Abatacept for systemic lupus erythematosus: the outlook

Abstract

Abatacept is a selective T-cell co-stimulation modulator that inhibits full T-cell activation and subsequent antibody production by B cells. Despite the efficacy of abatacept in murine lupus, randomized controlled trials in human SLE do not reveal benefit of abatacept in non-renal and renal lupus. While problems in the study design and the primary efficacy end points may contribute to the negative results of these trials, post hoc analyses using alternative definitions for clinical response suggest the possibility that abatacept may have beneficial effects in active lupus arthritis and proliferative nephritis. Future clinical trials of abatacept should target on defined subsets of SLE patients, utilize multiple pre-defined outcomes based on experience from previous studies and determine the best timing of adding abatacept on a background of minimal immunosuppressive therapies.

Keywords::

• abatacept
• co-stimulation
• CTLA4-Ig
• lupus
• lupus nephritis
• outcome

1. Introduction

Activation of T cells requires two distinct signals: the first signal involves the binding of antigen to the T-cell receptor; and the second signal occurs with the interaction between receptor–ligand pairs on the T cells and antigen-presenting cells that include the B cells. The interaction of CD80 or CD86 molecule on B cells and CD28 on T cells provides an important second co-stimulatory signal for T-cell activation, which is important for the subsequent production of antibodies by the B cells. Abatacept (CTLA4-Ig) is a recombinant fusion protein consisting of the extracellular domain of CTLA4 and a modified fragment of Fc domain of human IgG1 that binds CD80 or CD86 with a higher affinity than CD28, and thus inhibits this co-stimulatory pathway and subsequent full T-cell activation.

Currently, abatacept has been approved for the treatment of moderately and severely active rheumatoid arthritis that is refractory to conventional disease-modifying anti-rheumatic drugs or the anti-TNF biologics. This biological agent has not yet been approved for the treatment of systemic lupus erythematosus (SLE). The outlook of abatacept in the treatment of SLE is hereby discussed.

2. Efficacy of CTLA4-Ig in murine lupus

CTLA4-Ig has long been tested in the murine lupus models. Wofsy and collaborators [1] demonstrated that administration of CTLA4-Ig to NZW/B mice delayed the onset of autoantibody production and mortality. Blockade of both the B7/CD28 and the CD40/gp39 pathways produced long-lasting ameliorating effect on autoantibody production, glomerulonephritis and mortality [2]. A subsequent study showed that combining cyclophosphamide (CYC) and CTLA4-Ig was more effective than either agent alone in reducing renal disease and mortality in these mice [3]. More recent studies and different murine SLE models have confirmed the mechanism of abatacept in reducing the autoreactive B-cell population, inhibiting immunoglobulin switching and production of autoantibodies, leading to less glomerular inflammation and renal infiltration by inflammatory cells on light microscopy [4-6].

3. Efficacy of abatacept in human lupus

A Phase IIb exploratory trial (IM101-042) was carried out to evaluate the safety and efficacy of abatacept in non-renal SLE [7]. In this 12-month multicenter, double-blind, placebo-controlled trial, 175 patients with predominantly active lupus arthritis (54%), discoid skin lesions (34%) or serositis (11%) were randomized in a 2:1 ratio to abatacept (10 mg/kg, i.e., 500 mg for body weight < 60 kg, 750 mg for weight of 60 – 100 kg, and 1000 mg for weight > 100 kg) or placebo given by intravenous infusion on days 1, 15, 29 and then every 4 weeks. Prednisone (30 mg/day or equivalent) was given for 1 month and tapered per protocol.

At 12 months, the proportion of patients with the primary end point (new BILAG A or B flare) following corticosteroid taper was nearly identical (79.7% for abatacept and 82.5% for placebo). Figures for the secondary efficacy end points (proportion of patients with a new SLE flare within the initial 6 months of the double-blind period, total number of SLE flares per patient, time to occurrence of a SLE flare and proportion of patients with no SLE flares while receiving low-dose corticosteroids (≤ 7.5 mg/day) for any 2 consecutive months during the double-blind period) were also similar between the abatacept and placebo groups of patients. Thus, the primary and secondary efficacy end points of abatacept in this study were not met. Post hoc analyses revealed that when only BILAG A flare was considered, 40.7% in the abatacept and 54.4% in the placebo group experienced flares over 12 months. The treatment difference was greatest in patients who had polyarthritis as the primary manifestation. Improvement in patient-reported outcomes, such as quality of life, fatigue and sleep problems score was greater in abatacept than placebo-treated patients. Serious adverse events were more frequent in abatacept than placebo-treated patients (19.8 vs 6.8%). The open-label extension phase of this study was discontinued by the developing company. Nevertheless, post hoc analyses and improvement in certain exploratory measures as described above suggest that abatacept may have some efficacy in moderate to severe non-renal manifestations of SLE.

