Idiopathic Pulmonary Fibrosis

Idiopathic pulmonary fibrosis (IPF) is the prototypic chronic, progressive fibrosing interstitial lung disease. In 2011, new ATS/ERS guidelines on the diagnosis, prognosis, and therapy of IPF were published (Raghu et al.). Its prevalence in the United States approaches 100,000 cases with an incidence of 15,000 to 30,000 cases per year. The disease generally affects patients in their sixth decade of life or later, and in patients older than 75 the prevalence increases to 175 per 100,000. There is a slight male predominance.

The cause of IPF is unknown, although risk factors include tobacco use, environmental and occupational exposures, infectious agents such as Epstein-Barr virus (EBV) and hepatitis C virus (HCV), and gastroesophageal reflux disease (GERD). Symptoms are frequently insidious and include cough and shortness of breath, especially with exertion. Symptomatic progression typically occurs over years, although abrupt and otherwise unexplained declines in respiratory symptoms, physiology, and imaging studies (acute exacerbation of IPF) can occur. On examination, digital clubbing may be present. Chest auscultation reveals mid- to end-inspiratory crackles, most pronounced at the bases. During late stages, patients may show evidence of right-sided heart failure caused by the development of pulmonary artery hypertension.

The diagnostic utility of laboratory testing is limited because no biologic fluid-based biomarkers are currently useful in making a definitive diagnosis. Serum Krebs von den Lugen-6 (KL-6), serum surfactant A and D, serum CCL18, serum brain natriuretic peptide (BNP), plasma and bronchoalveolar lavage (BAL) fluid, and matrix metalloproteinases (MMPs) 1 and 7 are all potential biomarkers. Ongoing studies are evaluating the role of these and other potential biomarkers in diagnosis and disease progression. Laboratory testing is useful in defining the clinical context and excluding alternatives to explain the ILD. In particular, serologic testing for connective tissue diseases (CTDs) is recommended in all patients because the distinction between CTD and idiopathic ILD may have prognostic significance. Pulmonary functions tests (PFTs) reveal a restrictive ventilatory defect with a decreased diffusion capacity. BAL findings are nonspecific, and the procedure is not normally required in the diagnostic evaluation. If performed, a mild neutrophilia and mild (<20%) eosinophilia are frequently found. Lymphocytosis, especially greater than 40%, favors an alternate diagnosis.

When a surgical lung biopsy is performed, the histopathologic pattern seen is termed usual interstitial pneumonia (UIP). A definite UIP histopathologic pattern diagnosis requires all four of the following features:

The chest imaging or radiologic pattern seen on high-resolution computed tomography (HRCT) scanning in about 50% of patients is also termed UIP. In the other 50% of patients, less well-characterized patterns are seen. When present, a UIP radiographic pattern has a positive predictive value of 90% to 100% for the presence of a UIP histopathologic pattern. Therefore, when this pattern is present, a surgical lung biopsy is rarely if ever needed. A definite HRCT radiologic pattern diagnosis of UIP can be made if all the following characteristics are present:

In the clinical context of an idiopathic ILD, the identification of a UIP chest imaging pattern confirms the diagnosis of IPF. In patients with discordant findings in their clinicoradiologic appearances, or if they have atypical features on HRCT, a surgical biopsy is recommended in those who need a definitive diagnosis and are suitable for surgery. Transbronchial biopsies are too small to be diagnostic; therefore when a definitive diagnosis is necessary, a surgical biopsy should be obtained, preferably using video-assisted thoracoscopy, with biopsies from more than one lobe. A UIP histopathologic pattern confirms the diagnosis.

In summary, the new guidelines allow the clinician to establish the diagnosis of IPF based on the following criteria:

Because the diagnosis of IPF requires combining clinical, radiologic, and often pathologic information, multidisciplinary discussions with chest imagers and pathologists are strongly recommended, especially since a definitive diagnosis cannot be made based on radiologic and histologic findings alone. Figure 47-2 outlines an algorithm for the diagnosis of IPF.

