Thursday, 29 May 2014

MERS-CoV did not just transmit via 2 meetings and a handshake...the retraction [UPDATED with CDC Press Release]

There must be a couple of internally relieved people around the US CDC today. Relieved. Why? Because they have got off their chest something they must have had a growing inkling about for at least a few days now. That being the news result that the Illinois 3rd US case was in fact not infected by MERS-CoV via a handshake and 2 meetings, 1 lasting 40min. 

So that didn't happen. 

And the implications for much simpler transmission of MERS-CoV did not result. And that all kinda makes more sense in the broader scheme of things MERS-related. This result always looked like an outlier.
Not a fun thing to have to report. Kudos to the team though for going back to correct an error. MERS reporting could definitely do with some more dynamic editing, and ownership, of it's mistakes.

That said, for the life of me this morning, I cannot work out why the CDC announced part of the antibody (Ab) testing result without having put such an important preliminary piece of diagnostic information, with so many epidemiological implications, through an even more rigorous testing pipeline first. I had certainly assumed that had happened when I previously wrote in support of antibody-testing on the back of this result over a week ago...and made note that that hiccups in the Ab testing process could follow! 

Let's look at what we know publicly about this test method. The CDC team have published 2 different papers [1,2] where they use an enzyme-linked immunosorbent assay (ELISA) first (the same one as used in the Illinois case I presume), then confirmed those results with either an imunofluorescence assay (IFA) or a virus neutralization (NT) test. The latter is the most specific method of showing that the MERS-CoV antibodies in a patients serum, if present, can block, or "neutralize", the ability of a virus to infect permissive cells in the lab. However, there has been a previously recorded issue with sera from SARS-CoV positive people cross-reacting in a MERS-CoV neutralization test. [3]

So the 2 papers have the following definitions relevant to antibody testing (my highlighting).

In the study of a possible MERS-CoV related stillbirth, the CDC team used the definition..

"MERS-CoV antibody positivity was defined as having positive a serologic result from the HKU5.2N Enzyme Immunoassay (EIA) and a correlated test-positive result from either the MERS-CoV Immuno-fluourescent assay (IFA) or MERS-CoV microneutralization titer assay (MNt) developed at CDC."[1]
*The HKY5.2N is a bat CoV antigenically related to MERS-CoV.

When they went back to the original Jordan cluster from 2012, the same definition was used..

"To maximize specificity, we defined MERS-CoV antibody positivity as subjects having correlated, positive laboratory results from the HKU5.2N screening ELISA as well as confirmed positive results by either the MERS-CoV immunofluorescence assay (IFA) or the MERS-CoV microneutralization assay (MNT)."[2]

What we know in the Illinois retraction story is that the IFA results did support the ELISA (less specific test) results (both were positive) before that result was announced.[4] They were clearly not supported by virus-specific MNT though. So the definitions above will need to be changed, perhaps to include all 3 results for a definitive answer or definitely have MNT in the tetsing mix somewhere. This has an impact on a result from the Jordan retrospective study[2] since 1 of those "positive" cases was defined using only ELISA and IFA-reactive without support from MNT (see Outbreak member 11; Table 1).[2]

Oh well. Just goes to show, no-one is perfect and everyone is subject to a little hysteria when the pressure is on.

I've gone back to strike-through the text relating to this retraction in my previous posts (might take me a little while to complete). I'm leaving the text in place as it was, but adding new comments in red. I've also deleted this line from my personal MERS-CoV line list. 

While this sadly incident does nothing to help people trust antibody testing in the future, at least for MERS-CoV, the literature for MERS-CoV antibody testing contains good examples of well-validated assays that require and conduct multiple tests to yield robust results. I still think rigorously determined positive antibody test results should still be considered as valid indications of a MERS-CoV positive result. Clearly not in this instance because this seems to be a pipeline "in process".

  1. Stillbirth During Infection With Middle East Respiratory Syndrome Coronavirus
  2. Hospital-associated outbreak of Middle East Respiratory Syndrome Coronavirus: A serologic, epidemiologic, and clinical description
  3. Cross-reactive antibodies in convalescent SARS patients' sera against the emerging novel human coronavirus EMC (2012) by both immunofluorescent and neutralizing antibody tests.
  4. CDC concludes Indiana MERS patient did not spread virus to Illinois business associate