- Split View
-
Views
-
Cite
Cite
M. L. Nowlan, E. Drewe, H. Bulsara, N. Esposito, R. A. Robins, P. J. Tighe, R. J. Powell, I. Todd, Systemic cytokine levels and the effects of etanercept in TNF receptor-associated periodic syndrome (TRAPS) involving a C33Y mutation in TNFRSF1A, Rheumatology, Volume 45, Issue 1, January 2006, Pages 31–37, https://doi.org/10.1093/rheumatology/kei090
- Share Icon Share
Abstract
Objective. To investigate the levels of the pro-inflammatory cytokines IL-6, TNF-α, IL-1β, IL-8, IL-10 and IL-12p70 in the plasma of patients with TNF receptor-associated periodic syndrome (TRAPS) in relation to CRP levels and treatment with etanercept.
Methods. Cytokine concentrations were measured in sequential plasma samples obtained from eight patients with a C33Y mutation in TNFRSF1A and diagnosed with TRAPS, using cytokine bead array. The TRAPS samples were compared with samples from normal controls and rheumatoid arthritis patients.
Results. Levels of IL-6 were significantly elevated in C33Y TRAPS patients and these correlated with CRP levels in some of the patients. IL-8 levels were also significantly elevated in the TRAPS patients. However, neither TNF-α nor IL-1β demonstrated a similar increase. This differed from the patients with rheumatoid arthritis, for whom levels of IL-6, IL-8, TNF-α, IL-1β and IL-10 were significantly elevated. The levels of detectable TNF-α in the TRAPS patients’ plasma were elevated during etanercept treatment.
Conclusions. The cytokine profile of C33Y TRAPS differs from that of a typical autoimmune inflammatory condition such as rheumatoid arthritis, as only IL-6 and IL-8 were elevated in C33Y TRAPS patients, as distinct from a generalized elevation of pro-inflammatory cytokines. However, only some of the C33Y patients tested showed a relationship between elevated IL-6 and CRP. This is consistent with clinical observations that there is marked heterogeneity between individuals with TRAPS, including those in the same family cohort. Although etanercept has a therapeutic effect in some TRAPS patients, it induces increased plasma concentrations of TNF-α, possibly by increasing TNF-α stability.
TNF receptor-associated periodic syndrome (TRAPS) is an auto-inflammatory condition, characterized by recurrent fevers, myalgia, skin rashes and joint and abdominal pain. It is associated with autosomal dominant mutations in the gene encoding the 55 kDa tumour necrosis factor receptor (TNFRSF1A, TNFR1, CD120a, p55 TNFR) [1]. Studies on the pro-inflammatory cytokine profiles during attacks of TRAPS are limited but have revealed low levels of soluble TNFRSF1A in some cases, and elevated interleukin-6 (IL-6) during attacks [2]. Actions of other pro-inflammatory cytokines may relate to the clinical variability notable across the spectrum of TRAPS patients. TRAPS patients can also develop amyloidosis in association with a persisting acute-phase response. Treatment with the TNFRSF1B (TNFR2):Fc fusion protein etanercept (Enbrel) has therapeutic benefits in some TRAPS patients and can control the clinical attacks, and results in improvement of renal function and reduction in amyloid deposition [3]. Associated cytokine changes in TRAPS patients have not been fully investigated. Other periodic fever syndromes, such as familial Mediterranean fever (FMF) and hyperimmunoglobulinaemia-D with periodic fever syndrome (HIDS), are also characterized by an elevation of IL-6 during attacks, high C-reactive protein (CRP) levels, and in some instances this is followed by amyloidosis and renal failure [4–8]. Other cytokines that are elevated during attacks in FMF include TNF-α [8], IL-8 [9] and interferon-γ [10, 11].
The acute-phase inflammatory response is associated with the production of pyrogenic cytokines, namely IL-6, IL-1β and TNF-α, and acute-phase proteins such as CRP and serum amyloid A. CRP is a clinically recognized marker of systemic inflammation and its production by the liver is modulated by IL-6. The pro-inflammatory effects of TNF-α are mediated predominantly through binding to TNFRSF1A [1] inducing NFκB activation, increased gene transcription and cytokine production. The identification of numerous mutations in the ectodomain of TNFRSF1A in TRAPS patients, and low serum levels of soluble TNFRSF1A in certain cases further suggest a role for the receptor in the generation of clinical attacks of TRAPS [12]. Our studies with transfected cell lines have shown that the ability of the various mutant forms of TNFRSF1A to be expressed on the surface of cells and to bind TNF-α varies and also that the shedding of mutant receptors from the cell surface differs between cell types [13, 14].
