COVID-19 in Egypt after a year: the first and second pandemic waves from the radiological point of view; multi-center comparative study on 2000 patients

Background One year has passed since the announcement of COVID-19 as a pandemic and two waves had already stricken Egypt. The authors witnessed several atypical radiological features through the second pandemic wave, either early at the active infective stage or delayed at the post-infectious convalescent period. They believed every radiologist should be familiar with these features. Therefore, they performed this comparative study on 2000 Egyptian patients using multi-slice computed tomography (MSCT) to highlight the radiological differences between the first and second pandemic waves and correlate them to the clinical status. Results This random multi-center comparative study was retrospectively conducted on 2000 COVID-19 Egyptian patients; 1000 patients were registered at the first pandemic wave from April 2020 till September 2020, while the other 1000 patients were registered at the second pandemic wave from October 2020 till March 2020. Follow up CT examinations were performed for 49 and 122 patients through the first and second pandemic waves respectively. MSCT examinations were carefully evaluated by four expert consulting radiologists who came to a consensus. Meanwhile, the correlation with the clinical outcome was performed by two consulting pulmonologists. During the second pandemic wave, the prevalence rate of the “crazy-paving” pattern had significantly increased by 1.3 times (P value = 0.002). Additionally, the prevalence rate of the “air-bubble” sign had significantly increased by 1.9 times (P value = 0.02). Similarly, the presence of enlarged mediastinal lymph nodes (> 1 cm in short-axis diameter) had significantly increased by 1.7 times (P value = 0.036). Furthermore, the prevalence rate of pericardial effusion had significantly increased by 2.5 times (P value = 0.003). The above-mentioned signs were correlated to increased clinical severity and higher rates of hospitalization. Unexpectedly, other atypical radiological signs were only encountered through the second pandemic wave, including bronchiectatic changes (2.5%), “head-cheese” pattern (0.8%), cavitation (0.5%), and “bulls-eye” sign (0.2%). The prevalence rate of post-COVID fibrosis had doubled through the second wave but not in a significant way (P value = 0.234). Secondary fungal infection was only encountered throughout the second pandemic wave in four patients. COVID-19 reinfection was encountered in a single patient only during the second pandemic wave. Conclusion After 1 year from the announcement of COVID-19 as a pandemic, the radiological presentation of COVID-19 patients showed some significant differences between its first and second waves.


Background
Around 1 year has passed since Egypt had reported the first case of COVID-19 infection in Africa on February 14, 2020 [1] and since WHO declared it as a pandemic in March 2020 [2]. On April 21, 2020, 3333 cases of COVID-19 were reported by the Egyptian Ministry of Health [3]. On May 31, 2020, 24,985 cases were furtherly reported [2]. On June 13, 2020, 41,303 cases and 1422 deaths were reported nearly at the peak of the first pandemic wave [4]. On September 2, 2020, 99,280 cases and over 5400 deaths were reported [5] and in mid-October, 2020, the second pandemic wave was announced [6] and reached its peak in December 2020 and January 2021.
Over the past year, the clinical, laboratory, and radiological characteristics of COVID-19 have been thoroughly evaluated by multiple observational studies, systematic reviews, and meta-analysis [7][8][9]. CT was known as a rapid and sensitive tool for diagnosis of COVID-19 infection but with low specificity [10,11]. Several typical, indeterminate, and atypical CT-signs of COVID-19 were described [12,13].
The authors witnessed several atypical radiological features through the second pandemic wave, either early at the active infective stage or delayed at the postinfectious convalescent period. They believed every radiologist should be familiar with these features. Therefore, they performed this comparative study on 2000 Egyptian patients using multi-slice computed tomography (MSCT) to highlight the radiological differences between the first and second pandemic waves and to report their impact on the clinical status.

