Table 1 Overview of studies measuring IL-1 in patients reporting fatigue
From: Interleukin-1 as a mediator of fatigue in disease: a narrative review
Reference | Disease activity | Number of patients | Fatigue questionnaire | IL-1 measurement | Main outcome |
|---|---|---|---|---|---|
Rheumatoid arthritis | |||||
 Lampa et al., 2014 | No neurological disease or generalized pain nor of swollen joints 4.9 ± 3.8 | Patients (n = 14), controls (n = 12) | VAS-fatigue | CSF IL-1 and IL-1Ra | Higher IL-1β and lower IL-Ra in RA vs controls (p < 0.001, p < 0.05), positive correlation between IL-1β and fatigue (R = 0.55, p < 0.05) |
Sjögrens syndrome | |||||
 Harboe et al., 2009 | No acute illness in the week prior to or after sampling, no CRP/ESR elevation | Patients (n = 54), controls (n = 53) | FSS, VAS-fatigue | CSF IL-1β, IL-1Ra, and IL-1sRII | Higher IL-1Ra in patients (p = 0.026), correlation IL-1Ra and VAS-fatigue (R 2 = 0.11, p = 0.015). |
Sarcoidosis | |||||
 Korenromp et al., 2011 | No disease activity | Fatigued patients (n = 34), non-fatigued patients (n = 38) | CIS-fatigue (severe fatigue when ≥35) | Plasma IL-1α, IL-1β, and IL-1Ra | No significant differences |
|  |  |  |  | Whole blood production of IL-1α and IL-1β after stimulation with LPS (1 ng/ml) |  |
 Baydur et al., 2010 | Pulmonary sarcoidosis | Patients (n = 22), controls (n = 22) | MFI-20 | Plasma IL-1β before, directly after and 4–6 h after exercise | Higher fatigue scores in sarcoidosis patients (p < 0.0001). IL-1β not different between patients and controls or among the three collection times. Correlation between pre-exercise IL-1β and MFI-20 in patients receiving immunomodulatory medication (R 2 0.63, p = 0.03). |
Cancer | |||||
Mixed (cancer survivors plus advanced cancer) | |||||
 De Raaf et al., 2012 | Advanced cancer or 1–5 years post cancer treatment | Advanced cancer (n = 45), cancer survivors (n = 47) | MFI | Plasma IL-1Ra | Advanced cancer patients had higher IL-1Ra concentrations (p < 0.01). In these patients, physical fatigue was correlated with IL-1Ra (r = 0.32, p = 0.03). In cancer survivors, IL-1Ra was related to both physical (r = 0.24, p = 0.10) and mental fatigue (r = 0.35, p = 0.02). |
Prostate cancer | |||||
 Greenberg et al.,1993 | Men undergoing localized radiotherapy | Patients (n = 15) | VAS-fatigue/daily during 8 weeks | Serum IL-1β, at baseline and weekly thereafter | A rise in fatigue was seen between weeks 1 and 4, fatigue stabilized during week 5 and increased again in weeks 6 and 7. Rise in fatigue during the first 4 weeks was accompanied by increased IL-1β concentrations (p-value not reported). |
 Bower et al., 2009 | Patients undergoing external beam radiation therapy | Prostate cancer (n = 20), breast cancer (n = 28) | FSI/fatigue during the past week/baseline, after 5/10/20 days of treatment, final week of treatment, and 2 weeks and 2 months after treatment | Serum IL-1β, IL-1Ra in a subset of patients, at same time-points as the questionnaires | Fatigue increased in both groups during treatment. Significant quadratic trend for IL-1β during treatment (p = 0.034). Treatment dose was not associated with IL-1β and IL-1Ra concentrations. There was no correlation between IL-1β and fatigue severity. IL-1Ra was associated with fatigue (β = 0.63, p = 0.016). |
 Dirksen et al., 2014 | Non-metastatic cancer prior to radiation therapy | Patients (n = 30) | POMS fatigue (inertia subscale)/pre-treatment en post-treatment | Serum IL-1β, pre-treatment and post-treatment (<2 weeks after radiotherapy, <10 weeks after brachytherapy) | Fatigue was increased post-treatment (p = 0.027). No differences in IL-1β concentrations, no correlation with fatigue severity |
 Jim et al., 2012 | Non-metastatic or asymptomatic metastatic prostate cancer | Patients (n = 53) | FSI (fatigue over the past week)/at baseline and after 6 months | SNP in IL1B gene (rs16944) | IL1B had no significant effect on fatigue-related outcomes |
Breast cancer | |||||
