A phase II, randomized, doubleblind ... - Wiley Online Library

11 downloads 4375 Views 199KB Size Report
Dec 1, 2011 - American College of Rheumatology 20% improvement criteria (achieving an ... Supported by Medarex (since acquired by Bristol-Myers. Squibb) and ...... Sharline Chen and Fei Cao for technical assistance. AUTHOR ... 177:1809–14. 6. Schulthess FT, Paroni F, Sauter NS, Shu L, Ribaux P, Haataja L, et al.
ARTHRITIS & RHEUMATISM Vol. 64, No. 6, June 2012, pp 1730–1739 DOI 10.1002/art.34330 © 2012, American College of Rheumatology

A Phase II, Randomized, Double-Blind, Placebo-Controlled Study Evaluating the Efficacy and Safety of MDX-1100, a Fully Human Anti-CXCL10 Monoclonal Antibody, in Combination With Methotrexate in Patients With Rheumatoid Arthritis Michael Yellin,1 Igor Paliienko,2 Andra Balanescu,3 Semen Ter-Vartanian,4 Vira Tseluyko,5 Li-An Xu,6 Xiaolu Tao,7 Pina M. Cardarelli,7 Heidi LeBlanc,7 Geoff Nichol,1 Codrina Ancuta,8 Rodica Chirieac,9 and Allison Luo7 Objective. CXCL10 (also known as interferon-␥– inducible 10-kd protein [IP-10]) is a chemokine that potentially plays a role in the immunopathogenesis of rheumatoid arthritis (RA). We undertook this phase II study to evaluate the efficacy and safety of MDX-1100, a fully human, anti-CXCL10 (anti–IP-10) monoclonal antibody, in RA patients whose disease responded inadequately to methotrexate (MTX). Methods. Patients with active RA receiving stable doses of MTX (10–25 mg weekly) were randomized to receive intravenous doses of 10 mg/kg MDX-1100 (n ⴝ 35) or placebo (n ⴝ 35) every other week. The primary

end point was the proportion of patients meeting the American College of Rheumatology 20% improvement criteria (achieving an ACR20 response) on day 85, and patients were followed up for safety to day 141. Results. The ACR20 response rate was significantly higher among MDX-1100–treated patients than among placebo-treated patients (54% versus 17%; P ⴝ 0.0024). Statistically significant differences in the ACR20 response rate between treatments were observed starting on day 43 (P < 0.05). The ACR50 and ACR70 response rates on day 85 did not differ between the groups. Overall, 51.4% of MDX-1100–treated patients and 30.3% of placebo-treated patients experienced at least 1 adverse event (AE). No study drug–related serious AEs were reported. Conclusion. MDX-1100 was well tolerated and demonstrated clinical efficacy in RA patients whose disease responded inadequately to MTX. This is the first study to demonstrate clinical efficacy of a chemokine inhibitor in RA and supports the notion of a potential role of IP-10 in the immunopathogenesis of RA.

ClinicalTrials.gov identifier: NCT01017367. Supported by Medarex (since acquired by Bristol-Myers Squibb) and Bristol-Myers Squibb. 1 Michael Yellin, MD (current address: Celldex Therapeutics, Phillipsburg, New Jersey), Geoff Nichol, MB ChB (current address: Sangamo Biosciences, Richmond, California): Medarex, Bloomsbury, New Jersey; 2Igor Paliienko, MD, PhD, DMSc: O. O. Bogomolets National Medical University, Kiev, Ukraine; 3Andra Balanescu, MD, PhD: St. Maria Clinical Hospital, Bucharest, Romania; 4Semen TerVartanian, PhD: Kiev Central Municipal Clinical Hospital, Kiev, Ukraine; 5Vira Tseluyko, MD, PhD: City Clinical Hospital No. 8, Kharkiv, Ukraine; 6Li-An Xu, PhD: Bristol-Myers Squibb, Hopewell, New Jersey; 7Xiaolu Tao, PhD, Pina M. Cardarelli, PhD, Heidi LeBlanc, PhD, Allison Luo, MD: Bristol-Myers Squibb, Princeton, New Jersey; 8Codrina Ancuta, MD, PhD: University of Medicine and Pharmacy “Grigore T. Popa” and Clinical Rehabilitation Hospital, Iasi, Romania; 9Rodica Chirieac, MD, PhD: Clinical Rehabilitation Hospital, Iasi, Romania. Drs. Xu, Tao, Cardarelli, LeBlanc, and Luo own stock or stock options in Bristol-Myers Squibb. Dr. Cardarelli is a co-inventor on patents for anti–interferon-␥–inducible 10-kd protein antibodies and their uses (US patent nos. 20100077497 and 20050191293). Address correspondence to Allison Luo, MD, Bristol-Myers Squibb, Immunology, PO Box 4000, Princeton, NJ 08543. E-mail: [email protected]. Submitted for publication March 10, 2011; accepted in revised form December 1, 2011.

Multiple studies in patients with rheumatoid arthritis (RA) have identified increased chemokine levels in blood, synovial fluid, and synovial membrane compared to patients with osteoarthritis or healthy controls (1–3). Interferon-␥ (IFN␥)–inducible 10-kd protein (IP10) (also known as CXCL10) is a chemokine that can potentially regulate inflammation at several levels. It induces integrin activation and generates directional migration of multiple cell types, including activated T cells, monocytes, and natural killer (NK) cells (4,5). In addition to mediating chemotaxis, IP-10 induces apopto1730

