Abstract
Objective Soluble transferrin receptor (sTfR) is considered to be a useful biomarker for the diagnosis of iron deficiency, especially in the setting of inflammation, as it is thought to not be affected by inflammation. We analyzed the relationship between sTfR levels and inflammatory markers in patients with known or suspected inflammatory rheumatic disease (IRD).
Methods Blood samples of 1001 patients with known or suspected IRD referred to a tertiary rheumatology center were analyzed. Study participants were classified as patients with active IRD and patients with inactive IRD or without IRD. Correlation analyses were used to explore the relationship between sTfR levels and inflammatory markers (ie, C-reactive protein [CRP], erythrocyte sedimentation rate [ESR]). We applied multiple linear regression analysis to evaluate the predictive value of CRP levels for sTfR concentrations after adjustment for potential confounding factors.
Results There were positive correlations between inflammatory markers (CRP, ESR) and serum sTfR levels (ρ 0.44, ρ 0.43, respectively; P < 0.001), exceeding the strength of correlation between inflammatory markers and the acute phase reactant ferritin (ρ 0.30, ρ 0.23, respectively; P < 0.001). Patients with active IRD demonstrated higher serum sTfR levels compared to patients with inactive or without IRD (mean 3.99 [SD 1.69] mg/L vs 3.31 [SD 1.57] mg/L; P < 0.001). After adjustment for potential confounding factors, CRP levels are predictive for serum sTfR concentrations (P < 0.001).
Conclusion The study provides evidence against the concept that sTfR is a biomarker not affected by inflammation.
Anemia of inflammation, also known as anemia of chronic disease (ACD), is the most frequent anemic entity observed in patients experiencing chronic inflammatory rheumatic diseases (IRDs) with prevalence rates varying among different rheumatic diseases.1 In rheumatoid arthritis (RA) as a prototype of a chronic inflammatory disease, for example, anemia is one of the most common extraarticular manifestations, with a reported prevalence between 15% and 60%, despite the declining prevalence of anemia in the era of biologic disease-modifying antirheumatic drugs (DMARDs).2-4 It is well established that anemia in IRD is related to higher disease activity, disease progression, worse patient-related outcome measures, and increased mortality.5-7 Although the decisive treatment for ACD in systemic rheumatic diseases is remission of the underlying disease, anemia in rheumatic diseases because of iron deficiency, vitamin deficiency, or treatment-related anemia requires different diagnostic and therapeutic interventions.8 Therefore, it is critical to differentiate between ACD and anemia because of other or coexisting causes beyond inflammatory immune-driven mechanisms.
Patients with chronic inflammatory diseases frequently experience a combination of ACD and iron deficiency anemia (IDA). In RA, for example, it is estimated that iron deficiency contributes to anemia prevalence in 30% to 50% of cases, and in certain populations of patients with well-controlled disease, IDA was found to be more common than ACD.2,9 In the absence of inflammation, serum ferritin as an indicator of total body iron stores is the most useful parameter to differentiate ACD from IDA.1,10,11 However, in acute and chronic inflammatory disorders, high concentrations of serum ferritin result from increased secretion by iron-retaining macrophages. Further, serum ferritin is an acute phase protein that is induced by inflammatory mediators.1,3 Thus, in inflammatory states, ferritin loses its diagnostic value as an indicator of total iron body stores.
The main challenge in ACD is identifying patients with concomitant true iron deficiency, as these patients need specific evaluation for gastrointestinal blood loss and iron-targeted management strategies.
