Bildnachweis: (c) Promatix.
Promatix Biosciences presented new data at American Association for Cancer Research (AACR) Annual Meeting 2026, addressing a key bottleneck in the development of antibody–drug conjugates (ADCs): the scarcity of truly tumour-selective targets. The UK-based biotech company demonstrates how its integrated proteomics-driven discovery platform systematically identifies complementary antigen pairs, significantly expanding the target landscape for next-generation bispecific ADCs.
At the core of the approach is the combination of large-scale proteomic profiling (TxPro) and computational modelling (CipherPro). This enables a shift from a limited set of monospecific targets to thousands of potential bispecific combinations. By applying so-called logic gates – particularly AND and OR configurations – the platform identifies antigen pairs that promise improved tumour selectivity and, consequently, a more favourable therapeutic index. The data were presented in a poster entitled “Overcoming the Limited Monospecific Target Landscape in Cancer via Discovery of Tumor-Selective AND- and OR-Gate Bispecifics”. It outlines how hybrid avidity using an AND-gate design – where both antigens are required for binding – can increase selectivity, enhance efficacy and reduce toxicity. Despite advances in payloads, linkers and conjugation technologies, target selection remains a key limiting factor for ADC performance. As a result, interest in bispecific approaches is growing, as they aim to enable more precise tumour targeting. “Across the ADC field, target selection is increasingly recognized as the critical bottleneck for achieving meaningful improvements in efficacy and safety,” said Dr. Michael Hunter, CEO and Co-Founder of Promatix. “Our integrated platform is designed to overcome this limitation through a systematic, data-driven approach to identifying complementary antigen pairs. It greatly expands the universe of viable tumour-selective targets beyond what conventional monospecific approaches can achieve and provides a scalable foundation for advancing next-generation ADC performance. In ADC development, we believe success will be determined by better targets, which will ultimately lead to better drugs.”
Monospecific targets remain scarce
Analysis of the company’s oncology-focused proteomics database, TxPro, highlights the structural limitations of conventional ADC approaches. Of 2,768 surface proteins analysed, only 575 showed high tumour expression, and ultimately just 24 demonstrated strong differential expression between tumour and normal tissue. Notably, targets currently used in approved ADCs for solid tumours do not exhibit strong differential expression profiles, underlining the limited availability of highly tumour-specific antigens and the associated safety challenges. By contrast, TxPro identifies several highly selective targets with emerging clinical relevance, including CLDN6, for which encouraging early clinical activity has been reported, for example in ovarian cancer.
Limited innovation in the current bispecific landscape
The current bispecific landscape continues to be dominated by a small number of established antigens, with EGFR and MET featuring prominently across many programmes. Overall, target novelty remains low: only 15.3% of constructs include at least one novel target, and just 2.5% incorporate two novel targets. Among explored programmes, EGFR × MET shows the most robust complementary expression and serves as a reference benchmark.
Scalable expansion of the target landscape
Using the CipherPro modelling framework, Promatix evaluated 165,025 possible bispecific combinations derived from the 575 tumour-expressed surface proteins identified in TxPro. Applying a differential expression threshold of greater than tenfold, the company identified 2,164 potentially tumour-selective bispecific combinations, compared with only 24 monospecific targets. Even under more stringent filters for expression and normal tissue, the number of promising candidates remains substantial. While AND-gated designs deliver the strongest predicted gains in tumour selectivity, OR-gated approaches support broader tumour coverage, particularly in heterogeneous disease settings.
Experimental validation confirms predictive power
As a case in point, Promatix highlights the EGFR × EphA2 combination, identified as a compelling AND-gated candidate based on complementary expression patterns. Experimental validation – including immunofluorescence staining of patient biopsy samples and flow cytometry in patient-derived xenograft models – confirms strong tumour co-expression with minimal overlap in normal tissue. The resulting lead programme, PBS293, is currently being developed for metastatic colorectal cancer and is designed to address a broad patient population – irrespective of RAS/BRAF mutation status and including right-sided tumours, where existing therapies such as EGFR-targeting antibodies (e.g. cetuximab) have limited efficacy. “As the ADC field continues to evolve, expanding the universe of tumour-selective targets will be essential to sustaining innovation,” said Dr. Roy Pettipher, Chief Scientific Officer of Promatix. “Our data demonstrate that logic-gated bispecific discovery provides a viable and scalable path toward safer and more effective therapies across many cancer indications. The platform doesn’t just predict new bispecific targets; it identifies combinations that work biologically and can be validated experimentally. Importantly, this approach has already been used to identify a promising ADC candidate, our lead programme, PBS293, targeting EGFR and EphA2 for the treatment of colorectal cancer, demonstrating how predictive discovery can translate into actionable therapeutic programmes.”
