In January 2009, UST (CNTO 1275, Stelara; Janssen Cilag) was granted marketing authorization by the European Commission for the treatment of moderate to severe chronic plaque psoriasis in adults who failed to respond to, who have a contraindication to or who are intolerant of systemic oral immunosuppressants. Unlike the other biological anti-psoriatic agents already brought to market, which all targeted TNFα, UST was the first-in-class anti-IL agent for psoriasis, representing an important milestone in rational drug design. UST is a fully human IgG1κ monoclonal antibody that inhibits IL-12 and IL-23 activity by binding with high affinity and specificity to their shared p40 subunit. IL-12/23 bioactivity is thus inhibited by preventing their binding to IL-12 receptor β1 (IL-12Rβ1) on the surface of immune cells.
Phase III Clinical Trials
Following on from the promise shown in the phase II study, the safety and efficacy of UST were further assessed in three large phase III clinical trials involving 2,899 adult patients with moderate to severe psoriasis (PASI >12, PGA ≥3 or 10% body surface area involvement) of at least 6 months duration and who were candidates for systemic immunosuppression or phototherapy. Run in parallel, both PHOENIX I (ClinicalTrials.gov #NCT00267969) [
28] and PHOENIX II (ClinicalTrials.gov #NCT00307437) [
29] were multicenter, randomized, double-blind, placebo-controlled trials with similar objectives and methods and a primary endpoint of PASI 75 response from baseline at week 12. The same primary endpoint was selected for the third, the ACCEPT (ClinicalTrials.gov #NCT00454584) trial, but differed in that it compared UST with etanercept in place of placebo [
30].
PHOENIX I, a 76 week study, involved 766 patients, 53% of which were either non-responsive to, intolerant of or had a contraindication to other systemic therapy [
28]. Participants were initially randomized (1:1:1) to placebo or active treatment with UST at either 45 mg or 90 mg subcutaneously at week 0, 4, and then every 12 weeks thereafter. Baseline randomization was stratified by study site, weight (≤90 or >90 mg) and the number of systemic therapies to which the patient had had an inadequate response, intolerance or contraindication (<3 or ≥3).
Patients in the active treatment group who achieved a PASI 75 response at both weeks 28 and 40 were re-randomized at week 40 to maintenance UST (same dose as initial stage) or withdrawal from treatment (placebo administered) until loss of response. Randomization at week 40 was also based on study site and patient weight. Patients in this group who had achieved a partial (PASI 50–PASI 74) response at week 40 were adjusted to a dosing interval of every 8 weeks. Patients randomized to receive placebo at week 0 and week 4 crossed over to receive UST (45 or 90 mg) at weeks 12 and 16, followed by dosing every 12 weeks thereafter. This study design allowed not only comparison of UST against placebo, but also long-term efficacy, duration of therapeutic effect after drug withdrawal and possible dose escalation in partial responders [
28].
At the primary endpoint (week 12), 67.1% of those receiving the 45 mg dose and 66.4% of those receiving 90 mg achieved PASI 75, compared to 3.1% of the placebo group (
P < 0.0001). Improvement was rapid, with many in the active treatment groups, regardless of dose, achieving PASI 50 by week 2. Maximum efficacy was observed at week 24 for both dosing regimens (76.1 and 85.0% PASI 75 response for 45 and 90 mg, respectively), with similar findings in the group initially assigned to placebo after crossing over to active treatment at week 12 [
28].
After re-randomization at week 40, either maintenance therapy or withdrawal, preservation of PASI 75 was significantly greater in those receiving continuous UST therapy (84% at week 76) compared with the treatment withdrawal group (19% at week 76). In the latter, PASI scores began to deteriorate by week 44 (16 weeks after last injection), and accelerated after week 52. Withdrawn patients were re-treated with their initial dose after their PASI 50 response was lost. A total of 195 patients needed to restart therapy and 85.6% regained their PASI 75 after 12 weeks of re-treatment [
28].
