Does COVID-19 pneumonia signify secondary organizing pneumonia?: A narrative review comparing the similarities between these two distinct entities

Multiple observational studies have described the similarities between COVID-19 pneumonia and organizing pneumonia (OP). These two entities clinically manifest with mild and subacute respiratory symptoms, often with a delayed diagnosis due to the atypical ARDS and silent hypoxemia presentation. Radiological features are often indistinguishable between the two. With the increase in antemortem lung biopsies and autopsies being performed, more histopathological findings of OP and its variant, acute fibrinous and organizing pneumonia (AFOP), are being diagnosed. These entities are known complications of viral infections as a delayed immunological process, explaining the favorable response to corticosteroids. Clinicians should be vigilant to diagnose this under-recognized entity of secondary OP in people with COVID-19 when clinical deterioration occurs, especially with compatible radiologic findings and recent cessation of corticosteroids. Despite the proven benefits of corticosteroids in treating COVID-19, treatment approaches can be more effective as OP often requires higher doses and a more prolonged therapy duration for remission and preventing relapses. The purpose of our narrative review is to compare the similarities between COVID-19 pneumonia and OP, emphasizing the clinical, radiological, and histopathological features based on the evidence available in the literature.


Introduction
Coronavirus disease 2019 (COVID-19) results from severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infection leading to the ongoing worldwide pandemic. As the wave of COVID-19 disease continues to spread across the globe, many medical publications have followed as part of scientific efforts to characterize, treat, and prevent this disease. COVID-19 can present with a wide variety of respiratory complications that range from self-limiting upper respiratory tract infection to acute respiratory failure from diffuse bilateral lung infiltrates, resulting in acute respiratory distress syndrome (ARDS). 1À3 By their very nature, histopathology manuscripts have lagged behind those describing the clinical and imaging characteristics, but we are now seeing an abundant amount of publications giving us insight into the histopathological findings in COVID-19 pneumonia. Organizing pneumonia (OP) is defined pathologically as the presence of intra-alveolar buds of granulation tissue consisting of fibroblasts and myofibroblasts mixed with a loose connective matrix extending from the alveoli into the lumen of distal bronchioles ( Fig. 1). OP is a rare distinctive form of interstitial lung disease (ILD) that is often contributed to idiopathic causes, termed cryptogenic organizing pneumonia (COP). The "cryptogenic" nature of OP could be related to the limited knowledge about all entities leading to secondary organizing pneumonia (secondary OP). Various known etiologies such as infections (commonly viral-induced), drugs, rheumatological diseases, aspiration, radiation, drugs, and toxins have been well recognized to cause secondary OP. 4À7 Historically, a strong correlation exists between secondary OP and various infections involving adenovirus, cytomegalovirus, herpes virus, human immunodeficiency virus, parainfluenza virus, and influenza virus, perhaps via stimulation of the immune system by viral antigens. 5À8 This entity has even been reported among highly contagious viruses responsible for previous global pandemics such as severe acute respiratory syndrome (SARS), Middle East respiratory syndrome (MERS), and influenza A. 9À12 The association between COVID-19 pneumonia and secondary OP has been called into question considering the increased recognition of the similarities between these two entities despite being limited to several case reports/series and post-mortem findings. Based on the current evidence in the medical literature, the purpose of our narrative view is to compare and describe the similarities between COVID-19 pneumonia and OP, with a strong emphasis on the clinical characteristics, radiological features, histopathological findings, and available therapies.

