Idiopathic Pulmonary Fibrosis

Idiopathic pulmonary fibrosis (IPF) and cryptogenic fibrosing alveolitis (CFA) 


The synonymous terms idiopathic pulmonary fibrosis (IPF) and cryptogenic fibrosing alveolitis (CFA) refer to a relentlessly progressive fibrotic lung disorder that is the underlying diagnosis in over one-half of patients presenting with typical clinical features of the ‘CFA clinical syndrome’. Incidence is about 10 to 15 per 100 000, men are more often affected than women, and it most commonly presents in the seventh and eighth decades. Aetiology remains uncertain.

Clinical features—typical presentation is with progressive exertional dyspnoea, without wheeze, and a nonproductive cough. Digital clubbing is present in over 50% of patients. Very fine end-inspiratory crackles are usually heard bilaterally at the lung bases and become widespread in advanced disease. Central cyanosis and clinical evidence of pulmonary hypertension are late features.

Diagnosis—this requires a surgical biopsy revealing a usual interstitial pneumonia (UIP) histological pattern in association with (1) the absence of other known causes of interstitial lung disease, (2) a restrictive lung function profile, and (3) compatible features on chest radiography or high-resolution CT scans. Diagnosis in the absence of a surgical biopsy requires (1) the absence of other known causes of interstitial lung disease, (2) a restrictive lung function profile, (3) high-resolution CT appearances of predominantly basal reticular abnormalities with honeycombing and little or no ground-glass attenuation, and (4) no features of an alternative diagnosis in transbronchial lung biopsy or bronchoalveolar lavage; together with at least three of the following: (a) age over 50 years, (b) insidious unexplained exertional dyspnoea, (c) duration of illness exceeds 3 months, and (d) predominantly basal or widespread crackles on auscultation of the chest.

Differential diagnosis—the distinction between IPF and fibrotic nonspecific interstitial pneumonia (NSIP)  poses particular difficulty and is crucial because of their very different prognoses: the 5-year survival is 10 to 15% in IPF compared to over 60% in fibrotic NSIP.

Management—in definite IPF there is little evidence that traditional treatment regimens have a major impact on outcome. The use of antioxidant (N-acetylcysteine) and low-dose steroid therapy (prednisolone 10 mg/day), with or without an immunosuppressive agent (azathioprine), appears reasonable and is the therapeutic approach preferred by the authors, and the routine use of such a regimen can be justified when a diagnosis of fibrotic NSIP is possible. In 10 to 15% of patients with IPF there is an accelerated deterioration over several weeks that often leads rapidly to a fatal outcome: intravenous high-dose corticosteroids together with intravenous cyclophosphamide are often used in this circumstance.

Diffuse parenchymal lung disease

Idiopathic pulmonary fibrosis in great detail


The disorder previously known as fibrosing alveolitis, first described in 1907, was increasingly recognized following the description of a small group of patients with rapidly progressive fatal disease, grouped as the Hamman–Rich syndrome. Until late in the 20th century, a stereotypical clinical presentation of idiopathic interstitial lung disease was termed idiopathic pulmonary fibrosis (IPF) or cryptogenic fibrosing alveolitis (CFA), and a number of histological patterns were unified under this term. However, it became increasingly clear that the clinical presentation of IPF/CFA (‘CFA clinical syndrome’) was shared by a number of diseases, including predominantly inflammatory and predominantly fibrotic disorders, now known collectively as the idiopathic interstitial pneumonias. Their separation is essentially pragmatic, being justified by large differences in treated outcome. The classification proposed by an American Thoracic Society/European Respiratory Society (ATS/ERS) nomenclature committee is now widely accepted and can be readily applied to routine practice, with increasing recognition of characteristic patterns of disease on high-resolution CT. Histological evaluation is reserved for a minority of patients in whom accurate management cannot be based on clinical and high-resolution CT findings alone. The synonymous terms IPF and CFA now refer to a relentlessly progressive fibrotic disorder, associated with a histological pattern of usual interstitial pneumonia (UIP) or typical high-resolution CT and clinical features in nonbiopsied cases.

The remainder of this article is devoted to IPF. Epidemiological and aetiological data are briefly reviewed and the clinical picture is summarized. Key clinical issues are then discussed, including diagnosis, prognostic evaluation, routine monitoring and treatment.