IL-6 is not only a pro-inflammatory cytokine but can also have anti-inflammatory feedback effects by inducing cytokine antagonists such as IL-1 receptor antagonist (IL-1RA) and soluble TNF receptors [15]. IL-1β can also act as a pro-inflammatory cytokine, inducing the production of several cytokines, including itself, TNF-α, IL-6, IL-8 and IL-12, and can be induced by acute-phase proteins such as CRP [16]. Increased plasma IL-1β has been correlated to disease activity in rheumatoid arthritis [17].
Monocytes and fibroblasts produce IL-8, a chemoattractant for neutrophils during inflammation, and this chemokine may have a role in the dermatological manifestations notable in TRAPS. Monocytic fasciitis has also been observed in affected muscles during attacks of TRAPS [18]. IL-10 is usually characterized as an anti-inflammatory cytokine, and has been shown to inhibit NF-κB activity [19]. Anti-inflammatory responses and a feedback mechanism may explain the periodicity of TRAPS.
In this study, we have simultaneously measured CRP, IL-6, IL-1β, TNF-α, IL-8, IL-10 and IL-12p70 levels in sequential plasma samples from patients with C33Y mutation in the TNFRSF1A gene (referred to here as C33Y TRAPS). These patients are from the prototypical TRAPS family, whose condition was originally called familial Hibernian fever (FHF) [20]. We have also investigated the effect of etanercept treatment on the plasma cytokine levels in these TRAPS patients, and changes to their cytokine levels during disease attacks in relation to changes in the CRP levels.
Methods
Patient samples
Plasma samples of eight TRAPS patients with a C33Y mutation were collected at consecutive time points over a period of about 2 yr (aged 25–75 yr; five male, three female). The symptomatic assessment of the patients at the time of obtaining each sample was noted and classified as ‘well’, ‘mild attack’ or ‘attack’. The medication was also recorded; specifically whether they were receiving etanercept or not. Etanercept (Enbrel; Wyeth, UK) was given subcutaneously at 25 mg twice weekly to certain patients. The patients frequently received prednisolone during attacks. Peripheral blood was collected into 4 ml EDTA (ethylenediamine tetraacetate) Vacutainers (BD Biosciences, Oxford, UK) during out-patient attendances. The blood was centrifuged at 700 g for 8 min within 30 min of collection; the plasma was then aspirated and frozen at −80°C until analysis. Samples were also collected anonymously from seven patients with recently diagnosed rheumatoid arthritis (RA) and 34 normal controls. All the patients gave informed consent and the Nottingham Ethics Committee approved this study. Immediately prior to the analysis of the samples, they were thawed and aliquoted to prevent repeated freeze/thawing. Some of the plasma samples required microcentrifugation to remove fibrin debris.
Cytokine bead array
A cytokine bead array kit (CBA; BD Biosciences) was used to measure the inflammatory cytokines in the plasma samples. The cytokines measured by this kit were IL-6, IL-1β, TNF-α, IL-8, IL-10 and IL-12p70. The data were acquired on an Altra flow cytometer (Beckman Coulter, High Wycombe, UK). The data were then interpreted using BD cytometric bead array software (BD Biosciences).
Clinical assays
CRP levels were measured as part of the routine testing performed on blood samples taken in the out-patient clinic.
Etanercept and TNF-α
To determine whether TNF-α can be detected by CBA or ELISA when bound to etanercept, various dilutions of etanercept (10–0.05 μg/ml) were incubated at room temperature for 1 h with different concentrations of TNF-α (31–1000 pg/ml; R & D Europe, Abingdon, UK) in Tris-buffered saline with 0.1% bovine serum albumin (Sigma-Aldrich, Poole, UK). These samples were frozen at −20°C until assayed by CBA or ELISA (Duoset kit; R & D Europe). A TNF-sensitive L929 bioassay was used to assess the bioactivity of the TNF-α when bound to etanercept [21]. The L929 fibroblast cell line was obtained from the European Collection of Animal Cell Cultures (Salisbury, UK); the cells were grown in 96-well plates (Nunc; SLS, Nottingham, UK) at 4 × 104/100 μl overnight in Dulbecco's Modified Eagle Medium (DMEM; Invitrogen, Paisley, UK) supplemented with 10 mM HEPES, 100 U/ml penicillin, 100 μg/ml streptomycin, 2 mM l-glutamine (all from Sigma) and 10% fetal bovine serum (FBS; Harlan SeraLab, Loughborough, UK). Etanercept at 0.5 μg/ml was diluted in supplemented DMEM with TNF-α at 1 ng/ml and incubated for 1 h at room temperature followed by 3 h at 4°C. A TNF-α-only control was also prepared. The etanercept/TNF-α solutions were diluted on the L929 plate in doubling dilutions, 50 μl per well; 50 μl actinomycin-D–mannitol solution at 8 μg/ml in supplemented DMEM was also added. The plate was then incubated overnight at 37°C, 5% CO2. It was washed with phosphate-buffered saline and the cells were stained with 100 μl 0.05% crystal violet (SLS) in 20% ethanol for 10 min. The plate was then washed with tap water and dried. The crystal violet was eluted with 100 μl methanol and the optical density was measured at 550 nm on an Emax Microplate Reader (Molecular Devices, Wokingham, UK).
