Patients With Pulmonary Tuberculosis Are Frequently Positive for Anti–Cyclic Citrullinated Peptide Antibodies, but Their Sera Also React With Unmodified Arginine-Containing Peptide potx

ARTHRITIS & RHEUMATISM Vol. 58, No. 6, June 2008, pp 1576–1581 DOI 10.1002/art.23514 © 2008, American College of Rheumatology Patients With Pulmonary Tuberculosis Are Frequently Positive for Anti–Cyclic Citrullinated Peptide Antibodies, but Their Sera Also React With Unmodified Arginine-Containing Peptide Prasanthi Kakumanu, 1 Hajime Yamagata, 2 Eric S. Sobel, 1 Westley H. Reeves, 1 Edward K. L. Chan, 1 and Minoru Satoh 1 Objective. The anti–cyclic citrullinated peptide (anti-CCP) enz yme-linked immunosorbent assay (ELISA) has high sensitivity and specificity for rheu- matoid arthritis (RA). However, detection of anti-CCP in patients with active pulmonary tuberculosis (TB) has recently been reported. To determine whether this ac- tivity was specific for the citrullinated residue, the specificity of anti-CCP–positive sera for CCP versus that for unmodified arginine-containing peptide (CAP) was examined in patients with TB and compared with that in patients with RA. Methods. Anti-CCP and anti-CAP in sera from patients with pulmonary TB (n ؍ 49), RA patients (n ؍ 36), and controls (n ؍ 18) were tested by ELISA. Sera were available at diagnosis from most TB patients. All TB patients were treated with a combination of 2–4 antibiotics for at least 6 months, and sera were collected over time. Results. Anti-CCP was found in 37% of TB pa- tients and in 43% of RA patients. CAP reactivity was more common in TB than in RA. High anti-CCP:anti- CAP ratios (>2.0) were seen far more commonly in anti-CCP–positive RA patients than in anti-CCP– positive TB patients (94% versus 22%). Anti-CCP was inhibited by CCP peptide in sera from RA patients, but not in sera from TB patients. A slight increase in anti-CCP was common after initiating treatment for TB, although the anti-CCP level decreased after 1–2 months. Conclusion. Anti-CCP is frequently present in patients with active TB. However, many anti-CCP– positive TB sera also reacted with CAP, and anti-CCP: anti-CAP ratios in TB sera were low. Anti-CCP:anti- CAP ratios should be useful clinically for distinguishing CCP-specific reactivity seen in RA from reactivity with both CCP and CAP frequently seen in pulmonary TB. Anti–cyclic citrullinated peptide (anti-CCP) us- ing a peptide of filaggrin has been used extensively as a new serologic marker of rheumatoid arthritis (RA) (1,2). Many studies have confirmed that anti-CCP enzyme- linked immunosorbent assay (ELISA) is as sensitive as rheumatoid factor (RF) and much more specific for RA among various rheumatic diseases (2). However, anti- CCP in other diseases has been less well studied. Pa- tients with certain infectious diseases, such as hepatitis C and tuberculosis (TB), can present with arthritis and frequently also have serum RF, making it difficult to distinguish infectious disease–related arthritis from RA. Thus, the utility of anti-CCP testing under these clinical circumstances is important. Studies in patients with hepatitis C viral infection showed the absence of anti- CCP, in contrast to frequent RF positivity (3). However, a recent study showed the presence of anti-CCP in 32% of patients with active pulmonary TB (4). Patients with TB can exhibit various forms of arthropathy, including tuberculous arthritis and reactive arthritis (5). Some of these patients’ diseases may mimic early RA and may be difficult to diagnose. Thus, anti- CCP–positive test results in these patients can be mis- leading. Early studies showed that anti-CCP antibodies in RA patients are specific for citrullinated peptide and do not react with unmodified peptide (1). However, the Supported by the NIH (grants R01-AR-40391 and M01-R- 00082) and the Lupus Foundation of America. 