IgG and IgM Anticardiolipin Antibodies Following Treatment With Infliximab Plus Methotrexate in Patients With Early Rheumatoid Arthritis

Transcription

1 ARTHRITIS & RHEUMATISM Vol. 54, No. 9, September 2006, pp DOI /art , American College of Rheumatology IgG and IgM Anticardiolipin Antibodies Following Treatment With Infliximab Plus Methotrexate in Patients With Early Rheumatoid Arthritis Sudha Visvanathan, 1 Carrie Wagner, 1 Josef Smolen, 2 E. William St.Clair, 3 Ron Hegedus, 1 Daniel Baker, 1 and Gregory Keenan 1 Objective. To assess the occurrence of anticardiolipin antibodies (acl) in patients with early rheumatoid arthritis (RA) receiving treatment with plus methotrexate (MTX) versus MTX alone. Methods. The first 299 patients enrolled in the randomized, Active-Controlled Study of Patients Receiving Infliximab for the Treatment of Rheumatoid Arthritis of Early Onset (ASPIRE) trial who had baseline (week 0) samples available for acl testing were included in this study. Sera were collected at weeks 0, 30, and 54 from 110 patients taking 3 mg/kg plus MTX, 98 patients taking 6 mg/kg, and 91 patients taking placebo. IgG and IgM acl were measured using an anticardiolipin assay. Results. IgG and IgM acl positivity at baseline was similar in all treatment groups. Most patients were negative for IgG acl at baseline and remained so at the last followup evaluation. One percent (2 of 208) of patients who received and were Supported by Centocor, Inc., a subsidiary of Johnson & Johnson. 1 Sudha Visvanathan, PhD, Carrie Wagner, PhD, Ron Hegedus, BA, Daniel Baker, MD, Gregory Keenan, MD: Centocor, Inc., Malvern, Pennsylvania; 2 Josef Smolen, MD: University of Vienna and Lainz Hospital, Vienna, Austria; 3 E. William St.Clair, MD: Duke University Medical Center, Durham, North Carolina. Drs. Visvanathan, Hegedus, Baker, and Keenan own stock and/or hold stock options in Johnson & Johnson, of which Centocor is a subsidiary. Dr. Wagner owns stock and/or holds stock options in Johnson & Johnson and has filed for a patent for a method to assess immune responses to. Dr. Smolen has received consulting fees or honoraria (less than $10,000 each) from Centocor, Schering- Plough, Wyeth, Roche, Bristol-Myers Squibb, Aventis, and UCB. Dr. St.Clair has received consulting fees or honoraria (less than $10,000 each) from Centocor, Human Genome Sciences, Genentech, Xoma/ Chiron, and MedImmune. Address correspondence and reprint requests to Sudha Visvanathan, PhD, Centocor, Inc., 200 Great Valley Parkway, Mailstop RA-1-4, Malvern, PA svisvana@cntus.jnj.com. Submitted for publication December 2, 2005; accepted in revised form May 24, negative for IgG acl at baseline were positive for IgG acl at weeks 30 and 54. A slightly higher proportion of patients who received and were negative for IgM acl at baseline were positive for IgM acl at weeks 30 and 54 (4.8% [10 of 208]) as compared with patients who received placebo (1.1% [1 of 91]), but the difference was not significant. Conclusion. There was a low incidence of the development of acl in patients with early RA who received in combination with MTX, and the difference was not significant compared with patients who received placebo. Naturally occurring antibodies to membrane phospholipids, including anticardiolipin, antiphosphatidylserine, antiphosphatidylethanolamine, and antiphosphatidylinositol antibodies, can play an important role in human disease (1). Anticardiolipin antibodies (acl) were first identified in association with systemic lupus erythematosus (2). Subsequently, antiphospholipid syndrome (APS; primary and secondary) was described in patients with thromboembolic events (and other characteristic clinical features) in the presence of acl (3,4). Increased concentrations of acl have since also been observed in patients with rheumatoid arthritis (RA) (5 7), although they are not typically associated with APS. More recently, several investigators have described the development of acl in RA patients treated with anti tumor necrosis factor (anti-tnf) agents. In all of these cases, however, there was no association with clinical manifestations of APS (8 11). Nevertheless, elevated levels of acl have been reported to be associated with an increased risk of venous and arterial thrombosis (12,13); therefore, the development of acl is a potential safety concern. 2840

