Easy-to-treat and difficult-to-treat radiological phenotypes in coronavirus disease 2019 pneumonia: A single-center experience : Radiology of Infectious Diseases

Secondary Logo

Journal Logo
Advanced Search
Original Article

Easy-to-treat and difficult-to-treat radiological phenotypes in coronavirus disease 2019 pneumonia

A single-center experience

Patil, Shital1,; Dhumal, Uttareshvar2; Patil, Deepak1; Acharya, Abhijit3

Author Information

1Department of Pulmonary Medicine, MIMSR Medical College, Latur, Maharashtra, India

2Department of Radiodiagnosis, MIMSR Medical College, Latur, Maharashtra, India

3Department of Pathology, MIMSR Medical College, Latur, Maharashtra, India

Address for correspondence: Dr. Shital Patil, Department of Pulmonary Medicine, MIMSR Medical College, Latur, Maharashtra, India. E-mail: [email protected]

Received December 26, 2022

Received in revised form February 20, 2023

Accepted March 10, 2023

This is an open access journal, and articles are distributed under the terms of the Creative Commons Attribution-NonCommercial-ShareAlike 4.0 License, which allows others to remix, tweak, and build upon the work non-commercially, as long as appropriate credit is given and the new creations are licensed under the identical terms.

Radiology of Infectious Diseases 10(1):p 19-29, Jan–Mar 2023. | DOI: 10.4103/RID.RID_47_22
  • Open

Abstract

INTRODUCTION: 

Radiological phenotypes are observable radiological patterns or characteristics. Robust data are available regarding the role of high-resolution computed tomography (HRCT) in coronavirus disease 2019 (COVID-19) pneumonia. We evaluated the role of radiological phenotyping in assessing severity and predicting the response to therapy, as well as its association with outcomes in COVID-19 pneumonia.

METHODS: 

This prospective observational study included 3000 COVID-19 reverse transcription polymerase chain reaction-confirmed cases with lung involvement who underwent thoracic HRCT on hospital admission and were categorized as mild, moderate, or severe according to lung segment bilateral involvement (mild 1–7, moderate 8–15, and severe 16–25). Follow-up thoracic CT imaging was also conducted 6 months after hospital discharge. Response to treatment phenotypes was categorized as “easy to treat” or “difficult to treat” based on the response and interventions required in indoor settings, including ventilatory support. Age, gender, comorbidities, laboratory parameters, the use of bilevel-positive airway pressure/noninvasive ventilation, and outcomes (with or without lung fibrosis) were key observations. The Chi-square test was used for statistical analysis.

RESULTS: 

Easy-to-treat and difficult-to-treat radiological response phenotypes were observed in 20% and 80% of the cases, respectively. There were significant associations between the radiological phenotypes and the duration of illness at hospital admission. The duration of illness (<7 days, 7–14 days, and >14 days) could predict the radiological phenotype (P < 0.00001). Laboratory parameters at hospital admission (C-reactive protein, interleukin-6, ferritin, lactate dehydrogenase, and D-dimer) were significantly associated with the radiological phenotypes (P < 0.00001), as were interventions required in indoor units (P < 0.00001). The HRCT severity score at admission was significantly correlated with the radiological phenotype (P < 0.00001). Post-COVID lung fibrosis or sequelae were also significantly associated with the radiological phenotype (P < 0.00001).

CONCLUSION: 

Easy-to-treat and difficult-to-treat phenotypic differentiation had a crucial role during the initial assessment of COVID-19 cases on hospitalization and was used for planning targeted intervention treatments in intensive care units. In addition, phenotypic differentiation had an important role in analyzing the radiological sequelae and predicting final treatment outcomes.

Introduction

Coronavirus disease 2019 (COVID-19) has constituted a global pandemic and is caused by severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2). COVID-19 has been the most recent coronavirus-caused pandemic following SARS and Middle East respiratory syndrome reported in the last two decades. The first COVID-19 radiological images were published from early cases and documented classical radiological patterns; thereafter, further diagnostic trends were identified.[1] Although the reverse transcriptase polymerase chain reaction (RT-PCR) test is the gold standard for confirming COVID-19, high-resolution computed tomography (HRCT) of the thorax has played a role in assessing COVID-19 cases in indoor and outdoor settings. Initially, radiology experts observed a limited role of HRCT imaging in COVID-19 in the presence of atypical opacities because they were similar to opacities caused by other respiratory infections that can lead to acute lung injury. However, the sensitivity of HRCT has been shown to be higher than that of RT-PCR.[2,3]

Typical COVID-19 lung parenchymal involvement is predominantly ground glass opacities (GGOs) and consolidations in the peripheral and subpleural portions of any lobe, but predominantly in the lower lobes.[4] Atypical radiological patterns in COVID-19 have been documented as bronchopneumonia, multifocal consolidations, necrotizing pneumonia, and cavitations with GGOs with or without consolidations.[5] The opacities are usually GGOs, sometimes with areas of consolidation, and are often nodular or mass-like, resembling an organizing pneumonia pattern.[6,7]

Pan et al. described four temporal stages of acute and subacute COVID-19 on CT images. In the initial phase, abnormalities manifest as GGOs, which can be unilateral and tend to lack the characteristic peripheral lung distribution. Patients often experience progression from days 5 to 8, during which pulmonary opacities become more extensive and confluent, often with bilateral lung involvement.[8] The peak stage occurs around 9–13 days and features more extensive consolidation, paralleling the evolution of acute lung injury.[9] This is consistent with chest radiograph abnormalities being most extensive 10–12 days after symptom onset.[10] Beginning at approximately 2 weeks, many patients enter the absorption stage, although this can vary. During this period, consolidation may wane, and other manifestations that were previously absent can emerge, such as linear opacities, a reverse halo sign, and a crazy-paving pattern.[11]

CT severity scoring is an accepted tool for assessing COVID-19 pneumonia in conjunction with laboratory and clinical parameters.[12] Many asymptomatic patients have parenchymal involvement on CT that is similar in severity with that of symptomatic patients, and the CT severity scores of severe COVID-19 pneumonia can overlap with those of moderate clinical severity, underscoring the limitations in drawing clinical conclusions from CT severity alone.[13,14]

Here, we investigated characteristic radiological phenotypes by categorizing them as either easy-to-treat or difficult-to-treat and correlated them with an accepted CT severity grading tool, clinical parameters, laboratory inflammatory markers, interventions required in indoor units, and outcomes.