Methods
A literature search was performed through MEDLINE, Pubmed, and SCOPUS databases using keywords of "coronavirus disease 2019 (COVID- 19)," "severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2)," "organizing pneumonia (OP)," "cryptogenic organizing pneumonia (COP)," and "secondary organizing pneumonia (secondary OP)." All specified keywords were combined using the "OR" operator and "AND" operator for searching the literature. Articles published up to February 1 st , 2021, were included from peerreviewed journals. We simultaneously searched the reference lists of all recovered articles for potentially eligible studies. All research study types, such as case series, case reports, and observational studies, were included. No document restriction was applied, and no methodology filters were used. Two authors (W.H. and B.S.) independently reviewed the titles and abstracts of all search results for eligibility. If an article was considered potentially eligible, both authors independently examined the full article for inclusion. A consensusbased discussion resolved disagreements between the authors. If no consensus was reached, a third reviewer (A.C.) made the final decision.
We identified 5 case reports and 3 case series describing secondary OP diagnosis among hospitalized COVID-19 patients with histopathological confirmation (Table 1). These studies were analyzed for month, year, and country of publication; age and gender; hospital course (radiological and histopathological findings); and outcomes. Data extraction was based on conclusions found in reviews, methods and results gathered from observational studies, and results reported in case reports and case series. Observational studies identified were autopsies describing histopathological findings of OP in the lungs of people with COVID-19 or descriptive radiological studies. Case reports and case series described the antemortem diagnosis of OP based on either histopathological findings and/or radiological features in people hospitalized with COVID-19. As the literature data are limited and heterogeneous, it was not feasible to perform a qualitative assessment, statistical analysis, or provide high-quality, evidence-based recommendations. Still, consensus, recommendations and opinions are provided where appropriate.

