One hundred ninety-five rheumatoid arthritis (RA) cases, 87 healthy controls, and 97 systemic lupus erythematosus (SLE) controls were studied. Cutoffs to define positivity were ≥5 units/ml for anti–cyclic citrullinated peptide (anti-CCP), ≥10.8 IU/ml for IgA rheumatoid factor (IgA-RF), ≥9.5 IU/ml for IgM-RF, and ≥9.6 IU/ml for IgG-RF. Of the 75 RA cases who were negative for anti-CCP, 19 (25%) were positive for IgA-RF, 21 (28%) were positive for IgM-RF, and 12 (16%) were positive for both; of the 49 RA cases who were negative for both IgA-RF and IgM-RF, only 2 were positive for anti-CCP; of the 195 total RA cases, 7 were positive for IgA-RF and negative for both IgM-RF and anti-CCP, and 9 were positive for IgM-RF and negative for both IgA-RF and anti-CCP. 95% CI = 95% confidence interval.
Anti–cyclic citrullinated peptide antibody and rheumatoid factor isotypes in African Americans with early rheumatoid arthritis
Article first published online: 31 AUG 2006
Copyright © 2006 by the American College of Rheumatology
Arthritis & Rheumatism
Volume 54, Issue 9, pages 3057–3059, September 2006
How to Cite
Mikuls, T. R., Holers, V. M., Parrish, L., Kuhn, K. A., Conn, D. L., Gilkeson, G., Smith, E. A., Kamen, D. L., Jonas, B. L., Callahan, L. F., Alarcón, G. S., Howard, G., Moreland, L. W. and Bridges, S. L. (2006), Anti–cyclic citrullinated peptide antibody and rheumatoid factor isotypes in African Americans with early rheumatoid arthritis. Arthritis & Rheumatism, 54: 3057–3059. doi: 10.1002/art.22200
- Issue published online: 31 AUG 2006
- Article first published online: 31 AUG 2006
- NIH. Grant Numbers: N01-AR-02247, K23-AR-050004
- Arthritis Foundation
- NIH. Grant Numbers: U19-A150864, R01-AR-051394
- General Clinical Research Center (National Center for Research Resources). Grant Number: M01-RR-00032
Although measurements of rheumatoid factor (RF) and anti–cyclic citrullinated peptide (anti-CCP) are frequently used in the evaluation of patients with inflammatory arthritis, studies examining the metric properties of both RF isotypes and anti-CCP antibodies have almost exclusively involved populations of Caucasian/European ancestry, and thus their generalizability to non-Caucasian patients is largely anecdotal. In the study described herein, we examined the metric properties of RF isotypes and anti-CCP antibody in a well-characterized population of African American patients with RA.
Rheumatoid arthritis (RA) cases were participants in the ongoing Consortium for the Longitudinal Evaluation of African-Americans with Early Rheumatoid Arthritis (CLEAR) (1). Subjects were enrolled through 4 sites in the southeastern US. All cases satisfied the American College of Rheumatology (ACR) classification criteria for RA (2), were of self-reported African American race/ethnicity, and had a disease duration of <2 years.
Two groups of control subjects were studied. Healthy African American controls included volunteers evaluated as part of the CLEAR. African American systemic lupus erythematosus (SLE) controls were active participants in the ongoing SLE in Gullah Health Study (3) and satisfied ACR classification criteria for SLE (4). Healthy controls were recruited from a pool of control subjects in the Birmingham, Alabama area or from a random-digit dialing sample from throughout Alabama, with both groups of controls matched to cases based on age and sex. The African American cases with RA (mean ± SD age 51 ± 13 years) were older than the healthy controls (45 ± 14 years) and those with SLE (39 ± 12 years) and, as with controls, were predominantly women (81% of the RA group, 82% of the healthy controls, and 90% of the SLE controls).
Anti-CCP (IgG) antibodies were measured using a commercially available second-generation (anti-CCP2) enzyme-linked immunosorbent assay (ELISA) kit (Diastat; Axis-Shield Diagnostics, Dundee, UK). The assay was performed according to the manufacturer's instructions. Anti-CCP antibodies were measured in arbitrary units per milliliter and were considered to be positive if present at levels of ≥5 units/ml. The cutoff value for anti-CCP antibody positivity was confirmed at the laboratory of one of the authors (VMH), using serum collected from a de-identified control population (200 organ/tissue donors) that demonstrated a positivity rate of 1.5%. RF isotypes (IgA, IgM, and IgG) were measured in international units per milliliter and were assessed using a commercially available ELISA kit (Inova Diagnostics, San Diego, CA). Positivity was defined as a level of ≥10.8 IU/ml for IgA-RF, ≥9.5 IU/ml for IgM-RF, and ≥9.6 IU/ml for IgG-RF. These cutoff values were established using serum collected from the control population described above, with the assumption of a 5% false-positive rate for each assay. To examine the specificity of anti-CCP antibody responses to citrulline-containing antigen in subjects with SLE, we repeated the assays using a commercially available arginine-based control plate (Axis-Shield Diagnostics) in a subset of RA cases and in all SLE subjects who had positive results on the Anti-CCP2 assay.