The soluble transferrin receptor (sTfR), a truncated monomer of tissue transferrin receptor 1 (TfR1) expressed by almost all proliferating cells, is the biomarker most frequently used in clinical practice to detect iron deficiency in the presence of inflammation. Although sTfR is increased in IDA, it is considered to be normal in ACD. Thus, elevated sTfR levels in the setting of ACD suggest the presence of additional IDA.12-15 This interpretation is based on the predominant scientific view that sTfR levels are not affected by inflammation and remain within the normal range in acute or chronic inflammatory diseases.11,16-21
Several clinical studies have reported increases in sTfR during inflammation,22-24 contrasting with the established view of sTfR as a biomarker not affected by inflammation. Additionally, more recent large-scale epidemiologic and cross-sectional studies demonstrate significant correlations between sTfR levels and markers of inflammation and the necessity of adjusting sTfR concentrations for inflammation to accurately assess the prevalence of iron deficiency in a population.25-28
Against the background of the conflicting information in the literature regarding the effect of inflammation on serum sTfR levels, the present study analyzed the relationship between markers of inflammation and serum sTfR concentrations in patients with IRDs.
METHODS
Patient characteristics. The study population consisted of 1001 inpatients and outpatients referred to a tertiary care rheumatology center with known IRDs or for diagnostic workup of a suspected IRD between June 2019 and November 2021.
Patients were classified as patients with an active IRD (“active patients”) and patients with an inactive IRD or without an IRD (“inactive patients”). Activity of an IRD was defined by the presence of clinical or laboratory manifestations of the disease requiring the initiation, intensification, or adjustment of antiinflammatory or immunosuppressive therapy with the purpose of better disease control. Subjects who had received concentrated red blood cells within the past 3 months were excluded from the study.
Ethics. The local ethics committee of the University of Regensburg approved the study (approval number 12-101-0074) and written informed consent was obtained from all participants. The study was conducted in compliance with the Declaration of Helsinki, International Conference on Harmonization Good Clinical Practice Guideline, and local country regulations.
Laboratory analysis. Nonfasting blood samples were collected. sTfR serum levels were determined using a commercial particle-enhanced immunoturbidimetric assay (Tina-quant Soluble Transferrin Receptor II, Roche Diagnostics). The assay was performed on a cobas c 501 analyzer (Roche Diagnostics). The limit of detection of the test was 0.40 mg/L (4.72 nmol/L). The average intraassay and interassay coefficient of variation was 1.5% and 1.7%, respectively. Serum ferritin levels were measured using a commercial particle-enhanced immunoturbidimetric assay (Tina-quant Ferritin Gen 4, Roche Diagnostics) performed on a cobas c 501 analyzer (Roche Diagnostics). Iron, plasma transferrin concentration, and C-reactive protein (CRP) were measured on a cobas c 501 analyzer (Roche Diagnostics; iron by colorimetric assay, and transferrin and CRP by immunoturbidimetric assay). The percent plasma transferrin saturation was calculated using the following formula: (serum iron [μg/dL]/serum transferrin [mg/dL]) × 70.9. Blood counts were measured with an automated hematology analyzer (XN-1000-analyzer, Sysmex). The erythrocyte sedimentation rate (ESR) was determined by the Westergren method using an SRS 100/II analyzer (Electa-Lab S.r.l.).
Statistical analysis. Results were analyzed using the Statistical Package for Social Sciences for Windows, version 25.0 (SPSS Inc.). Correlations between sTfR, traditional measures of iron metabolism, and inflammatory markers were analyzed using Spearman ρ correlation analysis. Correlation coefficients between 0.0 and 0.3 indicate a weak positive relationship, values between 0.3 and 0.7 indicate a moderate positive relationship, and values between 0.7 and 1.0 indicate a strong positive linear relationship. Mann-Whitney U test was used to analyze differences of the median between patients with active and inactive disease. Multilinear regression analysis was employed with sTfR as a dependent variable and hemoglobin, standard measures of iron metabolism, CRP, creatinine, and age as independent variables. P < 0.05 was considered significant.
RESULTS
Characteristics of patients. Serum samples of 1001 patients referred to a tertiary care rheumatology center with known IRDs or for diagnostic workup of a suspected IRD were analyzed for sTfR and standard measures of iron deficiency. Characteristics of patients under study are displayed in Table 1.
Characteristics of patients under study.
