Chest CT manifestations with emphasis on the role of CT scoring and serum ferritin/lactate dehydrogenase in prognosis of coronavirus disease 2019 (COVID-19)

Background In March 2020, the World Health Organization announced coronavirus disease 2019 (COVID-19) a pandemic, and because of the primary pulmonary manifestations of the disease, chest CT is essential in the evaluation of those patients. The aim of the study was to evaluate the role of chest CT findings and chest CT scoring along with serum ferritin and LDH in the prognosis of COVID-19 patients in a cohort of the Egyptian population. Results This retrospective study included 250 patients with positive RT-PCR for COVID-19, 138 males [55.2%] and 112 females [44.8%], age range 17–82 years with median 49.5. Two hundred patients had a positive significant correlation between age, serum ferritin, serum LDH, and CT score. Bilateral affection was 88% while unilaterality was 12%, and peripheral chest CT findings were stratified as follows: mild [score from 1 to 10], 114 patients [57%]; moderate [score from 11 to 19], 65 patients [32.5%]; and severe [score from 20 to 25], 21 patients [10.5%]. In severe cases, males constitute 85.7% while females were only 14.3%. Statistical and central distribution was 67%, peripheral was 31%, and central was 2%. Ground glass opacity (GGO) was the highest pattern 39.2%, consolidation 31.2%, fibrosis 15.2%, and CP 13.7%, with lymph nodes only 0.6%. Fifteen cases [6%] were critical; all showed severe scores ranging from 21 to 23 with three times increase in serum ferritin and four times increase in LDH. A follow-up study done to 8 cases [3.2%] showed an increase in CT scoring, serum ferritin, and serum LDH. Conclusion Chest CT findings are crucial for early diagnosis of COVID-19 disease especially for asymptomatic patients with old age and male sex considered risk factors for poor prognosis. Chest CT score, serum ferritin, and serum LDH help in predicting the short-term outcome of the patients aiming to decrease both morbidity and mortality.

Background A viral pneumonia of unknown etiology appeared in China in late December 2019, and the Chinese scientific centers declared that the cause is a new strain of coronavirus named severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), and the disease is named coronavirus disease 2019 (COVID-19) [1]. It is a highly contagious disease, and by March 2020, the World Health Organization declared it as a pandemic causing a global health emergency [2] making early diagnosis mandatory for proper management of the patients and rapid detection of contacts for quarantine purpose; real-time reverse transcriptasepolymerase chain reaction of the viral nucleic acid (RT-PCR) within body specimens mainly nasopharyngeal or oro-phangeal swab is the gold standard for diagnosis [3]. Many publications appeared mainly focusing on the diagnosis of the disease either comparing chest CT findings with RT-PCR results [4][5][6] or describing the disease imaging features mainly chest CT [7][8][9] with paucity of articles investigating the role of chest CT in disease prognosis regarding helping the physician to speed up the diagnostic work flow by properly selecting patients in need of intensive medical care, or only hospital ward care or just home quarantine. The purpose of the study was to investigate the role of chest CT findings and CT scores along with serum biomarkers namely serum ferritin and lactate dehydrogenase (LDH) in the proper effective management of COVID-19 patients.

Methods
This retrospective study was done in the university hospital which included 250 patients [138 males 55.2%, 112 females 44.8%, age range 17-82 years, median is 49.5] from first of April till 15 June 2020. The study was approved by the hospital's ethical committee. Inclusion criteria included patients with positive RT-PCR for COVID-19 in the nasopharyngeal or oro-phangeal swab; either having symptoms of viral pneumonia namely fever, cough, dyspnea, chest pain, or gastrointestinal symptoms mainly diarrhea and abdominal pain; or asymptomatic but with a history of contact to a positive COVID-19 patient. Exclusion criteria included any patient with chronic lung disease.

CT scanning protocol
All the patients underwent non-enhanced chest CT examination using either 16 or 128 multislice CT scanner. The scan was taken with the patient in the supine position, head first, from above the first rib till below the diaphragm during breath holding at end of inspiration with the following parameters: tube voltage 120 kV, adaptive tube current, pitch [0.3-1.5] with slice thickness and interslice gap both of 5 mm, 0 gantry tilt, and FOV depending on the size of the patient. Images were reconstructed in both coronal and sagittal planes and were sent to a compatible PACS system. CT images were evaluated by two experienced radiologists [more than 10 years of experience] independently and any disagreement was resolved by discussion. Axial images were reviewed along with the coronal and sagittal reformatted images on both lung window [window width 1500HU, window level 700HU] and mediastinal window [window width 350HU, window level 40HU] settings.

