Article Text

Download PDFPDF

Fibrosing organising pneumonia
  1. Brooke Beardsley1,
  2. Doris Rassl2
  1. 1Department of Anatomical Pathology, Canterbury Health Laboratories, Christchurch, New Zealand
  2. 2Department of Histopathology, Papworth Hospital, Cambridge, UK
  1. Correspondence to Dr Doris Rassl, Department of Histopathology, Sims Woodhead Building, Papworth Hospital, Papworth Everard, Cambridge CB23 3RE, UK; doris.rassl{at}


Organising pneumonia (otherwise referred to as bronchiolitis obliterans organising pneumonia) is characterised histologically by plugs of granulation tissue, which are present predominantly within small airways, alveolar ducts and peri-bronchiolar alveoli. This pattern is not specific for any disorder or cause, but is one type of inflammatory response to pulmonary injury, which may be seen in a wide variety of clinical conditions. Typically, organising pneumonia responds very well to corticosteroid treatment; however, a small percentage of patients appear to develop progressive fibrosis.

  • Lung
  • Inflammation
  • fibrosis

Statistics from


Organising pneumonia (OP) is a non-specific response to acute lung injury seen in a wide variety of clinical settings. It may be encountered in cases of aspiration, following bacterial, viral and fungal infections, distal to obstruction, following inhalation of toxic compounds, as a drug reaction and as a reaction to chemotherapy or radiotherapy.112 The OP pattern has also been observed in bone marrow transplant recipients, and in lung transplant patients it may represent an acute rejection response.13 ,14 Furthermore, OP can occur in patients with connective tissue diseases and it may be associated with other pathological processes such as hypersensitivity pneumonitis, eosinophilic pneumonia, eosinophilic granuloma, Wegener's granulomatosis and non-specific interstitial pneumonia (NSIP).1 ,3 ,10 ,12 ,15–20 The OP pattern sometimes represents a non-specific reaction at the periphery of unrelated pathological processes such as neoplasms or infarcts.1 ,3 ,10 Occasionally, associated histological features may give guidance as to the aetiology of the OP (such as foreign body giant cells and exogenous material in aspiration, or viral inclusions in viral pneumonias); however, the underlying cause is often not apparent from the pathological findings alone and clinical correlation is usually required.2 The term cryptogenic OP (COP), previously referred to as idiopathic bronchiolitis obliterans OP (BOOP), is reserved for those patients in whom no aetiology is found, and who present with a distinct clinicopathological syndrome. COP (idiopathic BOOP) is therefore a clinicopathological diagnosis of exclusion. At least 50% of OP cases fall under the heading of COP, which is classified within the idiopathic interstitial pneumonias.1 ,11 ,13 ,15 ,21

The separation of COP from secondary OP is of clinical importance, as the management of patients with secondary OP requires both the treatment of the OP, as well as the management of the underlying disorder or removal of any injurious agents.

Clinical and radiological features

OP shows no gender predilection and commonly affects people aged 40–60 years, with an average age of approximately 55 years.1 ,9–11 ,13 ,15 ,18 ,19 ,22–27 It can, however, occur at any age, including the paediatric age group.5 ,11 ,28–30 Studies that have examined the association of OP with cigarette smoking have produced conflicting results, and some studies have even suggested a protective effect of tobacco use.19 ,23 ,25 ,31–33

Typically, patients with OP experience a sub-acute presentation with symptoms occurring over a period of time, usually between 1 and 12 weeks, which include progressive dyspnoea, a non-productive cough and weight loss. A number of patients experience a flu-like illness with sore throat, fever and malaise. Other less common symptoms include haemoptysis and chest pain.1 ,9–13 ,15 ,18 ,19 ,22–27 ,30 ,34 ,35 Vasu et al9 noted that patients with COP were older, were more likely to be female, were less symptomatic and had longer symptom duration prior to diagnosis when compared with patients with secondary OP; however, these findings were not replicated by a later study.27

On examination, those affected typically have bi-basilar crackles, and pulmonary function tests generally show a predominantly restrictive ventilatory defect.1 ,10 ,13 ,19 ,22–27 ,30 ,33 ,34 Some patients, however, may have normal lung function studies, or may exhibit obstructive or mixed obstructive and restrictive patterns, the latter two possibly due to prior or concurrent patient smoking.1 ,9 ,13 ,23 ,26 ,27 ,30

