The Role of Imaging in COVID-19

WHAT ROLE DOES IMAGING PLAY in assessing COVID-19? And what are the effects of COVID-19 in cancer patients? These are two central questions that researchers aimed to elucidate in a scientific session presented at the Society of Nuclear Medicine and Molecular Imaging meeting titled “Integral Role of Imaging in the Management of COVID-19 Patients.”

The Use of Imaging to Assess the Impact of COVID-19 in Cancer Patients

The first presenter of this session, Michael Knopp, MD, PhD, the Ohio State University Wexner Medical Center, Columbus, Ohio, delved into the observation of COVID- 19 as a growing problem in cancer patients in a natural history study titled “Update on the NCI-NCTN Clinical Trial: Natural History of COVID-19 in Cancer Patients.”

What exactly is a natural history study? Dr. Knopp, in the content of this analysis, defined a natural history study as an observation of “the biological effects of COVID-19 as a rising comorbidity in cancer patients,” noting its importance as a reason for the NCI initiating the study “in record time.”

Dr. Knopp and colleagues are conducting this with three objectives in mind: first, to analyze information with the intent of characterizing the association between patient factors (such as existing pre-comorbidities, cancer types, treatment type, and demographics) and both short- and long-term outcomes of COVID-19 (such as disease symptoms, severity, and mortality). The second objective is to describe potential modifications to cancer treatment in response to COVID-19 infection in both adult and pediatric patients, including dose adjustments, changes in symptom management, and even temporary or permanent discontinuation of therapy. Third, this study aims to assess the association between COVID-19 outcomes and patients’ clinical-pathological characteristics.

Imaging efficacy is a focal part of Dr. Knopp’s analysis, and the goal of imaging in this study is to assess the potential impact of COVID-19 on cancer patients and to evaluate the response of malignant lesions to COVID-19 infection.

“We know about inflammatory processes from flare observation and also, we have been speculating about potential accelerated progression,” Dr. Knopp said. “We also wanted to determine the potential diagnostic difficulties in managing patients who have been diagnosed or have been exposed to COVID-19 due to their residual abnormalities.”

With this natural history study, Dr. Knopp explained, he and colleagues “are looking at baseline images prior to confirmed exposure of COVID-19 and the radiological findings to understand its relationship to the exposure.” He added that, while only approximately 2% of pediatric patients have tested positive for COVID-19 infection, the study included a pediatric population.

As of May 2021, this natural history study has enrolled 1,172 patients from 253 different clinics in North America and collated data from 826 patients with 127 positron emission tomography/computed tomography (PET/CT) studies included for analysis. Dr. Knopp provided two examples that demonstrate the breadth of information captured in this study: in one, example of a patient with metastatic cancer and a pulmonary lesion, imaging revealed that, two months after COVID-19 infection, the patient had progressive oncologic disease and pulmonary sequelae, with some collapse of the left lung.

In another example, a cancer patient who underwent imaging with fluorodeoxyglucose (FDG)-PET at the beginning of COVID-19 infection showed active lymph nodes that have elevated FDG uptake in the area of COVID-19 inflammation.

“In conclusion, this is a unique natural history trial that typically (would not have) included imaging and will give us a unique insight into the correlation of imaging and other biomarkers,” Dr. Knopp said. “We have excellent participation from a large number of U.S. sites and [we have shown] the feasibility of staging such a study in record time.”

When asked about the ultimate plan for analyzing this large amount of data, Dr. Knopp responded: “Our initial focus or our primary hypothesis was determining whether we see different diagnostic and therapeutic patterns compared to matched populations. Within the NCI’s National Clinical Trials Network, there is a massive repository of clinical trials, so we were anticipating that we would match the COVID-19 population to a matched population that we have on hand to see differences.” Initially, he and his research team were “concerned about the flare and now we are basically adjusting the questions as we better understand those challenges.”

“While not initially anticipated, one of our biggest challenges from the community is the ‘long haulers,’ who have long-term effects from COVID-19 infection,” Dr. Knopp added. “So, we do have that initial question, but this is one of the trials where we’ll have to adjust, but having this embedded in the structure allows us to have matched populations so we can do that assessment.”

