SEARCH

SEARCH BY CITATION

Abstract

  1. Top of page
  2. Abstract
  3. PATIENTS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. AUTHOR CONTRIBUTIONS
  7. ROLE OF THE STUDY SPONSOR
  8. Acknowledgements
  9. REFERENCES

Objective

To assess the efficacy and safety of rilonacept (Interleukin-1 [IL-1] Trap), a long-acting and potent inhibitor of IL-1, in patients with cryopyrin-associated periodic syndromes (CAPS), including familial cold autoinflammatory syndrome (FCAS) and Muckle-Wells syndrome (MWS).

Methods

Forty-seven adult patients with CAPS, as defined by mutations in the causative NLRP3 (CIAS1) gene and pathognomonic symptoms, were enrolled in 2 consecutive phase III studies. Study 1 involved a 6-week randomized double-blind comparison of weekly subcutaneous injections of rilonacept (160 mg) versus placebo. Study 2 consisted of 9 weeks of single-blind treatment with rilonacept (part A), followed by a 9-week, randomized, double-blind, placebo-controlled withdrawal procedure (part B). Primary efficacy was evaluated using a validated composite key symptom score.

Results

Forty-four patients completed both studies. In study 1, rilonacept therapy reduced the group mean composite symptom score by 84%, compared with 13% with placebo therapy (primary end point; P < 0.0001 versus placebo). Rilonacept also significantly improved all other efficacy end points in study 1 (numbers of multisymptom and single-symptom disease flare days, single-symptom scores, physician's and patient's global assessments of disease activity, limitations in daily activities, and C-reactive protein and serum amyloid A [SAA] levels). In study 2 part B, rilonacept was superior to placebo for maintaining the improvements seen with rilonacept therapy, as shown by all efficacy parameters (primary end point; P < 0.0001 versus placebo). Rilonacept was generally well tolerated; the most common adverse events were injection site reactions.

Conclusion

Treatment with weekly rilonacept provided marked and lasting improvement in the clinical signs and symptoms of CAPS, and normalized the levels of SAA from those associated with risk of developing amyloidosis. Rilonacept exhibited a generally favorable safety and tolerability profile.

Cryopyrin-associated periodic syndromes (CAPS) are a group of inherited inflammatory disorders, consisting of familial cold autoinflammatory syndrome (FCAS; also known as familial cold urticaria [FCU]), Muckle-Wells syndrome (MWS), and neonatal-onset multisystem inflammatory disease (NOMID; also known as chronic infantile neurologic, cutaneous, articular [CINCA] syndrome). While varying in severity and phenotype (NOMID/CINCA syndrome is the most severe), all of these disorders are associated with considerable hardship. These disorders are typically associated with heterozygous mutations in the NLRP3 (CIAS1) gene, which encodes the cryopyrin (NALP3) protein, and are inherited in an autosomal-dominant manner (1, 2), although in some patients with CAPS, no detectable NLRP3 mutations can be demonstrated (3).

CAPS are classified as hereditary autoinflammatory disorders because of the absence of high-titer autoantibodies and antigen-specific T cells, and they are characterized by genetically determined dysregulation of the innate immune system (4). In particular, the inflammation in CAPS is driven by excessive release of interleukin-1β (IL-1β) (5). IL-1β release is normally regulated by an intracellular protein complex known as the inflammasome. The inflammasome, which is composed of cryopyrin, caspase 1, Cardinal, and apoptosis-associated speck-like protein containing a caspase activation and recruitment domain, cleaves proIL-1β, allowing the release of mature IL-1β (5–7). NLRP3 mutations are hypothesized to encode an aberrant cryopyrin protein that dysregulates the inflammasome, generating the inappropriate release of IL-1β and, thus, leading to the excessive multisystem inflammation that is responsible for the symptoms associated with CAPS (5). Elucidation of the pathway involved in IL-1 action has allowed for the development of IL-1–targeted therapies with different modes of action, such as reducing IL-1 maturation and release via inhibition of caspase 1 (e.g., VX-765) or blocking the interaction of IL-1 with the IL-1 receptor complex with the use of a soluble decoy receptor (e.g., rilonacept), anti–IL-1 antibody, or IL-1 receptor antagonist (IL-1Ra; anakinra).

CAPS are rare and are not well recognized among physicians, with only several hundred cases identified in the US (8). FCAS, MWS, and NOMID/CINCA syndrome all involve fever, urticaria-like rash, arthralgia, myalgia, fatigue, and conjunctivitis (3, 8). In patients with FCAS, substantial chronic symptoms are punctuated by acute flares triggered by exposure to cool ambient temperatures (e.g., air conditioning or a cool breeze), whereas in patients with MWS, the symptoms are more constant, and flares are unpredictable. In addition, patients with MWS often develop progressive neurosensory hearing loss (8, 9).

