Preventing Primate Gvhd Using a Novel Antagonistic Anti-CD28 Antibody Plus Rapamycin: Downregulation of CD8 Proliferation Predicts Gvhd-Free Survival

Introduction:  We have previously shown, using a non-human primate (NHP) model, that inhibition of CD28:CD80/86 costimulation with CTLA4-Ig could prevent GVHD, and this strategy is now being tested in a Phase 2 clinical trial. Despite the potential efficacy of CTLA4-Ig, there is concern that it may have off-target effects, given its inhibition of both (+) CD28 signaling and (-) CTLA4 signaling.  In order to specifically target CD28, we have developed an antagonistic humanized anti-CD28 monovalent Fab' antibody (FR104). Here we investigate the efficacy and mechanism of action of FR104 in NHP, to potentially facilitate its most rapid clinical translation.

Methods:  NHP underwent MHC-mismatched HCT after myeloablative TBI.  They were transplanted with GCSF-mobilized PBSCs (4 ± 1x10⁸ TNC/kg and 2 ± 0.5x 10⁷ CD3+ T cells/kg). GVHD prophylaxis was with FR104 monotherapy (5mg/kg/wk IV, resulting in <1% CD28 expression) or with FR104 + mTOR inhibition with rapamycin, and were compared to two controls groups (untreated recipients and rapamycin monotherapy). Clinical and histologic GVHD was monitored and longitudinal immunologic analysis performed.

Results:  Untreated controls (n = 5) developed rapid, severe AGVHD (MST=7 d).  Rapamycin alone (n = 6) partially protected recipients, with GI-predominant AGVHD (MST = 14 d, Fig. 1). FR104 monotherapy showed statistically significant prolongation in survival (MST = 27 days, n = 3) compared to rapamycin (p = 0.05) and untreated controls (p=0.02), with breakthrough GVHD occurring in two animals (liver and skin). In contrast, FR104 + rapamycin  (n = 3) controlled GVHD, with all recipients reaching the timed terminal analysis at day +33-35 without clinical disease (p = 0.017 vs. untreated, p = 0.013 versus rapamycin alone).

Our previous work has documented a major role for T cell proliferation (by Ki-67 expression) in GVHD, with untreated controls demonstrating rampant CD8+ proliferation (89 ± 5% Ki-67+ CD8+ T cells at terminal analysis (d +6) vs. 4% pre-transplant, Fig. 2A).  Rapamycin and FR104 monotherapy both partially controlled proliferation (9% ± 4% and 14% ± 8% Ki-67+ CD8+ on d +6) with rapamycin + FR104 combination therapy resulting in further control (3.5% ± 0.3% Ki-67+ CD8+ on d +6). Moreover, the degree of CD8+ proliferation correlated closely with GVHD-free survival (Fig. 2B), suggesting that Ki-67 expression may be a predictive biomarker of this disease.

Implications:  These results show that specific blockade of CD28 can inhibit NHP GVHD and when combined with rapamycin, can effectively control this disease for the length of dual therapy.  Importantly, clinical control of GVHD with FR104 correlated with normalization of T cell proliferation.  Our results suggest that selective CD28 blockade may be a safe and effective adjunctive strategy to inhibit GVHD-associated T cell activation, and is deserving of clinical evaluation in patients undergoing HCT.