Methods

Ethical approval

This study was approved by the Institutional Review Board (IRB)/Ethics Committee of the Venkatesh Hospital and Critical Care Center, Latur, India and by MIMSR Medical College, Latur, India (approval number: VCC/50-2020-2021; approval date November 12, 2020).

Data source

This prospective, observational, 24-week follow-up study was conducted from July 2020 to June 2021 at MIMSR Medical College Latur and Venkatesh Hospital in Latur, India. The study included 3000 COVID-19 cases confirmed by RT-PCR who were admitted to the critical care unit. The primary objective was to investigate the role of easy-to-treat and difficult-to-treat radiological phenotypes and examine their correlations with radiological presentations. As secondary objectives, we also investigated the use of radiological phenotype categorization in predicting illness severity, assessing the response to therapy, and its association with the outcomes of post-COVID fibrosis or sequelae. In total, 3000 cases were enrolled in the study after IRB approval, and written informed consent of all included cases was obtained at either Venkatesh Hospital or MIMSR Medical College Latur. [Figure 1].

F1-4
Figure 1:
Flowchart of the study

Inclusion criteria

COVID-19 cases confirmed by RT-PCR who were older than 18 years and were hospitalized in one of the study centers were included. Cases with comorbidities and all severity and oxygen saturation statuses were included.

Exclusion criteria

People unwilling to give consent, those not able to undergo thoracic HRCT upon hospital admission and those unwilling to remain in follow-up or undergo HRTC during follow-up were excluded. COVID-19 cases who died during hospitalization or within 24 weeks of hospital discharge were also excluded.

Methodology

Case definitions of the radiological phenotypes were formulated by a group of experts in the teaching faculties at the two tertiary care institutes:

  1. Easy-to-treat phenotype:
    1. Radiological criteria – predominant unilateral or bilateral consolidations that are peripheral and subpleural, with minimal or negligible GGOs. Consolidations that have the classical “limiting sign,” i.e. a line of demarcation between normal lucent and abnormal opaque lung parenchyma. [Figures 2 and 3]
    2. Laboratory criteria – abnormal (less than fourfold above the upper limit of the normal range) inflammatory markers such as interleukin (IL)-6, C-reactive protein (CRP), lactate dehydrogenase (LDH), ferritin, and D-dimer
    3. Clinical assessment and interventions required – hypoxia (oxygen saturation <90% in room air), tachypnea (respiratory rate >24 breaths/min), and tachycardia (heart rate >100 beats/min). These conditions were present in a few cases. A clinical improvement duration of <2 weeks. Very few cases required at-home oxygen supplementation to treat hypoxia at discharge
    4. Interventions required in cases with typical HRCT findings: oxygen supplementation and ventilatory support such as high flow nasal canula (HFNC), noninvasive ventilation (NIV), and invasive ventilatory support in indoor units. The majority of cases responded to oxygen supplementation with or without ventilatory support.
    F2-4
    Figure 2:
    Easy-to-treat phenotype showing bilateral consolidations
    F3-4
    Figure 3:
    Easy-to-treat phenotype showing consolidations with line of demarcation
  2. Difficult-to-treat phenotype:
    1. Radiological criteria – predominant GGOs, unilateral or bilateral, multifocal, peripheral, and subpleural with minimal or negligible consolidations. GGOs without the limiting sign, i.e. no line of demarcation between normal lucent and abnormal opaque lung parenchyma. Some of these cases had GGOs and interstitial marking due to thickened interlobular septa, called the crazy paving pattern. [Figures 4-6]
    2. Laboratory criteria – abnormal (more than fourfold above the upper limit of the normal range) inflammatory markers such as IL-6, CRP, LDH, ferritin, and D-dimer
    3. Clinical assessment – hypoxia (oxygen saturation <90% in room air), tachypnea (respiratory rate >24 breaths/min) and tachycardia (heart rate >100 beats/min). These conditions were present in the majority of cases. A clinical improvement duration of longer than 2 weeks. The majority of cases required at-home oxygen supplementation owing to hypoxia at discharge
    4. Interventions required in cases with typical HRCT findings – oxygen supplementation and ventilatory support such as HFNC, NIV, and invasive ventilatory support in indoor units. The majority of cases required oxygen supplementation with ventilatory support.
F4-4
Figure 4:
Difficult-to-treat phenotype showing GGOs and crazy paving. GGOs = Ground-glass opacities
F5-4
Figure 5:
Difficult-to-treat phenotype showing bilateral GGOs with dense consolidations. GGOs = Ground-glass opacities
F6-4
Figure 6:
Difficult-to-treat phenotype showing bilateral GGOs with consolidations with necrosis. GGOs = Ground-glass opacities

All study cases were tested following assessment but before enrolling in the study

COVID-19 RT-PCR testing was performed with nasopharyngeal samples collected in accordance with all standard institutional infection control policies. If the first test results were negative and radiological features clearly documented pneumonia, we repeated the RT-PCR test and enrolled all cases with a positive test result. Thoracic HRCT was conducted to assess the severity of lung involvement as per the COVID-19 Reporting and Data System[15] and categorized as follows: mild: <7, moderate: 8–15, and severe >15 (i.e. 15–25). A clinical assessment and routine biochemistry and hematological testing were conducted to assess inflammatory markers (CRP, ferritin, LDH, and IL-6 concentrations). The inflammatory markers were analyzed on hospital admission and repeated every 72 h to assess the response to treatment. Follow-up thoracic HRCT was conducted 24 weeks after hospital discharge to investigate the presence of post-COVID lung fibrosis or sequelae in cases with abnormal inflammatory markers at discharge, those who required bilevel positive airway pressure/NIV during hospitalization, and those who required oxygen supplementation at home.