Patients with CAPS have chronically elevated levels of acute-phase proteins, most notably, serum amyloid A (SAA) and C-reactive protein (CRP) (10–12). Elevated SAA levels are associated with reactive amyloidosis and renal failure, a severe complication of CAPS and other hereditary autoinflammatory disorders. The risk of developing amyloidosis is elevated in patients with CAPS. While the incidence of amyloidosis is ∼2% in patients with FCAS (9), this potentially life-threatening condition affects 25% of patients with MWS, reflecting the especially intense and prolonged acute-phase response in this disorder (8, 10, 11).

Patients with FCAS and MWS experience frequent, intermittent episodes of incapacitation, which may last from hours to days (9, 13). The symptoms of FCAS and MWS wax and wane, and the severity varies from day to day (9). The often unpredictable occurrence of disease flares and the necessity of focusing on limiting one's exposure to cool temperatures in order to avoid disease flares interfere with the patient's ability to lead a productive life and to participate in normal work, social, and family activities (11, 14, 15). To date, no medications have been approved for the treatment of CAPS. Many patients typically are prescribed or self-medicate with corticosteroids, nonsteroidal antiinflammatory drugs, or antihistamines, which generally do not adequately relieve symptoms (8, 9).

Small, open-label studies of treatment with IL-1Ra have indicated that inhibition of the action of IL-1 reduces symptoms and levels of biomarkers of inflammation in patients with these and other closely related conditions (11, 12, 14). Recently, an open-label pilot study of rilonacept in patients with FCAS showed clinical evidence of efficacy (16). Rilonacept (also known as IL-1 Trap), a long-acting (∼1 week half-life) potent inhibitor of IL-1 action, is a fully human dimeric fusion protein that incorporates the extracellular domains of both of the IL-1 receptor components (IL-1 receptor type I and IL-1 receptor accessory protein) required for IL-1 signaling, which are linked to the Fc portion of IgG1 (17). Rilonacept blocks IL-1 signaling by acting as a soluble decoy receptor that binds IL-1 and prevents its interaction with cell surface receptors.

Based on the results of the pilot study, we hypothesized that rilonacept would reduce the signs, symptoms, and chronic inflammation in patients with CAPS. The 2 phase III studies reported herein, which are the first randomized, double-blind, placebo-controlled trials of any therapy for CAPS, were designed to assess the efficacy and safety of rilonacept in the treatment of the clinical signs, symptoms, and inflammation of FCAS and MWS. Unlike previous studies of CAPS treatments, these trials used a validated patient-reported outcomes instrument (the Daily Health Assessment Form [DHAF]) to assess primary efficacy (18) and were large relative to the total North American population of CAPS patients, enrolling 47 patients.

PATIENTS AND METHODS

  1. Top of page
  2. Abstract
  3. PATIENTS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. AUTHOR CONTRIBUTIONS
  7. ROLE OF THE STUDY SPONSOR
  8. Acknowledgements
  9. REFERENCES

Study design.

Enrolled patients were evaluated in 2 sequential studies (Figure 1). Study 1 was a 6-week, randomized, double-blind, placebo-controlled comparison of the efficacy, safety, and tolerability of rilonacept versus placebo. After a 3-week screening/baseline period, patients with active FCAS or MWS who met the eligibility criteria were randomized in a 1:1 ratio to receive a loading dose of 320 mg of rilonacept (Regeneron Pharmaceuticals, Tarrytown, NY) or placebo via two 2-ml subcutaneous injections on the same day, with subsequent weekly single 2-ml self-administered subcutaneous injections of rilonacept 160 mg or placebo.

thumbnail image

Figure 1. Design of the 2 studies of rilonacept versus placebo in the treatment of patients with cryopyrin-associated periodic syndromes.

Download figure to PowerPoint

Patients who completed study 1 entered directly into study 2, of which there were 2 parts. Part A consisted of 9 weeks of weekly single-blind treatment with subcutaneous rilonacept 160 mg. This was followed by part B, which consisted of 9 weeks of weekly double-blind subcutaneous injections of rilonacept 160 mg or placebo in a randomized (1:1 ratio) withdrawal procedure. Throughout both studies, patients were evaluated in the clinic at 3-week intervals. Patients who completed study 2 were eligible to enter an open-label rilonacept extension period.

The trials were reviewed and approved by the Institutional Review Board at each participating site and were conducted in accordance with the ethical principles articulated in the Declaration of Helsinki, consistent with Good Clinical Practice and applicable regulatory requirements. All patients gave their informed consent.

Study population.

