Sensitization to inorganic but not to organic antigens, was associated with sarcoidosis in a cohort of over 100 Dutch patients. Metal and silica sensitized patients were at greater risk of developing fibrosis 5 years after diagnosis.
Keywords: metals, P. acnes, sarcoidosis, silica, triggers
Summary
Organic and inorganic antigens were studied simultaneously in the same cohort of sarcoidosis patients to investigate whether correlations between clinical characteristics and immunological sensitization could reveal new phenotypes. Sensitization to antigens of mycobacteria, Propionibacterium acnes catalase and vimentin was investigated in 201 sarcoidosis and 51 obstructive sleep apnoea patients, serving as control group. Sensitization to aluminium, beryllium, silica and zirconium was also studied in 105 of the sarcoidosis patients and in 24 of the controls. A significantly higher percentage of sarcoidosis patients (27·6%) than controls (4·2%) had an immunological response to metals or silica (P = 0·014). A higher percentage of these sarcoidosis patients showed fibrosis on chest X‐ray 5 years after the diagnosis (69·2 versus 30·3%, P = 0·016). No significant differences in mycobacterial or vimentin enzyme‐linked immunospot (ELISPOT) assay results were observed between sarcoidosis and control patients. A significantly lower percentage of sarcoidosis patients (3·5%) than control patients (15·7%) had a positive ELISPOT for P. acnes catalase (P = 0·003). However, sarcoidosis patients sensitized to P. acnes catalase were more likely to have skin involvement, while sarcoidosis patients sensitized to mycobacterial antigens were more likely to have cardiac involvement. Our study suggests a more prominent role for inorganic triggers in sarcoidosis pathogenesis than previously thought. Immunological sensitization to inorganic antigens was associated with development of fibrotic sarcoidosis. No association was found between sensitization to bacterial antigens or vimentin and sarcoidosis in Dutch patients. However, our data suggest that trigger‐related phenotypes can exist in the heterogeneous population of sarcoidosis patients.
Introduction
Sarcoidosis is a systemic disease characterized by formation of non‐caseating granulomas, mainly affecting the lungs, skin, eyes and lymph nodes [1]. The course of the disease is unpredictable, meaning that the disease can spontaneously resolve, while a chronic disease course can also develop [2]. Currently, no curative treatment exists for sarcoidosis patients. Patients with symptomatic organ involvement or at risk of permanent damage are treated with immunosuppressive drugs. This type of treatment is associated with several side effects, and approximately 10% of patients still develop progressive disease while receiving immunosuppressives [2].
Considerable research effort has focused on identification of triggers of sarcoidosis, including antigens of mycobacteria [3, 4], Propionibacterium acnes [5, 6], vimentin [7, 8], metals and silica [9, 10]. Although all these antigens have shown potential, results demonstrate that sarcoidosis probably has multiple triggers. Importantly, none of the above organic and inorganic antigens have been studied simultaneously in one group of sarcoidosis patients. Identification of triggers in sarcoidosis will allow personalized treatment options, which may include anti‐bacterial drugs. In addition, increased knowledge concerning sensitization to metals and silica can lead to identification of sarcoidosis patients for whom exposure prevention may be more beneficial than increasing immune suppression.
The aim of this study was to examine possible triggers of sarcoidosis in Dutch patients. Correlations between clinical characteristics and immunological response to microbial antigens or metals and silica may result in the identification of new clinical phenotypes. Identification of these subgroups can lead the way to clinical trials in the future, to explore more accurate treatment options.
Materials and methods
Study subjects
Biopsy‐proven sarcoidosis patients or Löfgren syndrome patients, seen for the first time at the Interstitial Lung Diseases (ILD) outpatient clinic of the St Antonius Hospital (Nieuwegein, the Netherlands) from 1 May 2016, were asked to participate until at least 200 patients had been included. The diagnosis of sarcoidosis had been established according to the criteria of the American Thoracic Society/European Respiratory Society [11]. Patients with obstructive sleep apnoea (OSA) are also seen at the pulmonary outpatient clinic, and this disorder has not been associated with the antigens tested in this study [12]. OSA patients were therefore included as a control group. The study was approved by the Medical research Ethics Committees United (MEC‐U) of the St Antonius Hospital (R14.023), and written informed consent was obtained from all participants.
IFN‐γ enzyme‐linked immunospot assay (ELISPOT)
A blood sample was drawn from all participants. Peripheral blood mononuclear cells (PBMCs) were collected and stored at −150°C until time of analysis. Interferon (IFN)‐γ ELISPOT assays were performed following the manufacturer’s instructions (Mabtech, Nacka Strand, Sweden). Briefly, 2 × 105 PBMCs were seeded per well, together with 33·33 µg/ml early secretory antigenic target (ESAT)‐6 peptides, 33·33 µg/ml Mycobacterium tuberculosis catalase‐peroxidase (mKatG) peptides, 10 µg/ml MKatG protein, 10 μg/ml P. acnes catalase or 20 µg/ml vimentin in duplicate. Responses were calculated by subtracting the number of spot‐forming‐cells (SFCs) of the negative control wells from that of the antigen‐specific wells. Following Drake et al. [13], positive responses were defined as at least 50 SFC per 106 PBMCs, at least three times higher than the negative control well. Details of PBMC isolation, the antigens used and ELISPOT are provided in the Supporting information.
Clinical tuberculosis IGRAs
Medical records of sarcoidosis patients were searched for data of IFN‐γ release assays (IGRAs) (TB ELISPOT, or Quantiferon tests). When data of those tests were not available, medical records were searched for results of tuberculin skin tests (TSTs).
