending on exactly what is happening. If GSK submits a supplement and it goes through the usual review process and is judged to pass the relevant standard criteria and gets approved for CFD, that seems unobjectionable. If the labeling is changed for treating CFD and/or ASD before that process, as the press release suggests, that would be a notable departure from the usual process and understandably raise concern. Time will tell, although one will likely need to pay close attention, and to more than just announcements and media coverage. In the interim, it is always good practice to discuss any questions you might have about medication with your physician.

This is a longer than usual post; if you stuck with me so far, thank you. Given how much confusion has been generated on these topics, I thought it was worth getting into the details. If you wish this were simpler, I understand; I wish I didn’t have to write this much. And if it feels like medicine and health used to be simpler, you are probably right. There are many illnesses where we can say it is caused by X, we have drugs Y and Z to treat it, and then you are well. A condition possibly caused by hundreds of genes interacting with dozens of chemical and behavioral factors to produce a wide spectrum of presentations is much harder to understand. But that’s not because anybody wants it to be complicated. It’s because the simpler conditions were easier to identify and treat. At some point, all we will have left is the complicated stuff higher up in the proverbial fruit tree. It may be tempting to try to make do with what has fallen onto the ground, but to get to the good stuff we’ve got to climb.