Study design

  • A total of 290 cases were excluded (248 cases excluded because they did not attend the 24-week follow-up, and 42 cases died) [Figure 1]
  • HRCT phenotyping conducted as per criteria
  • HRCT phenotyping evaluation
  • Final assessment of easy-to-treat or difficult-to-treat phenotype.

Inflammatory marker analysis

The inflammatory markers were analyzed using a Rosch automated biochemistry analyzer. The inflammatory marker concentrations were considered statistically significant if they were fourfold greater than the upper limit of the normal range, as follows:

  1. CRP: 0–6 mg/L
  2. LDH: 90–470 mg/L
  3. Ferritin: 14–250 ng/ml for men; 6–160 ng/ml for women <45 years, and 5–200 ng/ml for women ≥45 years
  4. D-dimer: 70–470 mg/dl
  5. IL-6: 0–7 pg/ml

We correlated the inflammatory markers that had at least a fourfold increase above the normal value with radiological phenotypes and the interventions required during hospitalization.

Statistical analysis

The statistical analysis was conducted using the Chi-square test in R-3.4 software. Significant χ2 values were identified based on a probability table and depended on the required degrees of freedom. P values were considered statistically significant if they were <0.05 and highly significant if they were <0.001.

Results: Covariates

In this study of 3000 RT-PCR-confirmed COVID-19 pneumonia cases, there were 1835/3000 men and 1,165/3,000 women; 1285 cases were >50 years, and 1715 cases were <50 years. Significant associations were observed between the easy-to-treat and difficult-to-treat radiological phenotypes and the variables of age (P < 0.0042), gender (P < 0.0087), diabetes mellitus (P < 0.00019), Ischemic Heart Disease (IHD) (P < 0.00001), hypertension (P < 0.00001), Chronic Obstructive Pulmonary Disease (COPD) (P < 0.0006), and obesity (P < 0.0067). Similarly, the HRCT severity score at hospital admission was significantly correlated with the radiological phenotypes of COVID-19 pneumonia (P < 0.00001) [Table 1].

T1-4
Table 1:
Variables and C-reactive protein concentrations in coronavirus disease 2019 pneumonia cases (n=3000)

Results: Core observations

The easy-to-treat and difficult-to-treat radiological response phenotypes were observed in 20% and 80% of the sample, respectively [Table 2]. There was a significant association between the radiological phenotype and the duration of illness at hospital admission (P < 0.00001). The duration of illness (<7 days, 7–14 days, and >14 days) played a crucial role in predicting the radiological phenotype, i.e. as the duration of illness increased, we observed a shift from the easy to difficult phenotype [Table 3]. Laboratory parameters at hospital admission (d-dimer, CRP, IL-6) were significantly associated with the two radiological responses to treatment phenotypes (P < 0.00001) [Table 4], as were interventions required in indoor care units (P < 0.00001) [Table 5]. Post-COVID lung fibrosis or sequelae were also significantly associated with the radiological phenotypes (P < 0.00001) [Table 6].

T2-4
Table 2:
Radiological response to treatment phenotypes
T3-4
Table 3:
Duration of illness upon hospital admission in coronavirus disease 2019 pneumonia cases (n=3000)
T4-4
Table 4:
Association between radiological response to treatment phenotypes and inflammatory markers
T5-4
Table 5:
Association between radiological response to treatment phenotypes and supplemental oxygen requirements
T6-4
Table 6:
Association between radiological phenotype at hospital admission and postcoronavirus disease lung fibrosis (n=622)

Discussion

Easy-to-treat and difficult-to-treat phenotypes

The easy-to-treat and difficult-to-treat radiological phenotypes were documented in 20% and 80% of cases, respectively. Clinical and laboratory parameters were used to define these patterns, such as predominant GGOs, crazy paving with minimal consolidation without a line of demarcation between normal and abnormal lung in the difficult-to-treat phenotype, and predominant consolidation without GGOs or crazy paving with a line of demarcation between normal and abnormal lung in the easy-to-treat phenotype. We observed the important role of thoracic HRCT in the early diagnosis of COVID-19 pneumonia by analyzing GGOs in the early stage of illness and in cases with a duration of illness longer than 2 weeks. In both scenarios, the chance of a false-negative RT-PCR result is high, and in those cases, we repeated the RT-PCR test to confirm the diagnosis. CT is a sensitive diagnostic tool, but it has limited specificity.[2,3] However, the CT findings can help in assessing disease severity in conjunction with inflammatory markers and other clinical parameters. Many studies have documented the role of thoracic CT in predicting severity.[15,16,17,18,19,20] Our study found that the anatomical extent of involvement correlates well with the disease severity and laboratory parameters.[6,18,19,20,21,22,23,24,25,26]

Correlation of the easy-to-treat and difficult-to-treat phenotypes with the duration of illness

There was a significant association between the radiological phenotype and duration of illness at hospital admission (P < 0.00001). The duration of illness (<7 days, 7–14 days, and >14 days) plays a crucial role in predicting the radiological phenotype, i.e. as the duration of illness increased, we observed a shift from the easy to difficult phenotype. There was a temporal association between the radiological pattern and the duration of illness. In the early phase of the disease, the predominant radiological abnormalities were GGOs; in the later phase, the GGOs progressed and evolved into consolidations.[27] We observed that as the duration of illness increased, the radiological pattern showed more consolidations and fewer GGOs. Other patterns were reticulations characteristically described as interlobular and intralobular septal thickenings,[16] and some showed crazy paving, which is characteristically described as thickened interlobular septa with minimal GGOs.[27]