4. Efficacy of abatacept in lupus nephritis

A recent pharmacokinetic study of abatacept showed a linear relationship between the dosage of administration and systemic exposure to the drug (measured by maximum concentration (C_max), area under the curve (AUC) and minimum concentration (C_min)) in patients with active lupus nephritis [8]. The systemic exposure was lower in patients with nephrotic syndrome caused by active renal disease than those without, but the improvement in proteinuria was greater in patients with nephrotic syndrome.

A 12-month Phase II/III multicenter study on the efficacy of abatacept in active proliferative lupus nephritis has just been released (IM101-075) [9]. Two hundred and ninety-eight patients with active ISN/RPS class III or IV lupus nephritis (urine protein/creatinine ≥ 50 mg/mmol and active urinary sediments) were randomized in a double-blind manner to receive abatacept (30 mg/kg loading for 3 months, followed by 10 mg/kg), abatacept (10 mg/kg) and placebo infusion in a 1:1:1 ratio on top of corticosteroids and mycophenolate mofetil (MMF). The initial dosage of prednisone (or equivalent) was 30 – 60 mg/day which tapers to 10 mg/day at week 12 and the target dosage of MMF was 2 g/day for Caucasians or Asians, and 3 g/day for Africans at day 57. Complete response was defined as a urine protein/creatinine of < 30 mg/mmol, inactive urinary sediment and an estimated glomerular filtration rate of no more 10% of the baseline value. The pre-defined primary efficacy end point was the time to reach a complete response.

After 52 weeks, no significant difference in the time to reach a complete response was observed among the two groups of patients who received abatacept and the group of patients who received placebo (primary end point not met). The proportion of patients who achieved complete response and renal improvement criteria was also similar in the three groups. However, exploratory and post hoc analyses indicated that a modified complete response and renal response rate was higher in patients who received abatacept than placebo. The incidence of serious adverse events was similar in the abatacept and placebo groups of patients.

Wofsy et al. [10]. performed a repeat analysis of the rates of complete response of this study using the definitions from the American College of Rheumatology (ACR) [11] and other randomized controlled trials such as the ALMS (Aspreva Lupus Management Study) [12] and LUNAR (Lupus Nephritis Assessment with Rituximab) studies [13]. It was demonstrated that a greater difference in complete response rate between the abatacept and placebo groups could be detected by using, in part, less stringent criteria for response that included a more reasonable cut-off for proteinuria improvement and change in estimated glomerular filtration rate from baseline to define stable renal function.

Finally, a Phase II trial testing the efficacy of abatacept/CYC combination in comparison with standard CYC in the treatment of active lupus nephritis (ACCESS study) is underway (ClinicalTrials.gov NCT00774852).

5. Expert opinion: the outlook of abatacept in the treatment of lupus

Despite the efficacy of abatacept in halting the onset and progression of renal disease, and reducing mortality in the murine models of lupus, results from recent randomized controlled trials of abatacept in renal and non-renal human SLE are disappointing. Issues related to the study design and the definition of the primary efficacy end point may have contributed to the negative results of the abatacept/SLE trials.

In the abatacept/non-renal lupus trial [7], one issue of the study protocol is the intensification of the background immunosuppression (moderate dose of prednisone) at study entry that is highly effective on its own for the lupus manifestations studied. The discordance between flares defined by the BILAG criteria and the treating physicians' opinion in this study (lower rate of flares as assessed by physicians) suggests that in a clinical trial setting, physicians are less likely to consider normal waxing and waning of less serious lupus manifestations as flares. Thus, the inclusion of both severe (BILAG A) and moderate (BILAG B) SLE flares for the definition of the primary efficacy outcome may have created noise that weakens the signal of interest. In fact, when more serious BILAG A flares were considered, a greater difference was observed between the treatment and placebo groups of patients.

In the abatacept/renal SLE trial [9], there appears to be a problem of the pre-defined primary efficacy end point, being in part, too stringent in terms of the change in glomerular filtration rate and the target proteinuria level [10]. Post hoc analyses of data from the abatacept/SLE studies [7,9] examining the active lupus manifestations in different systems at study entry and using alternative definitions for clinical and renal response suggest the possibility that abatacept may be effective in patients with active arthritis and proliferative lupus nephritis.

Future clinical trials of abatacept should focus on subsets of SLE patients that would benefit most from treatment (e.g., arthritis and renal disease), utilize multiple pre-defined outcomes based on post hoc analyses of data from previous trials, and determine the best timing of add-on therapy with abatacept on top of minimal background immunosuppressive therapies. However, one should be cautious about the increase in the incidence of serious adverse events with the use of abatacept. The role of abatacept/CYC combination in the treatment of more severe lupus manifestations has to be explored. Data of the ongoing randomized controlled trial comparing the efficacy of combined abatacept/CYC with standard CYC therapy in severe active lupus nephritis are eagerly awaited.

Declaration of interest

The author states no conflicts of interest and has received no payment in preparation of this manuscript.