Study protocol
This random multi-center comparative study was retrospectively conducted on 2000 COVID-19 Egyptian patients; 1000 patients were registered at the first pandemic wave from April 2020 till September 2020, while the other 1000 patients were registered at the second pandemic wave from October 2020 till March 2020. MSCT examinations were carefully evaluated by four expert consulting radiologists who came to a consensus. They have long time experience in chest imaging (15-25 years) and they were informed with all data. Meanwhile, the correlation with the clinical outcome regarding the rate of hospitalization and impact on respiratory functions later was performed by two consulting pulmonologists, who have long time experience in the management of chest diseases (16 and 20 years).
The Ethics Committee of our University hospital had approved the study protocol. The patient consent was waived by the Research Ethics Board with the assurance of the respect of the confidentiality of patients and medical records.
Inclusion criteria were as follows: (1) positive PCR and MSCT results for COVID-19, and (2) available clinical records. Exclusion criteria were as follows: (1) poor CT images with motion artefacts, and (2) absent clinical records.
The CT Scanning parameters were as follows: slice thickness = 1-2 mm, FOV = 350 mm × 350 mm, tube rotation = 0.6-0.9 second and detector collimation = 1 mm. The irradiation dose parameters were as follows: 120-130 kVp and 100-200 mA (according to the machine type as well as patient age and weight). Intravenous contrast administration was not utilized.

CT evaluation
Multi-planar reconstruction (MPR) was performed for CT images. Additionally, the maximum intensity projection (MIP) reconstruction was utilized for nodular and reticular assessment, while the minimum intensity projection (Min-IP) reconstruction was utilized for airway assessment and mosaic attenuation characterization.
The following CT features were compared between the patients throughout the first and second pandemic waves: A. Preexisting lung comorbidity; including COPD, ILDs, lung neoplasms (whether lung cancer or metastatic disease), and miscellaneous conditions (such as TB and sarcoidosis). B. MSCT findings during the first two weeks of infection, including: Ground-glass opacities with or without consolidative changes [14]. "Atoll sign": Central ground-glass attenuation with peripheral consolidation [14]. "Bulls-eye sign" or "Target sign" or "Double-halo": Central nodule or consolidation and peripheral organization rim with ground-glass in between [15]. "Crazy-paving pattern": Ground-glass attenuation with septal thickening [14]. "Air-bubble sign": Small foci of air trapping or bronchiolectasis sequel to septal thickening and fibrosis [16,17]. Cavitary changes. Airway involvement: bronchial wall thickening, bronchiectasis and tree in bud nodules [14]. "Head-cheese pattern": ground-glass attenuation with alternating areas of air trapping [14]. Peri-lobular fibrosis. Significant mediastinal or hilar lymph node enlargement (Short axis diameter > 1 cm) [14]. Pleural effusion. Pneumo-mediastinum: spontaneous if not preceded by oxygen therapy. Pericardial effusion.
D. Post-infectious sequel or complications (after 1 month of infection): including residual ground-glass opacities, residual air trapping, or mixed headcheese pattern [14] as well as post-COVID fibrosis. E. Signs of secondary bacterial or fungal infection. F. COVID-19 reinfection.

Statistical methods
The prevalence rate was estimated. Data were compared using online calculators (https://www.socscistatistics. com) with chi-square test and p value analysis (p value < 0.05 was considered as statistically significant).