We observed a surprising distribution of illness duration in the easy-to-treat and difficult-to-treat phenotypes. As the duration of illness increased, there was typically more consolidation in the imaging or more resolved GGOs and consolidations in the easy-to-treat phenotype than in the difficult-to-treat phenotype. Of the 600 cases with the easy-to-treat phenotype, 251/600 (41.83%) were in their 1st week of illness, 90/600 (15%) were in their 2nd week and 259/600 (43.16%) were in their 3rd week. For cases with the difficult-to-treat radiological phenotype, 1311/2400 (54.62%) were in their 1st week of illness, 792/2,400 (33%) were in their 2nd week, and 297/2400 (12.37%) were in their 3rd week. In this study, the duration of illness was defined from symptom onset reported by the patient and/or attendants. Many studies have reported similar observations to ours and have found that GGOs with crazy paving with or without consolidation are a predictor of severe pneumonia and indicate advanced disease.[28,29,30] Previous studies of postmortem examinations have described the hyaline membrane as a probable mechanism of GGOs and have identified GGOs as a marker of advanced disease with a higher risk of mortality and a poor outcome.[31] One study identified GGOs with consolidations as a marker of organizing pneumonia with acute respiratory distress syndrome.[1]

Interestingly, the majority of cases in the 2nd week of illness (days 7–14) were easy to treat, which was in agreement with the inflammatory markers and clinical parameters. This may be because the immune dysregulation due to COVID-19 manifested as GGOs that were predominantly observed in the 1st week of illness and changed to consolidations with minimal GGOs during the 2nd and 3rd week of illness, and this change correlated with the inflammatory markers monitored during hospitalization in indoor units. In the 3rd week of illness, with the evolution of COVID-19 pneumonia, the pattern showed more consolidations without or with minimal GGOs. During the 3rd week, 43.16% of cases showed the easy-to-treat phenotype and 12.37% showed the difficult-to-treat phenotype. We observed that cases with the difficult-to-treat phenotype in the 3rd week showed more lung parenchymal involvement with GGOs plus consolidation with lucencies or breakdown suggestive of necrotizing pneumonia and were more likely to require ventilator support and aggressive interventions in indoor units. Previous studies have observed more GGOs with consolidation and crazy paving in advanced cases later in the course of illness, and these signs are an indicator of a poor outcome.[32] Several studies have documented observations similar to ours and have reported that GGOs with consolidations and a 1–3 week duration of illness indicate severe disease.[33,34] Crazy paving is found in 5%–30% of COVID-19 cases and is a hallmark of the peak of disease, with consolidations and GGOs being markers of severe disease.[11,35]

Easy-to-treat and difficult-to-treat phenotypes and their correlations with inflammatory markers during hospitalization

Inflammatory parameters upon hospital admission (CRP, IL-6, ferritin, LDH, and D-dimer) were significantly associated with the easy-to-treat and difficult-to-treat radiological response to treatment phenotypes (P < 0.00001). There was a fourfold increase in inflammatory markers in 448/600 (74.66%) of cases with the easy-to-treat phenotype and in 2302/2400 (95.91%) of cases with the difficult-to-treat phenotype. The fourfold increase in laboratory parameters in the difficult-to-treat cases was mainly because of the greater amount of GGOs and other features such as more anatomical involvement and the associated consolidation with necrosis. Previous studies have found that inflammatory markers (CRP, ferritin, LDH, IL-6, and D-dimer) are positively correlated with the radiological pattern, severity assessment, interventions required, and outcomes in COVID-19 pneumonia.[36,37,38,39,40]

There were 152/600 (25.33%) cases with the easy-to-treat phenotype and 98/2400 (4.08%) cases with the difficult-to-treat phenotype with a less than fourfold increase in inflammatory markers (CRP, IL-6, ferritin, LDH, and D-dimer) at hospital admission. The inflammatory marker concentrations were proportional to inflammation. Greater inflammation owing to GGOs in the difficult-to-treat phenotype and less inflammation and minimal GGOs in the easy-to-treat phenotype probably led to the observed inflammatory patterns. Similarly, previous studies have reported the role of inflammatory markers in association with CT findings in predicting the severity of COVID-19 pneumonia.[41,42,43,44]

Few cases (4.08%) with the difficult-to-treat phenotype had a less than fourfold increase in inflammatory markers. This might have resulted from a lower level of ongoing inflammation in the presence of early GGOs. One study found a similar observation and reported a modest increase in inflammatory markers, which is an indicator of controlled inflammation.[45]

Easy-to-treat and difficult-to-treat phenotypes and their correlation with interventions required during hospitalization

There were significant associations between the radiological response to treatment phenotypes and interventions required in indoor units (P < 0.00001). In the easy-to-treat and difficult-to-treat phenotypes, the interventions required during hospitalization were oxygen in 17.5% (105/600) and 43.29% (1039/2400) of cases, oxygen plus ventilatory support in 3.33% (20/600) and 53.58% (1286/2400) of cases, and no oxygen or ventilatory support in 79.16% (475/600) and 3.21% (75/2400) of cases, respectively (P < 0.00001). Previously published data support our results because CT findings and inflammatory markers have been correlated with interventions required in indoor units.[36,37,38,39,40]

Compared with the easy-to-treat phenotype cases, more cases with the difficult-to-treat phenotype required oxygen or oxygen plus ventilatory support because they had more GGOs with a crazy paving pattern and higher levels of inflammatory factors, resulting in greater levels of hypoxia. The HRCT results correlated with more than fourfold increases in inflammatory markers. Few cases with the difficult-to-treat phenotype did not require oxygen or ventilatory support. In these cases, the early GGOs and less than fourfold increases in inflammatory markers resulted from controlled inflammation and less oxygen diffusion deficiency manifesting as hypoxia. Many previous studies have reported similar findings.[46,47,48]

