Introduction
Sarcoidosis is a multisystem granulomatous disease of unknown etiology. The lung is the most common organ affected with sarcoidosis. Pulmonary involvement can range from an asymptomatic state, where a radiographic abnormality is detected on a screening chest radiograph, to end-stage fibrocystic disease that may result in respiratory failure and death. The diagnosis is sometimes straight-forward when the chest radiograph shows typical findings of pulmonary sarcoidosis, but many times the diagnosis is unsuspected and the patient may not be correctly diagnosed for years after the onset of pulmonary symptoms. Treatment of pulmonary sarcoidosis has not been standardized. It is often problematic to decide which patients with pulmonary sarcoidosis should be treated with corticosteroids and how to monitor patients whether or not they have received therapy. This manuscript will outline clinical aspects of pulmonary sarcoidosis and hopefully will be a useful aid to clinicians caring for these patients. The Immunologic Basis of Pulmonary Involvement
The early sarcoid reaction is characterized by the accumulation of activated CD4 T-lymphocytes and macrophages causing an inflammatory alveolitis (1). These activated T-lymphocytes secrete interferon-gamma, IL-2 and other cytokines that lead to the formation of granulomas. The production of this set of cytokines has been described as a T-helper 1 (Th1) response that has been observed in other granulomatous diseases (2). So it is clear that macrophages and T-lymphocytes are first activated causing an inflammatory alveolitis, then cytokines are produced that result in granulomas. But what causes the initial activation of the macrophages and lymphocytes? This answer to this question remains unknown, and the cause of sarcoidosis remains a mystery. It is thought that exposure to some antigen triggers a cell-mediated immunity response mediated through antigen presenting cells and antigen-specific T-lymphocytes which eventually leads to a more nonspecific inflammatory response with granuloma formation. However no "sarcoidosis antigen" has yet been found. Natural Course
The natural course of pulmonary sarcoidosis is highly variable. Unlike most other interstitial lung diseases, remission and resolution are common so that many patients are best served by avoiding use of potentially toxic therapy. Thirty to 60 percent of patients are asymptomatic such that sarcoidosis is detected by an incidental finding on a chest radiograph (3). Abnormalities seen on the chest radiograph of patients with untreated pulmonary sarcoidosis resolve, improve or stabilize in 60 to 90 percent (4). The radiographic response is better and complete resolution occurs more frequently in patients with hilar adenopathy alone on the initial chest radiograph (radiographic Stage I) than in those with varying degrees of parenchymal infiltration (radiographic Stages II and III) (5). Remissions often occur within the first six months after diagnosis, although it may take two to five years (6). Ten to thirty percent of patients develop chronic pulmonary disease. Some of these patients have not only persistent but progressive disease resulting in severe fibrotic and cystic changes, volume loss, lung distortion, and end-stage honeycombed lung (radiographic Stage IV) (3). Fatalities ascribed to sarcoidosis occur in 1 to 6 percent of patients, and three-fourths of these deaths are related to chronic progressive pulmonary disease (7) including complications from aspergilloma (8). Because sarcoidosis is a systemic disease, the prognosis of sarcoidosis cannot be determined solely on the outcome of pulmonary involvement without considering the extrapulmonary manifestations of the disease. As mentioned above, three-fourths sarcoidosis-related deaths are the result of pulmonary involvement. Sarcoidosis of the heart and central nervous system account for most of the remaining deaths (9). The prognosis of pulmonary sarcoidosis correlates with several extrapulmonary factors. The presence of erythema nodosum, arthritis, or fever at presentation, known as Lofgren’s Syndrome, is associated with a good prognosis, while disfiguring cutaneous lesions, known as lupus pernio, splenomegaly, or bone involvement is associated with a poor prognosis (5,6). The disease has a worse prognosis in black patients (10) and patients with disease onset after age 40 (6). The Location of Pulmonary Involvement