Potential study participants were identified through their previous participation in a study of the natural history of CAPS. Patients were at least 18 years of age, had genetic evidence of an NLRP3 mutation upon DNA sequencing (GeneDx, Gaithersburg, MD), and had classic signs and symptoms of FCAS (recurrent, intermittent fever and rash that were often exacerbated by exposure to generalized cool ambient temperature [natural, artificial, or both]) or MWS (chronic fever and rash of waxing and waning intensity, sometimes exacerbated by exposure to generalized cool ambient temperature). Women of childbearing potential were required to use an accepted contraceptive method. The ability to read, understand, and complete study-related questionnaires was required. DHAFs were to be completed for at least 11 of the 21 days during the screening/baseline period before randomization into study 1.

Key reasons for exclusion were known hypersensitivity to Chinese hamster ovary cell–derived therapeutics or proteins or any of the components of rilonacept; recent vaccination with a live (attenuated) virus; recent treatment (<5 half-lives) with a tumor necrosis factor inhibitor or investigational agent; concurrent treatment with IL-1Ra; chest radiography results consistent with current or previous tuberculosis infection, positive results on an intradermal tuberculin test in the absence of previous tuberculosis prophylaxis, history of listeriosis or active tuberculosis, a persistent chronic or active infection(s) requiring parenteral treatment with antibiotics, antivirals, or antifungals within the previous 4 weeks, or oral treatment with antibiotics, antivirals, or antifungals within 2 weeks before screening; any other active systemic inflammatory condition; evidence of current infection with human immunodeficiency virus, hepatitis B virus, or hepatitis C virus; malignancy within 5 years of screening; a creatinine level >1.5 times the upper limit of normal or an alanine aminotransferase (ALT) or aspartate aminotransferase (AST) level >2.0 times the upper limit of normal, a white blood cell (WBC) count <3.6 × 103/mm3, or a platelet count <150,000/mm3 at the screening visit; and lactation or pregnancy.

Efficacy assessment.

Disease activity was evaluated with the DHAF, a validated 1-page self-administered questionnaire (18) on which patients rated the severity of their key disease-related symptoms (rash, feeling of fever/chills, joint pain, eye redness/pain, and fatigue) over the previous 24 hours and provided their global assessment of disease activity. Two additional questions on the DHAF queried about exposures to cold temperatures and limitations in daily activities because of CAPS. Ratings were provided using linear rating scales on which circles were marked in half-step units from 0 (none; no severity) to 10 (very severe). The investigators also provided their assessments of the patients' disease activity (physician's global assessment). Levels of high-sensitivity CRP (hsCRP) and SAA were also measured.

Safety assessment.

Adverse events (AEs) and vital signs were recorded, along with the results of physical examinations, screening tuberculin skin tests, screening chest radiographs, electrocardiograms (EKGs), routine clinical laboratory tests (Medical Research Laboratories International, Highland Heights, KY), plasma rilonacept determinations, and serum antirilonacept antibody levels.

Data analysis.

Randomization and evaluation periods.

Randomization in both studies included stratification by baseline disease activity in study 1. Stratum 1 included patients with active CAPS, as evidenced by a score of ≥3 on at least 1 key diary symptom during the 21-day screening/baseline period. Stratum 2 included all patients with scores of <3 during screening.

Efficacy end points.

In both studies, the primary end point was the mean change in the mean key symptom score derived from the DHAF from the baseline evaluation period (3-week period prior to randomization) to the end point evaluation period (last 3 weeks of double-blind treatment). For each day, the 5 scores (1 for each symptom) were summed and divided by 5 (daily mean score). For each evaluation period (Figure 1), the daily mean scores were summed and divided by 21 (the number of days in the observation period) to yield a mean key symptom score for the evaluation period.

In both studies, secondary and other end points included mean changes from baseline to the end point period in the number of multisymptom disease flare days (i.e., days when the daily mean key symptom score was >3), the number of single-symptom disease flare days (i.e., days when any single symptom had a score >3), and the maximum severity of any symptom. Other end points included the mean changes from baseline to the end point period in the physician's and patient's global assessments of disease activity, in each symptom score, and in limitations of daily activities, as well as the median changes in hsCRP and SAA levels from the day of randomization to the last day of double-blind treatment. In study 1, an analysis classified patients as 30%, 50%, or 75% responders according to improvements of at least 30%, 50%, or 75%, respectively, in the mean key symptom score from baseline to the end point.

Analysis of safety data.

Standard analyses of safety were performed on AEs, clinical laboratory results, vital signs, and EKG results.

Determination of sample size.