The majority of cases (79.16%) with the easy-to-treat pneumonia phenotype did not require either oxygen or oxygen plus ventilatory support. This was probably because more of the consolidations were either without or with minimal GGOs or crazy paving, and many had a less than fourfold increase in inflammatory markers, i.e. these cases had less inflammation and less oxygen diffusion issues manifesting as hypoxia. However, a few cases (3.33%) required ventilatory support, and some (17.5%) required oxygen support; these cases were also correlated with inflammatory markers and the presence of minimal GGOs with crazy paving. Similarly, previous studies have reported the role of CT findings in predicting interventions required in intensive care units and their correlation with inflammatory markers.[46,47,48]

Correlation of computed tomography severity with easy-to-treat and difficult-to-treat phenotypes

The HRCT severity score at hospital admission was significantly correlated with the radiological phenotypes (P < 0.00001). An HRCT severity score of mild was found in 190/600 (31.66%) of cases with the easy-to-treat phenotype and 210/2,400 (8.75%) cases with the difficult-to-treat phenotype; a moderate score was found in 192/600 (32%) of cases with the easy-to-treat phenotype and 1258/2400 (52.41%) cases with the difficult-to-treat phenotype; and a severe score was found in 218/600 (36.33%) cases with the easy-to-treat phenotype and 932/2400 (38.83%) cases with the difficult-to-treat phenotype. Previous studies have reported a correlation between the CT severity and inflammatory markers, interventions in indoor units, and GGOs and consolidations with or without crazy paving.[36,37,38,39,40]

We observed that the universally accepted HRCT severity scoring tool based on the topographical involvement of lung segments in bilateral lung lobes is not suited for routine care in indoor units. An equal proportion of COVID-19 cases with the easy-to-treat (36.33%) and difficult-to-treat (38.83%) radiological phenotypes had a severe score, i.e. CT severity >15. Thus, there were disproportionate inflammatory patterns, durations of illness, oxygen statuses, and interventions required between the easy-to-treat and difficult-to-treat categories despite having the same CT severity score. Thus, radiological phenotyping is superior to CT severity scoring, as it will be more accurate in the overall assessment of severe COVID-19 cases.

In the mild CT category (CT score <7), the severity assessment was significantly correlated with the easy-to-treat and difficult-to-treat radiological phenotypes. Because less anatomical area is involved in the mild category, a large number of these cases had the easy-to-treat phenotype (31.66%), while only 8.75% had the difficult-to-treat phenotype. This might have been because there was less inflammation owing to less lung parenchymal anatomical area involvement. Nevertheless, a few cases had significantly increased inflammatory markers with hypoxia and required oxygen supplementation with or without ventilatory support. This may be explained by GGOs being in the early stage and reported as mild on the CT severity tool. Previous studies have reported similar findings.[1,46,47,48]

In the moderate CT category (CT score 7–15), there were more cases with the difficult-to-treat phenotype (52.41%) than the easy-to-treat phenotype (32%), and the CT severity was significantly correlated with these radiological phenotypes. This could be explained by more GGOs, crazy paving with or without consolidation, and necrosis in the difficult-to-treat cases and fewer GGOs with predominant consolidation and without crazy paving in the easy-to-treat cases. We further analyzed the duration of illness, inflammatory markers, and oxygen interventions required with or without ventilatory support in indoor units. Previous studies have documented observations similar to ours.[46,47,48]

During the pandemic, the CT severity scoring tool helped us during the initial assessment and triage of cases upon hospitalization and during treatment planning. However, when performed upon hospitalization, this methodology was unable to predict the disease course during hospitalization, the interventions required during care in indoor units, and final outcomes in the easy-to-treat and difficult-to-treat radiological phenotypes. The CT severity scoring tool primarily depends on anatomical involvement of the lung lobes and segments, and this severity was not correlated with the duration of illness, inflammatory markers, interventions required in indoor units or final outcomes. Our observations were similar to those of Pan et al.[8] who documented the importance of radiological patterns such as GGOs and consolidations with or without crazy paving in predicting disease severity, interventions required and final outcomes.

Does the radiological phenotype predict long-term radiological outcomes?

We found that post-COVID lung fibrosis or sequelae were significantly associated with the radiological phenotype (P < 0.00001). Lung fibrosis was found in 10.83% (105/600) of cases with the easy-to-treat phenotype and in 23.20% (557/2400) of cases with the difficult-to-treat phenotype. We assessed the cases who required oxygen supplementation for more than 1 week during hospitalization (such as ventilatory support or high-flow nasal cannula) and those who required at-home oxygen and/or NIV for post-COVID lung sequelae or fibrosis at 24 weeks after hospital discharge. Studies have reported similar observations that post-COVID fibrosis is correlated with CT severity, inflammatory markers, and interventions required in indoor units.[36,37,38,39,40]

Compared with the easy-to-treat phenotype, post-COVID sequelae were more common in the difficult-to-treat phenotype (23.20%; 557/2400). A retrospective analysis of radiological patterns showed a considerable amount of post-COVID lung fibrosis. Radiological patterns with more GGOs, crazy paving with or without consolidation and necrosis were predominantly associated with post-COVID lung fibrosis. The indoor unit records showed that these cases often had a greater than fourfold increase in inflammatory markers. A retrospective analysis of the cases with post-COVID lung fibrosis revealed that nearly all required ventilatory support to treat hypoxia. Another reason for the post-COVID sequelae in these cases is that hypoxia and inflammation typically coexist, and they work synergistically to release profibrogenic inflammatory markers such as hypoxia-induced transcription factors, which can be indirectly assessed as increased LDH concentrations. Previous studies have reported similar observations.[8,49,50,51]

Post-COVID sequelae were less common in those with the easy-to-treat phenotype (10.83%; 105/600). A retrospective analysis of these cases found more consolidation and fewer GGOs, which was well correlated with the inflammatory markers and interventions required during hospitalization. We further analyzed these cases and found hypoxia and aggressive interventions in cases with typical consolidation without a crazy paving pattern and with a more than fourfold increase in inflammatory markers. Similarly, previous studies have documented that inflammatory markers and patterns of lung disease on CT imaging can predict post-COVID sequelae.[8,49,50,51,52]