1 Prasanthi Kakumanu, MD, Eric S. Sobel, MD, PhD, Westley H. Reeves, MD, Edward K. L. Chan, PhD, Minoru Satoh, MD, PhD: University of Florida, Gainesville; 2 Hajime Yamagata, MD, PhD: National Hospital Organization, Murayama Medical Center, Musashi- Murayama, Tokyo, Japan. Address correspondence and reprint requests to Minoru Satoh, MD, PhD, Division of Rheumatology and Clinical Immunology, University of Florida, PO Box 100221, Gainesville, FL 32610-0221. E-mail: satohm@medicine.ufl.edu. Submitted for publication October 11, 2007; accepted in revised form February 25, 2008. 1576 majority of clinical studies on anti-CCP are based on commercial anti-CCP ELISA kits and do not use an unmodified peptide as a negative control (2). In the present study, the specificity of anti-CCP–positive sera for CCP versus unmodified arginine-containing peptide (CAP) was examined in patients with TB and compared with that in patients with RA. PATIENTS AND METHODS Patients. Sera were obtained from Japanese patients with pulmonary TB (n ϭ 49) or RA (n ϭ 36) or from controls (n ϭ 18) at the National Hospital Organization, Murayama Medical Center, Musashi-Murayama, Tokyo, Japan. In the majority of patients with TB, the diagnosis was confirmed by culture for Mycobacterium tuberculosis, and all were treated with a combination of 2–4 antibiotics (streptomycin, isoniazid [INH], rifampin [RIF], ethambutol, para-aminosalicylic acid) for at least 6 months. None of the patients had atypical Mycobacterium. The initial sera from 36 TB patients were obtained before or within 1 month of treatment. In 6 TB patients, the first serum sample was obtained after 2–6 months of treatment. Detailed information was not available for 7 TB patients. Serial monthly sera were available from 25 patients with TB. The protocol was approved by the Institutional Review Board. Anti-CCP/anti-CAP ELISA. CCP (cfc1-cyc; amino ac- ids 306–324 of filaggrin, where arginine at position 312 is replaced with citrulline) and CAP (cf0-cyc; same peptide with arginine at position 312) peptides (1) were synthesized, cy- clized with Tl(CF 3 CO 2 ) 3 in dimethylformamide/anisole (19:1), and purified using high-performance liquid chromatography at the Interdisciplinary Center for Biotechnology Research Pro- tein Core Facility of the University of Florida. ELISA was performed as described elsewhere (6). Briefly, half of the microtiter plate wells (Immobilizer; Nunc, Naperville, IL) were coated with 2 ␮ g/ml CCP and the other half with CAP in 0.1M NaH 2 PO 4 , pH 9.0, at 4°C overnight. After blocking with 0.5% bovine serum albumin in NET (50 mM Tris, pH 7.5, 0.15M NaCl, 2 mM EDTA)/0.3% Nonidet P40 (NP40) (blocking buffer), wells were incubated with 1:500-diluted sera or the serially diluted high-titer anti-CCP–positive standard in the blocking buffer for 2 hours at 22°C. After washing 3 times with Tris buffered saline–Tween 20 (20 mM Tris, pH 7.5, 150 mM NaCl, 0.1% Tween 20), wells were incubated with alkaline phosphatase– conjugated goat anti-human IgG (1:1,000, ␥ -chain specific; Southern Biotechnology, Birmingham, AL) in blocking buffer, washed, and developed. The optical density at 405 nm (OD 405 nm ) of each sample was converted into units using the SoftMax Pro 4.7 program (Molecular Devices, Sunnyvale, CA), with 4-para- meter analysis. The standard curve was established by arbi- trarily defining the OD from a 1:312,500-diluted standard serum as 1 unit and applying units to each dilution so that the units correlated with the amount of antibodies as follows: 1:500 dilution, 625 units; 1:2,500 dilution, 125 units; 1:12,500 dilu- tion, 25 units; 1:62,500 dilution, 5 units; 1:312,500 dilution, 1 unit; and 1:1,562,500 dilution, 0.2 units. Both anti-CCP and anti-CAP units were interpolated from the same anti-CCP standard