Figure 47-2 Algorithm for diagnosis of idiopathic pulmonary fibrosis (IPF). The initial evaluation of a patient with suspected IPF (typically a patient with cough and chronic progressing dyspnea who has radiographic changes consistent with ILD) rules out other causes of interstitial lung disease (ILD). If no other causes are identified, high-resolution computed tomography (HRCT) should be done. If HRCT shows usual interstitial pneumonia (UIP), this is diagnostic. If HRCT is nondiagnostic, surgical biopsy is obtained, followed by multidisciplinary discussion (MDD) to make the diagnosis.

(Modified from Raghu G et al: Am J Respir Crit Care Med 183:788–824, 2011.)

Patients have complications directly and indirectly related to their IPF. Although the natural history of IPF is generally slow symptomatic and physiologic progression, otherwise unexplained rapid deterioration can occur. Acute exacerbations of IPF are defined as the presence of an acute respiratory decompensation of unknown cause and occur in 5% to 10% of patients yearly. Patients tend to present with a viral prodrome and worsening baseline hypoxemia. Invasive mechanical ventilation is often required to maintain oxygenation, but its use is not generally recommended given the poor prognosis associated with acute exacerbations, although it may be appropriate in a minority of patients. Workup should include exclusion of pulmonary embolism, infection, congestive heart failure, volume overload, pneumothorax, and other causes of lung injury. Imaging studies may show new onset of bilateral diffuse ground-glass opacity or dense consolidation. Lung biopsies show a diffuse alveolar damage (DAD) pattern. Although strong evidence is lacking, treatment with high-dose steroids is provided to most patients, in addition to broad-spectrum antibiotics. The prognosis is poor, with mortality rates of up to 70% or more during the following year.

Pulmonary hypertension is a recognized complication affecting about 10% of IPF patients at diagnosis. Its presence correlates with increased early mortality. With progression of disease, a majority of patients may ultimately develop some clinical evidence of pulmonary hypertension. In patients with worsening symptoms without significant changes in lung imaging studies or in lung volumes or spirometry, pulmonary hypertension should be considered.

Other comorbidities associated with IPF include an increased risk of venous thromboembolic disease, lung cancer, emphysema, GERD, obesity, obstructive sleep apnea, and cardiovascular disease.

A study quantified the mortality rate associated with IPF in 2003 in the United States as 61.2 deaths per 1 million in men and 54.5 per 1 million in women, with 60% of deaths attributed to progressive lung disease. Increased rates are seen in older patients and in the winter months. Figure 47-3 compares mortality rates (per 100,000 population) from various countries. Although the median survival of symptomatic patients is generally accepted as 3 to 5 years, the survival of individual patients varies widely.

Figure 47-3 Mortality estimates for patients with idiopathic pulmonary fibrosis (IPF) in seven countries (1979–1991).

(Modified from Coultas DB et al: Am J Respir Crit Care Med 150:967–972, 1994.)

Figure 47-4 shows the natural history of IPF. Relatively stable lung function and slow (most common) or rapid disease progression can all occur. A patient’s course is unpredictable, and risk factors for decompensation are poorly understood. No clear consensus exists on a valid staging system for IPF. However, the specific characteristics used at diagnosis to identity patients with risk factors for early mortality include severity of dyspnea, lower forced vital capacity (FVC) or diffusion capacity (DLCO), significant desaturation on 6-minute walking distance (6MWD) testing, and the extent of fibrosis on HRCT. Other suggested risk factors include a shorter 6MWD, presence of significant emphysema, and pulmonary hypertension. Important longitudinal factors associated with an increased risk of early mortality in IPF patients are worsening dyspnea, decline in FVC, decline in DLCO, and increasing extent of disease on HRCT.

