What is the role of imaging and biomarkers within the current testing strategy for the diagnosis of Covid-19?

April 8, 2020

Kile Green¹, A. Joy Allen¹, Jana Suklan¹, Fiona R. Beyer², D. Ashley Price³, Sara Graziadio³

¹NIHR Newcastle In Vitro Diagnostics Co-operative, Newcastle University
Newcastle upon Tyne, NE2 4HH
²Population Health Sciences Institute
Newcastle University, NE2 4HH
³NIHR Newcastle In Vitro Diagnostics Co-operative,
Newcastle upon Tyne NHS Hospitals Foundation Trust, Newcastle upon Tyne, NE1 4LP

Correspondence to sara.graziadio@newcastle.ac.uk

VERDICT
Real time reverse transcriptase polymerase chain reaction (RT-PCR)remains the main technique for COVID-19 diagnosis, although chest X-ray, CT scans, and biomarkers (i.e. high CRP, low PCT, low lymphocyte counts, elevated IL6 and IL10) have been employed by some nations to aid diagnosis or to provide evidence of more severe disease progression. Many guidelines reviewed in this document relied upon RT-PCR for initial diagnosis, monitoring of disease and for aiding discharge decisions. Sometimes, patients initially have a negative RT-PCR test but with clinical suspicion of COVID-19, are kept in isolation and re-tested until positive or a clear alternative diagnosis was found. Discharge criteria in the analysed guidelines suggested two consecutive negative RT-PCR tests over 24-72h, to minimise the high false negative rate of the test. Considering the latest publications these criteria may not be stringent enough. 

BACKGROUND
The current standard for testing and direct identification of infection with SARS-CoV-2 is RT-PCR, but it has been observed that high false negative rates may be common. A number of factors could contribute to a false negative result, such as the technique of sample collection, poor quality/low sample volume of respiratory samples collected, time when the sample was collected in the course of disease, handling and storage of the sample or technical limitations of the test. This implies a high level of uncertainty in clinical decisions linked to single isolation, cohort isolation as well as de-isolation of the patients and healthcare professionals. It also leads to uncertainty regarding the diagnosis and can mean that antibiotics are not reviewed and stopped in patients with initial negative RT-PCR.

CURRENT EVIDENCE
We described the approaches to clinical investigations following a negative RT-PCR results as described in guidelines from China, Italy, Spain, UK and US. Then we looked at possible strategies to understand how to increase the accuracy of the current testing pathway to identify patients with COVID -19, while focusing on the role of imaging and of additional tests, including blood biomarkers.

Actions followed the initial negative RT-PCRs tests:

The Chinese handbook¹ included multiple tests in its ‘screening criteria’, including RT-PCR, CT abnormalities, fever and reduced white cell counts or lymphopenia. Furthermore, it states that any patient suspected of COVID-19 who tests RT-PCR negative should be kept in single isolation and re-tested 24 hours later. After two consecutive negative nucleic acid tests and no clinical manifestation or suspicion of disease, patients may be discharged from hospital. However, if COVID-19 cannot confidently ruled out in those patients based on clinical manifestations, they must remain isolated and tested every 24 hours until the COVID-19 is excluded or confirmed. Any instances of positive RT-PCR will result in the patient being admitted and treated collectively based on severity of condition.

Italian guidelines² mentioned Chest X-ray as a useful first-line radiological examination, alongside RT-PCR for follow-up and rapid assessment of patients with suspected COVID-19.

WHO guidance³ promotes the collection of lower respiratory tract samples (expectorated sputum, endotracheal aspirate, or bronchoalveolar lavage in ventilated patients) where available from patients who have provided negative upper respiratory samples (nasopharyngeal / oropharyngeal swabs), but where clinical suspicion of COVID-19 remains.

Could CT scan decrease the false negative rate of RT-PCR?

Despite Chest X-ray (CXR) and CT chest features being common in COVID-19 patients and procedures being relatively simple and quick to produce results, CXR is not advised as a first-line diagnostic tool on its own for COVID-19 as it lacks specificity (other common illnesses may share the same features). This is reported in the CDC guidance⁴, amongst others. CXR may also be normal in the early stages of COVID-19.