curve. The cutoff values for anti-CCP positivity (1.7 units) and anti-CCP:anti-CAP ratios (2.0) were determined based on the receiver operating characteristic curves compar- ing RA patients and controls. Inhibition of ELISA reactivity with CCP or CAP peptide, and effects of different concentrations of NaCl on antibody binding. Inhibition of ELISA reactivity with CCP or CAP was evaluated by incubating diluted serum with CCP or CAP peptide and then applying it to wells coated with CCP or CAP. Each serum was diluted (1:125–1:2,500; final concentration 1:250–1:5,000) so that the reactivity was at the low linear range of the standard curve (10–50 units) for sensitive detection of inhibition. CCP or CAP peptide (serially diluted 1:10, from 5,000 ng/ml to 0.5 ng/ml) in blocking buffer or buffer alone and appropriately diluted sera were mixed at a 1:1 ratio. After incubation for 1 hour at 22°C, 100 ␮ lofthe mixture was added to the wells coated with CCP or CAP, followed by secondary antibodies. OD 405 nm values were con - verted into anti-CCP or anti-CAP units. The percentage inhibition of each sample was calculated as 100 ϫ (units of the serum incubated with buffer Ϫ units of the serum incubated with an inhibitor)/(units of the serum incubated with buffer). In other experiments, the effects of different concen- trations of NaCl on antibody binding to CCP or CAP were evaluated. Following incubation with serum samples and wash- ing, wells were incubated for 30 minutes with NET/NP40 containing 0.15M, 0.375M,or0.5M NaCl, and the protocol was then continued as above. RESULTS Anti-CCP and anti-CAP ELISA. Anti-CCP (Ͼ1.7 units) was frequently present in TB patients (37%) and RA patients (43%), consistent with the previous report (4). Although the levels of anti-CCP in RA patients appeared to be generally higher than those in TB patients (mean Ϯ SD 31.0 Ϯ 99.8 units versus 4.1 Ϯ 8.6 units) and the percentage of high-positive values (Ͼ10 units) was greater in RA patients (22% versus 10%), neither of the differences was statistically significant. High anti-CAP levels (Ͼ10 units) were found only in TB patients (4%). The reactivity of sera from RA patients was specific for CCP, and serum from only 1 RA patient was also reactive with CAP (P Ͻ 0.0001 for anti-CCP reactivity versus anti-CAP reactivity in RA patient sera). In marked contrast, sera from patients with TB reacted to CCP and CAP (Figure 1A). Moreover, while the frequency of anti-CCP in RA patients (43%) and TB patients (37%) was similar, anti-CAP was more fre- quently present in TB patients than in RA patients (27% versus 3%; P Ͻ 0.005 by Fisher’s exact test) (Table 1). Since anti-CCP in RA patients appeared to be specific for CCP while anti-CCP in TB patients also reacted with CAP, anti-CCP:anti-CAP ratios were cal- culated to test whether this parameter could be used ANTI-CCP IN TUBERCULOSIS 1577 diagnostically to distinguish citrulline-dependent anti- CCP reactivity in RA from citrulline-independent anti- CCP reactivity in TB. High anti-CCP:anti-CAP ratios (Ͼ2.0) were seen far more commonly in anti-CCP– positive RA patients than in anti-CCP–positive TB patients (94% versus 22%; P Ͻ 0.0001) (Figure 1B and Table 1). Time course of anti-CCP antibodies. Certain drugs that are commonly used to treat TB, such as INH and RIF, are known to be associated with drug-induced lupus and the production of antinuclear and antihistone antibodies. The majority of TB patients (36 of 49) had sera available for testing that were obtained before or within 1 month of treatment. These sera were still positive for anti-CCP, indicating that the production of anti-CCP was related to TB and was not the result of treatment. However, to address the question of whether anti-CCP levels increased during treatment, we analyzed sequential sera that were available from 25 patients