Epidemiology and aetiology

IPF is the underlying diagnosis in over one-half of patients presenting with typical clinical features of the CFA clinical syndrome, hence the epidemiology is necessarily that of the CFA clinical syndrome as epidemiological studies cannot be confined to younger patients undergoing a surgical biopsy.

The CFA clinical syndrome may develop at any time but most commonly presents in the seventh and eighth decades: in childhood, it has a largely benign outcome and is not associated with UIP. The syndrome has a slight male predilection and exhibits considerable geographical variation. The incidence and prevalence have risen steadily in recent decades, with the incidence now about 10–15 per 100 000, based on evaluation of death certificates and registry studies in the United States of America, the United Kingdom, and elsewhere. A recent study, using case definitions more reliably indicative of IPF, has suggested an incidence of 5 to 10 per 100 000.

The aetiology remains uncertain, but a single cause appears unlikely, even in patients with biopsy-proven IPF. Case–control studies of the CFA clinical syndrome have consistently identified an association with smoking, although the strength of the association has diminished in recent years as smoking itself has become less prevalent. A number of viruses, including especially the Epstein–Barr virus, have been proposed as trigger factors, but data are contradictory. Occupational exposure to metal dusts and wood fires has also been implicated in controlled evaluations. Thus, it appears that a number of airborne insults may trigger the CFA clinical syndrome, in keeping with the prominent lung epithelial damage observed in IPF in particular and in the idiopathic interstitial pneumonias in general.

Genetic factors have been widely evaluated but remain obscure. In 10% of patients with IPF there is a family history of a pulmonary fibrotic disorder, but although IPF is the most frequent disorder, other idiopathic interstitial pneumonias may develop in family members, suggesting a general predisposition to pulmonary fibrotic disease. Recent data have suggested that smoking and advancing age are the cardinal risk factors for the development of familial disease. The Hermansky–Pudlak syndrome is a rare genetically transmitted pulmonary fibrotic disorder characterized by oculocutaneous albinism and abnormal platelets.

Diagnostic criteria

The diagnostic criteria for IPF, stated by the ATS/ERS nomenclature committee, are now widely accepted.

At surgical biopsy, a usual interstitial pneumonia (UIP) histological pattern is required in association with:

  • the absence of other known causes of interstitial lung disease (drug toxicity, occupational or environmental exposures, rheumatological disorders)
  • a restrictive lung function profile (reduced vital capacity (VC) often with an increased FEV1/FVC ratio) or an isolated reduction in the carbon monoxide diffusing capacity (D LCO))
  • compatible features on chest radiography or high resolution CT

In patients not undergoing a surgical lung biopsy, diagnostic criteria are more rigorous. See box below:

Box: diagnostic criteria

All major criteria and at least three of the four minor criteria are required.

Major criteria
  •  Absence of other known causes of interstitial lung disease (drug toxicity, occupational or environmental exposures, rheumatological disorders)
  •  A restrictive lung function profile (reduced vital capacity (VC) often with an increased FEV1/FVC ratio) or an isolated reduction in the carbon monoxide diffusing capacity (D L CO))
  •  High-resolution CT appearances of predominantly basal reticular abnormalities with honeycombing and little or no ground-glass attenuation
  •  No features of an alternative diagnosis in transbronchial lung biopsy (granulomas) or bronchoalveolar lavage (an excess of lymphocytes)
Minor criteria
  •  Age >50 years
  •  Insidious unexplained exertional dyspnoea
  •  Duration of illness >3 months
  •  Predominantly basal or widespread crackles on auscultation of the chest

It should be appreciated that a requirement for bronchoscopy is unrealistic in many elderly patients and in those with advanced disease and major respiratory limitation. In these contexts, a diagnosis of IPF may be made without bronchoscopic support.

Histological features and pathogenesis

In UIP, the histological pattern underlying IPF, temporal and spatial heterogeneity of disease is the cardinal feature. Normal lung is seen adjacent to regions of fibrosis, with enlarged cystic air-spaces (honeycomb lung) and areas of milder interstitial fibrosis. A patchy chronic inflammatory cell infiltrate is variably present. Subepithelial foci of proliferating fibroblasts (‘fibroblastic foci’) are a characteristic feature: these occur occasionally and sparsely in nonspecific interstitial pneumonia (NSIP) but are not seen in other idiopathic interstitial pneumonias.