The granulomatous inflammation of sarcoidosis can occur anywhere in the lung, but it has a propensity for certain locations. Granulomas are seen throughout both lungs but tend to occur in the upper two thirds of the lung (11). Granulomas are commonly found along the bronchovascular bundles, in subpleural locations, and in the intralobular septa (12). Granulomas are found in airways in 45 to 70 percent of patients (13, 14), and sometimes cause significant airway distortion resulting in airflow obstruction (15). Sarcoidosis commonly affects intrathoracic lymph nodes, and bilateral hilar lymphadenopathy (often with concomitant right paratracheal adenopathy) is found in 50 to 85 percent of cases (3). Although bilateral hilar adenopathy may be found with malignancy (particularly lymphoma), malignancy rarely causes bilateral hilar adenopathy in the absence of symptoms, and sarcoidosis is almost always the correct diagnosis in these cases (16). As mentioned above, 10 to 30 percent of patients with sarcoidosis develop progressive fibrocystic disease with lung distortion and honeycombing. Bronchiectasis may be seen in these cases as this lung distortion affects airways. This may lead to severe airflow obstruction (17). Sarcoidosis may involve the pulmonary vasculature and occasionally cause significant pulmonary hypertension (18). Although one third of sarcoidosis patients who undergo pleural biopsy have evidence of pleural involvement, only 1 percent have radiographic evidence of pleural disease with pleural effusion or thickening (11). Clinical Features
The symptoms of pulmonary sarcoidosis are nonspecific. Cough and dyspnea are common. Chest pain is also common and is usually substernal and pleuritic (19). Although sarcoidosis is thought of as an interstitial lung disease, as previously mentioned the disease can involve the airways and wheezing may be a predominant symptom. Systemic symptoms such as fever, night sweats and weight loss are frequent. Patients may also present with extrapulmonary symptoms of disease such as Lofgren’s syndrome (erythema nodosum, arthritis, fever with hilar adenopathy on chest radiograph), uveitis or other eye manifestions, or skin lesions. End-stage fibrocystic sarcoidosis patients may have profound dyspnea, copious sputum production if they have bronchiectasis, and hemoptysis if they have bronchiectasis or an aspergilloma. The chest exam is usually unrevealing in sarcoidosis. Unlike other interstitial lung diseases, the chest is usually silent in pulmonary sarcoidosis and crackles are rarely heard (20). The physical exam is often most useful in identifying extrapulmonary manifestations of disease such as skin lesions, eye abnormalities, hepatosplenomegaly, parotid enlargement, peripheral lymphadenopathy, and neurologic abnormalities (e.g. Bell’s Palsy). Chest radiographs are abnormal in more than 90 percent of patients with pulmonary sarcoidosis (3). The most characteristic finding is bilateral hilar adenopathy in 50 to 85 percent of cases. Concomitant right paratracheal adenopathy is common. Lymph node enlargement is less common at other nodal sites but can often be detected on chest CT. Thoracic lymphadenopathy can be identified on chest CT in most interstitial lung diseases (21); but when the thoracic lymph nodes are greater than 1.5 cm, the diagnosis is usually sarcoidosis (21). Obviously malignancy (e.g. lymphangitic spread of carcinoma) needs to be excluded in these cases. Parenchymal infiltrates are present in 25 to 50 percent of cases (3). These infiltrates are often bilateral and preferentially involve the upper lobes. These infiltrates are usually central which is consistent with the fact that the disease predominates along the bronchovascular bundles ( Figure 1). These infiltrates are usually reticular, reticulonodular, or focal alveolar opacities (3). A radiographic staging system for sarcoidosis was developed four decades ago (Table 1, Figure 2 & Figure 3) (22). When examining groups of patients of different radiographic stages, patients of higher radiographic stage have more severe pulmonary dysfunction, lower remission rates, and higher mortality (3, 6, 9, 23). However there is significant overlap between the groups such that predictions for individual patients are not reliable by radiographic stage alone. One consistent finding is that patients with stage IV chest radiographs almost never undergo spontaneous remission (3, 24). The severity of disease on chest radiograph correlates poorly with clinical or functional impairment (3). Pulmonary function may be normal in patients with significant parenchymal disease on chest radiograph (stage II or III), although it is unusual for pulmonary function to be normal in patients with fibrocystic disease (radiographic stage IV). Pulmonary function tests may show a restrictive ventilatory defect and a decreased diffusing capacity typical of an interstitial lung disease (25). However some patients with significant airway involvement may demonstrate an obstructive ventilatory defect (26), and such patients are sometimes a misdiagnosed as having asthma. Diagnosis