Statistics
Graphpad Prism software (San Diego, California, USA) was used to analyse the data. The analyses used are described with the results.
Results
CRP levels in TRAPS patients
CRP level is measured clinically as a surrogate for systemic inflammation. The reference range for CRP levels is up to 10 mg/l. As shown in Fig. 1, the CRP levels of TRAPS patients can greatly exceed this level. However, the C33Y TRAPS patients can be suffering an ‘attack’ without an elevation of CRP because of localized inflammation, such as myalgia, skin rashes, abdominal, ocular and joint pain, rather than systemic symptoms, such as fever. Alternatively, CRP levels can be elevated without signs of clinical disease. As peripheral blood plasma cytokine levels were being measured, we used CRP in this study to reflect the systemic changes occurring during attacks. Figure 1 shows the CRP profiles for Patients 1, 2, 3, 4, 5, 6 and 8 in relation to etanercept treatment and clinical notes. CRP levels for Patients 5 and 6 did not rise above 10 mg/l (normal range), even during attacks: Patient 5 experienced relatively mild attacks only, whereas Patient 6 was atypical in having a renal transplant as a consequence of amyloidosis and received azathioprine and cyclosporin. For Patient 7, there were only two samples available, both with CRP levels above the normal range (31 and 87 mg/l, respectively): on both occasions the patient was described as ‘well’ and was not receiving any medication. Patient 1 did not have a sustained response to etanercept and hence was prescribed sirolimus instead. It can be seen that CRP did not always rise during an ‘attack’ and often remained substantially elevated even when the patient appeared clinically ‘well’. Most patients always had an abnormally high CRP level, except Patients 5 and 6, in whom the CRP levels were low, as noted above. Etanercept treatment generally resulted in a fall in CRP levels.
Correlation between CRP and proinflammatory cytokines
Table 1 shows the Spearman's r values and P values for the correlation between CRP and each cytokine for individual C33Y TRAPS patients. Bonferroni correction for multiple comparisons was performed and a P value of <0.008 was considered significant. The data are given for four of the eight patients when not receiving etanercept or sirolimus treatment. No data are given for Patients 2, 5, 6 or 7; Patients 2, 5 and 6 were on etanercept for the majority of the sampling period and only three samples were available for each patient while not receiving etanercept treatment; only two samples were taken for Patient 7 as this patient manages the condition well and does not attend clinic regularly. The data in this table demonstrate heterogeneity between the patients. Patients 3 and 4 showed a significant correlation between IL-6 and CRP, as would be expected in an inflammatory condition. There was a similar trend for Patient 1 (Spearman's r-value 0.75), although this was not statistically significant. IL-8 was also closely related to CRP in this patient, but again this did not reach statistical significance. However Patient 8 did not show any correlation between CRP and the cytokines. Apart from IL-6 in Patients 3 and 4, none of the other pro-inflammatory cytokines showed significant correlation with CRP in TRAPS.
