|Skip left side navigation and go to content||
Future Directions in Sarcoidosis Research
Summary of an NHLBI Working Group
Published in Am J Respir Crit Care Med Vol 170. pp 567–571, 2004
William J. Martin II, Michael C. Iannuzzi, Dorothy B. Gail, and Hannah H. Peavy
College of Medicine, University of Cincinnati, Cincinnati, Ohio; Pulmonary, Critical Care, and Sleep Medicine, Mount Sinai School of Medicine, New York, New York; and Division of Lung Diseases, National Heart, Lung, and Blood Institute, National Institutes of Health, Bethesda, Maryland
This working group, supported by the Division of Lung Diseases, National Heart, Lung, and Blood Institute (NHLBI), National Institutes of Health, Bethesda, Maryland, was held on August 22-23, 2002.
Sarcoidosis is a systemic granulomatous disease of unknown etiology that primarily affects the lungs. The etiology remains unclear; however, environmental, genetic, ethnic, and familial factors probably modify expression of the disease. As an example, African Americans are at greater risk of mortality and morbidity than are white Americans, and more often have a family history of sarcoidosis. Most patients with sarcoidosis recover spontaneously, but some develop chronic, debilitating disease. Corticosteroids and other drugs, although effective at controlling disease activity, may not influence the overall course of disease. Because of the many uncertainties about the pathogenesis, course, and management of sarcoidosis, the National Heart, Lung, and Blood Institute convened a working group to identify future research directions and opportunities for sarcoidosis. These include developing a tissue bank, using novel methods to identify genetic factors, studying the immunopathogenesis with human tissue and animal models, exploring new approaches to diagnose and manage disease, and, finally, conducting randomized controlled trials to assess new therapies.
Keywords: environmental; genetics; granuloma; lung disease; National Institutes of Health (United States)
Sarcoidosis, first described 125 years ago, is a systemic granulomatous disease of unknown etiology that primarily affects the lung, although multiorgan involvement frequently occurs. It is likely that expression of the disorder in response to one or more inciting agents is modifed by the genetics of the host immune response. In addition, an infectious component has long been suspected (1–3). Although many clinical features and associated syndromes have been described, the phenotype of sarcoidosis at diagnosis provides only limited information about the etiology or pathogenesis of the disorder.
Spontaneous remission or disease stabilization is reported to occur in approximately two-thirds of cases, but many patients have chronic progressive disease (2). About 4–7% of patients have serious extrapulmonary (i.e., heart, central nervous, liver, or eye) involvement at the time of presentation; this possibility increases as the disease progresses (2). Data on persistent disease and relapses from centers in the United States that treat a large number of patients with sarcoidosis vary from 20% in Iowa (4) to more than 50% among patients in Philadelphia whose treatment was stopped (5). Mortality rates at referral centers are high at 3–10% (6), whereas low mortality is reported from nonreferral centers (7). No good test for predicting disease progression exists and better means are needed to differentiate between remitting and chronic sarcoidosis. Much remains to be learned about determinants of susceptibility to sarcoidosis, optimal treatment strategies, and reasons for disease persistence.
The NHLBI convened a working group to identify the research directions that will promote better understanding of the etiology and pathogenesis of the disease, improve patient management, and develop novel therapeutic interventions. The participants were asked to briefly review the current state of knowledge about sarcoidosis; identify major gaps in understanding etiology and pathogenesis; identify major clinical problems, research needs, and opportunities; and develop specific recommendations for future sarcoidosis research.
PATHOGENESIS OF SARCOIDOSIS
Little is known about the mechanisms that induce the destructive inflammatory response, that is, noncaseating granulomas, in sarcoidosis. Granulomas occur as a result of a helper T cell Type 1 protective immune response and are dynamic structures. Under certain known (persistent antigen) and unknown circumstances, the exuberance of the granulomatous response becomes deleterious, resulting in organ damage and scar formation. The deleterious effects of the granulomatous inflammation in sarcoidosis depend on the extent and location of the inflammation. For example, limited granulomatous inflammation of the heart or nervous system may have devastating consequences, whereas extensive involvement of lymph nodes or spleen may have minimal consequences.