Cytokine measurements
Supernatant of ELISPOT plates was collected and stored at −80°C until time of analysis. Concentrations of tumour necrosis factor (TNF)‐α, IL‐2, IL‐17A and IL‐1β were measured using Human Magnetic Luminex Assays (Biotechne, Minneapolis, MN, USA), following the manufacturer’s instructions. Values beyond the range of the standard curve were given the lowest or highest value that the standard curve reached. Baseline cytokine concentrations were subtracted from cytokine concentrations of the antigen‐stimulated wells. Negative values were set to zero.
Lymphocyte proliferation tests to determine metal and silica sensitization
All sarcoidosis and control patients were asked to provide a second blood sample for lymphocyte proliferation tests (LPTs) on beryllium, aluminium, zirconium and silica. Blood was drawn in citrate tubes and sent within 24 h to Prohealth Medical (Nederweert, the Netherlands), where the memory lymphocyte immunostimulation assay (MELISA®) LPT was performed, as described in the Supporting information. Following MELISA® guidelines, a stimulation index (SI) > 2·0 was considered a weak positive test result and an SI > 3·0 a positive test result.
HLA‐DRB1 genotyping
Presence of HLA‐DRB1*0301 and HLA‐DRB1*1501 alleles was determined for sarcoidosis patients from whom residual DNA was available. Tag single nucleotide polymorphisms (SNPs) rs2040410A and rs3135388A were used to capture HLA‐DRB1*0301 and HLA‐DRB1*1501, respectively, using the methods described by Karakaya et al. [14].
Identification of trigger‐related phenotypes
Scadding stage at diagnosis and follow‐up, organ involvement and the use of immunosuppressive medication were collected from medical records. In addition, the clinical outcome status (COS) [15] was determined 2 and 5 years after diagnosis and classified into two groups: group 1: resolved, minimal or persistent disease without treatment (COS 1–6) and group 2: persistent disease with need for treatment (COS 7–9). Clinical characteristics and COS were related to ELISPOT and LPT results to explore trigger‐related phenotypes.
Statistical analysis
Study data were collected and managed using Research Electronic Data Capture (REDCap) tools hosted at St Antonius Hospital [16]. Data were analysed using IBM spss statistics version 24. An unpaired t‐test was used to compare numerical data. Non‐parametric tests were used for non‐normally distributed data (Mann–Whitney U‐test). Categorical data were compared using the χ2 test. If expected cell frequencies were below 5, Fisher’s exact test was used for categorical data up to two categories. P‐values < 0·05 were considered significant.
Results
Patient characteristics
A total of 205 sarcoidosis patients and 51 control patients were included into the study (Table 1). Control patients were older at inclusion, and a higher proportion of control patients were ever smokers compared to the sarcoidosis patients.
Table 1.
Demographics of sarcoidosis and control patients
| Sarcoidosis (n = 205) | Controls (n = 51) | P‐value * | |
|---|---|---|---|
| Sex (male/female) (%) | 53·2/46·8 | 62·7/37·3 | 0.219 |
| Age at diagnosis (years) | 43·74 ± 12·48 | n.a. | |
| Age at inclusion (years) | 48·93 ± 12·21 | 55·45 ± 13·71 | 0.001 |
| Ethnicity (W/NW) (%) | 89·2/10·8 | 92·2/7·8 | 0.535 |
| Smoking status (never/former or current) (%) | 51·2/48·8 | 29·4/70·6 | 0.005 |
| Scadding stage at inclusion (0/I/II/III/IV) (%) | 19·6/29·4/28·4/5·9/16·7 | n.a. | |
| Extra pulmonary involvement (%) | 59.5 | n.a. |
Age at diagnosis and age at inclusion are expressed as mean ± standard deviation.
W = white; NW = non‐white; 0 = normal chest radiograph; I = bilateral hilar lymphadenopathy (BHL); II = BHL with pulmonary infiltrates; III = pulmonary infiltrates without BHL; IV = fibrosis; n.a. = not applicable.
P‐value regarding sex, ethnicity and smoking status was calculated using Pearson’s χ2 test; an independent‐samples t‐test was used to calculate a P‐value for differences regarding age at inclusion.
ELISPOT results
An ELISPOT assay was performed on PBMCs of 201 sarcoidosis patients and 51 controls. Ninety‐one sarcoidosis patients (45·3%) were using immunosuppressive medication at the time of blood collection. No differences in number of spots upon stimulation with anti‐CD3 antibody were observed between sarcoidosis patients with and without immunosuppressive medication, while more controls had a low response (Supporting information, Figs. S1 and S2). Only one control and two sarcoidosis patients had a positive ELISPOT for ESAT‐6 and neither controls nor patients had a positive ELISPOT for MKatG peptide, although one sarcoidosis patient had a positive ELISPOT for MKatG protein. Eight controls (15·7%) compared to seven sarcoidosis patients (3·5%) had a positive ELISPOT for P. acnes catalase (P = 0·003). A vimentin‐positive ELISPOT was found for nine controls (17·6%) and 18 sarcoidosis patients (9·0%) (Fig. 1).
Fig. 1.
Percentage of positive interferon (IFN)‐γ enzyme‐linked immunospot (ELISPOT) tests of sarcoidosis and control patients. A positive response was defined as at least 50 spot‐forming cells per million cells after subtraction of the negative control, being at least three times higher than the negative control. A significantly higher percentage of the control patients (P = 0·003) had a positive ELISPOT for Propionibacterium acnes catalase. ESAT‐6 = peptides 14 and 15 of early secretory antigenic target produced by Mycobacterium tuberculosis; Mycobacterium tuberculosis catalase‐peroxidase (mKatG) pep. = peptide 31 of M. tuberculosis catalase‐peroxidase; MKatG prot. = whole M. tuberculosis catalase‐peroxidase protein; P. acnes Cat. = Propionibacterium acnes catalase‐peroxidase.