Historically it was believed that inflammation was the key pathogenetic process, preceding and leading to fibrotic disease, but this view has been largely abandoned (although inflammation may be an ancillary pathogenetic feature). Corticosteroid and immunosuppressive therapy, effective in primary inflammatory disorders, has at best a minor beneficial effect in slowing disease progression. There is increasing evidence that IPF has an epithelial fibrotic pathogenesis, with initial epithelial damage leading to the formation of fibroblastic foci and subsequently to more widespread thickening of the connective tissue matrix in advanced disease. Thus, IPF can be conceptualized as a disorder of abnormal wound healing. In established disease, lung injury—an immunological and inflammatory response—and fibrogenesis appear to occur in parallel. It is not known whether a single key mechanism is pivotal in pathogenesis. Oxidant–antioxidant imbalance and the release of damaging enzymes from inflammatory cells appear to amplify injury, but a wide variety of biological mechanisms interact in the lungs of patients with IPF, with up-regulation of tumour necrosis factor α (TNFα) and chemokines (interleukin (IL)-8, and growth factors, especially transforming growth factor β (TGFβ) and connective tissue growth factor), and activation of the coagulation cascade, known to promote fibrogenesis. However, it is not clear whether these mechanisms are primarily pathogenetic or represent physiological responses to some upstream abnormality. The profusion of fibroblastic foci and serum levels of protein markers of epithelial damage have both been linked to mortality and disease progression.

Presentation and features on investigation

Clinical features

The typical presentation is progressive exertional dyspnoea without wheeze and a nonproductive cough, although sputum production is present in few patients. Haemoptysis should prompt investigation for lung malignancy, which is approximately 10-fold more prevalent in IPF after the smoking history has been taken into account. Chest discomfort, fatigue, and weight loss are occasional features. Digital clubbing is present in over 50% of patients and has been an adverse prognostic determinant in some series. On auscultation, very fine end-inspiratory crackles are typically heard bilaterally at the lung bases and become widespread in advanced disease. Central cyanosis and clinical evidence of pulmonary hypertension, with or without right ventricular failure, are late features. Evidence of rheumatological disease, which is generally indicative of a much better outcome, should be sought from the history and examination.

Chest radiography

The chest radiograph typically shows small lung volumes and predominantly peripheral and basal reticulonodular shadowing, with obscuration of the heart borders and diaphragms in advanced disease and overt honeycombing in 10% of cases. However, this profile is very nonspecific, occuring in fibrotic NSIP, asbestosis, rheumatological disorders, and other fibrotic processes. Lymphadenopathy or pleural disease should suggest an alternative diagnosis or a concurrent pathological process. The heart may appear to be enlarged in the absence of cardiovascular disease as a result of reduced intrathoracic volume.

High-resolution CT

High-resolution CT appearances are virtually pathognomonic in up to 70% of patients. The disease is predominantly posterobasal and peripheral, becoming widespread in advanced disease, and consists of a reticular pattern, with or without honeycombing, and a minor component of ground-glass attenuation, usually indicative of fine fibrosis (rather than inflammation). It should be stressed that high-resolution CT appearances are atypical in at least 30% of cases, with the most frequent variant consisting of prominent ground-glass attenuation admixed with a fine reticular pattern and associated with traction bronchiectasis, which is an appearance also suggestive of NSIP. However, IPF is also diagnosed (i.e. a histological pattern of UIP at biopsy) in occasional patients with markedly atypical high-resolution CT appearances. Reactive mediastinal lymphadenopathy is usual on high-resolution CT and is not indicative of a coexisting disease process unless also present on chest radiography. In early disease, prone high-resolution CT sections may be required to distinguish abnormal appearances from normal increases in density due to gravity-related increases in perfusion in dependent areas.

Other imaging modalities

Ventilation–perfusion scans show ventilation mismatch due to vascular ablation in areas of cystic lung, which typically continue to ventilate normally. These appearances simulate pulmonary thromboembolism and probably account for a widespread misperception that pulmonary embolism is a frequent complication of IPF. CT pulmonary angiography is required when pulmonary embolism is suspected, especially when D LCO levels are disproportionately reduced, but is usually negative in this context.Gallium scanning almost invariably reveals abnormal signal, as in other forms of pulmonary fibrosis associated with macrophage activation, but does not contribute usefully to prognostic evaluation or management. Rapid clearance from the lung of inhaled technetium-99m diethylinetriamine pentacetate (DTPA) may be indicative of a more progressive course, but rapid clearance is also seen in healthy smokers and the test has no established role in routine management.

