Randomized, Double Blind, Placebo-Controlled Trial of a TNF Inhibitor (etanercept) for the Treatment of Idiopathic Pneumonia Syndrome (IPS) After Allogeneic Stem Cell Transplant (SCT). A Blood and Marrow Transplant Clinical Trials Network (BMT CTN) Study

IPS is a frequently fatal form of non-infectious pneumonia occurring early after allogeneic SCT. We conducted a phase III trial to investigate if a soluble TNF binding protein, etanercept (Enbrel^® Amgen), when given with corticosteroids, improved management of IPS.  The study was supported by the NHLBI and NCI, with etanercept supplied by Amgen. Eligible patients were ≥18 yrs, within 180 days post allo-SCT, with diffuse pulmonary infiltrates, and absence of active infection, sepsis syndrome, or cardiogenic shock. A negative BAL was required at study entry. Patients were randomized to receive etanercept (0.4 mg/kg/dose 2x/week x 4 weeks) or placebo, in addition to corticosteroids (2.0 mg/kg/day, with taper allowed after 7 days).  Response was defined as survival with discontinuation of supplemental oxygen for >72 hours by day 28 of study. The trial was initially designed to enroll 120 patients, with sufficient power to identify a 25% response difference (30% to 55%) between study arms. Due to suboptimal accrual, inclusion criteria were modified and expected enrollment decreased to 60 patients, 30 per arm. The trial did not meet accrual goals and was stopped after enrolling 34 patients. RESULTS: Between 2007 and 2011, 34 patients (median age 46.6y, range 22-70 y) were randomized to receive etanercept (n=16) or placebo (n=18). Study arms were balanced for patient demographics, except for a higher incidence of AML in the placebo arm (50% vs 6%, p=0.02).  50% of etanercept and 55% of placebo treated patients received all intended study doses, whereas 38% of etanercept and 6% of placebo treated patients received ≤25% of intended dosing (primarily investigator discretion).  Response at day 28 and day 56, and overall survival were comparable between groups (TABLE). There were no differences in time to discontinuation of supplemental oxygen, incidence of grades 3-5 adverse events, infections, cumulative incidence of relapse, or grades 2-4 or 3-4 acute GVHD. Patients receiving ≤40% FiO2 at study entry had response rates of 91% (placebo) and 73% (etanercept) respectively.  By comparison, responses in patients receiving >40% FiO2 at entry were 17% (1 of 6) on the placebo arm and 50% (2 of 4) on the etanercept arm. CONCLUSION:  No improvement in response or survival was noted when etanercept was added to corticosteroids for the treatment of IPS, though the small sample size significantly limits the statistical power of this study. Compared to historical reports, the response to corticosteroids alone was higher than anticipated supporting the importance of studying controls when evaluating a new therapy.

	Etanercept arm	Placebo arm	p
Day 28 Response	62.5% (95%CI: 35.4-84.8)	66.7% (95%CI: 41.0-86.7)	0.80
Day 56 Response	56.3% (95%CI: 29.9-80.3)	50.0% (95%CI: 26.0-74.0)	0.72
6 month OS	50.0% (95%CI: 24.5-71.0)	33.3% (95%CI: 13.6-54.5)
Median survival	171 days (95%CI: 11-362)	64 days (95%CI: 26-209)
Log-rank  (OS)			0.51