Expanding the Reach of Bispecifics

A trio of preclinical proof-of-concept studies is paving the way for bispecific antibodies to be deployed against mutations in two of cancer's most notorious genes, TP53 and RAS. The strategy, which may also prove effective against T-cell malignancies, was spearheaded at Johns Hopkins Kimmel Cancer Center in Baltimore, MD, by Ludwig Center codirectors Bert Vogelstein, MD, and Kenneth Kinzler, PhD.

“Genetic alterations are the most distinct marker discriminating between cancer and normal cells, but targeting them is not so easy,” explains Shibin Zhou, MD, PhD, a senior author on all three papers. Proteins from these mutant cancer driver genes, being largely intracellular, are inaccessible to conventional antibody-based treatments. “So, we decided to take advantage of nature,” Zhou says—specifically, routine protein turnover and processing into small peptides that HLA molecules can present on the surface as mutation-associated neoantigens (MANA).

The researchers screened for MANA-directed antibody fragments to develop therapeutically, but they encountered another problem: Most MANAs exist at levels too low to be readily recognized. As such, they turned to a bispecific antibody design wherein one end tethers CD3 on T cells and the other binds to a given MANA on tumor cells—which brings both into close proximity and unleashes cytotoxicity. Currently, the only FDA-approved T cell–engaging bispecific is Amgen's blinatumomab (Blincyto), which uses a proprietary BiTE format to target CD19 in blood cancers. After evaluating several different bispecific formats, Vogelstein and colleagues decided to go with a single-chain diabody (scDb) configuration.

Next, the team assessed their approach on TP53's most common mutation, R175H—presented as a MANA by HLA-A*02:01, an allele found in more than 40% of the U.S. population (Science 2021;371:eabc8697). A candidate antibody fragment, H2, was generated into a bispecific, dubbed H2-scDb. It showed efficacy against human multiple myeloma cells expressing this MANA in single-digit quantities. In mouse xenografts, H2-scDb induced major tumor regression and was well tolerated.

Structural biologist Sandra Gabelli, PhD, used X-ray crystallography to analyze the H2 fragment bound to the TP53 MANA. “We found that H2 forms a ‘cage’ around the mutant histidine and its adjacent amino acid,” Gabelli says. “This structure imparts stability, providing strong selectivity for the R175H, but not for the wild-type, peptide.”

The researchers also highlighted their bispecifics targeting two recurrent RAS mutations, G12V and Q61/H/L/R, respectively presented by HLA-A3 and HLA-A1 (Science Immunol 2021;6:eabd5515). Both agents provoked T cell–driven lysis of lung adenocarcinoma cells in vitro. There was a trend toward slowed tumor growth in mice, although these results were modest and “suboptimal,” the researchers wrote. But, as Zhou cautions, “it's hard to compare in vivo data for our RAS and TP53 bispecifics—the MANA affinity could be different, or perhaps we didn't test the ‘right’ mouse models. There are a lot of unknowns to be further investigated.”

Meanwhile, Suman Paul, MBBS, PhD, investigated whether scDb bispecifics might also have utility in T-cell cancers, which have considerably fewer treatment options than B-cell malignancies (Science Transl Med 2021;13:eabd3595). He focused on the T-cell receptor's β chain variable gene (TRBV) segments—these comprise 30 families, TRBV1 to TRBV30, that are “more or less randomly distributed across normal T cells,” Paul says. However, each T-cell cancer, being clonal, expresses only one TRBV, so “we figured a TRBV-targeting bispecific antibody could selectively deplete malignant T cells while largely sparing the healthy ones.”

Paul and the team evaluated two bispecifics targeted at TRBV5 and TRBV12. Both proved cytotoxic in mouse models of human T-cell malignancies expressing the relevant TRBVs, inducing significant tumor regression that was associated with a survival benefit. Importantly, although the bispecifics “wiped out 15% to 20% of healthy T cells,” Paul observes, “clinically, we've learned from HIV/AIDS that this level of healthy T-cell depletion should be manageable for patients.”

Jon Weidanz, PhD, of The University of Texas at Arlington, says in a commentary that these studies “are promising for advancing scDbs into the clinic” (Science 2021;371;996–7). Noting that scDbs are small and rapidly cleared by the kidneys, he recommends exploring ways to extend the half-life of these bispecifics. Otherwise, patients would require continuous infusion to maintain therapeutic efficacy. Weidanz also suggests expanding the target search to MANAs presented by HLA-E and HLA-G, which have relatively uniform genotypes, whereas HLA-A's is highly variable. This may broaden the population of patients who could benefit.

“What we've developed is still a prototype; it needs to be optimized,” Zhou says. “The goal is to generate a suite of bispecifics targeting the top ten most frequent MANAs across all cancer types. Then, and because this is off-the-shelf, we should be able to treat not just small fractions of individuals with the relevant mutation and HLA allele, but many more patients per year in the U.S.” –Alissa Poh

For more news on cancer research, visit Cancer Discovery online at http://cancerdiscovery.aacrjournals.org/CDNews.