The diagnosis of pulmonary sarcoidosis requires a compatible clinical presentation, histologic demonstration of noncaseating granulomas and exclusion of other diseases capable of producing a similar histologic or clinical picture. Specifically, tuberculosis and fungal pulmonary disease need to be excluded, and stains and cultures for mycobacteria and fungi should routinely be obtained when performing a lung biopsy for the diagnosis of sarcoidosis. Another disease that needs to be excluded is chronic beryllium disease that can mimic pulmonary sarcoidosis radiographically and histologically (27). Patients with presumed sarcoidosis should be questioned about beryllium exposure, although patients may be unaware of exposure; a minor exposure can cause disease. A lymphocyte stimulation test for beryllium (blood test) is moderately sensitive and highly specific for chronic beryllium disease. Transbronchial lung biopsy is the recommended method to obtain adequate lung biopsy specimens for the diagnosis of pulmonary sarcoidosis. Five biopsy specimens are usually adequate to demonstrate noncaseating granulomas (28). The sensitivity of transbronchial biopsy ranges from 60 to 97 percent (3). Although the yield is usually higher in patients with parenchymal disease, the yield is over 70 percent in patients with normal lung parenchyma on chest radiograph and bilateral hilar adenopathy (stage I) (3). There are several instances where lung biopsy is not required to diagnose pulmonary sarcoidosis. The patient may have evidence of extrapulmonary sarcoidosis that may be more easily or safely biopsied. For example, a patient may have a chest radiograph consistent with sarcoidosis and a skin lesion that may be biopsied. Although there may not be histologic confirmation of pulmonary, pulmonary sarcoidosis can be assumed if the clinical presentation is compatible with sarcoidosis, a biopsy of an extrapulmonary site demonstrates noncaseating granulomas, and a chest radiograph shows diffuse infiltrates, bilateral hilar adenopathy, and/or upper lobe fibrosis (29). Another instance in which a lung biopsy may not be required for the diagnosis of pulmonary sarcoidosis is when the clinical presentation is typical of the disease. An example would be Lofgren’s Syndrome: hilar adenopathy on chest radiograph, erythema nodosum skin lesions, fever, and arthritis. Unless some other cause of this clinical syndrome can be identified (e.g. coccidiodmycosis), most clinicians would accept this clinical presentation as specific enough to make a clinical diagnosis of sarcoidosis. As mentioned previously, asymptomatic patients with bilateral hilar adenopathy almost always have sarcoidosis and a biopsy is usually not required in these patients (16, 30). An elevated angiotensin converting enzyme level is never diagnostic of sarcoidosis because it is nonspecific and elevated in several other diseases (31). Treatment
The decision to treat a medical condition depends upon several factors that include the natural history of the disease, the expected response to treatment, and the toxicity of therapy. Pulmonary sarcoidosis poses unique treatment dilemmas because, unlike most interstitial lung diseases, the disease may remit or stabilize without therapy. Since corticosteroids may cause significant toxicity, their use is not justified in asymptomatic patients or those with mild disease that may spontaneously remit. For patients with clinical findings that predict spontaneous remission (e.g. erythema nodosum), the benefits of treatment are often exceeded by the toxicity of systemic corticosteroids. Often these patients can be managed with palliative therapy: non-steroidal anti-inflammatory agents for arthralgias and fever, or bronchodilators and inhaled corticosteroids for wheezing and cough. Many patients with pulmonary sarcoidosis will have spontaneous improvement or remission within 6 months of diagnosis (32). Therefore, it is recommended that patients with mild to moderate pulmonary sarcoidosis be observed for two to six months, if possible (32,33). Patients who improve will have avoided the toxicity of corticosteroids, while patients who deteriorate over this period should be treated. Patients with pulmonary dysfunction who neither improve nor deteriorate during the observation period are often given a corticosteroid trial (32). Patients with severe pulmonary dysfunction or pulmonary symptoms causing significant impairment should be treated with corticosteroids. Patients with chronic pulmonary sarcoidosis are often treated for prolonged periods; sometimes therapy is life-long. These patients need long-term monitoring of their pulmonary status if corticosteroids are discontinued, as disease exacerbation can occur years later. Table 2 outlines general guidelines concerning corticosteroid treatment decisions for sarcoidosis. Corticosteroids have been shown to improve radiographic findings and spirometry relative to placebo for several months as initial treatment of pulmonary sarcoidosis (34,35). However, a long-term benefit of corticosteroids has not been demonstrated (five or more years after therapy) (34-37). A recent study (38) suggested that corticosteroids may promote relapse of sarcoidosis, although this study was retrospective and the untreated and corticosteroid groups were not matched. The granulomatous inflammation from sarcoidosis usually lasts from several months to years. The initial intent of anti-inflammatory therapy is suppression of the acute inflammation. Subsequently, anti-inflammatory therapy is required to avoid recrudescence of a clinically significant inflammatory