Beyond Genomics Alone: Why the Future of Early Detection Will Be Multi-Dimensional Biology

Beyond Genomics Alone

Why the Future of Early Detection Will Be Multi-Dimensional Biology

A Perspective from Breath Diagnostics, Inc.

As an investor managing a diversified portfolio, I have been repeatedly asked what the recent results from Grail’s NHS-Galleri trial mean for the future of biotechnology — and more specifically, for the field of early cancer detection.

The question is understandable. When a landmark, population-scale study reports that its primary endpoint was not met, it invites reflection, but it also invites deeper analysis.

The recently reported results from the NHS-Galleri trial represent a meaningful milestone in population-scale cancer screening. While the study did not meet its primary endpoint of demonstrating a statistically significant stage shift in identifying cancers, it confirmed something equally important: multi-cancer early detection (MCED) testing can be deployed at national scale and can identify cancers that would otherwise remain undetected.

This is not a failure of technology.

It is a logical consequence of biology.

Over the past decade, circulating free DNA (cfDNA)–based platforms have transformed oncology diagnostics. They have shown that tumor-associated genomic signals can be detected through a simple blood draw. That achievement is real, consequential, and foundational to precision medicine.

But early detection is not solely a sequencing problem.

It is a biological problem.

Large interventional trials are now defining the biological boundaries of DNA-based detection. Rather than diminishing genomics, these data illuminate where complementary biological signals may strengthen the field.

At Breath Diagnostics, we believe the next era of early detection will not be genomic alone.

The Biological Constraint of cfDNA: Shedding

cfDNA assays detect fragments of DNA in circulation, including tumor-derived DNA (ctDNA) released from cancer cells. These technologies have reshaped oncology by enabling:

• Minimally invasive tumor genotyping
• Minimal residual disease detection
• Therapy selection and monitoring
• Multi-cancer early detection approaches

However, detection of circulating tumor DNA is governed by fundamental biological parameters:

1. Tumor cell turnover (apoptosis and necrosis)
2. Tumor burden (absolute malignant cell mass)
3. Access of tumor-derived DNA to the bloodstream
4. The fraction of tumor DNA relative to background cfDNA

ctDNA can be released at all stages of tumor development. However, measurable levels in plasma generally increase with tumor size, cellular turnover, and vascular interface. As tumors grow and become more vascularized, the probability of detectable DNA fragments entering circulation rises.

In very early-stage tumors:

• Tumor volume is small
• Absolute tumor cell death is limited
• Tumor DNA fractions may be extremely low
• Variant allele frequencies may fall below assay detection thresholds

This is not a flaw in the technology, but rather a reflection of tumor biology.

cfDNA platforms detect what is present in circulation. If only minute quantities of tumor-derived DNA are shed, only minute quantities can be measured, irrespective of assay sophistication.

The sensitivity patterns observed in recent MCED trials are biologically coherent. They largely mirror expected differences in tumor burden and DNA shedding dynamics across stages.

The apparent ceiling in early-stage sensitivity reflects the timing and magnitude of tumor DNA release during tumor evolution — not a failure of genomic science.

A Complementary Biological Layer: Metabolism

If circulating tumor DNA reflects genomic material shed into blood, breath metabolomics reflects something different: the downstream biochemical consequences of disease biology.

Cancer is not only a genetic disease. It is also a metabolic disease.

Long before substantial tumor DNA accumulates in circulation, malignant transformation can alter:

• Cellular metabolism
• Oxidative stress pathways
• Lipid peroxidation
• Inflammatory signaling
• Host–tumor interactions

These processes generate volatile carbonyl compounds and related metabolites that diffuse into blood and are ultimately exhaled in breath.

Unlike ctDNA, which depends on sufficient tumor DNA shedding into plasma, volatile metabolites arise from active biochemical flux within tumor and host tissues. Detection therefore does not require identification of rare mutant fragments, but instead reflects dynamic metabolic activity.

This does not make one modality superior. They simply measure different dimensions of biology.

ctDNA captures genomic alterations, while breath metabolomics captures metabolic and oxidative activity.

Together, they describe different chapters of tumor evolution.

In early disease, when ctDNA fractions may be extremely low, metabolic perturbations and host responses may already be underway. In later disease, genomic burden and metabolic disruption intensify in parallel.

The opportunity lies not in replacing one with another, but in integrating them.