Data from a previous study of the natural history of CAPS (18) were used for estimates of disease activity. Assuming a mean change of –0.18 for the placebo group and –1.60 for the rilonacept group in the mean key symptom score from baseline to end point, a common standard deviation of 1.4, and an alpha value of 0.05 (2-sided), a sample size of 22 patients per group in stratum 1 (score of ≥3 on at least 1 key diary symptom during the 21-day screening/baseline period) would provide at least 90% power to detect this difference using a 2-sample t-test. Approximately 60 patients were to be screened, and ∼50 patients were to be randomly allocated to receive rilonacept 160 mg or placebo in the initial study 1.

Statistical analysis.

For the primary efficacy analyses, a conditional sequence of hypothesis tests was performed to permit the initial evaluation of efficacy in a population with adequate disease activity (stratum 1). Conditional on statistical significance in stratum 1 patients, subsequent testing was to include all patients.

Multiple testing was controlled for through the conditional sequence of hypothesis testing. Analysis of covariance with the initial baseline mean key symptom score as the covariate was used. Efficacy variables that were proportions were analyzed by means of Fisher's exact test. AEs were summarized for study 1 and for both parts of study 2.

RESULTS

  1. Top of page
  2. Abstract
  3. PATIENTS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. AUTHOR CONTRIBUTIONS
  7. ROLE OF THE STUDY SPONSOR
  8. Acknowledgements
  9. REFERENCES

Characteristics of the study patients.

Of the 53 patients screened for study 1, a total of 47 patients (44 with FCAS and 3 with MWS) at 22 sites across the continental US were randomly assigned to receive study medication (23 rilonacept and 24 placebo). Forty-six patients were stratified into stratum 1 (encompassing patients with active CAPS, as evidenced by a score of ≥3 on at least 1 key diary symptom during the 21-day screening/baseline period); a single patient was enrolled in stratum 2 (encompassing patients with scores of <3 on key diary symptoms during screening; this patient was assigned to placebo in study 1). Of the 47 patients who enrolled, 44 of them completed both studies (Figure 2). Three patients receiving rilonacept withdrew from the studies: 1 during study 1 because of hepatitis C virus infection that predated the initial dose of study drug (rilonacept), 1 during study 2 part A because of noncompliance, and 1 during study 2 part B because of an AE (worsening finger joint pain during treatment with rilonacept). Table 1 summarizes the demographic and clinical characteristics of the study patients.

thumbnail image

Figure 2. Flow chart showing the progression of the patients from randomization into study 1 through completion of parts A and B of study 2.

Download figure to PowerPoint

Table 1. Demographic and clinical characteristics of the enrolled patients*
ParameterStudy 1Study 2 part B
Rilonacept (n = 23)Placebo (n = 24)Rilonacept (n = 22)Placebo (n = 23)
  • *

    All patients were white and were positive for the CIAS1 gene mutation. Forty-four patients had familial cold autoinflammatory syndrome, and 3 patients had Muckle-Wells syndrome.

Age, mean (range) years46 (22–76)56 (24–78)52 (26–78)50 (22–78)
Sex, no (%)    
 Men8 (35)8 (33)8 (36)7 (30)
 Women15 (65)16 (67)14 (64)16 (70)
Height, mean (range) cm168 (155–190)169 (158–183)170 (155–190)167 (158–179)
Weight, mean (range) kg72 (50–114)76 (50–119)76 (50–119)74 (50–114)
Baseline key symptom score, mean (range)3.1 (0.73–8.2)2.4 (0.64–5.4)0.3 (0–1)0.2 (0–2.1)

Compliance with the medication regimen.

The percentage of patients taking all scheduled doses of study medication was >90% during study 1. The mean numbers of doses taken were 6.9 of 7.0 in the placebo group and 7.0 of 7.0 in the rilonacept group. During study 2 part A, 89% of patients took all scheduled doses of rilonacept, and the mean number of doses taken was 8.9 of 9.0. During study 2 part B, 84% of patients took all scheduled doses of study medication, and the mean number of doses taken was 8.4 of 9.0 in the rilonacept group and 9.0 of 9.0 in the placebo group.

Efficacy.

Study 1.

Rilonacept was significantly superior to placebo in the primary analysis of change from baseline in the composite mean key symptom score for the 46 patients in stratum 1 (P < 0.0001 versus placebo) and for the subsequent conditional analysis of the 47 patients in strata 1 and 2 combined (P < 0.0001 versus placebo). Because of this finding, and because only 1 patient was randomized into stratum 2, only changes for the combined strata are reported for both studies.

The mean change from baseline in the key symptom score was –2.6 (group mean change of –84%) in the rilonacept group and –0.3 (group mean change of –13%) in the placebo group (Table 2). Relative to placebo, patients who received rilonacept noted a reduction in the key symptom score within 1 day of starting treatment (Figure 3A). Subgroup evaluations (data not shown) by sex (men versus women), age (≥51 years versus <51 years), and key symptom score at baseline (<2.4 versus ≥2.4) indicated the significant superiority of rilonacept for the primary efficacy variable in all subgroups (all P ≤ 0.002).