EFFICACY AND SAFETY OF MDX-1100 IN RHEUMATOID ARTHRITIS

sis of pancreatic beta cells and inhibits the proliferation of both epithelial and endothelial cells (6–9). Other proinflammatory functions of IP-10 include induction of molecules, such as interleukin-8 (IL-8) and CXCL5, and up-regulation of costimulatory cell surface molecules, such as CD54, CD80, and CD86, on monocytes (Cardarelli PM: unpublished observations). Multiple receptors have been implicated in these processes, including CXCR3, CXCR3b, glycosaminoglycans, Toll-like receptor 4, and an unidentified highaffinity receptor on epithelial cells (6–11). Expression of IP-10 and CXCR3 is increased in murine models of arthritis, and neutralizing anti–IP-10 or anti-CXCR3 antibodies ameliorate disease manifestations in these models (12,13). IP-10 and CXCR3 expression is also increased in the synovial membrane of RA patients (14,15). Taken together, these data suggest a potential role of IP-10 as a therapeutic target in RA. MDX-1100 is a fully human anti–IP-10 (antiCXCL10) monoclonal antibody (produced by Medarex, since acquired by Bristol-Myers Squibb) that binds to IP-10 with high affinity but not to other CXCR3 ligands, CXCL9, or CXCL11. In vitro studies demonstrated that MDX-1100 neutralizes the functional activity of IP-10, including calcium influx, leukocyte migration, and up-regulation of IP-10–responsive genes. Phase I single-dose studies of MDX-1100 in healthy volunteers or patients with ulcerative colitis demonstrated that the drug was well tolerated at all dose levels tested (0.1–10 mg/kg) and had a half-life of 10 days. The current study was a randomized, double-blind, placebocontrolled, phase II proof-of-concept study to determine the clinical efficacy and safety of MDX-1100 in RA patients whose disease responded inadequately to methotrexate (MTX). PATIENTS AND METHODS Patients. We enrolled male and female adult patients (age ⱖ18 years) in Ukraine and Romania who were diagnosed as having RA based on the American College of Rheumatology (ACR) 1987 revised criteria (16) and whose disease responded inadequately to MTX. Inclusion criteria were at least 6 swollen joints and 6 tender joints and at least 2 of the following: a C-reactive protein (CRP) level of ⬎10 mg/liter, an erythrocyte sedimentation rate (ESR) of ⱖ28 mm/hour, or morning stiffness lasting ⱖ45 minutes. All patients had to be seropositive for rheumatoid factor (RF) and/or anti–cyclic citrullinated peptide (anti-CCP) antibody. Patients were included if they had been taking MTX at a dosage of 10–25 mg weekly for at least 6 months and the dosage had been stable for at least 42 days prior to randomization. No change in MTX treatment was permitted during the study unless neces-

1731

sitated by toxicity. Concomitant treatment with low-dose corticosteroids (ⱕ10 mg/day of prednisolone or equivalent) and nonsteroidal antiinflammatory drugs was permitted provided no change in dosage was anticipated during the study. Besides MTX, all other disease-modifying antirheumatic drugs (DMARDs) or biologic agents had to have been discontinued at least 28 days prior to randomization, with the exception of leflunomide, which had to have been discontinued at least 60 days prior to randomization. Patients were excluded if they had prior B cell–depleting therapy, were immunodeficient, had complications of RA such as pulmonary fibrosis or vasculitis, or had a history of malignancy or acute or chronic infections. Study design and treatment administration. Patients were randomly assigned in a 1:1 ratio to receive either 10 mg/kg of MDX-1100 or placebo every 2 weeks for a total of 12 weeks (6 infusions). MDX-1100 or placebo was administered intravenously for 60 minutes on days 1, 15, 29, 43, 57, and 71 (weeks 0, 2, 4, 6, 8, and 10). Premedication was not routinely administered. ACR core set assessments (17) and data on the primary end point for clinical efficacy (meeting the ACR 20% improvement criteria [achieving an ACR20 response] [18]) were obtained at each visit up to day 85 (week 12), and the followup visit occurred on day 113 (week 16). Peak and trough blood samples for pharmacokinetic studies were obtained at each dosing visit, and single samples were obtained at nondosing visits up to day 113. Routine chemistry, hematology, and other safety laboratory tests were performed at each visit. Blood samples for immunogenicity testing were obtained at baseline and on days 29, 57, 85, and 113. All patients were followed up for safety until day 141. Reported efficacy and safety parameters are based on the visit days when the parameters were assessed. Written informed consent was obtained from each patient prior to screening, and the ethics review board of each participating center approved the study. Safety assessments. All patients were monitored for adverse events (AEs), and these were recorded at each visit and graded according to the Common Terminology Criteria for Adverse Events, version 3 (19). A serious AE was defined as an event that was fatal or life-threatening, required prolonged hospitalization, was significantly or permanently disabling or incapacitating, was associated with a congenital anomaly or birth defect, or required medical or surgical intervention to prevent one of the above-mentioned outcomes. An independent data monitoring committee reviewed unblinded safety data during the study. Safety analyses were conducted on all subjects administered at least 1 dose or partial dose of MDX-1100 or placebo. Safety assessments, including the incidence and severity of treatment-emergent AEs, were summarized by descriptive statistics. A treatment-emergent AE was defined as any AE occurring within 70 days of the last dose of study drug or any study drug–related AE occurring at any time point in the study. Efficacy end points. The primary end point of the study was the proportion of patients having achieved an ACR20 response on day 85. Secondary end points included the proportion of patients achieving ACR50 and ACR70 responses (20) as well as improvements in the individual components of the ACR core set of measures (17). A 0–100-mm visual analog scale was used for the patient’s assessment of pain and the patient’s and physician’s global assessments of disease activity.

1732

The proportion of patients having achieved disease remission or a European League Against Rheumatism (EULAR) good response (21) was determined using the Disease Activity Score in 28 joints (DAS28) (22), which assessed the number of swollen and tender joints, CRP levels, and the patient’s global assessment of disease activity. Physical function was determined using the modified Health Assessment Questionnaire (23). Pharmacokinetics assessment. Serum MDX-1100 concentrations were determined by quantitative enzyme-linked immunosorbent assay. Pharmacodynamics assessment. Samples for pharmacodynamic assessments were obtained for exploratory analyses at baseline and on study days 8 and 85. Measurement of gene expression by quantitative polymerase chain reaction (qPCR). Total RNA was purified from peripheral blood collected in PAXgene Blood RNA tubes (Qiagen) according to the manufacturer’s protocol. Complementary DNA (cDNA) was synthesized using the Invitrogen SuperScript III Reverse Transcriptase Kit and oligo(dT) primers (Invitrogen) according to the manufacturer’s instructions. Gene expression was measured by qPCR using gene-specific primers and Applied Biosystems Power SYBR Green PCR Master Mix. Copy number per 5 ng cDNA was determined using a standard curve of gene-specific PCR product, and then normalized to the copy number of a housekeeping gene, UBE2D2. Measurement of serum cytokine levels. Cytokines were analyzed using the Luminex multiplexing system and Millipore Milliplex Human Cytokine/Chemokine Immunoassay kits. Kits were run according to the manufacturer’s recommended protocol. Levels of 30 cytokines/chemokines, including IP-10 and eotaxin, were assayed, and concentrations were determined from a standard curve run with each assay. Sixteen of the 28 other cytokines/chemokines were below the detection range of the assay (granulocyte colony-stimulating factor, IFN␥, IL-1␤, IL-2, IL-3, IL-4, IL-5, IL-7, IL-8, IL-10, IL-13, IL-15, IL-17, IL-12p70, tumor necrosis factor ␣ [TNF␣], and TNF␤), and 12 showed no significant differences among treatment groups (granulocyte–macrophage colony-stimulating factor, IFN␣, IL-6, IL-12p40, macrophage inflammatory protein 1␣ [MIP1␣], MIP-1␤, IL-1␣, soluble shed CD25, monocyte chemotactic protein 1, vascular endothelial growth factor, macrophagederived chemokine, and soluble CD40L). Immunogenicity assessment. Serum samples along with positive and negative controls were tested in an electrochemiluminescence assay using the Meso Scale Discovery Sector PR 100 Plate Reader. Briefly, biotinylated MDX-1100, ruthenylated MDX-1100, and patient serum samples were added to wells of a polypropylene microtiter plate. Upon incubation, anti–MDX-1100 antibodies bind to both the biotinylated MDX-1100 and ruthenylated MDX-1100 molecules to form an antibody complex bridge. Next, the samples were added to microwell plates coated with streptavidin (Meso Scale Discovery). After washing, a tripropylamine-containing read buffer was added, and upon voltage application, a chemiluminescent signal was triggered. Samples were screened at a dilution of 1:20. Samples with signal intensity that was less than or equal to the calculated assay cut point (as determined by a low probability for presence of any anti–MDX-100 antibodies) were reported as negative.