Chest X-ray and CT have been used in China¹ and Italy² alongside RT-PCR for both screening and monitoring of patient condition. In the Chinese Handbook of COVID-19 prevention and treatment¹ screening criteria included fever, CT abnormalities or reduced white cell or lymphocyte counts. Patients meeting this criteria would be tested by RT-PCR for confirmation. From the Italian guidelines², chest X-ray or Chest CT was suggested for stable or unstable symptomatic patients, but was not considered for asymptomatic patients. WHO guidance³ highlighted the benefits of CXR and/or CT for patients with severe pneumonia, but advised against utilising these techniques as sole first-line diagnostic tools due to the conflicting published evidence.

From KCH ITU guidelines for critical care⁵, comparisons between CT and RT-PCR highlighted greater sensitivity of CT and clinical criteria vs RT-PCR at an early stage of disease, indicating CT abnormalities may appear before PCR positivity, although not all cohort studies have shown this pattern. KCH ITU guidelines⁵ indicated that Guan et al, 2020⁸ had found CT to have greater sensitivity than chest X-ray for subtle opacities although these subtle changes may not have clinical significance⁹.

Spanish guidelines⁷ promoted the use of portable chest radiography equipment alongside PCR, blood screening and PCT for the diagnosis of COVID-19. Portable chest radiography was preferred due to ease of cleaning and decontamination compared to standard radiology suites with the added benefit of keeping radiography available for other non-COVID affected patients.

Could additional tests decrease the false negative rate of the RT-PCR?

Multiple testing in terms of both repeated measures and utilisation of multiple technologies simultaneously improved the specificity of diagnosis. The clinical information gathered for a COVID-19 diagnosis and rule-out of diseases with similar clinical symptoms was deemed imperative in a number of guidelines. It is still unclear how tests can be combined to guide diagnosis, especially in those who are RT-PCR negative¹⁰.

The Chinese guidance¹ indicated nucleic acid testing is the preferred method for diagnosis COVID-19, with combined detection of nucleic acids from multiple sites or specimens improving diagnostic test accuracy. When testing multiple sites at once, 30-40% of patients with positive respiratory tract RT-PCR also had detectable vRNA in the blood, with 50-60% expressing vRNA in faecal samples. CDC guidance⁴ reported similar findings and indicated that lower respiratory samples may provide greater yields than upper respiratory samples.

Serum antibody testing though immunochromatography, enzyme-linked immunosorbent assays (ELISA) and chemiluminescence were suggested for diagnosis and monitoring of disease with viral load decreasing as antibody level increased. IgM was detectable 10 days after symptom onset, with IgG detectable 12 days after symptom onset.

The Chinese guidance¹ further recommended testing for C-reactive protein (CRP) and procalcitonin (PCT) – most COVID-19 cases had normal procalcitonin and significantly elevated CRP, with higher CRP indicative of poor prognosis and more severe disease. This was also reported by the CDC guidelines⁴.

Lymphocyte counts, IL6 and IL10 testing were also mentioned as useful tools to aid in the diagnosis and management of COVID-19. Patients with low total lymphocyte counts at the beginning of disease onset had a poorer prognoses, while IL6 and IL10 expression was regarded as ‘helpful to assess the risk of progression to severe condition’¹.

Italian guidance² outlined a similar array of techniques and measures for accurate diagnosis and management of the condition. Repeated RT-PCRs tested via rhinopharyngeal swab on diagnosis and throughout hospital admission ‘every 48-72h until persistently negative’ was the standard approach, alongside utilisation of Chest X-ray and CT for assessment of lung parenchymal involvement. Initial screening for influenza virus was promoted alongside COVID-19 testing to reduce the risk of misdiagnosis, with the caveat that positive influenza testing should not exclude patients from being tested for COVID-19 as well. SARS-CoV-2 serology was used for confirmation of disease and monitoring of disease progression where available.

Public Health England (PHE) guidance⁶ states influenza testing should be considered in severe cases and immunocompromised patients where COVID-19 testing has returned negative, or where the test will inform clinical management.

WHO guidance³ recommends performing routine haematology and biochemistry laboratory testing, and that ECG should be performed at admission and as clinically indicated to monitor for complications, such as acute liver injury, acute kidney injury, acute cardiac injury or shock.

KCH ITU guidelines⁵ highlighted lymphopenia as a common feature of COVID-19 positive patients¹¹, occurring in ~80% of cases (data pulled from Guan et al 2020⁸ and Yang et al 2020¹¹).

KCH guidelines⁵ also made a connection with CRP, correlating it to disease severity and prognosis (elevated CRP in patients that became hypoxemic and those with a poorer prognosis). They also mentioned Influenza testing as a method for explaining symptom causes with the same comment that testing for COVID-19 should be performed regardless of influenza positivity.