Table 2. Results of primary, secondary, and other efficacy measures*
 Study 1PStudy 2 part BP
Rilonacept (n = 23)Placebo (n = 24)Rilonacept (n = 22)Placebo (n = 23)
BaselineEnd pointBaselineEnd pointBaselineEnd pointBaselineEnd point
  • *

    For disease symptoms, score ranges of 1–3, >3–7, and >7–10 were categorized as mild, moderate, and severe, respectively. Values are the mean, except for the values for high-sensitivity C-reactive protein (hsCRP; reference range 0.0–8.4 mg/liter) and serum amyloid A (SAA; reference range 0.7–6.4 mg/liter), which are median. NS = not significant.

  • Between-group comparison based on analysis of covariance main-effects model.

Key symptom score3.10.52.42.1 0.30.40.21.2 
 P for change from baseline<0.0001NS<0.0001NS<0.00010.0002
No. of multisymptom disease flare days8.60.16.25.0 000.11.9 
 P for change from baseline<0.0001NS<0.0001NS0.0070.003
No. of single-symptom disease flare days13.21.111.610.4 0.82.10.86.3 
 P for change from baseline<0.0001NS<0.0001NS0.00050.01
Maximum score for any symptom8.12.78.17.6 2.52.21.55.0 
 P for change from baseline<0.0001NS<0.0001NS<0.0001<0.0001
Individual key symptoms          
 Feeling of fever/chills3.00.42.01.7 0.10.20.11.1 
  P for change from baseline<0.001NS<0.0001NS0.00150.008
 Rash4.00.53.53.3 0.40.60.42.3 
  P for change from baseline<0.001NS<0.0001NS<0.0001<0.0001
 Eye redness/pain1.70.21.31.2 0.20.20.10.4 
  P for change from baseline<0.001NS0.0001NS0.0270.03
 Fatigue3.60.82.72.3 0.50.50.31.1 
  P for change from baseline<0.0010.03<0.0001NS0.00050.0005
 Joint pain3.10.52.62.0 0.40.50.30.9 
  P for change from baseline<0.001NS<0.0001NS0.00130.02
Physician's global assessment5.61.54.75.0 1.31.41.04.3 
 P for change from baseline<0.001NS<0.0001NS<0.0001<0.0001
Patient's global assessment3.60.93.12.7 0.50.70.41.7 
 P for change from baseline<0.001NS<0.0001NS0.00040.003
Limitation of daily activities3.00.82.41.6 0.50.50.10.8 
 P for change from baseline<0.0010.0040.006NS0.02610.05
hsCRP, mg/liter20.11.325.221.8 1.51.71.516.3 
 P for change from baseline<0.001NS<0.0001NS0.00020.0001
SAA, mg/liter49.52.563.539.7 2.52.83.828.3 
 P for change from baseline<0.001NS0.006NS0.01910.01
thumbnail image

Figure 3. Changes in key symptom scores from baseline with rilonacept or placebo therapy during study 1 and parts A and B of study 2.

Download figure to PowerPoint

Table 2 summarizes the results of the primary, secondary, and other efficacy outcome measures by treatment group. Rilonacept was significantly superior to placebo in decreasing the numbers of multisymptom and single-symptom disease flare days (P < 0.0001 for each comparison), the maximum score for any single symptom during the entire study period (P < 0.0001), and the mean scores for each of the 5 key symptoms (P = 0.0001). Relative to placebo, rilonacept also significantly improved the physician's and patient's global assessments of disease activity (P < 0.0001 for each comparison), decreased limitations in patients' daily activities (P = 0.006), and decreased the levels of both hsCRP (P < 0.0001) and SAA (P = 0.006). Of the patients taking rilonacept, 96% experienced at least a 30% reduction in the mean key symptom score, as compared with 29% of the patients taking placebo (P < 0.0001). A total of 87% of the rilonacept-treated patients and 8% of the placebo-treated patients experienced at least a 50% reduction in the mean key symptom score (P < 0.0001), with 70% and 0%, respectively, experiencing at least a 75% reduction (P < 0.0001) (data not shown).

Study 2.

Patients who were assigned to rilonacept therapy during study 1 and who continued to receive rilonacept during study 2 part A maintained the benefits demonstrated in study 1. Patients who were assigned to placebo therapy during study 1 and who were switched to rilonacept during study 2 part A showed rapid improvements in their symptoms (Figure 3A) and reductions in their hsCRP and SAA values during this period of study (Table 2).

