Why Are There So Many False Negatives in COVID-19 PCR Testing?
As we contend with the COVID-19 pandemic, we are hearing lots of news about “false negatives” in patient testing. What are false negatives, what causes them, what are the consequences, and what might be done to minimize them?
In COVID-19, the infectious agent is the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2). Viruses gain access into the cell by fooling the cells into ingesting them. Once inside, they hijack the cell’s manufacturing machinery by introducing their own blueprints (DNA or RNA) and using the cell’s machinery to make more viruses (creating the virus’ proteins). These new viruses then leave the cell, often resulting in its death, and repeat the process, resulting in the viral infection. Modern means of detecting the viruses rely on the technique of quantitative polymerase chain reaction or qPCR.
PCR is a technology based on the “amplification” of polynucleic acids (so-called information molecules) using the polymerase chain reaction.
There are two information molecules. DNA is the long term archive of genetic information. RNA, which is the working copy of the DNA, converts the “blueprints” to proteins: the structural materials, sensors, and chemical factories of all cells.
Some viruses use DNA as their “blueprints”, as do most higher life forms. Other viruses, like the COVID-19 virus, use RNA as their blueprint. The PCR process only amplifies DNA, not RNA. Thus, for COVID-19 diagnostic testing using PCR technology, a pre-step is required to convert the viral RNA to DNA using another naturally occurring enzyme, reverse transcriptase.
So what generates a “positive” signal in patient testing? A sample of material is taken from the patient, thought to contain the virus. This can be a biological fluid — such as blood, stool, sputum, or urine — or a tissue, such as a biopsy sample. For COVID-19, the most commonly used method for screening patients is to obtain a mucus sample from the nasopharyngeal (NP) region located in the back of the throat. The sample is obtained with a long “swab” introduced through the nose. Depending on the system used, the sample can be stored in a special fluid (viral transfer media or VTM) until it is “read”, or directly introduced into the reader.
The sample is processed to remove non-nucleic acid contaminants and then submitted for the test. At each “cycle” of the PCR test, the signal from double-stranded DNA creates doubles, so the anticipated shape of the curve of fluorescence signal, the mark of amplification, is exponential. A curve that grows too fast is due to contamination and is ignored. For an acceptable curve shape, the number of PCR cycles required for the fluorescence signal to cross a threshold intensity, Ct, is used to determine whether the sample is positive or negative[1].
False signals are readings from the PCR assay that incorrectly diagnose the patient, relative to a “gold standard”. In the case of a false positive, the PCR reading incorrectly suggests that the patient has the virus.
False negatives are a far more frequent — and far more serious — outcome for COVID-19. In this case, the test result suggests that a patient does not have the virus, when the person is in fact infected. What causes these false negatives? A negative read of the assay is due to not enough fluorescence signal growing before 40 PCR cycles have occurred. If there is too small an amount of virus in the sample site, then not enough signal is generated. There are several potential contributors to a sample with low viral load:
● For some point-of-care diagnostics systems, a swab is meant to go from the nose straight to the reader. Before clarification by the manufacturer and FDA, some of these point-of-care systems were being used for samples that were first stored in VTM and then read, strongly diluting the samples and lowering the amount to be amplified. Unfortunately, given the large volume of samples being collected for COVID-19 testing and staffing issues, there was not a practical way to use the machine as originally intended [2].
● In some patients, the virus has migrated from the nasopharyngeal region further down the airway, so the “viral titer” is also too low. The patient is still infected, but the virus is elsewhere in the body.
● Finally, the sampling swabs may have an impact. The complex biochemical reactions used to convert the RNA to DNA, and then also to perform the amplification, are sensitive to contaminants. Natural products, like cotton, can contain enzymes that break down the RNA or interfere with the processing enzymes. Moreover, some surfaces of natural and synthetic materials are charged positively and can stick to the negatively charged RNA molecules, not releasing them for the assay. Finally, the geometry of the swab can have an impact. Does it have enough surface area to collect sufficient virus?
