… And why they will continue to fail until we rethink how biology is analyzed.
The recent ASCENT-07 readout reverberated across the oncology community, not because Trodelvy is an ineffective drug, and certainly not because antibody–drug conjugates (ADCs) are a flawed modality. The failure reflects something far more fundamental: the industry still does not understand the mechanistic basis of ADC response and resistance.
And this is not a Gilead problem, or any one team’s problem. It is a systemic failure rooted in outdated analytical thinking and an entrenched reliance on descriptive methods that were never built to resolve complex therapeutic behavior in heterogeneous cancers.
ASCENT-07 is simply the latest demonstration that the standard analytical playbook (bulk RNA, limited gene panels, classical enrichments, mutation-centric interpretations, and black-box ML models) cannot explain why some patients respond, why others do not, and how that heterogeneity evolves under treatment pressure. These tools were designed for correlation, not causation. They identify patterns but cannot interpret mechanisms. And every time the industry attempts to stretch them beyond their limits, we see the same pattern: early promise, diminishing returns, an expansion miss, and internal confusion over “what happened?”
The ADC Development Loop
Early Signal → Plateau → Expansion Miss → Confusion
This loop is now predictable across companies. A compelling early signal is followed by diminishing returns, then an expansion failure, and finally a scramble to interpret superficial patterns that never answered the biological question in the first place.
The tragedy is not the failure itself. It is that the underlying mechanisms were likely solvable, had the right analytical system been applied. The limiting factor was not the drug; it was the analytical frameworks used to interrogate the data.
Despite unprecedented data availability, the field remains anchored to pipelines that cannot resolve the mechanistic interplay determining ADC response: tumor microenvironmental suppression, lineage plasticity, treatment-induced transcriptional adaptation, intracellular trafficking failures, pathway rewiring, and stress-response dynamics that bulk workflows cannot detect.
The Dangerous Assumption Underpinning ADC Analytics
A pervasive (and increasingly dangerous) assumption persists across pharma R&D:
“If we run the same standard pipelines harder, deeper, or at larger scale, the biology will eventually reveal itself”.
ASCENT-07 is definitive proof that this belief is false.
These tools cannot resolve:
- microenvironmental resistance circuits
- therapeutic stress adaptation
- intracellular processing defects
- lineage state transitions
- pathway rewiring
- or multi-omic escape mechanisms
They were never designed for this level of biological complexity.
The Structural Problem: Siloed Tools for a Non-Siloed Disease
Another fundamental issue is structural: most organizational workflows remain siloed.
- Genomics done here
- Transcriptomics there
- Clinical trajectories elsewhere
- Spatial and proteomic layers missing
- No cross-modal coherence
But diseases do not behave in silos, and tools that do cannot uncover the biology needed to guide ADC strategies.
Equally problematic is the near-total absence of reproducibility frameworks. Many signatures that appear promising collapse under external validation, platform changes, or cohort shifts. Without a reproducibility engine, findings (no matter how elegant) cannot be trusted.
The Core Analytical Failure: Inability to “Peel the Onion”
Understanding ADC biology requires structured, sequential, multi-layered interrogation:
- broad pathway signals
- ontology-level refinement
- network-level resolution
- reproducibility simulation
- isolation of true causal drivers
Yet most pipelines stop at the very first layer. They surface associations that are interesting but not actionable, leaving the mechanistic heart of the problem unexplored. As long as this continues, expansion studies will keep failing, even when the drugs themselves remain clinically powerful.
Why ASCENT-07’s Failure Should Alarm the Industry
The concerning part is not that ASCENT-07 missed its endpoint. The concerning part is that it likely missed for reasons that were scientifically solvable, but invisible to standard workflows.
And no pharma company, no matter how large, is protected from this. Not because they lack talent, but because they lack a system designed to interrogate biology at the depth today’s therapeutics require.
What Oncology Actually Needs (and It’s Definitely Not More Sequencing)
The solution is not deeper sequencing, larger cohorts, or more expensive assays. Oncology needs a fundamentally different analytical paradigm, one capable of:
- integrating all omics layers coherently
- resolving heterogeneity down to causal architecture
- peeling the onion from surface phenomena to mechanistic cores
- extracting minimal, assay-ready biomarkers
- producing insights that survive cross-platform, multi-cohort validation
- mapping resistance mechanisms to therapeutic strategies
- informing sequencing, combinations, and patient stratification with clinical fidelity
This requires abandoning correlation-driven analytics and embracing a causal, systems-based framework.
Where Bioada Fits: A System Built for Mechanistic Depth
Bioada has spent the past years building precisely this framework. Across glioblastoma, pancreatic cancer, lung cancer, breast cancer, chemotherapy resistance, and metastatic microenvironment studies, our multi-layered, reproducibility-first approach has consistently revealed the drivers, networks, and mechanistic signatures that standard workflows fail to detect.
Each study followed the same blueprint: broad → refined → mechanistic → reproducible → clinically aligned.
These are exactly the insights ADC development requires.And they are precisely what standard pipelines, regardless of scale, cannot deliver.
ASCENT-07 Must Be a Wake-Up Call
ASCENT-07 does not need to be the end of Trodelvy’s story, unless the industry chooses not to learn from it.
The greatest threat to ADC innovation is not the biology; it is the analytical mindset used to interrogate that biology. Expansion programs will continue to fail until we replace descriptive, surface-level analytics with causal, multi-omics discovery frameworks capable of resolving therapeutic mechanisms with precision.
The era of descriptive analytics is over. The era of mechanistic, clinically aligned discovery must begin.
Only through such a shift can we prevent future ASCENT-07s, unlock the true potential of ADCs, and finally deliver on the promise these drugs were designed to fulfill.