Granuloma formation in organs and lymph nodes is dependent on several factors: local IFN-gamma secretion; selected expression of potent monocyte and lymphocyte chemotactic cytokines (chemokines); and chemoattraction of unique receptor-bearing T cells and monocytes, which may determine the differentiation or activation state of the cell (e.g., activated CD4+ memory cells). The cause of the destructive lesions may be determined by the persistent presence of antigen; by excessive synthesis of chemotactic factors; or by the subsequent influx, immobilization, apoptosis, efflux, and recirculation of recruited cells. Any of these steps would be reasonable targets for new therapies.
Granuloma formation in sarcoidosis may be a stereotypical response to multiple infectious and noninfectious agents. Research to elucidate the destructive tissue inflammatory response is needed to identify which factors are responsible for accumulation of destructive T cells and monocytes in organs and how best to target therapies that will neutralize chemotactic factors or block their interaction with receptors. If successful, this approach might obviate the need for identification of a specific etiologic agent or host HLA genotype. Two challenges in controlling the destructive granulomatous response will be to identify selective inhibitors that do not affect the immune response to other agents and to specifically target therapies to the organs involved with sarcoidosis.
A possible clue to the etiology of sarcoidosis is the Kveim– Siltzbach reaction, a local delayed granulomatous response to the intradermal inoculation of sarcoidosis splenic or lymph node tissue (1). The Kveim reagent is thought to contain antigens that induce an immune-mediated granulomatous response (1, 8–10). To date, identification of specific antigens in the Kveim reagent that induce a sarcoid-like reaction has not been accomplished. A proteomic approach to identify unique proteins in sarcoidosis tissues might be a valuable opportunity to learn more about the role of antigens. Identification of a disease-related antigen in sarcoidosis tissue, including Kveim reagent, would be an important step to furthering understanding of the pathogenesis of this disease and should be encouraged through support of high-risk research projects.
The working group concluded that tissue banks are necessary to provide substantial amounts of lung tissue (surgical and not transbronchial biopsies) as well as tissue from other organs and sources of tissue including cell pellets from bronchoalveolar lavages, blood, and DNA. Availability of these resources would facilitate defining the set of chemotactic factors that contribute to granuloma pattern and would make it easier to conduct microarray and proteomic analyses to discover other unknown factors that can be used to target downstream inflammatory events.
SEARCHING FOR AN INFECTIOUS CAUSE OF SARCOIDOSIS
Infectious origins are suspected for many human diseases of unknown etiology, on the basis of epidemiologic and clinical features (11, 12). These diseases include cancers, autoimmune disorders, and inflammatory diseases. Because sarcoidosis is similar in some ways to diseases caused by mycobacteria and fungi, the possibility of a causative role for a microbial agent seems high. This is supported by the predilection for cases to occur in spring months, the clustering of cases, and the observation that certain bacteria, viruses, and fungi induce granulomatous host responses similar to those seen in sarcoidosis. It is important to determine whether the granulomas seen in sarcoidosis represent ongoing disease from an unknown infectious agent. It is possible that the lung is the primary site of entry for this agent with subsequent systemic infection and dissemination or, alternatively, that the sites of chronic inflammation result from inadequate clearance or tropism of some infectious agent. Improved techniques to isolate nontraditional infectious agents are needed. Clearly, therapeutic opportunities for a putative infectious agent would include antimicrobial therapy or vaccination.