Clinical characteristics of COVID-19 and OP
The symptoms of COVID-19 infection are primarily respiratory with cough, dyspnea, myalgia, and fever, although often indistinguishable from other viral illnesses. 1À3,13 Many presents with mild and non-specific symptoms. Only a few seek medical attention, especially during the early phase of the disease. In a large multi-center observational study involving 1,099 people with COVID-19, the overall incidence of ARDS is 3.4% but increases to 17-29% among those requiring hospital admission. 1À3 The median time from illness onset to development of dyspnea is 8 (5)(6)(7)(8)(9)(10)(11)(12)(13) days, whereas the time to develop ARDS and ICU admission is 12 (8)(9)(10)(11)(12)(13)(14)(15) days. 2,13 In OP, the mean age of diagnosis is 59.0 +/-13.6 years, with non-specific symptoms often described as a flu-like illness. 5,8 In a similar fashion to COVID-19, OP symptoms are usually mild at the initial onset with subacute presentation over a few weeks. 6,7 Due to the mild and nonspecific nature of presentation, it may take approximately 2-3 months before the diagnosis of OP can be made. 5 Fever and shorter duration of symptoms are often helpful clues to indicate an underlying infectious etiology for secondary OP. 4,6,14 The majority of people diagnosed with COVID-19 pneumonia and OP are non-smokers or ex-smokers. 1À3,5,13,15 These clinical characteristics have been described in the many case reports and series included in Table 1.
Pathophysiology COVID-19-related ARDS does not present and progress in a typical ARDS pattern with clinical symptoms often inconsistent with the severity of laboratory and radiologic findings. 16 Furthermore, lung compliance may be relatively normal to high, and the onset of COVID-19-related ARDS of 8-12 days is inconsistent with ARDS Berlin criteria, defining ARDS onset within a week from known insult. 2,13,16 COVID-19-related ARDS have a time-related disease spectrum with two primary phenotypes of type L and type H that explains the delayed subacute symptomatic presentation (Fig. 2). 17 Type L phenotype is characterized by low elastance, low ventilation to perfusion (V/Q) ratio, low lung weight and lung recruitability. In the early course of the disease (phenotype L), elastance remains low, but compliance is high due to the incomplete alveolar filling, which partially reduces ventilation in these regions with less impact on elastic recoil. 18 Severe hypoxemia occurs due to the loss of lung perfusion regulation and hypoxic vasoconstriction. At this time, people with COVID-19 typically show no signs of respiratory distress, giving the term "silent hypoxemia or happy hypoxemia" from the lack of compensatory respiratory drive as a minimal increase in minute ventilation will cause a rapid reduction in PaCO2 in the setting of low elastance. 19 The respiratory centers are more sensitive to changes in PaCO2, where minimal changes in PaCO2 will outweigh PaO2 in influencing the respiratory drive. 20 Similar phenomena are observed during the early stage of OP where marked hypoxemia occurs from defective pulmonary vasoconstriction (Type L phenotype) in the areas of poorly ventilated and well-perfused lung regions due to intra-alveolar buds occupying the entire lumen of alveoli. 6,7 Over time, the progression and accumulation of edema become more prominent secondary to high permeability from ongoing lung inflammation, causing an increase in lung weight and dependent atelectasis. Clinical deterioration will ensue with the evolution from type L to type H phenotype, especially in critically ill patients. Type H phenotype is characterized by a high elastance, high V/Q ratio, high lung weight and lung recruitability. This will result in further rightto-left shunting, dead space ventilation, and lower tidal volumes generated. As a result, respiratory distress will occur from normalization and retention of PaCO2, and activation of both hypoxic and hypercapnic ventilatory drive in the respiratory center. 17,19,20 Radiological features of COVID-19 and OP Multiple COVID-19 radiological manifestations have been described in the current literature to determine the typical radiological features of COVID-19 pneumonia. Furthermore, several observational studies have assessed the different stages of COVID-19 pneumonia from the time of illness onset. 21À26 In the early phase of COVID-19 disease (0-4 days), alveolar opacities [predominantly  ground-glass opacities (GGOs) and reticulations] favoring the peripheral, bronchovascular, and basilar lung regions occur that will increase in lung distribution on day 5-9 (intermediate stage) by diffusely involve multiple lung regions and interstitium giving rise to interlobular and crazy-paving patterns. In the later stage (day ten and more), the evolution of alveolar opacities into consolidation occurs where GGO interspersed with consolidations and reticulations plus reverse halo signs (central GGO surrounded by complete/incomplete ring of consolidation) are common findings on CT chest. 21À26 The typical radiological pattern of OP is characterized by bilateral, patchy, perilobular alveolar opacities (with a density ranging from GGO to consolidation) that are frequently subpleural (Fig. 3) or bronchovascular in distribution with predominance for the middle and lower lung zones. 4,27 In a similar fashion to COVID-19, the opacities in OP will evolve into a more consolidative appearance with reverse halo signs. 28À31 Peripheral and lower-lobe predominant GGOs concerning for OP have been described on chest CT during influenza A pandemic. 12,32 With the wide variety of radiological manifestations described, it is not surprising why COVID-19 pneumonia can be an excellent mimicker of a variety of different infectious and inflammatory respiratory diseases including OP. 33 Horii et al. 83 November 2020, Japan (70, Female) Radiological: Admission chest CT showed bilateral GGOs superimposed with interlobular reticulations and crazy-paving pattern that worsen on day 10 but improved on day 17. Treatment: Favipiravir, ciclesonide, nafamostat for 5-7 days but no improvement so oral prednisolone on day 13 with tapering regimen over 3 months.
[Survived] John et al. 84 September 2020, USA  85 August 2020, Japan (60, Male) Radiological: Admission chest CT showed bilateral patchy GGOs, predominantly in the subpleural areas but worsen on day 6 before improving on day 23 of hospitalization with corticosteroids. Treatment: Ciclesonide, favipiravir, levofloxacin with no improvement before 3 days of 1 gram IV methylprednisolone on day 6 and 80 mg/day oral prednisolone after for 3 weeks.
[Survived] (60, Female) Radiological: Bilateral consolidations with band-like shapes and distributed in subpleural and bronchovascular regions but worsen on day 6 before improving on day 17 with corticosteroids. Treatment: Azithromycin, ceftriaxone, hydroxychloroquine, favipiravir with no improvement before starting prednisolone 50mg/day on day 6 with taper over 3 weeks.
[Survived] Rea et al. 86 October   34 In that study, the radiological findings of OP conveyed better outcomes with less likelihood of ICU admission, invasive mechanical ventilation (IMV) requirement, mortality, and healing with pulmonary fibrosis than with a non-OP pattern. A case series by Vasu et al. reported that pleural effusions were notably more common in secondary OP (60% vs. 0%; p = 0.004), although pleural effusions were generally small in size. 14 Multiple observational studies assessing the COVID-19 radiological characteristics revealed that pleural effusion incidence, although highly variable (4.2-23.5%), increases in frequency after 12 days from onset of symptoms. 22,35À37 It is possible that pleural effusions likely represent a more severe inflammatory radiological evolution of COVID-19 pneumonia with increasing frequency during the hospital course and concomitant progression of opacities from GGOs to consolidations. This correlation of secondary OP has been seen in people with SARS, MERS, and influenza A, demonstrating radiological findings of alveolar opacities (GGOs and consolidations) predominantly in the peripheral, basilar, and bronchovascular lung regions with pleural effusion. However, no lung histopathological confirmation exists to solidify this hypothesis. 9À12