YELLIN ET AL

Statistical analysis. A sample size of 35 patients per cohort was needed to provide 80% power to detect a difference of 40% in the ACR20 response between the MDX-1100– treated and placebo-treated cohorts at the 5% significance level (2-tailed) adjusted for a discontinuation rate of 10%. The ACR20 response rate in the placebo-treated cohort was assumed to be 30%. The intent-to-treat (ITT) population, defined as all patients who were randomized to receive MDX1100 or placebo, was the primary population for determining efficacy. The difference in response rates was analyzed by Fisher’s exact test. Subjects who discontinued prior to day 85 were imputed as nonresponders. Secondary end points and baseline demographics were summarized using descriptive statistics. For pharmacokinetics assessments, peak and trough concentrations were summarized using descriptive statistics. Table 1.

Baseline characteristics in the intent-to-treat population*

Age, mean (range) years Weight, mean (range) kg Female, % Caucasian, no. (%) Disease duration, mean (range) years Baseline medications Corticosteroids, % Prednisolone equivalent dose, mean (range) mg/day MTX, % MTX dose, mean (range) mg/week Prior RA therapy, %† DMARDs Anti-TNF RF positive, % Anti-CCP positive, % CRP, mean (range) mg/liter ESR, mean (range) mm/hour No. of tender joints‡ No. of swollen joints‡ Patient’s assessment of pain, 0–100-mm VAS§ Patient’s global assessment of disease activity, 0–100-mm VAS§ Physician’s global assessment of disease activity, 0–100-mm VAS§ M-HAQ score, 0–3¶

MDX-1100 (n ⫽ 35)

Placebo (n ⫽ 35)

47.5 (24–73) 67.5 (47–116) 82.9 35 (100) 6.6 (1–23)

50.7 (21–74) 66.1 (40–86) 77.1 35 (100) 7.6 (0–29)

51.4 8.2 (3–13)

54.3 7.4 (4–10)

100 11.8 (10–20)

100 11.8 (10–20)

34.3 0 88.6 94.3 18.3 (0–109) 44.0 (22–95) 22.5 ⫾ 11.30 15.4 ⫾ 6.61 63.9 ⫾ 12.43

31.4 2.9 94.3 94.3 11.6 (1–41) 47.3 (18–86) 20.5 ⫾ 9.79 13.5 ⫾ 6.16 62.2 ⫾ 17.9

66.6 ⫾ 13.21

67.1 ⫾ 16.36

61.1 ⫾ 11.76

60.3 ⫾ 14.32

1.5 ⫾ 0.47

1.5 ⫾ 0.66

* Except where indicated otherwise, values are the mean ⫾ SD. RA ⫽ rheumatoid arthritis; DMARDs ⫽ disease-modifying antirheumatic drugs; anti-TNF ⫽ anti–tumor necrosis factor; RF ⫽ rheumatoid factor; anti-CCP ⫽ anti–cyclic citrullinated peptide; CRP ⫽ C-reactive protein; ESR ⫽ erythrocyte sedimentation rate; VAS ⫽ visual analog scale; M-HAQ ⫽ modified Health Assessment Questionnaire. † Does not include methotrexate (MTX). ‡ Sixty-eight joints were evaluated for tenderness and 66 joints were evaluated for swelling. § High scores indicate more severe abnormalities. ¶ Low scores indicate low disease activity and high scores indicate greater disease activity.

EFFICACY AND SAFETY OF MDX-1100 IN RHEUMATOID ARTHRITIS

Table 2. Patient disposition from enrollment to completion of the study*

Completed through day 85 No Yes Primary reason for discontinuation prior to day 85 Adverse event Subject withdrew consent Unsatisfactory therapeutic effect Entered extended followup to day 141 No Yes Completed extended followup No Yes

MDX-1100 (n ⫽ 35)

Placebo (n ⫽ 35)

2 (5.7) 33 (94.3)

6 (17.1) 29 (82.9)

1 (2.9) 1 (2.9) 0

0 5 (14.3) 1 (2.9)

2 (5.7) 33 (94.3)

5 (14.3) 30 (85.7)

0 33 (94.3)

0 30 (85.7)

1733

The majority of patients in each cohort completed the day 85 efficacy assessment (Table 2). Two patients in the MDX-1100–treated cohort withdrew early from the study; 1 patient withdrew consent for unknown reasons and 1 withdrew due to an AE. Six

* Values are the number (%) of patients.

For pharmacodynamics assessments, a paired t-test was used to compare FoxP3 gene expression in patient samples collected on day 1 and day 85. The Mann-Whitney U test was used to compare mean expression levels of cytokines or genes between 2 patient groups. When more than 2 groups were compared, the Kruskal-Wallis test with Dunn’s correction was used. Data were analyzed using GraphPad Prism software. The reported P values for the pharmacodynamic analyses were not adjusted for multiplicity, although multiple pharmacodynamic biomarkers were analyzed.