Lung function tests

Lung function tests reveal a restrictive ventilatory defect, as shown by reductions in vital capacity, total lung capacity, residual volume and pulmonary compliance. However, the wide range of normal premorbid lung volumes sometimes results in apparent normality (when lung volumes have fallen from the upper to the lower end of the normal range). Thus, measures of gas transfer, especially D LCO levels, may be reduced in isolation in early disease. Adjustment of D LCO for reduced alveolar volume (K CO) has been advocated as a more specific index of interstitial fibrosis, but K CO levels are disproportionately reduced by coexistent emphysema, which is present in over 30% of IPF patients. The combination of emphysema and IPF may result in spurious preservation of lung volumes, even in advanced disease, and disproportionate reduction in D LCO levels. Overall, the severity of disease is most accurately captured by D LCO levels, which correlate best with the extent of IPF as judged by high-resolution CT.

In early disease arterial gases may be normal, but mild arterial hypoxia with widening of the alveolar-arterial gradient and normal or low PaCO 2 levels are usual. Severe hypoxia is a late feature and increased PaCO 2 levels occur in terminal disease.

Blood tests

Blood tests contribute little to the management of IPF, except in rare cases in which an unsuspected underlying cause is identified. Mild increases in the erythrocyte sedimentation rate, serum immunoglobulins, rheumatoid factor and antinuclear antibodies are frequent, and secondary polycythaemia may occur in severe disease. A high neutrophil count may be indicative of infection but a moderate increase is also seen in association with corticosteroid therapy. However, striking increases in autoantibodies may be indicative of a hitherto undiagnosed rheumatological disorder. Precipitin tests against fungal and avian antigens should be performed when there is suggestive exposure history because chronic extrinsic allergic alveolitis occasionally presents with the CFA clinical syndrome, high-resolution CT appearances suggestive of IPF, and a pattern of UIP at surgical biopsy.

Bronchoalveolar lavage

Bronchoalveolar lavage is a useful ancillary diagnostic test when a surgical biopsy is not performed. Typically, there is an increase in total cell counts with an excess of neutrophils and/or eosinophils. A mild lymphocytosis is not infrequent, but striking rises in lymphocyte counts are not generally a feature of IPF and suggest an alternative disorder such as NSIP, extrinsic allergic alveolitis, fibrotic sarcoidosis, cryptogenic organizing pneumonia complicated by interstitial fibrosis, or drug-induced lung disease. Bronchoalveolar lavage is occasionally useful in excluding opportunistic infection in treated patients.

Surgical lung biopsy

A surgical lung biopsy is the histological diagnostic procedure of choice. Video-assisted thoracoscopic biopsy is the most widely used procedure, but minithoracotomy is occasionally required in advanced disease. It is strongly recommended that at least two sites are biopsied, with high-resolution CT findings taken into account to ensure that the full spectrum of morphological abnormalities is sampled and to avoid areas of endstage disease which seldom yield diagnostic tissue. The diagnosis of IPF and other idiopathic interstitial pneumonias cannot be based upon appearances at transbronchial biopsy: larger biopsies are required to determine whether abnormalities are spatially heterogeneous or truly homogeneous (as in NSIP), a crucial discriminatory diagnostic feature.


Based upon recent reports of a high prevalence of pulmonary hypertension in IPF, routine echocardiography is warranted at presentation and in patients subsequently developing disproportionate hypoxia or a selective serial reduction in D LCO. In some patients with IPF the development of pulmonary hypertension is a feature of endstage disease, but in other cases early pulmonary hypertension occurs, not associated with major functional impairment due to interstitial lung disease.


Once suspected in the symptomatic patient, IPF is usually easy to detect using lung function tests and chest radiography, but in early disease high-resolution CT may be required to confirm or exclude interstitial lung disease. However (as discussed earlier), clinical, chest radiographic and physiological features are highly nonspecific in discriminating between individual idiopathic interstitial pneumonias, and high-resolution CT plays a crucial role in this regard. High-resolution CT appearances are diagnostic in an appropriate clinical setting in most patients with IPF, hence it is seldom necessary to confirm the diagnosis at surgical biopsy, especially when a typical course of relentless progression is already apparent. However, in a few patients diagnostic imprecision leads to major prognostic uncertainties and inaccurate management and a diagnostic biopsy is warranted. Thus, biopsy should not be performed by protocol in all cases but should be reserved for situations in which it appears realistic that clinician perceptions of best management, including treatment and the approach to monitoring, might change significantly with additional information.