response. At some point, a decision to withdraw anti-inflammatory therapy is made and the patient is monitored for relapse. Therefore, treatment of pulmonary sarcoidosis with corticosteroids can be categorized into six different phases (Figure 4): initial dosing, taper to maintenance, maintenance dosing, corticosteroid taper, monitoring off therapy, and treatment of relapse. Neither the doses nor the time period of these six phases has been standardized. Initial dosing: The initial starting prednisone dose is usually 30 to 40 mg daily (33,39), although 1 mg/kg/day has also been recommended (40). It has been suggested that higher initial dosing may be required if concomitant neurologic, myocardial, or severe ocular lesions are present or for severe hypercalcemia (39). Taper to maintenance: Symptomatic improvement usually occurs within a month, and tapering from these corticosteroid doses is usually begun within 2 to 6 weeks (33,40). Failure to improve within one month suggests either that the disease is steroid-resistant or pulmonary sarcoidosis is not responsible for the pulmonary dysfunction. It is unusual for longer trials at these high doses to be successful (33). Tapering to a maintenance dose is usually completed in one to three months (39-41), although a slower taper to maintenance over 6 months has also been suggested (42). The decision to taper corticosteroids is based on improvement in pulmonary symptoms and pulmonary function (33,40). Spirometry is adequate to assess pulmonary function in most cases. Measurement of diffusing capacity rarely adds additional information, since spirometry and diffusing capacity have a concordant response in two-thirds of cases; these tests move in opposite directions only 5 percent of the time (25). Some authors (32) have used the chest radiograph to monitor the clinical status of pulmonary sarcoidosis patients. However, pulmonary function and chest radiographic findings are insufficient indications for corticosteroid modification without compelling pulmonary symptoms (38). Since most patients with acute pulmonary sarcoidosis will improve with the initial high doses of corticosteroids, monitoring for relapse first becomes an issue during the "taper to maintenance" phase. In terms of monitoring, it is important to note that measurements of "active" granulomatous inflammation such as 1) serum angiotensin converting enzyme (SACE) 2) bronchoalveolar lavage (BAL) cell count differentials and 3) gallium radionuclide scans are not routinely used to guide treatment decisions because no data exist to support the contention that therapy based solely on the results of any of the indices of activity will alter the eventual outcome of a patient with sarcoidosis. This is because the granulomatous inflammation detected by these tests often spontaneously resolves in sarcoidosis (36,37,42). Most permanent organ dysfunction in sarcoidosis relates to the development of fibrosis, and it is unclear if this is dependent solely on the presence or degree of the granulomatous inflammation or if other additional factors are required (1,43). Maintenance dosing: The recommended maintenance corticosteroid doses is 10 to 15 mg daily (39,41), although 0.25 mg/kg/day has been suggested (40). The duration of maintenance is controversial. Some have suggested that corticosteroids can be discontinued as early as 6 months after starting therapy (33). However, others have shown that when corticosteroid therapy is continued for at least one year relapse is less common (40) and long-term pulmonary function is slightly improved (25). These benefits must be weighed against the increased risk of side-effects from prolonged therapy (25), almost all of which are transient (e.g. weight gain, gastrointestinal disturbance, acne). Alternate day dosing may reduce the prevalence of corticosteroid side-effects, but patient compliance with this dosing schedule is often poor (39). Corticosteroid taper: The maintenance dose is usually continued until the decision to taper the patient completely off corticosteroids is made, which is usually done over a 6 week to 3 month period (40). Doses are not routinely tapered more rapidly than every two weeks, since this is usually the minimal amount of time for clinical relapse to occur if anti-inflammatory therapy is inadequate. Monitoring off therapy: Pulmonary symptoms and spirometry are the most useful parameters to monitor when deciding if corticosteroid therapy should be reinstituted. Tests of active granulomatous inflammation such as SACE, BAL and gallium scans are not useful guides for therapy, relapse, or prognosis and should not be used for monitoring purposes. Although it has been recommended that patients be monitored for one year off therapy until sarcoidosis is considered arrested (44), 20 percent of relapses occur more than one year after cessation of corticosteroid therapy, and 10 percent beyond two years (38). In a retrospective review (38), chest radiographic stage and race did not correlate with the rate of relapse. Relapse was more