 Geinitz et al., 2000 | Women undergoing postoperative radiotherapy (no chemotherapy), without metastatic disease | Patients (n = 41) | FAQ, and VAS-fatigue/during previous week/at baseline, end of weeks 1–5, and 2 months after treatment | Serum IL-1β, same time points as questionnaires | VAS-fatigue increased until week 4 (p < 0.001). During weeks 4 and 5 FAQ physical (p = 0.035 and 0.015) and cognitive (p = 0.008 and 0.007) subscales were significantly elevated. IL-1β did not increase during treatment. |
 Von Ah et al., 2008 | Stage 0–IIIa breast cancer before adjuvant therapy | Patients (n = 44) | Piper-fatigue scale/at baseline and at 3 months (during adjuvant therapy) and 6 months after baseline (initial recovery) | Whole blood production of IL-1β after stimulation with PHA (10 μg/ml) | IL-1β predicted fatigue before adjuvant therapy (β = 0.30, p < 0.05). |
 Liu et al., 2012 | Stage I–III breast cancer prior to ≥4 3-week cycles of chemotherapy | Patients (n = 53) | MFSI-SF/fatigue during past week/at baseline and during cycles 1 and 4 of chemotherapy (last 2 weeks) | Plasma IL-1Ra, at the same time points as questionnaires | Fatigue significantly increased over time (p < 0.05). IL-1Ra dropped at cycle 1 week 3 (p < 0.0001). There was no association between IL-1Ra and fatigue. |
 Schmidt et al., 2015 | Stage 0–III breast cancer prior to adjuvant radiation therapy | Patients (n = 92) | FAQ/at baseline, after completion of radiotherapy (week 7), and the end of the intervention (week 13, resistant exercise/relaxation) | Serum IL-1Ra, at the same time points as questionnaires | Moderate correlation between IL-6/IL-1Ra at the end of radiotherapy with physical fatigue at the same time (r = 0.25, p = 0.022) and at 6 weeks after chemotherapy (r = 0.23, p = 0.046). |
 Bower et al., 2002 | Stage 0–II breast cancer 1–5 years after diagnosis, after completion of treatment | Fatigued (n = 20), non-fatigued (n = 20) | Energy/fatigue subscale RAND-36 (score 0–50 = high fatigue, score 70–100 = low fatigue)/fatigue during past 4 weeks FSI/fatigue during past week | Serum IL-1β and IL-1Ra | Fatigued women had significantly higher IL-1Ra concentrations (p = 0.006). |
 Bower et al., 2011 | Stage 0–IIIA breast cancer, after completion of primary cancer therapy (within past 3 months) i.e., surgery, radiation, and/or chemotherapy | Patients (n = 103) | FSI (cut-off 3)/fatigue during the past week | Plasma IL-1Ra | 64% scored above 3 on the FSI; these patients did not have a higher IL-1Ra concentration. There was no significant association between IL-1Ra and fatigue or chemotherapy exposure. |
 Bower et al., 2007 | Stage 0–II breast cancer survivors (6.5–10 years after diagnosis) | Fatigued (n = 10), non-fatigued (n = 15) | Vitality scale SF-36 (<50 = significant fatigue, >70 = absence of significant fatigue) | Whole blood production of IL-1β after stimulation with LPS (100 pg/ml) or cortisol (0, 10−8, 10−7, 10−6M), at baseline, directly after TSST, and after 30 min recovery | No differences at baseline. IL-1β increased significantly in fatigued patients after completion of the TSST (p = 0.02). |
 Collado-Hidalgo et al., 2006 | Stage 0–III breast cancer survivors, 1–5 years post-diagnosis | Fatigued (n = 32), non-fatigued (n = 18) | Vitality scale SF-36 (<50 = significant fatigue, >70 = absence of significant fatigue) | Plasma IL-1Ra | IL-1Ra was significantly higher in fatigued breast-cancer survivors (p = 0.05). |
 Orre et al., 2011 | Stage II–III breast cancer patients, 2.7–7.2 years after postoperative locoregional radiotherapy | Patients (n = 299) | Fatigue questionnaire | Serum IL-1Ra | There was no significant association between IL-1Ra and fatigue. |
 Collado-Hidalgo et al., 2009 | Stage 0–III breast cancer survivors, 1–5 years post-diagnosis | Fatigued (n = 33), non-fatigued (n = 14) | Vitality scale SF-36 (≤55 = significant fatigue, >70 = absence of significant fatigue), MFSI | IL-1B-511 (CT) polymorphism | Fatigued survivors had a substantial overrepresentation of CC alleles, and underrepresentation of TT alleles. The prevalence of at least one cytosine was more frequent among fatigued patients (p = 0.007) and associated with fatigue in multiple regression (p = 0.021). Which was no longer significant after controlling for depressive symptoms (p = 0.052). |