Association of covariates with radiological phenotypes

In the present study, 61.16% (1835/3000) of the study population were men and 38.83% (1165/3000) were women. Those older than 50 years constituted 42.83% (1285/3000) of cases, and those younger than 50 constituted 57.16% (1715/3000). We found that age and gender were significantly associated with the radiological phenotype. In the easy-to-treat phenotype group, 56.5% (339/600) were men and 43.5% (261/600) were women; in the difficult-to-treat phenotype group, 62.33% (1496/2400) were men and 37.66% (904/2400) were women (P < 0.0042). It is unknown why men are more susceptible to COVID-19 illness. Accordingly, it is also unknown why there were more men in the difficult-to-treat phenotype group; whether it was by chance or because of hormonal factors that protect women from severe COVID-19 illness requires further research. Previous studies have documented similar observations.[36,37,38,39,40,49,50,51,52]

There were significant associations between covariates and the easy-to-treat and difficult-to-treat phenotypes. The distribution of the easy-to-treat and difficult-to-treat phenotypes was examined according to the presence and absence of comorbidities. Diabetes mellitus was present in 18.33% (110/600) and absent in 81.66% (490/600) of easy-to-treat cases, and was present in 25.87% (621/2400) and absent in 74.12% (1779/2400) of difficult-to-treat cases (P < 0.00019). Ischemic Heart Disease (IHD) was present in 15.16% (91/600) and absent in 84.83% (509/600) of easy-to-treat cases, and was present in 8.62% (207/2400) and absent in 91.37% (2193/2400) of difficult-to-treat cases (P < 0.00001). Hypertension was present in 16.66% (100/600) and absent in 83.33% (500/600) of easy-to-treat cases, and was present in 8.75% (210/2400) and absent in 91.25% (2190/2400) of difficult-to-treat cases (P < 0.00001). COPD was present in 1.5% (9/600) and absent in 98.5% (591/600) of easy-to-treat cases, and was present in 4.54% (109/2400) and absent in 95.45% (2291/2400) of difficult-to-treat cases (P < 0.0006). Finally, obesity was present in 9.83% (59/600) and absent in 90.16% (541/600) of easy-to-treat cases, and was present in 6.62% (159/2400) and absent in 93.37% (2241/2400) of difficult-to-treat cases (P < 0.0067). Several studies have documented similar observations, and the observed covariates were predictors of poor outcomes because they were associated with more GGOs and crazy paving associated with COVID-19 pneumonia.[36,37,38,39,40,49,50,51,52]

Study limitations

Our study had a sufficient sample size to analyze the correlations between the radiological phenotypes and laboratory markers and management in indoor units. However, there are some study limitations. First, confounding factors that may have led to increased inflammatory markers were not analyzed. Second, the duration of illness may have been inexact because it was based on information from the patients and/or their attendants. Third, a multivariate analysis using individual inflammatory markers was not conducted to investigate their correlations with the radiological phenotypes. Fourth, while we considered the interventions used (such as oxygen and ventilatory support) at hospital admission and during hospitalization in the indoor unit, we considered thoracic CT findings only at hospital admission. We did not perform repeated thoracic CT to examine CT findings during hospitalization owing to cost constraints. Thus, trends in the timing of interventions and CT findings were not assessed. Finally, post-COVID radiological outcomes were examined only for those cases who required aggressive indoor unit interventions or who required at-home oxygen supplementation after hospital discharge (with or without comorbidities).

Learning points

  1. Radiological patterns or phenotypes have an important role in assessing disease severity in COVID-19 pneumonia. Easy-to-treat and difficult-to-treat phenotypes help to triage cases at hospital admission and are correlated with inflammatory marker concentrations. Phenotypic categorization is simple, sensitive, and can help guide treatment planning in indoor units
  2. The easy-to-treat phenotype, in conjunction with laboratory and clinical parameters, can predict a reduced need for aggressive interventions compared with the difficult-to-treat phenotype. Radiological phenotyping can be used to assess the state of COVID-19 pneumonia at hospital admission. The presence or absence of GGOs, consolidations and crazy paving with necrosis are key radiological markers when categorising these phenotypes
  3. Radiological phenotyping should be used in conjunction with clinical and laboratory parameters to obtain an accurate severity assessment and to predict the illness duration. Phenotyping will also help in monitoring COVID-19 pneumonia cases and guide necessary and timely interventions in indoor units to achieve the best possible treatment outcomes. Post-COVID fibrosis is reversible and should be labeled as a sequela owing to its near-total reversible nature.

Conclusion

Radiological phenotyping can play a crucial role in the initial assessment of COVID-19 pneumonia. The differentiation between easy-to-treat and difficult-to-treat phenotypes is an important step in managing these cases in indoor and outdoor settings. Phenotypic differentiation helps in the initial assessment and can be used to guide treatment planning and targeted interventions in intensive care units and in predicting treatment outcomes.

Financial support and sponsorship

Nil.

Conflicts of interest

There are no conflicts of interest.

Acknowledgment

We thank Katherine Thieltges from Liwen Bianji (Edanz) (http://www.liwenbianji.cn/) for editing the English text of a draft of this manuscript.

