top of page
  • Donald J. Pagel

Testing for covid-19


By now, most people have heard that testing for infection by the SARS-CoV-2 virus is an important step in controlling the spread of the Covid-19 pandemic and in determining how quickly social distancing rules can be relaxed before there is a vaccine for the virus (or the virus recedes for some other reason). In view of this, it’s useful to understand how such tests work, and how complex the underlying technology is. Understanding the technology also helps answer questions like: Why did other countries have more widespread testing before the U.S? What went wrong with the original CDC test that caused the delay in U.S. testing? And why does it take hours or days to get the results from some Covid-19 tests? Here’s a brief attempt to answer these questions.

Most of the assays (tests) for the 2019-nCoV virus (SARS-CoV-2) are a standard (but complex) type of test known as a polymerase chain reaction (PCR) assay. This general type of assay has been used for years in laboratories that perform medical research or diagnostics, and most of the components and equipment for such assays are commercially available from multiple sources. That’s why countries around the world, like Korea, Germany and China, were able to quickly develop assays for SARS-CoV-2.

The biggest hurdles in developing a PCR assay for a specific disease are that the pathogen that causes the disease has to be identified, the genome of the pathogen needs to be sequenced, and a few of the reagents (chemicals) used in the assay need to be custom formulated to bind to a part of the pathogen’s genome (RNA or DNA). In the case of Covid-19, scientists identified the pathogen and quickly sequenced and published the genome for the RNA of SARS-CoV-2 (published by Chinese scientists and other researchers in the January 10-12, 2020 period, see timeline here). After that, many biomedical research labs around the world could (and did) develop experimental PCR assays for the disease using commercially available technology and a some skillful lab work to customize the critical reagents (for example, German scientists developed an assay on about January 23, 2020 that was distributed by the World Health Organization (WHO), as explained here). Of course, government approval is usually needed before an assay can be used in a country for medical purposes.

In the U.S., the Centers for Disease Control and Prevention (CDC) developed an assay for SARS-CoV-2 and received an emergency use authorization (EUA) from the FDA on February 4, 2020 (this EUA was amended on March 30, 2020), according to the FDA's EUA webpage. According to the user manual released by the CDC on March 30, 2020, the CDC test is a real-time reverse transcriptase polymerase chain reaction (rRT-PCR) assay. The CDC supplies users with two boxes containing the customized reagents needed to run the assay for SARS-CoV-2, and the rest of the reagents and equipment needed for the assay are acquired by the user from commercial sources.

The first box contains three vials, two of which contain a primer/probe mix for the SARS-CoV-2. These first two vials (called N1 and N2 in the CDC kit) are the custom designed reagents that are specific to the Covid-19 assay. The third vial contains a Human RNase P Forward primer/probe mix (RP) which is used to determine if the patient sample was collected and processed properly. The second box contains a control substance called 2019-nCoV Positive Control (nCoVPC) that was designed to yield a positive result in the 2019-nCoV Real-Time RT-PCR Diagnostic Panel. A negative result means that something went wrong with the assay.

Understanding the basic way that rRT-PCR assays work helps in understanding what the customized reagents are for and what can go wrong with an assay. In a preliminary step, a sample is collected from a patient in the form of a nasal or throat swab. Then in the lab, the patient sample is subjected to standard laboratory techniques that separate and purify RNA that comes from any viruses that are present in the patient sample. After this stage, there would only be a small amount of RNA present, if any, and you wouldn’t know what type of microorganism it came from. So, this is where the rRT-PCR part of the assay comes into play.

To get the rRT-PCR process started, an enzyme called reverse transcriptase (commercially available as a reagent) is used to convert any RNA that is present in the sample into cDNA (cDNA is DNA formed from an RNA template). This is done because viruses, such as the SARS-CoV-2, contain RNA genetic material, but the PCR reaction only works with DNA (or cDNA).

Next, the primer/probe mixes (N1 and N2) supplied by the CDC come into play. The primer comprises a pair of customize designed oligonucleotides that bind to the ends of a cDNA sequence that could only come from SARS-CoV-2 RNA and that is to be replicated in the PCR process. The probe is an oligonucleotide that was custom designed to bind to a specific section of the cDNA between the two primers in a set. The probe includes a fluorescent molecule and a quencher molecule. The quencher molecule prevents the fluorescent molecule from emitting light while it’s near the quencher molecule. Later, the fluorescent molecule is separated from the quencher, so it emits light that indicates the SARS-CoV-2 was present in the patient sample. In the CDC assay, two of these primer/probe sets are used to add certainty to the identification of SARS-CoV-2 because each primer/probe set binds to a different section of the cDNA.