Expanding on the initial PHOENIX I trial data reported at week 76, all patients were subsequently followed to week 244 (5 years) to assess longer-term safety and efficacy [
31,
32]. Overall, 68.7% (
n = 517) of the initial overall population of 753 (who had received at least one dose of UST in PHOENIX I) were evaluated. Initial clinical responses were generally maintained through week 244 (PASI 75: 63.4 and 72.0%; PASI 90: 39.7 and 49.0%; PASI 100: 21.6 and 26.4% for patients receiving 45 or 90 mg, respectively) [
31]. At week 264, analysis of 8998 patient years of follow-up demonstrated event rates (per 100 patient years; 45 and 90 mg, respectively) for all adverse events (242.6, 225.3), serious adverse events (7.0, 7.2), serious infections (0.98, 1.19), non-melanoma skin cancers, or NMSCs (0.56, 0.36), other malignancies (0.59, 0.61), and major adverse cardiovascular events, or MACE (0.56, 0.36), that were comparable between the two dosing groups. No increasing trend in any adverse events was seen over time, and the rates of overall mortality and other malignancies were comparable with the general population of the United States [
32].
Nail involvement may be present in up to 80% of patients with psoriatic disease and is notoriously difficult to treat, leading to high psychosocial embarrassment [
33] and in severe cases, functional limitation [
34]. Improvements in fingernail psoriasis were assessed in the PHOENIX I cohort using the Nail Psoriasis Severity Index (NAPSI) on a target fingernail in addition to a nail PGA and assessment of the mean number of nails involved [
35]. Of the 766 randomized, 545 had nail psoriasis. By week 24, the percentage improvement in NAPSI from baseline was 46.5 and 48.7% for UST 45 and 90 mg, respectively. Improvements in the less sensitive nail PGA scores were generally not observed in the overall nail psoriasis cohort at week 12; however, substantial improvements were noted at week 24, with the majority of patients with a PGA ≥3 at baseline achieving improvement by at least one point. In the 45 and 90 mg groups, 77.0 and 75.0%, respectively, of patients with moderate nail disease (PGA 3) improved to mild (PGA 2) or clear (PGA 1) by week 24.
The second large phase III clinical trial, PHOENIX II, recruited 1,230 patients and 61% were either non-responsive to, intolerant of or had a contraindication to other systemic therapy [
29]. Like PHOENIX I, patients were randomized to one of three arms; 45 mg or 90 mg subcutaneously at week 0, 4 and every 12 weeks, or placebo at weeks 0 and 4 and then crossover to active therapy (randomized 1:1 to either 45 or 90 mg) at week 12 (with loading doses at week 12 and 16, followed by injections every 12 weeks thereafter). At week 28, patients were deemed responders (PASI 75 response achieved), partial responders (PASI 50–74) or non-responders (PASI < 50). Responders continued treatment at the same dose every 12 weeks, non-responders discontinued therapy, and partial responders were re-randomized to either continue their current regimen or reduce their dosing interval to every 8 weeks. Stratification was as described for PHOENIX I. The partial responder group permitted analysis of the number of visits between weeks 28 and 52 where PASI 75 was achieved for the two different dosing schedules.
At week 12 (primary endpoint), 66.7% of patients receiving UST 45 mg and 75.7% of patients receiving 90 mg every 12 weeks achieved PASI 75 (
P < 0.0001), compared with 3.7% of participant receiving placebo. Maximum efficacy was seen around week 20 for both doses (PASI 75 in 74.9 and 83.5% for 45 and 90 mg, respectively), with similar outcomes seen in the placebo group after crossing over to active therapy. In those who achieved PASI 75 by week 28, the improvement was maintained until the end of the study (week 52). In all, the median clinical response at the end of the study was PASI 95 for those in the 45 mg group and PASI 96 for those in the 90 mg group [
29].
Partial responders accounted for 22.7% of those receiving 45 mg every 12 weeks, and 15.8% of those receiving 90 mg. At baseline, compared to responders, these individuals were of a greater body weight, had more severe PGA scores, a longer duration of psoriasis, a greater incidence of PsA, a higher failure rate with previous systemic immunosuppressants and lower serum drug levels at week 28. For those receiving 90 mg, a reduction in the dosing interval from 12 to 8 weeks did equate to a greater number of visits where a PASI 75 response was achieved, but this was not the case for those receiving 45 mg [
29].