 Reinertsen et al., 2011 | Stage II–III breast cancer survivors | Fatigued (n = 101), non-fatigued (n = 201) | Fatigue questionnaire (cut-off 4, clinical significant fatigue), chronic fatigue was defined as fatigue being present for at least 6 months | IL-1B rs16944 (A/G) SNP, and IL-1β mRNA expression | There was no association between chronic fatigue and the IL-1B SNP or IL-1β mRNA expression. |
Testicular cancer | |||||
 Orre et al., 2009 | Patients 5–20 years after unilateral orchiectomy | Fatigued (n = 92), non-fatigued (n = 191) | Fatigue questionnaire (cut-off 4, clinical significant fatigue), chronic fatigue was defined as fatigue being present for at least 6 months | Plasma IL-1Ra | Fatigued patients had significant higher IL-1Ra (p = 0.002). In multiple regression analysis, IL-1Ra corrected for age had an OR of 1.93 (95%CI 1.21–3.08). Although age an IL-1Ra explained only 4% of the variance. IL-1Ra was not included in the final model. |
Uterine cancer | |||||
 Ahlberg et al., 2004 | Patients receiving external radiation therapy after hysterectomy | Patients (n = 15) | MFI-20/at baseline, after 30Gy (+3 weeks) and after 46Gy (+5–6 weeks) | Plasma IL-1 (α or β unknown), same time-points as questionnaires | Fatigue increased during treatment, IL-1 remained below the detection limit during the entire study period (4 pg/ml). |
AML/MDS | |||||
 Meyers et al., 2005 | Newly diagnosed AML/MDS before undergoing chemotherapy. | Patients (n = 54) | Brief fatigue inventory (cut-off score ≥4, moderate-severe fatigue)/fatigue in the past 24 h/baseline and after 1 month of treatment | Plasma IL-1 (α or β unknown) and IL-1Ra, at baseline. | There was a positive correlation of IL-1Ra and fatigue (r = 0.52, p value not reported). |
Post-stroke fatigue | |||||
 Ormstad et al., 2011 | Acute stroke patients | Patients (n = 45) | FSS (dichotomized as a score ≥4 or <4)/at 6, 12, and 18 months after stroke | Serum IL-1β and IL-1Ra, <24h (n = 35), 24–48 h (n = 7), and 48–72 h (n = 3) after stroke onset | Significant correlation between IL-1β and fatigue at 6 months (r = 0.37, p = 0.015). Negative correlation between IL-1Ra and fatigue at 12 months (r = −0.38, p = 0.013). Fatigued patients had significant lower IL-1Ra concentrations. |
 Becker et al., 2015 | Acute stroke patients | Patients (n = 39) | FAS/30/90/180/365 days after stroke | IL1RN SNP rs4251961 | Carriers of a C allele reported more fatigue (p = 0.03). At 30 and 90 days, patients with at least one C allele had higher scores on fatigue (p < 0.05). |
CFS | |||||
 Hornig et al., 2015 | CFS | Patients (illness duration ≤3 years n = 52, illness duration >3 years n = 246), controls (n = 348) | MFI | Plasma IL-1α, IL-1β and IL-1Ra | There were no differences when comparing all patients combined to controls. However, patients with a short illness duration had significantly higher IL-1α (p < 0.05) and IL-1Ra (p < 0.05) compared to controls. In patients with a long illness duration, IL-1β was significantly lower compared to controls (p < 0.05). IL-1α, IL-1β and IL-1Ra were higher in short illness patients compared to long illness patients (p < 0.01). |
 Russell et al., 2016 | CFS (female) | Patients; 1. ≤18/illness duration ≤2 years (n = 18), 2. age 18–50/average illness duration 7 years (n = 22), 3. age ≥50 and average illness duration 11 years (n = 28), controls (n = 81) | Chalder fatigue in adolescents, and MFI in other patients | Plasma IL-1α and IL-1β | Looking at individual expression, there were no differences between patients and controls. IL-1α appeared in a linear classification model in the adolescent group, but not in the other 2 groups. |
 Hardcastle et al., 2015 | Moderate (mobile) or severe (housebound) CFS | Moderate CFS (n = 22), severe CFS (n = 19), controls (n = 22) | FSS | Serum IL-1β and IL-1Ra | Significant IL-1β increase in moderate compared with severe CFS patients (p = 0.002). For other subgroups and IL-1Ra there were no differences. |