A recent systematic review searched and analysed all available diagnostic and prognostic models concluding that they were at high risk of bias and they needed to be validated and their results replicated. Early studies, mainly based on data from China and Italy, identified as predictor factors  age, body, temperature, and (respiratory) signs and symptoms from diagnostic models; age, sex, CRP, lactic dehydrogenase, lymphocyte count, and features derived from CT-scoring from prognostic models. Other predictors to validate in future studies were albumin (or albumin/globin), direct bilirubin, and red blood cell distribution width¹⁰.

Impact of sampling location on accuracy of RT-PCR

Some studies have been performed to determine RT-PCR positivity in a range of samples from individual patients. Wang et al, 2020¹² indicated RT-PCR from lower respiratory tract samples were most often testing positive of SARS-CoV-2 in known positive patients (n=205, with 1070 samples collected in total). BAL (in ventilated patients) was the most sensitive (93% positive results), with sputum positivity in 72% of tests, followed by 63% positive tests from nasal swabs. Pharyngeal swabs were positive in 32% of tests. Very few patients had RT-PCR positive blood and none had positive tests from urine in this cohort.

In contrast, Wölfel et al¹³ noted no discernible differences in viral load or detection rates between naso- vs oropharyngeal swabs with all returning positive tests within 1-5 days of onset of symptoms with an average viral load of 6.76×10⁵ copies per swab, compared to post-day 5 swabs, with an average viral load of 3.44×10⁵ copies per swab and a detection rate of 39.93%. The final positive swab was taken 29-days post onset on patients with no positive culture. Wölfel found varied viral loads when comparing paired swabs and sputum samples, with some patients exhibiting higher viral load in one than the other. Average viral load of sputum samples was 7×10⁶ copies per mL.

Even though more research is needed in this area, testing of multiple sites in COVID-19 patients simultaneously may improve sensitivity and reduce false-negative rates. CDC guidance⁴ supported the use of nasopharyngeal swabs for SARS-CoV-2 testing by RT-PCR over other upper respiratory specimens, with oropharyngeal swabs recommended if nasopharyngeal specimens could not be collected.

CONCLUSIONS
Lan et al, 2020¹⁴ provides anecdotal evidence of some patients becoming consistently RT-PCR positive 5-13 days after meeting the criteria for discharge. The patients mentioned by Lan et al¹⁴ had received antiviral treatment during their hospital stay, although no comment was made on the effect this might have had on RT-PCR testing. These findings were confirmed by Li et al¹⁵ who studied a cohort of 610 patients and found that 27.5% of patients diagnosed with COVID-19 were RT-PCR positive after their first test, with 12.5% of the initially negative patients becoming RT-PCR positive in the second test. A further 12 patients in the study were RT-PCR positive after their 3^rd to 5^th test. Furthermore, RT-PCR positive patients that eventually became RT-PCR negative for several days returned to RT-PCR positive suggesting that RT-PCR positivity fluctuates throughout disease. This may be caused by insufficient viral material in the specimen, laboratory error on sampling or another issue. Li suggests that to improve survival rates in critically ill patients with signs of respiratory failure, timely transfer to ICU should be considered even if RT-PCR results are negative. Li also observes that 18 patients were found to have positive RT-PCR after meeting discharge criteria, risking viral transmission if released from isolation. Recommendations were made to isolate patients for several days after discharge to minimise the risk of transmission. Although this study was not paired with cell culture to look for viable virus.

There do not appear to be any ‘gold standard’ tests for SARS-CoV-2 causing COVID-19. RT-PCR has limitations as false negative tests appear to be high (20-30%) in the region. CXR changes may not be present on admission and although CT chest may be sensitive to pick up early changes these may not relate to clinical severity. In patients with COVID-19 a number of abnormalities may be predictive of disease severity including lymphopenia, CRP and IL-6 levels. These may not be helpful in initial diagnosis. It is possible that a pre-test probability score could be developed that would help inform the RT-PCR result particularly if this was negative. It may be that this would also be helpful in identifying patients at risk of developing severe COVID-19 disease and allowing early discharge of patients with milder disease.

False negatives in RT-PCR results are occurring both at admissions and at discharge. At admission, sampling from different locations could decrease the false negative rate, as well as using CT scan in conjunction with chest X-ray. At discharge, common guidelines suggesting two negative RT-PCR tests in 24h and absence of clinical symptoms may not be stringent enough to ensure discharged patients are no longer infectious or shedding the virus. Although practical clinical guidance for discharge may need to be developed where the ability to do RT-PCR is limited and taking into account that RT-PCR positivity may not relate to infectivity in patients who have clinically improved to point of discharge¹³.

