Dear Editor,

Long-term sequelae of COVID-19 pneumonia are still unknown. Han et al. [1] were the first to describe chest CT findings with pulmonary fibrosis-like changes (FLC) in about one third of patients at six-months follow-up; in this study FLC were defined by the presence of traction bronchiectasis, parenchymal bands and/or honeycombing. Similar results were observed by our group at six months in a cohort of patients hospitalized for severe Covid-19 pneumonia [2]. However, whether these lesions represent a marker of ongoing fibrosis and their clinical meaning remain uncertain.

The aim of this study is to evaluate the radiologic findings at two-years follow-up chest CT-scans in patients who presented FLC on chest-CT performed 6 months after severe COVID-19 pneumonia.

This study obtained ethical approval by the Ethics Committee of our Hospital. We retrospectively evaluated two-years CT scans of patients who had evidence of FLC at 6-months follow-up after severe COVID 19 pneumonia.

All CT images were evaluated by a radiologist with 20-years experience in thoracic radiology, who was blinded to patients’ clinical data. CT lesions were defined according to Fleishner Society classification system. We assessed the extent of pulmonary abnormalities by a semi-quantitative score (CT-score) and calculated the Warrick score, as previously described [2].

Recorded clinical data and pulmonary function tests data were also evaluated. Statistical analysis: categorical variables were reported as counts (percentages) whereas continuous variables were reported as mean (standard deviation, SD) or median (interquartile range, IQR), as appropriate.

Eighteen patients (15 men; 3 women; median age 70—IQR 64.75–77.75)-out of 24 in which FLC had been found at six months [2]—underwent a two years CT scan and follow-up clinical evaluation at our hospital.

The worst, median P/F during hospitalization for Covid-19 pneumonia was 171 (IQR 97.25–259); 10 patients underwent non-invasive and 1 invasive mechanical ventilation.

As compared to 6-months’ findings, 2-years CT scans showed a marked reduction in ground glass opacities (GGO) while reticular lesions remained the predominant pattern. Of notice, none of the patients showed honeycombing and/or traction bronchiectasis as new findings at 2-years follow-up. Persistence of FLC was confirmed in 18/18 patients, although a reduction in density and number of lesions was observed in most of the patients (Fig. 1). At two years, CT and Warrick scores improved respectively in 6 and 5 patients, leading to a trend toward reduction in mean scores values (Table 1).

Fig. 1
figure 1

Serial non-contrast axial chest CTs of three studied patients with prior severe COVID-19 pneumonia. Upper line: chest CT of a 61-year-old woman showed ground-glass opacities, reticular alterations and parenchymal bands (Warrick Score: 8; CT score: 5) in both lower lobes at the 6-month CT follow-up (Panel A). At the 2-years follow-up (Panel B) complete resolution of GGO with persistency of reticulations and bands was noticed (Warrick score: 5; CT score: 3). Middle line: panel C shows six-month follow-up chest CT of a 65-year-old woman demonstrating ground-glass opacities, reticular alterations, parenchymal bands and traction bronchiectasis in middle lobe, lingula and in both lower lobes (Warrick score: 20; CT score: 9). At the 2-years follow-up (Panel D) almost complete resolution of traction bronchiectasis, bands and GGO with persistency of reticulations was noted (Warrick score: 13; CT score: 6). Bottom line: panel E shows chest CT of a 57-year-old woman with reticular alterations and parenchymal bands in the lung lower lobes at the 6-month follow-up (Warrick score: 8; CT score: 8). At 2-years follow-up (Panel F) complete resolution of GGO and parenchymal bands with persistency of tiny residual reticulations are observed (Warrick score: 5; CT score: 5).

Table 1 The table describes CT findings with a comparison between 6-months and 2-years follow-up CT examinations

In more than half (72%) of the patients we observed persistency of symptoms; in particular, 39% reported fatigue and/or dyspnea. In 83% of patients a reduction in DLCO (< 80%) was detected; two patients showed a DLCO < 60%.

To the best of our knowledge, this is the first report to investigate the evolution of lung CT abnormalities after two years from severe COVID-19 pneumonia.

In most patients we observed a progressive reabsorption of GGO with persistency of reticulations and parenchymal bands, thus confirming findings reported in studies with shorter follow-up.

Furthermore, we noticed a persistency of FLC in all the patients with a slight reduction in their extent over time, similarly to what reported at one-year evaluation [3, 4].

A careful longitudinal analysis of CT findings suggests that these lung abnormalities may not be clues of a develo** fibrosis. In fact, thin reticulations and parenchymal bands were the most common findings, while none of the patients had a new diagnosis of traction bronchiectasis and/or honeycombing which are considered true markers of fibrotic evolution.

These observations are supported by a trend toward a reduction in quantitative scores together with fading of lesions in most of the patients.

In conclusion, we suppose that the CT alterations reported may represent residual radiological scars-as suggested by Wells et al. [5]—rather than markers of ongoing fibrosis. Future studies on larger populations and longer follow-up period are needed to verify this hypothesis, which could impact on follow-up management of patients with COVID-19 pneumonia.