TL;DR

A peer-reviewed study by Intelligencia AI and GSK, analyzing ~4,000 oncology programs, reveals that single-agent activity is the most reliable predictor of combination therapy success. While the baseline approval rate for oncology combinations is only 4%, assets with prior monotherapy efficacy, including those that “failed” to meet primary endpoints but showed objective response rates, have significantly higher success rates when repositioned. This data-driven approach allows R&D leaders to de-risk portfolios and identify hidden value in deprioritized assets.

What Drives Success in Oncology Combination Therapies? Insights from Our New Peer-Reviewed Study

Oncology drug development remains one of the most complex and capital-intensive endeavors in biopharma. With development timelines often exceeding a decade and costs per asset surpassing the billion-dollar mark, the ability to make earlier, better-informed decisions is no longer a competitive advantage, it is a necessity.

In a newly published peer-reviewed study, Intelligencia AI, in collaboration with a leading global pharmaceutical partner, applied large-scale clinical trial intelligence to one of the most critical strategic questions in oncology R&D: what truly drives success in combination therapy development?

Turning Historical Data into Forward-Looking Strategy

Combination therapies are now the backbone of modern oncology pipelines. Yet despite their promise, historical approval rates remain stubbornly low. Using Intelligencia AI’s proprietary clinical trial database—covering nearly 4,000 historical oncology programs—the study systematically analyzed how early signals of single-agent activity influence the probability of eventual regulatory approval for combination regimens.

The findings are both intuitive and actionable. Across solid tumor indications and development phases, combination programs built on agents that had already demonstrated meaningful single-agent activity consistently outperformed those that had not. Overall, oncology combination programs achieved an approval rate of just over 4%. However, when at least one component drug had previously shown efficacy as a monotherapy, approval rates increased materially. Even more striking, drugs that failed as monotherapies but demonstrated strong objective response rates still meaningfully improved the odds of success when repositioned in rational combinations.

In practical terms, this means that “failure” in early development does not necessarily signal the end of value. When interpreted through the right analytical lens, early clinical data can reveal hidden optionality—assets that may be better suited for combination strategies rather than single-agent development.

From Evidence to Portfolio Action

For R&D and portfolio leaders, the implications of these findings are both strategic and operational. The analysis provides empirical validation for a more disciplined approach to combination strategy—one that explicitly links early single-agent signals to downstream regulatory outcomes.

Rather than treating monotherapy failure as a binary stop signal, the data demonstrates that assets with measurable biological activity can retain significant strategic value when redeployed in rational combinations. By quantifying how different levels of single-agent activity correlate with approval outcomes across indications and phases, the analysis offers a concrete, decision-ready framework that supports:

• Earlier identification of viable combination opportunities, including assets that may have been deprioritized based on monotherapy performance alone
• Reduction of late-stage attrition risk, by filtering out combinations with structurally low probabilities of regulatory success before capital is heavily committed
• More effective capital allocation, prioritizing programs with empirically higher likelihoods of approval and clearer paths to value creation

At a portfolio level, this shifts decision-making away from isolated case studies and therapeutic intuition toward evidence-based risk stratification at scale. In an oncology landscape characterized by crowded pipelines, rising development costs, and heightened regulatory scrutiny, the ability to connect early clinical signals with long-term outcomes becomes a decisive competitive capability.

Looking Ahead

As oncology continues to evolve toward more personalized and combinatorial treatment paradigms, the importance of leveraging historical data intelligently will only grow. This study reinforces a simple but powerful message: better use of existing data can materially improve future outcomes.

At Intelligencia AI, we remain committed to working alongside our partners to transform clinical data into decisions that matter—reducing uncertainty, optimizing development paths, and helping promising therapies reach patients faster.

Read the Full Study

The full peer-reviewed article, Strategic risk assessment in oncology: Utilizing single-agent activity to boost combination therapy approvals, was published in Contemporary Clinical Trials Communications and is available open access.

The study was developed through a close collaboration between Intelligencia AI and GSK, combining large-scale clinical trial intelligence with deep oncology development expertise. Together, the teams examined thousands of historical oncology programs to translate real-world clinical evidence into actionable insight for portfolio and development decision-makers.

For readers interested in the underlying data, methodology, and indication-level results, the full publication provides a detailed and transparent view of the analysis and its implications for modern oncology R&D.

Read the full publication here.

For more tailored, data-rich insights, let’s talk.