COVID-19 Associated Interstitial Lung Disease-An Emerging Entity

The authors have requested that this preprint be removed from Research Square.


Background
The Coronavirus disease  caused by the novel coronavirus severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) has posed a global economic, psychosocial, political and medical challenge.
As patients recover from COVID-19 disease, we are approaching an era where physicians would encounter COVID associated pulmonary sequelae. These could be infectious, like COVID associated pulmonary aspergillosis (CAPA) [1]; or non-infectious which could include COVID associated interstitial lung disease (ILD), an increasingly recognized entity. The previous epidemics of coronavirus due to severe acute respiratory distress syndrome coronavirus (SARS-CoV) and middle east respiratory syndrome coronavirus (MERS-CoV) also led to the development pulmonary brosis [2,3].
ILD covers a wide spectrum of pulmonary parenchymal disorders of both known and unknown etiology.
Different radiological and histopathologic patterns of ILD have been described, some of which include usual interstitial pneumonitis (UIP), nonspeci c interstitial pneumonitis (NSIP), and organizing pneumonia (OP) [4]. Similar patterns can also occur as a result of pulmonary infections, like pneumocystis pneumonia or cytomegalovirus pneumonitis [5,6].
The diagnosis of ILD during COVID-19 pandemic remains challenging because invasive testing like bronchoscopies, open lung biopsies, or autopsies are rarely performed in COVID-19 patients due to risk of disease transmission. Although there are multiple reports on the importance of CT in diagnosing COVID-19 infection, there is little or no data on the clinical presentation and management of patients with ILD associated with COVID-19.
Some patients after recovery from COVID-19 infection might present with new, persistent or worsening respiratory symptoms due to COVID-19 associated Interstitial Lung disease. Our study aims to describe this underdiagnosed entity and underscore the importance of appropriate follow-up and prompt diagnosis of these patients, to facilitate early management and prevent fatal outcomes.

Study Design and Setting:
This single-center observational study was performed between April 1, 2020 and September 15, 2020 at the Aga Khan University Hospital, the largest tertiary care center located in Karachi, Pakistan. We retrospectively collected the demographic, clinical, laboratory and radiological data of patients presenting with COVID-19 associated ILD from medical records. Disease severity of patients with COVID-19 infection was classi ed according to the WHO classi cation [7]. High resolution computed tomography (HRCT) of the chest ndings, treatment and outcomes were recorded.

Study Subjects:
Patients who were seen in the outpatient Pulmonology clinic or inpatient consultation service at the Aga Khan University Hospital, owing to persistent respiratory symptoms after recovery from COVID-19 infection, were included in the study. Follow-up imaging with either HRCT or chest X-ray performed 6 to 8 weeks after treatment was interpreted and compared with previous imaging for disease progression, improvement or resolution.
Proposed Diagnostic Criteria for COVID-19 associated ILD (COVILD): The COVILD diagnosis in our study was based on 'new, persistence and/or worsening of respiratory symptoms and identi cation of ILD pattern on HRCT imaging of the chest after the initial recovery phase of acute COVID-19 infection de ned as 6 to 8 weeks after the onset of infection with no previous history of ILD. They were diagnosed to have ILD by a specialist in the ILD clinic or on inpatient consultation'.

Inclusion and Exclusion Criteria:
Adult patients (age 18 and above) who were con rmed for SARS-CoV-2 by nasopharyngeal and/or oropharyngeal swabs for real-time RT-PCR at initial presentation and underwent HRCT on follow-up visit were included in our study. Patients with pre-existing ILD and those with incomplete medical records were excluded.
Operational De nition of Outcomes: We have de ned outcomes as complete recovery, improvement and progression of the disease process. Patients were labelled as completely recovered if they returned to their baseline functional status and chest imaging showed clearance of lung in ltrates after the diagnosis of COVID-19 associated ILD. Improvement was de ned as subjective improvement in functional status but not to the baseline and at least 50% clearance of radiological in ltrates. The patients whose symptoms persisted with interval worsening of functional status and no signi cant improvement or had worsening of lung in ltrates had progression of the disease process.
HRCT Chest Analysis: The key HRCT chest ndings of COVID-associated ILD were de ned using standard taxonomy described in the ILD literature with interstitial patterns including but not limited to diffuse ground glass opacities (GGOs) with or without traction bronchiectasis (NSIP), basal and peripheral opacities with honey combing (UIP) and peripheral and peribronchovascular consolidation with or without ground-glass opacities (OP) [8]. The main HRCT ndings were described as GGOs, consolidation, honeycombing/ brosis and interlobular septal thickening/reticulation. Other HRCT ndings included crazy paving, reverse halo sign, traction bronchiectasis and emphysematous cysts. The distribution of pulmonary involvement was reported as either peripheral or diffuse.

Statistical Analysis:
Statistical analyses were performed using Stata version 12.1. Quantitative data were presented as mean ± standard deviation (SD), while frequencies and percentages were used to represent qualitative (categorical) data. The P value < 0.05 was considered statistically signi cant.

