Coronavirus disease 2019 (COVID-19) pneumonia has had catastrophic effects worldwide. Radiology, in particular computed tomography (CT) imaging, has proven to be valuable in the diagnosis, prognostication, and longitudinal assessment of those diagnosed with COVID-19 pneumonia. This article will review acute and chronic pulmonary radiologic manifestations of COVID-19 pneumonia with an emphasis on CT and also highlighting histopathology, relevant clinical details, and some notable challenges when interpreting the literature.
Key points
- •
Acute computed tomography (CT) manifestations of coronavirus disease 2019 (COVID-19) pneumonia most often reflect organizing pneumonia in mild resolving cases and diffuse alveolar damage in progressive severe disease, though there is substantial overlap.
- •
Chronic CT manifestations are predominately seen in those with moderate to severe disease.
- •
Specific CT features of post-COVID-19 pneumonia fibrosis include traction bronchiectasis, architectural distortion, and rarely honeycombing.
Introduction
Novel coronaviruses have led to 3 outbreaks of severe acute respiratory syndrome (SARS) in the last 20 years. The most recent was termed coronavirus disease 2019 (COVID-19) caused by the SARS-CoV-2 virus, which began in Wuhan, China in December 2019. The virus quickly spread leading to a worldwide pandemic that is now approaching its fourth year. The lungs are a frequent site of involvement, with a small subset of mostly at-risk patients progressing to acute respiratory distress syndrome (ARDS) and death. COVID-19 survivors, especially those who had moderate to severe disease, have a relatively high rate of persistent pulmonary symptoms and computed tomography (CT) abnormalities. This article will review acute and chronic pulmonary radiologic manifestations of COVID-19 pneumonia with an emphasis on CT, while also highlighting histopathology, relevant clinical details, and some notable challenges when interpreting the literature.
Histopathological appearance
The virus SARS-2-CoV causes lung epithelial injury predominately in 2 ways: direct viral toxicity to alveolar epithelial cells and the resultant (and sometimes exaggerated) host immune response. Broadly, COVID-19 lung injury can be categorized into “mild resolving” and “severe progressing” though there is substantial overlap pathologically and by imaging. In both disease courses, the immune response causes vascular membrane permeability and the transmigration of fluid, various proteins, and inflammatory cytokines. In mild cases, nonspecific or organizing pneumonia (OP) predominates and corresponds to the classically-described peripheral basilar opacities seen on CT ( Fig. 1 ).
In severe progressive disease, there is basement membrane disruption and progression to diffuse alveolar damage (DAD) and acute and fibrinous OP (AFOP). The expected evolution of patients with DAD begins with an exudative phase characterized by hyaline membrane formation and worsening diffuse pulmonary opacification (often from ground-glass to consolidation) on CT, followed by a proliferative phase, where there is organization of lung injury manifesting with worsening volume loss, architectural distortion, and frequently accompanied ventilator related injury. The final fibrotic phase is hallmarked by intense fibroblast activity and varying degrees of worsening architectural distortion and traction bronchiectasis. The phases of DAD are not linear or homogeneous as there are frequently spared areas of lung initially, which can become involved at later timepoints. Endothelial injury is also a key factor in the pathogenesis of patients with DAD, pulmonary embolism (PE), and perfusion abnormalities.
Thus, the pathologic (and corresponding radiographic) appearance of COVID-19 pneumonia is not necessarily unique, rather the expected consequence of viral induced lung injury patterns such as OP and DAD, with the long-term and potentially permanent manifestations linked to the severity of acute disease. It should be noted that some histologic features may be more common in COVID-19 such as pulmonary microthrombi and certain vascular lesions. ,
Acute COVID-19 pneumonia
Background
The incubation period of SARS-CoV-2 is typically up to 14 days with most cases occurring 4 days to 5 days after exposure, though up to a third of patients never develop symptoms. In the majority of those that did, symptoms were usually mild, especially if infected by the Omicron variant (B.1.529) or after being vaccinated. However, during the ancestral strain, 14% of infected individuals developed dyspnea, hypoxia, and/or lung opacities and 5% went into respiratory failure. , In those with more severe disease, the median time to hospital admission was around 1 week with ARDS developing in up to a third of inpatients about 1 week later. Nearly 70% of those in intensive care units (ICUs) required mechanical ventilation and had fatality rates as high as 50%. , Case fatality rates scale by age, ranging from less than 1% at age 50 to 20% at age 90 and are higher when subcategorized by other risk factors.
