A 47-Year-Old Woman With Sarcoidosis

DR REYNOLDS: Ms K is a 47-year-old Caucasian woman with sarcoidosis. She lives in Massachusetts and has commercial insurance. Ms K's diagnosis was made incidentally in spring 2005 when she was found to have cutaneous melanoma. A staging chest x-ray showed a small opacity in the right mid lung; a follow-up computed tomographic (CT) scan showed multiple bilateral lung nodules (Figure 1). Biopsy revealed non-necrotizing granulomas, which were acid-fast bacilli negative (Figure 2).

Since learning the diagnosis, Ms K notes several chronic symptoms that she now attributes to sarcoidosis. She experiences variable but sometimes intense fatigue, general achiness (especially in her legs), and dyspnea on exertion. Ms K also recalls “annual” wintertime episodes of shortness of breath diagnosed as bronchitis.

After Ms K was diagnosed with sarcoidosis, she consulted with a pulmonologist who recommended a trial of prednisone. Ms K chose not to take the prednisone, as she tries to avoid medications. She currently uses a fluticasone-salmeterol combination inhaler twice daily and a proton pump inhibitor to control her cough. She tried naproxen for her aches, but found that this caused bloating and constipation that outweighed any benefit it provided.

On the whole, Ms K's symptoms have not changed dramatically over the past year. She finds that her dyspnea is worst when climbing stairs and that the episodes of lower extremity achiness and weakness come and go.

Ms K's past medical history is significant only for 2 cesarean deliveries and a moderate anemia attributed to the sarcoidosis. She has never smoked and drinks alcohol in moderation. Ms K's brother was recently diagnosed with sarcoidosis after presenting with night sweats and weight loss, which were initially thought to be due to lymphoma. Her paternal grandfather was sent to a sanatorium for presumed tuberculosis but was later found not to have this diagnosis; she wonders whether he too had sarcoidosis.

On physical examination, Ms K appeared well. Her blood pressure was 128/94 mm Hg, her heart rate was 80/min, her oxygen saturation at rest was 99%, and her weight was 186 lb (84 kg). She did not have any lymphadenopathy. Her chest was symmetrical with expansion and resonant to percussion bilaterally. Breath sounds were clear to auscultation bilaterally. The remainder of her physical examination was within normal limits.

Ms K's laboratory tests showed a normocytic anemia with a hematocrit of 29%. Her electrolytes and liver-associated enzymes were within normal limits.

A recent CT scan showed small mediastinal lymphadenopathy with nodular septal thickening and nodules along the bronchovascular bundles and interlobular septa consistent with the diagnosis of sarcoidosis and was without change from 6 months prior. Her spirometry and diffusing capacity were within normal limits. Her blood pressure was in the high normal range after nonpharmacologic intervention.

We found out about the sarcoidosis only incidentally after I was diagnosed with a melanoma. The good news was that the nodule on chest x-ray wasn't what they thought it was. And then that helped explain a lot of the other symptoms, which I think were otherwise confusing. Some of them could be premenopausal, such as the tiredness and the aches, and you could attribute the shortness of breath to just being out of shape. But it is a severe shortness of breath, and it mostly comes on when climbing stairs or something like that.

It can be a little embarrassing when you're walking and talking with somebody, not even running upstairs, and can already be out of breath.

Usually I feel the achiness primarily in my legs. I sometimes just get exhausted for no particular reason, and that is when the achiness becomes more apparent.

The symptoms really haven't changed over time. The shortness of breath and the joint aches are still there and are very noticeable. I don't know whether I’ve just gotten used to feeling tired and compensate for it by going to bed earlier. Fortunately, the symptoms don't really prevent me from doing anything. However, it can also be frustrating, because you can't see it and you don't look different, so I almost feel like I’m being a baby when I complain about it.

My brother was diagnosed with sarcoidosis this summer. It seemed to lend a little legitimacy to what I was going through.

My doctor wasn't anxious for me to go on prednisone, nor was I keen on that myself. The symptoms didn't seem to be ones that I couldn't live with, so we decided to wait and see. I started using an inhaler, which isn't really helping. I’m due to go back to the pulmonologist and review with her whether I should start other medication.

