Rheumatoid arthritis (RA) is the most common chronic, systemic inflammatory arthritis, manifested by inflammation of synovial joints with progressive joint destruction which ultimately leads to chronic pain, bone erosions, and progressive functional disability [
Calcitonin (CT), a 32-amino-acid monomeric peptide physiologically produced mainly from the thyroid C-cells, results from cleavage and posttranslational processing of procalcitonin (PCT) [
PCT, a 116-amino-acid precursor protein, is usually generated and cleaved to CT in the CT cells of the thyroid gland [
The activation of TLR is a double-edged sword, which is the cause of cancer, autoimmune diseases, chronic inflammation, and neurodegenerative diseases [
Cases and healthy controls were retrospectively collected from October 2018 to November 2019 by reviewing the electronic medical records. According to the study of the clinical charts, included patients were classified into 3 groups. On a total of 180 patients, 82 patients were diagnosed with RA, according to the 2010 RA Classification Criteria by the American College of Rheumatology (ACR), and 98 patients were classified as non-RA. The RA group was subdivided according to disease duration in the early RA group with disease duration ≤1 year and established RA group with disease duration >1 year. The non-RA group included 37 patients diagnosed with SLE, 30 with OA, and 31 with GA. The last group included 80 healthy individuals. All patients’ medical records were reviewed, and the relevant clinical and serological data were collected. The study was endorsed by the local ethics committee, and each patient provided informed consent before entering the study.
Serum samples were collected from the patients on medical visits to our hospital and taken aseptically by venipuncture and centrifuged within 1 h. The samples were stored frozen at -80°C until further analysis. CT and PCT were measured on a Roche Cobas e601 system (BRAHMS, Berlin, Germany) by ECLIA. The assay has a functional sensitivity of 0.06 ng/ml. RF was detected by Roche Cobas c311 system (BRAHMS, Berlin, Germany) with a range of 10-130 IU/ml, and cutoff set at 14 IU/ml. Anti-CCP and was analyzed by a commercial ELISA kit (YHLO, Shenzhen, China) and considered positive at a cutoff value of 30 AU/ml. Anti-RA33 was monitored using a commercial ELISA kit (YHLO, Shenzhen, China), and normal ranges were 0 to 25 AU/ml for anti-RA33.
We performed a statistical analysis that compared the general information of each group. Qualitative items were presented as numbers and percentages for the description of the basic characteristics, and quantitative items were identified as
In order to determine whether there is an advantage in diagnosing RA when using both tests compared to using a single RA biomarker, we undertook a ROC (receiver operating characteristic) analysis and calculated the areas under the curve (AUC). The areas under ROC curves were estimated by the nonparametric method of Mann–Whitney statistics. We evaluated the detailed diagnostic performance of PCT and CT combined with the present biomarkers of RA according to the AUC, sensitivity, specificity, positive likelihood ratios (LR), and negative likelihood ratios (LR). Confidence intervals (95%) for the AUC were performed in Medcalc statistical software version 15.2.2 (MedCalc Software bvba, Ostend, Belgium). In all analyses, a
A total of 260 patients were involved. In RA, 48 cases had early RA, 34 of them (70.8%) were females, and 34 cases with established RA aged between 44 and 84 years, 32 of them (94.1%) were females. In non-RA, 37, 30, and 31 patients had SLE, OA, and GA, respectively. The corresponding male/female ratios were 3/34, 4/26, and 30/0, respectively, and the corresponding average ages were
Demographic and clinical characteristics of the participants.