Imaging of Acute Lung Parenchymal Manifestations
Common CT abnormalities (<2 weeks): Imaging manifestations early on are mostly those of OP, commonly with bilateral peripheral and lower lung ground glass or consolidative opacities (see Fig. 1 ). , Other imaging findings that may be seen include patchy ground glass opacities (GGO) without a clear distribution, sometimes appearing rounded, the reverse halo sign, and the crazy paving pattern ( Fig. 2 ). Notably, solid nodules (large, small, or tree-in-bud), cavitation, and a perihilar or lobar distribution of disease are very infrequently (or not) seen and should be presumed to be a concurrent or different process. Pleural effusions and significant lymphadenopathy are also infrequent. , Common early CT findings and distribution are summarized in Table 1 .
| CT Findings First 2 Weeks | ||
|---|---|---|
| Common | Less Common | Low or Not Seen |
| GGO with consolidation | Consolidation without GGO | Cavitation |
| GGO without consolidation | Crazy paving | Nodules |
| Linear opacities | Tree in bud | |
| Rounded GGO | RHS | |
| Distribution | ||
| >2 lobes | Single lobe | Lobar |
| Peripheral | Diffuse | Perihilar |
| No clear distribution | Upper lobe | |
| Lower | ||
Differential diagnosis : During the pandemic phase of COVID-19, peripheral and lower lung predominant opacities strongly suggested COVID-19 pneumonia, mainly due to high prevalence but also when compared to other common community acquired viral infections. However, the imaging differential diagnosis for this distribution of disease also includes OP (cryptogenic or due to other factors including other viral infections), non-specific interstitial pneumonia (NSIP), and pulmonary infarcts , ( Fig. 3 ). When GGOs are diffuse, the differential is less specific and includes pulmonary edema, alveolar hemorrhage, some opportunistic infections, and DAD of any cause. , Differential diagnoses and distinguishing features are summarized in Table 2 .
| Differential Diagnosis | |
|---|---|
| Peripheral Opacities a | Distinguishing Features |
| Influenza | Tree in bud nodules, pleural effusions |
| Infarcts | “Bubbly” appearance on soft tissue windows, absence of air bronchograms, lobar |
| Non-infectious OP | Nodular opacities, unilateral, lobar, recent drug initiation |
| CEP | Upper lobe, eosinophilia |
| NSIP | Subpleural sparing, chronicity, connective tissue disease |
| Diffuse opacities a | Distinguishing features |
| Pulmonary edema | Septal thickening, perihilar, pleural effusions |
| Alveolar hemorrhage | Centrilobular opacities |
| Pneumocystis pneumonia | Intrapulmonary cysts, immunodeficiency |
| CMV pneumonia | Ill-defined small nodules, transplant setting |
a Opacities in this table imply GGO with or without consolidation.
Temporal changes (2–6 weeks): Chest CTs in symptomatic SARS-CoV-2 positive patients are usually normal within 24 hours and up to a third of patients within the first 5 days. Rarely, CT findings may become apparent prior to symptom onset and/or in patients with negative upper respiratory tract RT-PCR. CT findings of COVID-19 pneumonia typically develop by day 5 and peak around day 10. In those with mild to moderate disease, GGOs usually predominate throughout the first month and then gradually resolve, typical of OP. During resolution, GGO may appear to paradoxically increase though are often paired with a decrease in attenuation, a finding some have called the “tinted sign” ( Fig. 4 ). However, the frequency of opacities and time to resolution is strongly affected by disease severity. In those with progressive disease, DAD ensues and GGOs often become denser, appearing more consolidative during the second and third weeks where development of volume loss and bronchial dilation is common during the organizing phase ( Fig. 5 ). By 1-month, consolidation usually decreases, and organization becomes evident, hallmarked by well-defined opacities, reticulation, architectural distortion, parenchymal bands, and possibly features of early fibrosis.