Disclaimer:  the article has not been peer-reviewed; it should not replace individual clinical judgement and the sources cited should be checked. The views expressed in this commentary represent the views of the authors and not necessarily those of the host institution, the NHS, the NIHR, or the Department of Health and Social Care. The views are not a substitute for professional medical advice.

SEARCH STRATEGY
We used the available resource that is automatically downloading and capturing keywords (along with other data) from Pubmed (https://docs.google.com/spreadsheets/d/e/2PACX-1vRIJ0lZhFEJScTMqP4x7F1aAfxWbAMu-zjYaWwjnhPMLa-ypsX16s9NE5KMbG8B8bOibNc-e1L_0ko8/pubhtml?gid=1028682965&single=true) and selected only results that are keyworded ‘guidelines’ or ‘recommendations’. From this source we obtained 14 guidelines and 32 recommendations. Furthermore, we accessed TRIP (https://www.tripdata) and UpToDate websites (https://www.uptodate.com/contents/society-guideline-links-coronavirus-disease-2019-covid-19) on 30/03/2020 and extracted the list of guidelines available using the search terms: “novel coronavirus” OR 2019nCov OR nCoV OR 2019-nCoV OR covid19 OR covid-19 OR SARS-CoV-2. We obtained 39 and 46 guidelines/recommendations respectively and sifted them to including only guidelines on general patient management in English, Italian or Spanish (the authors were able to translate from Italian and Spanish). We also looked for guidelines for the main national and international organizations (World Health Organization – WHO; Centre of Disease Control – CDC; National Institute for Health and Care Excellence, NICE) and included the relevant ones. We included 7 guidelines for the analysis listed in the References.

REFERENCES:

1. Handbook of COVID-19 Prevention and Treatment. Chinese guidelines, 18/03/2020. alnap.org/help-library/handbook-of-covid-19-prevention-and-treatment.
2. National Institute for the Infectious Diseases “L. Spallanzani” IRCCS. Recommendations for COVID-19 Clinical Management, 16/02/2020. pagepress.org/journals/index.php/idr/article/view/8543.
3. WHO guidelines, 11/03/2020. who.int/emergencies/diseases/novel-coronavirus-2019/technical-guidance
4. Interim Clinical Guidance for Management of Patients, CDC guidelines, 30/03/2020. www.cdc.gov/coronavirus/2019-ncov/hcp/clinical-guidance-management-patients.html
5. King’s Critical Care – Clinical Management of COVID-19. UK summary of evidence, 09/03/2020. wspidsoc.kenes.com/wp-content/uploads/sites/95/2020/03/KCC-Covid19-evidence-summary.pdf
6. PHE guidelines, 25/03/2020. gov.uk/government/publications/wuhan-novel-coronavirus-initial-investigation-of-possible-cases/investigation-and-initial-clinical-management-of-possible-cases-of-wuhan-novel-coronavirus-wn-cov-infection.
7. Clinical Management of COVID-19 (Spanish). Spanish guidelines, 20/03/2020. diariofarma.com/2020/03/20/actualizado-el-documento-con-las-recomendaciones-de-tratmiento-para-pacientes-con-covid-19
8. Guan et al, ‘Clinical Characteristics of Coronavirus Disease 2019 in China’, NEJM, 2020.
9. Inui et al, ‘Chest CT Findings in Cases from the Cruise Ship “Diamond Princess” with Coronavirus Disease 2019 (COVID-19)’, Radiology: Cardiothoracic Imaging, 2020
10. Wynants et al, ‘Prediction models for diagnosis and prognosis of covid-19 infection: systematic review and critical appraisal’. BMJ
11. Yang et al, ‘Viral Dynamics in mild and severe cases of COVID-19’, Lancet, 2020
12. Wang et al, ‘Detection of SARS-CoV-2 in Different Types of Clinical Specimens, JAMA, 2020.
13. Wölfel et al, ‘Virological assessment of hospitalized patients with COVID-2019’, Nature, 2020.
14. Lan et al, ‘Positive RT-PCR Test Results in Patients Recovered From COVID-19’, JAMA, 2020.
15. Li et al, ‘Stability issues of RT-PCR testing of SARS-CoV-2 for hospitalized patients clinically diagnosed with COVID-19’