Lung Parenchymal Abnormalities and Outcomes in Hospitalised Patients with COVID-19 Pneumonia
Pulmonary fibrosis is one of the commonly reported sequelae with varying frequencies among the survivors of COVID-19. While cases have been reported of parenchymal lung abnormalities (PLA) on high-resolution computed tomography (HRCT) scan in patients with COVID-19 pneumonia with varying periods of follow-up, pulmonary consequences are clinically significant for potential therapeutic intervention and is highlighted as a key research priority.
To better understand the evolution of the patterns of PLA on HRCT over 12 weeks, we undertook a prospective observational study in adult patients of COVID-19 whose infection was confirmed via an RT-PCR test, and were hospitalised from June 2020 to September 2020 at Medanta - Gurugram.
Consecutive, consenting patients with acute symptoms of COVID-19 and no pre-existing awareness of PLA – both confirmed with an RT-PCR test and HRCT, respectively – were included in the study. Patients with clinical and/or radiological features suggestive of an alternative diagnosis were excluded. The supportive care and treatment regimen for individual patients were in accordance with the standard of care at the time.
A non-contrast, supine, inspiratory HRCT of the lungs was done at admission (CT1); it was repeated 4-8 weeks after admission (CT2), and then again at 10-12 weeks (CT3) in a subgroup of patients with persisting symptoms. Interpretation of HRCT image pattern was done by a team of 3 experienced chest radiologists blinded to the clinical severity and each other’s interpretation.
The distribution and patterns of PLA were pre-specified and agreed upon by 2 senior
pulmonologists and chest radiologists. The PLA patterns and distribution were visualised at the same time as CT1, CT2, CT3 and the findings were interpreted without discussions among the radiologists. The final diagnosis was based on a multidisciplinary discussion with pulmonologists and the COVID-19 team in Medanta - Gurugram. Patients were followed up over telephone at 6 months post discharge to enquire about overall status of well-being using a structured questionnaire. Temporal changes in clinical severity, PLA and extent of lung involvement across CT1, CT2 and CT3 were presented and compared using the Chi-Square Test.
The study was approved by the Institutional Review Board and Ethics committee at
Medanta - Gurugram.
Of the 179 patients who met the inclusion criteria, 145 were male (81%) with a median ± interquartile range (IQR) age of 57.9 years ± 9.7 years. There was pre-existing hypertension in 78 (44%) and diabetes in 72 (40%). At presentation, cough in 111 patients (63%) and fever in 153 (85%) were the predominant symptoms, with 15 (8%) complaining of dyspnea and 74 (41%) had a hospital stay of more than 10 days. Only one patient had pre-existing interstitial lung disease (ILD) while 17 patients had pre-existing chronic obstructive pulmonary disease (COPD) or asthma.
At presentation, 88 (44.7%) were on low-flow oxygen (<4L/min), 60 (33.5%) were on high-flow oxygen (>8L/min) and 7 (3.9%) on invasive ventilation. At 4-8 weeks, none of the patients needed to be on high-flow oxygen or ventilator support.
At baseline, the predominant feature was ground glass opacities (GGOs) in 144 (80.4%) patients; consolidation in 23 (12.9%) and reticulation in 7 (4%). At 4-8 weeks, there was a reduction in the number of GGOs to 123 (68.7%), consolidation to 10 (5.6%) and an increase in reticulation to 23 (12.9%). Of the 44 patients who returned for follow-up at 10-12 weeks due to persisting symptoms, GGOs were seen in 26 (59%) and reticulations were seen in 9 (20.5%) patients.
The above image shows the extent of lung involvement across 3 serial CT scans. The extent of lung involvement showed a substantial reduction over time in the 3 sets of HRCTs (CT1 vs CT2 p < 0.005; CT2 vs CT3 p < 0.0001; CT1 vs CT3 p < 0.008).
Our observations of the GGOs as the most common PLA, followed by consolidation at baseline is similar to the report of a systematic review of 109 reports involving 2,908 adults with COVID-19 that showed GGOs in 68% followed by consolidation in 48% of cases.
At 4-8 weeks, our study showed patterns of fibrosis in HRCT images of 12% cases. At 10-12 weeks (CT3), fibrosis was seen in 9 out of the sub-group of 44 patients. We interpret the latter finding with caution as CT3 was obtained in the sub-group of patients prompted by persistent respiratory symptoms and clinical decision of the treating physician.
Nevertheless, it is apparent that at least 5% (9 out of 179) patients hospitalised with COVID-19 pneumonia had HRCT evidence of pulmonary fibrosis at 3 months. Two recent meta-analyses have reported higher prevalence of fibrotic sequelae of 29% (95% CI: 22%–37%) and 45% (range: 9.3-84%) with a median follow-up of 3 months.
Our factual observations at 4-8 weeks, interpreted prospectively and precisely based on pre-specified criteria are in keeping with the interim analyses of the UK Interstitial Lung Disease (UKILD) Consortium-Long COVID study that estimated residual lung abnormalities in up to 11% people following COVID-related hospitalisation. In essence, our study confirms that our findings are complementary to the estimate from the
UKILD study.
A total of 144 out of 176 survivors who responded over the telephone at the 6-month follow-up, reported good overall sense of well-being and return to their pre-COVID health status. Studies from other cohorts have reported exertional dyspnea in more than 1/3rd of hospitalised COVID-19 patients at 12 months. Another large prospective multi-centric study (post-hospitalisation COVID-19 study) found that only 29% were fully recovered at 6 months. While these differences could be due to varying ethnic backgrounds, clinical severity, management protocols etc., we believe our results from patients’ cohort of Caucasoid origin, born and raised in India, not in Europe, are useful to understand the sequelae and consequences of COVID-19 pneumonia globally.
Limitations of our prospective observational study include findings in hospitalised patients at a single tertiary care hospital, shorter duration of follow-up, lack of pulmonary function tests to correlate with the PLA, 6-month follow-up was limited to phone calls in 144 out of the 176 survivors who responded, and third HRCT was obtained in the subgroup of 44 patients with persisting respiratory symptoms at 10-12 weeks. Acknowledging the challenges and difficulties associated with obtaining lung function test during the peak of the COVID-19 pandemic, our observations of PLA are a prospective study of
a relatively large cohort of hospitalised patients with COVID-19 in the pre-vaccination era.
In summary, our study documents resolving or completely resolved PLA in 88% of 179 hospitalised patients with COVID-19, fine reticulations indicative of pulmonary fibrosis in 12% at 4-8 weeks, and fibrotic sequelae in at least 5% of patients at 10-12 weeks from baseline. At 6 months, 144 of the 176 survivors, who responded to follow-up phone calls, had returned to their normal pre-COVID health status. Contrary to most studies that have reported a significantly higher prevalence (30%-45%) of fibrotic sequelae post COVID-19 pneumonia, our findings of much lower prevalence are reassuring and send a positive message to the medical fraternity.
Results from our study and lessons from other studies related to COVID-19 pneumonia during this pandemic warrant consideration of prospective studies proactively to determine the clinical significance of long-term PLA, lung function impairment and the potential role of anti-fibrotics among those who manifest persistent pulmonary fibrosis after new variants of COVID-19 or other similar viral pneumonia, if and when that occurs
in future.