The search for unrecognized pathogens is difficult, as many microorganisms are resistant to cultivation in the laboratory, cultures are frequently contaminated by normal flora, and even when an agent is present it may not be causing disease. The investigators participating in a multicenter Case-Control Etiologic Study of Sarcoidosis (ACCESS) attempted to amplify bacterial DNA (including mycobacterial DNA) from blood samples with polymerase chain reaction primers, but could not confirm the presence of bacteria. They also cultured blood specimens from 197 patients with recently diagnosed sarcoidosis and 150 control subjects and found that cell wall–deficient forms of mycobacteria grew with equal frequency from case subjects (38%) and control subjects (41%). A major limitation of these studies was that only blood and not tissues from involved organs was available for study (13). Similarly, the ACCESS investigators looked for, but were not able to demonstrate, in vitro immune responses to Kveim preparations. This may have been due to suboptimal in vitro assay conditions or insufficient bioavailable antigen.
Another obstacle to linking an infectious agent with sarcoidosis is that the prevalence and incidence as well as the seasonal and geographical distribution of sarcoidosis in the population are not well defined (1). Better tracking, for example, by state programs in collaboration with the CDC might provide the ability to identify clusters of cases that could be important in uncovering potential associations with infectious or environmental agents.
A role for pathogens in the development of sarcoidosis has been suggested. Proposed candidates include mycobacteria, human herpes virus, retroviruses, Chlamydia pneumoniae, Borrelia burgdorferi, and Rickettsia helvetica. However, none of these agents has been unequivocally proven to be associated with the pathogenesis of sarcoidosis in a well designed case-control study. Molecular techniques, including the sequence-based computational subtraction technique (14, 15), could be used to search for pathogen-specific sequences in this chronic inflammatory disorder.
Sequence-based computational subtraction relies on comparing the sequence of the human genome with genetic sequences obtained from infected tissues containing RNA and genomic DNA (or RNA) from infectious agents. The presence of nonhuman transcripts can be detected by sequencing cDNA libraries derived from infected tissue, followed by elimination of the sequences matching the human genome. The normal human sequences can thus be identified and “subtracted.” The remaining “filtered” sequences contain genes from unsequenced regions of the human genome, poor quality sequences, known microbial sequences, and unknown microbial sequences. Subsequently, the disease specificity of the subtracted, nonhuman sequences can be tested by employing a follow-up polymerase chain reaction analysis. Using the sequence-based computational subtraction technique, it has been possible to identify known pathogens in filtered sequences; these include hepatitis B virus, Heliobacter pylori, Salmonella typhimurium, human papillomaviruses, Epstein-Barr virus, Kaposi’s sarcoma virus (HHV-8), cytomegalovirus, and hepatitis C virus. Novel disease-causing organisms might be identified by similar approaches. Limitations to using this technique include the possibility of contamination by normal flora, which could lead to identifying the wrong agent and the possibility that even if an agent is present, it may not be causing disease (14, 15).
Detection of a pathogen-specific sequence fragment, if one can be identified, could lead to the identification of the entire genome and protein products of a pathogen. What constitutes suitable specimens and sufficient tissue requirements and appropriate handling needs to be determined. Clearly, samples from involved tissues are needed. Samples could be analyzed in parallel by other methods, in particular by polymerase chain reaction amplification of bacterial ribosomal DNA sequences (16).
GENETICS OF SARCOIDOSIS
While environmental factors remain to be defined, race and family history of disease are, at present, the most strongly identified risk factors for developing sarcoidosis, thus supporting the notion that there is a genetic susceptibility for development of this disease (17). Familial aggregation and racial differences in disease incidence have been reported worldwide. In the United States, ACCESS found a fivefold increase in the relative risk of developing sarcoidosis among the first-degree relatives of patients with sarcoidosis (17). African Americans are at least four times more likely to develop the disease than are white Americans (18).