For each autoantibody, we calculated the sensitivity, specificity, test accuracy, and predictive values (positive and negative). Specificity, accuracy, and predictive values for each assay were calculated in 3 ways, using 3 comparator groups (versus all controls combined, versus SLE subjects only, and versus healthy controls only). Ninety-five percent confidence intervals were calculated for each estimate, using normal approximation.
Results are summarized in Table 1. Elevated levels of anti-CCP antibody were observed in 120 of 195 African American RA patients (sensitivity 62%). Assay sensitivities in RA patients were slightly higher for IgM-RF (70%) and IgA-RF (65%) but lower for IgG-RF (39%).
|Anti-CCP†||IgA-RF||IgM-RF||IgG-RF||Anti-CCP + IgA-RF||Anti-CCP + IgM-RF|
|Sensitivity, % (95% CI)||62 (55–68)||65 (58–72)||70 (63–76)||39 (32–46)||55 (48–62)||59 (52–66)|
|Specificity, % (95% CI)|
|Vs. all controls (healthy controls plus SLE)||94 (91–97)||87 (82–92)||80 (75–86)||85 (80–90)||98 (96–100)||98 (96–100)|
|Vs. healthy controls||98 (95–100)||94 (89–99)||84 (76–92)||80 (72–89)||99 (97–100)||99 (97–100)|
|Vs. SLE||91 (85–96)||80 (73–88)||77 (69–86)||90 (84–96)||97 (93–100)||97 (93–100)|
|Accuracy, % (95% CI)|
|Vs. all controls (healthy controls plus SLE)||77 (72–83)||76 (70–81)||75 (69–81)||61 (56–67)||76 (71–81)||78 (73–83)|
|Vs. healthy controls||73 (66–80)||74 (67–81)||74 (67–82)||52 (45–59)||69 (62–76)||71 (64–78)|
|Vs. SLE||71 (64–78)||70 (63–78)||72 (65–80)||56 (49–63)||69 (62–76)||72 (65–78)|
|Positive predictive value, % (95% CI)|
|Vs. all controls (healthy controls plus SLE)||92 (87–96)||84 (78–90)||79 (73–85)||74 (65–82)||96 (93–100)||97 (93–100)|
|Vs. healthy controls||98 (96–100)||96 (93–99)||91 (86–95)||82 (74–90)||99 (97–100)||99 (97–100)|
|Vs. SLE||93 (89–97)||87 (73–88)||86 (81–91)||88 (82–95)||97 (94–100)||97 (95–100)|
|Negative predictive value, % (95% CI)|
|Vs. all controls (healthy controls plus SLE)||70 (64–75)||70 (64–76)||71 (65–78)||57 (51–63)||67 (62–73)||69 (64–75)|
|Vs. healthy controls||53 (45–61)||55 (47–63)||55 (47–64)||37 (30–44)||50 (42–57)||52 (44–59)|
|Vs. SLE||54 (46–62)||53 (45–62)||56 (48–64)||42 (35–49)||52 (45–59)||54 (47–61)|
The highest assay specificities were generally observed when comparisons were limited to healthy volunteers, and the lowest when comparisons were limited to those with SLE. Depending on the referent population used, the specificity of anti-CCP antibody ranged from 91% to 98%. In comparison, the specificity of IgA-RF ranged from 80% to 94%, while the specificity of IgM-RF and IgG-RF was 77–84% and 80–90%, respectively. With the use of a higher cutoff value of 50 IU/ml to define RF positivity, as has been suggested by others (5), the specificity of IgA-RF and IgM-RF approached or exceeded that observed for anti-CCP antibody (94–100% for IgA-RF and 98–100% for IgM-RF), although resulting in substantial declines in test sensitivity (32% for IgA-RF and 46% for IgM-RF). Of the 195 RA cases, 108 were positive for both anti-CCP antibody and IgA-RF (≥10.8 IU/ml), yielding a composite sensitivity of 55%. Seropositivity for both anti-CCP antibody and IgA-RF was observed almost exclusively in subjects with RA (specificity 97–99%), occurring in only 1 healthy control and 3 subjects with SLE. Composite seropositvity for both anti-CCP antibody and IgM-RF (≥9.5 IU/ml) had nearly identical metric characteristics (sensitivity 59%, specificity 97–99%). One hundred five of the RA cases (54%) were positive for all 3 antibodies (anti-CCP, IgA-RF, and IgM-RF), whereas only 1 healthy control and 2 SLE controls were positive for all 3 antibodies.