The most common finding on chest x-ray is patchy airspace consolidation. The opacities may be multiple and bilateral, often wax and wane, and may appear migratory.1 ,9–12 ,15 ,19 ,23–27 ,30 ,33 ,34 High-resolution CT (HRCT) scans typically show similar features with bilateral patchy consolidation or ground glass opacities, which have a predominantly sub-pleural and peri-bronchiolar distribution. Unilateral areas of consolidation, reticular changes and localised nodules can also occur (figure 1). Air bronchograms are a frequent finding when consolidation is present.3 ,9 ,11 ,18 ,24 ,26 ,27 ,30 ,36 ,37

Figure 1

CT images demonstrating a range of CT appearances encountered in patients with organising pneumonia: (A) bilateral nodular areas of consolidation with associated ground glass changes; (B) localised ground glass opacity; (C) bilateral irregular areas of consolidation (courtesy of Dr N Screaton, Papworth Hospital).

Differential cell counts in bronchoalveolar lavage (BAL) fluid usually show a lymphocytosis with lesser increases in neutrophils and eosinophils.6 ,11 ,15 ,22–25 ,30 ,33 A decreased CD4/CD8 ratio may be seen when compared with normal controls.11 ,24 ,25 ,33 Drakopanagiotakis et al27 found that patients with secondary OP had a higher mean lymphocyte count.

Pathological features

The initiating event is alveolar epithelial injury with denudation of the basal laminae, intra-alveolar oedema and fibrin. Fibroblasts and inflammatory cells migrate through gaps in the denuded basement membranes, forming fibro-inflammatory buds (figure 2), organisation of which leads to the formation of polypoid plugs of granulation tissue within the lumen of bronchioles, alveolar ducts and peri-bronchiolar alveoli, recognised histologically as OP. The granulation tissue is composed of fibroblasts and myofibroblasts within an oedematous or myxoid, pale staining stroma (figure 3), often with admixed inflammatory cells, including lymphocytes, plasma cells and macrophages.1–3 ,10 ,22 ,32 ,35 Increased numbers of capillaries can also be demonstrated. The rounded masses of granulation tissue are sometimes referred to as ‘Masson bodies’ and may extend between alveoli via the pores of Kohn, giving a butterfly appearance.22 ,35 There is regeneration of the surface epithelium, with re-epithelialisation of the basal laminae, important for maintaining alveolar structural integrity. Epithelial cells sometimes also extend over the surface of the granulation tissue plugs, which then may become incorporated into the interstitium.2 ,35 A mild degree of interstitial chronic inflammation is often apparent; however, there is little interstitial fibrosis and the normal lung architecture is generally preserved.3 ,22 The process is usually patchy with sharp demarcation from the adjacent normal parenchyma, and the changes of OP are temporally homogeneous—that is, the luminal granulation tissue is of the same age throughout, which suggests damage occurring at a single point in time.1 ,10 This is in contrast to usual interstitial pneumonia (UIP), where the process is temporally heterogeneous in that there are immature interstitial fibroblastic foci adjacent to areas of more mature, well established interstitial fibrosis. The fibroblastic foci in UIP are interstitial rather than intraluminal, and lack the serpiginous shapes and peribronchiolar distribution.

Figure 2

Fibrin and inflammatory cells within airspaces, forming early fibroinflammatory buds, organisation of which leads to the formation of polypoid plugs of fibroblastic granulation tissue (H&E ×200).

Figure 3

Light staining fibroblast plugs filling airspaces (H&E ×100).

Some cases of secondary OP may display features of a primary process, for example granulomatous inflammation in hypersensitivity pneumonitis or foreign body giant cells in aspiration (figure 4).

Figure 4

Focal organisation with a foreign body type giant cell, suggestive of aspiration (H&E ×100).