Image analysis
CT images were evaluated for the following features: pulmonary manifestation localization either involving one lung or both lungs; the distribution either peripheral/subpleural, central/broncho-vascular, or both; and lesion pattern that was described in line with the Fleischner Society glossary of terms for thoracic imaging (2008) [10] including ground glass opacity [GGO] which is an area of increased attenuation not obscuring the underlying vasculature, consolidation which is an area of increased attenuation obscuring underlying pulmonary vessels, crazy paving [CP] which is an area of ground glassing with interlobular septal thickening, or mediastinal or hilar lymph nodes [with short-axis diameter more than 10 mm]. Each patient was given a severity score using a semi-quantitative method for each lobe involvement as follows: 0, no involvement; 1, < 5% involvement; 2, 5-25% involvement; 3, 26-50% involvement; 4, 51-75% involvement; and 5, > 75% involvement. The total CT score was the sum of each individual lobar score and it ranges from 0 to 25 [11]; then a severity grading was given to each patient as follows: mild if score is from 1 to 10, moderate from 11 to 19, and severe from 20 to 25.

Laboratory markers
For each patient, serum ferritin and serum LDH were measured within 1 day of CT.

Follow-up
A follow-up chest CT was done to 8 patients in time interval ranging from 3 to 10 days from the first scan along with serum ferritin and serum LDH due to deterioration of the clinical condition.

Data collection
Epidemiological data, laboratory results, and CT results were obtained from the hospital information system [HIS] and image storage and transmission system [PACS].

Statistical analysis
IBM SPSS statistics (V. 26.0, IBM Corp., USA, 2019) was used for data analysis. Data were expressed as median and percentiles for quantitative non-parametric measures and both number and percentage for categorized data. Comparison between sex and different CT stages showed no difference in the mild stage, but in the moderate stage, females were 35.4% compared to 64.6% for males and in the severe stage, females were only 14.3% compared to 85.7% for males (Table 2); also, a comparison between age and CT scoring showed that there was a statistically significant correlation between the two factors using the Kruskal-Wallis test (Tables 3 and 4).   Regarding laterality of positive chest CT cases, bilateral affection was 88% while unilaterality was only 12% [7.5% right lung and 4.5% left lung] (Figs. 1 and 2). Regarding the distribution of lesions, both peripheral and central were the highest percent 67% while peripheral distribution alone was 31% (Fig. 1) and central was only 2% (Fig. 2).
There was a high statistical significance between both elevated serum ferritin and serum LDH and CT staging using the Kruskal-Wallis test where ferritin was increased in 18.4% in the mild stage, 63% in the moderate, and 100% in the severe stage (Table 6) while LDH was increased in 39.4%, 92.3%, and 100% mild, moderate, and severe stages, respectively (Table 7).
Fifteen cases [6%] were critical cases that needed ICU admission, and all were in the severe stage while all the negative chest CT patients, mild, moderate, and the rest  A follow-up chest CT, serum ferritin, and serum LDH were done for 8 cases [3.2%] in the interval of 3-10 days from the first chest CT due to deterioration of the clinical picture; their data were as follows: 4 cases of mild category, three of which became moderate and one remained mild but with a higher score; 3 moderate cases became severe; and one severe case remained in the severe category but with a higher score. All showed an increase in both serum ferritin and LDH compared to the first result (Table 9).