In less typical cases the findings at bronchoalveolar lavage may play an important ancillary role in excluding alternative disorders such as extrinsic allergic alveolitis and respiratory bronchiolitis with associated interstitial lung disease (characterized by a striking lymphocytosis and a marked increase in pigmented macrophages respectively).

It should be stressed that the distinction between IPF and fibrotic NSIP , based on clinical and high-resolution CT features, poses particular difficulty. Even when high-resolution CT appearances are considered typical for NSIP, there is a significant likelihood that a surgical biopsy will disclose a pattern of UIP, indicative of a worse outcome. In difficult cases, it is essential to review the diagnosis in a multidisciplinary meeting, with the reconciliation of clinical and radiological features, in order to confirm that a diagnostic surgical biopsy is truly required. This decision is often difficult when IPF is likely, because of patient age (typically advanced), disease severity, and the presence of comorbidity, especially cardiovascular disease. The threshold for performing a biopsy is increased in patients aged over 70 years and when D LCO levels are less than 35% of predicted, both factors being associated with a significant increase in morbidity and very occasional fatalities following biopsy.

It is also important that histological findings are no longer viewed as a diagnostic ‘gold standard’ in interstitial lung disease, although usually more diagnostically influential than clinical and high-resolution CT features when the diagnosis is uncertain. A multidisciplinary diagnosis, made by negotiation between clinicians, radiologists, and pathologists, is now considered optimal. A histological pattern other than UIP is considered to exclude IPF, with one important caveat: ‘sampling error’ (i.e. a biopsy taken from a nonrepresentative site) should be kept in mind when high-resolution CT findings and the subsequent clinical course are strongly suggestive of IPF. Conversely, when UIP is disclosed at biopsy, the final consensus diagnosis sometimes differs from the histological diagnosis. This applies especially to patients with clinical evidence of extrinsic allergic alveolitis or a rheumatological disorder.


Accurate diagnosis is central to prognostic evaluation. The 5-year survival is approximately 10 to 15% in IPF, as compared to over 60% in fibrotic NSIP, and over 90% in patients with predominantly inflammatory idiopathic interstitial pneumonias. In established IPF a number of adverse prognostic factors have been identified (summarized in Table 1 below).

Table 1  Features associated with a worse outcome in idiopathic pulmonary fibrosis, with evidence graded as possible but uncertain (±), definite and moderately useful in routine practice (+), or definite and highly predictive (++)
Features associated with a worse outcome Grade of evidence
Increasing age ++
Male gender ±
Former or current smoking ±
High profusion of fibroblastic foci at biopsy +
Prominent honeycombing on high-resolution CT ++
Presence of pulmonary hypertension ++
Moderate impairment of lung function +
Resting hypoxia ++
Major desaturation on maximal exercise testing +
Major desaturation during a 6-min walk test ++
Increasing dyspnoea +
Serial decline in FVC or D LCO ++

Increasing age has consistently been an adverse prognostic determinant, although it is not clear whether disease is on average more progressive in older people, or—as seems more likely—comorbidity (cardiac disease and malignancy) is largely responsible for an adverse outcome. Disease severity at presentation is a crucial consideration. Increased mortality is associated with severe functional impairment, with D LCO levels providing the most accurate guidance to likely outcome amongst lung function tests performed at rest. A composite physiological index, containing D LCO, FVC, and FEV1 levels, has recently been shown to predict survival more accurately than any single lung function test in isolation. Severe resting hypoxia is indicative of a very poor outcome.

Maximal exercise testing is advocated as a superior prognostic determinant by some authorities but, in reality, there are no convincing data establishing that maximal exercise data are superior to D LCO levels in this regard. However, desaturation below 88% during a 6-min walk test has consistently identified IPF patients with a much worse outcome in several series. It is not yet clear whether desaturation during exercise is primarily linked to incipient pulmonary hypertension, but the presence of moderate to severe pulmonary hypertension is indicative of a very poor outcome.

