Bunyaviruses and birth defects: are we due for some follow-up studies yet?
This year’s unprecedentedly large Oropouche virus epidemic marked the first known cases of human vertical transmission of the virus, causing miscarriage and birth defects. In addition to the first two reported cases, four likely cases were retrospectively identified from a microcephaly biobank, and a number of additional potential cases in Brazil remain under investigation. Other countries affected by the epidemic have not reported finding – or investigating – such incidents.
While many closely related orthobunyaviruses cause neurological birth defects and spontaneous abortion in other animals, and preliminary studies have suggested that similar issues may occur in humans, I wasn’t aware of any prior work on this for Oropouche virus, so I was surprised that the July PAHO Epi Alert of the incident referenced a prior study:
In 1982, IEC (Instituto Evandro Chagas) professionals together with the Institute of Tropical Medicine of Manaus and the University of Manaus, Brazil reported the occurrence of nine cases of OROV infection in pregnant women. Among these cases, two, who were in the second month of gestation, resulted in spontaneous abortion. Although the technique used for diagnosis in pregnant women was a serological test (due to the unavailability of molecular tests at the time), this finding is suggestive of vertical transmission, and was recorded as part of the characterization of the first outbreak of Oropouche in the state of Amazonas, Brazil between 1980 and 1981.
After a bit of digging, I found the source cited: a 1982 study by Borborema et al. In its discussion of patient symptom profiles, it does, in fact, note:
Given that Oropouche virus epidemics have historically affected relatively remote areas with limited access to real-time testing, and resolved at the end of that year’s rainy season, the surveillance tools and follow-up appointments needed to actively monitor for congenital infections – much less complications down the line – haven’t consistently been available in most transmission foci. An additional factor likely at play: the reassortant virus driving the ongoing epidemic does seem to replicate much more prolifically than the prototype in mammalian cells, including in a glial, or neural immune, cell line (U-251, for any other nosy pipette jockeys).
This might, unfortunately, parallel the story of Zika’s association with microcephaly: now suspected to have flown under the radar historically, but also to have been exacerbated by mutations that arose shortly before it emerged in Brazil. It’s also unlikely that this year’s handful of Oropouche vertical transmissions would have been detected, especially so early, without the microcephaly surveillance biobank program the IEC developed after the Zika epidemic.
So, if we want to avoid being caught off guard like this: how hard is it to anticipate ‘new’ harmful congenital infections?
Thinking about Zika always reminds me of Cache Valley virus. I was a freshman in college, working in an arbovirus surveillance lab, when Zika virus was proven to cause human birth defects. At the time, I was starting my own research project on Cache Valley orthobunyavirus, which causes a similar enough set of birth defects in sheep that, during the spike in Cache Valley cases that year, a Toronto Star article compared the two.
It was from this article that I learned about a study, published the year before I was born and never really followed up on, that maternal Cache Valley virus antibodies may be associated with macrocephaly in human infants. That study’s author was interviewed for the story:
“What surprises me is that nobody has followed this up,” said lead author and arbovirologist Charles Calisher, who is now retired. “Cache Valley is a representative of a problem that we refuse to face.”
We haven’t gotten much further on the subject in the near decade since. This is particularly troubling given the ubiquity of Cache Valley (and related orthobunyavirus) infections in humans, which we don’t really test for but assume are mostly mildly or non-symptomatic: serological studies of different groups find anywhere from 2 to 50% of individuals tested have been infected with it, but fewer than a dozen human disease cases have ever been reported. (Another useful review.)
Other evidence, with varying degrees of support, exists linking vertical transmission of various bunyaviruses to adverse human birth outcomes. Of course, primates, including humans, have differently structured placentas than many of the animals that commonly suffer congenital defects due to bunyavirus infections, in ways that should make it trickier for viruses to cross them. This unfortunately means that few avenues for proactive study of potential causal pathogens are available, unless the problem one works on is well-funded enough to support long-term non-human primate studies for each pathogen of interest. (If you have also worked on bunyaviruses for any length of time, you likely find this as funny as I do.)
Left to penniless daydreaming about observational studies of rare outcomes of rare infections, one might consider the IEC model, which is quite useful when screening for a handful of specific pathogens that might cause one outcome of interest, or perhaps long-term chart review of a given health system’s patients, which could take decades to accumulate the statistical power we’d need to support causality. I somehow always find myself imagining a new and improved version of the NIH congenital rubella syndrome serum bank that Charlie Calisher used for his correlation-based study: a massive dataset of systematic maternal serosurveillance and birth outcomes, ideally with multiple data points to track maternal seroconversion, and matched newborn cord serum or tissue from miscarriages. 1
(Even with that information, considering the trends in the U.S.’s maternal mortality rate, it’s fair to ask: at a societal level, would we do anything with it? I’m not sure. Given, however, that we currently believe there are no vector-borne teratogenic viruses in the U.S., and some potential ones are pretty ubiquitous, it’d probably be useful for expecting parents to know if they should take extra precautions to avoid mosquito, tick, midge, or sandfly bites in their own backyards. But I digress.)
Where does this leave us with Oropouche virus? I worry that this epidemic might only be getting started: it’s been introduced to new regions of Brazil and other South American countries this year, and may not have fully ceased transmission this dry season. It’s also gained quite a foothold in the Caribbean, if the 10 new travel cases from Cuba reported by the Florida Department of Health, the week after the CDC reported 21 travel cases from Cuba so far in 2024, are anything to go by. As we approach fall and winter in the northern hemisphere, we’re coming up on the southern hemisphere’s spring, summer, and rainy season(s), when Oropouche transmission has historically peaked. We know that humans are sufficient vertebrate amplifiers to spread and perpetuate epidemics; the biting midge Culicoides paraensis, the virus’s primary known vector, is already found throughout mainland North, Central, and South America, so the range of potential spread could be quite large. 2
So, if this virus consistently causes miscarriage or birth defects, we are almost certainly going to see more cases in the fairly near future. We might want to implement more surveillance systems outside of Brazil before we miss them. (We also don’t have great vector control tools for midges yet, so that’d be an important intervention to start strategizing.)
We should at least be willing to finally face this problem. I’d like to be ready to do something about what we find when we do.
1 It’d be an expensive undertaking, though it’d be more affordable than ever with massively multiplexable serosurvey tools like Luminex or VirScan.
2 Culicoides paraensis is curiously thought to be absent from most of the Caribbean – the question of which vectors are currently driving Oropouche transmission is an open one, and was the post I was assembling the research to write when I came across the material for this one.
If you want first dibs what I find in the stacks, subscribe for future updates.
Feedback is welcome, but enabling direct newsletter comments costs the author money. As this author is a graduate student, she can be reached for free via email at iott@unc.edu, or on BlueSky as @isabelott.bsky.social.
(If you hit an article paywall, I’m always happy to share PDFs – just send me an email.)