Although most patients with sarcoidosis have a good prognosis, a significant proportion runs a more severe and prolonged disease course. There is no marker to distinguish these subpopulations of patients, however. To investigate the relationship between HLA haplotype and clinical course, 122 Scandinavian patients with sarcoidosis were genomically typed for HLA-DR, -DQA1 and -DQB1 alleles using PCR amplification with sequence-specific primers. Control subjects were 250 healthy Swedish volunteers. Patients were carefully clinically monitored for up to 10 yr. We found that HLA-DR17(3) was overrepresented among sarcoidosis patients (33%) compared with control subjects (17%, p < 0.001). Ninety-one patients were followed for more than 2 yr and classified into chronic or nonchronic patients, according to disease outcome. Among the 34 patients with a nonchronic form of sarcoidosis, 65% were DR17(3)-positive (p < 10− 5 versus control subjects). On the other hand, DR14(6) and DR15(2) were significantly associated with chronic disease. Even in patients with clinical manifestations that are normally associated with good prognosis, HLA typing enabled a subgrouping into two categories with significantly different clinical courses. Therefore, HLA class II typing is a valuable tool in predicting the outcome of the disease in Scandinavian sarcoidosis patients.
Sarcoidosis is a systemic disease of unknown etiology characterized by the presence of noncaseating granulomas and several immunological abnormalities (1, 2). Although most patients have a favorable prognosis, a significant proportion of patients develop a more severe and sometimes even fatal course of disease (3, 4). To distinguish this group of patients, many attempts to find a prognostic marker have been made (5-11). However, a reliable marker for prognosis is still lacking.
The human leukocyte antigen (HLA) molecules play a fundamental role in immune responses by presenting antigens for T cell recognition (12). In sarcoidosis, CD4+ T cells accumulate at the site of active disease, i.e., in the lung, where they are considered to interact through their T cell receptor (TCR) for antigen with HLA molecules expressed by antigen presenting cells (APC). The finding of certain TCRs expressed by lung T cells (13, 14), in some cases strongly associated with particular HLA haplotypes (15, 16), suggested the presence of a specific sarcoidosis-associated antigen in these patients.
Previous studies of the HLA frequencies in patients with sarcoidosis have usually been limited by small numbers of patients, study groups consisting of mixed ethnic origins, no or only short follow-up times, and/or HLA-typing by serological methods (17-24).
In this study, we investigated the correlation between HLA class II alleles and the clinical outcome in Caucasian patients with sarcoidosis. The patients were typed for HLA-DR broad specificities, in selected cases with DRB1 subtypes, and DQA1 and DQB1 alleles using PCR amplification with sequence-specific primers (PCR-SSP). Clinical activity and involvement of the lung were evaluated by symptoms, chest X-ray, and lung function tests and monitored for up to 10 yr. We found strong correlations between HLA-DR17(3) and good prognosis, and between DR15(2) and DR14(6) and a chronic form of the disease, and we therefore argue that HLA typing should be performed routinely in the clinical evaluation of Scandinavian patients with sarcoidosis.
All patients (n = 122) with sarcoidosis, biopsy proven and/or with typical clinical features that attended the Department of Respiratory Medicine at Karolinska Hospital, Stockholm, Sweden, were tested for their HLA class II alleles. All patients were Caucasians and born in the Nordic countries (Sweden, Norway, Denmark, or Finland). They were 45 women and 77 men, median age at the time of diagnosis was 36.0 yr (interquartile ranges [iqr] 30.0–45.5). Disease duration and activity were assessed by looking for progression on chest X-ray within the last 6 mo and/or some of the following symptoms: irritating cough, periods of low grade fever, dyspnea on effort, tiredness, and recent loss of weight. Patients who recovered within 2 yr were classified as nonchronic, whereas those with disease activity beyond that time point were regarded as having chronic disease (25). Thirty-four of the patients had an acute onset of the disease, whereas 75 had a more insidious illness, and 13 had the disease discovered by chance. Twenty-five patients presented with classic Löfgren's syndrome, i.e., acute onset with fever, ankle arthritis/erythema nodosum, and bilateral hilar lymphoma.