Characteristic | Early RA | Established RA | Non-RA | Healthy controls | ||
---|---|---|---|---|---|---|
SLE | OA | GA | ||||
Total | 48 | 34 | 37 | 30 | 31 | 80 |
Age (years) | ||||||
Females (%) | 34 (70.8%) | 32 (94.1%) | 34 (91.9%) | 26 (86.7%) | 0 (0%) | 42 (52.5%) |
Disease duration (years) | 0.9 (0.6–1.0) | 6.0 (3.8–11.0) | — | — | — | — |
Tender joint count | 12 (4–17) | 10 (5–20) | — | — | — | — |
Swelling joint count | 3 (0–13) | 3 (1–10) | — | — | — | — |
CRP (mg/L) | 17.91 (5.81–45.51) | 23.40 (11.11–66.43) | — | — | — | — |
ESR (mm/h) | — | — | — | — | ||
Radiographic progression, median | ||||||
Joint-space narrowing score | 1 (1–1) | 1 (1–2) | — | — | — | — |
Erosion score | 1 (1–2) | 3 (2–4) | — | — | — | — |
For normally distributed data, values were expressed as
A Mann–Whitney
(a) Distribution of the serum PCT level in early RA, established RA, SLE, OA, GA, and healthy control groups. (b) Distribution of the level of serum PCT in early RA, established RA, SLE, OA, GA, and healthy control groups. (c) Distribution of the serum RF level in early RA, established RA, SLE, OA, GA, and healthy control groups. (d) Distribution of the serum anti-CCP level in early RA, established RA, SLE, OA, GA, and healthy control groups. (e) Distribution of the serum anti-RA33 level in early RA, established RA, SLE, OA, GA, and healthy control groups. RA: rheumatoid arthritis; SLE: systemic lupus erythematosus; OA: osteoarthritis; GA: gouty arthritis; CRP: C-reactive protein; ESR: erythrocyte sedimentation rate; CT: calcitonin; PCT: procalcitonin; anti-CCP: anti-cyclic citrullinated peptide; anti-RA33: anti-RA33 antibodies.
Differences between the serum biomarkers level in the study groups.
Group | PCT, ng/ml | CT, pg/ml | RF, IU/ml | Anti-CCP, AU/ml | Anti-RA33, AU/ml |
---|---|---|---|---|---|
Median (IQR) | Median (IQR) | Median (IQR) | Median (IQR) | Median (IQR) | |
Early RA, | 0.065 (0.040–0.099) | 0.880 (0.675–1.470) | 18.8 (7.2–29.5) | 20.0 (11.8–50.6) | 6.6 (3.3–16.3) |
Established RA, | 0.076 (0.044–0.171) | 0.845 (0.659–1.515) | 90.2 (38.7–192.4) | 57.8 (10.4–177.1) | 3.3 (1.6–6.1) |
All controls, | 0.029 (0.021–0.060) | 2.480 (1.205–3.910) | 11.9 (8.8–19.7) | 4.5 (3.3–10.3) | 4.3 (2.9–6.2) |
All disease controls, | 0.056 (0.027–0.101) | 2.110 (0.873–3.070) | 12.3 (9.5–30.3) | 3.6 (2.8–4.6) | 4.1 (2.7–6.6) |
SLE, | 0.093 (0.054–0.169) | 2.480 (1.235–3.210) | 31.3 (13.2–41.3) | 3.1 (2.1–5.0) | 5.4 (2.9–9.4) |
OA, | 0.025 (0.020–0.031) | 0.586 (0.500–2.045) | 9.4 (7.5–11.1) | 4.2 (3.6–4.9) | 4.6 (3.1–5.9) |
GA, | 0.072 (0.045–0.120) | 2.550 (1.970–3.490) | 10.8 (9.2–15.8) | 3.5 (2.2–4.1) | 2.9 (1.9–4.1) |
Healthy, | 0.024 (0.017–0.028) | 3.159 (1.757–4.827) | 11.7 (6.6–18.8) | 10.4 (5.8–18.5) | 4.6 (3.4–6.1) |
The data was divided into 5 groups. Data are expressed as medians. RA: rheumatoid arthritis; SLE: systemic lupus erythematosus; OA: osteoarthritis; GA: gouty arthritis; PCT: procalcitonin; CT: calcitonin; RF: rheumatoid factor; anti-CCP: anti-cyclic citrullinated peptide; anti-RA33: anti-RA33 antibodies.
The correlation matrix provided in Tables
Correlation coefficients of PCT and CT with clinical and serological measures in early RA.
Variable | Age | Disease duration | Sharp scores | ESR | CRP | RF | Anti-CCP | Anti-RA33 | PCT |
---|---|---|---|---|---|---|---|---|---|
48 | 48 | 48 | 48 | 48 | 48 | 38 | 48 | 48 | |
PCT | 0.050 | 0.024 | -0.188 | 0.442 | 0.414 | -0.245 | 0.105 | 0.263 | — |
CT | 0.033 | -0.034 | 0.679 | -0.073 | 0.067 | -0.073 | 0.047 | 0.070 | 0.092 |
Correlation coefficients of PCT and CT with clinical and serological measures in the whole course of RA.