Histopathological findings of COVID-19 and OP
The histopathological findings that have been described during post-mortem examinations of patients with COVID-19 pneumonia are commonly diffuse alveolar damage (DAD), less frequently OP, and rarely acute fibrinous and organizing pneumonia (AFOP). 38À41 DAD in classic ARDS is divided into acute/exudative (1-8 days) and organizing/proliferative (9-17 days) phase. 42 The early acute/exudative phase is defined as alveolar injury leading to leakage of fluid and protein accumulation in the alveoli giving rise to the hyaline membrane. This will be followed by the late organizing/proliferative phase, with resolving hyaline membrane and interstitial infiltration by myofibroblasts with collagen deposition. 39,41,43À45 In OP, leakage of coagulative proteins occurs during alveolar injury that will trigger an accumulation of fibrin from diminished fibrinolytic activity with fibroblast activation and proliferation ensuing giving the intra-alveolar buds of granulation tissue (Masson bodies) appearance (Fig. 1). AFOP often termed a severe form of OP, is characterized by an extensive intra-alveolar fibrin deposition called "fibrin balls" as opposed to hyaline membranes seen in DAD. 41,46,47 Historically, AFOP was considered a variant of DAD due to the similarities shared involving clinical course and outcome; however, recent studies have refuted this claim to be closer to those of OP. 48,49 Clinically, AFOP likely represents a rapidly progressive course of OP with symptomatic presentation and exhibit extensive radiological findings of diffuse, bibasilarpredominant lung infiltrates. 6,7,43 It is reasonable to hypothesize that DAD, the most common lung histopathological finding reported, predominates in the early phase of COVID-19 pneumonia before the organizing phase of DAD ensues with transformation into OP and AFOP. 4,39,40,44,45 Hence, it is not infrequent to see areas of OP, AFOP, and DAD in the same autopsy specimen from a person dying of COVID-19 pneumonia. 39,41,42,45,47 In all of these different histopathological phenotypes, healing may occur in the form of fibrosis (fibrotic phase), weeks to months after initial injury with distortion of lung architecture depending on the severity of lung inflammation. In extreme cases, these changes may resemble usual interstitial pneumonia (UIP). 39,41,43À45 Two large post-mortem lung examinations of 100 people diagnosed with COVID-19 reported the histopathological findings of OP in up to 44% of cases. 45,50 The deceased had a prolonged illness duration [median 20 (5-58) days]. Autopsies of six people with COVID-19 who died later in their disease course (20 days after onset of symptoms) described histopathological findings of AFOP. 38 Similarly, postmortem histopathological OP and AFOP patterns of acute lung injury were also noted during the SARS and influenza A epidemic, especially in people suffering from prolonged illness duration (around 20 days and more). 47,51À54 It is important to note that the lung parenchymal involvement in OP is typically patchy, with normal lung architecture immediately adjacent to the disease. Therefore, it can be present as a bystander in many pulmonary disease processes, such as lung malignancy, lymphoma, connective tissue disease-associated ILD (CTD-ILD), vasculitis, non-specific interstitial pneumonia, chronic eosinophilic pneumonia, hypersensitivity pneumonitis (HP), and DAD. 4,6 The diagnosis of OP cannot be made unless it represents the predominant histopathologic abnormality. Thus, a transbronchial biopsy can be misleading but remains useful if clinical and radiological findings are compatible. The histopathological evidence of OP was

