IMPS-40ENHANCED VERSIONS OF THE ANTI-EGFRvIII VACCINE REVEAL A POTENTIAL NEW PARADIGM FOR ANTIGEN RECOGNITION AND VACCINE OPTIMIZATION

An anti-EGFRvIII peptide vaccine for treating glioblastoma (Rindopepimut) has shown significant clinical potential, recently receiving Breakthrough Therapy Designation from the FDA. Interestingly, we created this vaccine before MHC binding principles were understood so the sequence was arbitrarily chosen and never optimized for immune response. X-ray diffraction of the EGFRvIII peptide reveals that it has structure, unusual for a 13 amino acid peptide, and the novel glycine at the junction, thought to be essential, is not critical for immune recognition. However, the glycine is at a hinge so we hypothesized this position is important for secondary structure. In our EGFRvIII/NIH-3T3 syngeneic tumor model, we tested substitution of the glycine by all 19 other amino acids, length variations, synthetic amino acids and deletion of the glycine (-Gly) (N = 16 or greater for 24 peptides tested). Several substitutions, including Ala, Val, Pro, show a >58% improved survival rate over Rindopepimut. Most substitutions were similar to Rindopepimut, but some, including His, Cys and -Gly, were comparable to control. All peptides elicited antibodies specific for EGFRvIII with a general trend towards high titer (>1:2x105) in the higher performing vaccines. CTL responses by Elispot revealed that the highest performing vaccines elicited a >42% increase in spots over Rindopepimut (p 19 cleavage products, including recombination to form a larger products at ∼2600Da. Rindopepimut produced ∼14 fragments with lower amounts of the 2600Da, while the poorest vaccines showed almost no cleavage and no 2600Da. Our results show that more robust vaccines against EGFRvIII can be produced, that 3D structure is important, and the vaccine shows unexpected complex processing. Further work is being performed to bring these vaccines to clinical trial.