To determine sensitization to mycobacterial antigens, the results of our ELISPOT and clinical data of diagnostic IGRAs or TSTs were combined. A total of five sarcoidosis patients had a positive test for mycobacterial antigens.
Antigen‐specific cytokine measurements
Median baseline concentrations of IL‐1β, IL‐17, IL‐2 and TNF‐α were significantly higher in sarcoidosis patients (Supporting information, Table S1). Median concentrations of IL‐1β were significantly higher in sarcoidosis patients after PBMC stimulation with ESAT‐6 (Supporting information, Fig. S3a, P = 0·024), P. acnes catalase (Supporting information, Fig. S3c, P = 0·001) and vimentin (Supporting information, Fig. S3d, P = 0·012).
Metal and silica LPT results
Of all patients, 24 controls and 105 sarcoidosis patients agreed to provide a second blood sample for an LPT. Fifty‐nine sarcoidosis patients (56·2%) were using immunosuppressive medication at the time of LPT. No differences in the SI upon stimulation with pokeweed (used as positive control) were observed between sarcoidosis patients with and without immunosuppressive medication, or between sarcoidosis patients and controls (Supporting information, Fig. S4).
Five sarcoidosis and no control patients had a positive LPT. A significantly higher percentage of sarcoidosis patients (27·6 compared to 4·17% controls, P = 0·014) had a weakly positive LPT (Fig. 2 and Table 2). Four sarcoidosis patients had a weakly positive LPT for two antigens (Supporting information, Table S2). The single weakly positive control patient had a weakly positive LPT for all the antigens tested. Occupations at time of diagnosis of the 29 sensitized sarcoidosis patients and one sensitized control patient are provided in Supporting information, Table S5.
Fig. 2.
Percentage of positive and weakly positive lymphocyte proliferation tests of sarcoidosis and control patients. (a) Percentages of sarcoidosis patients and controls with a weakly positive LPT. A stimulation index > 2·0 was defined as a weakly positive result. A significantly higher percentage of sarcoidosis patients had a weakly positive LPT (P = 0·014). Four sarcoidosis patients had a weakly positive LPT for two antigens. The only weakly positive control patient had a weakly positive LPT for all the antigens tested. (b) Percentages of sarcoidosis patients and controls with positive LPT, which was defined as a stimulation index > 3·0. Alu = aluminium; Be = beryllium; Sil = silica; Zir = zirconium; LPT = lymphocyte proliferation test, SI: stimulation index.
Table 2.
Percentage of positive and weakly positive lymphocyte proliferation tests of sarcoidosis and control patients
| Controls, n = 24 (%) | Sarcoidosis, n = 105 (%) | P * | |
|---|---|---|---|
| SI > 2.0 | |||
| Alu | 1 (4·17) | 10 (9·52) | 0·688 |
| Be | 1 (4·17) | 8 (7·62) | 1·000 |
| Sil | 1 (4·17) | 12 (11·43) | 0·460 |
| Zir | 1 (4·17) | 3 (2·86) | 0·566 |
| Any | 1 (4·17) | 29 (27·62) | 0·014 |
| SI > 3·0 | |||
| Alu | 0 | 2 (1·90) | 1·000 |
| Be | 0 | 1 (0·95) | 1·000 |
| Sil | 0 | 2 (1·90) | 1·000 |
| Zir | 0 | 0 | n.a. |
| Any | 0 | 5 (4·67) | 0·583 |
A stimulation index > 2.0 was defined as a weakly positive result. A positive LPT was defined as a stimulation index > 3.0
Alu = aluminium, BE = beryllium, Sil = silica; Zir: zirconium; LPT = lymphocyte proliferation test; SI = stimulation index; n.a. = not applicable.
P‐value of any positive LPT SI > 2·0 was calculated using Pearson’s χ2 test; the other P‐values were calculated using Fisher’s exact test. The two silica‐positive patients were also included in a previous study [19].
Trigger‐related phenotypes
Bacterial antigens
One sarcoidosis patient with a positive mycobacterial IGRA also had a positive ELISPOT for vimentin (Supporting information, Table S3), and two sarcoidosis patients with a positive LPT also had a positive vimentin ELISPOT (Supporting information, Table S4). No other overlap between ELISPOT tests or between ELISPOT and LPT was observed. A significantly higher proportion of sarcoidosis patients with a positive response to mycobacterial antigens had cardiac involvement (three of five) compared to sarcoidosis patients without sensitization to mycobacterial antigens (34 of 196, P = 0·044). Although not statistically significant, patients with a positive ELISPOT for P. acnes catalase showed a trend towards more skin involvement (three of seven, compared to 27 of 194 P. acnes catalase‐negative patients, P = 0·069) (Table 3). Furthermore, P. acnes catalase‐positive patients were significantly younger at diagnosis than sarcoidosis patients without a positive P. acnes catalase ELISPOT (median age 27·94 compared to 42·76, P = 0·002).
Table 3.