In terms of safety, both PHOENIX I and II reported similar outcomes during the placebo-controlled phase. Adverse events occurred in 278 (54.5%) of the 510 patients receiving UST in PHOENIX I [
28] and 414 (50.5%) of 820 patients in PHOENIX II [
29]. This is compared with 48.2 and 49.8% in their respective placebo groups. Serious adverse events occurred in similar proportions in both trials and with similar low frequencies between the UST and placebo treated arms (1.2% UST vs. 0.8% placebo in PHOENIX I; 1.6% UST vs. 2.0% placebo in PHOENIX II). In PHOENIX I, the pattern of adverse events was much the same in the placebo crossover and randomized withdrawal phases as it was in the placebo-controlled phase [
28]. Rates of antibody formation to UST were found in 5.1% of patients by the end of week 76 (PHOENIX I) and 5.4% of patients by the end of week 52 (PHOENIX II), and in both trials, these were mostly of low titer.
The ACCEPT phase III clinical trial differed from the PHOENIX trials in that the safety and efficacy of UST were compared with an active comparator (etanercept) rather than placebo [
30]. In this 64 week trial, 903 patients were randomized (3:5:5 ratio) to receive subcutaneous injections of UST (45 or 90 mg) at weeks 0 and 4, or etanercept (50 mg) twice weekly for 12 weeks. Randomization was stratified according to study site and baseline weight (<90 or ≥90 mg). Patients were aware of their treatment, but study assessors remained blinded. At week 12, patients in the etanercept group who did not respond (classified as moderate, marked or severe psoriasis on the PGA) were given 90 mg UST at weeks 16 and 20, and those who did not respond in the UST group were given one further additional dose of UST at week 16. For those who did respond (classified as clear, minimal, or mild) at week 12, treatment was withdrawn. If psoriasis recurred and was graded moderate, marked or severe, patients were retreated with UST, regardless of initial therapy.
At week 12, 67.5 and 73.8% of patients receiving 45 and 90 mg UST respectively achieved PASI 75, compared with 56.8% of those receiving etanercept (
P = 0.01 and
P < 0.001, respectively), and the time to improvement was more rapid in those treated with UST. PASI 90 responses were achieved in 36.4% of patients receiving 45 mg UST, 44.7% of patients receiving 90 mg UST and 23.1% of patients receiving etanercept (
P < 0.001 for both). Amongst those patients who were deemed non-responders to etanercept, 48.9% achieved PASI 75 and 23.4% achieved PASI 90 12 weeks after crossing over to UST. For those who were graded as responders at week 12 and had therapy withdrawn, the median time to recurrence was 14.4 weeks (45 mg UST), 18.1 weeks (90 mg UST) or 7.3 weeks (etanercept). Of the 633 patients who were retreated after re-emergence of moderate to severe psoriasis, 534 were classed as having mild, minimal or no psoriasis within 12 weeks [
30].
Adverse events occurred with similar frequency across all treatment groups, with at least one event in 70.0% of etanercept-treated participants, 66.0% in the 45 mg UST group and 69.2% in those receiving 90 mg UST. Most adverse events were classed as minor, with only 12 patients (4 in each group) from the 903 recruited having a major event. Overall, discontinuation of therapy was necessary in similar proportions, ranging from 1.2 to 2.3%. A noticeable discrepancy was seen in injection site reactions (24.8% of patients who received etanercept as compared with 4.3% (45 mg) and 3.7% (90 mg) of patients receiving UST), although it is worth acknowledging the higher number of injections necessitated by the dosing schedule of the etanercept arm. Through to week 12, infections occurred at comparable rates in the three treatment groups (29.1, 30.6, and 29.7% in the groups that received etanercept, 45 mg UST, and 90 mg UST, respectively) and this was relatively consistent to the end of the trial. NMSCs occurred only in patients treated with UST but at low numbers (three by week 12 and a further nine by week 64). Quality of life indices were not recorded in the ACCEPT trial [
30].