Once the primers and probes have bonded to the cDNA, the PCR process can begin. A commercially available PCR machine (an Applied Biosystems 7500 Fast Dx Real-Time PCR System is used in the CDC assay) subjects the cDNA sample to a series of thermal cycles in which the temperature of the cDNA sample is raised and then lowered according to a predetermined protocol. Generally, each thermal cycle includes several stages run at different temperatures, with the specific nature of the stages being dependent on the particular assay. However, regardless of the specific thermal cycling process used, PCR assays usually duplicate the cDNA using a general mechanism.

For example, during a high temperature denaturation stage the two DNA strands separate to form two single DNA strands. During a lower temperature annealing stage, a primer molecule binds to a complimentary region of each single DNA strand, the probe binds to a different region of each single DNA strand, and the polymerase binds to the DNA/primer complex to begin DNA synthesis. Finally, during an intermediate temperature extension stage, the polymerase causes double stranded DNA synthesis to occur by forming a complementary DNA strand to each of the single DNA strands.

In the CDC assay, during the extension stage, as the new double stranded DNA is formed, the probe is degraded by a Taq polymerase causing the fluorescent molecule part of the probe to break away from the DNA strand it was attached to and begin emitting light. This fluorescence intensity is monitored by the Applied Biosystems PCR System. This process continues for a predetermined number of thermal cycles (e.g., forty-five thermal cycles) during which time millions or billions of identical copies of the cDNA are produced, and more fluorescent molecule is released each time a copy of cDNA is produced. If the measured fluorescent light from the molecule exceeds a predetermined threshold during this period, the sample is considered positive for Covid-19, provided that the controls that are run with the patient samples don’t void the assay because of anomalous assay results.

This brings up the question of what went wrong with the original CDC Covid-19 test kits released on about February 5, 2020. Many labs using these original assays reported that the kits generated inconsistent results, causing crucial testing to slow to a crawl for about a month. Various news reports have simply described the problem with the original CDC assay as being caused by contaminated reagents. However, a more specific explanation is found by looking at the letter from the FDA to the CDC (dated March 15, 2020) which authorized changes to the CDC’s Emergency Use Authorization for the Covid-19 test kits. Footnote two in the letter states that the “N3” vial of primer and probe is being deleted from the original kits. This implies that the original kits contained a faulty (or contaminated) primer/probe reagent. Since the original kits contained three different primer/probe reagents (N1, N2 and N3), packaged in three separate vials, simply deleting the problematic N3 reagent vial was an easy fix for the problem because it still left two functioning primer/probe reagents.

So, what was in the defective vial? Looking at the user manual for the original assay (dated February 4, 2020), shows that the N1 and N2 vials contained primer/probe sets for the specific detection of Covid-2019, while the N3 vial contained a primer/probe set for the detection of SARS-like coronaviruses. Hence, deleting the N3 vial didn’t weaken the detection of Covid-19, it just eliminated the ability to detect other coronaviruses (i.e. non-Covid-19 coronaviruses) in the patient sample.

It doesn’t appear that this fix required any modification of the original test (other than changing the instructions to reflect the deletion of the N3 vial), so it can be assumed that the N1 and N2 vials in the original assay were the same reagents as in the final assay. The fix did slow down the ability to test widely for Covid-19 in the U.S. by at least a month. However, by February 29, 2020, the FDA had begun allowing other labs (like Quest Diagnostics and LabCorp) to begin testing for Covid-19 in the U.S., and to begin developing their own tests, thereby mitigating the delay caused by the CDC’s defective assay, as explained in an article from Wired. By April 23, 2020, the FDA had granted about forty EAUs for companies to market RT-PCR assays and/or reagent kits, according to the FDA's EAU webpage.

A problem with rRT-PCR assays is that they take a while to run. In the CDC assay, the thermal cycling stage of the assay takes about forty minutes by itself, and it is estimated that the whole assay takes about four to six hours to run when you include all of the instrument set-up and sample and reagent preparation times. This explains why it sometimes takes a few days to get the results back from an rRT-PCR assay. On the plus side, rRT-PCR assays are highly automated, so many samples can be run during a single assay.

Recently, a flurry of product development activity has produced several new types of assays for Covid-19 that produce results faster than a rRT-PCR assay. For example, on February 27, 2020, Abbott Diagnostics Scarborough, Inc. received an EUA from the FDA for its ID NOW assay for the qualitative detection of SARS-CoV-2. The ID NOW assay reportedly produces positive test results in about five minutes. The ID NOW assay uses a different type of technology called isothermal nucleic acid amplification technology, which is faster than PCR technology because, among other things, it doesn’t require thermal cycling. On the downside, the ID NOW assay, as currently designed, processes one sample at a time.

Even more recently, scientists at the University of California, San Francisco and Mammoth Biosciences announced a prototype assay that uses CRISPR-Cas12 technology to detect the Covid-19 in samples from nose or throat swabs. The assay uses reverse transcription and isothermal amplification using loop-mediated amplification (RT–LAMP) for RNA amplification and takes roughly 40 minutes to complete, with positive test results being indicated by a color change on a dipstick.

8 views0 comments


bottom of page