with TB. Among the 8 TB patients who were anti-CCP positive (Ͼ1.7 units) and the additional 6 TB patients with borderline anti-CCP positivity (1.2–1.7 units) who were tested over time, a transient increase of anti-CCP after initiating treatment was common (11 of 14 patients had an increase after 1 month, 2 of 11 patients had an increase at 2 months). However, in 9 patients examined Figure 1. A, Levels of anti–cyclic citrullinated peptide (anti-CCP) antibodies compared with anti–unmodified arginine-containing peptide (anti-CAP) antibodies in patients with rheumatoid arthritis (RA), patients with tuberculosis (TB), and in healthy controls. Anti-CCP and anti-CAP antibodies in sera from RA patients (n ϭ 37), TB patients (n ϭ 49), and healthy controls (n ϭ 18) were tested by enzyme-linked immunosorbent assay (ELISA). Optical densities were converted into units. The cutoff value of anti-CCP and anti-CAP was 1.7 units (shaded area). Horizontal bars indicate the median. P values for anti-CCP reactivity versus anti-CAP reactivity were obtained by Wilcoxon’s matched pairs test. NS ϭ not significant. B, Anti-CCP:anti-CAP ratios in anti-CCP–positive (Ͼ1.7 units) sera from patients with RA and TB. The ratios were high (Ͼ2.0) in 94% of RA patients, but in only 22% of TB patients. P values were obtained by Mann-Whitney U test. The cutoff value of the anti-CCP:anti-CAP ratio was 2.0 (shaded area). Horizontal bars indicate the median. C, Changes in anti-CCP levels in patients with TB over time. Sera obtained monthly from 14 TB patients (8 anti-CCP positive, 6 borderline anti-CCP positive) were tested by anti-CCP ELISA. A transient increase in anti-CCP levels after initiation of treatment was common (11 of 14 patients had an increase after 1 month, 2 of 11 patients had an increase at 2 months). However, in 9 patients examined for 2 months or more, the anti-CCP level decreased after 1–2 months. Table 1. Anti-CCP and anti-CAP levels and anti-CCP:anti-CAP ratios in RA patients, TB patients, and controls* RA patients (n ϭ 37) TB patients (n ϭ 49) Controls (n ϭ 18) Anti-CCP Ͼ1.7 units 43 37 6 Anti-CAP Ͼ1.7 units 3† 27 6 Anti-CCP:anti-CAP ratio Ͼ2.0 (all subjects) 59‡ 12 17 Anti-CCP:anti-CAP ratio Ͼ2.0 (subjects with anti-CCP Ͼ1.7 units) 94‡ 22 0 * Values are the percentage of patients or controls. Anti-CCP ϭ anti– cyclic citrullinated peptide; anti-CAP ϭ anti–unmodified arginine-containing peptide; RA ϭ rheumatoid arthritis; TB ϭ tuber- culosis. † P Ͻ 0.005 versus TB patients, by Fisher’s exact test. ‡ P Ͻ 0.0001 versus TB patients, by Fisher’s exact test. 1578 KAKUMANU ET AL for 2 months or more, the anti-CCP level decreased after 1–2 months (Figure 1C). In addition, none of the remaining 11 patients who were initially anti-CCP negative became anti-CCP positive (Ͼ1.7 units) after treatment. Inhibition of ELISA reactivity using CCP or CAP peptide and the effects of different concentrations of NaCl on antibody binding. We examined whether solu- ble CCP or CAP peptides could inhibit antibody binding to CCP or CAP using 5 sera from RA patients (all were anti-CCP positive and anti-CAP negative) and 4 sera from TB patients (all were positive for anti-CCP and anti-CAP). In sera from all 5 RA patients, antibody binding to CCP was nearly completely (87.8–99.8% at 2,500 ng/ml CCP) inhibited in a dose-dependent manner (Figures 2A–D). CAP did not have clear effects on anti-CCP binding in sera from 3 patients (Figures 2A– C), consistent with a lack of cross-reactivity of anti-CCP and anti-CAP in these sera. However, in addition to nearly complete inhibition of anti-CCP binding by CCP, CAP also showed weak but reproducible inhibition of anti-CCP binding in sera from 2 patients (15.3% and 22.2% inhibition) (Figure 2D). These 2 patients were classified as anti-CAP negative (Ͻ1.7 units) in the screening; however, their