Treatment and outcome

Approximately 60–80% of patients will respond to corticosteroid therapy and have complete clinical and physiological resolution of their disease.1 ,9 ,12 ,23 ,31 ,32 Generally, patients also have resolution of disease on follow up x-rays; however, serial high resolution CT scans may show some residual changes on follow-up studies.18 ,38 In most of these cases, the CT appearances resemble fibrotic NSIP.38 Typically, clinical improvement is rapid following treatment with corticosteroids; however, relapses, even during initial steroid therapy, do occur with different series reporting relapse rates between 9% and 80%. Relapses may occur if the steroid dose is decreased too quickly, and multiple and late relapses have also been reported.1 ,6 ,11 ,13 ,16 ,18 ,25–27 ,30 ,31 ,34 ,38 ,39 The vast majority of these patients respond to further steroid therapy and relapses do not appear to affect long term outcome in most cases.1 ,13 ,31 ,34 Recurrences may be more likely in patients with severe hypoxaemia at presentation and in patients with an underlying disease. Laboratory abnormalities, such as high erythrocyte sedimentation rate, low serum albumin and severe anaemia, are more common in patients with secondary OP and can be associated with a worse prognosis.27

Multiple relapses have been reported to be encountered more frequently in patients with mild cholestasis or where treatment is delayed.16 ,31 ,39 Nodular forms of OP have a very good prognosis, as the lesion is usually surgically removed because it simulates a tumour. In most cases this treatment is curative, although some patients may experience recurrent disease.2 ,11 ,12 ,22 ,26 ,37 ,40

Spontaneous improvement without corticosteroid therapy has also been reported in a number of patients.1 ,15 ,25 ,30 ,33 ,35

Unfortunately, not all patients will respond to treatment and some authors estimate 10–15% will experience progressive disease.16 ,22 ,32 Death from OP has been reported to occur in approximately 4–14.5% of patients; however, in a number of series it is difficult to ascertain whether this was purely due to progression of the OP or whether in some patients other complicating conditions may have contributed.11 ,13 ,15 ,16 ,30 ,41

In patients who did not respond to initial corticosteroid treatment, alternative agents have been trialled, with some reports of success. These agents include azathioprine, cyclophosphamide and inhaled triamcinolone.12 ,23 ,42 However, the use of these drugs has mainly been described in case reports, and larger studies are required.

In some patients, who were not treated with corticosteroids, a response to tetracycline or erythromycin was reported, even though infection was never confirmed, which raises the possibility that in some instances OP may be an atypical response to an unidentified infectious agent.13 Macrolides, such as clarithromycin and azithromycin, may also be used as alternative or adjuvant agents, and have an immunomodulatory effect.43

Progressive OP

As discussed above, not all patients share the generally good prognosis of OP and there appears to be a small subset who go on to develop progressive fibrosis.

There are some authors who believe that clinically aggressive OP can be explained by poor sampling and/or incorrect morphological diagnosis or categorisation, particularly prior to the 2002 publication of the ATS/ERS diagnostic criteria.21 ,42 Early papers are likely to have included entities other than what we would classify as OP today, particularly NSIP cases, thereby affecting descriptions of outcomes in pure OP.42 Additionally, OP frequently coexists with other interstitial pneumonias, especially when associated with connective tissue diseases.1 ,10 ,32 ,42 Therefore, for patients who do not respond to treatment it is important to determine if the OP pattern is secondary to some other underlying process.

Clinical aspects

COP generally shows a good response to corticosteroid treatment with a 5-year survival of 73%, but patients with OP secondary to another chronic disease process, especially connective tissue disease, autoimmune disease, or exposure to drugs or environmental agents, appear to have a poorer prognosis.1 ,10 ,16 ,19 ,23 ,26 ,32 Those patients with associated connective tissue disorders seem to have lower complete recovery rates and a tendency towards higher recurrence when compared with COP patients.16 ,26 One important factor appears to be ongoing lung injury. Those cases secondary to an adverse drug reaction may progress if the drug therapy is maintained and patients with connective tissue disorders may have chronic lung injury as part of their disease.32 This recurrent injury to the lung parenchyma is likely to be the driving force that prevents normal regression of the OP and promotes development of alveolar septal inflammation and fibrosis.19 ,32

