ATLANTA –
Sharon Worcester/MDedge News
Dr. Eric F. Morand
In TULIP-1, which compared intravenous anifrolumab at doses of 300 or 150 mg and placebo given every 4 weeks for 48 weeks, the primary endpoint of SLE Responder Index (SRI) in the 300 mg versus the placebo group was not met, but in post hoc analyses, numeric improvements at thresholds associated with clinical benefit were observed for several secondary outcomes, Richard A. Furie, MD, a professor of medicine at the Hofstra University/Northwell, Hempstead, N.Y., reported during a plenary session at the annual meeting of the American College of Rheumatology.
The findings were published online Nov. 11 in Lancet Rheumatology.
TULIP-2 compared IV anifrolumab at a dose of 300 mg versus placebo every 4 weeks for 48 weeks and demonstrated the superiority of anifrolumab for multiple efficacy endpoints, including the primary study endpoint of British Isles Lupus Assessment Group (BILAG)-based Composite Lupus Assessment (BICLA), Eric F. Morand, MD, PhD, reported during a late-breaking abstract session at the meeting.
The double-blind, phase 3 TULIP trials each enrolled seropositive SLE patients with moderate to severe active disease despite standard-of-care therapy (SOC). All patients met ACR criteria, had a SLE Disease Activity Index (SLEDAI)-2K of 6 or greater, and BILAG index scoring showing one or more organ systems with grade A involvement or two or more with grade B. Both trials required stable SOC therapy throughout the study except for mandatory attempts at oral corticosteroid (OCS) tapering for patients who were receiving 10 mg/day or more of prednisone or its equivalent at study entry.
The trials followed a phase 2 trial, reported by Dr. Furie at the 2015 ACR meeting and published in Arthritis & Rheumatology in 2017, which showed “very robust” efficacy of anifrolumab in this setting.
“The burning question for the last 20 years has been, ‘Can type 1 interferon inhibitors actually reduce lupus clinical activity?’ ” Dr. Furie said. “The problem here [is that] you can inhibit interferon-alpha, but there are four other subtypes capable of binding to the interferon receptor.”
Anifrolumab, which was first studied in scleroderma, inhibits the interferon (IFN) receptor, thereby providing broader inhibition than strategies that specifically target interferon-alpha, he explained.
In the phase 2 trial, the primary composite endpoint of SRI response at day 169 and sustained reduction of OCS dose between days 85 and 169 was met by 51.5% of patients receiving 300 mg of anifrolumab versus 26.6% of those receiving placebo.
TULIP-1
The TULIP-1 trial, however, failed to show a significant difference in the primary endpoint of week 52 SRI, although initial analyses showed some numeric benefit with respect to BICLA, OCS dose reductions, and other organ-specific endpoints.
The percentage of SRI responders at week 52 in the double-blind trial was 36.2% in 180 patients who received 300 mg anifrolumab vs. 40.4% in 184 who received placebo (nominal P value = .41), and in a subgroup of patients who had high IFN gene signature (IFNGS) test results, the rates were 35.9% and 39.3% (nominal P value = 0.55), respectively.
Sustained OCS reduction to 7.5 mg/day or less occurred in 41% of anifrolumab and 32.1% of placebo group patients, and a 50% or greater reduction in Cutaneous Lupus Erythematosus Disease Activity Severity Index (CLASI) activity from baseline to week 12 occurred in 41.9% and 24.9%, respectively. The annualized flare rate to week 52 was 0.72 for anifrolumab and 0.60 for placebo.
BICLA response at week 52 was 37.1% with anifrolumab versus 27% with placebo, and a 50% or greater reduction in active joints from baseline to week 52 occurred in 47% versus 32.5% of patients in the groups, respectively.
The 150-mg dose, which was included to provide dose-response data, did not show efficacy in secondary outcomes.
“We see a delta of about 10 percentage points [for BICLA], and about a 15-percentage point change [in swollen and tender joint count] in favor of anifrolumab,” Dr. Furie said. “So why the big difference between phase 2 results and phase 3 results? Well, that led to a year-long interrogation of all the data ... [which revealed that] about 8% of patients were misclassified as nonresponders for [NSAID] use.”
The medication rules in the study automatically required any patient who used a restricted drug, including NSAIDs, to be classified as a nonresponder. That means a patient who took an NSAID for a headache at the beginning of the study, for example, would have been considered a nonresponder regardless of their outcome, he explained.
“This led to a review of all the restricted medication classification rules, and after unblinding, a meeting was convened with SLE experts and the sponsors to actually revise the medication rules just to make them clinically more appropriate. The key analyses were repeated post hoc,” he said.
The difference between the treatment and placebo groups in terms of the week 52 SRI didn’t change much in the post hoc analysis (46.9% vs. 43% of treatment and placebo patients, respectively, met the endpoint). Similarly, SRI rates in the IFNGS test–high subgroup were 48.2% and 41.8%, respectively.
However, more pronounced “shifts to the right,” indicating larger differences favoring anifrolumab over placebo, were seen for OCS dose reduction (48.8% vs. 32.1%), CLASI response (43.6% vs. 24.9%), and BICLA response (48.1% vs. 29.8%).
“For BICLA response, we see a fairly significant change ... with what appears to be a clinically significant delta (about 16 percentage points), and as far as the change in active joints, also very significant in my eyes,” he said.
Also of note, the time to BICLA response sustained to week 52 was improved with anifrolumab (hazard ratio, 1.93), and CLASI response differences emerged early, at about 12 weeks, he said.
The type 1 IFNGS was reduced by a median of 88% to 90% in the anifrolumab groups vs. with placebo, and modest changes in serologies were also noted.
Serious adverse events occurred in 13.9% and 10.8% of patients in the anifrolumab 300- and 150-mg arms, compared with 16.3% in the placebo arm. Herpes zoster was more common in the anifrolumab groups (5.6% for 300 mg and 5.4% for 150 mg vs. 1.6% for placebo).
“But other than that, no major standouts as far as the safety profile,” Dr. Furie said.
The findings, particularly after the medication rules were amended, suggest efficacy of anifrolumab for corticosteroid reductions, skin activity, BICLA, and joint scores, he said, noting that corticosteroid dose reductions are very important for patients, and that BICLA is “actually a very rigorous composite.”
Importantly, the findings also underscore the importance and impact of medication rules, and the critical role that endpoint selection plays in SLE trials.
“We’ve been seeing discordance lately between the SRI and BICLA ... so [there is] still a lot to learn,” he said. “And I think it’s important in evaluating the drug effect to look at the totality of the data.”