Mortality and a more progressive course have both been linked to a higher profusion of fibroblastic foci at surgical biopsy in IPF, but this finding has yet to be applied to routine practice. High-resolution CT features have also been linked to outcome, with prominent honeycombing associated with a high short-term mortality, although this finding may partially reflect an association between severe honeycombing and extensive disease. Patients with biopsy-proven IPF and high-resolution CT appearances suggestive of NSIP have a better treated outcome than patients with high-resolution CT appearances typical of IPF.

Smoking status may also be important. The provocative observation that current smokers have a higher survival than former smokers or non-smokers is likely to represent merely a “healthy smoker effect”. However, recent analyses have suggested that lifelong non-smokers have a better survival in IPF than current and former smokers.

Serial observations are more prognostically accurate than observations made at a single point in time in patients with IPF. Changes in FVC over time have consistently predicted mortality more reliably than baseline data, and serial D LCO trends have been similarly predictive in some but not all reports. The distinction between stability and significant decline at 12 months is particularly useful. Once this information is known the histological diagnosis provides no additional prognostic information in mixed patient populations with UIP or fibrotic NSIP.

Routine monitoring

Lung function tests have traditionally been used to identify treatment responsiveness (in inflammatory disorders) and deterioration (in IPF and other fibrotic disorders). However, measurement variation is a major limitation which requires the use of thresholds for ‘significant change’. A 10% change from baseline FVC levels, or a 15% change from baseline D LCO levels, is required to identify definite regression or progression of disease: the greater measurement variation in D LCO may explain the fact that serial FVC trends are more predictive of longer-term outcome than serial D LCO trends. However, the interpretation of serial lung function testing must be modified in some contexts. Concurrent emphysema often has a major confounding effect, with spurious preservation of FVC levels but a disproportionate reduction in D LCO (reduced in both disorders), in which case a selective serial decline in D LCO levels may be seen with no change in FVC despite significant progression of disease. A selective reduction in D LCO may also be indicative of incipient pulmonary hypertension. Thus, serial lung function trends must be integrated with clinical, high-resolution CT and—when indicated—echocardiographic information.

A marginal reduction in lung function indices (a 5–10% change in FVC levels, a 10–15% change in D LCO levels) commonly causes difficulties for clinicians. These changes may indicate true disease progression in some patients, but lie within the measurement variation of lung function tests. Symptomatic change is sometimes a useful guide in this difficult scenario, but is sometimes misleading. Exertional dyspnoea may increase because of disease progression, loss of fitness, comorbidity, or weight gain and myopathy due to corticosteroid therapy. Serial high-resolution CT is sometimes informative, with clear evidence of disease progression in the context of marginal lung function decline. However, serial high-resolution CT should be reserved for situations in which the demonstration of disease progression is likely to influence management: it is difficult to assign significance to minor change on high-resolution CT in the absence of lung function deterioration.

Detailed lung function tests are often impracticable in advanced disease with increasing hypoxia, when serial tests tend to be less informative than observations of changes in oxygen saturation (in the steep component of the oxygen dissociation curve).

In IPF the intensity of monitoring is critically dependent upon the therapeutic goal. Regular monitoring at 3- to 4-monthly intervals is especially important in patients receiving treatment, especially novel therapies, and when referral for lung transplantation is contemplated. In other cases, in which no change in therapy is contemplated, less frequent monitoring may be appropriate. However, the importance of best supportive care, including the correct use of oxygen in advanced disease, justifies continued monitoring in the long term.


Treatment in IPF can be broadly subdivided into treatment of definite IPF and treatment of patients in whom IPF is likely but fibrotic NSIP is a realistic differential diagnosis. This distinction is important because in definite IPF it may be possible to slow decline in some cases, but long-term stability is very seldom observed. By contrast, prevention of disease progression is a realistic therapeutic aim in fibrotic NSIP, justifying vigorous treatment.

Treatment decisions should be made before the development of limiting dyspnoea, which is indicative of the loss of pulmonary reserve and—usually—a reduction in D LCO to less than 50% predicted. The outcome is better when disease is less severe and when there is less honeycombing, considerations that are especially important when IPF and fibrotic NSIP are both possible, the outcome being better in NSIP except when disease is severe (at which point mortality is high and differs little between the two disorders). To delay therapy until disease is advanced, in the belief that IPF is the likelier diagnosis, is to risk the loss of an important window of opportunity in fibrotic NSIP.