Organ involvement and ELISPOT results
| Mycobacteria | P. acnes catalase | Vimentin | |||||||
|---|---|---|---|---|---|---|---|---|---|
| Neg. n = 196 | Pos. n = 5 | P‐value | Neg. n = 194 | Pos. n = 7 | P‐value | Neg. n = 183 | Pos. n = 18 | P‐value | |
| Extra pulm. | 116 (59·2) | 3 (60) | 1·00 | 113 (58·2) | 6 (85·7) | 0·244 | 108 (59) | 11 (61·1) | 0·863 |
| Skin | 30 (15·3) | 0 | 1·00 | 27 (13·9) | 3 (42·9) | 0·069 | 27 (14·8) | 3 (16·7) | 0·736 |
| Eyes | 24 (12·2) | 0 | 1·00 | 23 (11·9) | 1 (14·3) | 0·595 | 21 (11·5) | 3 (16·7) | 0·457 |
| Heart | 34 (17·3) | 3 (60) | 0·044 | 36 (18·6) | 1 (14·3) | 1·00 | 33 (18) | 4 (22·2) | 0·749 |
Data are shown in absolute numbers with percentages in brackets. P‐values were calculated using the χ2 test. If expected cell frequencies were below 5, Fisher’s exact test was used.
ELISPOT = enzyme‐linked immunospot.
Vimentin
A higher percentage of Löfgren’s syndrome patients were positive for HLA‐DRB1*0301 than non‐Löfgren patients (50 versus 16·5%, P = 0·006). None of the vimentin‐positive sarcoidosis patients were diagnosed with Löfgren’s syndrome. No difference in the presence of HLA‐DRB1*0301 was observed between vimentin‐positive and ‐negative patients (11·8 and 19·7%, respectively, P = 0·75). A significantly higher percentage of vimentin‐positive patients were positive for HLA‐DRB1*1501 than vimentin‐negative patients (47·1 versus 23·1%, P = 0·041) (Fig. 3).
Fig. 3.
Percentage of vimentin‐positive patients positive for human leucocyte antigen (HLA)‐DRB1*0301 and HLA‐DRB1*1501.
Inorganic antigens
No relationship was observed between organ involvement and sensitization to metals or silica (data not shown), although a higher percentage of the LPT‐positive sarcoidosis patients showed fibrosis on chest X‐ray 5 years after the diagnosis than LPT‐negative sarcoidosis patients (69·2 versus 30·3%, P = 0·016). This difference was not observed 2 years after the diagnosis of sarcoidosis (Table 4). No relations were observed between HLA types and LPT‐positive sarcoidosis patients (Supporting information, Fig. S5).
Table 4.
Fibrosis, COS and LPT results
| 2 years after diagnosis | Silica | Metals | Silica and/or metals | ||||||
|---|---|---|---|---|---|---|---|---|---|
| Neg. n = 81 | Pos. n = 10 | P‐value | Neg. n = 73 | Pos. n = 18 | P‐value | Neg. n = 65 | Pos. n = 26 | P‐value | |
| Fibrosis on chest X‐ray a | 12 (14·8) | 1 (10) | 1·00 | 9 (12·3) | 4 (22·2) | 0·28 | 8 (12·3) | 5 (19·2) | 0·51 |
| Neg. n = 89 | Pos. n = 11 | P‐value | Neg. n = 81 | Pos. n = 19 | P‐value | Neg. n = 72 | Pos. n = 28 | P‐value | |
| Persistent disease with medication b | 47 (52·8) | 3 (27·3) | 0·11 | 41 (50·6) | 9 (47·4) | 0·80 | 38 (52·8) | 12 (42·9) | 0·37 |
| 5 years after diagnosis | Neg. n = 40 | Pos. n = 6 | P‐value | Neg. n = 38 | Pos. n = 8 | P‐value | Neg. n = 33 | Pos. n = 13 | P‐value |
| Fibrosis on chest X‐ray c | 15 (37·5) | 4 (66·7) | 0·21 | 14 (36·8) | 5 (62·5) | 0·25 | 10 (30·3) | 9 (69·2) | 0·016 |
| Neg. n = 40 | Pos. n = 6 | P‐value | Neg. n = 39 | Pos. n = 7 | P‐value | Neg. n = 34 | Pos. n = 12 | P‐value | |
| Persistent disease with medication d | 25 (62·5) | 4 (66·7) | 1·00 | 26 (66·7) | 3 (42·9) | 0·40 | 22 (64·7) | 7 (58·3) | 0·74 |
Data are shown in absolute numbers with percentages in brackets. P‐values were calculated using the χ2 test. If expected cell frequencies were below 5, Fisher’s exact test was used.
Chest X‐rays 2 years after diagnosis were not available from 14 of 105 patients (11 of the 76 patients without any positive LPT test, two of the 12 silica‐positive patients, one of the 19 metal‐positive patients and three of the 29 LPT‐positive patients).
COS 2 years after diagnosis could not be determined for five of 105 patients (four of the 76 patients without any positive LPT, one of the 12 silica‐positive patients and one of the 29 LPT‐positive patients).
Chest X‐rays 5 years after diagnosis were not available from 59 of 105 patients (43 of the 76 patients without any positive LPT test, six of the 12 silica‐positive patients, 11 of the 19 metal‐positive patients and 16 of the 29 LPT‐positive patients).
COS 5 years after diagnosis could not be determined for 59 of 105 patients (42 of the 76 patients without any positive LPT, six of the 12 silica‐positive patients, 12 of the 19 metal‐positive patients and 17 of the 29 LPT‐positive patients).
LPT = lymphocyte proliferation test; COS = clinical outcome status.