REFERENCES

1. Kanne JP. Chest CT findings in 2019 novel coronavirus (2019-nCoV) infections from Wuhan, China: Key points for the radiologist Radiology. 2020;295:16–7
2. Ai T, Yang Z, Hou H, Zhan C, Chen C, Lv W, et al Correlation of chest CT and RT-PCR testing for coronavirus disease 2019 (COVID-19) in China: A report of 1014 cases Radiology. 2020;296:E32–40
3. Fang Y, Zhang H, Xie J, Lin M, Ying L, Pang P, et al Sensitivity of chest CT for COVID-19: Comparison to RT-PCR Radiology. 2020;296:E115–7
4. Kanne JP, Bai H, Bernheim A, Chung M, Haramati LB, Kallmes DF, et al COVID-19 imaging: What we know now and what remains unknown Radiology. 2021;299:E262–79
5. Patil S, Gondhali G, Patil R. Radiological phenotypes' in covid-19 pneumonia: Retrospective, multicentric study of 6000 cases in tertiary care setting in India Chest. 2022;161:A537
6. Chung M, Bernheim A, Mei X, Zhang N, Huang M, Zeng X, et al CT imaging features of 2019 novel coronavirus (2019-nCoV) Radiology. 2020;295:202–7
7. Simpson S, Kay FU, Abbara S, Bhalla S, Chung JH, Chung M, et al Radiological Society of North America Expert Consensus statement on reporting chest CT findings related to COVID-19. Endorsed by the Society of Thoracic Radiology, the American College of Radiology, and RSNA – Secondary Publication J Thorac Imaging. 2020;35:219–27
8. Pan F, Ye T, Sun P, Gui S, Liang B, Li L, et al Time course of lung changes at chest CT during recovery from coronavirus disease 2019 (COVID-19) Radiology. 2020;295:715–21
9. Bernheim A, Mei X, Huang M, Yang Y, Fayad ZA, Zhang N, et al Chest CT findings in coronavirus disease-19 (COVID-19): Relationship to duration of infection Radiology. 2020;295:200463.
10. Wong HY, Lam HY, Fong AH, Leung ST, Chin TW, Lo CS, et al Frequency and distribution of chest radiographic findings in patients positive for COVID-19 Radiology. 2020;296:E72–8
11. Pan Y, Guan H, Zhou S, Wang Y, Li Q, Zhu T, et al Initial CT findings and temporal changes in patients with the novel coronavirus pneumonia (2019-nCoV): A study of 63 patients in Wuhan, China Eur Radiol. 2020;30:3306–9
12. Xu PP, Tian RH, Luo S, Zu ZY, Fan B, Wang XM, et al Risk factors for adverse clinical outcomes with COVID-19 in China: A multicenter, retrospective, observational study Theranostics. 2020;10:6372–83
13. Li K, Chen D, Chen S, Feng Y, Chang C, Wang Z, et al Predictors of fatality including radiographic findings in adults with COVID-19 Respir Res. 2020;21:146.
14. Zhang R, Ouyang H, Fu L, Wang S, Han J, Huang K, et al CT features of SARS-CoV-2 pneumonia according to clinical presentation: A retrospective analysis of 120 consecutive patients from Wuhan city Eur Radiol. 2020;30:4417–26
15. Liu KC, Xu P, Lv WF, Qiu XH, Yao JL, Gu JF, et al CT manifestations of coronavirus disease-2019: A retrospective analysis of 73 cases by disease severity Eur J Radiol. 2020;126:108941.
16. Zhang N, Xu X, Zhou LY, Chen G, Li Y, Yin H, et al Clinical characteristics and chest CT imaging features of critically ill COVID-19 patients Eur Radiol. 2020;30:6151–60
17. Xie X, Zhong Z, Zhao W, Zheng C, Wang F, Liu J. Chest CT for typical coronavirus disease 2019 (COVID-19) pneumonia: Relationship to negative RT-PCR testing Radiology. 2020;296:E41–5
18. Zhao W, Zhong Z, Xie X, Yu Q, Liu J. Relation between chest CT findings and clinical conditions of coronavirus disease (COVID-19) pneumonia: A multicenter study AJR Am J Roentgenol. 2020;214:1072–7
19. Hani C, Trieu NH, Saab I, Dangeard S, Bennani S, Chassagnon G, et al COVID-19 pneumonia: A review of typical CT findings and differential diagnosis Diagn Interv Imaging. 2020;101:263–8
20. Luo N, Zhang H, Zhou Y, Kong Z, Sun W, Huang N, et al Utility of chest CT in diagnosis of COVID-19 pneumonia Diagn Interv Radiol. 2020;26:437–42
21. Wang J, Xu Z, Wang J, Feng R, An Y, Ao W, et al CT characteristics of patients infected with 2019 novel coronavirus: Association with clinical type Clin Radiol. 2020;75:408–14
22. Guo L, Wei D, Zhang X, Wu Y, Li Q, Zhou M, et al Clinical features predicting mortality risk in patients with viral pneumonia: The MuLBSTA Score Front Microbiol. 2019;10:2752.
23. Yuan M, Yin W, Tao Z, Tan W, Hu Y. Association of radiologic findings with mortality of patients infected with 2019 novel coronavirus in Wuhan, China PLoS One. 2020;15:e0230548.
24. Wan S, Yi Q, Fan S, Lv J, Zhang X, Guo L, et al Relationships among lymphocyte subsets, cytokines, and the pulmonary inflammation index in coronavirus (COVID-19) infected patients Br J Haematol. 2020;189:428–37
25. Wu J, Wu X, Zeng W, Guo D, Fang Z, Chen L, et al Chest CT findings in patients with coronavirus disease 2019 and its relationship with clinical features Invest Radiol. 2020;55:257–61
26. Salehi S, Abedi A, Balakrishnan S, Gholamrezanezhad A. Coronavirus disease 2019 (COVID-19): A systematic review of imaging findings in 919 patients AJR Am J Roentgenol. 2020;215:87–93
27. Ye Z, Zhang Y, Wang Y, Huang Z, Song B. Chest CT manifestations of new coronavirus disease 2019 (COVID-19): A pictorial review Eur Radiol. 2020;30:4381–9
28. Han X, Cao Y, Jiang N, Chen Y, Alwalid O, Zhang X, et al Novel coronavirus disease 2019 (COVID-19) pneumonia progression course in 17 discharged patients: Comparison of clinical and thin-section computed tomography features during recovery Clin Infect Dis. 2020;71:723–31
