| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1 Pulmonary, Allergy, and Critical Care Division, Department of Medicine, Hospital of the University of Pennsylvania, Philadelphia, Pennsylvania
Correspondence and requests for reprints should be addressed to Milton D. Rossman, M.D., 834 West Gates Building, Hospital of the University of Pennsylvania, 3400 Spruce Street, Philadelphia, PA 19104. E-mail: rossmanm{at}mail.med.upenn.edu
ABSTRACT
ACCESS (A Case Control Etiologic Study of Sarcoidosis) was funded by the National Institutes of Health and collected data on 704 newly diagnosed, biopsy-proven cases of sarcoidosis and control subjects matched by age, sex, race, and geographic area. The goal of this study was to generate hypotheses about the etiology of sarcoidosis. The major hypothesis of the ACCESS investigators was that sarcoidosis occurs in genetically susceptible individuals through alteration in immune response after exposure to an environmental, occupational, or infectious agent. Strict criteria were used for the diagnosis of sarcoidosis and definitions of specific organ involvement were developed. The patients recruited for ACCESS represent the best clinical description of sarcoidosis at presentation in the United States. The study investigated the following: occupational/environmental triggers using a detailed questionnaire, infectious agents in the blood by polymerase chain reaction of 16s rDNA of microorganisms and cultures for cell wall–deficient mycobacteria, and genetic associations using a questionnaire to determine familial aggregation and candidate gene analysis. No single cause of sarcoidosis was identified. The results of this study are reviewed and possible lessons learned are discussed.
Key Words: granuloma • lung • genetics • epidemiology • exposure
Sarcoidosis is a systemic granulomatosis that was first described in 1877. The hallmark of this disease is the noncaseating granuloma (1). The granulomas are sharply circumscribed and consist of epitheliod cells, Langerhans giant cells, lymphocytes, monocytes, and fibroblasts. Although they occur in every organ, in over 90% of cases there is intrathoracic involvement. Every racial or ethnic group has been described with this disease, but it appears to be more common in African Americans in the United States. The diagnosis of sarcoidosis usually requires a clinical presentation consistent with the disease, demonstration of typical lesions, and the exclusion of other disorders known to cause granulomas. Although prior epidemiologic studies have been done in sarcoidosis, they have been problematic because of the lack of a precise and consistent case definition, variable methods of case ascertainment, and variability of disease presentation. The lack of a sensitive and specific diagnostic test raises concerns of both underrecognition and misdiagnosis. Because of these problems, there was a paucity of systematic epidemiologic investigation in sarcoidosis. Thus, in 1992, the National Institutes of Health (NIH) selected 10 clinical centers and 1 clinical coordination center to conduct a study to generate hypotheses about etiologic agents in sarcoidosis. The study was not intended or designed to prove any specific hypothesis.
The clinical centers, the clinical coordinating center, and the NIH spent a year designing the study (2). During that time, it was decided that a matched case-control study would have the greatest power to generate a hypothesis about etiologic agents in sarcoidosis. The study was called ACCESS (A Case Controlled Etiologic Study of Sarcoidosis). A strict definition of sarcoidosis was adopted. To be accepted, cases had to be clinically consistent with sarcoidosis, have tissue confirmation of noncaseating granulomas, and no evidence of other diseases that cause granulomas. In addition, only newly diagnosed cases (within 6 mo of tissue diagnosis) would be recruited into the study so that if an occupational or environmental agent was involved in the disease, it would be more likely recalled by the patient. Finally, definitions (3) were adopted to define specific organ involvement if a definite diagnosis of sarcoidosis was determined by the above criteria (Table 1).
|
The major hypothesis of the ACCESS investigators was that sarcoidosis occurs in genetically susceptible individuals through alteration in immune response after exposure to an environmental, occupational, or infectious agent(s). A questionnaire was developed to elicit occupational and environmental exposures and an extensive job history was obtained. A detailed family history was obtained to test for a possible genetic predisposition. In addition, a medical and social history was recorded. Finally, blood DNA and plasma were obtained on all individuals to investigate possible infectious causes and genetic predisposition by candidate gene polymorphisms. Bronchoalveolar lavage (BAL) fluid was obtained from a subset of patients but not control subjects, and cell counts were performed but BAL fluid was not routinely collected because of financial restraints. The clinical course of the first 215 patients was observed after 2 years by determining organ involvement and activity, treatment, and a short follow-up questionnaire.