. | Patient 1 . | . | Patient 3 . | . | Patient 4 . | . | Patient 8 . | . | ||||
---|---|---|---|---|---|---|---|---|---|---|---|---|
. | r value . | P value . | r value . | P value . | r value . | P value . | r value . | P value . | ||||
IL-12 | 0.393 | 0.396 | 0.561 | 0.151 | −0.030 | 0.946 | −0.358 | 0.336 | ||||
TNF-α | −0.107 | 0.840 | −0.342 | 0.389 | 0.103 | 0.785 | −0.368 | 0.336 | ||||
IL-10 | 0.667 | 0.110 | 0.464 | 0.243 | 0.188 | 0.607 | −0.122 | 0.744 | ||||
IL-6 | 0.750 | 0.066 | 0.881 | 0.007 | 0.877 | 0.002 | −0.050 | 0.912 | ||||
IL-8 | 0.800 | 0.034 | 0.431 | 0.299 | 0.377 | 0.279 | −0.025 | 0.948 | ||||
IL-1β | 0.324 | 0.498 | −0.027 | 0.977 | 0.059 | 0.880 |
. | Patient 1 . | . | Patient 3 . | . | Patient 4 . | . | Patient 8 . | . | ||||
---|---|---|---|---|---|---|---|---|---|---|---|---|
. | r value . | P value . | r value . | P value . | r value . | P value . | r value . | P value . | ||||
IL-12 | 0.393 | 0.396 | 0.561 | 0.151 | −0.030 | 0.946 | −0.358 | 0.336 | ||||
TNF-α | −0.107 | 0.840 | −0.342 | 0.389 | 0.103 | 0.785 | −0.368 | 0.336 | ||||
IL-10 | 0.667 | 0.110 | 0.464 | 0.243 | 0.188 | 0.607 | −0.122 | 0.744 | ||||
IL-6 | 0.750 | 0.066 | 0.881 | 0.007 | 0.877 | 0.002 | −0.050 | 0.912 | ||||
IL-8 | 0.800 | 0.034 | 0.431 | 0.299 | 0.377 | 0.279 | −0.025 | 0.948 | ||||
IL-1β | 0.324 | 0.498 | −0.027 | 0.977 | 0.059 | 0.880 |
P<0.008 is considered evidence of a significant difference.
. | Patient 1 . | . | Patient 3 . | . | Patient 4 . | . | Patient 8 . | . | ||||
---|---|---|---|---|---|---|---|---|---|---|---|---|
. | r value . | P value . | r value . | P value . | r value . | P value . | r value . | P value . | ||||
IL-12 | 0.393 | 0.396 | 0.561 | 0.151 | −0.030 | 0.946 | −0.358 | 0.336 | ||||
TNF-α | −0.107 | 0.840 | −0.342 | 0.389 | 0.103 | 0.785 | −0.368 | 0.336 | ||||
IL-10 | 0.667 | 0.110 | 0.464 | 0.243 | 0.188 | 0.607 | −0.122 | 0.744 | ||||
IL-6 | 0.750 | 0.066 | 0.881 | 0.007 | 0.877 | 0.002 | −0.050 | 0.912 | ||||
IL-8 | 0.800 | 0.034 | 0.431 | 0.299 | 0.377 | 0.279 | −0.025 | 0.948 | ||||
IL-1β | 0.324 | 0.498 | −0.027 | 0.977 | 0.059 | 0.880 |
. | Patient 1 . | . | Patient 3 . | . | Patient 4 . | . | Patient 8 . | . | ||||
---|---|---|---|---|---|---|---|---|---|---|---|---|
. | r value . | P value . | r value . | P value . | r value . | P value . | r value . | P value . | ||||
IL-12 | 0.393 | 0.396 | 0.561 | 0.151 | −0.030 | 0.946 | −0.358 | 0.336 | ||||
TNF-α | −0.107 | 0.840 | −0.342 | 0.389 | 0.103 | 0.785 | −0.368 | 0.336 | ||||
IL-10 | 0.667 | 0.110 | 0.464 | 0.243 | 0.188 | 0.607 | −0.122 | 0.744 | ||||
IL-6 | 0.750 | 0.066 | 0.881 | 0.007 | 0.877 | 0.002 | −0.050 | 0.912 | ||||
IL-8 | 0.800 | 0.034 | 0.431 | 0.299 | 0.377 | 0.279 | −0.025 | 0.948 | ||||
IL-1β | 0.324 | 0.498 | −0.027 | 0.977 | 0.059 | 0.880 |
P<0.008 is considered evidence of a significant difference.
Examples of the relationships between IL-6 and CRP levels in Patients 1, 3 and 8 are shown in Fig. 2. For Patient 3, the IL-6 levels closely mimicked changes in CRP, even when the patient was on etanercept. As predicted by the absence of correlation in Patient 8 (Table 1), IL-6 and CRP levels did not mirror each other closely. Patient 8 had never required etanercept treatment. In Patient 1, the profiles for CRP and IL-6 were similar, whether on or off treatment with sirolimus.