Discussion
To the best of our knowledge, this is the first study that simultaneously investigated organic and inorganic antigens in the pathogenesis of sarcoidosis within one cohort of patients. We found immunological sensitization, defined by an LPT SI above 3·0, to inorganic antigens in 5% of sarcoidosis patients. In addition to beryllium, we identified sarcoidosis patients sensitized to aluminium and silica. Even though only observed in a small percentage of patients, these findings imply clinical significance. It makes sense, in our opinion, to carefully reassess work or home exposure in this subgroup of patients. In patients with chronic beryllium disease (CBD), improved lung function and radiographic findings can be observed when exposure is terminated or decreased [17, 18]. It seems plausible to assume that this will also be the case for other inorganic triggers. In a previous paper, we showed that occupational exposure to metal/silica determined by a job exposure matrix (JEM) is not predictive for a positive LPT [19]. A possible explanation may be that the JEMs are designed to be very specific, and will not capture individuals experiencing coincidental high‐risk exposure while carrying out low‐risk jobs. However, even very low occasional exposures can be relevant, as bystander beryllium exposure can lead to CBD [20, 21]. Perhaps a simple occupational history is not detailed enough, and additional (job‐specific) questionnaires will be required to identify occasional/recreational exposures that can be sufficient to initiate sarcoidosis in genetically predisposed individuals. Considering the above‐mentioned explanations, in the current study we decided to test all sarcoidosis patients for metal and silica sensitivity, without determining exposure beforehand.
Regarding the LPT, an SI between 2·0 and 3·0 is considered weakly positive, making interpretation difficult. A weakly positive test does not prove, but rather suggests sensitization. However, when comparing LPT results using an SI > 2·0, we found a remarkable difference between patients and controls, with 27·6% of sarcoidosis patients showing an immunological response compared to only 4·2% of controls. Clinical relevance is further supported by our finding that sensitization for inorganic triggers, suggested by an SI > 2·0, was associated with development of fibrotic sarcoidosis 5 years after diagnosis. An unresolved issue is why some sarcoidosis patients experience spontaneous remission while others develop chronic disease. It has been proposed that difference in antigen clearance could play a role [22]. Patients in whom the antigen cannot be cleared are likely to develop chronic disease. Our results of a higher prevalence of fibrotic disease among the subgroup of sarcoidosis patients sensitized to inorganic antigens fit this hypothesis extremely well.
Interpretation of our findings regarding P. acnes are also quite challenging. P. acnes is a commensal of the human skin but can also be detected in lymph nodes or lung tissue of sarcoidosis, as well as non‐sarcoidosis patients [23]. Unlike T cells specific for self‐antigens, commensal‐specific T cells do not undergo negative selection in the thymus and can be found in healthy individuals [24]. Indeed, in our study a P. acnes catalase‐specific T cell response was present in approximately 15% of controls. However, the percentage of sarcoidosis patients with such a T cell response was significantly lower. Our observations are not in line with previous publications, which described an increased response to P. acnes in Japanese patients with sarcoidosis [25, 26]. Differences in ethnicity and genetic susceptibility might explain why an increased P. acnes immune response is observed in Japanese but not in Dutch sarcoidosis patients. It seems too early, however, to exclude a role for P. acnes in the pathogenesis of sarcoidosis in Dutch patients. First, the lower percentage of P. acnes T cell responses in our sarcoidosis patients could reflect homing of these cells to the lungs. In inflammatory bowel disease it has been demonstrated that circulating microbiota‐reactive CD4‐positive T cell frequencies are decreased, which might indeed reflect their selective recruitment to the inflamed gut [27]. One may, therefore, argue that peripheral blood is not the correct compartment to study P. acnes sensitization in order to clarify its role in sarcoidosis pathogenesis. Accordingly, in our opinion, future sarcoidosis research should focus either on testing sensitization for P. acnes in the alveolar compartment or on detection of P. acnes in biopsy material. Secondly, our results demonstrate that sarcoidosis patients sensitized to P. acnes catalase are significantly younger at the time of diagnosis and are more likely to have skin involvement. While, in theory, it seems reasonable to relate P. acnes to cutaneous sarcoidosis, more research is needed to confirm this possible association. Finally, it could be that P. acnes is a mitogen rather than a specific trigger in sarcoidosis, explaining the relative absence of a specific T cell response. It has been reported that P. acnes has a mitogenic effect on T lymphocytes and consequently promotes an inflammatory response, activates specific T cell subsets and stimulates production of cytokines [28]. In this way P. acnes may, rather, serve as an adjuvant, by providing an inflammatory environment during presentation of other antigens to antigen‐specific immune cells [29].
Our results suggest that mycobacterial antigens do not play a significant role in sarcoidosis pathogenesis of Dutch patients, which is in line with the fact that the Netherlands is a country with a low tuberculosis incidence: approximately five per 100 000 [30]. Other studies investigating the role of mycobacterial antigens in sarcoidosis have shown conflicting results. In cohorts of Danish and Japanese sarcoidosis patients no or a very low percentage of latent tuberculosis infection (LTBI) was found (0 and 3%, respectively [31, 32]). However, in a study using US and Swedish patients with sarcoidosis, 50% of patients demonstrated T cell reactivity against MKatG [33]. The contradicting results regarding MKatG between this and our study may be explained by a different definition of a positive ELISPOT result. Whereas we defined it as 50 SFCs per 106 PBMCs, being at least three times higher than background, Chen et al. used the 95% confidence interval of the control group as a cut‐off, making comparison of results difficult. In our search for trigger‐related phenotypes, a remarkable observation was that sarcoidosis patients sensitized to mycobacterial antigens were more likely to have cardiac involvement. As only five sarcoidosis patients were sensitized for mycobacterial antigens this association requires further investigation, because cardiac involvement is a major cause of death in sarcoidosis patients.