I guess I’m hanging onto the idea I’ve heard that sarcoidosis can just crop up and then go away. I don't know how you actually know it has gone away other than by an improvement in symptoms. This is one of the things I would like to know more about. Also, what should I watch for, is there any time limit and what are the odds for it to go away, and what are the odds that it becomes an ongoing, chronic disease?

What are the usual presenting signs or symptoms of sarcoidosis and when should the diagnosis be considered? What is the pathophysiology of sarcoidosis? What are the treatment options available, when should they be used, and how does one choose between them? What is the prognosis for those with sarcoidosis? And finally, what do you recommend for our patient?

DR WEINBERGER: Ms K illustrates some of the typical features of sarcoidosis as well as the common issues that confront us as clinicians. As is often the case, her diagnosis was made because of abnormalities found incidentally on a chest radiograph performed for another reason. However, she did have some nonspecific symptoms, namely fatigue and achiness, which may be attributable to the disease. Whether her exertional dyspnea is due to sarcoidosis is unclear, but the combination of her anemia and her normal pulmonary function makes me suspect that other factors may be responsible. She has a positive family history of sarcoidosis, which I will address when discussing the genetics of the disease. Finally, I will address issues related to steroid treatment, both for her and more generically for patients with sarcoidosis.

Sarcoidosis is a difficult disease to define, in part because its cause is unknown and in part because its manifestations are so protean. The American Thoracic Society, the European Respiratory Society, and the World Association of Sarcoidosis and Other Granulomatous Disorders published a combined statement on sarcoidosis in 1999, in which they provided a descriptive definition of the disease.¹ According to this description, sarcoidosis is a multisystem disorder of unknown cause that commonly affects young and middle-aged adults. It frequently presents with bilateral hilar lymphadenopathy, pulmonary infiltration, skin lesions, and/or ocular involvement. The diagnosis is based on compatible clinical and radiographic features as well as noncaseating granulomas on tissue histopathology, with the caveat that no other explanations are available for the presentation or the granulomatous inflammation. Typical immune features of the disease include depressed cutaneous delayed hypersensitivity despite the presence of a heightened T_H1 helper cell lymphocytic response at the sites of disease.¹

Patients with sarcoidosis present in many ways.² Some patients, like Ms K, are diagnosed incidentally because of an abnormal chest radiograph, whereas other patients present with a variety of symptoms. Patients can develop respiratory symptoms, such as cough and dyspnea, or symptoms that are more systemic, such as fever. Symptoms or dysfunction of organ systems other than the lungs can occur as a result of granulomatous involvement of the particular organ. Alternatively, patients can have nonspecific symptoms that do not result from granulomatous involvement, such as chest pain, erythema nodosum (which represents a panniculitis), or arthralgias.¹

The epidemiology and demographics of sarcoidosis have been extensively studied.^{3 - 4} The disease has a worldwide distribution but with a particularly high prevalence in Scandinavia and Ireland. It is reasonably common in the United States, with a lifetime estimated risk of 2.4% for blacks and 0.85% for whites.⁵ One problem with assessing the frequency of the disease is that the methodology for case identification greatly affects the prevalence data. In those parts of the world in which sarcoidosis has been reported to be less prevalent (eg, China, Russia), detection of asymptomatic cases is likely less frequent than in the United States, where imaging studies and other diagnostic tests are performed more frequently. In a study of US Navy recruits screened with chest radiographs, nearly half the individuals found to have sarcoidosis were asymptomatic.⁶

Recent data on more than 700 patients with newly diagnosed sarcoidosis have been reported from the multicenter ACCESS (A Case Control Etiologic Study of Sarcoidosis) study sponsored by the National Institutes of Health.⁷ Because the ACCESS study involved only 10 academic centers and was not designed as a true incidence study, these demographic data are not necessarily generalizable to the total population of patients with sarcoidosis. Although there was a wide age distribution at the time the ACCESS patients presented with sarcoidosis, the peak age occurrence in both men and women was 35 to 39 years. However, new cases of sarcoidosis presented throughout the age spectrum, including in patients older than 65 years. Two thirds of affected patients were women. The racial distribution of new cases was 53% whites, 44% blacks, and 2% other. The higher relative risk of developing sarcoidosis for blacks than for whites in the United States is evident when comparing these rates with the US population distribution (75% whites and 12% blacks, according to US census data in 2000⁸ ).