Molecular Scale Matters

The physical properties of cfDNA and volatile metabolites differ dramatically.

cfDNA

• ~150–200 base pair fragments
• ~100,000–200,000 Daltons
• Large macromolecular fragments
• Diluted in plasma
• Low permeability through cell membranes

Volatile Carbonyl Compounds

• <300 Daltons
• Often <200 Daltons
• Most are small, diffusible molecules
• Rapid systemic equilibration
• Most cross the alveolar membrane easily

These molecules differ by several orders of magnitude in size with one (DNA) being biologically designed to remain within intact cells.

Where cfDNA fragments represent structural debris from cellular breakdown, VOCs are byproducts of active biochemical processes.

The physics alone create distinct biological windows.

Analytical Evolution: Breath Is Different Today

Historically, the field of breath diagnostics faced legitimate technical challenges, including sorbent variability, thermal desorption degradation, contamination, and semi-quantitative fingerprinting.

Our microreactor platform was engineered to overcome those limitations.

Using a patented microreactor, volatile carbonyl metabolites are captured from a single natural exhale. Surface-bound aminooxy chemistry forms stable covalent adducts with carbonyl compounds at the moment of interaction with breath samples, effectively stabilizing otherwise labile metabolites. This reaction also introduces a permanent positive charge, enhancing downstream LC-MS sensitivity.

The stabilized analytes are quantified using UHPLC-MS/MS, enabling true molecular measurement rather than pattern recognition. For many captured carbonyl species, we achieve linear dynamic ranges spanning six orders of magnitude, with multiple orthogonal qualifier ions confirming molecular identity. This provides quantitative robustness and analytical specificity consistent with modern bioanalytical standards.

In other words, this is not a conventional “breath test” adapted for diagnostics — it is a molecular assay, built on analytical chemistry principles, that uses breath as its biological matrix.

Clinical Evidence

Across multiple peer-reviewed studies, we have demonstrated strong clinical signal in idiopathic pulmonary fibrosis (IPF), lung cancer, and COVID-19.

In lung cancer, prospective data demonstrated high sensitivity, including detection of early-stage disease, as well as normalization of carbonyl levels following tumor resection, supporting biological specificity and dynamic responsiveness.

Machine learning classification has demonstrated strong performance in larger datasets. When integrated with CT-derived imaging features, diagnostic accuracy and AUC further improved, suggesting that metabolic profiling may enhance organ-focused detection workflows.

Most recently, our platform received FDA Breakthrough Device Designation for pre-operative risk assessment of postoperative pneumonia in elective cardiac surgery patients. This designation reflects FDA recognition of the platform’s potential clinical impact and supports its broader translational trajectory.

The Strategic Implication

Large MCED trials have validated blood-based screening and clarified the biological window of DNA-based detection.

Rather than viewing recent results as setbacks, they define the boundaries of a powerful modality: that biology is not one-dimensional.

The future of early detection is unlikely to be single-axis.

It may be:

• Genomic
• Metabolic
• Imaging-integrated
• Clinically contextual

The question is not whether genomics works, but how to broaden the biological aperture of detection.

The Core Thesis

Liquid biopsy measures what tumors release when they fragment.

Metabolic breath analysis measures how disease alters biology while it is still functionally active.

One reads structure, while the other reveals activity.

The more biological layers we integrate, the more complete the narrative becomes — and the earlier and more confidently we may act.

The Pharma Opportunity: Biology in Motion

Beyond screening, metabolic profiling introduces an additional dimension: dynamic biological monitoring.

Because volatile carbonyl metabolites reflect ongoing oxidative stress, inflammatory signaling, and metabolic flux, they offer the potential to:

• Track biological response to therapy in near real time
• Provide non-invasive pharmacodynamic readouts
• Support mechanism-of-action validation
• Enable longitudinal monitoring without repeated invasive sampling

Where genomics often captures structural change, metabolomics can capture functional response.

For pharmaceutical partners, this represents an opportunity to observe biology in motion: not only whether a tumor harbors a mutation, but whether targeted intervention is altering the underlying metabolic state.

We believe the next step forward is not replacement, but combination.

If you are exploring:

• Multi-omic integration
• Complementary early detection strategies
• Dynamic pharmacodynamic monitoring
• Or real-time biological readouts in clinical development

We welcome the conversation.

Early detection — and therapeutic monitoring — will not be defined by a single signal.

It will be defined by dimensional biology.