Variable | Age | Disease duration | Sharp scores | ESR | CRP | RF | Anti-CCP | Anti-RA33 | PCT |
---|---|---|---|---|---|---|---|---|---|
82 | 82 | 82 | 82 | 82 | 82 | 64 | 71 | 82 | |
PCT | 0.064 | 0.183 | 0.025 | 0.360 | 0.371 | 0.063 | 0.223 | 0.090 | — |
CT | -0.132 | 0.009 | 0.039 | -0.097 | 0.232 | -0.015 | 0.014 | 0.194 | 0.205 |
All correlations were established with a Spearman rank correlation as variables were nonnormally distributed according to Shapiro-Wilk normality testing. PCT: procalcitonin; CT: calcitonin; ESR: erythrocyte sedimentation rate; CRP: C-reactive protein; RF: rheumatoid factor; anti-CCP: anti-cyclic citrullinated peptide; anti-RA33: anti-RA33 antibodies;
All correlations were established with a Spearman rank correlation as variables were nonnormally distributed according to Shapiro-Wilk normality testing. PCT: procalcitonin; CT: calcitonin; ESR: erythrocyte sedimentation rate; CRP: C-reactive protein; RF: rheumatoid factor; anti-CCP: anti-cyclic citrullinated peptide; anti-RA33: anti-RA33 antibodies;
Furthermore, as shown in Figure
(a) Distribution of the serum PCT level in high-value groups and low-value groups. (b) Distribution of the serum CT level in high-value groups and low-value groups. PCT: procalcitonin; CT: calcitonin.
Correlation coefficients of the high-value groups of PCT with clinical and serological measures in early RA.
Variable | Disease duration | RF | Anti-CCP | Anti-RA33 |
---|---|---|---|---|
Low groups of PCT | 0.292 | -0.215 | 0.006 | 0.085 |
High groups of PCT | -0.093 | -0.161 | 0.225 |
Correlation coefficients of the low-value groups of CT with clinical and serological measures in early RA.
Variable | Disease duration | RF | Anti-CCP | Anti-RA33 |
---|---|---|---|---|
Low groups of CT | 0.065 | 0.291 | 0.120 | -0.174 |
High groups of CT | 0.374 | 0.018 | 0.029 | 0.072 |
All correlations were established with a Spearman rank correlation as variables were nonnormally distributed according to Shapiro-Wilk normality testing. PCT: procalcitonin; RF: rheumatoid factor; anti-CCP: anti-cyclic citrullinated peptide; anti-RA33: anti-RA33 antibodies;
All correlations were established with a Spearman rank correlation as variables were nonnormally distributed according to Shapiro-Wilk normality testing. PCT: procalcitonin; RF: rheumatoid factor; anti-CCP: anti-cyclic citrullinated peptide; anti-RA33: anti-RA33 antibodies;
ROC analysis was conducted on a single serum biomarker and combinations of serum biomarkers, and we found an impressive additional diagnostic value of PCT and CT compared with the single use of the present biomarkers alone. We compared early RA patients with controls including all controls, disease controls (SLE groups, OA groups, and GA groups), and healthy individuals.
As shown in Figure
Corresponding ROC curve for a single serum biomarker and combinations of serum biomarkers in early rheumatoid arthritis along with their respective area under the curve (AUC), comparing with all controls. (a) ROC analysis of RF, RF combined with PCT and CT. (b) ROC analysis of anti-CCP, anti-CCP combined with PCT and CT. (c) ROC analysis of anti-RA33, anti-RA33 combined with PCT and CT. (d) ROC analysis of RF and anti-CCP, RF and anti-CCP combined with PCT and CT. (e) ROC analysis of RF and anti-RA33, RF and anti-RA33 combined with PCT and CT. (f) ROC analysis of anti-CCP and anti-RA33, anti-CCP and anti-RA33 combined with PCT and CT.
As shown in Figure
Corresponding ROC curve for a single serum biomarker and combinations of serum biomarkers in early RA along with their respective area under the curve (AUC), compared with disease controls (SLE groups, OA groups, and GA groups). (a) ROC analysis of RF, RF combined with PCT and CT. (b) ROC analysis of anti-CCP, anti-CCP combined with PCT and CT. (c) ROC analysis of anti-RA33, anti-RA33 combined with PCT and CT. (d) ROC analysis of RF and anti-CCP, RF and anti-CCP combined with PCT and CT. (e) ROC analysis of RF and anti-RA33, RF and anti-RA33 combined with PCT and CT. (f) ROC analysis of anti-CCP and anti-RA33, anti-CCP and anti-RA33 combined with PCT and CT.