2 acl IN PATIENTS WITH EARLY RA TREATED WITH INFLIXIMAB PLUS MTX 2841 We report herein our findings regarding the presence of acl (IgG and IgM) in patients with early RA participating in the Active-Controlled Study of Patients Receiving Infliximab for the Treatment of Rheumatoid Arthritis of Early Onset (ASPIRE) trial (14). Analyses were performed before and after treatment with (3 or 6 mg/kg) plus methotrexate (MTX) or treatment with MTX alone. These study patients were naive to MTX treatment prior to study entry. This allowed us to determine the presence of acl both before and after initiation of therapy. PATIENTS AND METHODS Role of the funding source. The overall ASPIRE study was designed by a committee composed of Centocor staff members and the ASPIRE steering committee members. Centocor collected data from all clinical sites to create the clinical database. Centocor staff members and members of the ASPIRE steering committee analyzed and interpreted the data, wrote the manuscript, and agreed to submit the manuscript for publication. All authors approved the content of the manuscript prior to submission. ASPIRE study population and protocol. Details of the ASPIRE study have been reported previously (14). Briefly, 1,049 patients with early, active RA who had not previously been treated with MTX or a TNF inhibitor were randomly assigned at a ratio of 4:5:5 to 1 of 3 treatment groups: placebo, 3 mg/kg, or 6 mg/kg. Details of the treatment protocol were described previously (14). Treatment with oral MTX was initiated at a dosage of 7.5 mg/week, and the dosage was escalated in a graduated manner to 15 mg/week by week 4 and 20 mg/week by week 8. Infusions of or placebo were given at weeks 0, 2, and 6, and every 8 weeks thereafter through week 46. The ASPIRE study protocol was approved by the Institutional Review Board at each study center. The study was performed in accordance with the Declaration of Helsinki and all subsequent revisions. Population assessed in the present study. Our study population comprised the first 299 patients enrolled in the randomized controlled ASPIRE study who had baseline (week 0) samples available for acl testing. In the main ASPIRE study, patients were randomized in order to control for selection bias. Sera from the same 299 patients were also collected for acl testing at weeks 30 and 54. Of the 299 patients, 208 were receiving either 3 mg/kg (n 110) or 6 mg/kg (n 98) of, and 91 patients were receiving placebo. All 3 groups were taking MTX. While baseline samples were available for all 299 patients, followup samples at weeks 30 and 54 were not obtained from all of them. Measurement of acl. Antibodies to cardiolipin with specificity for 2 -glycoprotein I are considered more relevant in relation to autoimmune diseases (15,16). IgG and IgM acl were measured by Mayo Medical Laboratories (Rochester, MN), using an enzyme-linked immunosorbent assay method that requires the presence of 2 -glycoprotein I. The assay was fully validated by Mayo Medical Laboratories and, overall, had an 83% concordance with the Helix apl Cardiolipin assay (Helix Diagnostics, Sacramento, CA). Values between 15 and 29.9 IgG phospholipid (GPL) units and between 10.1 and 29.9 IgM phospholipid (MPL) units were considered weakly positive. Values between 30 and 79.9 GPL or MPL units were considered positive. Values 80 GPL or MPL units were considered strongly positive. The most conservative approach was used to summarize changes in acl status, such that the 3 categories of acl positivity were combined into a single category. Lupus anticoagulant was not measured. Assessment of acl results. The treatment groups were compared with regard to the development of IgG and IgM acl, using Fisher s exact 2-sided pairwise test. The acl data were collected using a 4-point descriptive scale: negative, weakly positive, positive, and strongly positive. As noted above, we conservatively