Normal cytokine ranges
Table 2 shows the median and range of cytokine levels in 34 normal controls and seven RA patient samples in comparison with samples taken from eight TRAPS patients without etanercept treatment. For the TRAPS patients, the median and range of cytokine levels are given for all samples assayed, and for the median cytokine values for each patient. We used Kruskal–Wallis test followed by Dunn's multiple comparison post-test to compare the normal controls, RA patients and the median TRAPS samples. The RA patients had a generalized elevation of most of the cytokines: the levels of IL-6 (P<0.001), TNF-α (P<0.001), IL-8 (P<0.001) and IL-10 (P<0.001) were significantly higher than in controls. IL-6 and IL-8 were also significantly elevated in the median samples from TRAPS patients compared with the normal controls (P<0.01 for both); there was no significant difference between the levels of these cytokines in the TRAPS patients and the RA patients. The IL-1β levels in the TRAPS patients were significantly lower than in the RA patients and normal controls. Unlike the RA patients, TNF-α levels were not significantly raised in the TRAPS patients. No significant difference in IL-12 levels was seen between any of the groups.
. | . | . | TRAPS . | . | |
---|---|---|---|---|---|
Cytokine . | Normal . | RA . | All samples . | Median samples . | |
IL-6 | 4.0 (0–9.3) | 15.8c (8.8–31.2) | 15.5 (0–260) | 12.5b (3.8–34.7) | |
TNF-α | 3.2 (0–7.3) | 10.0c (6.0–45.8) | 3.1 (0–29.6) | 3.1 (0–8.9) | |
IL-1β | 57.3 (0–265.5) | 129.8 (0–711) | 0 (0–98.0) | 0a (0–23.5) | |
IL-8 | 2.6 (1.4–5.4) | 9.3c (3.2–28.7) | 5.3 (2.1–17.1) | 4.8b (3.6–16.3) | |
IL-10 | 4.6 (0–8.6) | 15.5b (7.4–24.4) | 4.1 (0–31.6) | 4.1 (2.8–14.8) | |
IL-12 | 6.2 (0–27.9) | 20.2 (0–39.7) | 4.8 (0–17.9) | 4.6 (3.4–10.9) |
. | . | . | TRAPS . | . | |
---|---|---|---|---|---|
Cytokine . | Normal . | RA . | All samples . | Median samples . | |
IL-6 | 4.0 (0–9.3) | 15.8c (8.8–31.2) | 15.5 (0–260) | 12.5b (3.8–34.7) | |
TNF-α | 3.2 (0–7.3) | 10.0c (6.0–45.8) | 3.1 (0–29.6) | 3.1 (0–8.9) | |
IL-1β | 57.3 (0–265.5) | 129.8 (0–711) | 0 (0–98.0) | 0a (0–23.5) | |
IL-8 | 2.6 (1.4–5.4) | 9.3c (3.2–28.7) | 5.3 (2.1–17.1) | 4.8b (3.6–16.3) | |
IL-10 | 4.6 (0–8.6) | 15.5b (7.4–24.4) | 4.1 (0–31.6) | 4.1 (2.8–14.8) | |
IL-12 | 6.2 (0–27.9) | 20.2 (0–39.7) | 4.8 (0–17.9) | 4.6 (3.4–10.9) |
Data are presented as median cytokine level and range for normal controls and RA patients, for all the TRAPS patient samples and for the median TRAPS patient samples. aP<0.05, bP<0.01, cP<0.001 in Kruskal–Wallis test with Dunn's multiple comparison post-test is considered evidence of a significant difference from normal controls.
. | . | . | TRAPS . | . | |
---|---|---|---|---|---|
Cytokine . | Normal . | RA . | All samples . | Median samples . | |
IL-6 | 4.0 (0–9.3) | 15.8c (8.8–31.2) | 15.5 (0–260) | 12.5b (3.8–34.7) | |
TNF-α | 3.2 (0–7.3) | 10.0c (6.0–45.8) | 3.1 (0–29.6) | 3.1 (0–8.9) | |
IL-1β | 57.3 (0–265.5) | 129.8 (0–711) | 0 (0–98.0) | 0a (0–23.5) | |
IL-8 | 2.6 (1.4–5.4) | 9.3c (3.2–28.7) | 5.3 (2.1–17.1) | 4.8b (3.6–16.3) | |
IL-10 | 4.6 (0–8.6) | 15.5b (7.4–24.4) | 4.1 (0–31.6) | 4.1 (2.8–14.8) | |
IL-12 | 6.2 (0–27.9) | 20.2 (0–39.7) | 4.8 (0–17.9) | 4.6 (3.4–10.9) |
. | . | . | TRAPS . | . | |
---|---|---|---|---|---|
Cytokine . | Normal . | RA . | All samples . | Median samples . | |
IL-6 | 4.0 (0–9.3) | 15.8c (8.8–31.2) | 15.5 (0–260) | 12.5b (3.8–34.7) | |
TNF-α | 3.2 (0–7.3) | 10.0c (6.0–45.8) | 3.1 (0–29.6) | 3.1 (0–8.9) | |
IL-1β | 57.3 (0–265.5) | 129.8 (0–711) | 0 (0–98.0) | 0a (0–23.5) | |
IL-8 | 2.6 (1.4–5.4) | 9.3c (3.2–28.7) | 5.3 (2.1–17.1) | 4.8b (3.6–16.3) | |
IL-10 | 4.6 (0–8.6) | 15.5b (7.4–24.4) | 4.1 (0–31.6) | 4.1 (2.8–14.8) | |
IL-12 | 6.2 (0–27.9) | 20.2 (0–39.7) | 4.8 (0–17.9) | 4.6 (3.4–10.9) |
Data are presented as median cytokine level and range for normal controls and RA patients, for all the TRAPS patient samples and for the median TRAPS patient samples. aP<0.05, bP<0.01, cP<0.001 in Kruskal–Wallis test with Dunn's multiple comparison post-test is considered evidence of a significant difference from normal controls.