frequent in patients who were treated with corticosteroids, although the corticosteroid group may have had more severe disease at baseline. Erythema nodosum at presentation was associated with a lower rate of relapse. Based on these data, it seems intuitive to observe patients who have been treated with corticosteroids for pulmonary sarcoidosis for at least two years. Spirometry and pulmonary symptoms should monitored. Relapse: Relapses occur in 20 to 50 percent of patients in whom corticosteroids are discontinued (39). Relapses are treated identically to initial treatment with reinstitution of high dose corticosteroids for 2 to 6 weeks followed by an identical tapering scheme (40). Some authors have recommended higher maintenance doses or prolonging the maintenance period in patients who relapse (40). Alternative anti-inflammatory medications, either used alone or as steroid-sparing agents, should be considered if daily prednisone doses of 20 mg or more are required to avoid exacerbations of pulmonary sarcoidosis. Treatment of Chronic Pulmonary Sarcoidosis
As mentioned, most deaths from sarcoidosis result from progressive pulmonary disease (45). Three-quarters of pulmonary sarcoidosis patients who require corticosteroids for 5 or more years will relapse when corticosteroids are withdrawn (46). Almost all of these relapses occur within 1 to 2 months of corticosteroid withdrawal. Over 90 percent of these patients can be maintained on 15 mg or less of daily prednisone, and 65 percent on 10 mg or less (46). In general, complications from corticosteroids are not frequent or serious in nature (weight gain is most common) (46). Corticosteroid-dependence should not be considered corticosteroid failure. The relatively successful long-term use of corticosteroids and relatively few side-effects must be kept in mind when considering alternative therapies, such as other toxic pharmacotherapy or lung transplantation. Low-dose corticosteroid dependence may be superior to alternative therapies, or to no therapy, which has a high likelihood of severe relapse. Alternative Therapy
Alternative agents should be considered for patients with pulmonary sarcoidosis who fail to respond to corticosteroid therapy or cannot tolerate corticosteroid side effects. Table 3 outlines these medications. Of the alternative agents, methotrexate has been studied in most detail. Acute pulmonary sarcoidosis patients given methotrexate (10 mg/week) were shown to have similar improvement in spirometry as patients given corticosteroids (40 mg daily for 2 months, then 20 mg daily) (47). Methotrexate is the only alternative agent that has been shown to be effective for chronic sarcoidosis, but long-term treatment with this drug may be hampered by the cumulative risk of hepatotoxicity (48). Other alternative medications for acute sarcoidosis that have shown some usefulness include azathioprine (49, 50) and pentoxifylline (51). Chlorambucil is also effective for acute flares of sarcoidosis (52, 53) but cumulative toxic reactions, including the risk of malignancy, prevent its long-term use. Hydroxychloroquine has been found useful for cutaneous sarcoidosis (54), but its efficacy for pulmonary sarcoidosis is limited (55). It may be useful as a corticosteroid-sparing agent. Several studies of inhaled corticosteroids for pulmonary sarcoidosis (56-62) have suggested that these agents are beneficial in pulmonary sarcoidosis, although other studies have failed to demonstrate a benefit (63). Summary
Pulmonary sarcoidosis has a variable natural course from an asymptomatic state to a progressive life-threatening condition. Radiographic abnormalities are frequently an important clue to the diagnosis. The diagnosis usually requires a tissue biopsy that demonstrates noncaseating granulomas that cannot be ascribed to another clinical condition. The lung may be biopsied, but extrapulmonary sites may be biopsied for the diagnosis if such sites are involved with disease. When the lung is biopsied, a transbronchial lung biopsy with a flexible bronchoscope is the procedure of choice, even if the chest radiograph shows thoracic adenopathy alone without obvious parenchymal infiltration. On occasion the diagnosis can be made on clinical grounds without biopsy when the presentation is highly specific for sarcoidosis, such as Lofgren’s Syndrome. Treatment for pulmonary sarcoidosis has not been standardized. Since many patients have spontaneous remissions and the benefits of therapy do not affect the long-term outcome, therapy is reserved for patients with severe or progressive pulmonary symptoms and/or pulmonary dysfunction. Corticosteroids are the primary therapy for pulmonary sarcoidosis. Corticosteroid therapy involves 6 phases: initial high dose, taper to a maintenance dose, a maintenance dose, taper off corticosteroids, monitor off therapy, and treatment of relapse if it occurs. Referrals
If you would like to make a referral, you may either call MEDULINE at 1-800-922-5250, locally at 792-2200 or email Dr. Judson at judsonma@musc.edu. References | 