combined the 3 positive categories into 1 positive category to further simplify these data. Patients were also categorized as being positive for IgG acl or IgM acl if the corresponding antibody status had been negative at baseline but was positive at week 30, week 54, or both weeks 30 and 54. Anticardiolipin status was assessed in 3 groups of patients: those who had baseline samples available for testing, those who had evaluations at both baseline and at least 1 followup time point, and those who had followup samples obtained at both followup time points (weeks 30 and 54). Assessment of infections. We also assessed the incidence of infections in this cohort of 299 patients. Specifically, we assessed the occurrence of general infections (pharyngitis, sinusitis, and upper respiratory tract infections) and serious infections. RESULTS Findings at baseline. Baseline patient and disease characteristics were similar across the treatment groups (Table 1). At baseline (week 0), 7 of 208 patients (3.4%) in the combined groups and 2 of 91 patients (2.2%) in the placebo group were positive for IgG acl (Table 1). The proportion of patients who were positive for IgM acl at baseline was slightly higher than the proportion who were positive for IgG acl at baseline in both the combined (36 of 208 [17.3%]) and placebo (11 of 91 [12.1%]) groups. Thus, prior to receiving study medication, the randomized treatment groups were similar with regard to their IgG acl and IgM acl status. The proportion of patients positive for IgG acl was lower than the proportion who were positive for IgM acl at week 30, week 54, or both across treatment groups (Table 1). Positive IgG acl status was observed in fewer than 2% of patients across treatment groups and followup time points. No patients who were treated with placebo were positive for IgM acl at week 30, whereas 6 of 91 patients (6.6%) were positive for IgM acl at week 54. There was a similar trend in patients in the combined groups,

3 2842 VISVANATHAN ET AL Table 1. Baseline characteristics and positivity for IgG or IgM acl, by treatment group* Infliximab Placebo (n 91) 3 mg/kg of (n 110) 6 mg/kg of (n 98) Both groups (n 208) Sex, no. (%) Female 67 (73.6) 77 (70.0) 59 (60.2) 136 (65.4) Male 24 (26.4) 33 (30.0) 39 (39.8) 72 (34.6) Age, mean SD years Disease duration, mean SD years acl-positive at baseline, no. (%) IgG 2 (2.2) 3 (2.7) 4 (4.1) 7 (3.4) IgM 11 (12.1) 18 (16.4) 18 (18.4) 36 (17.3) IgG and IgM combined 0 1 (0.9) 1 (1.0) 2 (1.0) acl-positive at week 30, no. (%) IgG 0 1 (0.9) 1 (1.0) 2 (1.0) IgM 0 3 (2.7) 5 (5.1) 8 (3.8) acl-positive at week 54, no. (%) IgG 0 1 (0.9) 1 (1.0) 2 (1.0) IgM 6 (6.6) 9 (8.2) 5 (5.1) 14 (6.7) acl-positive at weeks 30 and 54, no. (%) IgG 0 2 (1.8) 0 2 (1.0) IgM 1 (1.1) 5 (4.5) 5 (5.1) 10 (4.8) * acl anticardiolipin antibody; MTX methotrexate. with 8 of 208 patients (3.8%) positive for IgM acl at week 30, 14 of 208 patients (6.7%) positive at week 54, and 10 of 208 patients (4.8%) positive at both time points (Table 1). Patterns of IgG and IgM acl positivity before and after treatment. Table 2 provides a summary of the various patterns of IgG and IgM acl positivity and negativity before and after treatment with Table 2. Patterns of acl findings at baseline and the last assessment, by treatment group* IgG acl IgM acl Placebo Infliximab Placebo Infliximab acl status at baseline and at last assessment (n 91) (n 208) (n 91) (n 208) Change in acl status Negative at baseline; positive at week (2.9) 7 (7.7) 32 (15.4) or week 54 Positive at baseline; negative at week 30 2 (2.2) 6 (2.9) 7 (7.7) 17 (8.2) or week 54 Negative at baseline; positive at weeks (1) 1 (1.1) 10 (4.8) and 54 No change in acl status Negative at baseline and at week 30 or 89 (97.8) 195 (93.8) 78 (85.7) 158 (76.0) week 54 Positive at baseline and at week 30 or 2 (2.2) 3 (1.4) 7 (7.7) 24 (11.5) week 54 Positive at baseline and at weeks 30 and (0.5) 2 (2.2) 12 (5.8) * Values are the number (%) of patients. P values for all between-group comparisons were not significant, by Fisher s exact 2-sided pairwise test. acl anticardiolipin antibody; MTX methotrexate.