anti-CAP antibodies were at detectable levels, suggesting that they might have low levels of antibodies to CAP. Both CCP and CAP weakly (16.5–25.6%) inhibited anti-CAP binding in sera from these 2 patients (Figure 2D). In sera from all 4 TB patients with anti-CCP and anti-CAP positivity that were tested, inhibition by either peptide was not detected (Figures 2E and F). Unlike the reactivity of certain low-affinity antibodies, anti-CCP or anti-CAP reactivity was not affected significantly (Ͻ5%) by incubation of wells with higher concentrations of NaCl (0.375M or 0.5M) (data not shown), suggesting that these interactions are of relatively high avidity. DISCUSSION In the present study, anti-CCP was detected in 37% of Japanese patients with active pulmonary TB, similar to findings in a previous study from Israel, which showed that 32% of patients with active TB were anti-CCP positive (4). There are several situations in which anti-CCP positivity associated with TB may mis- lead or confuse clinical judgments on the diagnosis and treatment of arthritis. First, monarthritis or oligoarthri- Figure 2. Inhibition of antibodies to CCP or CAP by preincubation with CCP or CAP peptide. Sera from anti-CCP–positive RA patients (A–D)or anti-CCP–positive and anti-CAP–positive TB patients (E and F) were incubated with serially diluted CCP or CAP peptide and then added to wells coated with CCP or CAP (4 sets of data for each sample). Anti-CAP reactivity was undetectable or too weak for reliable calculation of the percentage inhibition in samples represented in A–C; therefore, anti-CAP data are not shown. The anti-CAP value in the sample represented in D was below the cutoff level; however, it was still considered the level at which the percentage inhibition could be calculated. The optical density value at 405 nm in each sample was converted into units as described in Patients and Methods. The percentage inhibition of each sample was calculated as 100 ϫ (units of the serum incubated with buffer – units of the serum incubated with an inhibitor)/(units of the serum incubated with buffer). See Figure 1 for definitions. ANTI-CCP IN TUBERCULOSIS 1579 tis caused by Mycobacterium can mimic synovitis seen in early RA (7,8). Tuberculous arthritis used to be consid- ered to predominantly affect weight-bearing joints such as the hip and knee. However, changing clinical charac- teristics with more frequent involvement of wrist and other peripheral joints have been reported (7). Second, patients with Mycobacterium infection can exhibit non- infectious reactive polyarthritis, known as Poncet’s dis- ease or tuberculous rheumatism (5). Thus, a patient could be misdiagnosed as having early RA based on polyarthritis and anti-CCP positivity. A third situation arises when a patient with RA contracts or experiences a reactivation of a TB infection and develops increas- ing levels of anti-CCP and signs of inflammation, such as fever, a higher erythrocyte sedimentation rate, or a higher C-reactive protein level, and these findings are mistakenly attributed to an RA disease flare. Biologic agents that inhibit tumor necrosis factor ␣ have rapidly become a part of standard therapy in RA, but they are associated with an increased risk of developing, or having a relapse of, Mycobacterium infection (9). Whether nonpulmonary TB such as isolated tu- berculous arthritis or an atypical Mycobacterium infec- tion can also induce anti-CCP will need to be evaluated in future studies. The effects of preventive vaccination with BCG also need to be considered. In many Asian countries, including Japan, BCG vaccination is manda- tory in childhood when the skin test for purified protein derivative (PPD) yields a negative result. Thus, positive PPD skin test results can be seen in high frequency among these populations, making the diagnosis