Over the past months, HRCT had participated effectively in the diagnosis of COVID-19 and assessment of disease severity. The major imaging features of COVID-19 pneumonia have been discussed in detail in many publications . However, the post-recovery outcome of the disease and its long-term effects on lung parenchyma remain unanswered questions.
The previous experience with SARS and MERS showed that follow-up CT is recommended in individuals recovering from COVID-19 to evaluate which group of patients is more likely to develop pulmonary fibrosis .
In this study, we aimed to evaluate the radiological findings on follow-up HRCT of patients recovered from COVID-19 and identify the possibility of pulmonary fibrosis in discharged patients after treatment.
At the current study, patients were classified into two groups according to the presence of fibrosis on follow-up CT after discharge: “fibrotic group” (with evident fibrosis) and “non-fibrotic group” (without evident fibrosis). Out of 210 patients, 101 patients (48.1%) showed fibrosis on follow-up CT while 109 patients (51.9%) had no evident fibrosis.
Our results showed a statistically significant difference between the two groups regarding the patient’s age where patients in the fibrotic group were significantly older than those in the non-fibrotic group indicating that elderly patients are more liable to develop fibrosis following COVID-19.
In line with our study, Yu et al.  found that patients who developed fibrosis on follow-up after discharge were older than those without fibrosis, suggesting that fibrosis was more common in elderly patients, similar to SARS. They stated that patients with fibrosis had a longer length of stay in the hospital with a higher rate of ICU admission and a higher level of CRP than those without fibrosis; they also reported that patient with fibrosis had received more pulsed steroid therapy and antiviral therapy and for a longer period compared with patients without fibrosis; thus, these clinical parameters during acute disease may help in the prediction of the risk of developing pulmonary fibrosis after discharge.
Similarly, our results showed that patients in the fibrotic group showed a longer length of stay in the hospital and a longer duration of ICU admission and steroid therapy (p ˂0.001). In comparison with both groups regarding the laboratory results including the lowest lymphocytic level, CRP level, serum ferritin, high-sensitivity troponin, and d-dimer levels, there were statistically significant differences between both groups (p ˂0.001) with higher serum levels detected among patients in the fibrotic group suggesting that higher inflammatory markers are more associated with developing fibrosis.
Shi et al.  found that on COVID-19 patients, cardiac troponin is a prognostic marker with a strong association with mortality observed in the currently available reports of patients hospitalized with COVID-19, with some evidence suggesting cardiac troponin T/I even as an independent predictor of mortality.
At the current study, higher values for CT severity score and consolidation/crazy-paving score on the initial CT were found in the fibrotic group compared to the non-fibrotic group suggesting that patients with severe disease are more liable to fibrosis after discharge.
In the same context, Wei et al.  included 59 patients who were treated for COVID-19 pneumonia in a multi-center study and had a follow-up CT within 1 month after being discharged from four hospitals in China; their study showed that 39% of patients had residual fibrosis while 61% had no evidence of fibrosis on HRCT. They found that elderly patients had a higher chance of developing fibrosis, and patients who developed fibrosis had a higher CT score, higher ICU admission, and higher C-reactive protein. They also cited previous work by Antonio et al.  in studying severe cases of SARS who stated that early lung fibrosis rate reaches as high as 62% in SARS. COVID-19 showed a lower rate of fibrosis than SARS. In addition, SARS causes severe lung parenchymal damage than COVID-19.
Likewise, Das et al.  stated that 33% of patients with MERS show lung fibrosis on follow-up CT. These patients were older in age, had a longer duration of ICU admission, and had more severe lung involvement in the acute stage of the disease.
Additionally, we performed multivariate analysis for predictors of post-COVID-19 fibrosis, and we found that among these previous significant variables, patient’s age, initial CT severity score, consolidation/crazy-paving score, and ICU admission were independent risk factors associated with post-COVID-19 fibrosis (p<0.05, Table 4). Further analysis of the ROC curve for independent factors was done and showed the highest AUC for chest CT severity score reflecting a good predictive value for post-COVID-19 fibrosis with a sensitivity of 86.1% and a specificity of 78% at a cutoff point of 10.5.
Yu et al.  compared the imaging features between a group of patients with fibrosis and without fibrosis regarding the initial CT; they found that more patients in the fibrosis group had interstitial thickening, coarse reticulations, and subpleural/parenchymal bands on the initial CT. Also, more lung segments were involved on the initial CT in patients in the fibrosis group than in the non-fibrosis group. They suggested that these findings on the initial CT as interstitial thickening, reticulations, and parenchymal bands might be predictors of pulmonary fibrosis in recovered patients since they have similar pathogenesis.
This was matching with our results that showed more lung segment affection on the initial CT in the fibrotic group compared to the non-fibrotic group (p<0.001). Pure GGO was statistically higher in the non-fibrotic group, while pure consolidation or GGO with consolidation, crazy paving, air bronchogram, and fibrotic changes were significantly higher in the fibrotic groups (p<0.001).
At the current study, follow-up CT after discharge was performed at an average time of 41.5 days (range 20–65 days) after discharge, and it showed persistent parenchymal abnormalities with fibrotic changes in 48.1% of patients while 51.9% of patients had no residual parenchymal changes or fibrosis.
These results were in accordance with the findings reported by Liu et al.  who performed a 3-week follow-up study to determine the cumulative percentage of complete radiological resolution of pulmonary changes in discharged patients recovering from COVID-19. They stated that in 53% of patients, the pulmonary changes were completely absorbed at the 3rd week after discharge, reflecting that pulmonary damage induced by COVID-19 could be potentially repaired without permanent sequelae. However, more than 40% of patients at the 3rd week radiological follow-up showed residual parenchymal abnormalities, including GGO and fibrous parenchymal bands. In their study, younger age was associated with more complete radiological resolution.
Zaho et al.  in a multi-center cohort study found that residual pulmonary abnormalities and abnormal CT scores persisted for 3 months after discharge in 70.91% of COVID-19 survivors. Bilateral lung involvement was still found in 23.6% of patients and in about half of patients (54.55%); 1–3 lung segments were involved. Typical parenchymal features were almost resolved, but evidence of fibrosis was observed. The most common CT feature found on the latest follow-up was interstitial thickening (27.27%).
The current study has limitations to be acknowledged. First, small sample size relative to the disease burden. Therefore, studies in larger samples should be considered. Second, fibrosis was not confirmed by histopathology even though imaging manifestations were diagnostic. Third, follow-up time was relatively short, and it is unknown whether these pulmonary changes will resolve on further follow-up or permanently remain, so long-term follow-up studies are recommended for further research.