4 acl IN PATIENTS WITH EARLY RA TREATED WITH INFLIXIMAB PLUS MTX 2843 Table 3. group* Infections in the cohort of 299 patients, by treatment Infection Placebo (n 91) Infliximab (n 208) Pharyngitis 8 (8.8) 24 (11.5) Sinusitis 11 (12.1) 29 (13.9) Upper respiratory tract infection 18 (19.8) 64 (30.8) Serious infections 3 (3.3) 11 (5.3) * Values are the number (%) of patients. MTX methotrexate. and with placebo. Note that not all patients had followup samples obtained at both weeks 30 and 54. Thus, we first examined the patterns of acl positivity/negativity using the last available followup sample. The majority of patients were negative for IgG acl at baseline and remained negative at week 30 or 54. Fewer than 3% of patients who received and who were negative for IgG acl at baseline were positive for IgG acl at week 30 or 54 (Table 2). No patients who received placebo and were negative for IgG acl at baseline were positive for IgG acl at either followup time point. For IgM acl, a higher proportion of patients receiving had positive antibody status at week 30 or 54 (15.4%), as compared with the proportion of patients receiving placebo (7.7%) (Table 2). Similar findings were observed with regard to the proportion of patients who were positive for IgM acl at both baseline and week 30 or 54 (7.7% of those taking placebo plus MTX and 11.5% of those taking ). We also assessed patterns of acl positivity/ negativity among the 172 patients taking plus MTX and the 74 patients taking placebo who had followup samples obtained for acl determinations at both weeks 30 and 54. Of the -treated patients, 5.8% were persistently positive at baseline, week 30, and week 54. Of the placebo-treated patients, 4.8% were negative at baseline and positive at both weeks 30 and 54 (data not shown). Furthermore, none of the patients receiving placebo were persistently positive or had a change in status from negative at baseline to positive at both weeks 30 and 54. Occurrence of infections. In this cohort of 299 patients, the occurrence of both general and serious infections was slightly higher in patients receiving than in those receiving placebo plus MTX (Table 3). DISCUSSION Rheumatoid arthritis is associated with an increased frequency of acl (5 7), as well as thromboses (17), as compared with normal healthy donors. The acl that are detected in patients with RA and other autoimmune diseases are directed against negatively charged phospholipids associated with 2 -glycoprotein, whereas acl associated with infection are directed against negatively charged phospholipids alone (15,16). Recently, there have been anecdotal reports of the development of acl in RA patients treated with the anti-tnf agents and etanercept (8 11). However, no data have been presented to date on the development of acl in RA patients participating in randomized controlled studies of TNF blockers. To this end, we investigated the occurrence of IgG and IgM acl in patients with early RA who were enrolled in the ASPIRE trial, a randomized controlled study (14). Prior to therapy, the treatment groups were similar with regard to the occurrence of IgG and IgM acl. The incidence of IgM acl at baseline was higher than the incidence of IgG acl in both the combined group (17% versus 3%, respectively) and the placebo group (12% versus 2%, respectively). The predominance of IgM acl is consistent with previous observations in RA patients receiving. In one study (8), all 5 patients who were positive for acl (21% [5 of 24 patients]) had acl of IgM isotype. In another study (10), 47% of patients who were positive for acl (21 of 45 patients) had acl of IgM isotype and 22% (10 of 45 patients) had acl of IgG isotype. The clinical significance of these acl remains to be clarified. Among patients who had followup samples obtained for acl determinations at both weeks 30 and 54, 2 9% of those in the combined group who had negative findings at week 30 showed a conversion to positive findings at week 54 or vice versa. These findings highlight the importance of obtaining serial samples over at least 1 year in order to ascertain a patient s acl status, and they underscore the waxing and waning of these antibodies. Relevant publications relating to the occurrence of acl in RA patients treated with suggest an incidence of 21 47% after initiation of treatment (8 10). These reports involved small groups of non randomly selected -treated patients, ranging from a total of 24 to 82 patients. In our subanalysis of 299 patients from the ASPIRE study, 208 patients were treated with and 91 were treated