of active TB more difficult (10). The incidence of TB is also much higher in Japan than in the US. However, the percentage of anti-CCP positivity in normal individuals or in pa- tients with diseases other than RA among Japanese individuals does not appear to be higher than that in the US or in European countries (ref. 11 and the present study), suggesting that BCG vaccination during child- hood does not cause long-lasting anti-CCP positivity. However, it is not known whether preventive BCG vac- cination can cause a transient increase in anti-CCP. Intravesicular BCG immunotherapy has been used worldwide in patients with bladder cancer (5,12). Whether these patients develop anti-CCP will also need to be evaluated, because the chronic granulomatous in- flammation in the bladder after BCG therapy is quite different from the inflammation arising from preventive BCG vaccination. Reactive arthritis following BCG im- munotherapy is well described (12), and if anti-CCP is induced, it could also mislead diagnosis. The mechanism of induction of anti-CCP in TB or even in RA is not known. However, it does not appear to be the result of treatment, as is seen for antihistone antibodies in drug-induced lupus. The detection of IgG anti-CCP antibodies before starting therapy and the reduction of anti-CCP levels while continuing treatment support this interpretation. Anti-CCP in TB may be induced and maintained by the immune response to chronic granulomatous inflammatory tissue rather than by the immune response directed at live Mycobacterium, because anti-CCP levels in patients with TB do not drop rapidly after treatment is started (Figure 1C). Both hepatitis C and TB patients’ sera have RF, but the difference in the inflammatory process may explain the absence of anti-CCP (3) in patients with hepatitis C. Various cellular proteins are citrullinated during apoptosis (13), and protein citrullination is a process associated with various inflammatory conditions regard- less of the diagnosis (14), suggesting that inflammatory tissue and apoptotic cells in TB granuloma likely contain citrullinated proteins as well. Thus, the citrullinated proteins in inflammatory tissue may play a role in the development of anti-CCP in TB. Nevertheless, the pat- tern of host response to citrullinated versus uncitrulli- nated peptide in RA versus TB is quite different: anti-CCP in RA is specific for CCP, whereas anti-CCP in TB reacts with both CCP and CAP. The poor capability of liquid-phase CCP or CAP to cross-inhibit the reactiv- ity of sera from patients with TB compared with sera from patients with RA (Figure 2D) may be due to the presence of low-avidity antibodies or to the differences between epitopes expressed on solid-phase peptides bound to the plastic surface (considered partially dena- tured) and epitopes expressed on inhibitors in liquid phase. Thus, anti-CCP–positive sera can be classified into 3 categories: 1) sera that have anti-CCP but little anti-CAP (as in most RA patients), 2) sera that have antibodies that bind to CCP and CAP but are not polyreactive, and 3) sera with polyreactive antibodies that bind to CCP and CAP. A recent study showed that anti-CCP in patients with autoimmune hepatitis is often citrulline independent and suggested that anti-CCP in nonrheumatic diseases should be interpreted with care (15). The present data on anti-CCP in pulmonary TB support that interpretation. In summary, anti-CCP is frequently present in patients with active TB, and this may cause confusion and clinical misjudgment in the diagnosis and treatment of patients with RA and/or TB. However, many anti- CCP–positive TB sera also reacted with CAP, and anti-CCP:anti-CAP ratios in TB sera were low, very 1580 KAKUMANU ET AL different findings from those in RA sera. Determination of anti-CCP:anti-CAP ratios or performance of