Figure 47-4 Most patients with idiopathic pulmonary fibrosis display a slowly progressive natural history. In a minority of patients, IPF can be stable or rapidly progressive. Some slowly progressive or stable patients may also have acute decompensation (lightning bolt), from a known infection, other lung insult, or an unknown cause, generally with high mortality or resulting in worsened lung function.

(Modified from Raghu G et al: Am J Respir Crit Care Med 183:788–824, 2011.)

Nonspecific Interstitial Pneumonia

Historically, nonspecific interstitial pneumonia (NSIP) has been the catchall phrase for ILD that does not fit into other categories. In the mid-1990s, NSIP was defined as a specific histopathologic pattern that did not match histopathologic criteria for other idiopathic ILDs. Since then, NSIP has expanded from simply describing a histopathologic pattern to a specific form of an otherwise idiopathic ILD. Because the histopathologic NSIP pattern is not unique, a concerted effort should identify a potentially causative exposure or associated medical condition. When no alternative explanation is identified, the diagnosis of “idiopathic NSIP” can be made.

The epidemiology of NSIP is unknown. Significantly less common than IPF, estimates of the prevalence of idiopathic NSIP range from 1 to 9 per 100,000 population. NSIP presents in a younger population than IPF, with patients in their 40s to 50s and with a female predominance. Symptoms are nonspecific and include cough and dyspnea with a presentation that ranges from subacute (6-12 months) to chronic (up to 3 years); median duration is 18 to 31 months. History and physical examination should be used to evaluate for features suggestive of an alternative explanation for the presence of ILD. For example, the presence of Raynaud’s phenomenon, arthritis, and myositis suggests an underlying CTD. Pulmonary examination typically reveals basilar crackles.

Laboratory testing offers limited information and cannot make a definitive diagnosis. However, testing helps define the clinical context and exclude alternative explanations for ILD. Similar to IPF, serologic testing for a CTD is recommended in all patients because the distinction between idiopathic NSIP and the histopathologic pattern of NSIP associated with a CTD may have prognostic significance. PFTs reveal a restrictive ventilatory defect with a decreased DLCO. A minority of patients may present with a mild airflow obstruction. In contrast to IPF, BAL fluid will show lymphocyte predominance in 50% of suspected NSIP patients, generally with a minor proportion of neutrophils and/or eosinophils.

The histopathologic pattern of NSIP is a homogeneously abnormal parenchyma with diffuse interstitial involvement. It can be subdivided based on the amounts of lung fibrosis and inflammation, with cellular NSIP consisting almost exclusively of cellular inflammation and fibrotic NSIP showing some component of acellular fibrosis. Cellular NSIP has mild to moderate chronic interstitial inflammation with associated type II pneumocyte hyperplasia and few areas of lymphoid aggregates. Fibrotic NSIP results in uniform involvement with dense or loose interstitial fibrosis and a lack of honeycombing or fibroblast foci.

On imaging studies, cellular NSIP has diffuse areas of ground-glass infiltrate, whereas fibrotic NSIP is associated with reticular changes, volume loss, and traction bronchiectasis. The findings are typically symmetric, lower lung predominant, and often the subpleural zone is relatively free of fibrosis. Honeycombing and consolidation are rare findings. The sensitivity and specificity of imaging studies in NSIP are not to the level of being diagnostic. Therefore, when the diagnosis of NSIP is suspected, a surgical lung biopsy should be considered.

Treatment and Prognosis

Prospective, controlled therapeutic trials are lacking in the treatment of patients with NSIP, and expert opinion provides the basis for most recommended regimens. Corticosteroids are typically initiated in these patients, with an anticipated transition to steroid-sparing agents such as azathioprine, mycophenolate mofetil, or cyclophosphamide. The prognosis of NSIP is variable, with a minority of patients with progressively fatal disease and others with nearly complete recovery. Patients with a predominantly fibrotic NSIP have a worse prognosis than those with predominantly cellular NSIP. The majority of patients will have relatively stable disease or some mild improvement with therapy. Overall, mortality is better than with UIP. Figure 47-5 compares the survival curves of patients with NSIP, UIP, and other idiopathic ILD.