Initially, the goal was to recruit 720 cases and matched control subjects over 4 years. Because of financial concerns, the recruitment time was shortened; however, between November 1996 and April 1999, 738 cases were enrolled and 727 attempted matches were made (2). Seven hundred and four control subjects were successfully matched by age, sex, race, and geographic area. Thus, the first lesson that was learned from the ACCESS study was that 10 university medical centers (predominately East Coast and Midwest) in the United States could successfully recruit over 700 newly diagnosed subjects with sarcoidosis and matched control subjects in fewer than 3 years. This success paved the way for a second study investigating the genetics of sarcoidosis (the SAGA (Sarcoidosis Genetics Analysis) study [4, 5]), and the relationships among the investigators also led to the development of two randomized double-blind placebo-controlled drug trials in sarcoidosis, one of which has been published (6).
Although the cases collected within the geographic areas around each center were collected consecutively, all the newly diagnosed cases within a geographic area were not enrolled. Thus, the ACCESS study could not determine the incidence of sarcoidosis in the United States. Nevertheless, the age, sex, and self-designated racial distribution of these cases and the organ involvement at presentation (Table 2) offer the best description to date of sarcoidosis in the United States (7). Because most of the principal investigators in the study were pulmonary physicians, the distribution of cases may have been biased toward pulmonary cases and against cases of other organ involvement.
|
Finances limited the study so that only 215 patients could be reevaluated 2 years after their initial presentation (9). Eighty percent had improved or had stable function as assessed by forced vital capacity, dyspnea, or chest radiographic stage. Improvement of chest radiographic stage was associated with erythema nodosum. Less improvement in forced vital capacity was associated with African-American race. Lower family income was associated with new organ involvement and worse dyspnea. Finally, new organ involvement was associated with being African American or having extrapulmonary involvement at presentation.
Thus, the second lesson learned from the ACCESS study was about the clinical description of sarcoidosis at presentation in the United States and the role of social economic status, and some other possible barriers to care of these individuals. However, finances played an important role in how much we learned. Bronchoscopy and BAL were not routinely performed, computed tomographic chest X-rays were not performed, and follow-up was performed on only the first 215 patients.
The major goal of ACCESS was to generate hypotheses about the etiology of sarcoidosis. One of the major attempts to do this was through a detailed occupational and environmental questionnaire (10). Occupations associated with sarcoidosis involved raising birds (odds ratio [OR], 3.73; 1.10–12.59), automobile manufacturing (OR, 13.38; 1.48–120.98), secondary or middle-school teacher (OR, 1.80; 1.14–2.83), cotton ginning (OR, 4.98; 1.19–20.89), and work involving radiation exposure (OR, 2.28, 1.17–4.47). Exposures associated with sarcoidosis included occupational insecticide exposure (OR, 1.61; 1.13–2.28), mold and mildew (OR, 1.62; 1.24–2.11), use of home central air conditioners (OR, 1.49; 1.00–2.21), and bird exposure (OR, 1.49–2.21). In addition, certain exposures that had previously been associated with sarcoidosis were not associated with sarcoidosis in the ACCESS study. These exposures included nonoccupational wood dust exposure, lumbering, occupational metal and silica exposure, and dusty trades. There was insufficient power in the study to test whether or not firefighting or service in the military was associated with sarcoidosis. Finally, sarcoidosis was negatively associated with smoking, job as a data processor/typist or programmer, service as a hospital volunteer, having a fish tank greater than 10 gallons, and the use of foam pillows. In addition, contrary to a prior hypotheses, unpaid childcare, cats, use of feathered pillows/down, hot tub use, gold exposure, and welding were negatively associated with sarcoidosis.
A detailed occupational history was taken as part of the ACCESS study and each individual was characterized as to standard industrial codes (SICs) and standard occupational codes (SOCs) (11). Associations with sarcoidosis were found for building materials, hardware, garden supplies, and mobile homes (SIC, 52) (OR, 3.20; 1.12–11.17), industrial organic dusts (SIC combination) (OR, 2.57; 1.35–5.16), elementary and secondary schools (SIC, 821) (OR, 1.43; 1.03–2.00), and educators (SOC combination) (OR, 1.42; 1.02–1.99). Jobs and industries with negative associations with sarcoidosis were personal service (SOC, 526) (OR, 0.48; 0.29–0.78), electric energy (SIC, 36) (OR, 0.50; 0.25–0.96), social and rehabilitation services (SIC, 83) (OR, 0.62; 0.44–0.87), childcare provider (SIC combination) (OR, 0.65; 0.45–0.93), childcare provider (SOC combination) (OR, 0.65; 0.42–0.99), information clerks (SOC, 464) (OR, 0.65; 0.42–1.00), and metal dust/metal fumes in whites (SOC combination) (OR, 0.69; 0.48–0.98).
Thus, the third lesson from the ACCESS study was that a detailed occupational and environmental questionnaire could not identify a unifying exposure as a trigger (cause) of sarcoidosis. Most associations had ORs of less than 2.0, and when they were greater than 2.0, they occurred in small numbers, which accounted for the wide range of the confidence intervals. Although this might suggest that there are many triggers for sarcoidosis in a manner similar to asthma, this also does not exclude the possibility that a single or predominant trigger might still exist.
