NIH starts serological survey testing anti-SARS-CoV-2 among the US population 2

Using both traditional blood draw and an innovative home-collection device, the NIH recruits 1000 healthy volunteers nationwide

The SARS-CoV-2 virus (otherwise known as coronavirus, the causative virus of COVID-19 disease currently rampaging worldwide) is a difficult pathogen to battle. While clearing out my bursting-at-the-seams ‘downloads’ folder on my computer this morning, I came across two documents from the United States Government that indicate what a difficult problem we face.

Coronavirus SARS-CoV-2 is not Influenza

The first one goes back to 2006, from the Bush White House site, titled “National Strategy for Pandemic Influenza”; the 233-page PDF can be downloaded here. A second document that I stumbled across was from the CDC, called “Interim pre-pandemic planning guidance – community strategy for pandemic influenza mitigation in the United States”, with the subtitle that may sound familiar: “Early, Targeted, Layered Us of Nonpharmaceutical Interventions”. The 97-page PDF is available here.

I bring this up here to bring home the point that SARS-CoV-2 is not the influenza virus, and has affected so many aspects of our response to it. The virus has been sequenced in record time and molecular reverse-transcription real-time polymerase chain reaction tests (rRT-PCR tests) have been quickly developed (fumbling by the CDC and FDA notwithstanding). Testing has ramped up to over 2.5M tests as of this writing in the US, and public health laboratories, major industrial testing laboratories, and many vendors of different types of rRT-PCR testing reagents and instrumentation are approved under Emergency Use Authorization (EUA, the updated FDA EUA page for coronavirus tests is located here).

Is molecular sensitivity the problem?

The IFUs located at the FDA page give a lot of detailed information – for example an LDT such as LabCorp documents lack of cross-reactivity with closely-related viruses, and sensitivity levels down to 6.5 molecules of viral RNA per mL.

Yet stories would crop up like this one: “Questions about accuracy of coronavirus tests sow worry” in the WSJ, and “If you have coronavirus symptoms, assume you have the illness, even if you test negative” in the New York Times.

Digging a little further, you find out that the coronavirus is not like the flu. It is not due to faulty swabbing, not a problem with the preservative solution, nor any of the molecular biology of reverse transcription or real-time PCR amplification. It has to do with the natural biology of the disease. Now up to this point whenever I’d read a report of negative results from individuals with clear cases of COVID-19, I assumed it was technical error.

A clinical study provides the needed data

A research letter dated 11 Mar 2020 in JAMA titled “Detection of SARS-CoV-2 in different types of clinical specimens” examined eight different types of specimens (totalling 1070 in all) from 205 patients with COVID-19 disease.

The results were somewhat shocking – of the swabs, only 32% tested positive, with the limit of detection set at 40 cycles (for those unfamiliar with rRT-PCR values, a Ct of 40 is very, very high, which means a very low amount of virus present). Think about it – these were hospitalized COVID-19 patients with severe disease, and 68% of them test negative with a molecular PCR test. Looking at the other sample types, the best positive results were from Bronchoalveolar lavage fluid at 93%, however the n=15 is small (from what I have heard a lung lavage is sounds as unpleasant as it is sounds). Here’s the table for you.

The paper doesn’t give discrete copy numbers or a standard curve to measure Ct against, but the range of Ct values for pharyngeal swabs is 20.8 to 38.6. In other words, among the 32% positive by rRT-PCR, the measured amount of viral copies in the throat varied from single copy (a Ct 38.6 may be considered untrustworthy though by some) up to 2^(38.6-20.8) = 2^17.8 = 2.6 x 10^5 copies.

So you have 398 patients with COVID-19, and 272 of them test negative for SARS-CoV-2. Of the 126 left, the amount of virus ranges from single-digits to 100’s of thousands. This is illustrated in the figure reproduced from the Wang et al. JAMA paper.

Thus the collection and laboratory work can be done to the highest standards, but the biology of the disease (especially with the importance of the immune response, and the evident respiratory ailments directly related to the individual immune response to infection) dictates the effectiveness of the molecular test.

The first serology (serosurvey) by the NIH

It was only yesterday that I wrote about the interesting preliminary results from 500 of the intended 1000 individuals from the town of Gangelt in Germany. (Their announcement in German is here.) And yesterday afternoon, the National Institutes of Health announced a 1000-person serological study of the healthy population of their own.

This study is the first serology study undertaken by NIAID (National Institute of Allergy and Infectious Disease) and the NIBAB (National Institute of Biomedical Imaging and Bioengineering); their announcement is here. They will analyze both IgG and IgM by ELISA, and the antibody they are using for detection was developed by NIAID and NIBAB. (For those unfamiliar, a serology antibody test can test for the presence of anti-SARS-CoV-2 antibodies using… another antibody.)

A unique home-collection device for blood

What is exciting and unusual however, is the NIH plans to enroll a subset from around the country using a home-collection method called Mitra by a company called Neoteryx.

It is not clear what the regulatory status is of this device; it is likely an EUA will be issued in the gap between recruiting (apparently) healthy volunteers and the shipment of these devices. You may remember the difficulty Theranos had in their characterization of their nanocollection devices, which were clearly medical devices, running afoul of the FDA.

Having readily-available technology for sampling blood to be analyzed at a central laboratory will open up the possibility of scaling in a massive way, which will be needed quickly to understand the scope of asymptomatic and pre-symptomatic spread across the United States.

In addition the ability to ‘certify’ individuals as virus-free, having protective antibodies for at least six months, will be a great aid in getting the economy started again. This kind of ‘serosurvey’ is vitally needed; the NIH announced they are looking for 1,000 volunteers, to sign up you only need to email clinicalstudiesunit@nih.gov per their press release, and the clinical trial information can be found here at clinicaltrials.com.

Additional data-collection underway

Stanford University has already collected 2500 samples from healthy volunteers (and about 500 from their children) in the Santa Clara area, according to lead investigator Jay Bhattacharya from Stanford School of Medicine. They expect to release their results soon.

In Los Angeles, a company called Elevated Health has started fee-based antibody testing last week at Westminster Mall. A sister told me the wait time for drive-up testing was 4 to 6 hours.

If you have read this far, I’d like to let everyone know it is time for me to find my next challenge. If you are in an organization that could use my skills in marketing, business development, key-account sales, technical writing and content marketing, do reach out to me. And do feel free to let others know who may be looking for someone. (Alas I’m not able to relocate from Maryland, so it would need to be a remote position.) My email is dale at yuzuki dot org. Thanks!