New Multi-Omics Framework Aims to Transform Cancer Treatment Decisions

Precision oncology has long been driven by the identification of individual genetic mutations, yet translating vast genomic data into actionable clinical insights remains a persistent challenge. A growing body of research now suggests that this mutation-centric approach may be inherently limited, as it overlooks the complex, system-level biology that governs tumour behaviour.

In a significant development, Akrivia Biosciences has introduced a novel framework that redefines how cancer genomes are interpreted. By analysing multi-omics data across thousands of breast cancer samples, the company has demonstrated that tumour mutations are not random or isolated, but organised into distinct biological programmes that influence disease progression, therapeutic response, and resistance.

This discovery underpins T-OMICS, a next-generation decision-support framework designed to address one of the biggest bottlenecks in oncology today: interpretation. Moving beyond static mutation lists, T-OMICS seeks to provide a comprehensive, system-level view of tumour biology by integrating genomic, transcriptomic, and microenvironmental data into clinically meaningful insights.

In this interview with MedTech Spectrum, Dr Amit Gupta, Founder of Akrivia Biosciences, discusses the scientific breakthroughs behind decoding cancer genomic heterogeneity, the limitations of current precision oncology approaches, and how T-OMICS could reshape clinical decision-making, drug development, and patient stratification particularly in complex conditions such as metastatic ER+/HER2– breast cancer.

Your research challenges the long-held view of cancer mutations as isolated events. What were the key scientific insights that enabled you to uncover this functional organisation of tumour genomes?

That is a very important question because it strikes at the core of what we believe has constrained precision oncology for years.

Most current approaches treat cancer mutations as isolated biomarkers. But cancer does not operate as a list of independent events — it behaves as an integrated biological system. That was the key insight behind our work.
For years, tumour DNA sequencing has generated increasingly detailed catalogues of alterations, but interpretation has remained limited because these alterations were still being read one by one. We approached the problem differently – instead of asking what each mutation means in isolation, we asked whether frequent genomic alterations consistently group into broader biological patterns that reflect how tumours actually function.

By integrating DNA, RNA, and proteomic data across more than 5,000 breast tumours, we found that cancer genomic complexity is not random, but organised into distinct biological programmes that shape tumour behaviour, prognosis, and therapeutic response. This marked the conceptual turning point: the tumour genome is not just a collection of separate mutations, but part of a structured functional architecture.

This also helped explain one of the field’s longstanding puzzles: why major mutations in breast cancer, such as PIK3CA, can appear biologically inconsistent from tumour to tumour. Our findings show that its meaning depends on the wider biological context in which it occurs. In other words, once tumour genomes are viewed at the level of biological state rather than isolated mutations, their functional organisation becomes visible.

This insight ultimately led to T-OMICS™, our framework designed to translate this functional structure into a clinically interpretable tumour profile.

How does the T-OMICS framework differ from existing multi-omics and precision oncology approaches currently used in clinical practice?

That is also a very important question, because the key issue in oncology today is no longer just data generation — it is interpretation. Existing approaches have become very good at scaling data generation. T-OMICS is designed to solve the interpretation bottleneck.

Most current workflows still return long lists of mutations and interpret them as isolated biomarkers. Even many so-called multi-omics approaches sequence both DNA and RNA, but use the RNA mainly qualitatively to detect events such as gene fusions or splice variants. T-OMICS uses RNA very differently: it relies on quantitative gene expression measurements, together with a set of critical DNA alterations, to assign a tumour to a biologically meaningful programme.

As a result, T-OMICS does not just list abnormalities. It identifies the tumour programme, measures its activity state, captures the immune and microenvironmental context, and reveals the driver and buffering systems that influence response, resistance, and rational combination strategies. Built on our discovery that decodes cancer genomic heterogeneity, it is designed to resolve ambiguity and guide precision oncology more confidently.

The key distinction is not simply that T-OMICS uses both DNA and RNA. It is how that data is used through proprietary RNA and DNA signatures developed from our underlying discovery. Sequencing both DNA and RNA does not, by itself, constitute biologically interpretable multi-omics. T-OMICS uses a five-tier framework, built on foundational genomic discovery, to define the functional state of the tumour.

You highlight the context-dependent role of PIK3CA mutations. How could this reshape ongoing and future drug development strategies targeting the PI3K pathway?

You have asked a very important question, because PIK3CA is one of the most common mutations in breast cancer, a major drug target, and yet still biologically ambiguous. Our findings show that PIK3CA-mutant disease is not a single biological or therapeutic entity. Most often, these mutations occur in a biologically buffered context associated with more favourable outcomes, but in more aggressive tumours shaped by other key alterations, they can signal higher-risk disease with different treatment needs.