At the time of diagnosis, the chest radiographic stages were: stage 0, normal radiograph (n = 6); stage I, bilateral lymphadenopathy (n = 53); stage II, bilateral lymphadenopathy with parenchymal infiltration (n = 38); stage III, solely parenchymal infiltration (n = 25).
Ninety-one patients (38 women, 53 men; median age 36.0 yr, iqr 31.0–43.0) were followed for 2 yr or more. Fifty-seven of these were regarded as having a chronic disease and 34 as running a nonchronic course. Forty-four patients were followed for 5 yr and 26 for 10 yr. Twenty-eight of 34 patients with an acute onset of disease, and 23 of 25 patients with Löfgren's syndrome were followed for 2 yr or more.
After an initial period of 6 mo expectancy without any steroid treatment, all of the chronic patients were treated with oral and/or inhaled corticosteroids periodically or continuously. Eight of the DR17(3)- positive patients were periodically on cortisone treatment. The control population consisted of 250 healthy Caucasians born in the Nordic countries.
Vital capacity (VC) was determined with a body plethysmography. FEV1 was measured with a Sensor-Medics 2400 spirometer (Sensormedics Ltd., BV Bitthoven, The Netherlands) connected to an IBM computer. Reference values were obtained from established regression equations (26, 27).
HLA-DR specificities, DRB1, DQA1, and DQB1 alleles were determined by PCR-SSP (28-30). DNA was extracted from peripheral blood leukocytes by the salting-out technique. PCR reaction mixture components, cycling parameters, and the detection of PCR products by agarose gel electrophoresis are detailed in references (28) and (31).
Data were analyzed by the chi-square test or in the case of small numbers by Fisher's Exact Test. When looking for secondary positive or negative associations, the gene frequencies in patients and control subjects were compared after subtraction of the first identified associated allele(s), the relative predispositional effects method, not to conceal positive associations with more than one genetic marker or to overrate decreased antigen frequencies as negative associations (32). The p values of previously unreported associations were multiplied with the number of detected specificities/alleles of the respective locus. Relative risks (RR) were calculated according to Woolf (33). Because the observed values were not normally distributed, medians and interquartile ranges are given.
At disease onset, we found significantly increased frequencies of the HLA-DR17(3) specificity (33%) in sarcoidosis patients compared with healthy control subjects (17%, p < 0.01; Table 1).
|DR Allele*||All Patients (n = 122)||Chronic (n = 57)†||Nonchronic (n = 34)†||Control Subjects (n = 250)|
Patients who recovered within 2 yr were classified as nonchronic, whereas those with disease activity beyond that time point were regarded as having chronic disease. In the nonchronic group, there was a marked overrepresentation of DR17(3)-positive patients (65% compared with 17% in control subjects, p < 10−5; Table 1). Patients developing chronic disease had, on the other hand, significantly increased frequencies of the DR14(6) (18% compared with 6% in control subjects, p < 0.005) and DR15(2) alleles (60% compared with 30% in control subjects, p < 0.001) (Table 1).
Among patients that still had an active disease at the 5 yr follow-up (n = 44), only 9% were DR17(3)-positive, whereas 18% were DR14(6)-positive and 66% were DR15(2)-positive. At the 10 yr follow-up (n = 26), none was DR17(3)-positive, 27% were DR14(6)-positive, and 62% were DR15(2)-positive (Figure 1).
Of all DR17(3)-positive patients, 76% recovered within 2 yr. As shown in Figure 2, a similar figure was found for the complete group of patients with an acute onset (75%), while among all patients with Löfgren's syndrome 91% recovered within 2 yr. DR17(3)-positive and DR17(3)-negative patients in both groups had dramatically different prognoses; among patients with acute onset, 95% (19 of 20) of DR17(3)-positive patients, but only 25% (two of eight) of DR17(3)-negative patients, recovered within 2 yr (p < 0.01). For patients with Löfgren's syndrome, 100% (19 of 19) of DR17(3)-positive and 50% (two of four) of DR17(3)-negative patients recovered within 2 yr (Figure 2).