Radiological findings

Several studies have noted that patients with radiographic or CT evidence of reticulo-nodular interstitial opacities on radiology have worse outcomes with persistent radiological abnormalities and some with progressive disease, compared with patients who have localised infiltrates or alveolar opacities.10 ,22 ,44 These patients had a longer duration of symptoms prior to diagnosis compared with those with the classical findings of diffuse interstitial opacities.22 A study correlating CT findings in OP with clinical, radiographic and histological features, found that a reticular pattern on CT was associated with histological findings atypical for OP, including thickening and fibrosis of the interstitium.18

Pathological features

Rarely, death may result from a rapidly progressive form of OP.12 ,19 ,41 ,45 It is difficult to interpret these reports, however, as in some cases there appears to be an overlap with organising diffuse alveolar damage, and often the histological features described are atypical for pure OP, such as marked acute and chronic interstitial inflammation and interstitial widening.19 ,41 ,45 ,46 Many patients who subsequently underwent a postmortem examination were found to have more severe interstitial inflammation and alveolar septal fibrosis, some with honeycombing, than identified in their pre-mortem samples.19 ,41 The authors of one study suggested that this meant that COP could represent an early phase in the temporal spectrum of interstitial lung disease.19 However, the features described are suggestive of a UIP pattern and raise the possibility that at least some of these patients may have had an underlying fibrotic process, which may ultimately have led to their death. Authors of a later study specifically comment that in their series the patients who died of a rapidly progressive and steroid non-responsive OP (6.6%) had no evidence of underlying fibrotic disease on HRCT and no history of connective tissue disease, so a rapidly progressive form of OP may exist.16 Whether these clinical differences are due to a difference in the extent of lung injury or a difference in the response to injury is unclear.

King et al23 suggested that ample lung tissue must be reviewed to rule out other diseases when diagnosing cryptogenic forms of OP, and the same could be said for ruling out underlying disorders that may lead to progressive fibrosis. In this regard transbronchial biopsies may not be adequate, and video assisted thorascopic surgery (VATS) or open lung biopsies remain the gold standard. However, others have advocated that if a patient presents with clinical features and radiological evidence suggestive of OP, with characteristic pathological material obtained via a transbronchial biopsy, then the risks of a more invasive lung biopsy are not justified and a trial of steroids should be considered.4

Yousem et al32 also found that steroid non-responsive patients with OP were far more likely to have histological evidence of remodelling of the pulmonary parenchyma with thickening and fibrosis of the alveolar septa. The changes consisted of increased septal collagen and reduction of elastic tissue fibres, with reactive changes in alveolar lining cells. No honeycomb changes were identified. It was noted that eight of the nine patients in the steroid non-responsive group showed dense hyalinisation and fibrosis of the central cores of the Masson bodies (figures 5 and 6). Ultimately there was restructuring of the lung architecture, however it was not possible to determine if the interstitial fibrosis was secondary to the COP or was present prior to the COP episode. A possible explanation for the progressive nature of these cases is that there was a pre-existing chronic interstitial lung disease with a superimposed COP reaction.

Figure 5

(A) Dense hyalinisation and fibrosis of Masson bodies (H&E ×100). (B) Collagen deposition in buds of fibroblastic tissue (Elastic Van Gieson (EVG) ×200).

Figure 6

Collagenised Masson bodies becoming incorporated into the interstitium (Elastic Van Gieson (EVG) ×100).

There is some evidence, however, to support a progressive OP as a primary process, in that two cases from Yousem et al's32 group of patients with progressive COP underwent an autopsy which showed similar changes to that in the pre-mortem biopsy, and perhaps more importantly did not show any evidence of UIP. Furthermore, the length of symptoms in these patients was short (average 3.4 weeks), rather than the prolonged history of symptoms usually elicited from UIP patients, and no patient had honeycomb changes identified on radiography.

As there appears to be a proportion of patients with pure OP who develop progressive fibrosis, several factors have been identified as being associated with a poor outcome. These include a lack of lymphocytosis in BAL fluid differential cell counts, OP associated with other disorders (particularly connective tissue diseases), a predominantly interstitial pattern on imaging, and a finding on histological examination of scarring and remodelling of the lung parenchyma in addition to OP.1 ,2 ,10 ,11 ,19 ,22 ,23 ,26 ,32 ,39 ,44

Some have suggested that the finding of a more marked increase in neutrophils or eosinophils in BAL fluid, as opposed to a predominant lymphocytosis, is indicative of increased disease activity and progression, and may predict a poor response.2 ,11 ,19 However, this correlation was not confirmed by Drakopanagiotakis et al,27 who did not find a significant correlation between BAL lymphocyte levels and relapse rates.