Given the shortages of conventional NP swabs[3], which consist of a synthetic nylon “flock” glued to a flexible shaft, a number of manufacturers and academic clinicians began to use additive manufacturing to rapidly fill the need[4]. Polymers typically used for the 3D printing of dental appliances were employed; given that such dental materials have already been tested and cleared by FDA as safe for contact with the mucosal lining of the mouth, they are therefore appropriate for contact with the throat and nose. Through a collaboration with Beth Israel Deaconess Medical Center in Boston, various designs were tried in clinical studies and tested for their ability to capture viral material without compromising the PCR test. They were also tested for patient and clinician acceptance. All were shown to not interfere with the PCR test in side-by-side comparisons with conventional swabs. The only inaccuracies observed were a small number of false positives. But on further scrutiny, these proved to be true positives, due to the “standard swab” incorrectly delivering a negative result from a patient showing clinical symptoms of COVID-19. So, in addition to filling the gap, these 3D printed swabs may result in being more precise, decreasing false negative rates and giving patients and clinicians better clarity on patients’ true testing status. A larger study, powered to determine superiority, would be needed to draw definitive conclusions.
Carbon, working in collaboration with Resolution Medical, an established medical device manufacturer, and Stanford University through Stanford Health Care, undertook a large clinical trial [5] to establish the efficacy of the Resolution Medical Lattice Swab. These swabs were fabricated using Carbon Digital Light Synthesis™ (DLS™) technology by several of Carbon’s production partners, then shipped to Resolution Medical for final processing and distribution. About 400 patients were tested in parallel with conventional swabs, and the results confirmed that the Resolution Medical Lattice Swab performed as well as conventional flocked swabs. There was only one false negative result, where the Lattice Swab read negative and the conventional swab read positive, and for the two false positive results, where the conventional swab was read as negative and the Lattice Swab as positive, the patient was clinically determined to be a positive patient. In these cases, the viral titer was low (a Ct of 38–39 cycles). Thus, the Lattice Swab might be more effective in the diagnosis of COVID-19 and other viral or bacterial infection at low pathogen titer.
Based on the compelling non-inferiority data and positive clinical experience across multiple institutions, Resolution Medical, using Carbon DLS technology and KeySplint Soft® Clear material from Keystone® Industries, is manufacturing and shipping hundreds of thousands of swabs to healthcare facilities around the country.
- FDA EUA website, SARS-CoV-2 RNA, Qualitative Real-Time RT-PCR (Test Code 39433), https://www.fda.gov/media/136231/download
2) a) Atreyee Basu, Tatyana Zinger, Kenneth Inglima, Kar-mun Woo, Onome Atie, Lauren Yurasits, Benjamin See, Maria E. Aguero-Rosenfeld, “Performance of Abbott ID NOW COVID-19 rapid nucleic acid amplification test in nasopharyngeal swabs transported in viral media and dry nasal swabs, in a New York City academic institution”, BioR𝝌iv, https://doi.org/10.1101/2020.05.11.08989; b) Study Raises Questions About False Negatives From Quick COVID-19 Test, https://www.npr.org/sections/health-shots/2020/04/21/838794281/study-raises-questions-about-false-negatives-from-quick-covid-19-test
3) Lauren Webber, Christina Jewett, Testing Swabs Run In Short Supply As Makers Try To Speed Up Production, NPR, https://www.npr.org/sections/health-shots/2020/03/18/817801222/testing-swabs-run-in-short-supply-as-makers-try-to-speed-up-production.
4) Cody J Callahan, Rose Lee, Kate Zulauf, Lauren Tamburello, Keneth P Smith, Joe Previtera, Annie Cheng, Alex Green, Ahmed Abdul Azim, Amanda Yano, Nancy Doraiswami, James Kirby, Ramy Arnaout, Open Development and Clinical Validation of Multiple 3D-Printed Sample-Collection Swabs: Rapid Resolution of a Critical COVID-19 Testing Bottleneck, medRxiv 2020.04.14.20065094; doi: https://doi.org/10.1101/2020.04.14.20065094
5) Ian Bennett, Philip L. Bulterys, Melody Chang, Joseph M. DeSimone, Jennifer Fralick, Marie Herring, Hardik Kabaria, Christina Kong, Blane Larson, Owen Lu, Arielle Maxner, Elle Meyer, Shawn Patterson, Steve Pollack, Jason Rolland, Steven Schmidt, Sridhar Seshadri, Keshav Swarup, Chelsey Thomas, Ryan Van Wert, “The Rapid Deployment of a 3D Printed Latticed Nasopharyngeal Swab for COVID-19 Testing Made Using Digital Light Synthesis”, MedR𝟀iv, https://doi.org/10.1101/2020.05.25.20112201
Steve Pollack, Ph.D. is a Science Fellow at Carbon, Inc., and former Director, Office of Science and Engineering Laboratories, Center for Devices and Radiological Health, FDA