Loss of lung function due to fibrosis is irreversible and thus improvement in lung function is seldom seen with treatment in IPF. It is essential to focus on the prevention of disease progression as a valid primary aim, underlining the importance of initiating therapy before severe lung function impairment has developed. Too often stability without improvement is incorrectly taken to indicate failure of therapy, both by the patient and by medical colleagues not used to treating fibrotic lung disease.

Definite IPF

In definite IPF there is little evidence that traditional treatment regimens have a major impact on outcome. Corticosteroid therapy, often in association with other immunosuppressive treatment, has been widely used, although not based on prospective, placebo-controlled, randomized trials. In the absence of a definitive evidence base, the ATS/ERS consensus committee made a weak recommendation in 2000, supporting the use of a combination of low-dose prednisolone (such as 10 mg daily) and azathioprine (2.5 mg/kg per day up to a maximum dose of 150 mg/day). During the first month a test dose of azathioprine (50 mg/day) is usual, with weekly full blood count monitoring to avoid enhanced bone-marrow toxicity due to the rare methyltransferase deficiency. A change to full dosage is made at four weeks, with full blood counts and liver function tests performed 6- to 8-weekly thereafter. Other traditional IPF treatments, including high-dose prednisolone and cyclophosphamide, have been largely discredited due to unacceptable toxicity without a worthwhile therapeutic benefit. Colchicine and ciclosporin have also been used to treat IPF, without evidence of efficacy.

More recently, attention has turned to number of antifibrotic agents, many of which are currently under study. Pirfenidone, a pleiotrophic agent with antifibrotic, anti-inflammatory and antioxidant activity, appears promising, based on early evidence of a reduction in the rate of decline of FVC levels, but is not yet available for routine use and requires further study. IPF patients receiving antioxidant therapy, consisting of N-acetylcysteine (600 mg three times a day), given in combination with low-dose prednisolone and azathioprine (at doses detailed above), were recently shown to have a significant reduction in decline of FVC and D LCO over 12 months, compared with patients receiving prednisolone and azathioprine alone. It is not clear whether this benefit was due to antioxidant therapy in isolation or represented synergism between the three agents. Pending further data, the use of antioxidant and low-dose steroid therapy, with or without an immunosuppressive agent, appears reasonable and is the therapeutic approach preferred by the authors. However, the lack of definitive evidence should be freely acknowledged, the patient should be encouraged to take part in decision making, and a firm decision by the patient to decline therapy in definite IPF should be supported (although most patients prefer a trial of treatment).

IPF and fibrotic NSIP both possible

There is often diagnostic uncertainty in routine practice in nonbiopsied patients. A realistic differential diagnosis of NSIP is important, even when IPF is more probable, because long-term stability has been observed in over 50% of NSIP cases with treatment (steroid therapy, alone or in combination with an immunosuppressive agent). The routine use of low-dose prednisolone and azathioprine at doses suggested for IPF (detailed above) can be justified when a diagnosis of fibrotic NSIP is possible. However, progression of disease despite treatment in the next 6 to 12 months is associated with an equally poor outcome, whether the histological diagnosis is NSIP or UIP. Thus, if significant deterioration occurs despite treatment, a therapeutic approach as for IPF becomes appropriate, with acceptance of a lower likelihood of arresting further decline, and a lower threshold for reducing treatment for reasons of toxicity.

Acute exacerbations of IPF

In 10 to 15% of patients there is an accelerated deterioration occurring over several weeks and often leading rapidly to a fatal outcome. The pathogenesis of this phenomenon is not well understood and it is important to exclude supervening infection, heart failure and thromboembolism. If an acute exacerbation of IPF is diagnosed, intravenous corticosteroids (e.g. 1 g/day methylprednisolone for 3 days) together with intravenous cyclophosphamide (600 mg/m2 as a single dose, repeated at roughly 2-week intervals if blood counts are satisfactory) are often used, although no controlled treatment data exist. Noninvasive ventilation is sometimes useful, but mechanical ventilation should be avoided because of its uniformly poor outcome.


Single lung transplantation is now the preferred procedure. As in other endstage lung diseases a 3-year survival rate of over 50% can be achieved, but a worse outcome is seen in severely deconditioned patients and those over the age of 65. The rapidly progressive nature of IPF, compared to other chronic lung diseases, demands the early referral of suitable cases to a transplantation centre, ideally before D LCO levels fall below 30% of predicted normal.