A commonly used radiographic staging system for sarcoidosis categorizes patients based on the presence or absence of hilar adenopathy and parenchymal lung disease on chest radiograph (Figure 3).¹ This system is summarized in the Table, along with data from more than 700 newly diagnosed patients in the ACCESS study regarding the percentage of patients in each radiographic stage at the time of presentation.⁷ Ms K's chest radiograph presents a challenge in terms of the usual staging system, since the parenchymal lung disease in either stage 2 or 3 disease is typically more widespread than just a single nodule. Her CT scan (Figure 1) reveals more scattered nodules that are too small to be seen on the chest radiograph, but the radiographic staging system was defined by chest radiographic rather than CT appearance.

Accompanying the chest radiographic changes of sarcoidosis is a typical pattern of pulmonary pathophysiological changes.⁹ Involvement of the lung parenchyma can produce a restrictive pattern on pulmonary function testing, characterized by a relatively symmetrical decrease in lung volumes, and often accompanied by a decrease in the diffusing capacity for carbon monoxide (DL_CO). Hypoxemia, typically exacerbated by exercise, may be present in some cases. Some patients with sarcoidosis also have airflow obstruction, generally due either to endobronchial granulomas or to parenchymal scarring with resultant traction and distortion of airways. Interestingly, the degree of physiological impairment is often not well correlated with the extent of radiographic abnormalities, as evidenced by some patients with dramatically abnormal chest radiographs but normal pulmonary function.⁹

Fortunately, pulmonary function is relatively preserved in the majority of patients with sarcoidosis, as in the case of Ms K. At the time of presentation, 69% of patients in the ACCESS study had a normal forced vital capacity (FVC) 80% or more of predicted. In contrast, FVC was between 70% and 79% of predicted in 18% of patients, between 50% and 69% of predicted in 11% of patients, and below 50% of predicted in only 3% of patients.⁷

The multisystem nature of sarcoidosis has been well characterized by many older studies as well as the newer ACCESS study.^{1 ,7} More than 90% of patients with sarcoidosis have involvement of intrathoracic lymph nodes and/or the pulmonary parenchyma. The next most commonly involved organs or tissues are peripheral lymph nodes, skin, liver, and eye, each being involved clinically in approximately 10% to 33% of patients. Involvement is less frequent in a variety of other organs or tissues, including spleen, bones, and salivary glands. Although the heart and the nervous system are involved relatively infrequently, disease in these organ systems is potentially more serious. Cardiac sarcoidosis is often characterized by conduction disturbances, arrhythmias, and if the myocardium is extensively involved, by congestive heart failure. Neurosarcoidosis is typically manifest by cranial nerve (especially facial nerve) involvement, by peripheral neuropathy, or by basilar meningitis or spinal cord involvement.¹⁰ Ms K does not appear to have extrapulmonary involvement at this time.

When a patient such as Ms K presents with pulmonary sarcoidosis, whether to evaluate for extrapulmonary sarcoidosis is guided primarily by the presence of suggestive findings on history or physical examination. However, based on consensus,¹ several routine baseline studies are recommended at the time of diagnosis, including hemogram (for possible bone marrow or splenic involvement), routine blood chemistries (for hypercalcemia or hepatic sarcoidosis), electrocardiogram (for cardiac sarcoidosis), urinalysis (for renal involvement), and ophthalmological examination (for ocular sarcoidosis).¹ Subsequent studies for extrapulmonary sarcoidosis are done primarily for follow-up of previously documented extrapulmonary involvement or for evaluation of new symptoms or physical findings.

Biopsy demonstration of noncaseating granulomas confirms the diagnosis in a compatible clinical setting and in the absence of other causes of granulomatous disease as it did in Ms K's case (Figure 2). The most common biopsy sites are the lung, lymph nodes, or skin, depending on the patient's clinical presentation. However, biopsy is not necessary in all patients, particularly those who have symmetrical bilateral hilar lymphadenopathy and are asymptomatic or those with Löfgren syndrome (bilateral hilar adenopathy, arthralgias, and erythema nodosum).¹

The cause and the underlying pathogenesis of sarcoidosis have remained elusive. Although sarcoidosis is currently thought to represent a response to 1 or more exogenous agents in a genetically susceptible host, causative agents have not yet been identified and genetic susceptibility factors have not been well characterized.^{1 ,11 - 12} In fact, whether the protean manifestations of sarcoidosis represent a single disease or the unexplained granulomatous changes that link patients with various presentations of sarcoidosis represent a nonspecific, final common pathway to more than 1 underlying process or disease is not clear.