It is likely that the genetic susceptibility to sarcoidosis involves several genes, each with a small to moderate effect. Investigators have evaluated many attractive candidate genes, but nearly all reported associations with sarcoidosis risk remain unconfirmed (19). The most likely explanation for inconsistent results relates to the problem of population stratification inherent in case-control studies. The major histocompatibility complex (MHC) region, on Chromosome 6p, has been and continues to be an important target of investigation (20–26). Association studies and linkage analyses support the idea that candidate disease susceptibility gene(s) occupy a 5 million–base pair–long stretch of the human genome located on the short arm of Chromosome 6. This is also the locus of the MHC, which is a critical part of the human immune system (23–26). Studies have suggested that genes that encode the human leukocyte antigen (HLA) portion of the MHC play an important role in determining the risk and the clinical course of sarcoidosis. For example, HLADQB*0602 is transmitted more often than expected to affected offspring and appears to be associated with progression of disease, whereas HLA-DQB1*0201 is found only half as often as expected (24). Besides the HLA genes conferring susceptibility, they may also be involved in disease outcomes. Numerous investigators have been able to link certain HLA Class II genes with favorable clinical outcomes for patients with sarcoidosis (27). In addition, polymorphisms in genes of the interleukin-1 family and the promoter region of the gene for tumor necrosis factor-alpha have been reported to be associated with the pathogenesis of sarcoidosis (28–30).
One attractive approach to identifying likely candidate genes is to use positional cloning that relies on evaluating candidates in linked chromosomal regions. Because susceptibility to sarcoidosis probably involves many genes, a genome-wide scan offers the most systematic and efficient approach to unraveling the genetics of sarcoidosis. So far, only one genome-wide search for sarcoidosis-predisposing genes has been reported. On the basis of 225 microsatellite markers tested in 63 German families with affected siblings, linkage at the major histocompatibility complex was found, with additional suggested linkage to markers on Chromosomes 1, 3, 9, and X (22). The U.S. Sarcoidosis Genetic Analysis Consortium is about to complete a linkage analysis of 360 African-American families with affected siblings, using a 300-microsatellite marker scan. The U.S. Sarcoidosis Genetic Analysis Consortium study has focused on African Americans as they are more likely to report a family history of sarcoidosis, present at an earlier age, and have more severe disease (31).
CROHN’S DISEASE AS A MODEL OF SARCOIDOSIS
The working group members discussed possible lessons applicable to sarcoidosis that may be learned from research on inflammatory bowel disease, including Crohn’s disease and ulcerative colitis. Both are chronic inflammatory conditions of the intestines and are of unknown cause. The focus was mostly on Crohn’s disease, which can affect any region of the gastrointestinal tract, although the ileum and colon are the sites most frequently involved. Crohn’s disease becomes apparent during the second or third decades of life. It can begin gradually or suddenly, usually without known precipitating factors. In some patients disease can be precipitated by certain intestinal infections. Crohn’s disease is believed to result from an overresponse of the mucosal immune system to subsets of usually harmless intestinal bacteria that normally reside in the gut.
In considering approaches to investigating the pathogenesis of Crohn’s disease as a possible model for basic research on sarcoidosis, the following similarities between the two diseases need to be kept in mind: (1 ) both Crohn’s disease and sarcoidosis are chronic relapsing/remitting granulomatous inflammations; (2 ) both diseases generate helper T cell Type 1 inflammation; and (3 ) both are probably manifestations of interactions between environmental and genetic factors. In addition, there are linkages between the bowel and the lung in inflammatory bowel disease, as ulcerative colitis and to a lesser extent Crohn’s disease can be associated with inflammatory processes in the lung, although this is rare and the pathogenesis is poorly understood.
Differences between Crohn’s disease and sarcoidosis include the following: (1 ) Crohn’s disease is common in a white, relatively affluent population, but sarcoidosis is seen mostly in an African-American, less affluent population; (2 ) unlike sarcoidosis, Crohn’s disease is common in the North and less common in the South; (3 ) Crohn’s disease occurs with equal frequency in men and women, whereas sarcoidosis occurs more frequently in females; (4 ) cigarette smokers are at increased risk of developing Crohn’s disease (32), but smoking cigarettes is associated with a decreased risk of developing sarcoidosis (1); (5 ) patients with sarcoidosis are not prone to Crohn’s disease and vice versa; and (6 ) sarcoidosis of the gut is unusual and looks histologically different from Crohn’s disease; and finally, (7 ) lung manifestations are rarely seen in Crohn’s disease.
