A novel platform that can identify tumour-specific alternative splicing (AS)–derived neoantigens with high confidence level could potentially overcome key stumbling blocks in cancer neoantigen vaccine development, according to Ms Bui Que Tran Nguyen of the Medical Genetics Institute, Ho Chi Minh City, Vietnam, who presented her team’s findings at ESMO TAT Asia 2026.
Nguyen’s team sought to identify a new source of shared neoantigens using AS, and developed the OncoSpliceNeo platform that enabled high-specificity identification of recurrent, immunogenic AS-derived neoantigens through tumour-normal transcriptomic comparisons and multilayer filtering, establishing a scalable foundation for off-the-shelf (OTS) immunotherapies, particularly in low tumour mutational burden cancers such as breast cancer (BC). [Nguyen BQT, et al, ESMO TAT Asia 2026, 183RO]
“Unlike conventional mutations, AS also generates many shared neoantigens, and identifying the spliced variants remain challenging because they are detectable in normal tissues as well. Therefore, a robust bio line is needed to identify tumour-specific splicing neoantigens,” explained Nguyen.
OncoSpliceNeo was thus developed. It was benchmarked using a published dataset of 88 BC patients from the Fudan University Shanghai Cancer Center (FUSCC) and applied to a clinically annotated cohort of 49 patients to construct an OTS neoantigen (OTS-neoTST) panel. The applicability of the OTS panel was prospectively validated in five patients using ELISpot assays with long peptides and patient-derived peripheral blood mononuclear cells, with responses further characterized by 10x single-cell RNA and T-cell receptor (TCR) sequencing.
OncoSpliceNeo outperformed existing pipelines, including RMATs and ScanExitron, identifying 106 neo-tumour specific target (TST)-human leukocyte antigen (HLA) pairs and achieving a high tumour specificity rate of 92.3 percent in the FUSCC cohort. In the clinical validation cohort, the OTS-neoTST panel achieved coverage of approximately 74 percent among the 49 Vietnamese BC patients. Functional validation showed that all five patients harboured at least one immunogenic AS-derived neoantigen from the OTS panel, with approximately 35.3 percent of predicted peptides eliciting measurable immune responses.
“Notably, these immunogenic OTS-neoTSTs could trigger CD8+ T-cell responses and promote the development of memory phenotypes, highlighting their potential as ready-to-use cancer vaccines. Nevertheless, further studies are needed to evaluate the safety and efficacy of this panel in BC patients,” said Nguyen.
“From the point of view as a BC medical oncologist, I think this is potentially a breakthrough and highly exciting,” commented discussant Dr Janice Tsang of the University of Hong Kong. “This could perhaps have a further role in prevention of ductal carcinoma in situ [DCIS] or prevention of malignant transformation in patients diagnosed with DCIS.”
“Although the read-out did not actually match any clinical outcomes, we still congratulate the team because the AS generated a rich source of neoantigens, and these are immunogenic,” Tsang added. “This is also timely, innovative, and potentially strategic, especially in this part of the world, in bridging equal access and unmet needs of accessibility [high cost and long waiting time] to personalized vaccines.”
Nevertheless, Tsang cautioned that broader validation across larger cohorts, diverse HLA backgrounds (eg, non-Chinese or non-Asian populations), and different BC subtypes will be required (whether subtype-specific AS neoantigen panels could improve coverage and immunogenicity).
“We also need additional evidence on quantitative cytotoxicity, clone expansion, and antitumour activity to strengthen the translational implications of the work,” noted Tsang. “Furthermore, as a proportion of AS neoantigens are expressed in adjacent normal tissues, additional validation using large-scale normal tissue resources would be needed to confirm safety and specificity of these targets.”