We did not observe a significant difference in vimentin‐positive ELISPOTs between sarcoidosis and control patients, which is in line with an Australian study that found no difference in IFN‐γ‐producing cells after stimulation with pooled peptides of vimentin and lysyl tRNA synthetase [34]. Based on a previous study reporting a higher number of IFN‐γ‐producing cells after vimentin stimulation in DRB1*0301‐positive sarcoidosis patients, [35] we also performed human leucocyte antigen (HLA)‐DRB1 genotyping. In contrast to DRB1*0301, we observed a significantly higher percentage of DRB1*1501 carriers among vimentin‐positive patients. This can be explained by findings of Wahlström et al. who identified, by peptide binding predictions, that apart from DRB1*0301, other HLA molecules are also possible binders for vimentin, including DRB1*1501 [35]. Where we used the whole protein in ELISPOT assays, Wahlström et al. used a vimentin peptide. The observed association with DRB1*1501 instead of DRB1*0301 might be caused by difference in vimentin epitopes. Unfortunately, no residual DNA was available from controls, so we were not able to investigate whether sensitization to vimentin in the control group was also related to DRB1*1501 or whether this was sarcoidosis‐specific.
To additionally study other cytokines, concentrations of TNF‐α, IL‐2, IL‐17A and IL‐1β were measured in supernatant of ELISPOT plates. Higher median IL‐1β concentrations were found in sarcoidosis patients. However, as higher concentrations were observed after stimulation with vimentin, mycobacterial and P. acnes antigens, this seems to be an unspecific inflammatory reaction rather than a specific T cell response.
There are some limitations to our study. Presence of fibrosis was determined by chest X‐ray. Only moderate agreement on Scadding stages is reported in the literature [36]. Furthermore, we are aware that HRCT is preferable to chest X‐ray to determine fibrosis. However, follow‐up data of HRCT were available only for a proportion of patients, which would have decreased the cohort size and so the power of the analysis. A second limitation was that only half of all patients agreed to provide a second blood sample required for an LPT. Hence, overlap between sensitization to organic and inorganic antigens could not be analysed for the whole cohort. However, as only a few of the 105 patients demonstrated more than one potential trigger, it is unlikely that this would have been the case in the patients not analysed for inorganic agents.
In conclusion, our study suggests a more prominent role for inorganic triggers such as metals and silica in sarcoidosis pathogenesis than previously thought. Immunological sensitization to inorganic antigens seems related to the development of fibrotic sarcoidosis. Future studies including a higher number of patients and controls are warranted to clarify the meaning of a weak positive LPT in sarcoidosis and to definitely confirm the role of metal and silica sensitization in its pathogenesis. We did not find an association between sarcoidosis and an immunological response to bacterial antigens or vimentin. However, our data on LTBI and the presence of cardiac sarcoidosis, and the possible relation between sensitization to P. acnes and cutaneous involvement, further suggest that trigger‐related phenotypes can exist in the heterogeneous population of sarcoidosis patients.
Disclosures
The authors declare that they have no competing interests.
Author contributions
E. B. contributed to the design of the study, contributed to PBMC isolation and ELISPOT assays, data acquisition and analysis, statistical analysis and drafted and edited the manuscript. R. K. performed the ELISPOT and luminex experiments and edited the manuscript, C. R. isolated PBMCs, performed the ELISPOT and luminex experiments and edited the manuscript, J. G. supervised the project and edited the manuscript. B. M. participated in the design of the study, data interpretation, supervision and editing the manuscript. M. V. designed the study, was involved in data interpretation, supervised the project and edited the manuscript. The data sets used and/or analysed during the current study are available from the corresponding author on reasonable request.
Supporting information
Fig. S1. Elispot response after stimulation with the positive control (anti‐CD3 mAb), compared between sarcoidosis patients with and without immunosuppressive medication at time of PBMC isolation for the elispot assay. No significant differences were observed between patients with and without medication (P = 0·09).
Fig. S2. Elispot response after stimulation with the positive control (anti‐CD3 mAb), compared between sarcoidosis patients and controls. A significant higher percentage of controls had a low response (<100 spots) after stimulation with anti‐CD3 mAb used as a positive control (P = 0·020).
Fig. S3. Median cytokine concentration in cell supernatant after subtraction of baseline cytokine levels. A: Il‐1β, Il‐17, Il‐2 and TNF‐α concentrations in supernatant after PBMC stimulation with ESAT‐6. B: Il‐1β, Il‐17, Il‐2 and TNF‐α concentrations in supernatant after PBMC stimulation with the whole MKatG protein. C: Il‐1β, Il‐17, Il‐2 and TNF‐α concentrations in supernatant after PBMC stimulation with P. acnes catalase. D: Il‐1β, Il‐17, Il‐2 and TNF‐α concentrations in supernatant after PBMC stimulation with vimentin.
Fig. S4. Median stimulation index after lymphocyte stimulation with pokeweed, used as a positive control in de LPT. A. No difference was observed in median stimulation index between sarcoidosis patient with and without immunosuppressive medication at time of LPT (191·80 vs. 155·40, P = 0·344). B. No difference was observed in median stimulation index between sarcoidosis patients and controls (152·45 vs. 173·40, P = 0·492).
Fig. S5. Percentage of sarcoidosis patients positive for HLA DRB1*03 (A) and *15 (B).
Table S1. Median baseline cytokine levels.
Table S2. Overlap in weakly positive LPT results of sarcoidosis patients.
Table S3. Overlap in positive elispot results of sarcoidosis patients.
Table S4. Overlap in positive elispots and LPTs of sarcoidosis patients.
Table S5. Occupations of the LPT positive and weakly positive sarcoidosis patients and control patient.