Figure 47-5 Survival curves based on histopathologic subgroups comparing usual interstitial pneumonia (UIP), idiopathic pneumonia (NSIP), and other idiopathic interstitial lung diseases. Patients with UIP pattern have worse prognosis than those with NSIP or other ILDs.

(Modified from Bjoraker JA et al: Am J Respir Crit Care Med 157:199–203, 1998.)

Respiratory Bronchiolitis Interstitial Lung Disease

The histopathologic pattern respiratory bronchiolitis (RB) is a common finding in current or former heavy smokers. However, the overwhelming majority of patients appear to have no related clinical sequelae and it exists as an incidental finding when a patient has lung pathology obtained for other reasons. In contrast, respiratory bronchiolitis interstitial lung disease (RBILD) occurs when a patient with clinical and physiologic evidence of interstitial lung disease has RB as the sole histopathologic finding to account for their ILD. Given the overlapping clinical context, imaging, and pathologic findings, RBILD appears to be on a spectrum of disease that includes desquamative interstitial pneumonia (DIP).

Patients with RBILD present in their 30s and 40s, with a male predominance of almost 2 : 1. Patients are almost invariably heavy smokers of at least 30 pack-years. They generally present with dyspnea and cough. Pulmonary findings are similar to DIP, with a mild to moderate reduction in DLCO and a mixed restriction and obstruction. Up to 10% of patients may present with normal physiology, and another 10% show significant bronchodilator response. On chest imaging, as many as 20% of patients may have normal plain chest radiographs. HRCT shows significant airway wall thickening, reticular and centrilobular abnormalities, and small patches of ground-glass opacity. The findings are bilateral and can be found in all five lung lobes and tend to be less extensive and involve less of the lung parenchyma than in DIP. BAL fluid typically shows an excess number of pigmented macrophages with black, golden, or brown inclusions, similar to that seen in otherwise healthy smokers. There may also be a modest neutrophilia.

The diagnosis of RBILD is typically made from the clinical context of a young, heavy smoker and the chest imaging pattern seen on HRCT. If definitive diagnosis is needed, surgical lung biopsy is indicated because transbronchial biopsy and BAL findings cannot provide this level of confidence. The histopathologic pattern is bronchiolocentric, involving first-order and second-order respiratory bronchioles with the peribronchiolar alveoli filled with tightly packed pigmented macrophages. These cells stain positive for periodic acid–Schiff (PAS). The lumen may also be filled with mucus (mucostasis). There is minimal fibrosis, limited to the peribronchial areas.

Desquamative Interstitial Pneumonia

Desquamative interstitial pneumonia is thought to be on a continuum of disease with RBILD, and is seen in a similar clinical context, heavy smokers in 60%-87% of cases. In contrast to RBILD, 20% of cases occur in the setting of CTD, viral infections or drug/toxin exposure. The DIP patient population is younger than those with IPF, typically in their 30s to 50s, with a 2 : 1 male predominance.

The most common pulmonary physiologic finding is a reduced DLCO. The patient may also have mild restriction or mild mixed restrictive-obstructive patterns. HRCT typically shows ground-glass opacities in the lung periphery and bases without significant fibrosis or honeycombing. Other findings include traction bronchiectasis, centrilobular nodules, and cysts, which are typically less than 2 cm in diameter.

Surgical biopsy is required for a definitive diagnosis because BAL fluid and transbronchial biopsy show nonspecific findings. The histopathologic pattern reveals homogeneous and diffuse interstitial thickening. Alveoli are filled with iron-laden macrophages, resulting in the ground-glass pattern. Type II cell hyperplasia and diffuse alveolar septal thickening also are seen. In contrast to RBILD, the patient with DIP has more interstitial fibrosis, lymphoid follicles, giant cells, and eosinophils, although the patterns can show considerable overlap. Unlike UIP, fibroblast foci and honeycombing are not present in DIP. Given the smoking history, some emphysematous changes may also be seen.