RESULTS Baseline characteristics. Baseline demographic characteristics and disease activity were generally comparable between the treatment cohorts (Table 1). The majority of patients in each cohort were female, all were Caucasian, and the mean age of patients was 47.5 years and 50.7 years in the MDX-1100–treated and placebotreated cohorts, respectively. The mean weekly MTX dose was the same in each cohort (11.8 mg), and only 1 patient had previously been treated with an anti-TNF agent. The majority of patients in each cohort were RF positive (88.6% and 94.3% in the MDX-1100–treated and placebo-treated cohorts, respectively) and/or antiCCP positive (94.3% in both cohorts). Patient disposition. Seventy patients were randomized to receive 10 mg/kg MDX-1100 plus MTX (n ⫽ 35) or placebo plus MTX (n ⫽ 35), and 68 received ⱖ1 doses of study drug. Two patients withdrew from the study after randomization and prior to administration of the first dose of study drug. Most patients received all 6 infusions of study drug; 32 patients (91.4%) in the MDX-1100–treated cohort and 28 patients (84.8%) in the placebo-treated cohort completed all infusions.

Figure 1. Efficacy responses to treatment with MDX-1100 or placebo in the intent-to-treat population. Top, Proportion of patients meeting the American College of Rheumatology 20% improvement criteria (achieving an ACR20 response) by week. Differences in ACR20 response rates in patients treated with MDX-1100 versus those treated with placebo were observed starting at week 4 and achieved statistical significance by week 6 and at all subsequent visits. Bottom, Mean ⫾ SD change in the Disease Activity Score in 28 joints (DAS28) by week. Mean change from baseline in the DAS28 was similar between cohorts up to week 2, and there were nonsignificant differences from week 4 to week 12.

1734

YELLIN ET AL

Table 3. Mean ⫾ SD change from baseline in the ACR core components on day 85* ACR core component Tender joint count (66 joints assessed) Swollen joint count (68 joints assessed) Patient’s assessment of pain, 0–100-mm VAS Patient’s global assessment of disease activity, 0–100-mm VAS Physician’s global assessment of disease activity, 0–100-mm VAS HAQ disability index ESR, mm/hour CRP, mg/dl

MDX-1100 (n ⫽ 35)

Placebo (n ⫽ 35)

⫺9.9 ⫾ 15.07

⫺5.1 ⫾ 11.09

⫺10.3 ⫾ 7.06

⫺7.0 ⫾ 6.81

⫺16.6 ⫾ 14.23

⫺11.1 ⫾ 19.69

⫺18.6 ⫾ 15.94

⫺14.1 ⫾ 19.69

⫺18.1 ⫾ 13.11

⫺11.5 ⫾ 12.83

⫺0.5 ⫾ 0.41 ⫺5.0 ⫾ 20.65 ⫺3.2 ⫾ 13.11

⫺0.2 ⫾ 0.47 ⫺10.4 ⫾ 14.17 4.2 ⫾ 22.49

* ACR ⫽ American College of Rheumatology (see Table 1 for other definitions).

patients withdrew from the placebo-treated cohort prior to day 85; 5 withdrew consent for unknown reasons and 1 withdrew due to lack of efficacy. Efficacy. In the ITT population, 19 of 35 patients (54%) in the MDX-1100–treated cohort and 6 of 35 patients (17%) in the placebo-treated cohort had achieved an ACR20 response on day 85. The ACR20 response differed significantly (P ⫽ 0.0024) between the 2 cohorts on day 85. Analysis of response by visit demonstrated a difference in ACR20 response rates from day 29 that achieved statistical significance by day 43 (P ⬍ 0.05) and all subsequent visits (Figure 1). While the response rates were greater in the MDX-1100–treated cohort than in the placebo-treated cohort, none of the secondary efficacy end points differed significantly. The ACR50, ACR70, and EULAR good responses in the MDX-1100–treated and placebotreated cohorts on day 85 were 8.6% versus 2.9% (P ⫽ 0.61), 2.9% versus 0 (P ⫽ 1.00), and 5.7% versus 0 (P ⫽ 0.49), respectively. Mean change from baseline in the DAS28 was similar in the 2 cohorts up to day 15, and there were nonsignificant differences from day 29 to day 85 (Figure 1). For each of the ACR core set components with the exception of the ESR, a greater change in improvement from baseline was observed on day 85 in the MDX-1100–treated cohort (Table 3). There was no change in RF levels. AEs. MDX-1100 had an acceptable toxicity profile. Overall, 51.4% of patients treated with MDX-1100 and 30.3% of patients treated with placebo experienced ⱖ1 AEs during the study (Table 4). The difference in

AE frequency was primarily driven by a higher proportion of patients reporting peri-infusional events (defined as potentially infusion-related AEs occurring on the day of or the day following infusion) in the MDX-1100– treated cohort (25.7%) compared to the placebo-treated cohort (3.0%). With the exception of 1 patient with a grade 3 infusion reaction of bronchospasm and allergic rhinitis which led to early discontinuation from the study, all of the other AEs in the MDX-1100–treated group were mild to moderate in intensity and none led to discontinuation. The most common peri-infusional AEs in the MDX-1100–treated group were pyrexia and hyperthermia, which were reported by 5 of the 9 subjects. The majority of the peri-infusional AEs occurred after the first infusion only and did not recur when the patient continued to receive the study drug for the remainder of the 3-month study. The majority of AEs seen in ⱖ2 patients and occurring more frequently in the MDX-1100–treated cohort than in the placebo-treated cohort were infusion-related AEs (Table 4). The frequency of infections was 8.6% and 12.1% in the MDX1100–treated and placebo-treated cohorts, respectively. With few exceptions, AEs occurring during the study were grade 1 or grade 2 in severity. One patient administered placebo developed grade 4 pancytopenia and died suddenly. One patient administered MDX1100 developed a grade 3 infusion reaction (broncho-

Table 4.