HLA-DR17(3) was overrepresented in patients with an acute onset of the disease, including those with Löfgren's syndrome. Twenty-two of 34 (65%) patients with acute symptoms and 20 of 25 (80%) of patients with Löfgren's syndrome were HLA-DR17(3)-positive.
Most DR17(3)-positive patients (70%) presented with X-ray stage I, while patients with the DR15(2) or DR14(6) alleles mainly presented with the more advanced radiographic stages II or III (p < 0.001 for both patient groups compared with DR17(3)-positive patients) (Figure 3).
The lung function, measured as VC and FEV1, was followed in 44 patients for 5 yr and in 26 patients for 10 yr. HLA-DR17(3)-positive patients had normal values at diagnosis and at 5 yr follow-up. DR14(6)-positive patients had already at diagnosis significantly decreased VC and FEV1 values compared with DR17(3)-positive patients (p < 0.01), but their values did not significantly decrease during follow-up. In contrast, while VC and FEV1 values were normal at diagnosis in the DR15(2)- positive patient group, they significantly decreased at 5 yr (p < 0.01) and 10 yr (p < 0.001) follow-up (Figure 4).
Subtyping was performed for DR17(3), DR15(2), and DR14(6), and identical subtypes were found in patients and control subjects, i.e., in both groups, 100% of DR17(3) were DRB1*0301, 100% of DR15(2) were DRB1*1501 and 100% of DR14(6) were DRB1*1401.
The HLA-DQA1*0501 allele was significantly overrepresented in the nonchronic group of patients (79% compared with 33% in control subjects, p < 0.005) (Table 2). The HLA- DQB1*0201/0202 allele was significantly overrepresented in the total group of patients (43%, p < 0.01) and even more strongly associated with the nonchronic group (74%, p < 0.001) compared with control subjects, (23%) (Table 3). The chronic group of patients was positively associated with the HLA-DQB1*0503 and HLA-DQB1*0602 alleles (p < 0.05 compared with controls for both) (Table 3).
|DQA1 Allele*||All Patients (n = 79)†||Chronic (n = 38)†||Nonchronic (n = 19)†||Control Subjects (n = 250)|
|DQB1 Allele*||All Patients (n = 79)†||Chronic (n = 38)†||Nonchronic (n = 19)†||Control Subjects (n = 250)|
In this study, we found a significantly increased risk for individuals in a Caucasian population with the HLA-DR17(3) specificity to develop sarcoidosis. HLA-DR17(3)-positive patients significantly more often had acute onset, chest X-ray stage I, and normal lung function tests. Importantly, the vast majority of DR17(3)-positive patients (76%) recovered completely within 2 yr. The DR17(3)-positive patient with the most prolonged disease duration (9 yr) proved to also have the chronic-associated HLA-DR15(2) specificity, suggesting why a small portion of DR17(3)-positive patients still may suffer an extended disease course.
We also found a positive association of DR15(2) and DR14(6) and chronic disease. These patients presented with more advanced chest radiographic stages, perhaps as a result of their more insidious disease onset, and they had impaired lung functions compared with DR17(3)-positive patients. In general, only approximately two of 10 DR15(2)-positive and one of 10 DR14(6)-positive patients, respectively, recovered within 2 yr. The only DR14(6)-positive patient that recovered within 2 yr proved to be also DR17(3)-positive. Clearly, the HLA class II type has a profound influence on the disease course in Scandinavian sarcoidosis patients.
Acute onset of sarcoidosis and Löfgren's syndrome have previously been associated with good prognosis and also with HLA-DR3 (17), although patients within this group also may run a chronic course (34). In our study, whereas patients with an acute onset in general had a good prognosis, HLA typing enabled a subgrouping of these patients into two categories with significantly different clinical courses. Thus, DR17(3)- positive patients with acute onset had an extremely good prognosis, since close to 100% of these patients recovered within two years. In contrast, only one of four DR17(3)-negative patients with an acute onset recovered within that time period. A similar tendency was found when using DR17(3) to subgroup patients with Löfgren's syndrome. Therefore, even in patients with clinical manifestations that are normally associated with good prognosis, HLA typing substantially adds information and enables the clinician to predict the outcome of the disease with an extremely high accuracy.