Eleven of the controls (2 healthy volunteers and 9 subjects with SLE) were positive for anti-CCP. Of the 2 healthy volunteers, 1 had a borderline elevation (5.6 units/ml) while the other had a much higher titer (101.5 units/ml). This latter subject had no prior history of RA and no joint symptoms at the time of evaluation. Of the 9 SLE controls who were positive for anti-CCP antibody, 1 was noted to have “possible RA” and in all but 1, “arthritis” was one of the SLE diagnostic criteria that were met (85% of anti-CCP negative SLE patients also had documented arthritis). All 9 of these study subjects were positive for antinuclear antibody, 3 were positive for anti–double-stranded DNA antibody, and 3 had documented renal involvement.
Recent studies in mouse models of SLE have shown that apparent anti-CCP2 ELISA reactivity may not be specific for the citrulline modification (6). Because of this, as well as to further characterize this reactivity in patients with SLE, we used arginine-based control plates provided by the same manufacturer. We found that, of the 4 individuals with SLE who had anti-CCP levels of >50 units/ml by the anti-CCP2 assay, all exhibited specific reactivity with the citrulline- as compared with arginine-containing plates (data not shown). Of the 5 SLE subjects with positive anti-CCP2 assay values ranging from 5 units/ml to 50 units/ml, only 1 exhibited specificity for the citrulline-containing plate, while the other 4 showed comparable reactivity with the 2 substrates. Therefore, a subset of patients with SLE exhibits “nonspecific” anti-CCP2 reactivity, while another subset has definitively positive anti-CCP2 reactivity. The relative importance of this distinction is not clear, since the arginine-containing control plate for the anti-CCP2 test has not been widely used. Nevertheless, it may explain in part our results as well as suggest that additional specificity studies should be performed in patients with SLE and perhaps other non-RA diseases in which the same effect may be found.
Similar to the results of previous studies involving subjects of predominantly European ancestry (7, 8), we found that anti-CCP antibody, IgA-RF, and IgM-RF were all moderately sensitive for early RA in this African American population, with sensitivities ranging from 62% to 70%. Although infrequent in healthy controls, false-positive results for anti-CCP antibody, IgA-RF, and IgM-RF were observed in ∼10–20% of African American SLE patients. While combined seropositivity for anti-CCP antibody and either IgA-RF or IgM-RF had a slightly lower sensitivity (55–59%) when compared with positivity on the individual tests, a positive result in both assays was found almost exclusively in patients with early RA, with a specificity of ≥97% regardless of the control group used.
These results differ from those of an earlier study showing a false-positive rate of only 2% for anti-CCP antibody among European subjects with SLE (7). It is possible that the divergent findings relate to differences in the anti-CCP antibody assays used (first-generation in the previous study versus second-generation in the present study), the race/ethnicity of the study subjects, or disease characteristics of the lupus populations examined.
In a more recent cross-sectional study (8), the diagnostic characteristics of anti-CCP antibody and RF isotypes were examined in a group of 196 European RA patients with established disease and compared with findings in 239 subjects with other rheumatic diagnoses. The sensitivity and specificity of anti-CCP antibody and RF isotypes (IgA-RF and IgM-RF) were similar to those observed in our study population, particularly when SLE comparators were used. Moreover, as in our study, combined seropositivity for anti-CCP antibody and either IgA-RF or IgM-RF was associated with a modest decrease in sensitivity (44–48%) but a substantial improvement in specificity (96–98%).
Taken together, our results and those reported by Bas and colleagues (8) suggest that the metric properties of anti-CCP antibody and RF isotypes are comparable in African Americans with RA and similar subjects of European/Caucasian descent. Additionally, our findings suggest that a positive anti-CCP antibody result does not necessarily exclude SLE in African American patients presenting with inflammatory arthritis. In such patients, the additional assessment of IgA-RF or IgM-RF isotypes may be of added value since composite seropositivity appears to be nearly exclusive to patients with RA.
The CLEAR Registry is supported by NIH grant N01-AR-02247. Dr. Mikuls' work was supported by NIH grant K23-AR-050004 and by the Arthritis Foundation. Dr. Holers' work was supported by NIH grants U19-A150864 and R01-AR-051394. Drs. Moreland and Bridges' work was supported by General Clinical Research Center (National Center for Research Resources) grant M01-RR-00032.
- 3Correlation of vitamin D deficiency with lupus disease measures among Sea Island Gullah African Americans [abstract]. Arthritis Rheum 2005; 52 Suppl 9: S180., , , , , , et al.