Mast cells and tryptase are increased in BAL fluid of patients with OP, and the lavage fluid has also been found to show increased monocyte chemotactic protein-1, IL-10, IL-12 and IL-18 levels with respect to controls and patients with UIP. This is consistent with marked macrophage and lymphocyte activation, with an expansion of a T-helper type-1 response in COP.44

Disease mechanisms

Attempts have been made to identify aspects of the granulation tissue plugs seen on histology in OP that may predict a worse prognosis for the patient. Yoshinouchi et al35 found that patients with fibrin negative and α-smooth muscle actin negative Masson bodies had complete resolution of their radiological abnormalities, whereas radiological abnormalities persisted, despite steroid therapy, in all five patients with fibrin or α-smooth muscle actin positivity. They suggested that this indicated two different disease types and that the differences in outcome were attributable to differences in the processes of organisation. The formation of fibrin results from an imbalance in the alveolar lumen between the coagulation and fibrinolytic cascades, and the coagulation and fibrinolysis factors and inhibitors (especially plasminogen activator inhibitor-1) play a complex role in fibrogenesis.44

Ranzani et al47 compared the granulation tissue plugs seen in idiopathic and secondary OP, and found that plugs in secondary OP showed less collagen deposition, but increased microvascular density, increased endothelial activity and increased numbers of α-smooth muscle actin positive myofibroblasts. Although this study did not address outcomes in these patients, it is possible that some of these features could predict behaviour given that in general secondary OP tends to have a worse prognosis than the idiopathic type.

Remodelling of the pulmonary parenchyma is a complex process, involving interactions between numerous cell types, which is still not completely understood.47 Many researchers have attempted to elucidate the differences between the processes involved in idiopathic pulmonary fibrosis (the histological correlate of which is UIP) and OP, as these two fibrosing diseases generally have very different prognoses. UIP is a disease which is marked by progressive interstitial fibrosis with architectural remodelling leading eventually to death or lung transplantation. OP however consists of granulation tissue plugs within alveolar spaces, usually does not involve significant interstitial fibrosis or architectural remodelling, and generally responds well to corticosteroids.1 It has been proposed that ordinarily OP represents a model of ‘normal’ parenchymal healing, with appropriate extracellular matrix (ECM) remodelling, fibroblast-myofibroblast apoptosis and alveolar re-epithelialisation. Conversely, UIP represents a model of ‘abnormal’ wound healing with production of excessive ECM, decreased fibroblast-myofibroblast cell death, ongoing epithelial cell apoptosis and abnormal re-epithelialisation.48 It is possible that in cases of OP, which go on to develop progressive fibrosis, the normal healing process is disrupted.

Some have suggested that insufficient or delayed re-epithelialisation of alveolar walls may be important in the development of UIP, and examination of fibromyxoid lesions of both COP and UIP has shown that re-epithelialisation is usually more abundant in COP.49 ,50 There is also evidence that disruption of the basement membrane in UIP may contribute to inadequate re-epithelialisation.51 ,52 In COP most basal laminae are not destroyed, although some gaps are present.44

Alveolar pneumocyte apoptosis has been identified adjacent to fibroblastic foci in UIP and it has been proposed that ongoing apoptosis following epithelial cell injury may be a key component in the progression of fibrosis.53 As well as demonstrating more abundant re-epithelialisation, COP fibromyxoid lesions also exhibit more abundant vascularisation, increased expression of vascular and fibroblast growth factors, and higher apoptotic activity when compared with fibromyxoid lesions in UIP. The authors of these papers have suggested that these features may play an important role in the resolution of the newly formed connective tissue in typical OP.50 ,54–56