Acute Interstitial Pneumonia

Acute interstitial pneumonia (AIP) has been known by a variety of names over the years. Hamman and Rich described the first cases in 1935. More recently it was considered an idiopathic form of acute respiratory distress syndrome (ARDS) because their presentation and clinical characteristics are indistinguishable, but they are differentiated by the absence of a known cause or association in AIP. Symptoms include a rapid onset of dyspnea over days to less than 3 weeks, typically following a viral prodrome. Patients routinely develop significant hypoxia and respiratory failure, requiring assisted ventilation. The epidemiology includes a wide age range, with a mean age of 50 years, and affects both genders equally. There is no known association with smoking.

Chest imaging in AIP patients typically shows bilateral ground-glass infiltrates sparing the costophrenic angles and otherwise patchy distribution with associated air bronchograms. Pleural effusions are rare. HRCT in the early phase of disease shows diffuse bilateral ground-glass opacities. In the later stages, bilateral consolidative opacities, especially at the lung bases and in dependent regions are found, as well as bronchial dilation and architectural distortion with honeycombing and cysts.

The diagnosis of AIP is a diagnosis of exclusion. BAL fluid is primarily used to exclude overwhelming infection and reveals increased total cell numbers with red blood cells, neutrophils, and lymphocytes. The histopathologic pattern is diffuse alveolar damage, with hyaline membranes, diffuse septal fibrosis, and type II pneumocyte hyperplasia.

Lymphoid Interstitial Pneumonia

Lymphoid interstitial pneumonia (LIP) rarely presents as an idiopathic lung disease and is most often seen in the setting of autoimmune disease, genetic abnormalities, infection (especially HIV, EBV, and HTLV-1), or dysproteinemia (e.g., polyclonal and monoclonal gammopathy). When seen as an idiopathic ILD, patients with LIP typically present with cough and progressive shortness of breath over months and rarely up to years. Symptoms can include fever, weight loss, and pleurisy.

Physical examination is consistent with crackles and less frequently digital clubbing (~10% of patients). Pulmonary physiology reveals a restrictive pattern with reduced DLCO. Chest imaging findings can vary and involve basilar reticular opacities, ground-glass opacities, or centrilobular nodularity. Suggestive perivascular cysts are seen in many patients. BAL fluid typically shows a lymphocytosis, but this is not diagnostic. For definitive diagnosis, surgical biopsy is recommended and shows confluent cellular infiltration of the alveolar septae (and less frequently peribronchial and perivascular septae) with lymphocytes, plasma cells, and histiocytes. Germinal centers, occasional non-necrotizing granulomas, and associated multinucleated giant cells are also seen. In later stages of LIP, fibrosis and even honeycombing can be seen. Genetic studies are important to rule out malignancy and monoclonality.

Controversies and Pitfalls

Establishing the correct diagnosis remains the crucial step in the assessment and management of patients with interstitial lung disease, because the diagnosis is the most important determinant of prognosis and appropriate treatment plan. The multidisciplinary approach involving the clinician, chest radiologist, and pulmonary pathologist is critical in piecing together the puzzle. However, this requirement for active multidisciplinary communication is often the rate-limiting step, and the clinician must expend considerable effort to make it work. Beyond the diagnosis, projecting the future disease activity remains an uncertain science, and multiple ongoing studies are underway in an attempt to identify biomarkers predictive of future disease behavior. A number of genetic risk factors for the development of pulmonary fibrosis have been identified, although the utility of these findings in clinical practice is uncertain at present.

Because no therapy used in the treatment of idiopathic pulmonary fibrosis has provided a survival, quality of life, or functional status benefit, the limitations of current treatment options leave plenty of room for controversy. As most interventional treatment trials use purported surrogate end points, the results of these trials will be difficult to interpret and will require the clinician to engage each patient fully to make informed decisions regarding treatment.