Two studies in ACCESS attempted to address the issue of infectious agents. The rates of positive blood cultures for cell wall–deficient mycobacteria were similar in patients and control subjects (12). In addition, polymerase chain reaction for 16s rDNA of microorganisms did not reveal any differences between patients and control subjects (manuscript in preparation). Thus, the fourth lesson from ACCESS was that an infectious agent as the cause of sarcoidosis was unlikely to be found using molecular techniques in the analysis of blood. Whether these techniques will demonstrate an organism in infected tissue is not known.
The role genetics played in the predisposition to sarcoidosis was assessed both by a questionnaire to determine familial aggregation and by candidate gene analysis. An increased risk for sarcoidosis was observed in sibs and parents of subjects with sarcoidosis (13). This increased risk was present in both whites and African Americans. High-resolution molecular typing was used to analyze the HLA class II genes in the first 476 cases and control subjects in ACCESS (14). The distribution of HLA DRB1 alleles was significantly different between cases and control subjects (p < 0.0001). The HLA DRB1 alleles that accounted for this difference were DRB1*1101 (OR, 2.68; 1.78–4.03), DRB1*0402 (OR, 4.59; 1.76–12.0), DRB1*1201 (OR, 2.62; 1.38–4.95), and DRB1*1501 (OR, 1.84; 1.22–2.77). When analyzed by race, DRB1 alleles continued to be differentially expressed between cases and control subjects, but in addition, DRP1 alleles were also observed to be differentially expressed in African Americans. HLA DPB1*0101 was strongly associated with sarcoidosis in African Americans (OR, 2.20; 1.38–3.53). Only one allele was differentially associated with sarcoidosis in whites and African Americans (HLA DRB1*1501), suggesting that some of the difference in susceptibility to sarcoidosis between whites and African Americans depends on the differential expression of the HLA alleles. Polymorphisms of the IL-1ß, tumor necrosis factor (TNF)-
, and the immunoglobulin genes Km and Gm were also investigated, but no association with sarcoidosis was observed (15). Thus, the fifth lesson learned from ACCESS was that genetic associations with sarcoidosis were observed and appeared to be stronger predictors of sarcoidosis than environmental factors.
ACCESS also investigated whether environmental exposures and genetic factors were associated with particular phenotypes of the disease. Whites with agricultural dust exposure and African Americans exposed to wood-burning stoves were less likely to have clinically evident extrapulmonary disease in sarcoidosis (16). Km was associated with African Americans without erythema nodosum (15), DRB1*0401 was associated with eye involvement, DRB3 was associated with bone marrow involvement in African Americans, and DPB1*0101 was associated with hypercalcemia in whites (14). Thus, the sixth lesson learned from ACCESS was that clinical phenotypes may be associated with specific environmental exposures and genetic risk. However, clinical phenotypes are difficult to define and the phenotype of resolved versus persistent sarcoidosis was not able to be determined because of the lack of follow-up on most of the subjects.
Finally, gene–environment interactions in the predisposition to sarcoidosis were investigated in the 476 cases and control subjects that had HLA typing (manuscript in preparation). Because of the lack of power to detect interactions in low-frequency events, only the exposures or genes that were present in 10% of the control population and associated with an OR of greater than 1.0 were investigated. Only four exposures and four HLA alleles met this criterion. A positive interaction was observed between HLA DRB1*1101 (present in 13% of control subjects) and occupational insecticide exposure (present in 15% of control subjects). The combined relative risk was 5.82 with a p = 0.074 (Table 3). In addition, the combination of HLA DRB1*1101 and occupational insecticide exposure was associated with hypercalcemia and cardiac sarcoidosis. Thus, the seventh lesson learned from ACCESS was that gene–environment interactions may be a useful approach in sarcoidosis but large populations or well-defined phenotypes will be needed.
|
However, DNA on these well-documented cases and control subjects will continue to be a resource to evaluate candidate genes associated with the presentation of sarcoidosis. The recent study investigating the BTNL2 gene is an example of the ongoing utility of this resource. As additional candidate genes are identified, the ACCESS DNA bank will be a useful resource to confirm proposed genetic associations in sarcoidosis.
FOOTNOTES
Conflict of Interest Statement: Neither author has a financial relationship with a commercial entity that has an interest in the subject of this manuscript.
(Received in original form July 2, 2006; accepted in final form April 24, 2007)
REFERENCES
, IL1-ß and immunoglobulin (GM and KM) gene polymorphisms in sarcoidosis. Human Immunology 2002;63:485–491.[CrossRef][Medline]
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