5 2844 VISVANATHAN ET AL with placebo. Fewer than 3% of patients who received and had negative findings for IgG acl at baseline were found to have positive results at week 30 or 54. For IgM acl, a slightly higher proportion of patients who received plus MTX and had negative findings at baseline were found to have positive results at week 30 or week 54 (11.6%), as compared with the proportion of patients who received placebo (7.6%). The occurrence of general and serious infections in this cohort of patients from the ASPIRE study was comparable with the findings previously reported for the overall ASPIRE study (14). Given that we observed no significant difference between treatment groups in terms of the proportion of patients with persistent acl, the somewhat higher occurrences of infections in treated patients are most likely not reflective of differences in acl status. In conclusion, our findings indicate that there was a low incidence of the development of acl in patients with early active RA who received plus MTX, but the incidence was not significantly different from that in patients who received placebo. ACKNOWLEDGMENTS The authors wish to thank Michelle Perate, MS, and Mary Whitman, PhD, of Centocor, for writing support and editorial assistance. REFERENCES 1. Beaman KD, Gilman-Sachs A, Cifuentes D, Miller ML, O Gorman MR. Presence of multiple anti-phospholipid antibody specificities in a pediatric population. Autoimmunity 1995;21: Harris EN, Gharavi AE, Boey ML, Patel BM, Mackworth-Young CG, Loizou S, et al. Anticardiolipin antibodies: detection by radioimmunoassay and association with thrombosis in systemic lupus erythematosus. Lancet 1983;2: Hanly JG. Antiphospholipid syndrome: an overview [review]. CMAJ 2003;168: Reddel SW, Krilis SA. Testing for and clinical significance of anticardiolipin antibodies. Clin Diagn Lab Immunol 1999;6: Seriolo B, Cutolo M, Fasciolo D, DeCesari F, Accardo S. Anticardiolipin antibodies in rheumatoid arthritis [letter]. Ann Rheum Dis 1992;51: Fort JG, Cowchock FS, Abruzzo JL, Smith JB. Anticardiolipin antibodies in patients with rheumatic diseases. Arthritis Rheum 1987;30: Merkel PA, Chang Y, Pierangeli SS, Convery K, Harris EN, Polisson RP. The prevalence and clinical associations of anticardiolipin antibodies in a large inception cohort of patients with connective tissue diseases. Am J Med 1996;101: Ferraro-Peyret C, Coury F, Tebib JG, Bienvenu J, Fabien N. Infliximab therapy in rheumatoid arthritis and ankylosing spondylitis-induced specific antinuclear and antiphospholipid autoantibodies without autoimmune clinical manifestations: a two-year prospective study. Arthritis Res Ther 2004;6:R Jonsdottir T, Bratt J, Klareskog L, van Vollenhoven R. Development of ACLA (anti-cardiolipin antibodies) in patients treated with (Remicade) [abstract]. Arthritis Rheum 2001;44 Suppl 9:S Morris AJ, Morris CR, Hernandez CR. Anticardiolipin antibodies developing during therapy [abstract]. Arthritis Rheum 2001;44 Suppl 9:S Ferraccioli G, Mecchia F, DiPoi E, Fabris M. Anticardiolipin antibodies in rheumatoid patients treated with etanercept or conventional combination therapy: direct and indirect evidence for a possible association with infections. Ann Rheum Dis 2002;61: Aichbichler BW, Petritsch W, Reicht GA, Wenzl HH, Eherer AJ, Hinterleitner TA, et al. Anticardiolipin antibodies in patients with inflammatory bowel disease. Dig Dis Sci 1999;44: Zuckerman E, Toubi E, Golan TD, Rosenvald-Zuckerman T, Sabo E, Shmuel Z, et al. Increased thromboembolic incidence in anticardiolipin-positive patients with malignancy. Br J Cancer. 1995;72: St.Clair EW, van der Heijde DM, Smolen JS, Maini RN, Bathon JM, Emery P, et al. Combination of and methotrexate therapy for early rheumatoid arthritis: a randomized, controlled trial. Arthritis Rheum 2004;50: Hunt JE, McNeil HP, Morgan GJ, Crameri RM, Krilis SA. A phospholipid- 2 -glycoprotein I complex is an antigen for anticardiolipin antibodies occurring in autoimmune disease but not with infection. Lupus 1992;1: McNeil HP, Chesterman CN, Krilis SA. Immunology and clinical importance of antiphospholipid antibodies. Adv Immunol 1991;49: Seriolo B, Accardo S, Garnero A, Fasciolo D, Cutolo M. Anticardiolipin antibodies, free protein S levels and thrombosis: a survey in a selected population of rheumatoid arthritis patients. Rheumatology 1999;38:675 8.