During the randomized withdrawal period (study 2 part B), rilonacept was significantly superior to placebo in maintaining the low mean key symptom scores that had been achieved during the single-blind rilonacept treatment period (Figure 3B and Table 2) (P = 0.0002), in maintaining low numbers of multisymptom disease flare days (P = 0.003) and single-symptom disease flare days (P = 0.01), and a low maximum score for any single symptom (P < 0.0001). Rilonacept was also significantly superior to placebo in maintaining a low score for each of the key symptoms. Compared with placebo, rilonacept was also significantly superior in maintaining improvements in the physician's and patient's global assessments of disease activity and in their limitations of daily activities, as well as maintaining low levels of hsCRP and SAA (Table 2).

Of note, these robust results were observed despite incorrect allocation of study medications to 11 of the 45 patients during the first 3 weeks of study 2 part B (misallocated according to study 1 randomization), which biased the study against demonstrating a difference between rilonacept and placebo because some individuals erroneously received rilonacept instead of placebo and vice versa. In both studies 1 and 2, the 3 patients with a diagnosis of MWS responded robustly to rilonacept in a manner similar to that of the remaining patients with a diagnosis of FCAS.

Safety.

Adverse events.

In study 1, 17 patients receiving rilonacept (74%) and 13 patients receiving placebo (54%) had treatment-emergent AEs (Table 3). The most common AEs were injection site reactions and upper respiratory tract infections. The most common AEs reported with rilonacept therapy during study 2 part B were injection site reaction, headache, arthralgia, and diarrhea. One patient discontinued during study 1 because of preexisting elevations of transaminase levels, which were subsequently determined to be due to hepatitis C virus infection. During study 2 part B, 1 patient discontinued rilonacept because of an AE (worsening finger joint pain), which was not considered by the investigator to be related to the study drug.

Table 3. Treatment-emergent adverse events reported by at least 2 patients
Adverse eventNo. (%) taking rilonaceptNo. (%) taking placebo
  • *

    Includes erythema, pruritus, swelling, bruising, inflammation, mass, or pain at the injection site.

  • Includes erythema, hemorrhage, edema, bruising, mass, or pruritus at the injection site.

Study 1  
 No. of patients2324
 Any treatment-emergent adverse event17 (74)13 (54)
  Injection site reaction*11 (48)3 (13)
  Upper respiratory tract infection6 (26)1 (4)
  Diarrhea1 (4)3 (13)
  Nausea1 (4)3 (13)
  Sinusitis2 (9)1 (4)
  Abdominal pain, upper abdomen02 (8)
  Cough2 (9)0
  Hypoesthesia2 (9)0
  Stomach discomfort1 (4)1 (4)
  Urinary tract infection1 (4)1 (4)
Study 2 part B  
 No. of patients2223
 Any treatment-emergent adverse event15 (68)13 (57)
  Injection site reaction8 (36)3 (13)
  Headache3 (14)2 (9)
  Arthralgia3 (14)0
  Diarrhea3 (14)0
  Rash2 (9)1 (4)
  Abdominal pain, upper abdomen2 (9)0
  Dyspepsia2 (9)0
  Hyperhidrosis1 (5)1 (4)
  Sinus congestion1 (5)1 (4)
  Sinusitis02 (9)
  Urinary tract infection02 (9)

No serious AEs were reported during study 1 or study 2 part B. One serious AE (worsening sciatica) was reported during study 2 part A, but this was not considered to be related to the study medication. No deaths occurred during the 2 study periods, although during the subsequent open-label extension period (after >1 year of treatment), 1 elderly patient died after developing sinusitis and Streptococcus pneumoniae meningitis, which were not considered to be related to the study drug.

Injection site reactions.

In study 1, 48% of the rilonacept group and 13% of the placebo group reported injection site reactions; these were most frequently reported as mild erythema, pruritus, and swelling (Table 3). This type of AE was reported in 35% of the patients during study 2 part A, and during study 2 part B, 36% of the rilonacept group and 13% of the placebo group reported this AE.

Infections.

No opportunistic infections or infections requiring treatment with parenteral antibiotics were reported during the 2 studies. No patients withdrew participation because of treatment-emergent infections, although 2 patients temporarily discontinued the study drug for this reason. The incidence of patients reporting any type of infection was higher during study 1 in patients treated with rilonacept as compared with patients treated with placebo (48% versus 17%), with upper respiratory tract infections being the most frequently reported infection (Table 3). However, in study 2 part B, 18% of patients receiving rilonacept reported infections, as compared with 22% of those receiving placebo. In study 2 part A, during which all patients were treated with rilonacept, 20% of patients reported infections, of which one was a Staphylococcus aureus furuncle. All cases of infection except one (severe bronchitis) were rated as mild or moderate in severity.