Etanercept treatment
As demonstrated in Fig. 3 (data from Patients 3, 4 and 5), levels of TNF-α detected in plasma were elevated during etanercept treatment, although the CRP levels were generally lowered. In this assay it is not possible to distinguish whether this was due to an increase in the production of TNF-α by the C33Y TRAPS patients during etanercept treatment or whether the assay detects TNF-α bound to, or stabilized by, etanercept, or indeed both. To investigate this further, we mixed TNF-α with etanercept in vitro and assessed whether the TNF-α could be detected by CBA, ELISA and the L929 bioassay. In a dose-dependent manner, less TNF-α was detected in the presence of etanercept than in its absence, by CBA (data not shown) and ELISA (Fig. 4), indicating that etanercept competitively inhibited interaction of TNF-α with the antibodies used in these assays. However, it is of note that etanercept competed only partially with the assay capture antibody. Thus, as seen in Fig. 4, etanercept at 10 μg/ml inhibited detection of TNF-α only slightly more than 1 μg/ml etanercept did and, at both of these etanercept concentrations, detection of 500 pg/ml TNF-α was reduced by only about 50%. This suggests that the TNF-α that is detected at high levels in the plasma of C33Y TRAPS patients during etanercept treatment is not all bound to etanercept. The L929 bioassay confirmed that TNF-α bound to etanercept was not biologically active death of the L929 cells occurred at above 31 pg/ml TNF-α alone, but no killing occurred when the cells were treated with etanercept and TNF-α together (data not shown). Unfortunately, a bioassay of the patients’ plasma could not be interpreted because of complement-mediated cytotoxicity, and heat inactivation of complement in these samples was not practical as this might also denature both the TNF-α and the etanercept. However, overall our findings indicate that detectable levels of TNF-α were elevated in the plasma of TRAPS patients on etanercept treatment.
Discussion
In this study, we investigated the relationship between systemic CRP levels and pro-inflammatory cytokines in patients with C33Y TRAPS. CRP is used clinically to assess systemic inflammatory responses within the patients, and consequently we anticipated that elevations in CRP would reflect increased levels of pro-inflammatory cytokines. Some TRAPS patients develop amyloidosis, and it would be interesting to measure serum amyloid A (SAA) levels as an additional acute-phase reactant. However, SAA is not measured routinely for purposes of clinical monitoring in the C33Y TRAPS patients, and was not undertaken as part of the present study. The relationship between CRP and pro-inflammatory cytokines varied markedly between individual patients and reflects the variation in severity of symptoms seen clinically. However, in some patients there was a strong correlation between CRP levels and the pro-inflammatory cytokine IL-6. This would be anticipated since IL-6 controls CRP production, although not all the C33Y TRAPS patients showed a significant correlation. As shown in RA patients, the levels of the other pro-inflammatory cytokines might also be expected to be elevated; however, this was rarely the case in the TRAPS patients. There is a suggestion that IL-8 is involved in the inflammatory response in C33Y TRAPS, but there was no correlation between IL-8 levels and CRP for most of the patients. However, for Patient 1 P = 0.034 (Table 1). IL-8 is a neutrophil chemoattractant and increased levels would be expected in an inflammatory condition.
IL-1β levels were significantly lower in the C33Y TRAPS patients’ samples than in samples from the RA patients or the normal controls. Circulating levels of IL-β are recognized to be low and difficult to detect [16]. Many of the C33Y TRAPS patient samples were given the value of 0 as they were below the standard curve. However, treatment with recombinant IL-1RA (anakinra) proved successful in C34Y TRAPS patients [22], implying a role for IL-1-mediated inflammation in TRAPS.