Subtyping of DR17(3), DR15(2), and DR14(6) specificities showed identical results in patients and controls, as expected in the homogeneous Scandinavian population. Interestingly the DRB1*1401 allele has previously been associated with sarcoidosis in Japan, although patients were not followed up in that study (23). Our finding of strong associations of DQ-A1*0501 and DQB1*0201/0202 alleles with a nonchronic form of the disease was anticipated, because of their strong linkage to the DR17(3) haplotype. Similarly, associations of DQ-B1*0503 and DQB1*0602 with chronic patients were expected since they are strongly linked to DR14(6) and DR15(2), respectively. In Japan, the DQA1*0501 allele was also previously described to be associated with sarcoidosis, likely because of its linkage to the sarcoidosis-associated DR52 haplotypes (23).
In Caucasians, HLA-DR3 has been suggested to be associated with patients having arthritis, which usually confers a good outcome of the disease (17, 18), as well as with a mild course and short disease duration (18, 22, 24). Other studies failed to show such an association (35). Most of these studies have used serological HLA typing, which may give as much as 25% incorrect results (36). Moreover some studies have analyzed small patient groups (17, 18), populations with a mixed ethnic background (24) or did not follow the patients clinically (23). In our study, genomic HLA-typing was performed on a large group of ethnically homogeneous patients and controls, and a majority of patients were carefully monitored clinically, some for as long as 10 yr. We could confirm an overrepresentation of DR17(3) in patients with an acute onset of the disease including those with Löfgren's syndrome, although approximately one-third of these patients in fact were DR17(3)-negative. In the distinct ethnic group of Japanese sarcoidosis patients, HLA-DR52 was shown to be overrepresented (19– 21), and one study indicated a correlation with HLA-DR5 and good prognosis (20). Still comparisons with our study on Caucasian patients is difficult because of the difference in HLA background, e.g., DR17(3) is virtually missing in Japanese individuals. This ethnical difference is also reflected in a different TCR usage by lung-accumulated T cells of Japanese compared to Scandinavian sarcoidosis patients (37).
The strong correlation between DR17(3) and good prognosis, and between DR15(2) and DR14(6) and a chronic form of the disease, propose that antigen presentation may be of vital importance in sarcoidosis. A putative specific sarcoidosis- associated antigen, suggested by lung-accumulated T cells expressing certain TCRs in sarcoidosis patients (13, 14), may be processed and presented differently in individuals depending on their HLA-type. Among patients with a chronic form of the disease, DR15(2) and DR14(6) were significantly associated, but also DR11(5), DR7, and DR8 were overrepresented. All of these haplotypes were noted to be linked to HLA-DQβ chains with an aspartic acid at position 57, i.e., at the edge of the antigen presenting groove in the DQ molecule. Thus, antigen presentation by these HLA-DQ haplotypes may modulate the immune response, eventually resulting in a more chronic form of the disease. DR17(3)-positive patients on the other hand, with a good prognosis, may through antigen presentation via DR17(3) molecules initiate an efficient immune response that eradicates any offending antigen. Alternatively, HLA associations may merely reflect overrepresentations of other, non-HLA molecules that contribute to the disease. Recently, one such class II located non-HLA molecule (low molecular weight polypeptides, LMP) was investigated but not found to be associated with sarcoidosis in Japan (38).
On the basis of our results in this study we are convinced that HLA typing is a useful tool in helping to predict the outcome of sarcoidosis. The patients with HLA specificities associated with chronic disease, i.e., DR15(2) and DR14(6), should be followed with frequency and carefulness and may require early treatment to prevent impairment of their lung functions. DR17(3)-positive patients may on the other hand not require such a cautious clinical monitoring.
The authors express their gratitude to Ms. Margitha Dahl for valuable contributions.