As shown in Figure
Corresponding ROC curve for a single serum biomarker and combinations of serum biomarkers in early rheumatoid arthritis along with their respective area under the curve (AUC), compared with healthy individuals. (a) ROC analysis of RF, RF combined with PCT and CT. (b) ROC analysis of anti-CCP, anti-CCP combined with PCT and CT. (c) ROC analysis of anti-RA33, anti-RA33 combined with PCT and CT. (d) ROC analysis of RF and anti-CCP, RF and anti-CCP combined with PCT and CT. (e) ROC analysis of RF and anti-RA33, RF and anti-RA33 combined with PCT and CT. (f) ROC analysis of anti-CCP and anti-RA33, anti-CCP and anti-RA33 combined with PCT and CT.
The sensitivity, specificity, positive likelihood ratios (LR), and negative likelihood ratios (LR) of RF, anti-CCP, and anti-RA33 with and without PCT and CT in early RA patients were summarized in Table
Evaluation of the diagnostic performance using single tests and test combinations.
Variable | Without PCT and CT | With PCT and CT | |||
---|---|---|---|---|---|
Sn/Sp | LR+/LR- | Sn/Sp | LR+/LR- | ||
Early RA vs. all controls | |||||
RF | 64.58/61.24 | 1.67/0.58 | 89.58/58.99 | 2.18/0.18 | <0.01 |
Anti-CCP | 81.58/76.97 | 3.54/0.24 | 92.11/62.92 | 2.48/0.13 | >0.05 |
Anti-RA33 | 58.33/74.72 | 2.31/0.56 | 91.67/59.55 | 2.27/0.14 | <0.05 |
RF and anti-CCP | 86.84/50.56 | 1.76/0.26 | 92.11/62.36 | 2.45/0.13 | <0.05 |
RF and anti-RA33 | 47.92/84.27 | 3.05/0.62 | 87.5/64.04 | 2.43/0.20 | <0.01 |
Anti-CCP and anti-RA33 | 76.32/82.02 | 4.25/0.29 | 97.37/58.99 | 2.37/0.05 | >0.05 |
Early RA vs. disease controls | |||||
RF | 64.58/60.20 | 1.62/0.59 | 81.25/58.16 | 1.94/0.32 | <0.01 |
Anti-CCP | 81.58/96.94 | 26.65/0.19 | 89.47/54.08 | 1.95/0.19 | <0.01 |
Anti-RA33 | 58.33/74.79 | 2.29/0.56 | 87.50/54.08 | 1.91/0.23 | >0.05 |
RF and anti-CCP | 92.11/66.33 | 2.74/0.12 | 89.47/54.08 | 1.95/0.19 | >0.05 |
RF and anti-RA33 | 66.67/62.24 | 1.77/0.54 | 83.33/58.16 | 1.99/0.29 | <0.01 |
Anti-CCP and anti-RA33 | 81.58/91.84 | 9.99/0.20 | 86.84/58.16 | 2.08/0.23 | <0.05 |
Early RA vs. healthy | |||||
RF | 33.33/100.00 | — | 83.33/100.00 | — | <0.01 |
Anti-CCP | 44.74/92.50 | 5.96/0.60 | 92.11/93.75 | 14.74/0.08 | <0.01 |
Anti-RA33 | 58.33/75.00 | 2.33/0.56 | 87.50/91.25 | 10.00/0.14 | <0.01 |
RF and anti-CCP | 57.89/86.25 | 4.21/0.49 | 92.11/93.75 | 14.74/0.08 | <0.01 |
RF and anti-RA33 | 50.00/96.25 | 13.33/0.52 | 83.33/100.00 | — | <0.01 |
Anti-CCP and anti-RA33 | 63.16/78.75 | 2.97/0.47 | 92.11/95.00 | 18.42/0.08 | <0.01 |
In the early RA group compared with disease controls, only the RF comparison group and the RF combined with the anti-RA33 group demonstrated to have substantial increases in sensitivity and only a modest decrease in specificity with the additions of PCT and CT. The sensitivities of the RF group and the RF combined with the anti-RA33 group were 64.58% and 66.67%, respectively, and the specificities were 60.20% and 62.24%, respectively. The additions of PCT and CT yielded high sensitivities of 81.25% and 83.33% and specificities of 58.16% and 58.16% (
In the early RA group compared with healthy controls, the additions of PCT and CT led to significant increases in sensitivity and specificity in all comparison groups. The sensitivities of RF, anti-CCP, and anti-RA33 for early RA were 33.33%, 44.74%, and 58.33%, respectively, and the specificities were 100.00%, 92.50%, and 75.00%. The additions of PCT and CT showed very high sensitivities of 83.33%, 92.11%, and 87.50% and modest specificities of 100.00%, 93.75%, and 91.25%. In addition, the combination of RF and anti-CCP, RF and anti-RA33, anti-CCP, and anti-RA33 showed a sensitivity of 57.89%, 50.00%, and 63.16% and a specificity of 86.25%, 96.25%, and 78.75%. The additions of PCT and CT showed very high sensitivities of 92.11%, 83.33%, and 92.11% and specificities of 93.75%, 100.00%, and 95.00%. Taking all factors together suggests that the combination of anti-CCP and anti-RA33 with the additions of PCT and CT is the most effective method in the diagnosis of early RA, which showed a sensitivity of 92.11% and a specificity of 95.00% (