Damaged alveolar epithelial cells in UIP have been shown to release a number of pro-fibrotic cytokines, growth factors and enzymes, suggesting that altered type 2 pneumocytes may play a significant role in ECM remodelling.52 Additionally, a process known as ‘epithelial–mesenchymal transition’ has been described whereby under certain conditions re-epithelialisation by type 2 pneumocytes by differentiation into type 1 cells is inhibited (eg, by ongoing injury) and instead type 2 cells may undergo a metaplastic transition to fibroblasts and myofibroblasts.48 ,57–59 The fact that this is a metaplastic process could explain the low proliferative activity of fibroblasts and epithelial cells found in UIP.60 These metaplastic myofibroblasts could induce ongoing alveolar epithelial cell apoptosis and participate in parenchymal remodelling and destruction via several mechanisms, including increased ECM synthesis and production of matrix metalloproteinases (MMPs).51 ,53 ,57 Therefore, ongoing injury to the alveolar epithelium could explain the poor prognosis and propensity for progressive fibrosis in cases of OP secondary to an underlying disorder. The metaplastic myofibroblasts may correspond to the α-smooth muscle actin positive cells described above, which were seen by Yoshinuchi et al35 in the granulation tissue plugs of patients with OP who did not respond to corticosteroid therapy.

El-Zammar et al60 found that proliferative rates of fibroblasts were similar in UIP and COP, however the proliferative rate of macrophages was significantly higher in UIP. This suggested that macrophages played a role in the pathogenesis of UIP. The fibroblasts in UIP and COP also showed similar proliferation rates to keloid scars rather than normal skin scars. They postulated that this may indicate that the fibroblasts are abnormal, leading to abnormal healing. This may be true of UIP, however other evidence of abnormal wound healing is generally not seen in typical COP.

ECM turnover is partly regulated by a balance between matrix metalloproteinases (MMPs) and tissue inhibitors of metalloproteinases (TIMPs). In UIP an imbalance between these two groups of molecules has been found, with a wider distribution of TIMPs in comparison with collagenases (such as MMP-1 and MMP-8), which leads to progressive deposition of dense fibrillar collagen.51 This is in contrast to OP where it is postulated that the corticosteroid sensitivity may be due both to the collagen being of type III, which is newly formed, flexible and susceptible to enzymatic digestion (rather than the more mature type I), and to the abundance of MMPs within the Masson bodies.46 ,51 ,61


OP is a pattern of lung injury, which is seen in a wide variety of clinical contexts, with cases divided into cryptogenic and secondary forms. The pattern comprises granulation tissue plugs within small airways and adjacent alveoli with minimal interstitial inflammation or fibrosis. Typically there is no associated architectural remodelling, and in general most patients will respond extremely well to corticosteroid therapy with complete symptomatic and physiological resolution of disease. There are however a small proportion of patients who will experience progressive fibrosis and a poor prognosis. Several factors have been identified which may suggest a poorer prognosis including OP secondary to an underlying disorder such as a connective tissue disease, a predominantly interstitial pattern on imaging, scarring and remodelling of the lung parenchyma on histological examination, and a lack of lymphocytosis on BAL fluid. Additionally, fibrin and α-smooth muscle actin positive cells present within the granulation tissue plugs may indicate a worse prognosis. Factors which may lead to abnormal healing and progressive fibrosis include ongoing alveolar epithelial cell injury with epithelial mesenchymal transition of epithelial cells to fibroblasts and myofibroblasts. These metaplastic cells may then contribute to further epithelial cell damage and ECM remodelling leading to fibrosis. Further research into understanding the mechanisms responsible for progressive fibrosis in this small percentage of OP cases may lead to the identification of novel therapeutic targets. It is possible that at least some of the mechanisms of fibrosis in UIP are shared by these OP patients, therefore any novel treatments used successfully in UIP may also translate into treatments for patients with fibrosing OP.

Take-home messages

  • OP is a pattern of lung injury characterised by plugs of granulation tissue within small airways and peribronchiolar alveoli.

  • OP may be crytogenic or associated with other conditions.

  • Most cases of OP resolve with corticosteroid therapy; a smaller number may show progression with fibrosis.

  • In cases of OP which progress the possibility of an underlying cause should be considered.



  • Dr Nick Screaton kindly provided CT images to be included in the article.

  • Contributors BB assisted in writing the review article .

  • Competing interests None.

  • Provenance and peer review Commissioned; externally peer reviewed.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.