Therapies for COVID-19 and OP
Generally, corticosteroids are the cornerstone of therapy for OP with regression of symptoms within weeks and radiographic improvement within a month after secondary causes have been ruled out or treated. Current consensus recommends an initial dose of 0.75-1.5 mg/kg/day of prednisone for four weeks with a slow taper over 3-6 months. 6,7 It is not uncommon for relapse to occur during weaning of corticosteroids in 13 to 68% of cases when the corticosteroid dosing is reduced below 20 mg/day of prednisone equivalents or following the cessation of corticosteroids. A dramatic response to resumed therapy is observed in most cases, with doses of 20 mg prednisone daily often sufficient to treat relapse. Roughly one-quarter who suffer a relapse will go on to experience two or more relapses. 7,15 The risk of relapse is at its highest within the first year after the initial diagnosis of OP despite corticosteroid therapy. 5,7 Factors associated with an increased risk of relapse include delayed treatment initiation during the initial OP diagnosis and elevated liver enzymes. 15 The identification of secondary OP is clinically significant, as the management consists of treating OP and underlying disease. Corticosteroid therapy has emerged as an effective therapy for COVID-19 pneumonia, and current guidelines recommend the use of corticosteroid, primarily dexamethasone, as the mainstay therapy. 57,58 In contrast, corticosteroids are not recommended for the treatment of respiratory failure in the setting of SARS, MERS, and influenza A infection due to association with delayed viral clearance, increased in superimposed infection, and mortality. 59À61 In those studies, it is unclear how many patients had radiological patterns consistent with OP.
Several case series and case reports have reported the strong association between COVID-19 infection and secondary OP requiring high doses of corticosteroid therapy over a prolonged treatment duration, typically weeks to months (Table 1). 62À64 Intravenous (IV) treatment with methylprednisolone daily over 5-12 days was shown to be insufficient in preventing relapse in respiratory status after completion of therapy. In contrast, clinical improvement was achieved with 60 mg oral prednisolone daily with a gradual taper over a few months. 62,65 A single-center prospective observational study by Myall et al. assessed 325 COVID-19 patients post-discharge around 6-8 weeks from their hospital. 66 24% of COVID-19 patients who recovered had recurring dyspnea with restrictive limitations on pulmonary function tests and ongoing radiologic lesions concerning for OP. Highdose oral prednisolone was resumed in the outpatient setting with slow taper resulting in clinical and radiological improvements. A case series even suggested using high-dose corticosteroids beyond the recommended 10-day course for people with COVID-19 if there is no clinical improvement. 67 A prolonged period of corticosteroid therapy is frequently required to achieve remission from acute respiratory failure in the setting of OP and AFOP secondary to MERS and influenza A viruses. 52,68À71 At present, there is a lack of evidence that treatment of OP secondary to COVID-19 infection with corticosteroids, especially after ARDS, can prevent or reduce the development of residual pulmonary fibrosis. Macrolide therapy has even been studied to treat both OP and COVID-19 pneumonia, owing to their anti-inflammatory properties. However, a recent large, multi-center randomized control trial (RCT) study demonstrated a lack of clinical improvement in hospitalized COVID-19 patients receiving macrolide therapy of azithromycin. 72 Furthermore, the benefit of using macrolide therapy as an alternative option for people with OP diagnosis who are intolerant to corticosteroids or have frequent relapses is limited to small case series and case reports. 6,7 Therefore, a large, multi-center RCT is required to determine the indication, optimal dosage, and duration of corticosteroid therapy among those with suspected OP secondary to COVID-19 infection.