Several epidemiological, immunologic, and genetic approaches to better understand the nature of sarcoidosis have been informative but not definitive. Epidemiological clues were suggested by a study performed on the Isle of Man, near the coast of England. In this study, approximately 40% of patients with sarcoidosis were found to have had prior contact with other sarcoidosis patients, compared with only 1% to 2% of controls (without sarcoidosis).^{13 - 14} However, identification of cases and the selection of controls may have been biased, confounding the interpretation.¹⁵ The case-control ACCESS study also suggested environmental associations in that a higher frequency of insecticide use at work or exposure to microbial bioaerosols was identified in patients with sarcoidosis compared with matched controls, whereas smoking was less common among sarcoidosis patients than in matched controls.¹⁶

Many immunologic features of sarcoidosis have now been characterized.^{17 - 19} Activated T_H1 CD4 lymphocytes infiltrate at sites of disease activity, with the T-cell receptor types suggesting an oligoclonal expansion of the cell population. Cytokines and chemokines possibly play a role in the disease through lymphocyte-macrophage interactions and the development of granulomas. Interest has focused on tumor necrosis factor α (TNF-α) and interferon γ (IFN-γ) as particularly important mediators in disease pathogenesis.^{17 - 19}

The contribution of genetic factors to sarcoidosis is not yet well-defined. Sarcoidosis clusters in families, but whether this is due to genetic factors, transmission of an infectious agent, or common source exposure is not clear.²⁰ Genes at HLA and other loci on the short arm of chromosome 6 are potential candidates for predisposing to sarcoidosis.^{12 ,21} In addition, HLA and other genetic polymorphisms have been associated with particular clinical manifestations, such as parenchymal lung disease or Löfgren syndrome.^{22 - 23}

Numerous environmental or infectious agents have been evaluated for possible associations with sarcoidosis. Chemicals and other noninfectious agents suggested in the past but poorly supported by subsequent studies have included beryllium, zirconium, pine pollen, and clay.¹¹ Granulomatous disease has developed in donated organs transplanted into sarcoidosis patients, and sarcoidosis has developed in recipients of organs taken from patients known or later found to have sarcoidosis, suggesting a possible transmissible infectious agent.^{24 - 25} The infectious agents generating most recent interest are mycobacteria and propionibacteria.²⁶ While initial studies to identify mycobacterial DNA using polymerase chain reaction (PCR) techniques in affected sarcoidal tissue had inconsistent results,^{11 ,27} a subsequent study demonstrated 1 or more mycobacterial sequences by PCR in 60% of sarcoidosis tissue specimens but no control tissue specimens.²⁸ Additionally, a proteomics approach has found poorly soluble mycobacterial antigens in 75% of sarcoidosis tissues but only 14% of control tissues.²⁹

Also using PCR, Japanese investigators identified DNA from Propionibacterium species in 98% of lymph nodes from Japanese and European patients with sarcoidosis compared with 0% to 60% of tuberculosis or control lymph node samples (depending on the center from which the nonsarcoidosis samples were obtained).³⁰ The same group subsequently reported that P acnes is a common commensal in peripheral lung and mediastinal lymph node tissue in patients without sarcoidosis,³¹ but it has been hypothesized that expression of disease in response to a common external exposure such as an infectious agent may depend on interaction with genetically determined host factors.²⁶

As shown in Ms K's case, whether to treat sarcoidosis is a common clinical question. The decision is not straightforward because of the variable natural history of the disease and lack of data from controlled trials.