Complications: Patients with COVID-19 may suffer from several complications including pulmonary embolism (PE) and superinfection. In one large systematic review, patients (predominately those in ICUs) had a bacterial superinfection rate of nearly 20%. Clues to superinfection include the development of nodules, lobar consolidation, or cavitation ( Fig. 6 ). The presence of a pleural effusion may also indicate either concurrent heart failure, or the development of an empyema related to superinfection ( Fig. 7 ). Patients who develop clinical ARDS are at risk for barotrauma, which manifests as pneumothorax, pneumomediastinum, or intra-pulmonary pneumatoceles ( Fig. 8 ).
Acute Pulmonary Vascular Manifestations
The association between COVID-19 pneumonia and pulmonary thrombi or emboli was raised relatively early during the pandemic. Several reports described pulmonary vascular microthrombi on pathology of patients with DAD, although such findings have long been described in patients with DAD from other causes including other infections like influenza. After these early pathology reports, case series were published in radiology journals describing a very high rate (>20%) of positive CT pulmonary angiograms, mostly in patients with clinically severe illness. A meta-analysis that was later published showed a lower rate of 16% PE in COVID-19 patients. Notably, most studies did not include a comparison group.
Several studies comparing the rates of pulmonary embolism or venous thromboembolism in COVID-19 patients to other similar patients found no significant differences, although a large meta-analysis did find a slightly increased risk in COVID-19 patients compared to controls. Taking all these findings into consideration, it is likely that the predominant driving factor for venous thromboembolism is the systemic inflammatory state and development of DAD. It is possible that COVID-19 infection concurrently presents a uniquely more pro-thrombotic state.
The extent to which pulmonary vascular manifestations contribute to respiratory failure in patients with COVID-19 is difficult to determine. As noted earlier, the vast majority of the patients die from extensive pneumonia. Several studies reported unusual perfusion abnormalities within spared lung, as seen with dual energy CT (DECT) ( Fig. 9 ). Some consolidated regions also demonstrate decreased perfusion, which may be more pronounced in COVID-19 than other pneumonias. The size of these abnormalities is much greater than could be seen from the microthrombi described in pathology reports. It is possible these could be related to distal pulmonary emboli not identified by CT angiography but could also be the result of (dysregulated) hypoxemic vasoconstriction. It can also be challenging to interpret DECT iodine perfusion blood volume (PBV) maps when pulmonary opacities are present. Reduced iodine to affected lung may also be the result of threshholding and thus influenced by the analysis method and specific post-processing parameters limiting interpretation.
CT Reporting and Diagnosis of Acute COVID-19 Pneumonia
Reporting : In order to standardize reporting of chest CTs during the pandemic phase of COVID-19, the Radiologic Society of North America (RSNA) developed consensus reporting guidelines, which included 4 categories ( Fig. 10 ). Other diagnostic scoring systems were also developed , ( Table 3 ). These guidelines provided a valuable service early on to disseminate common imaging manifestations across the globe and to introduce some reporting consistency. In each reporting system, the higher confidence categories demonstrated relatively high specificities and positive predictive values (even when assessed across continents), though with controlled cohorts. , , Lower confidence categories had decreased specificity and likelihood of a positive RT-PCR, though were more sensitive. Moderate to substantial inter- and intra-observer agreement was shown between radiologists of varying experience levels and expertise. However, it is important to realize that at tertiary referral centers with complex patients on multiple medications, there is substantial overlap with other infectious and non-infectious lung diseases, resulting in lower predictive values.