inhibi- tion ELISAs using CCP should be useful for distinguish- ing the CCP-specific reactivity seen in RA sera from the reactivity to both CCP and CAP seen in TB sera. AUTHOR CONTRIBUTIONS Dr. Satoh had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. Study design. Kakumanu, Yamagata, Satoh. Acquisition of data. Kakumanu, Yamagata, Satoh. Analysis and interpretation of data. Kakumanu, Yamagata, Sobel, Reeves, Chan, Satoh. Manuscript preparation. Kakumanu, Yamagata, Sobel, Chan, Satoh. Statistical analysis. Satoh. REFERENCES 1. Schellekens GA, de Jong BA, van den Hoogen FH, van de Putte LB, van Venrooij WJ. Citrulline is an essential constituent of antigenic determinants recognized by rheumatoid arthritis-specific autoantibodies. J Clin Invest 1998;101:273–81. 2. Avouac J, Gossec L, Dougados M. Diagnostic and predictive value of anti-cyclic citrullinated protein antibodies in rheumatoid arthri- tis: a systematic literature review. Ann Rheum Dis 2006;65:845–51. 3. Wener MH, Hutchinson K, Morishima C, Gretch DR. Absence of antibodies to cyclic citrullinated peptide in sera of patients with hepatitis C virus infection and cryoglobulinemia. Arthritis Rheum 2004;50:2305–8. 4. Elkayam O, Segal R, Lidgi M, Caspi D. Positive anti-cyclic citrullinated proteins and rheumatoid factor during active lung tuberculosis. Ann Rheum Dis 2006;65:1110–2. 5. Franco-Paredes C, Diaz-Borjon A, Senger MA, Barragan L, Leonard M. The ever-expanding association between rheumato- logic diseases and tuberculosis. Am J Med 2006;119:470–7. 6. Satoh M, Weintraub JP, Yoshida H, Shaheen VM, Richards HB, Shaw M, et al. Fas and Fas ligand mutations inhibit autoantibody production in pristane-induced lupus. J Immunol 2000;165: 1036–43. 7. Garrido G, Gomez-Reino JJ, Fernandez-Dapica P, Palenque E, Prieto S. A review of peripheral tuberculous arthritis. Semin Arthritis Rheum 1988;18:142–9. 8. Choi ST, Park MC, Lee SW, Park YB, Hahn SB, Lee SK. Hand involvement of Mycobacterium tuberculosis infection mimicking rheumatoid arthritis [letter]. Clin Exp Rheumatol 2006;24:347. 9. Gomez-Reino JJ, Carmona L, Angel Descalzo M, for the BIOBA- DASER Group. Risk of tuberculosis in patients treated with tumor necrosis factor antagonists due to incomplete prevention of reactivation of latent infection. Arthritis Rheum 2007;57:756–61. 10. Miyasaka N, Takeuchi T, Eguchi K. Proposed Japanese guidelines for the use of infliximab for rheumatoid arthritis. Mod Rheumatol 2005;15:4–8. 11. Takasaki Y, Yamanaka K, Takasaki C, Matsushita M, Yamada H, Nawata M, et al. Anticyclic citrullinated peptide antibodies in patients with mixed connective tissue disease. Mod Rheumatol 2004;14:367–75. 12. Murata H, Adachi Y, Ebitsuka T, Chino Y, Takahashi R, Hayashi T, et al. Reiter’s syndrome following intravesical bacille bilie´de Calmette-Gue´rin treatment for superficial bladder carcinoma: report of six cases. Mod Rheumatol 2004;14:82–6. 13. Van Venrooij WJ, Pruijn GJ. Citrullination: a small change for a protein with great consequences for rheumatoid arthritis. Arthritis Res 2000;2:249–51. 14. Makrygiannakis D, af Klint E, Lundberg IE, Lofberg R, Ulfgren AK, Klareskog L, et al. Citrullination is an inflammation-depen- dent process. Ann Rheum Dis 2006;65:1219–22. 15. Vannini A, Cheung K, Fusconi M, Stammen-Vogelzangs J, Drenth JP, Dall’Aglio AC, et al. Anti-cyclic citrullinated peptide positivity in non-rheumatoid arthritis disease samples: citrulline-dependent or not? Ann Rheum Dis 2007;66:511–6. ANTI-CCP IN TUBERCULOSIS 1581 . Rheumatology Patients With Pulmonary Tuberculosis Are Frequently Positive for Anti–Cyclic Citrullinated Peptide Antibodies, but Their Sera Also React With Unmodified. anti-CCP positive sera for CCP versus that for unmodified arginine-containing peptide (CAP) was examined in patients with TB and compared with that in patients