The implication is that PI3K-pathway development may need to move beyond mutation-defined enrolment and toward programme- and state-aware stratification. That could help distinguish truly PI3K-dependent tumours from those that are merely PI3K-marked, enable more rational combination strategies, and reduce the biological noise that may have diluted signals in earlier trials. It also suggests that some prior studies may benefit from retrospective re-analysis to identify hidden responder subsets and better understand resistance.

What are the biggest challenges in translating T-OMICS from a research framework into a clinically deployable decision-support tool for oncologists?

The main challenge is translating a biologically rich framework into something clinicians can quickly trust and use. That requires rigorous validation, clear clinical positioning, and highly intuitive outputs. We are focused on solving exactly that interface between complex biology and real-world decision-making.

We have already launched T-OMICS in research-use-only form as a decision-support tool built on proprietary RNA and DNA signatures, for molecular tumour boards, oncologists, pharma, and translational researchers working with available NGS and/or transcriptomic data. In its current form, T-OMICS does not provide treatment recommendations or replace clinician judgment.

At a time when cancer burden is rising rapidly in India and globally, we believe the field needs a programme- and state-level framework like T-OMICS that can provide 360-degree biological clarity. In that sense, the remaining challenges are less about whether such a system is needed and more about ensuring that sufficient support, investment, and translational infrastructure are in place to bring it into routine clinical use as quickly and responsibly as possible.

How do you see T-OMICS integrating into molecular tumour boards (MTBs) and influencing real-world treatment decisions for metastatic ER+/HER2– breast cancer patients?

That is a very important practical question, because the value of a framework like this ultimately depends on whether it helps clinicians better understand real patients.

We see T-OMICS integrating into molecular tumour boards as an interpretation layer rather than another data-generation tool. In metastatic ER+/HER2– breast cancer, clinicians often already have NGS panel reports and, in some cases, transcriptomic data. The bottleneck is interpretation: understanding what those findings mean biologically, especially when current reports often return long lists of mutations without explaining the tumour’s dominant programme or state.

T-OMICS is designed to address that gap by converting available molecular data into a structured and clinically interpretable profile of the tumour’s programme identity, activity state, immune context, and key modifier systems. This can help molecular tumour boards move beyond reviewing isolated biomarkers and toward understanding the biological architecture that is actually shaping response, resistance, and disease behaviour.

This is especially relevant in metastatic disease, where repeated biopsies are often interpreted independently, making it difficult to distinguish true biological evolution from sampling variability. Our work shows that T-OMICS can separate truncal tumour identity from lesion-level state shifts, making longitudinal interpretation more meaningful and helping clinicians understand whether a new biopsy reflects a fundamentally new biology or a state change within the same underlying tumour system.

In that way, T-OMICS could support more biologically informed discussions around treatment sequencing, resistance interpretation, rational combinations, and longitudinal sampling strategy. In its current research-use-only form, it does not provide treatment recommendations or replace clinician judgment. Its role is to help molecular tumour boards and oncologists interpret tumour biology more clearly, so that real-world decisions can be made with a deeper, more mechanistic understanding of the disease.

Beyond breast cancer, what is your roadmap for expanding this framework to other cancer types, and what timelines can we expect for broader clinical adoption?

Our broader vision is to extend this framework well beyond breast cancer, because the central problem it addresses — how to decode cancer genomic heterogeneity in a biologically and clinically meaningful way — is not unique to one tumour type. We believe the same core principle applies across oncology: cancer should not be interpreted as a list of isolated alterations, but as an integrated biological system organised into programmes and states. The platform is built on proprietary, patent-backed RNA and DNA signatures that we believe can be extended across multiple tumour types.

Breast cancer was the right place to start because of the scale of data and the biological depth available. From here, expansion will be guided by clinical need, data maturity, and translational readiness in other cancers.

Regarding timelines, the T-OMICS platform is currently accessible through a research-use-only platform for MTBs, oncologists, pharma, and translational researchers working with genomic and transcriptomic data. However, wider clinical adoption will depend on prospective validation and the necessary infrastructure for responsible implementation. The next phase focuses on translating the concept into broader clinical use in a reliable and scalable manner.

 What is clear to us is that the need is urgent. As cancer burden rises in India and globally, the field needs more comprehensive systems that can move beyond mutation lists and provide 360-degree mechanistic clarity. We believe T-OMICS is the start of that shift, and our roadmap is to expand that capability across cancer types in a deliberate, evidence-led way.