Results
Demographic criteria (Table 1) Throughout the first and second pandemic waves, the rate of infection was highest in the fourth decade of life. Despite that, the rate of infection through the fourth and fifth decades of life witnessed more than 50% reduction through the second pandemic wave (significant P value < 0.00001, for both). On the other hand, the rate of infection through the second and third decades of life during the second wave witnessed around 10 and 1.5 times elevation respectively (P value ≤ 0.00001 and 0.001). Additionally, a single 7-year-old patient was also registered ( Fig. 1).
Meanwhile, minimal and insignificant changes were encountered regarding the rate of infection through the sixth and seventh decades of life. Similarly, minimal and insignificant changes were documented regarding sex predilection. Males were more commonly infected  during the first and second pandemic waves (62.5% and 61%, respectively). Higher CT severity scores (> 50%) were depicted mainly in the 6 th and 7 th decades of life throughout both pandemic waves (93% and 83%, respectively). This coincided with corresponding higher clinical severity and increased rates of hospitalization. During the second pandemic wave, around 10% rise in the clinical and CT severity-scores as well as the hospitalization rates were reported through the 4 th and 5 th decades of life.
Preexisting lung comorbidity and clinical severity (Table 1) The prevalence rate of preexisting lung comorbidity increased four times throughout the second pandemic wave (significant P value = 0.000073). The prevalence rate of COPD and interstitial lung diseases (ILDs) had significantly elevated (P value = 0.005 and 0.025, respectively). Despite that the prevalence rate of sarcoidosis had doubled, it was statistically insignificant (P value = 0.164) (Fig. 2). A single patient with old TB and fibrothorax was encountered in the second pandemic wave (Fig. 3).
Hospitalization with oxygen support using high-flow nasal cannula up to mechanical ventilation was prescribed for 224/1000 patients (22.4%) during the first pandemic wave and 289/1000 patients (28.9%) during the second wave.
MSCT findings of the disease (Table 1) The ground-glass opacities with or without consolidative changes remain the gold standard CT-findings in all patients throughout the first and second pandemic waves.
During the second pandemic wave, the prevalence rate of the "crazy-paving" pattern and peri-lobular fibrosis with architectural distortion had significantly increased by 1.3 and 1.4 times respectively (P value = 0.002 and 0.02 respectively). Additionally, the prevalence rate of "air-bubble" sign had significantly increased by 1.9 times (P value = 0.02). Similarly, the presence of enlarged mediastinal lymph nodes (> 1 cm in short-axis diameter) had significantly increased by 1.7 times (P value = 0.036). Furthermore, the prevalence rate of pericardial effusion had significantly increased by 2.5 times (P value = 0.003). The above-mentioned signs were correlated to increased clinical severity. They were encountered in 98% and 93% of hospitalized patients throughout the first and second pandemic waves respectively.
Spontaneous pneumo-mediastinum, in patients who did not receive oxygen therapy, was encountered only in few patients through the first and second pandemic waves with minimal differences in between (0.2% and 0.7% respectively) (Fig. 4). All patients experienced rapid progressive dyspnea which necessitated hospitalization.
Tree in bud nodules that represent bronchiolitis were similarly encountered once throughout the first and second pandemic waves (Fig. 9). (Table 1) Four patients (0.4%) with secondary fungal infection were only documented during the second pandemic wave (significant P value = 0.045). Two patients of them had typical mycetoma balls (non-invasive pattern), while one patient had positive CT signs of semi-invasive aspergillosis including mucous plugging, air-fluid leveling, and air-trapping. One patient had mixed non-invasive and semi-invasive CT findings (Fig. 10). Sputum culture confirmed the diagnosis and anti-fungal therapy was prescribed.

Secondary infection
Post-infectious sequel or complications (after 1 month of infection) ( Table 1) Follow up CT examinations were only performed for 49 and 122 patients through the first and second pandemic waves respectively.
The prevalence rate of post-COVID fibrosis had doubled through the second pandemic wave but not in a significant way (P value = 0.234) (Fig. 11).
Persistent dyspnea was documented in 71% and 79% of patients with persistent ground-glass opacities and

COVID-19 re-infection
A single patient with COVID-19 re-infection was encountered during the second pandemic wave. He had his first infection at the beginning of October 2020 and reinfection around 6 months later at the end of March 2020 (Fig. 12).