Several other smaller phase III clinical trials have assessed the safety and efficacy of UST in non-western populations and found similar clinical responses. In the PEARL trial, 121 Taiwanese and Korean patients with moderate-to-severe psoriasis were enrolled into a 36 week, multicenter, double-blind, placebo-controlled trial to receive UST 45 mg at week 0, 4 and 16, or placebo at week 0 and 4, followed by UST at week 12 and 16 [
36]. At the primary endpoint (week 12), PASI 75 was achieved by 67.2% in the UST-treated group, and 5.0% in the placebo arm (
P < 0.001). Efficacy was maintained through to week 28 in the UST group. Adverse events were similar between the groups, with exception of abnormal hepatic function, which was related to concomitant isoniazid treatment for latent tuberculosis. No deaths, malignancies, or MACE were reported. An identical study design was employed in a 72 week phase II/III clinical trial involving 158 Japanese patients, with the addition of 90 mg UST arm [
37]. At week 12, 59.4% and 67.7% of UST 45 mg and 90 mg treated patients achieved PASI 75, compared with 6.5% in the placebo group (
P < 0.0001). By week 12, rates of infections were comparable amongst the groups (UST 45 mg, 20.3%; 90 mg, 24.2%; placebo, 18.8%), and only single cases of serious infections and non-cutaneous malignancies were recorded, both occurring in the 90 mg ustekinumab group. There were no reports of NMSC. Through to week 72, similar rates and types of adverse reactions and serious adverse were reported between the 45 and 90 mg ustekinumab-treated groups.
Quality of Life Response
In PHOENIX I, more than 97% of patients had a score of 1 or more on the DLQI at baseline, and the average score was greater than 10 out of a possible maximum of 30, indicating a significant impact on patients’ quality of life [
44]. Significantly greater proportions of patients receiving UST 45 mg and 90 mg achieved a normalization of DLQI (≤1) compared with placebo (53.2, 52.4, 6.0%, respectively, both
p < 0.001) at week 12. The SF-36 questionnaire revealed similarly impressive results for both the physical (45 mg, 23.1%; 90 mg, 33.7%; placebo, 15.6%) and mental (45 mg, 25.5%; 90 mg, 31.3%; placebo, 14.8%) component scores by week 12 (
P < 0.001). The greatest improvements were found in the bodily pain and social functioning domains. These quality of life improvements were sustained with maintenance UST therapy at one year.
In the PHOENIX II trial, the Hospital and Anxiety Depression Scale (HADS) replaced the SF-36, alongside the DLQI [
45]. At baseline, a high psychological impact of disease was apparent, with 40.3% in the group receiving UST 45 mg, and 26.7% receiving 90 mg, reporting symptoms of anxiety and depression, and 54.6% reporting a DLQI ≥10. By week 12, the absolute mean (±SD) reduction in DLQI was by 9.3 points (±7.1) in the 45 mg group, 10.0 (±6.7) points in the 90 mg group, compared with −0.5 (±5.7) in the placebo arm. The proportion of patients with baseline symptoms of mild to severe anxiety (as assessed by HADS-A) decreased from 38.2 to 25.7% by week 12 in the UST 45 mg group and from 41.0 to 27.1% in the UST 90 mg group (
P < 0.001 vs. placebo), representing a combined relative reduction of 34% from baseline (compared with a 1.4% increase in the placebo group). The prevalence of baseline symptoms of mild to moderate depression (as assessed by HADS-D) decreased from 24.7 to 12.8% by week 12 in the UST 45 mg group and from 31.1 to 12.5% in the 90 mg group (
P < 0.001 vs. placebo), representing a relative reduction of 55% from baseline (compared with an increase of 10% in the placebo group).
Sexual difficulties were specifically analyzed from the DLQI data collated in both PHOENIX I and II [
46]. Impaired sexual function was recorded if any patient scored ‘very much’ or ‘a lot’ for question 9 of the DLQI. 27.1% of women and 20.8% of men reported impaired sexual function at baseline, and this was significantly associated with increased psoriasis severity. At week 12, the overall proportion of patients with sexual difficulties decreased from 22.6 to 2.7%, compared to no change in the placebo arm (
P < 0.001). Patients with a greater mean improvement in PASI score experienced a greater reduction in sexual difficulties caused by psoriasis.