Imaging is Used to Understand the Extent of COVID-19

The guest speaker of this scientific session, Ali Gholamrezanezhad, Assistant Professor of Clinical Radiology at Keck School of Medicine, University of Southern California, spoke about COVID-19 from an imaging perspective. Dr. Gholamrezanezhad began his presentation by posing the question: What’s the role of imaging in the evaluation of COVID-19?

Early in the pandemic, countries like China were diagnosing COVID-19 using imaging because of an insufficient reserve of RT-PCR tests. Now, with enough supply of nasal swab test kits to meet the demand of the pandemic, imaging is used not to diagnosis COVID-19, but rather to provide a baseline understanding of the disease’s extent and to assess the level of lung involvement.

Therefore, physicians have two imaging options: chest x-ray, which is used for most patients and evaluates the extent of the disease, or chest CT, which is considerably more sensitive and used only when a complication in disease course is suspected. These complications could include “superimposed infection, abscess formation, pneumothorax, pneumomediastinum, pericardial effusion, or any complications of the disease,” Dr. Gholamrezanezhad said. “In most of our patients, we don’t do chest CTs.” To substantiate that claim, he referenced a previous retrospective study that he took part in consisting of more than 2,000 cases. The study found less than 10 chest CTs done in their university system.

Dr. Gholamrezanezhad then posed a second question: What do we see in imaging? To answer this question, Gholamrezanezhad and his colleagues conducted a systematic review in early February 2020 – the first such review in COVID-19 imaging. The review comprised 919 patients and found that COVID-19 causes multilobar, ground-glass opacities, indicating density around the lungs, which in the case of COVID-19, is usually found in both lungs and results in pain.

“You can see that the pericardiac area is preserved – it’s clean, especially on the left lung,” Dr. Gholamrezanezhad said of an image he showed of a young COVID-19 patient. “We also have the perihilar involvement, but as you see the disease process is mainly peripheral.” To standardize COVID-19 images, Dr. Gholamrezanezhad suggested using a RADS system specifically for COVID-19, fittingly called COVID- RADS. In this classification system, imaging findings are categorized from 0 (an almost normal chest CT) to 1, which indicates atypical imaging features of COVID-19, to 2A and 2B, meaning greatly suspicious for COVID-19, and 3, meaning highly suspicious for COVID-19.

To cite a specific COVID-19 occurrence, Dr. Gholamrezanezhad detailed the case of a 58-year-old man who presented with fever, cough, and shortness of breath for a few days. A PCR test confirmed COVID-19 infection, and imaging revealed multifocal ground-glass opacities in the peripheral region of both lungs. Three days later, the patient’s symptoms worsened. “This is how we understood the normal course of the disease,” Dr. Gholamrezanezhad said. “We still see all those ground-glass opacities, but now we have to play in position of consolidations. Those patchy, white areas superimposed on the prior ground-glass opacities [reveal] consolidation through which you cannot see the bronchovascular lung markings. This is how the disease progressed.”

No Clear Imaging Guidelines for COVID-19

Dr. Gholamrezanezhad was asked about his experience with fungal infections superimposed on infections in COVID-19 patients. “There are some reports about superimposed infections, not only fungal but also bacterial-like staph or trichomonad superimpositions,” he said. “It depends on the patient situation. In the outpatient setting [trichomonad superimpositions] are much less. In inpatient cases, in patients with intubation and ICU admission, there are reports that they say it’s up to 40% superimposition of infection, and that’s actually one of the main causes of death. This is unfortunate because it’s very easily overlooked. It’s a treatable component of COVID-19. We can just start them on antibiotics.”

In response to a question related to the chronic manifestations of COVID-19, such as fibrotic changes, Dr. Gholamrezanezhad was asked about the ideal length of follow-up on those changes. “For now, we try to start to get chest x-ray at 12 weeks. If we see any residual lung abnormality, we can go to high-resolution CT. Any future imaging depends on the clinical manifestation of the disease. If the patient is getting worse or if they have poor functional capacity, based on the clinical picture, we may need to do more imaging,” he said. “But there are no clear guidelines now because we are still in the earliest stages of the long-term COVID-19.”


Knopp M, et al. SNMMI 2021 Annual Meeting Presentation. Available at: http://www.snmmi. org/Video/Player.aspx?VideoID=903. July 2021.