Discussion
According to the Egyptian Ministry of health, the second pandemic wave of COVID-19 had stricken Egypt during the period from November 2020 till January 2021 (https://www.care.gov.eg). This period coincided with the winter season and followed the gradual re-opening of schools and universities. This fact could explain the striking elevation in the infection rate between the young Egyptian people in the second decade of life as well as the surge of infection between the children. This is similar to the findings in the study by Park JE et al. [18], in which low temperature and low humidity of the winter season were responsible for the more severe course of viral infections. At first, the authors thought that the virus predilects non-pathological lung parenchyma because of the very low incidence of preexisting lung comorbidity. But a significant rise of this incidence throughout the second pandemic wave was striking. The pulmonologists in this study tried to explain this notice. They hypothesized the following scenario. Meanwhile, those patients were considered as extremely vulnerable patients at high risk for COVID-19 complications, hence their physicians gave them strict warnings about the importance of COVID-19 precautions and this may decrease its incidence. Additionally, they minimized or delayed their chest CT needs to very limited situations (for fear of inadequate infection control measures). Even if their patients had suspicion about COVID-19, they prefer to depend on laboratory diagnosis with PCR instead of chest CT because radiological findings between the original disease and COVID-19 may overlap. They almost limit the role of CT to those patients with suspected false negative  PCR results or prior to hospitalization if needed. This can explain the very low incidence the first pandemic wave, but off course by time, the need for chest CT examinations increased and the confidence about infection control measures raised. Emphysema and ILDs were encountered during the second pandemic wave. Hussain MR et al. [19] was a case report study that similarly discussed the exacerbation of ILDs in a COVID-19 patient. Additionally, the incidental diagnosis of sarcoid disease was significantly reported in this study, which followed the wide use of chest CT examinations for either suspected or diseased COVID-19 patients. A single case of old TB was only reported in the second pandemic wave. Miller A et al. [20] hypothesized a theory about the immunity of people who received compulsory BCG vaccine against COVID-19 infection, but this study had not peer-reviewed and this theory had not been proved till writing this manuscript.
Coinciding with the universal criteria of COVID-19 radiological diagnosis [12,13], the typical CT signs of COVID-19 infection were encountered in this study, including ground-glass opacities (GGOs) with or without consolidative changes. This is consistent with the findings of other previous CT studies including Omar S et al. [21], Ali TF et al. [22], Emara DM et al. [23], Sabri YY et al, [24], and Mohamed IA et al. [25].
In this study, the clinical severity through the second pandemic wave had increased by around 6.5%. This coincided with the higher rates of pulmonary and extrapulmonary comorbidities or complications in addition to the increased prevalence of five CT findings of severity, including the "crazy-paving" pattern, "air-bubble" sign, significant mediastinal nodal enlargement, pericardial effusion, and spontaneous pneumo-mediastinum. Previous studies described the "crazy-paving" pattern in 100% of their severe patients, including Ali TF et al. [22], Ghweil AA et al. [26], Metwally M.I et al. [27], and Osman AM et al. [28]. Additionally, Li Y et al. [29] correlated the presence of pericardial effusion to the high clinical severity. Wang J et al. [30] also stated that 100% of patients with spontaneous pneumo-mediastinum suffered from severe respiratory distress and needed hospitalization. According to Lei P et al. [31], the spontaneous pneumo-mediastinum was encountered at the time of the resolution of the ground-glass opacities or consolidations and their replacement by peribronchiolar fibrosis or abscess formation with secondary interstitial emphysema.
In this study, some atypical CT findings were only encountered during the second pandemic wave. They included the bronchiectatic changes, the "head-cheese" pattern, the "bulls-eye" sign, and the cavitation. Therefore, the authors believe that radiologists everywhere should not exclude COVID-19 infection entirely if they found these CT findings.
The traction bronchiectatic changes, in this study, coincided with the early peri-lobular fibrosis and parenchymal distortion with high clinical severity. This was similar to the findings in the studies by Sabri YY et al. [2] and Zhao W et al. [32].
In this study, the second pandemic wave witnessed a doubling rise regarding the incidence of post-COVID fibrosis. Despite that, this rise was statistically insignificant. The post-COVID fibrosis throughout the first and second pandemic waves was correlated to the persistent dyspnea after the announcement of patient recovery. However and unexpectedly, 21-29% of patients with radiologically persistent lung fibrosis were asymptomatic. Spagnolo P et al. [38] and Tale S et al. [39] alarmed also to the bad prognosis of post-COVID fibrosis. Udwadia ZF et al. [40] dramatically described it as the tsunami that follows the earthquake. The super-added fungal infection at the second pandemic wave included either non-invasive or semi-invasive patterns. It is mostly explained by the mild decrease in lung immunity. It could be an additional explanation for the increased clinical severity.
This study adds to the literature the radiological differences between the first and second pandemic waves in one of the largest countries in the Middle East, Egypt, highlighting important unexpected findings.
This study was limited by the short time interval after subside of the second pandemic wave in Egypt; hence, further future researches about its long term effect are recommended.

Conclusion
After one year from the announcement of COVID-19 as a pandemic, the radiological presentation of COVID-19 patients showed some significant differences between its first and second waves.