Acknowledgements
We would like to thank the research group of Y. Eishi (Department of Human Pathology, Tokyo Medical and Dental University Graduate school, Tokyo, Japan) who kindly provided P.acnes catalase antigen. This study is part of the TopZorg Lung grant funded by ZonMw (no. 842002001).
Data availability statement
Data available on request from the authors. The data that support the findings of this study are available from the corresponding author upon reasonable request.
References
- 1. Baughman RP, Teirstein AS, Judson MA et al Clinical characteristics of patients in a case control study of sarcoidosis. Am J Respir Crit Care Med 2001; 164:1885–9. [DOI] [PubMed] [Google Scholar]
- 2. Prasse A. The diagnosis, differential diagnosis, and treatment of sarcoidosis. Dtsch Arztebl Int 2016; 113:565–74. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. Fang C, Huang H, Xu Z. Immunological evidence for the role of mycobacteria in sarcoidosis: a meta‐analysis. PLOS ONE 2016; 11:e0154716. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4. Gupta D, Agarwal R, Aggarwal AN, Jindal SK. Molecular evidence for the role of mycobacteria in sarcoidosis: a meta‐analysis. Eur Respir J 2007; 30:508–16. [DOI] [PubMed] [Google Scholar]
- 5. Eishi Y. Etiologic link between sarcoidosis and Propionibacterium acnes . Respir Invest 2013; 51:56–68. [DOI] [PubMed] [Google Scholar]
- 6. Zhou Y, Hu Y, Li H. Role of propionibacterium acnes in sarcoidosis: a meta‐analysis. Sarcoidosis Vasc Diffuse Lung Dis 2013; 30:262–7. [PubMed] [Google Scholar]
- 7. Grunewald J, Kaiser Y, Ostadkarampour M et al T‐cell receptor‐HLA‐DRB1 associations suggest specific antigens in pulmonary sarcoidosis. Eur Respir J 2016; 47:898–909. [DOI] [PubMed] [Google Scholar]
- 8. Eberhardt C, Thillai M, Parker R et al Proteomic analysis of kveim reagent identifies targets of cellular immunity in sarcoidosis. PLOS ONE 2017; 12:e0170285. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Fireman E, Shai AB, Alcalay Y, Ophir N, Kivity S, Stejskal V. Identification of metal sensitization in sarcoid‐like metal‐exposed patients by the MELISA(R) lymphocyte proliferation test – a pilot study. J Occup Med Toxicol 2016; 11:18–016‐0101‐1. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Rafnsson V, Ingimarsson O, Hjalmarsson I, Gunnarsdottir H. Association between exposure to crystalline silica and risk of sarcoidosis. Occup Environ Med 1998; 55:657–60. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. Costabel U, Hunninghake GW. ATS/ERS/WASOG statement on sarcoidosis. Sarcoidosis Statement Committee. American Thoracic Society. European Respiratory Society. World Association for Sarcoidosis and Other Granulomatous Disorders. Eur Respir J 1999; 14:735–7. [DOI] [PubMed] [Google Scholar]
- 12. Maspero C, Giannini L, Galbiati G, Rosso G, Farronato G. Obstructive sleep apnea syndrome: a literature review. Minerva Stomatol. 2015; 64:97–109. [PubMed] [Google Scholar]
- 13. Drake WP, Dhason MS, Nadaf M et al Cellular recognition of Mycobacterium tuberculosis ESAT‐6 and KatG peptides in systemic sarcoidosis. Infect Immun 2007; 75:527–30. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. Karakaya B, Schimmelpennink MC, Kocourkova L et al Bronchoalveolar lavage characteristics correlate with HLA tag SNPs in patients with Löfgren’s syndrome and other sarcoidosis. Clin Exp Immunol 2019; 196:249–58. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15. Baughman RP, Nagai S, Balter M et al Defining the clinical outcome status (COS) in sarcoidosis: results of WASOG Task Force. Sarcoidosis Vasc Diffus Lung Dis 2011; 28:56–64. [PubMed] [Google Scholar]
- 16. Harris PA, Taylor R, Thielke R, Payne J, Gonzalez N, Conde JG. Research electronic data capture (REDCap) – a metadata‐driven methodology and workflow process for providing translational research informatics support. J Biomed Inform 2009; 42:377–81. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17. Sood A, Beckett WS, Cullen MR. Variable response to long‐term corticosteroid therapy in chronic beryllium disease. Chest 2004; 126:2000–7. [DOI] [PubMed] [Google Scholar]
- 18. Sprince NL, Kanarek DJ, Weber AL, Chamberlin RI, Kazemi H. Reversible respiratory disease in beryllium workers. Am Rev Respir Dis 1978; 117:1011–7. [DOI] [PubMed] [Google Scholar]
- 19. Beijer E, Meek B, Bossuyt X et al Immunoreactivity to metal and silica associates with sarcoidosis in Dutch patients. Respir Res 2020; 21:141. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20. Infante PF, Newman LS. Beryllium exposure and chronic beryllium disease. Lancet 2004; 363:415–6. [DOI] [PubMed] [Google Scholar]
- 21. Newman LS, Kreiss K. Nonoccupational beryllium disease masquerading as sarcoidosis: identification by blood lymphocyte proliferative response to beryllium. Am Rev Respir Dis 1992; 145:1212–4. [DOI] [PubMed] [Google Scholar]
- 22. Panselinas E, Judson MA. Acute pulmonary exacerbations of sarcoidosis. Chest 2012; 142:827–36. [DOI] [PubMed] [Google Scholar]
- 23. Ishige I, Eishi Y, Takemura T et al Propionibacterium acnes is the most common bacterium commensal in peripheral lung tissue and mediastinal lymph nodes from subjects without sarcoidosis. Sarcoidosis, Vasc Diffus Lung Dis 2005; 22:33–42. [PubMed] [Google Scholar]