Based on consensus guidelines, extrapulmonary sarcoidosis affecting the heart or the nervous system typically warrants systemic therapy, whereas ocular sarcoidosis is treated either with topical or systemic therapy, depending on the nature and severity of the involvement.¹ For patients with pulmonary but without extrapulmonary involvement, the decision to treat is typically based on whether patients have abnormal pulmonary function or significant symptoms, and whether the disease is stable, progressing, or regressing over time. Treatment is not indicated for abnormalities on chest radiograph alone. The Table shows the expected spontaneous remission rates by radiographic stage, derived from the consensus article.¹ In a British Thoracic Society study, approximately half of those patients who did not immediately need therapy based on symptoms or extrapulmonary involvement had spontaneous improvement in their chest radiograph over a 6-month period.³² In addition, patients who present with Löfgren syndrome typically have a good prognosis with spontaneous resolution.³³ Whether to treat patients with progressive radiographic changes in the absence of symptoms or abnormal pulmonary function is more controversial.^{34 - 35} Ms K's chest x-ray does not readily permit classification of stage according to the Table, but with her limited parenchymal lung involvement and apparent lack of systemic disease, she would be expected to have a relatively high likelihood of spontaneous remission.

For patients with more severe disease possibly warranting treatment, it is still not known with certainty whether steroids alter the long-term natural history of the disease. A systematic Cochrane review of randomized controlled studies of oral or inhaled corticosteroid treatment identified 12 studies involving 1051 patients.³⁶ Although oral steroid treatment was associated with improved chest radiographs over 3 to 24 months (relative risk for improvement, 1.46 [95% confidence interval, 1.01-2.09]), other measured outcomes including pulmonary function were not significantly different. In addition, studies rarely extended beyond 2 years of follow-up. The few data on use of inhaled corticosteroids demonstrated no clear beneficial effect on either chest radiographs or pulmonary function.^{36 - 37} Inhaled steroids might be expected to be more useful for endobronchial disease and for treatment of cough due to endobronchial sarcoidosis, but the Cochrane review identified only a small number of patients treated with inhaled corticosteroids vs placebo and did not find an objective benefit of such treatment on cough.³⁶

Two specific studies illustrate the conundrum about therapy. In a double-blind, placebo-controlled trial of oral prednisolone for 3 months followed by inhaled budesonide for 15 months with follow-up at 18 months and 5 years,³⁸ no differences were found with treatment in patients who had stage 1 disease on chest radiography. Patients with stage 2 (and some with stage 3) disease had a slight (5%-10%) but statistically significant benefit of steroid treatment in terms of improved FVC and DL_CO; however, there was no difference in the effect on radiographic appearance. In contrast, an observational study using registry data found that patients who had disease remission associated with steroid treatment had a 74% subsequent relapse rate, in contrast to an 8% relapse rate for patients who had a spontaneous remission without steroids.³⁹ The authors noted 2 potential explanations for these findings—either that treated patients had a more severe form of disease that was more likely to have a progressive course, or that steroid treatment contributed to the propensity to relapse.

The usual duration of treatment of sarcoidosis is approximately 9 to 12 months, although this recommended duration is based on expert opinion rather than objective data.⁴⁰ Patients are typically treated initially with moderate doses of steroids (≤40 mg/d), and the dose is tapered to a lower maintenance dose (either daily or every other day) for most of the treatment duration. Patients may need to be retreated if suppressed disease relapses following discontinuation of therapy.

Alternative agents to steroids have also been used for specific clinical indications, for steroid-refractory disease, or for patients who do not tolerate steroids.^{17 ,40} Cytotoxic agents such as methotrexate or azathioprine may have an effect, but the data are not conclusive and they generally are used as second-line agents or for a steroid-sparing effect.^{41 - 42} Despite the effect of cyclosporine on T lymphocytes, which are thought to be important in the pathogenesis of the disease, cyclosporine has not been particularly effective and what benefits it may have are outweighed by its risks and adverse effects.⁴³ Antimalarials, particularly hydoroxychloroquine, have been used, most commonly for cutaneous sarcoidosis or sarcoid-related hypercalcemia, but controlled studies on their efficacy are limited.^{44 - 45} A recent Cochrane review analyzed the limited information available about some alternative forms of therapy and found that information was inconclusive about the benefit of methotrexate, cyclosporine, chloroquine, and pentoxifylline, but adverse effects were noted for all of these agents.⁴⁶

Agents that block the effect of TNF-α, such as infliximab, have been the focus of recent research.^{47 - 48} Tumor necrosis factor α is produced by alveolar macrophages in patients with sarcoidosis and is thought to be important in granuloma formation; elevated levels have been found in active and refractory disease. Anecdotal reports suggested a beneficial response to infliximab, including when used for refractory disease,⁴⁷ and a recently published phase 2 double-blind, placebo-controlled study found a small (2.5%) but statistically significant improvement in absolute percentage of predicted FVC after 24 weeks of therapy with infliximab compared with placebo.⁴⁸ In contrast, etanercept, the soluble TNF receptor that blocks the action of TNF-α through a different mechanism than infliximab, has not proven effective, with a phase 2 study discontinued early because of the lack of an effect.⁴⁹

