Lancet Medical Journal Correspondence re: Wastewater Data
Commentary from Italian researchers highlight both the potential and shortfalls found in utilizing wastewater sampling for early detection of high consequence pathogens
I came across this link this morning from the Lancet sharing correspondence from Italian disease surveillance experts (link) regarding their analyses of the U.S.-based “WastewaterSCAN, a monitoring system developed by Stanford University in collaboration with Emory University that analyses solid wastewater samples from municipal treatment plants in the USA for monitoring H5 avian influenza in wastewater (among other pathogens)”.
I’ve previously written blogs covering wastewater surveillance on many occasions, and I keep a regular watch on both the CDC wastewater site (which monitors matrix gene levels in wastewater (NOT H5), as well as the WastewaterSCAN site referenced in the commentary. They monitor different sites and states in many cases; neither one is totally comprehensive across the entire United States.
Here is the latest “2-week maximum” results from the CDC site, with all sites shown on the map. While it shows total potential coverage (and areas NOT covered), the low values and non-reporting sites can overwhelm looking at the aberrantly higher value sites:
Fortunately, we can click off percentile groups below the map on the website to eliminate them form view. Here are the same latest bi-weekly maximum results from the CDC site, with the non-reporting and less than moderate Matrix test results turned off:
This map shows collection sites with moderate and higher than average influenza Matrix gene PCR values across the U.S. What do these results mean? The short answer, as the commentary states, is that we don’t know! We can surmise that it is not likely to be human seasonal influenza A in mid-summer; however, local public health records should be checked.
While the WastewaterSCAN site is theoretically more specific for H5, it all comes down to which states and sites are actually tested and reported, which is a bit “hit and miss”. I also sense a bit of misperception on the part of many in public health regarding blips on either system, with a bias towards wanting to blame “hits” on some undiagnosed human H5N1 outbreak.
Regardless, it seems to me that the first prudent step is to sequence the samples with stronger matrix or H5 gene results! Any hemagglutinin gene segments generated through WGS, whether whole or even a substantial partial sequence would provide crucial evidence regarding the source of the virus. Diagnosing these to me is a lot like any other diagnostic process - look for horses, not zebras! The 2 big sources of H5 right now in the U.S. are avian wild bird (farmed poultry die and get reported) and dairy - farms in general do not drain into municipal sewage; however, shipped milk is hot with virus and leaves the farm every day in trucks for processing. So, until ruled out, differentials for high sewage summer influenza levels in the U.S. currently are: 1) drainage from dairy processing plants, bulk truck washes, milk product discard, etc.; and wild avian drainage into sewage stream, especially if there are associated avian mortality events. This assumes that the sample sequences out as a high pathogenic H5; if it sequences as a low path influenza, then back to the birds or other species.
If it sequences out as a likely high path H5 2.3.4.4b B3.13 clade, then officials need to further investigate dairy versus wildlife contaminants as the cause of the higher levels. Regardless of the live dairy herd “state status”, the sewer-shed should be analyzed for dairy production, processing, or washout inflow, given the current U.S. outbreak of unknown scope. Remember that processing facilities do not necessarily collect all milk from their state of residence, so a positive processor does not equal a positive dairy herd state. Similarly, wildlife, wet markets, or small abattoir inflows are additional possible sources to be checked.
The sequences generated should be useful in monitoring for genomic changes of concern in the clade. Whether the virus source can be identified due to lack of site specificity or confidentiality concerns are secondary to ongoing monitoring for viral changes and its resulting pathogenic and zoonotic potential.
That brings me to my final concern with both the article and the current lack of public health response to ongoing signals in both influenza wastewater surveillance streams as well as in retail milk (another related subject). The commentary referenced in this post aspires to open collaborative real-time surveillance involving farm samples. We just aren’t there yet! However, we DO have sewage and retail milk, albeit without geographical precision at a level lower than the sewer-shed or processing plant. Even those could be masked or aggregated if necessary to allow stakeholder- unchallenged testing. However, we must not just avoid sequencing because stakeholders don’t want possible unwanted exposure from positive results.
This virus is going nowhere; in fact, it seems to be spreading or reappearing in old haunts (like Michigan, South Dakota, Texas). Fall is coming with its associated cooler weather “flu season”, cattle movements and weaning, cattle off grass and moving south for warmer climate. Birds will also be migrating south. When cattle and birds move, H5N1 will move on to susceptible new locations. Do we even know (or care) where it is??
Get ready…
John