| Reporting Systems | ||||
|---|---|---|---|---|
| Confidence | RSNA | CO-RADS | COVID-RADS | BTSI |
| Low | Negative, Atypical | CO-RADS 1–2 | COVID-RADS 0–1 | Non-COVID |
| Intermediate | Indeterminate | CO-RADS 3–4 | COVID-RADS 2A, 2B | Indeterminate, probable |
| High | Typical | CO-RADS 5 | COVID-RADS 3 | Classic COVID-19 |
Application : Early in the pandemic, when viral molecular tests were in short supply and had long processing delays, some advocated for CT as an adjunct or even replacement for molecular testing. The Society of Thoracic Radiology and the American Society of Emergency Radiology issued a joint position statement that CT should not be used as a screening test in patients under investigation for COVID-19. The sensitivity of CT for patients with mild or asymptomatic cases of COVID-19 is quite low, so imaging would likely be insufficient to determine who should be isolated.
Certainly, imaging does play a role in helping to manage patients with acute respiratory diseases and may narrow or expand the differential diagnosis. For patients with a high clinical suspicion of COVID-19 pneumonia despite a negative nasopharyngeal reverse transcription-polymerase chain reaction test (RT-PCR), imaging can help suggest the diagnosis and drive additional testing. Interestingly, a substantial portion of those with typical CT features who initially had a negative RT-PCR eventually tested positive, allowing some in the early stages of the pandemic to use the ‘typical CT’ appearance (in the appropriate patient) for clinical diagnosis. However, in the current era of a low COVID-19 prevalence, caution is advised that other diagnoses are at least as likely, if not more likely than COVID-19 in patients with a ‘typical’ pattern on CT.
Computed Tomography as Prognostic and Triage Tool in Acute COVID-19 Pneumonia
A minority of patients progress to having severe diffuse opacities, representing DAD pathologically, and often requiring intubation. Predictably, the greater the extent of disease on CT, the more likely a patient is to die from COVID-19 pneumonia. As a result, several scoring systems that incorporated CT findings were developed to help triage patients and establish a prognosis. In one scoring system, Gangemi and colleagues showed that patients with diffuse opacities had an 11- fold increase in the ICU transfer rate at 24 hours and that the decision to transfer was based on the CT score in 84% of cases. The World Health Organization (WHO) went on to recommend “…using chest imaging in addition to clinical and laboratory assessment to decide on regular ward admissions versus ICU admission” in June of 2020. However, scoring systems should be viewed with caution. In a large retrospective review, the majority demonstrated a high risk of bias. In addition, these scoring systems are not necessarily unique to COVID-19 and probably could be applied to any cause of DAD.
Special Considerations
Viral variants : As the SARS-CoV-2 virus evolved, several variants developed, most recently the Omicron variant and its sublineages. COVID-19 disease severity has been less in the Omicron variant era, likely due to a combination of lower viral virulence, as well as vaccine-induced and natural immunity in the population. , In addition to patients developing pneumonia less commonly, those with pneumonia have been shown to less likely develop the typical imaging manifestations. , Despite this, there are unfortunately still many patients who die from COVID-19, the majority of whom develop COVID-19 pneumonia. Given the lower frequency of pneumonia and the lower rate of typical CT manifestations, it is possible that the RSNA guidelines will be less accurate in diagnosing COVID-19 in the Omicron era.
Immunocompromised patients : While the frequency and severity of pneumonia is less during the Omicron era than during the ancestral strain era, a minority of patients still develop COVID-19 pneumonia. These patients are disproportionately immunocompromised. The B-cell depleting agent rituximab has a particularly high association with worse COVID-19 outcomes. Some immunocompromised patients are unable to mount a sufficient immune response to clear the virus and may suffer from long-term circulation of virus and recurrent lung injury ( Fig. 11 ).
Related posts:
The Impact of Coronavirus Disease 2019 on Patients Living with Rheumatic Diseases
The Burden of Post-Acute Sequelae of Coronavirus Disease 2019 in Individuals with Rheumatic Diseases
Interventions to Improve COVID-19 Vaccine Hesitancy
Flares of Systemic Autoimmune Rheumatic Disease Following Coronavirus Disease 2019 Vaccination
Lessons on Telemedicine in Rheumatology
Review of Immunologic Manifestations of COVID-19 Infection and Vaccination
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