Limitations
Our review has several limitations. The majority of studies assessing OP secondary to COVID-19 are confined to case reports, case series, autopsies, and presumptive diagnosis in inadequate sample size observational studies, irrespective of histopathological confirmation (Table 1). Due to its rarity, OP is challenging to diagnose and likely under-recognized and under-reported, particularly in the setting of ARDS in which clinical features and radiological findings in OP resemble those of COVID-19 pneumonia. Furthermore, publication bias is likely for these reasons. The relationship between these two entities cannot be accurately assessed amid the current pandemic due to the scarcity of clinical data from the reluctance to perform aerosol-generating procedures such as lung biopsies and autopsies in order to reduce host transmission and conserve medical supplies. The use of bronchoscopy in people with COVID-19 has been recommended when an alternate diagnosis provided by BAL and/or transbronchial biopsies would significantly impact clinical management. 73 A low CD4+: CD8+ T-cell ratio on BAL can be seen in COVID-19 patients, but it does not contribute significantly to the diagnosis of secondary OP. 6,7,62 The BAL findings in OP are non-specific and often show mixed cellularity of neutrophils, lymphocytes, and eosinophils, although helpful in ruling out infectious and malignant etiologies, with similar sensitivity and negative predictive value compared to surgical lung biopsy. 6 BAL analysis may demonstrate lymphocytosis (>20%) only in those with secondary OP. 5,8 The mixed cellularity of white blood cells in BAL was observed in few case reports of COVID-19 patients diagnosed with biopsy-confirmed OP (Table 1). 62,64,65 Implications In view of the compatible clinical presentations, radiological abnormalities, and histopathological findings of secondary OP in COVID-19 patients (Table 1), concerns exist that the recommended dose of 6 mg dexamethasone daily for ten days is insufficient to achieve complete remission. COVID-19-induced dysregulated immune response termed cytokine storm typically manifests with elevated inflammatory markers, relapse in clinical symptoms, and multi-organ failures. 74 This phenomenon frequently occurs weeks after the initial injury from COVID-19 infection. As high inflammatory markers of erythrocyte sedimentation rate and C-reactive protein are also commonly seen in patients with OP, the timely administration of corticosteroids, especially in the early stage of the hyperinflammatory phase of cytokine storm, can effectively prevent and induce remission of COVID-19-related ARDS and secondary OP. 5À8, 75 Given the delayed onset of cytokine storm, these findings likely explain why respiratory symptoms and radiologic features of lung infiltrates continue for an extended period even after viral clearance was achieved and lengthy treatment with corticosteroids required. Upon ruling out superimpose and nosocomial infections, we recommend resuming high-dose corticosteroids in people with COVID-19induced respiratory failure when relapses occur, possibly at initial doses of 0.75-1.5 mg/kg/day of prednisone or equivalent for 2-4 weeks with a slow taper over several months. Although rare, spontaneous remission can be achieved without corticosteroid therapy in OP (Table 1). 76À81 There is currently no large-scale epidemiological data assessing this condition as a complication of COVID-19 infection, including the incidence, clinical course, outcomes, and treatment. For people diagnosed with secondary OP from COVID-19, a large, welldesigned RCT is required to determine: 1) the ideal duration and dosing of corticosteroids therapy such as short-term (10 days and less) versus long-term with a slow tapering regimen, 2) effective noninvasive means of diagnosing secondary OP from COVID-19 based on serum inflammatory markers and BAL findings, and 3) radiological and laboratory markers that are indicative of good corticosteroid response allowing weaning to occur.

Conclusion
Based on our findings, secondary OP should be considered as part of the differential diagnosis among patients with COVID-19 pneumonia, especially in the setting of persistent or relapsing respiratory symptoms and ongoing radiologic features of lung infiltrate even during the recovery phase. It is possible that, as has been described with other viruses, secondary OP in the setting of COVID-19 represents an immunological process after initial infection explaining the favorable response to immunosuppressive therapy of corticosteroids. We suspected that OP is much more common in COVID-19 infection despite the few reports to this day. OP can be present in isolation or combination with DAD and/or AFOP during post-mortem examinations of the lungs in deceased COVID-19 patients. Moreover, the timely diagnosis and treatment initiation may lead to less IMV requirement, redundant use of antibiotics, and improved overall survival. As our knowledge continues to grow during this current pandemic, with lung biopsies being increasingly warranted and performed in COVID-19 patients under the guidance of proper infection control protocol, we hope that a large, well-designed study can be implemented to determine the relationship of OP secondary to COVID-19 infection. A high clinical suspicion should exist for secondary OP when clinical deterioration occurs in people with COVID-19 when radiologic findings are compatible, especially in the setting of recent cessation of corticosteroid therapy.

Funding
None.

Author contribution
All authors had access to the data and were involved in writing the manuscript.

Declaration of Competing Interest
None.