Changes in vital signs and EKG results.

Vital signs showed modest transient fluctuations, with no discernible trends for rilonacept treatment compared with placebo treatment. A trend toward an increase in weight with rilonacept therapy (mean increase of ∼1 kg relative to placebo therapy) appeared to diminish with continued treatment. No deleterious effects of rilonacept on the EKG results were evident.

Clinical laboratory values.

Rilonacept was associated with small but statistically significant increases in hemoglobin levels and red blood cell (RBC) counts and with modest decreases in neutrophil and platelet counts. Small increases in mean ALT and AST levels, and small decreases in mean alkaline phosphatase levels were reported with rilonacept, but these changes were not clinically significant. Rilonacept was also associated with mean increases in total cholesterol levels of ∼15 mg/dl (data not shown). Many of the changes in these laboratory values may reflect corrections in the chronic acute-phase response induced by the systemic inflammation of CAPS (see Discussion).

Antirilonacept antibodies.

In assays based on full-length rilonacept receptor or only on the extracellular domain portions of the IL-1 receptor complex of rilonacept, 28% (13 of 46) and 43% (20 of 46), respectively, of the patients treated with rilonacept for up to 24 weeks demonstrated rilonacept-specific low-titer antibodies on at least 1 occasion. Some of these responses were transient. Development of antirilonacept antibodies did not appear to affect the efficacy (i.e., symptom scores or CRP or SAA levels) or the safety profile.

DISCUSSION

  1. Top of page
  2. Abstract
  3. PATIENTS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. AUTHOR CONTRIBUTIONS
  7. ROLE OF THE STUDY SPONSOR
  8. Acknowledgements
  9. REFERENCES

Findings of small, open-label treatment studies have suggested that inhibiting the action of IL-1 may produce significant therapeutic benefit in patients with CAPS. The results reported herein of 2 placebo-controlled trials of rilonacept in which a validated instrument for assessing CAPS symptoms (the DHAF) was used provide definitive evidence of the efficacy of the IL-1 inhibitor rilonacept in patients with CAPS. In both trials, rilonacept markedly decreased the signs and symptoms of CAPS.

In study 1, rilonacept produced rapid and profound improvements in, or resolution of, the signs and symptoms of CAPS as compared with placebo, whether determined by patient-reported outcomes, physician assessments, or objective measures of inflammation. These results are consistent with the findings of an open-label pilot study of rilonacept in 5 patients with FCAS (16). Rilonacept also reduced hsCRP levels and normalized elevated SAA concentrations, an important risk factor for amyloidosis. At the end of 9 weeks of study 2 part A, in which all patients were treated with rilonacept in single-blind manner, efficacy results for all patients were generally equivalent irrespective of previous assignment to rilonacept or placebo in study 1. In study 2 part B, withdrawal of rilonacept (treatment with placebo) led to a gradual return of disease activity and increases in hsCRP and SAA levels, whereas continued treatment with rilonacept maintained improvements in disease activity.

During study 1, rilonacept virtually abolished the occurrence of both multisymptom and single-symptom flare days, which occurred frequently with placebo treatment. Maintenance of the reduction in the numbers of these flare days with continued rilonacept treatment was demonstrated in study 2 part B. One would expect this degree of reduction of a significant disease burden to translate directly into decreases in limitations of work, social, and other activities. Indeed, along with the reported reduction in symptoms, patients treated with rilonacept reported a reduction in limitations in daily activities.

The majority of patients in the study had FCAS, in which flares are typically provoked by cool temperatures. Nonetheless, significant disease activity, as well as limitation of patients' activities, were documented quantitatively in these studies over periods that included warmer summer months as well as colder winter months. Of note, relative to placebo, reduction in the disease signs and symptoms and reduction in the limitations of activities were observed with rilonacept treatment during both cold and warm periods. These results are consistent with the recent suggestion that FCAS patients have substantial underlying disease-related symptoms between disease exacerbations induced by exposure to cold temperatures (19).

Treatment with rilonacept was generally well-tolerated. Injection site reactions, the most common AE, were reported in approximately one-third to one-half of patients, a rate that was 3-fold higher than that reported with placebo treatment. However, there were no severe injection site reactions and no study discontinuations for this reason. Patients treated with rilonacept or placebo reported infections, primarily upper respiratory, at similar rates, except during study 1 (the initial 6-week double-blind treatment phase). Although the treating physician considered the meningitis-related death that occurred during the extension phase of the studies to be unrelated to rilonacept treatment, the potential risk of serious life-threatening infections with therapies that target the IL-1 pathway should be taken into account.