This study was supported by the Swedish Heart-Lung Foundation, the Swedish Medical Research Council (project Nos. 10393 and 11953), the Children's Cancer Foundation of Sweden, and by donations of Mr. and Mrs. L. A. Koolberg, Vancouver, B.C.
|1.||Crystal R., Roberts W., Hunninghake G., Gadek J., Fulmer J., Line B.Pulmonary sarcoidosis: a disease characterized and perpetuated by activated lung T lymphocytes. Ann. Intern. Med.9419817394|
|2.||Campbell P., Tolson T.Immunoregulation in sarcoidosis. Am. Rev. Respir. Dis.134198610291035|
|3.||Hillerdal G., Nöu E., Osterman K., Schmekel B.Sarcoidosis: epidemiology and prognosis. Am. Rev. Respir. Dis.13019842932|
|4.||Milman N., Selroos O.Pulmonary sarcoidosis in the Nordic countries 1950–1982. II. Course and prognosis. Sarcoidosis71990113118|
|5.||Keogh B., Hunninghake G., Line B., Crystal R.The alveolitis of pulmonary sarcoidosis. Am. Rev. Respir. Dis.1281983256265|
|6.||Selroos O.Biochemical markers in sarcoidosis. CRC Critic Rev. Clin. Lab. Sci.241986185216|
|7.||Ward K., O'Connor C., Odlum C., Fitzgerald M.Prognostic value of bronchoalveolar lavage in sarcoidosis: the critical influence of disease presentation. Thorax441989612|
|8.||Foley N., Coral A., Ung K., Hudspith B., James D., Johnson N.Bronchoalveolar lavage cell counts as a predictor of short term outcome in pulmonary sarcoidosis. Thorax441989732738|
|9.||Verstraeten A., Demedts M., Verwilghen J., van den Eeckhout A., Marien G., Lacquet L., Ceuppens J.Predictive value of bronchoalveolar lavage in pulmonary sarcoidosis. Chest981990560567|
|10.||Blaschke E., Eklund A., Hernbrand R.Extracellular matrix components in bronchoalveolar lavage fluid in sarcoidosis and their relationship to signs of alveolitis. Am. Rev. Respir. Dis.141199010201025|
|11.||Bjermer L., Eklund A., Blaschke E.Bronchoalveolar lavage fibronectin in patients with sarcoidosis: correlation to hyaluronan and disease activity. Eur. Respir. J.41991965971|
|12.||Harding C.Cellular and molecular aspects of antigen precessing and the function of class II MHC molecules. Am. J. Respir. Cell Mol. Biol.81993461467|
|13.||Moller D., Konishi K., Kirby M., Balbi B., Crystal R.Bias toward use of a specific T cell receptor β chain variable region in a subgroup of individuals with sarcoidosis. J. Clin. Invest.82198811831191|
|14.||Grunewald J., Janson C. H., Eklund A., Öhrn M., Olerup O., Persson U., Wigzell H.Restricted Vα 2.3 gene usage by CD4+ T lymphocytes in bronchoalveolar lavage (BAL) fluid from sarcoidosis patients correlates with HLA-DR3. Eur. J. Immunol.221992129135|
|15.||Grunewald J., Olerup O., Persson U., Öhrn M., Wigzell H., Eklund A.T cell receptor variable region gene usage by CD4+ and CD8+ T cells in bronchoalveolar lavage fluid and peripheral blood of sarcoidosis patients. Proc. Natl. Acad. Sci. U.S.A.91199449654969|
|16.||Grunewald J., Hultman T., Bucht A., Eklund A., Wigzell H.Restricted usage of T cell receptor Vα/Jα gene segments with different nucleotide but identical amino acid sequences in HLA-DR3+ sarcoidosis patients. Mol. Med.11995287296|
|17.||Hedfors E., Lindström F.HLA-B8/DR3 in sarcoidosis. Tissue Antigens221983200203|
|18.||Gardner J., Kennedy H., Hamblin A., Jones E.HLA associations in sarcoidosis: a study of two ethnic groups. Thorax3919841922|
