Lead-in Dose Optimization to Mitigate Cytokine Release Syndrome in AML and MDS Patients Treated with Flotetuzumab, a CD123 x CD3 Dart® Molecule for T-Cell Redirected Therapy

Background:

Flotetuzumab (MGD006/S80880) is a novel CD123 x CD3 bispecific DART^® protein being tested in a Phase 1 study of patients with relapsed/refractory acute myeloid leukemia (AML) and myelodysplastic syndrome. Cytokine secretion with ensuing potential for cytokine release syndrome (CRS) is inherent in T-cell activation and is observed with T-cell redirecting therapies. Two lead-in dose (LID) strategies, in conjunction with early intervention with tocilizumab, were assessed for their ability to mitigate CRS.

Methods:

In LID-1 schema, flotetuzumab was administered at 100ng/kg/day(d) for 4d followed by 3d pause and resumption at 300 or 500ng/kg/d starting on Day 8. The LID-2 schema included 30ng/kg/d for 3d, 100ng/kg/d for 4d and 300 or 500ng/kg/d on Day 8. IL-2, IL-6, IL-8, IL-10, TNF-a, INF-ɣ, and GM-CSF were measured and CRS severity graded. Peak cytokine values during first reported CRS events, occurring within 10 days of start of first dose, were evaluated. Median peak cytokine levels were compared between patients with and without LID. Other potential CRS determinants were evaluated.

Results:

Infusion-related reactions (IRR)/CRS occurred in 34/45 (76%) of patients and were generally manageable and reversible, with most events (28/34; 82%) Grade (Gr) 1 and 2. Among 29 patients with complete cytokine data, 68% experienced CRS within 2 days of start of flotetuzumab, and an additional 8% within 10 days (14% Gr 1, 55% Gr2, and 7% Gr 3). Cytokine levels were generally higher in patients with CRS than in pts without (median IL-6, 116.2 vs. 67.9 pg/mL; IL-8, 191.1, vs. 144.6 pg/mL; IL-10, 867.6, vs. 348.7 pg/mL), and were generally higher with increasing CRS grade. LID reduced overall cytokine levels. Institution of a two-step LID in Week 1 decreased severity by mean 0.54 grade during cycle 1 (mean CRS grade week 1: 1.16 vs. 2; week 2: 1 vs. 1.33; week 3: 0.67 vs. 0.83; week 4: 0.13 vs 0.67 LID-2 vs. LID-1, respectively). Median peak cytokine levels observed with LID were lower during Week 1 and after achieving the maximum dose. Preliminary data show a relationship between baseline circulating T-cell number and maximum CRS grade during Week 1, with higher grade of CRS (≥2) in Week 1 associated with higher baseline levels of circulating T-cells. Other variables evaluated (including tumor burden) did not trend with CRS grade. Anti-leukemic activity with flotetuzumab has been reported (ESMO 2017); CRS grade and frequency did not correlate with response.

Conclusions:

CRS is commonly observed with T-cell-directing therapies. Two-step LID showed effectiveness in limiting IRR/CRS events and circulating cytokines over a single-step LID. Ability to predict which patients may develop CRS would be helpful when administering T-cell-activating therapies. Flotetuzumab data suggest that a patient's baseline circulating T-cell number may be a potential predictor of CRS severity. CRS severity did not correlate with anti-leukemic activity. clinicaltrials.gov NCT02152956