Consistent with the potential for immunogenicity of protein therapeutics, some patients treated with rilonacept developed detectable antirilonacept antibodies, as demonstrated by highly sensitive assays. Since these antibodies had no evident effect on the efficacy or safety parameters, their clinical significance is uncertain. Treatment with rilonacept was associated with small-to-modest increases in hemoglobin levels, RBC counts, and total cholesterol levels and with decreases in platelet, WBC, and neutrophil counts. Since chronic inflammation and the linked acute-phase response are known to be associated with erythropoietic suppression and elevation of WBC and platelet counts, these changes likely reflect a correction of the underlying inflammatory state. Consistent with the notion that the hematologic changes are likely to be class effects due to a treatment-related decrease in chronic inflammation, changes in these parameters have been reported in patients treated with IL-1Ra (20). The modest increase in total cholesterol levels observed with rilonacept therapy may also result from a treatment-related decrease in inflammation, since total cholesterol levels have been reported to be reduced as part of the acute-phase response (21), and other anti–proinflammatory cytokine agents have been reported to increase cholesterol levels in patients with other inflammatory diseases (22–25).

In conclusion, treatment with rilonacept at a dosage of 160 mg/week administered subcutaneously markedly reduced the clinical and laboratory signs and symptoms of FCAS and MWS and demonstrated a generally favorable safety and tolerability profile. The strength of these conclusions is enhanced by the relatively large percentage of the known total CAPS population that participated in the studies and the use of a validated instrument (the DHAF) for assessing disease symptoms and signs to determine efficacy.

AUTHOR CONTRIBUTIONS

  1. Top of page
  2. Abstract
  3. PATIENTS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. AUTHOR CONTRIBUTIONS
  7. ROLE OF THE STUDY SPONSOR
  8. Acknowledgements
  9. REFERENCES

Dr. Hoffman 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. Hoffman, Yancopoulos, Stahl, Mellis.

Acquisition of data. Throne, Sebai, Kivitz, Kavanaugh, Weinstein, Mellis.

Analysis and interpretation of data. Hoffman, Amar, Weinstein, Yancopoulos, Stahl, Mellis.

Manuscript preparation. Hoffman, Kivitz, Kavanaugh, Weinstein, Belomestnov, Yancopoulos, Stahl, Mellis.

Statistical analysis. Belomestnov, Yancopoulos.

ROLE OF THE STUDY SPONSOR

  1. Top of page
  2. Abstract
  3. PATIENTS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. AUTHOR CONTRIBUTIONS
  7. ROLE OF THE STUDY SPONSOR
  8. Acknowledgements
  9. REFERENCES

Regeneron Pharmaceuticals, Inc. proposed the study and was involved throughout the process of the study design, data acquisition and analysis, statistical analysis, and manuscript preparation. Regeneron Pharmaceuticals, Inc. approved the content of the manuscript and agreed to submit the manuscript for publication.

Acknowledgements

  1. Top of page
  2. Abstract
  3. PATIENTS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. AUTHOR CONTRIBUTIONS
  7. ROLE OF THE STUDY SPONSOR
  8. Acknowledgements
  9. REFERENCES

The authors wish to acknowledge the late Eugene Boling, MD (Upland, CA), an investigator who contributed substantially to patient recruitment in this study, as well as all of the other study investigators, the coordinators, and the patients. In addition to Dr. Boling and the authors, the study investigators were as follows: Bruce Berwald, MD (St. Louis, MO), Robert Cartwright, MD (Columbus, GA), Stanley Cohen, MD (Dallas, TX), W. Travis Ellison, MD (Greer, SC), Lansing Ellsworth, MD (Cedar City, UT), Darrell Fiske, MD, (Stuart, FL), Ronald Fogel, MD (Chesterfield, MI), Santosh Gil, MD (Aurora, IL), Susanna Goldstein, MD (New York, NY), Maria Greenwald, MD (Palm Desert, CA), Joe Hargrove, MD (Little Rock, AR), Wayne Larson, MD (Lakewood, WA), Michael Noss, MD (Cincinnati, OH), Stephen Pollard, MD (Louisville, KY), John Rubino, MD (Raleigh, NC), William Smith, MD (Chattanooga, TN), and Willard Washburne, MD (Shreveport LA). The authors also wish to thank Raphaela Goldbach-Mansky, MD (National Institute of Arthritis and Musculoskeletal and Skin Diseases, National Institutes of Health, Bethesda, MD) for important contributions from the pilot study and for helpful discussions and Richard Greenlaw for assistance with manuscript preparation.

REFERENCES

  1. Top of page
  2. Abstract
  3. PATIENTS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. AUTHOR CONTRIBUTIONS
  7. ROLE OF THE STUDY SPONSOR
  8. Acknowledgements
  9. REFERENCES