RA is a chronic systemic inflammatory disease of unclear etiology that is characterized by a progressive and destructive polyarthritis in company with serological evidence of autoreactivity. It is manifested by chronic pain and joint destruction, usually progressing from the distal end to the proximal joints [
One of the most crucial problems in RA diagnosis is the absence of sensitivity or specificity of the current indicators, anti-citrullinated protein antibodies (ACPA), and RF. Moreover, ACPA is an overlapping group of antibodies depending on the citrullination of an arginine residue, which includes antiperinuclear factor (APF) [
The first important finding from our data was that the serum PCT significantly increased (
Our study was designed to evaluate not only the clinical utility of single indicators such as RF, anti-CCP, and anti-RA33 in RA but also the physical effect of every possible combination with the additions of PCT and CT, among RF, anti-CCP, and anti-RA33. The result of the ROC curve indicated the combinations of PCT, CT and RF, PCT, CT and anti-CCP, PCT, CT, and RA33 further improved the diagnostic ability for RA with the AUC of 0.97, 0.98, and 0.97, while the AUC of single indicators of RF, anti-CCP, and anti-RA33 was 0.66, 0.73, and 0.64, respectively. With the combination of PCT and CT, the sensitivity has risen, at a moderate cost of a decline in specificity, compared to a single biomarker. In our study, out of all evaluated combinations, it showed that the most clinically useful test was the union of anti-CCP and anti-RA33 with the additions of PCT and CT that can produce not only the highest sensitivity (92.11%) but also higher specificity (95.00%), with an AUC of 0.98. So we can draw a conclusion that the combined use of PCT and CT is a powerful diagnostic tool and shows a higher value for clinical use than the single detection of conventional indicators only.
Most of the studies have stated that the levels of serum PCT [
There were some negative results in our study that might contribute to the application of PCT and CT in diagnosing RA. No correlation was observed between serum concentrations of PCT, CT, and the disease duration of early RA. This indicated that the diagnostic efficiency of PCT and CT may not decrease in early RA, whereas there could be a significant decline in the sensitivity of anti-CCP when diagnosing patients with early RA rather than patients with established RA [
Our study also has a few limitations. First, many studies did not report major methodological features, like the study setting, the inclusion and exclusion criteria, or disease severity. Lacking information reduced transparency of the methods and results and made it hard to exclude bias. Second, we collected the data from a single-center, so the number of RA patients was small. Consequently, there is a possibility that the results of our study could be different from those of other centers, and the predictive probability could be overestimated compared with a prospective study. In heterogeneous diseases such as RA, small sample sizes not only limit generalizability but also ignore important associations and may restrict the number of variables that could be contained in a multivariate analysis. As a result, additional prospective studies with larger populations including multiple centers are necessary in order to confirm our conclusion. Finally, we lack the follow-up data of the patients with RA. Long-term follow-up of RA patients with abnormal PCT or CT is necessary to assess the diagnostic value of this test in patients with RA.
In conclusion, our studies identified that the combination of PCT, CT, and clinically available RA-related biomarkers could further improve the diagnostic efficiency of RA, compared with the single use of the present biomarkers alone. In addition, we observed that the sensitivities of RF, anti-CCP, and anti-RA33 for RA significantly improved with the combination of PCT and CT. At last, we recommend that the combination of RF, anti-CCP, PCT, and CT is the most effective method in the diagnosis of early RA.
No data were used to support this study.
The authors have declared that no competing interests exist.
This project was supported by research grants from the National Natural Science Foundation of China (No. 81802447).