29. Wu X, Liu X, Zhou Y, Yu H, Li R, Zhan Q, et al 3-month, 6-month, 9-month, and 12-month respiratory outcomes in patients following COVID-19-related hospitalisation: A prospective study Lancet Respir Med. 2021;9:747–54
30. Zhang S, Bai W, Yue J, Qin L, Zhang C, Xu S, et al Eight months follow-up study on pulmonary function, lung radiographic, and related physiological characteristics in COVID-19 survivors Sci Rep. 2021;11:13854.
31. Xu Z, Shi L, Wang Y, Zhang J, Huang L, Zhang C, et al Pathological findings of COVID-19 associated with acute respiratory distress syndrome Lancet Respir Med. 2020;8:420–2
32. Li K, Wu J, Wu F, Guo D, Chen L, Fang Z, et al The clinical and chest CT features associated with severe and critical COVID-19 pneumonia Invest Radiol. 2020;55:327–31
33. Hansell DM, Bankier AA, MacMahon H, McLoud TC, Müller NL, Remy J. Fleischner Society: Glossary of terms for thoracic imaging Radiology. 2008;246:697–722
34. Shi H, Han X, Jiang N, Cao Y, Alwalid O, Gu J, et al Radiological findings from 81 patients with COVID-19 pneumonia in Wuhan, China: A descriptive study Lancet Infect Dis. 2020;20:425–34
35. Wong KT, Antonio GE, Hui DS, Lee N, Yuen EH, Wu A, et al Thin-section CT of severe acute respiratory syndrome: Evaluation of 73 patients exposed to or with the disease Radiology. 2003;228:395–400
36. Patil S, Gondhali G, Acharya A. Serial CRP (C-reactive protein) monitoring in COVID-19 pneumonia for the assessment of severity, ventilatory support requirement and predicting early lung fibrosis J Med. 2022;23:112–20
37. Patil S, Bhadake M, Narwade G, Patil R. Correlation of LDH with duration of illness, disease severity, ventilatory support and lung fibrosis in covid-19 pneumonia: A single center experience of 1000 cases in tertiary care setting in India Ital J Emerg Med. 2022;11:95–103
38. Patil S, Acharya A, Gondhali G, Narwade G. Serial interleukin-6 titer monitoring in COVID-19 pneumonia: Valuable inflammatory marker in the assessment of severity, predicting ventilatory support requirement, and final radiological outcome – Prospective observational study in tertiary care setting in India J Assoc Pulmonologist Tamilnadu. 2022;5:2–8
39. Patil S, Gondhali G, Acharya A. “Serial ferritin titer” monitoring in COVID-19 pneumonia: Valuable inflammatory marker in assessment of severity and predicting early lung fibrosis – Prospective, multicentric, observational, and interventional study in tertiary care setting in India Egypt J Intern Med. 2022;34:75.
40. Patil S, Acharya A, Gondhali G, Narwade G. Role of 'serial D-dimer level' in predicting severity and outcome in COVID-19 pneumonia: A prospective multicentric observational study of 1000 cases in tertiary care setting in India EJMA. 2022;2:73–80
41. Wang Y, Dong C, Hu Y, Li C, Ren Q, Zhang X, et al Temporal changes of CT findings in 90 patients with COVID-19 pneumonia: A longitudinal study Radiology. 2020;296:E55–64
42. Ojha V, Mani A, Pandey NN, Sharma S, Kumar S. CT in coronavirus disease 2019 (COVID-19): A systematic review of chest CT findings in 4410 adult patients Eur Radiol. 2020;30:6129–38
43. Di Meglio L, Carriero S, Biondetti P, Wood BJ, Carrafiello G. Chest imaging in patients with acute respiratory failure because of coronavirus disease 2019 Curr Opin Crit Care. 2022;28:17–24
44. Li J, Yan R, Zhai Y, Qi X, Lei J. Chest CT findings in patients with coronavirus disease 2019 (COVID-19): A comprehensive review Diagn Interv Radiol. 2021;27:621–32
45. Saeed GA, Gaba W, Shah A, Al Helali AA, Raidullah E, Al Ali AB, et al Correlation between chest CT severity scores and the clinical parameters of adult patients with COVID-19 pneumonia Radiol Res Pract. 2021;2021:6697677.1–7
46. Ackermann M, Verleden SE, Kuehnel M, Haverich A, Welte T, Laenger F, et al Pulmonary vascular endothelialitis, thrombosis, and angiogenesis in covid-19 N Engl J Med. 2020;383:120–8
47. Xie J, Covassin N, Fan Z, Singh P, Gao W, Li G, et al Association between hypoxemia and mortality in patients with COVID-19 Mayo Clin Proc. 2020;95:1138–47
48. Nuevo-Ortega P, Reina-Artacho C, Dominguez-Moreno F, Becerra-Muñoz VM, Ruiz-Del-Fresno L, Estecha-Foncea MA, et al Prognosis of COVID-19 pneumonia can be early predicted combining Age-adjusted Charlson Comorbidity Index, CRB score and baseline oxygen saturation Sci Rep. 2022;12:2367.
49. Yu M, Liu Y, Xu D, Zhang R, Lan L, Xu H. Prediction of the development of pulmonary fibrosis using serial thin-section CT and clinical features in patients discharged after treatment for COVID-19 pneumonia Korean J Radiol. 2020;21:746–55
50. Patil SV, Gondhali G, Patil R. Post-covid-19 lung fibrosis: Study of 600 cases in tertiary care setting in India Eur Respir J. 2021;58:PA3776
51. Antonio GE, Wong KT, Hui DS, Wu A, Lee N, Yuen EH, et al Thin-section CT in patients with severe acute respiratory syndrome following hospital discharge: Preliminary experience Radiology. 2003;228:810–5
52. Zhang D, Zhang C, Li X, Zhao J, An C, Peng C, et al Thin-section computed tomography findings and longitudinal variations of the residual pulmonary sequelae after discharge in patients with COVID-19: A short-term follow-up study Eur Radiol. 2021;31:7172–83
Keywords:

Coronavirus disease 2019 pneumonia; difficult to treat; easy to treat; inflammatory marker; radiological phenotype

© 2023 Radiology of Infectious Diseases | Published by Wolters Kluwer – Medknow