- 24. Sorini C, Cardoso RF, Gagliani N, Villablanca EJ. Commensal bacteria‐specific CD4+ T cell responses in health and disease. Front Immunol 2018; 9:2667. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25. Yorozu P, Furukawa A, Uchida K et al Propionibacterium acnes catalase induces increased Th1 immune response in sarcoidosis patients. Respir Investig 2015; 53:161–9. [DOI] [PubMed] [Google Scholar]
- 26. Furusawa H, Suzuki Y, Miyazaki Y, Inase N, Eishi Y. Th1 and Th17 immune responses to viable Propionibacterium acnes in patients with sarcoidosis. Respir Invest 2012; 50:104–9. [DOI] [PubMed] [Google Scholar]
- 27. Hegazy AN, West NR, Stubbington MJT et al Circulating and tissue‐resident CD4+ T cells with reactivity to intestinal microbiota are abundant in healthy individuals and function is altered during inflammation. Gastroenterology 2017; 153:1320–1337.e16. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28. Jappe U, Ingham E, Henwood J, Holland KT. Propionibacterium acnes and inflammation in acne; P. acnes has T‐cell mitogenic activity. Br J Dermatol 2002; 146:202–9. [DOI] [PubMed] [Google Scholar]
- 29. Wang Z‐B, Xu J. Better adjuvants for better vaccines: progress in adjuvant delivery systems, modifications, and adjuvant‐antigen codelivery. Vaccines 2020; 8:128. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30. Slump E, Erkens CGM, van Hunen MCJ, Schimmel HJ, van Soolingen D, de Vries G. Tuberculosis in the Nederlands 2018 – Surveillance Report. Utercht, the Netherlands: Rijksinstituut voor Volksgezondheid en Millieu [National Institute for Public Health and the Environment], 2019. [Google Scholar]
- 31. Milman N, Søborg B, Svendsen CB, Andersen ÅB. Quantiferon test for tuberculosis screening in sarcoidosis patients. Scand J Infect Dis 2011; 43:728–35. [DOI] [PubMed] [Google Scholar]
- 32. Inui N, Suda T, Chida K. Use of the QuantiFERON‐TB Gold test in Japanese patients with sarcoidosis. Respir Med 2008; 102:313–5. [DOI] [PubMed] [Google Scholar]
- 33. Chen ES, Wahlstrom J, Song Z et al T cell responses to mycobacterial catalase‐peroxidase profile a pathogenic antigen in systemic sarcoidosis. J Immunol 2008; 181:8784–96. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34. Ahmadzai H, Cameron B, Chui JJY, Lloyd A, Wakefield D, Thomas PS. Peripheral blood responses to specific antigens and CD28 in sarcoidosis. Respir Med 2012; 106:701–9. [DOI] [PubMed] [Google Scholar]
- 35. Wahlstrom J, Dengjel J, Winqvist O et al Autoimmune T cell responses to antigenic peptides presented by bronchoalveolar lavage cell HLA‐DR molecules in sarcoidosis. Clin Immunol 2009; 133:353–63. [DOI] [PubMed] [Google Scholar]
- 36. Baughman RP, Shipley R, Desai S et al Changes in chest roentgenogram of sarcoidosis patients during a clinical trial of infliximab therapy: comparison of different methods of evaluation. Chest 2009; 136:526–35. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Fig. S1. Elispot response after stimulation with the positive control (anti‐CD3 mAb), compared between sarcoidosis patients with and without immunosuppressive medication at time of PBMC isolation for the elispot assay. No significant differences were observed between patients with and without medication (P = 0·09).
Fig. S2. Elispot response after stimulation with the positive control (anti‐CD3 mAb), compared between sarcoidosis patients and controls. A significant higher percentage of controls had a low response (<100 spots) after stimulation with anti‐CD3 mAb used as a positive control (P = 0·020).
Fig. S3. Median cytokine concentration in cell supernatant after subtraction of baseline cytokine levels. A: Il‐1β, Il‐17, Il‐2 and TNF‐α concentrations in supernatant after PBMC stimulation with ESAT‐6. B: Il‐1β, Il‐17, Il‐2 and TNF‐α concentrations in supernatant after PBMC stimulation with the whole MKatG protein. C: Il‐1β, Il‐17, Il‐2 and TNF‐α concentrations in supernatant after PBMC stimulation with P. acnes catalase. D: Il‐1β, Il‐17, Il‐2 and TNF‐α concentrations in supernatant after PBMC stimulation with vimentin.
Fig. S4. Median stimulation index after lymphocyte stimulation with pokeweed, used as a positive control in de LPT. A. No difference was observed in median stimulation index between sarcoidosis patient with and without immunosuppressive medication at time of LPT (191·80 vs. 155·40, P = 0·344). B. No difference was observed in median stimulation index between sarcoidosis patients and controls (152·45 vs. 173·40, P = 0·492).
Fig. S5. Percentage of sarcoidosis patients positive for HLA DRB1*03 (A) and *15 (B).
Table S1. Median baseline cytokine levels.
Table S2. Overlap in weakly positive LPT results of sarcoidosis patients.
Table S3. Overlap in positive elispot results of sarcoidosis patients.
Table S4. Overlap in positive elispots and LPTs of sarcoidosis patients.
Table S5. Occupations of the LPT positive and weakly positive sarcoidosis patients and control patient.
Data Availability Statement
Data available on request from the authors. The data that support the findings of this study are available from the corresponding author upon reasonable request.