 Landi et al., 2016 | CFS | Patients (n = 100), controls (n = 79) | MFI | Plasma IL-1α and IL-1β | No significant differences. |
 Chao et al., 1991 | CFS | Patients (n = 9), controls (n = 7) | VAS-fatigue | Serum IL-1β | No differences in serum IL-1β. IL-1β production after LPS stimulation was significantly higher in CFS patients (p < 0.05) |
|  |  |  |  | PBMC production of IL-1β after stimulation with LPS (1 ng/ml) or PHA (4 μg/ml) |  |
 Swanink et al., 1996 | CFS | Patients (n = 76), controls (n = 69) | CIS | Plasma IL-1α, IL-1β, and IL-1Ra | No differences in circulating cytokine concentrations. Significant lower IL-1β production after LPS stimulation (p < 0.05), no correlation between production and fatigue severity. |
|  |  |  |  | Whole blood production of IL-1α, IL-1β, and IL-1Ra after stimulation with LPS |  |
 Mawle et al., 1997 | CFS | Patients (n = 26), controls (n = 50) | – | PBL production of IL-1α and IL-1β after stimulation with PHA | IL-1α production was lower in severely ill patients (n = 13) and those with a gradual disease onset (n = 17) compared to controls (p = 0.038, p = 0.011). IL-1β was also lower in patients with a gradual disease onset (p = 0.039). |
 Cannon et al., 1997 | Sudden onset CFS | Patients (n = 16), controls (n = 15) | – | PBMC production of IL-1β, IL-1Ra, and IL-1sRII after stimulation with LPS (1 ng/ml), indomethacin, or a combination, before and daily after a 15 min exercise on day 2 | At baseline, controls had a significant increase in IL-1β production during the luteal phase (unstimulated, p = 0.021). This increase was absent in CFS patients. In the follicular phase, control group had an increase IL-1β production 48 h after exercise. In CFS patients, there was no alteration over time. In the follicular phase, IL-1Ra secretion was higher in CFS patients (unstimulated, p = 0.023). IL-1sRII was higher in patients (unstimulated, p = 0.0002). |
 Tomoda et al., 2005 | CFS | Patients (n = 15), controls (n = 23) | – | IL-1β production of PBMCs after stimulation with PHA (5 μg/ml) or LPS (50 ng/ml) | No significant differences. |
 Lloyd et al., 1991 | CFS | Patients (n = 25), controls (n = 28) | – | Serum and CSF IL-1β | No significant differences. |
 Peterson et al., 2015 | CFS | Patients (n = 18), controls (n = 5) | – | CSF IL-1β and IL-1Ra | No significant differences. |
 Natelson et al., 2005 | CFS | Patients (n = 44), controls (n = 13) | MFI | CSF IL-1α and IL-1β | No significant differences. |
 Hornig et al., 2016 | CFS | Patients (n = 32), MS controls (n = 40), and controls (n = 19) | – | CSF IL-1α, IL-1β and IL-1Ra | CFS patients had significant lower IL-1β and IL-1Ra concentrations compared to normal controls (p = 0.003 and p = 0.014). And compared to MS patients IL-1α (p = 0.0007), IL-1β (p = 0.0018) and IL-1Ra (p = 0.0003) were decreased in CFS. |
(Post-)infectious fatigue | |||||
 Vollmer-Conna et al., 2004 | Patients with acute Q-fever, EBV, or RRV | Q-fever (n = 18), EBV (n = 24), RRV (n = 24) | Physical symptom checklist/fatigue in the past 2 weeks | Serum IL-1β | Fatigue was reported in 100% of Q-fever patients, >75% of EBV patients, and >50% of RRV patients. In Q-fever, IL-1β correlated significantly with fatigue (r = 0.47, p = 0.04), which was also found in the EBV/RRV combination group (r = 0.39, p = 0.01). All significant results were obtained from the unstimulated samples. |
|  |  |  |  | PBMC production of IL-1β after stimulation with LPS (10 ng/ml) |  |
 Vollmer-Conna et al., 2007 | Patients with post-infectious fatigue and post-infectious patients without fatigue | EBV (n = 11), RRV (n = 6), Q-fever (n = 5), and controls after EBV (n = 17), RRV (n = 14) or Q-fever (n = 11) | Somatic and psychological health report (fatigue was defined as a score ≥3 on the SOMA subscale)/at 1, 2, 3, 6, and 12 months after onset of the infection | Serum IL-1β | No significant differences. |
|  |  |  |  | PBMC production of IL-1β after stimulation with LPS (10 ng/ml), mouse anti-human or anti-CD3 |  |