There is strong evidence that Crohn’s disease is a complex genetic disease with important environmental modifying factors. Genome-wide scans have identified multiple chromosomal regions linked to Crohn’s disease, including the HLA region of Chromosome 6, the cytokine gene cluster on Chromosome 5, and a region on Chromosome 16 for which a specific gene has been identified, named Nod2/card15. The gene product of card15 appears to be an intracellular pattern recognition molecule for bacterial peptidoglycans. The card15 mutations in Crohn’s disease are loss-of-function mutations, which appear to interact with adapter molecules controlling NF-kappaB activation. Homozygous loss-of-function mutations in card15 are associated with a high relative risk of Crohn’s disease.
Crohn’s disease is more common in urban compared with rural areas, and people who live in less developed countries rarely develop it. The rapid rise in the prevalence of Crohn’s disease during the twentieth century also suggests the presence of important environmental modifying factors, only some of which have been identified (33). For example, smokers are twice as likely to develop Crohn’s disease as nonsmokers. It is possible that an increasingly hygienic environment has eliminated exposure to some particular organisms in early life, and that this predisposes individuals to Crohn’s disease.
The working group members agreed that mouse models of sarcoidosis are needed. This approach has been useful in investigating the immunomodulatory pathways of Crohn’s disease and could be useful in investigating poorly controlled inflammation in sarcoidosis. However, there are no animal models that completely simulate the disease process of Crohn’s disease and it is highly unlikely that a complete model for sarcoidosis could be developed. Several animal models of intestinal inflammation allow analysis of immunoregulatory mechanisms that limit intestinal inflammation. Interleukin-10 knockout mice, for example, exhibit spontaneous inflammation of the gut and are particularly useful. These animals exhibit enhanced helper T cell Type 1 responses associated with excess production of interleukin-12 and IFN-gamma. They also display difficulty in generating regulatory T cells that limit immune responses. The group concluded that similar mouse models might be useful for investigating inflammation in the lung and other organs and might provide clues about the pathogenesis of sarcoidosis.
DIAGNOSIS AND TREATMENT
Sarcoidosis remains a diagnosis based on exclusion. Currently, the diagnosis is based on a typical clinical presentation, compatible imaging, physiologic studies, blood analyses, a biopsy showing typical noncaseating granulomas, and the absence of a fungal or mycobacterial infection. New approaches to the diagnosis of sarcoidosis should be considered, perhaps based on evolving genetic and proteomic technology.
Assessment of disease activity in sarcoidosis is fraught with challenges as currently available measures are nonspecific and offer little prognostic value. It is unclear why some patients have favorable outcomes without apparent disease progression and why others develop extensive single or multiorgan damage. Sarcoidosis is also associated with an increased risk of depression, something medical personnel should be aware of, as depression may complicate the course of disease (34, 35). Laboratory assessment of disease activity is dependent on nonspecific tests including roentgenographic findings, physiologic testing, and blood studies such as angiotensin-converting enzyme activity. Improved biomarkers are needed to assess disease activity, to differentiate between remitting and chronic sarcoidosis, and to predict susceptibility and prognosis.