Supportive therapy

Supportive therapy is central to the management of advanced disease. Supplemental oxygen can be provided in the home through oxygen concentrators, and in some patients ambulatory oxygen (in small oxygen cylinders) may be beneficial in improving exercise tolerance. The prompt treatment of complications, including infection and heart failure (sometimes triggered by hypoxia), is also important. In terminal disease small doses of opiates alleviate the distressing severe dyspnoea associated with striking reductions in lung compliance.

It is difficult for patients and family members to come to terms with the chronic, relentlessly progressive nature of IPF. The input of medical and nonmedical health care professionals is indispensable to optimal supportive management: social workers, physiotherapists, and occupational therapists all have important roles to play. Rehabilitation programmes may benefit some patients, although less likely to be useful in preterminal disease.

Further reading  


ATS/ERS International Consensus Statement. (2000). Idiopathic pulmonary fibrosis: diagnosis and treatment. International consensus statement. Am J Respir Crit Care Med, 161, 646–64.
Azuma A, et al. (2005). Double-blind, placebo-controlled trial of pirfenidone in patients with idiopathic pulmonary fibrosis. Am J Respir Crit Care Med, 171, 1040–7.
Carrington CB, Gaensler EA, Coutu RE (1978). Natural history and treated course of usual and desquamative interstitial pneumonia. N Engl J Med, 298, 801–9. [Web of Science] [Medline] 
Collard HR, et al. (2003). Changes in clinical and physiologic variables predict survival in idiopathic pulmonary fibrosis. Am J Respir Crit Care Med, 168, 538–42.[Abstract/Full Text]
Demedts M, et al. (2005). High-dose acetylcysteine in idiopathic pulmonary fibrosis. N Engl J Med, 353, 2229–42.
Flaherty KR, et al. (2003). Radiological versus histological diagnosis in UIP and NSIP: survival implications. Thorax, 58, 143–8.[Abstract/Full Text]
Flaherty KR, et al. (2003). Prognostic implications of physiologic and radiographic changes in idiopathic interstitial pneumonia. Am J Respir Crit Care Med, 168, 543–8.
Flaherty KR, et al. (2004). Idiopathic interstitial pneumonia: what is the effect of a multidisciplinary approach to diagnosis? Am J Respir Crit Care Med, 170, 904–10.
Gay SE, et al. (1998). Idiopathic pulmonary fibrosis: predicting response to therapy and survival. Am J Respir Crit Care Med, 157, 1063–72.
Hunninghake GW, et al. (2001). Utility of a lung biopsy for the diagnosis of idiopathic pulmonary fibrosis. Am J Respir Crit Care Med, 164, 193–6.
Katzenstein AL, Myers JL (1998). Idiopathic pulmonary fibrosis: clinical relevance of pathologic classification (review). Am J Respir Crit Care Med, 157, 1301–15.
King TE, Tooze JA, Schwarz MI, Brown KR, Cherniack RM. Predicting survival in idiopathic pulmonary fibrosis: scoring system and survival model. Am J Respir Crit Care Med 2001, 164, 1171–81.
King TE, Jr., et al. (2001). Idiopathic pulmonary fibrosis: relationship between histopathologic features and mortality. Am J Respir Crit Care Med, 164, 1025–32.
Lama VN, et al. (2003). Prognostic value of desaturation during a six-minute walk test in idiopathic interstitial pneumonia. Am J Respir Crit Care Med, 168, 1084–90.
Latsi PI, et al. (2003). Fibrotic idiopathic interstitial pneumonia: the prognostic value of longitudinal functional trends. Am J Respir Crit Care Med, 168, 531–7.
Mogulkoc M, et al. (2001). Pulmonary function in idiopathic pulmonary fibrosis and referral for lung transplantation. Am J Respir Crit Care Med, 2001, 164, 103–8.
Raghu G, et al. (2006). Incidence and prevalence of idiopathic pulmonary fibrosis. Am J Respir Crit Care Med, 174, 810–16.
Steele MP, et al. (2005). Clinical and pathologic features of familial interstitial pneumonia. Am J Respir Crit Care Med, 172, 1146–52.
Selman M, King TE, Pardo A (2001). Idiopathic pulmonary fibrosis: prevailing and evolving hypotheses about its pathogenesis and implications for therapy (review). Ann Intern Med, 134, 136–51.
Wells AU, et al. (2003). Idiopathic pulmonary fibrosis: a composite physiologic index derived from disease extent observed on computed tomography. Am J Respir Crit Care Med, 167, 962–9.