Phase II Trial of Natalizumab (Tysabri®) with Corticosteroids As Initial Treatment of Gastrointestinal Acute Graft Versus Host Disease

Background: Graft-vs.-host disease (GVHD) is caused by allorecognition of host antigens by tissue-specific donor lymphocytes. We have previously demonstrated that α4ß7 integrin is upregulated on both naïve and memory T cell subsets in patients who subsequently developed intestinal acute GVHD (Chen, BMT 2013). This integrin interacts with endothelial cell adhesion molecules and mediates homing of lymphocytes to the gastrointestinal (GI) tract. Natalizumab (Tysabri®, Biogen Inc.) is a selective adhesion molecule inhibitor, acting directly against the α4 subunit of α4 integrins. Given the poor outcomes of GI GVHD when treated with corticosteroids alone, we initiated a phase II study to determine the safety and efficacy of natalizumab in combination with corticosteroids for the initial treatment of acute GI GVHD.

Methods: Subjects with new-onset, acute GVHD of the lower GI tract were eligible to participate. Other organ involvement was permitted but not required. Biopsy confirmation of GI GVHD and exclusion of infectious causes of diarrhea were required. Natalizumab was given at a dose of 300 mg IV, with a second dose permitted 4 weeks later in the event of an incomplete response. An initial corticosteroid dose of 2 mg/kg of methylprednisolone was suggested, but dosing and dose reductions were at the discretion of the treating physician. Steroids could be initiated up to 7 days prior to the initiation of natalizumab while biopsy and infectious disease specimens were being analyzed. The primary endpoint was GVHD-free survival rate 56 days from enrollment. Secondary endpoints included response rates at day 28 and 56, steroid dose at days 28, 56, 100, 180 and 365, incidence of systemic infections at day 56 and 180, and overall survival.

Results: 18 subjects were enrolled and received at least one dose of natalizumab. Subjects had a median age of 63.5 years, and 12 were male. Fifteen had a non-myeloablative transplant, 12 had a matched unrelated donor, and all had received PBSCs. All subjects had newly diagnosed acute GI GVHD at time of enrollment; 4 had concomitant cutaneous and 1 had concomitant hepatic involvement. Stage 1, 2, 3, and 4 GI involvement was noted in 7, 4, 4, and 3 subjects respectively, corresponding to overall grade II GVHD in 7 and grade III in 11 subjects. Sixteen patients had data for response assessment. The day 56 GVHD-free survival rate was 37.5%. The overall response rate (PR+CR) for GI GVHD at day 28 and day 56 was 75% and 62.5% respectively. All surviving subjects except one achieved steroid dose reduction by day 56, with a median reduction of 61.6%. Three patients experienced treatment-related toxicity possibly related to natalizumab; one subject had grade 2 hepatic toxicity, one had grade 4 encephalopathy, and one developed hepatic failure. Three subjects have developed chronic GVHD. Four subjects had documented infections with bacteria (2), yeast (1), and viral pathogens (1- polyomavirus). Six subjects had measurable JC viremia prior to initiation of natalizumab and 8 subjects had measurable JC viremia at day 56, although only 1 subject had a viral load greater than 1000 copies/ml. No subject has developed JC-associated disease. Seven subjects have died. Causes of death include GVHD (2), disease relapse (2), and organ failure (3). With a median follow-up of 180 days, the 6 month overall survival was 52% (95% CI 25-74%).

Conclusion: Natalizumab in combination with corticosteroids as initial treatment of acute gastrointestinal GVHD is feasible and well tolerated. These early results suggest a higher rate of response at day 56 than is typically observed with steroids alone. Inhibition of T-cell homing to the GI tract may improve GI GVHD response and potentially mitigate progression of acute GVHD.