Summary of AEs (safety population)*

Any AE AE grade ⱖ3 Serious AE Deaths Discontinued due to AE Infections Peri-infusional AEs (with 24 hours of infusion) AEs reported in ⱖ2 subjects Anemia Asthenia Hyperthermia Infusion-related reaction Pyrexia Alanine aminotransferase increased Aspartate aminotransferase increased Lymphocyte count decreased Neutrophil count increased

MDX-1100 plus MTX (n ⫽ 35)

Placebo plus MTX (n ⫽ 33)

18 (51.4) 1 (2.9) 0 0 1 (2.9) 3 (8.6) 9 (25.7)

10 (30.3) 1 (3.0) 1 (3.0) 1 (3.0) 0 4 (12.1) 1 (3.0)

2 (5.7) 2 (5.7) 2 (5.7) 2 (5.7) 3 (8.6) 0

0 0 0 0 1 (3.0) 2 (6.1)

0

2 (6.1)

2 (5.7) 2 (5.7)

0 1 (3.0)

* Values are the number (%) of patients. AEs ⫽ adverse events; MTX ⫽ methotrexate.

EFFICACY AND SAFETY OF MDX-1100 IN RHEUMATOID ARTHRITIS

1735

Figure 2. A, Mean serum interferon-␥–inducible 10-kd protein (IP-10) concentrations at baseline did not differ between MDX-1100–treated patients who would meet the American College of Rheumatology 20% improvement criteria (achieve an ACR20 response) (responders [R]) and those who would not (nonresponders [NR]). B, Shown are IP-10 mRNA copy numbers in MDX-1100–treated patients at baseline and at week 12. ACR20 responders had significantly higher baseline IP-10 levels than nonresponders. C, Shown are serum eotaxin levels in MDX-1100–treated patients at baseline and at week 12. ACR20 responders had significantly lower eotaxin levels than nonresponders. D, Shown are FoxP3 mRNA copy numbers in MDX-1100–treated patients at baseline and at week 12. ACR20 responders had significantly increased FoxP3 levels after treatment, whereas FoxP3 levels remained unchanged in nonresponders. Horizontal bars represent mean values. NS ⫽ not significant.

spasms) during the first infusion and withdrew from the study. In both cohorts there were no clinically significant changes from baseline during the study in relevant safety laboratory test results such as white blood cell, neutrophil, lymphocyte, monocyte, or platelet counts, or aspartate aminotransferase, alanine aminotransferase, creatinine, cholesterol, or triglyceride levels. Pharmacodynamics findings. Serum IP-10 protein levels were measured at baseline but could not be determined posttreatment because the anti–MDX-1100 antibody interferes with the assay (data not shown). Mean ⫾ SD baseline IP-10 protein levels were comparable between the MDX-1100–treated group (544 ⫾ 56 pg/ml) and the placebo-treated group (699 ⫾ 95 pg/ml). In the MDX-1100–treated cohort, mean ⫾ SD IP-10 protein levels at baseline were 593 ⫾ 78 pg/ml among ACR20 responders and 480 ⫾ 78 pg/ml among nonresponders (Figure 2A). IP-10 gene expression at baseline was significantly higher in ACR20 responders than in nonresponders (P ⫽ 0.026). The change from baseline in

IP-10 messenger RNA (mRNA) levels was not statistically significant on day 85 in patients administered MDX-1100 (Figure 2B). Of 30 soluble serum factors analyzed by Luminex assay, only eotaxin demonstrated a potentially interesting trend. In the MDX-1100–treated cohort, baseline serum eotaxin protein levels were markedly lower in ACR20 responders than in nonresponders, and levels did not change significantly over the course of treatment (Figure 2C). All the other soluble factors analyzed were either below the limit of detection or demonstrated no significant change during the study. Messenger RNA analysis of FoxP3, a marker for Treg cells, did not demonstrate any change in level of expression in the placebo-treated cohort (not shown) or in nonresponders in the MDX-1100–treated cohort. In contrast, a statistically significant increase in FoxP3 mRNA was observed in ACR20 responders on day 85 compared to baseline (Figure 2D). Pharmacokinetics findings. Limited pharmacokinetics assessments were performed. The serum concen-

1736

YELLIN ET AL

tration of MDX-1100 attained steady state after ⬃4 doses. The mean concentration at 30 minutes after the last dose on day 71 was 267.4 ␮g/ml, and the mean trough concentration on day 85 (2 weeks after the last dose) was 93.4 ␮g/ml. Exposure-response analysis was conducted to examine the relationship between ACR20 responses and serum MDX-1100 concentrations. Although there were differences in ACR20 responses between MDX-1100– treated and placebo-treated subjects, no differences were observed when ACR20 responders were stratified into 3 groups based on day 85 trough concentrations (bottom tertile, ⬃13.3–81.0 ␮g/ml; middle tertile, ⬃86– 111 ␮g/ml; top tertile, ⬃112–202 ␮g/ml). Immunogenicity findings. Based on both screening and titration tests, the results in 5 patients (14%) were inconclusive, and the rest of the patients in the MDX-1100–treated cohort were considered negative for anti-human antibody responses on the last postdose day (day 113). DISCUSSION MDX-1100 administered every 2 weeks for 12 weeks in combination with weekly MTX led to a statistically significant increase in the ACR20 response rate, the primary end point of the study, compared with placebo. Combination therapy with MDX-1100 and MTX had an acceptable toxicity profile. To our knowledge, this is the first randomized, double-blind study of a chemokine inhibitor to demonstrate clinical efficacy in RA. While the ACR20 response at 3 months conclusively favored MDX-1100 compared to placebo (54% versus 17%; P ⫽ 0.0024), the ACR50 and ACR70 responses did not differ significantly between treatment groups. Prior studies with inhibitors of chemokine pathways that are up-regulated in RA have failed to show evidence of pharmacodynamic or clinical efficacy. Clinical trials of antibodies targeting CCR2 (a chemokine receptor expressed on monocytes and T cell subsets), CCL2 (a chemokine that regulates monocyte/ macrophage and T cell trafficking), and IL-8 (a chemoattractant predominantly for neutrophils) failed to demonstrate clinical efficacy (24). Blocking of CCR1 (a receptor for several ligands with potential roles in RA immunopathogenesis) or CCR5 has also failed to demonstrate clinical, magnetic resonance imaging, or histologic improvements in RA patients (25). The redundancy in circulating chemokines possibly contributes to the lack of effect observed by others upon inhibition