Discussion
Our study found four distinct patterns of ILD associated with COVID-19 disease. Development of interstitial lung disease amongst survivors of COVID-19 disease has been reported during the ongoing pandemic, however data is limited to case reports [9][10][11]. Literature from the previous outbreaks of viral infections such as SARS and MERS, in 2002 and 2012 respectively, reported that clinico-radiological changes persisted in approximately one-third of patients even after 12 weeks of discharge [2,3,12].
The majority of our patients who developed COVID-associated ILD were males, which has also been previously reported in literature with certain ILDs like idiopathic pulmonary brosis [13,14]. COVID associated brosis is one of the lung insults already described with previous Coronavirus infections [2,3], and there are emerging studies now reporting COVID-19 associated early pulmonary brosis. In the initial studies, Pan et al. and Zhou et al. reported brotic changes in the imaging features of patients with COVID-19 pneumonia [12,15]. Since then, brotic lung parenchymal remodelling [16,17], brosing diffuse alveolar damage (DAD) [18] and honey combing [19] have also been con rmed after invasive testing such as cryobiopsies and autopsies in smaller cohorts. Li Yan et al. described DAD on autopsy of 30 patients with COVID-19, showing 43% developing brosing patterns while 25% showing organizing pattern [6].
Post-infectious secondary OP is a known entity, well described with certain viruses like Cytomegalovirus and In uenza A (H1N1) [6,20]. Pathology in patients who recovered from SARS-CoV has shown brogranulation tissue proliferation and organizing pneumonia like patterns [21] while MERS associated organizing pneumonia has also been documented [22,23]. Studies of COVID-19 CT imaging, along with postmortem lung biopsies and autopsies during the ongoing pandemic, suggest the development of a secondary OP, which at present remains an underrecognized complication [24].
A distinct feature of organizing pneumonia and NSIP is the remarkable resolution with corticosteroid treatment. Although the use of corticosteroids has been recommended in the treatment of COVID-19 disease [25], there is limited or no data on response of prolonged or higher dose corticosteroids in COVID-19 associated interstitial lung disease.
In our cohort, most patients developed a predominant OP or NSIP pattern with moderate to severe disease. Most patients with an OP and NSIP pattern improved signi cantly with steroids, showing both clinical and radiological improvement. UIP pattern however largerly remained static or progressed.
Disease severity did not seem to have a signi cant impact on development of any particular ILD pattern. This proves that these interstitial changes are not only a result of post-ARDS brosis or ventilator induced lung damage, but also a consequence of the direct virus induced injury and aberrant local immune response leading to ILD. Combet et al. recently described a case of a spontaneously breathing patient who developed rapid honeycombing following COVID-19 disease which responded to high-dose steroids and nintedanib [26]. Tale et al. also reported a similar case of a patient with persistent hypoxemia after recovery from moderate COVID-19 disease with 3 week follow-up HRCT showing artitechtural distortion, interlobar septal thickening and traction bronchiectasis [27]. These case reports reiterate our stance that predisposed patients who are moderately ill, and do not require mechanical ventilation can also develop early brotic changes. Post-viral pulmonary brosis associated with previous corona viruses has been seen in patients with critical disease leading to ARDS with longer duration of illness requiring ICU stay and invasive mechanical ventilation [2]. However, SARS-CoV-2 has shown to induce brosis in those patients suffering from moderate disease, which did not require invasive mechanical ventilation or ICU stay in our study.
The strength of our study is that we propose a new working de nition of COVID-19 associated ILD (COVILD). To the best of our knowledge this is a rst case series emphasizing and presenting a data of 30 patients with different patterns of COVID-19 associated ILD along with their follow up from a low to middle income country.
Our study has several limitations which includes the absence of histopathologic con rmation of ILD. Transbronchial and open lung biopsies were not performed due to the cost, invasive nature of procedure and risk of transmission of COVID-19; therefore our patients were diagnosed solely on clinical and radiological grounds. We did not have follow-up HRCT imaging on all the patients, due to nancial constraints. Limited number of pulmonary function tests and diffusion capacity of lung for carbon monoxide (DLCO) were performed to assess the physiologic function of the lung. Conclusion COVID-19 associated interstitial lung disease is a new entity. Close follow-up is essential in these patients, as they may require prolonged treatment with corticosteroids. The long-term effect of COVIDassociated ILD is yet to be determined, and a longitudinal follow-up will help us further explore the nature of disease. Availability of data and materials: The datasets used and analysed during the current study are available from the corresponding author on reasonable request ABSZ was responsible for the concept and design of the study. SA analysed the data and oversaw all analyses. ABSZ, AS, SMZ, ASA, SA and MI were responsible for data interpretation and drafting the manuscript. ABSZ, AS, SMZ, ASA, SA and MI revised the manuscript critically for intellectual content. All authors have approved the nal version to be published and are jointly accountable for all aspects of the work.  CT scan chest of a patient showing a) peripheral patchy areas ground glass opacities (GGOs) with consolidation consistent with an OP (organizing pneumonia) pattern (yellow arrow pointing towards peripheral patchy GGOs); and b) complete resolution of peripheral patchy GGOs after initiation of corticosteroids.