Correlation of sTfR with inflammatory markers and standard measures of iron metabolism. Both sTfR and standard measures of iron deficiency correlated significantly with inflammatory markers (CRP, ESR). sTfR, iron, transferrin, and transferrin saturation demonstrated moderate correlations with inflammatory markers, whereas ferritin demonstrated weak correlations (Table 2).
Correlation coefficients (Spearman ρ) of inflammatory markers (CRP, ESR) and sTfR and standard variables of iron deficiency.
Figure 1 shows the relation between sTfR levels and CRP deciles with increasing sTfR levels as the concentrations of CRP increase.
Serum soluble transferrin receptor levels of 1001 patients from a tertiary rheumatology center by CRP deciles. CRP: C-reactive protein.
Weak to moderate correlations between sTfR and standard markers of iron deficiency (ie, ferritin, iron, transferrin, and transferrin saturation), as well as between sTfR and hemoglobin, were determined as depicted in Table 3. There was no significant correlation between sTfR and sex (ρ −0.02, P < 0.55) or between sTfR and serum creatinine (ρ 0.05, P < 0.09). A weak correlation was found between sTfR and age (ρ 0.17, P < 0.001).
Correlation coefficients (Spearman ρ) of sTfR and standard variables of iron deficiency and hemoglobin.
Comparison of patients with active and inactive disease. An active IRD was present in 524 (52.3%) patients. Serum sTfR levels in patients with an active IRD (mean 3.99 [SD 1.69] mg/L) were significantly elevated compared to patients with an inactive rheumatic disease or without IRD (mean 3.31 [SD 1.57] mg/L; P < 0.001; Figure 2).
Serum levels of soluble transferrin receptor in patients with active IRD (“active”) and in patients with inactive or without IRD (“inactive”) as box plots with the 90th, 75th, 50th (median), 25th, and 10th percentiles. Group medians were compared by the Mann-Whitney U test. IRD: inflammatory rheumatic disease.
Patients with an active IRD had higher CRP and ESR levels, older age, and lower hemoglobin levels compared to patients with an inactive disease or without IRD (data not shown; P < 0.001 for each comparison). Serum creatinine and sex did not differ significantly between patients with an active vs an inactive or nonexisting IRD (data not shown).
Multilinear regression analysis. Multilinear regression analysis with hemoglobin, standard measures of iron metabolism, CRP, creatinine, and age as independent variables was employed to evaluate the predictive value of inflammation (operationalized by CRP) for sTfR levels as the dependent variable (Table 4). Hemoglobin, transferrin saturation, CRP, creatinine, and age were found to be predictors for sTfR levels (P < 0.001), independent of all included variables in the model. The 7 included variables in Table 4 explain 84.8% of the variation in sTfR serum levels (Table 4).
Predictors of serum sTfR as analyzed by multilinear regression analysis.
DISCUSSION
Although ACD is the prototypical type of anemia in chronic autoimmune inflammatory diseases, the detection of IDA alone or concomitant ACD and IDA in inflammatory diseases is critical as these types of anemia require different diagnostic and therapeutic interventions. As circulating ferritin, the hallmark indicator for body iron stores, is positively influenced by inflammation,1,3 the diagnosis of IDA or ACD with concomitant true iron deficiency in inflammatory states is challenging.
Among several markers studied for their potential to detect true iron deficiency in inflammatory states, sTfR is the most frequently used biomarker in clinical practice considered to be unaffected by inflammation and to remain within the normal range in acute or chronic inflammatory diseases.11,16-21,29 Thus, as sTfR is increased in IDA and considered to be normal in inflammatory states, elevated sTfR levels in the setting of inflammation are regarded as indicators of additional IDA.12-15
In the present study, we found that both sTfR and standard measures of iron deficiency correlated significantly with inflammation measures (ie, CRP, ESR). Moderate to strong correlations were observed between sTfR, transferrin, transferrin saturation, iron, and inflammation measures, and weak correlations between ferritin and markers of inflammation. Serum sTfR levels in patients with an active IRD were significantly elevated compared to patients with an inactive rheumatic disease or without an IRD. Serum creatinine and sex did not differ significantly between both groups of patients.