Ms K fortunately does not have pulmonary dysfunction, as evidenced by her normal spirometry and DL_CO. She does have symptoms that may be from sarcoidosis, but they are nonspecific and she has been clinically stable. Even if her disease does not regress, patients can often have stable radiographic abnormalities over time, without progression of disease and without a need for treatment.¹ I agree with avoiding systemic steroids in the presence of relatively nonspecific symptoms and the absence of organ system dysfunction and, considering the scant data to suggest a benefit of inhaled steroids and her lack of an objective response, I would discontinue their use.

Given her normal pulmonary function, sarcoidosis likely does not explain her dyspnea. Furthermore, since anemia is reported to occur in only 4% to 20% of patients with sarcoidosis,¹ I would not automatically assume that Ms K's anemia can be attributed to sarcoidosis and it may be contributing to her exercise intolerance. If her dyspnea still remains problematic or unexplained, a cardiopulmonary exercise test could distinguish whether her dyspnea is related to parenchymal lung disease or to other possible contributing factors, such as deconditioning or occult cardiac disease. An exercise training program may then be useful, particularly if deconditioning is contributing to the symptom. Ms K's symptom of achiness can best be treated by a nonsteroidal anti-inflammatory drug. Given the adverse effects she has had with naproxen, I would try a different nonsteroidal anti-inflammatory drug. I would follow up her sarcoidosis by assessing symptoms, pulmonary function, and a chest radiograph approximately every 6 months, unless there is symptomatic worsening to justify earlier follow-up. If she is stable for 2 or more years, her follow-up can then be less frequent.

QUESTION: In patients with exercise intolerance, is it advisable to follow quantitatively the amount of exercise that can be done and the associated level of oxygen desaturation? Is this also a way to monitor whether treatment is working?

DR WEINBERGER: Exercise testing in the dyspneic patient is often quite helpful. One of the more sensitive ways to evaluate and follow interstitial lung diseases such as sarcoid is to measure oxygen saturation with exercise even when the PO₂ is normal at rest. Additionally, if the cause of dyspnea is not clear in a particular patient, exercise testing can be diagnostically useful.

QUESTION: What about the downside of steroids? Patients often push for a treatment trial, but then it's very hard to get them off the drug.

DR WEINBERGER: I agree entirely, and it is not uncommon to see patients with sarcoidosis who are overtreated or treated prematurely with steroids.

QUESTION: I have seen a number of patients who present with a neoplasm and also have sarcoid, just as was demonstrated in this case. What is the statistical association, if any, between those 2 diseases?

DR WEINBERGER: There are conflicting data in the literature about whether there is an increased risk of cancer in patients who have sarcoidosis.^{50 - 51} However, any such increase in risk, if present, would appear to be modest. In my personal experience following many patients with sarcoidosis, I have not been impressed by an unusual or unexpected incidence of subsequent cancer.

QUESTION: What has happened to the Kveim test?

DR WEINBERGER:The Kveim test is an intradermal skin test that uses a suspension of ground-up spleen from a patient with sarcoidosis. The test site is biopsied after 4 to 6 weeks, and the finding of noncaseating granulomas at the test site represents a positive response.⁵² Patients with sarcoidosis typically respond to the test material but not to “control” splenic tissue, and individuals without sarcoidosis do not respond to the test material. At present, standardized test material is not readily available and should only be used with special Food and Drug Administration approval.

QUESTION: You discussed the possibility that sarcoidosis might be related to an infectious agent. Have there been any trials of antimicrobials?

DR WEINBERGER: I am not aware of any controlled trials of antimicrobial agents in sarcoidosis. However, an open-label trial of minocycline for cutaneous sarcoidosis conducted in France reported complete or partial remissions in 10 out of 12 patients.⁵³ Even if the response in these patients was due to minocycline, which has not been proven, it is not known whether anti-infective, anti-inflammatory, or immunomodulatory properties of the drug may have been responsible.