A further unexpected finding is the greatly elevated TNF-α levels detected in plasma during etanercept treatment, despite the fact that beneficial effects of etanercept are seen in some TRAPS patients [e.g. 23]. However, elevated circulating levels of TNF-α have also been detected in other groups of patients during etanercept treatment, e.g. patients with multiple myeloma [24] or OKT3-associated acute clinical syndrome [25]. Indeed, some of the early studies with etanercept in lipopolysaccharide (LPS)-treated mice showed that etanercept prolongs the half-life of TNF-α and reduces its loss from the murine circulation while at the same time protecting against LPS-induced shock [26]. This suggests that etanercept, like naturally produced soluble TNF receptors [27], binds and neutralizes TNF-α, but also stabilizes TNF-α, prolonging its half-life. There is also evidence that TNF-α dissociates from etanercept relatively rapidly [26, 28, 29]; it may then reassociate unless prevented from doing so. This may explain why, in the competitive inhibition assays that we performed (e.g. Fig. 4), etanercept could not fully inhibit detection of TNF-α in the CBA or ELISA assays, since TNF-α released by etanercept would bind effectively irreversibly to the capture anti-TNF-α antibodies. Similarly, the plasma of etanercept-treated TRAPS patients is likely to contain a depot of TNF-α bound to etanercept when this plasma is analysed in a CBA or ELISA assay, the TNF-α that dissociates from etanercept will be rapidly and effectively irreversibly bound by the anti-TNF-α capture antibodies used in the assays, resulting in the detection of apparently elevated levels of TNF-α. The therapeutic effect of etanercept in TRAPS indicates that it reduces the bioavailability of TNF-α in vivo, as we also found that it did in vitro in the L929 assay; but this TNF-α is detectable in antibody-based assays where the anti-TNF-α capture antibodies compete effectively with etanercept in TNF-α binding.
Despite the therapeutic effect of etanercept in TRAPS, our results do not clearly indicate a special role for TNF-α in the pathogenesis of TRAPS TNF-α levels did not correlate with CRP (Table 1) and were not significantly elevated compared with healthy controls, whereas they were elevated in RA patients (Table 2). It may be that various forms of inhibition of pro-inflammatory cytokine activity will be beneficial in TRAPS, as reported for the IL-1 antagonist anakinra [22] in addition to etanercept. In particular, we found that in some of the C33Y TRAPS patients only IL-6 levels are elevated and correlate with CRP levels, suggesting that IL-6 may be a promising target for therapy in TRAPS. Studies using anti-IL-6 monoclonal antibody in cancer showed a reduction in CRP levels [30], and anti-IL-6 receptor antibody has had clinical benefit in RA and Crohn's disease [31, 32].
In summary, we have shown marked heterogeneity between C33Y TRAPS patients in their cytokine profiles, in particular IL-6 levels in relation to the CRP levels. Systemic IL-6 and CRP levels were elevated in most C33Y TRAPS patients but did not always correlate. This indicates that TRAPS is a heterogeneous condition, even within a family with the same TNFRSF1A mutation. This heterogeneity suggests that the faulty TNFRSF1A is not the only factor in TRAPS and that other genetic and/or environmental factors may play a role. In addition, the cytokine response in the C33Y TRAPS patients did not appear to follow the pattern seen in autoimmune inflammatory conditions, such as RA: in RA patients, many of the pro-inflammatory cytokines were elevated, whereas in some of the C33Y TRAPS patients only IL-6 appeared to be truly elevated.
We are grateful to Dr Peter Lanyon for providing the RA patients’ samples. This work was funded by The Jones 1986 Charitable Trust and by RCPath and International Journal of Experimental Pathology Fellowship awarded to E.D.
The authors have declared no conflicts of interest.
References
McDermott M, Aksentijevich I, Galon J et al. Germline mutations in the extracellular domains of the 55kDa TNF receptor, TNFR1, define a family of dominantly inherited autoinflammatory syndromes.
McDermott E, Powell R.
Drewe E, Huggins ML, Morgan AG, Cassidy MJD, Powell RJ. Treatment of renal amyloidosis with etanercept in tumour necrosis factor receptor-associated periodic syndrome.
Drenth J, van Deuren M, van der Ven-Jongerkrijg J, Schalkwijk C, van der Meer J. Cytokine activation during attacks of the hyperimmunoglobulinemia D and periodic fever syndrome.
Gang N, Drenth J, Langevitz P et al. Activation of the cytokine network in familial Mediterranean fever.
Akcan Y, Bayraktar Y, Arslan S, Van Thiel DH, Zerrin BC, Yildiz O. The importance of serial measurements of cytokine levels for the evaluation of their role in pathogenesis in familial Mediterranean fever.