Besides expected elevated inflammatory measures in patients with an active IRD, there were also significant differences regarding hemoglobin and standard measures of iron metabolism with lower hemoglobin, transferrin, transferrin saturation, and serum iron levels, and higher serum ferritin concentration in patients with an active IRD compared to patients with an inactive IRD or without an IRD.
Using multilinear regression analysis to evaluate the relationship between sTfR serum levels and serum CRP concentrations after consideration of potential confounding factors, we demonstrated CRP concentration to be predictive for sTfR levels. The positive predictive value of hemoglobin for sTfR levels reflects that the serum concentration of sTfR is proportional to the total body cellular TfR1,18 and that the total cellular TfR1 depends on the number of erythroid precursors in the bone marrow.21
The positive predictive value of creatinine for sTfR levels is in accordance with the results of large-scale cross-sectional and cohort studies, revealing an association with chronic kidney disease and increased serum sTfR levels regardless of anemia and iron storage status and a negative correlation between estimated glomerular filtration rate and serum sTfR levels.28,30
Finally, age is clearly correlated with increasing inflammation, which has been previously demonstrated in the 1990s in a carefully selected group of healthy controls.31 Moreover, the higher inflammatory load of older people is well known. Although CRP and interleukin (IL)-6 are tightly interrelated because IL-6 induces CRP from hepatocytes,32 other age-related inflammatory factors add to the overall higher inflammatory load of older people. Thus, age might well be an independent factor for sTfR serum levels in our multilinear regression analysis, independent of IL-6-stimulated CRP.
Besides the evidence at the cellular level that TfR1 expression and serum sTfR levels, which are proportional to the cellular mass of TfR1,18 are affected by inflammation,33-39 there are clinical studies demonstrating a positive correlation between sTfR levels and markers of inflammation or increased sTfR levels in inflammatory diseases.23,24,40 Kasvosve et al found CRP levels to predict sTfR concentration in 147 children in an area of Zimbabwe where malaria and hookworm were not endemic.22 Additionally, more recent large-scale epidemiologic and cross-sectional studies demonstrate significant correlations between sTfR serum levels and markers of inflammation and the necessity of adjusting sTfR concentrations for inflammation to accurately assess the prevalence of iron deficiency in a population.25-28
Our study has limitations that should be taken into account. First, the study population consists largely of patients experiencing IRDs, questioning the transferability of the results of the present study to patient populations with nonautoimmune inflammatory or malignant diseases. Second, the study does not provide any information about the distribution of different types of anemia, nor to what extent the results gained by the study concerning the relationship between sTfR and inflammatory markers are reproducible in different types of anemia (eg, patients with ACD vs patients with IDA). Finally, despite the high proportion of patients receiving conventional or biologic DMARD therapy, the influence of immunosuppressive therapy on the relationship between sTfR concentration and inflammation—in particular, the effect of IL-6-antagonizing drugs interfering with the expression of hepcidin as key player in iron metabolism—was not addressed.
In contrast to the previous concept that sTfR levels are not affected by inflammation, our study revealed a positive correlation between sTfR and inflammatory markers exceeding the strength of the correlation between the acute phase reactant ferritin and inflammatory markers. Even after adjustment for potential confounding factors, CRP levels remained predictive for sTfR concentrations. The influence of inflammation on sTfR levels provides evidence against the argument for using sTfR concentrations as an indicator of iron deficiency in the setting of inflammation and for the necessity to adjust sTfR concentration for inflammation to allow its correct interpretation.
ACKNOWLEDGMENT
We wish to thank all the patients participating in this study for their support.
Footnotes
The authors declare no conflicts of interest relevant to this article.
- Accepted for publication December 3, 2023.
- Copyright © 2024 by the Journal of Rheumatology
REFERENCES
DATA AVAILABILITY
The datasets used and/or analyzed during the current study are available from the corresponding author upon reasonable request.