|19.||Kunikane H., Abe S., Tsuneta Y., Nakayama T., Tajima Y., Misonou J., Wakisaka A., Aizawa M., Kawakami Y.Role of HLA-DR antigens in Japanese patients with sarcoidosis. Am. Rev. Respir. Dis.1351987688691|
|20.||Abe S., Yamaguchi E., Makimura S., Okazaki N., Kunikane H., Kawakami Y.Association of HLA-DR with sarcoidosis: correlation with clinical course. Chest921987488490|
|21.||Ina Y., Takada K., Yamamoto M., Morishita M., Senda Y., Torii Y.HLA and sarcoidosis in the Japanese. Chest95198912571261|
|22.||Odum N., Milman N., Jakobsen B., Georgsen J., Svejgaard A.HLA Class II (DR, DQ, DP) in patients with sarcoidosis: evidence of an increased frequency of DRw6. Exp. Clin. Immunogenet.81991227232|
|23.||Ishihara M., Ohno S., Ishida T., Ando H., Naruse T., Nose Y., Inoko H.Molecular genetic studies of HLA class II alleles in sarcoidosis. Tissue Antigens431994238241|
|24.||Martinetti M., Tinell C., Kolek V., Cuccia M., Salvaneschi L., Pasturenzi L., Semenzato G., Cipriani A., Bartova A., Luisetti M.“The sarcoidosis map”: a joint survey of clinical and immunogenetic findings in two European countries. Am. J. Respir. Crit. Care Med.1521995557564|
|25.||Neville E., Walker A., James D.Prognostic factors predicting the outcome of sarcoidosis: an analysis of 818 patients. Q. J. Med.2081983525533|
|26.||Quanjer P.Standardized lung function testing. Bull. Eur. Physiopath. Respir.1919834551|
|27.||Cotes, J. 1979. Lung function assessment and application in medicine. Blackwell, Oxford.|
|28.||Olerup O., Zetterquist H.HLA-DR typing by PCR amplification with sequence specific primers (PCR-SSP) in 2 hours: an alternative to serological DR typing in clinical practice including donor-recipient matching in cadaveric transplantation. Tissue Antigens391992225235|
|29.||Olerup O., Aldener A., Fogdell A.HLA-DQB1 and -DQA1 typing by PCR amplification with sequence-specific primers (PCR-SSP) in 2 hours. Tissue Antigens411993119134|
|30.||Zetterquist H., Olerup O.Identification of the HLA-DRB1*04, -DRB1*07 and -DRB1*09 alleles by PCR amplification with sequence-specific primers (PCR-SSP) in 2 hours. Human Immunol.3419926474|
|31.||Aldener-Cannavá A., Olerup O.HLA-DPA1 typing by PCR amplification with sequence-specific primers (PCR-SSP) and distribution of DPA1 alleles in Caucasian, African and Oriental populations. Tissue Antigens471996153160|
|32.||Payami H., Joe S., Farid N., Stenszky V., Chan S., Yeo P., Cheah J., Thomson G.Relative predispositional effects (RPSs) or marker alleles with disease: HLA-DR alleles and Graves disease. Am. J. Hum. Genet.451989541546|
|33.||Woolf B.On estimating the relation between blood groups and disease. Ann. Hum. Genet.191955251253|
|34.||Gran J., Bøhmer E.Acute sarcoid arthritis: a favorable outcome? Scand. J. Rheumatol.2519967073|
|35.||Lenhart K., Kolek V., Bartova A.HLA antigens associated with sarcoidosis. Disease Markers819902329|
|36.||Olerup O., Hillert J.HLA-class II associated genetic susceptibility in multiple sclerosis: a critical evaluation. Tissue Antigens381991115|
|37.||Grunewald J., Shigematsu M., Nagai S., Mikuniya T., Wigzell H., Izumi T., Eklund A.T cell receptor V gene use in HLA-typed Japanese patients with pulmonary sarcoidosis. Am. J. Respir. Crit. Care Med.1511995151156|
|38.||Ishihara M., Ohno S., Mizuki N., Yamagata N., Ishida T., Naruse T., Ando A., Inoko H.LMP7 polymorphism in Japanese patients with sarcoidosis and Behçet's disease. Human Immunol.511996103105|