Discussion
COVID-19 is a new highly contagious infectious disease that causes lower respiratory inflammation; it starts by alveolar hyaline membrane injury or interstitial edema which appears on CT as GGO then progresses to alveolar filling with exudate which is the consolidation after which an autoimmune reaction begins by activation of both humeral and cellular immunity mediated by virus-specific B and T cells causing the formation of proinflammatory cytokines which are the cause of interlobular septal thickening, giving rise to first crazy paving then lung fibrosis [12].
The clinical picture of the disease is diverse ranging from asymptomatic or subclinical forms to critically ill patients with acute respiratory distress syndrome [ARDS] or multi-organ failure which can occur as short as 9 days from the onset of symptoms [9], so not only diagnosis of the disease is mandatory but also monitoring the clinical course of the disease is of paramount importance to triage patients who are in need of urgent intensive medical care. Extension of pulmonary affection is easily assessed by chest CT but this is not suitable for critically ill patients, so a laboratory biomarker of disease progression is needed and here comes the role of LDH and ferritin.
LDH is an enzyme that converts lactate to pyruvate in the process of tissue breaking down, so patients with severe COVID-19 pneumonia shows an increase in the serum LDH [13]; also, ferritin is increased because it is secreted either by the macrophages that produce cytokines and account for the main immune cells present in lung parenchyma or by cytokines themselves such as interleukin-6 [14].
From  [16] showed 56%; this big difference in the percentage between the current study/Yang and Yan [15] and Bernheim et al. [16] could be explained that the latter study was done only within the first 2 days from onset of symptoms in contrary to our study and Yang and Yan that were done within 1 week.  Thirty-seven [14.8%] patients were asymptomatic and showed positive chest CT findings; this clinicalradiological discrepancy was evident in previous studies like Chung et al. [17] which showed 14% while Inui et al. [18] which reported 39.2%; this was explained by a lot of theories, like those patients were exposed to COVID-19 virus before and hence developed immunity leading to a subclinical presentation, or they are having a recent infection and in the recovery period with subsiding of symptoms at the time of chest CT done but an absence of clear clinical history of infection and lacking CT features of healing COVID-19 pneumonia weaken this theory, and the only logical explanation is that there is a time interval between chest CT manifestation and developing symptoms with the former begins first but this hypothesis needs to be more investigated [17,18]; unfortunately, in the current study, none of those patients was followed up to see if they have developed symptoms or not.
Regarding the comparison between gender and different stages of COVID-19 pulmonary affection, there was no gender difference in the mild stage while in the moderate stage males represented 64.6% almost double of the females who were 35.4% while in the severe stage males were 85% compared to only 14.3% for females; this result was in concordance with Jin et al. [19] and Hiroki et al. [20] with the latter stating that male gender may be a predictor of a more severe form of COVID-19.
Regarding the age, there was a statistically significant correlation between older age and severity of chest CT manifestations as no patient in the severe stage was under 30 years while there was no difference between 31 and 50 years group and 51-82 years group showing 47.6% and 52.4%, respectively; this was in agreement with the US Centers for Disease Control and Prevention reporting that 8 in 10 COVID-19-related deaths have been among people aged 65 years or over [21].
Regarding the distribution of lung manifestations, unilateral lung affection was only observed in 24 [22].
Since the outbreak of the COVID-19 pandemic, many studies showed that the inflammatory cytokine storm is the main cause of complication of COVID-19 pneumonia leading to acute respiratory stress syndrome and multi-organ failure and even death, so laboratory detection of increased serum inflammatory markers is mandatory in order to provide those patients early proper treatment [23].
In the current study, there was a statistically high significance between increased levels of both serum ferritin  and LDH in correlation with CT staging, where ferritin was increased in 18.4% in the mild stage, 63% in the moderate, and 100% in the severe stage; these findings were in agreement with Lin et al. [24] who reported that patients with severe COVID-19 pneumonia showed higher levels of serum ferritin than the non-severe patients using a multivariate logistic regression analysis showing that the serum ferritin level was an independent risk factor for disease severity. LDH was increased in 39.4%, 92.3%, and 100% mild, moderate, and severe stages, respectively, and this was in concordance with Wu et al. [13] who reported that the higher LDH levels were found in patients with severe COVID-19. All the ICU patients showed a marked increase reaching up to more than four times the normal; these results were in agreement with the pooled study done by Henry et al. [25] who stated that increased LDH was associated with significantly increased odds of severe COVID-19 in both case-control studies and retrospective studies.
In spite that CT scoring is a subjective method for assessing the severity of COVID-19 pneumonia, it correlates well with the clinical manifestations and laboratory findings of the disease [26] helping to triage patients and reserving health care resources so only the ones with severe lung affection could be closely monitored in order to provide early aggressive treatment; in the current study, comparing the CT score for patients admitted to ICU and patients who were treated either in the hospital or at home, all the patients in ICU were in severe stage with CT scores ranging from 21 to 23, and this was in agreement with the study of 236 patients done by Colombi et al. [27] who reported a positive correlation between the extent of CT lung manifestations and intensive care unit admission or death. Also, Francone et al. [25] reported that CT score was significantly higher in critical COVID-19 patients ranging from 15 to 24 with 20.3±3 [mean value ± SD] than in the mild stage range 0-19 with 8.7±4 [mean value ± SD].
This study has limitations as none of the patients was in the pediatric age group; the study included only patients within the first week of onset of symptoms or positive RT-PCR, so late chest manifestations of the disease could not be investigated; also, there was no followup of cases that were asymptomatic with positive chest CT findings to find out if they developed symptoms or not or the patients that were symptomatic and with negative chest CT findings to see if they developed imaging manifestations or not, but this could be explained that the time of the study was at the peak period of the disease with a large number of cases and limited health care resources.

Conclusion
Chest CT findings are crucial for early diagnosis of COVID-19 disease especially for asymptomatic patients for early quarantine aiming to decrease the spread of the disease. Old age and male sex are considered risk factors for poor prognosis. CT score, serum ferritin, and LDH play an important role in predicting the short-term outcome of the patients allowing prompt effective clinical monitoring and intervention to decrease both morbidity and mortality.