of these mediator pathways. Positive results from the present study indicate that IP-10 is a dominant CXCR3 ligand in the immunopathogenesis of RA and possibly plays a more critical role in the inflammatory cascade. Another potential reason for lack of efficacy observed in the other studies cited above is their relatively short treatment duration, as discussed below. The kinetics of the MDX-1100 response are of interest. A difference in ACR20 response between the MDX-1100–treated cohort and the placebo-treated cohort was observed starting at 4 weeks. By 6 weeks, the differences were statistically significant and were maintained for the duration of the study. Biologic agents that specifically target proinflammatory molecules or their receptors, such as TNF␣ or IL-6 receptor, typically have a quicker onset of action, usually within just a week or two of treatment initiation (26,27). In contrast, the MDX-1100 response kinetics were similar to those of traditional DMARDs that have a slower onset of action of weeks to months. This may be due to the mechanism of action of MDX-1100 or to a suboptimal dosing schedule or regimen. It is possible that greater activity and a faster onset of action could be obtained with a higher MDX1100 dose or different dosing schedule. However, exposure-response analyses did not demonstrate an increase in clinical efficacy with higher exposure. The 10 mg/kg dose was based on a pharmacodynamic effect observed in a phase I study in which peripheral blood IP-10 mRNA levels were decreased following MDX1100 administration. The 12-week period for this study was chosen based on previous studies of blockers of proinflammatory molecules that demonstrate clinically and statistically significant activity in this time frame, particularly for the ACR20 response (26,27). Nevertheless, based on the kinetics of response discussed here, a longer dosing period with MDX-1100 may be required for optimal responses. This may also be relevant for the ACR50 and ACR70 responses, where 12–24 weeks of treatment may be needed to elicit higher responses (26,27). CXCR3, which is expressed on activated T cells, NK cells, plasmacytoid and myeloid dendritic cells, B cells, and mast cells and is the receptor for IP-10, can also bind 2 other ligands, CXCL9 and CXCL11 (28–30). CXCL9 and CXCL11 are both up-regulated during inflammation and regulate migration of T cells. It could be hypothesized that greater clinical efficacy in RA might be observed with an agent that neutralizes CXCL9 and/or CXCL11, in addition to IP-10. All 3 CXCR3 ligands are present in RA synovial fluid and produced by

EFFICACY AND SAFETY OF MDX-1100 IN RHEUMATOID ARTHRITIS

cytokine-stimulated synovial membrane fibroblasts in vitro (28,31). Serum CXCL9 and IP-10 levels correlate with RA disease activity and response to therapy (32). MDX-1100, a highly specific IP-10–neutralizing antibody that does not inhibit CXCL9 or CXCL11 function in vitro, was chosen for clinical development because IP-10 blockade is sufficient to ameliorate inflammation in multiple murine disease models (33,34), including an arthritis model (13), and because of theoretical concerns over potential safety issues with simultaneous blocking of all 3 CXCR3 ligands. It is currently unknown if other agents that effectively block multiple CXCR3 ligands or CXCR3 would demonstrate enhanced clinical efficacy with an acceptable safety profile. There are multiple potential mechanisms by which neutralization of IP-10 may have activity in RA. IP-10 is known to regulate trafficking of cell types found in RA synovial membrane, such as effector T cells, B cells, dendritic cells, and mast cells (4,5,29,35), and is thought to have key roles in initiating and perpetuating inflammation. CXCR3 is preferentially expressed on Th1 cells (36,37), and IP-10 or CXCR3 blockade can skew T cell responses to secretion of Th2 cytokines, which dampen inflammation in animal models of arthritis (38). IP-10 may also play a role in modulating the balance of Treg cells at sites of inflammation. In this regard, in an atherosclerosis model in mice genetically deficient in apolipoprotein E and IP-10, an increase in lesional Treg cells was observed in association with decreased T effector cells and decreased plaque size (39). IP-10 up-regulates RANKL and TNF␣ production by CD4⫹ T cells and induces osteoclastogenesis. Moreover, in the collagen-induced arthritis model, anti–IP-10 antibody therapy inhibited synovial T cell and macrophage migration and diminished bone erosions (40). In vitro, IP-10 can induce secretion of proinflammatory cytokines (such as IL-8, CXCL5, and IL-6) and up-regulate cell surface molecules (such as CD80, CD86, and CD54) thought to have roles in the immunopathogenesis of RA (Cardarelli PM, et al: unpublished observations). It is not known if MDX-1100 modulated the composition of synovial inflammatory cells since synovial biopsies were not performed. No effects of MDX1100 on leukocyte subset counts or on the levels of Th1 and Th2 cytokines in serum were observed (data not shown). In vitro data demonstrate that IP-10 induces IL-8 production (Cardarelli PM, et al: unpublished observations), and in this clinical study IL-8 mRNA levels decreased following MDX-1100 administration (data not shown). It is unlikely that modulating IL-8

1737

secretion would account for the observed clinical efficacy, since no differences in IL-8 levels between responders and nonresponders were observed after MDX1100 treatment. Also, an anti–IL-8 monoclonal antibody had no activity in a phase II study in a similar patient population (24). Eotaxin, a chemokine mostly associated with regulating eosinophil chemotaxis, is measurable in the serum of RA patients (41), and low serum levels were associated with rapid radiographic progression in patients with early RA (42). The finding that serum eotaxin levels were lower in responders than in nonresponders in the present study is of unclear significance. The most intriguing pharmacodynamic observation in this study is the finding of increased peripheral blood expression of mRNA for FoxP3, a transcription factor marker for Treg cells, in responders who received MDX-1100. CD4⫹CD25highFoxP3⫹ Treg cells have been identified in RA peripheral blood, synovial fluid, and synovial membrane (43). However, studies suggest that FoxP3-expressing cells may be relatively diminished in synovial membrane compared to T effector cells, and it has been hypothesized that increasing Treg cells in RA might have therapeutic potential (43–45). It is tempting to speculate that MDX-1100–mediated IP-10 neutralization modulates Treg cell–T effector cell balances, which would be consistent with the animal model findings discussed previously and would represent a novel mechanism of action in the treatment of RA. Finally, the MDX-1100 infusion was generally well tolerated without requirement for infusion reaction prophylaxis in the majority of patients, and most AEs were low grade. The need for administration of prophylaxis for infusion reaction is a topic for future investigations. The incidence of infection was comparable in the placebo-treated and MDX-1100–treated cohorts. However, this was only a 12-week study in a relatively small number of patients, and therefore definitive conclusions about the safety of MDX-1100 cannot yet be made. The immunogenicity result was inconclusive in 5 patients (14%) who were administered MDX-1100. However, this finding is likely to reflect suboptimization of the assay used to determine immunogenicity in a patient population with high-titer RF. MDX-1100 immunogenicity has not been observed in phase I single-dose studies or in a phase II multidose study in patients with ulcerative colitis. In summary, MDX-1100–mediated neutralization of IP-10 in this 12-week proof-of-concept study was safe and associated with a modest but statistically significant clinical response in RA patients with active disease

1738

YELLIN ET AL

who were also taking MTX. This is the first study to demonstrate clinical efficacy of a chemokine inhibitor in RA and supports the notion of a potential role of IP-10 in the immunopathogenesis of RA.