Treatment for patients with sarcoidosis remains problematic. The mainstay of therapy continues to be oral corticosteroids. Treatment of some manifestations of sarcoidosis with corticosteroids is clearly indicated, such as critical organ involvement (heart, cranial nerves, and eye), but for less severe or nonprogressive involvement (e.g., of the lung) the benefits are less obvious and it is not clear whether the potential benefits outweigh the risks. A review of five randomized controlled clinical trials indicated that steroid treatment for pulmonary involvement results in radiographic improvement and small benefit to vital capacity and diffusing capacity, but suggested that the long-term course of the disease is not altered (36). The British Thoracic Society study of long-term corticosteroid therapy, however, showed a benefit in Stage II and III disease (37). A “gold standard” test to identify which patients require treatment when noncritical organs are involved would be useful. A randomized, double-blind, placebo-controlled clinical trial reported that immediate treatment of pulmonary Stage II–III disease, but not Stage I, improved lung function indices at 5-year follow-up (38). The possibility that steroids might actually prolong the course of sarcoidosis has also been raised (5). It remains unproven whether the long-term course of sarcoidosis is altered by corticosteroid treatment. Corticosteroids are effective but not specific, and have many side effects. Development of novel, targeted treatment alternatives to corticosteroids that are more specific and less toxic should be encouraged. Cytotoxic agents are also being used, including methotrexate, azathioprine, and cyclophosphamide. Exaggerated tumor necrosis factor release from the alveolar macrophages of patients with sarcoidosis who do not respond to corticosteroids suggests a potential role for tumor necrosis factor in sarcoidosis (39). Tumor necrosis factor antagonists, including pentoxifylline (40), infliximab (41), and thalidomide (42), have been used in the treatment of sarcoidosis; however, the studies are limited and the ultimate benefits uncertain. Etanercept is another anti–tumor necrosis factor agent. A small Phase II study of etanercept for progressive pulmonary sarcoidosis was stopped early because the drug was frequently associated with early or late treatment failure. The investigators concluded that the data on etanercept would not support going ahead with a large randomized study of etanercept versus conventional treatment with corticosteroids (43). Development of collaborative treatment trials in sarcoidosis is necessary to validate current and proposed new therapies. The accurate clinical phenotyping of subjects with sarcoidosis will be critical to the success of any trial. Successful case-control studies of sarcoidosis such as ACCESS suggest the time is right to conduct such treatment trials.
The NHLBI Working Group formulated the following recommendations for future research directions.
Conflict of Interest Statement: W.J.M. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript; M.C.I. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript; D.B.G. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript; H.H.P. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript.
Acknowledgment: The authors thank the workshop participants, who provided the important ideas embodied in this report. These participants included: Chairman, William J. Martin II, Cincinnati, OH; Jeffrey S. Berman, Boston, MA; David Center, Boston, MA; Sheila Clark, Silver Spring, MD; Robinson Fulwood, Bethesda, MD; Glenda Fulton, Chicago, IL; Dorothy B. Gail, Bethesda, MD; Bernadette Gochuico, Bethesda, MD; Sandra Columbini Hatch, Bethesda, MD; Richard Helmers, Scottsdale, AZ; Gary Hunninghake, Iowa City, IA; Michael C. Iannuzzi, New York, NY; Stephen P. James, Bethesda, MD; Li Jin, Cincinnati, OH; James P. Kiley, Bethesda, MD; Linda Lanier, Cheltenham, MD; Stephen B. Liggett, Cincinnati, OH; Vincent C. Manganiello, Bethesda, MD; David R. Moller, Baltimore, MD; Joel Moss, Bethesda, MD; Pamela Mullins, Highland Park, MI; Robert Musson, Bethesda, MD; Dolores O’Leary, Sumner, WA; Cindy Palace, Bethesda, MD; Matthew K. Park, Bethesda, MD; Hannah H. Peavy, Bethesda, MD; Paula Yvette Polite, Memphis, TN; Ganesh Raghu, Seattle, WA; Milton Rossman, Philadelphia, PA; Marsha Thornhill, Cliffside Park, NJ; Joel V. Weinstock, Iowa City, IA; and Yaohui Xu, Boston, MA.
Correspondence and requests for reprints should be addressed to:
Hannah H. Peavy, M.D.
Posted: September 2004