Bagci S, Toy B, Tuzun A et al. Continuity of cytokine activation in patients with familial Mediterranean fever.
Baykal Y, Saglam K, Yilmaz MI, Taslipinar A, Akinci SB, Inal A. Serum sIL-2r, IL-6, IL-10 and TNF-alpha level in familial Mediterranean fever patients.
Direskeneli H, Ozdogan H, Korkmaz C, Akoglu T, Yazici H. Serum soluble intercellular adhesion molecule 1 and interleukin 8 levels in familial Mediterranean fever.
Aypar E, Ozen S, Okur H, Kutluk T, Besbas N, Bakkalogu A. Th1 polarization in familial Mediterranean fever.
Koklu S, Ozturk MA, Balci M, Yuksel O, Ertenli I, Kiraz S. Interferon-gamma levels in familial Mediterranean fever.
Galon J, Aksentijevich I, McDermott M, O’Shea J, Kastner D. TNFRSR1A mutations and autoinflammatory syndromes.
Todd I, Radford PM, Draper-Morgan KA et al. Mutant forms of tumour necrosis factor receptor I that occur in TNF-receptor-associated periodic syndrome retain signalling functions but show abnormal behaviour.
Huggins ML, Radford PM, McIntosh RS et al. Shedding of mutant tumor necrosis factor receptor superfamily 1A associated with tumor necrosis factor associated periodic syndrome (TRAPS): differences between cell types.
Tilg H, Trehu E, Atkins M, Dinarello C, Mier J. Interleukin-6 (IL-6) as an anti-inflammatory cytokine: induction of circulating IL-1 receptor antagonist and soluble tumor necrosis factor receptor p55.
Eastgate JA, Symons JA, Wood NC, Grinlinton FM, di Giovine FS, Duff GW. Correlation of plasma interleukin 1 levels with disease activity in rheumatoid arthritis.
Hull K, Wong K, Wood G, Chu W-S, Kastner D. Monocytic fasciitis: a newly recognized clinical feature of TNF-receptor dysfunction.
Driessler F, Venstrom K, Sabat R, Asadullah K, Schottelius AJ. Molecular mechanisms of interleukin-10-mediated inhibition of NF-kappaB activity: a role for p50.
Williamson L, Hull D, Mehta R, Reeves W, Robinson B, Toghill P. Familial Hibernian fever.
Hogan M, Vogel S.
Simon A, Bodar EJ, van der Hilst JC et al. Beneficial response to interleukin 1 receptor antagonist in TRAPS.
Drewe E, McDermott EM, Powell PT, Isaacs JD, Powell RJ. Prospective study of anti-tumour necrosis factor receptor superfamily 1B fusion protein, and case study of anti-tumour necrosis factor receptor superfamily 1A fusion protein, in tumour necrosis factor receptor associated periodic syndrome (TRAPS): clinical and laboratory findings in a series of seven patients.
Tsimberidou AM, Waddelow T, Kantarjian HM, Albitar M, Giles FJ. Pilot study of recombinant human soluble tumor necrosis factor (TNF) receptor (p75) fusion protein (TNFR:Fc; Enbrel) in patients with refractory multiple myeloma: increase in plasma TNF alpha levels during treatment.
Eason JD, Pascual M, Wee S et al. Evaluation of recombinant human soluble dimeric tumor necrosis factor receptor for prevention of OKT3-associated acute clinical syndrome.
Mohler KM, Torrance DS, Smith CA et al. Soluble tumor necrosis factor (TNF) receptors are effective therapeutic agents in lethal endotoxemia and function simultaneously as both TNF carriers and TNF antagonists.
Aderka D, Engelmann H, Maor Y, Brakebusch C, Wallach D. Stabilization of the bioactivity of tumor necrosis factor by its soluble receptors.
Scallon B, Cai A, Solowski N et al. Binding and functional comparisons of two types of tumor necrosis factor antagonists.
Evans T, Moyes D, Carpenter A et al. Protective effect of 55- but not 75-kD soluble tumor necrosis factor receptor-immunoglobulin G fusion proteins in an animal model of gram-negative sepsis.
Trikha M, Corringham R, Klein B, Rossi J. Targeted anti-interleukin-6 monoclonal antibody therapy for cancer: a review of the rationale and clinical evidence.
Nishimoto N, Yoshizaki K, Miyasaka N et al. Treatment of rheumatoid arthritis with humanized anti-interleukin-6 receptor antibody: a multicenter, double-blind, placebo-controlled trial.
Comments