9.

10.

ACKNOWLEDGMENTS The authors thank Dr. Cristina Tanaseanu for substantial contributions in data acquisition. We also give special thanks to Judy Maccarone and Einav Leiberknight for assistance in conducting the study and writing the manuscript, to Jason Tian and Jenny Zhu for statistical support, to Anu Santhanagopal, PhD for writing and editorial support, and to Sharline Chen and Fei Cao for technical assistance.

11.

12.

13.

AUTHOR CONTRIBUTIONS All authors were involved in drafting the article or revising it critically for important intellectual content, and all authors approved the final version to be published. Dr. Luo 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 conception and design. Yellin, Ter-Vartanian, Xu, Nichol. Acquisition of data. Paliienko, Balanescu, Ter-Vartanian, Tseluyko, Xu, LeBlanc, Nichol, Ancuta, Chirieac. Analysis and interpretation of data. Yellin, Ter-Vartanian, Xu, Tao, Cardarelli, LeBlanc, Nichol, Ancuta, Chirieac, Luo. Preclinical studies to identify lead antibody. Cardarelli.

14.

15. 16.

17.

ROLE OF THE STUDY SPONSORS Medarex (which has been acquired by Bristol-Myers Squibb) contributed to the study design, conduct of the study, data collection, and data analysis. All authors had access to the data. The manuscript was cowritten by the lead and senior authors, with additional editorial support provided by Bristol-Myers Squibb. All authors reviewed the initial manuscript and all revisions thereafter and provided approval to submit the manuscript for publication.

18.

19.

20.

REFERENCES 1. Koch AE. Chemokines and their receptors in rheumatoid arthritis: future targets? [review]. Arthritis Rheum 2005;52:710–21. 2. Loetscher P, Moser B. Homing chemokines in rheumatoid arthritis. Arthritis Res 2002;4:233–6. 3. Szekanecz Z, Szucs G, Szanto S, Koch AE. Chemokines in rheumatic diseases. Curr Drug Targets 2006;7:91–102. 4. Lee EY, Lee ZH, Song YW. CXCL10 and autoimmune diseases. Autoimmun Rev 2009;8:379–83. 5. Taub DD, Lloyd AR, Conlon K, Wang JM, Ortaldo JR, Harada A, et al. Recombinant human interferon-inducible protein 10 is a chemoattractant for human monocytes and T lymphocytes and promotes T cell adhesion to endothelial cells. J Exp Med 1993; 177:1809–14. 6. Schulthess FT, Paroni F, Sauter NS, Shu L, Ribaux P, Haataja L, et al. CXCL10 impairs beta cell function and viability in diabetes through TLR4 signaling. Cell Metab 2009;9:125–39. 7. Soejima K, Rollins BJ. A functional IFN-␥-inducible protein-10/ CXCL10-specific receptor expressed by epithelial and endothelial cells that is neither CXCR3 nor glycosaminoglycan. J Immunol 2001;167:6576–82. 8. Campanella GS, Lee EM, Sun J, Luster AD. CXCR3 and heparin

21.

22.

23.

24.

25.

binding sites of the chemokine IP-10 (CXCL10). J Biol Chem 2003;278:17066–74. Campanella GS, Colvin RA, Luster AD. CXCL10 can inhibit endothelial cell proliferation independently of CXCR3. PLoS One 2010;5:e12700. Lasagni L, Francalanci M, Annunziato F, Lazzeri E, Giannini S, Cosmi L, et al. An alternatively spliced variant of CXCR3 mediates the inhibition of endothelial cell growth induced by IP-10, Mig, and I-TAC, and acts as functional receptor for platelet factor 4. J Exp Med 2003;197:1537–49. Loetscher M, Gerber B, Loetscher P, Jones SA, Piali L, ClarkLewis I, et al. Chemokine receptor specific for IP-10 and Mig: structure, function, and expression in activated T-lymphocytes. J Exp Med 1996;184:963–9. Mohan K, Issekutz TB. Blockade of chemokine receptor CXCR3 inhibits T cell recruitment to inflamed joints and decreases the severity of adjuvant arthritis. J Immunol 2007;179:8463–9. Salomon I, Netzer N, Wildbaum G, Schif-Zuck S, Maor G, Karin N. Targeting the function of IFN-␥-inducible protein 10 suppresses ongoing adjuvant arthritis. J Immunol 2002;169:2685–93. Qin S, Rottman JB, Myers P, Kassam N, Weinblatt M, Loetscher M, et al. The chemokine receptors CXCR3 and CCR5 mark subsets of T cells associated with certain inflammatory reactions. J Clin Invest 1998;101:746–54. Patel DD, Zachariah JP, Whichard LP. CXCR3 and CCR5 ligands in rheumatoid arthritis synovium. Clin Immunol 2001;98:39–45. Arnett FC, Edworthy SM, Bloch DA, McShane DJ, Fries JF, Cooper NS, et al. The American Rheumatism Association 1987 revised criteria for the classification of rheumatoid arthritis. Arthritis Rheum 1988;31:315–24. Felson DT, Anderson JJ, Boers M, Bombardier C, Chernoff M, Fried B, et al. The American College of Rheumatology preliminary core set of disease activity measures for rheumatoid arthritis clinical trials. Arthritis Rheum 1993;36:729–40. Felson DT, Anderson JJ, Boers M, Bombardier C, Furst D, Goldsmith C, et al. American College of Rheumatology preliminary definition of improvement in rheumatoid arthritis. Arthritis Rheum 1995;38:727–35. National Cancer Institute, NIH. Common terminology criteria for adverse events: version 3.0. Bethesda: National Cancer Institute; 2003. American College of Rheumatology Committee to Reevaluate Improvement Criteria. A proposed revision to the ACR20: the hybrid measure of American College of Rheumatology response. Arthritis Rheum 2007;57:193–202. Van Gestel AM, Prevoo ML, van ’t Hof MA, van Rijswijk MH, van de Putte LB, van Riel PL. Development and validation of the European League Against Rheumatism response criteria for rheumatoid arthritis: comparison with the preliminary American College of Rheumatology and the World Health Organization/ International League Against Rheumatism criteria. Arthritis Rheum 1996;39:34–40. Prevoo ML, van ’t Hof MA, Kuper HH, van Leeuwen MA, van de Putte LB, van Riel PL. Modified disease activity scores that include twenty-eight–joint counts: development and validation in a prospective longitudinal study of patients with rheumatoid arthritis. Arthritis Rheum 1995;38:44–8. Pincus T, Summey JA, Soraci SA Jr, Wallston KA, Hummon NP. Assessment of patient satisfaction in activities of daily living using a modified Stanford Health Assessment Questionnaire. Arthritis Rheum 1983;26:1346–53. Abgenix. Abgenix releases results of a phase 2a study with ABX-IL8 in rheumatoid arthritis. 2002. URL: http://www.amgen. com/pdfs/abgenix/2002-01-03.pdf. Van Kuijk AW, Vergunst CE, Gerlag DM, Bresnihan B, GomezReino JJ, Rouzier R, et al. CCR5 blockade in rheumatoid arthritis:

EFFICACY AND SAFETY OF MDX-1100 IN RHEUMATOID ARTHRITIS

26.

27.

28. 29.

30. 31.

32. 33.

34.

35.

a randomized, double-blind, placebo-controlled clinical trial. Ann Rheum Dis 2010;69:2013–6. Weinblatt ME, Keystone EC, Furst DE, Moreland LW, Weisman MH, Birbara CA, et al. Adalimumab, a fully human anti–tumor necrosis factor ␣ monoclonal antibody, for the treatment of rheumatoid arthritis in patients taking concomitant methotrexate: the ARMADA trial. Arthritis Rheum 2003;48:35–45. Smolen JS, Beaulieu A, Rubbert-Roth A, Ramos-Remus C, Rovensky J, Alecock E, et al, for the OPTION Investigators. Effect of interleukin-6 receptor inhibition with tocilizumab in patients with rheumatoid arthritis (OPTION study): a doubleblind, placebo-controlled, randomised trial. Lancet 2008;371: 987–97. Lacotte S, Brun S, Muller S, Dumortier H. CXCR3, inflammation, and autoimmune diseases. Ann N Y Acad Sci 2009;1173:310–7. Lande R, Giacomini E, Serafini B, Rosicarelli B, Sebastiani GD, Minisola G, et al. Characterization and recruitment of plasmacytoid dendritic cells in synovial fluid and tissue of patients with chronic inflammatory arthritis. J Immunol 2004;173:2815–24. Farber JM. Mig and IP-10: CXC chemokines that target lymphocytes. J Leukoc Biol 1997;61:246–57. Ueno A, Yamamura M, Iwahashi M, Okamoto A, Aita T, Ogawa N, et al. The production of CXCR3-agonistic chemokines by synovial fibroblasts from patients with rheumatoid arthritis. Rheumatol Int 2005;25:361–7. Kuan WP, Tam LS, Wong CK, Ko FW, Li T, Zhu T, et al. CXCL 9 and CXCL 10 as sensitive markers of disease activity in patients with rheumatoid arthritis. J Rheumatol 2010;37:257–64. Singh UP, Singh S, Singh R, Cong Y, Taub DD, Lillard JW Jr. CXCL10-producing mucosal CD4⫹ T cells, NK cells, and NKT cells are associated with chronic colitis in IL-10⫺/⫺ mice, which can be abrogated by anti-CXCL10 antibody inhibition. J Interferon Cytokine Res 2008;28:31–43. Liu MT, Keirstead HS, Lane TE. Neutralization of the chemokine CXCL10 reduces inflammatory cell invasion and demyelination and improves neurological function in a viral model of multiple sclerosis. J Immunol 2001;167:4091–7. Ruschpler P, Lorenz P, Eichler W, Koczan D, Hanel C, Scholz R, et al. High CXCR3 expression in synovial mast cells associated with CXCL9 and CXCL10 expression in inflammatory synovial

36.

37.

38.

39.

40.

41.

42.

43. 44.

45.

1739

tissues of patients with rheumatoid arthritis. Arthritis Res Ther 2003;5:R241–52. Gangur V, Simons FE, Hayglass KT. Human IP-10 selectively promotes dominance of polyclonally activated and environmental antigen-driven IFN-␥ over IL-4 responses. FASEB J 1998;12: 705–13. Sallusto F, Lenig D, Mackay CR, Lanzavecchia A. Flexible programs of chemokine receptor expression on human polarized T helper 1 and 2 lymphocytes. J Exp Med 1998;187:875–83. Horsfall AC, Butler DM, Marinova L, Warden PJ, Williams RO, Maini RN, et al. Suppression of collagen-induced arthritis by continuous administration of IL-4. J Immunol 1997;159:5687–96. Heller EA, Liu E, Tager AM, Yuan Q, Lin AY, Ahluwalia N, et al. Chemokine CXCL10 promotes atherogenesis by modulating the local balance of effector and regulatory T cells. Circulation 2006;113:2301–12. Kwak HB, Ha H, Kim HN, Lee JH, Kim HS, Lee S, et al. Reciprocal cross-talk between RANKL and interferon␥–inducible protein 10 is responsible for bone-erosive experimental arthritis. Arthritis Rheum 2008;58:1332–42. Hueber W, Tomooka BH, Zhao X, Kidd BA, Drijfhout JW, Fries JF, et al. Proteomic analysis of secreted proteins in early rheumatoid arthritis: anti-citrulline autoreactivity is associated with up regulation of proinflammatory cytokines. Ann Rheum Dis 2007; 66:712–9. Syversen SW, Goll GL, Haavardsholm EA, Boyesen P, Lea T, Kvien TK. A high serum level of eotaxin (CCL 11) is associated with less radiographic progression in early rheumatoid arthritis patients. Arthritis Res Ther 2008;10:R28. Sarkar S, Fox DA. Regulatory T cells in rheumatoid arthritis. Curr Rheumatol Rep 2008;10:405—12. Behrens F, Himsel A, Rehart S, Stanczyk J, Beutel B, Zimmermann SY, et al. Imbalance in distribution of functional autologous regulatory T cells in rheumatoid arthritis. Ann Rheum Dis 2007; 66:1151–6. Han GM, O’Neil-Andersen NJ, Zurier RB, Lawrence DA. CD4⫹CD25high T cell numbers are enriched in the peripheral blood of patients with rheumatoid arthritis. Cell Immunol 2008; 253:92–101.