TL;DR
Global policy shifts are driving a move from serial to parallel drug development, increasing risk concentration and potentially reducing the number of new assets entering development. Orphan and rare diseases are becoming a central strategic focus, while PTRS and structured risk assessment are now essential for portfolio decision-making. AI-driven models enable dynamic scenario planning but are intended to complement, not replace, human judgment.
We recently welcomed Stephen Cho, PhD, to the Intelligencia AI’s Scientific Advisory Board. Cho’s career spans the intersection of science, policy, and portfolio decision-making, with leadership roles across research, development, and enterprise strategy. His career began at Eli Lilly & Company and Pfizer as a medicinal chemist and research project leader then pivoted his career advising top Pharma and MedTech clients at McKinsey & Company before joining Novartis, where he held senior roles bridging project management, clinical development, and portfolio strategy. Recently, he joined BeOne Medicines to lead their Global Portfolio & Program Management group.
At Novartis, Cho played a central role in advancing portfolio analytics and decision-making frameworks, including efforts to refine probability of technical and regulatory success (PTRS) assessments for late-stage development projects, and led enterprise-wide portfolio reviews to improve R&D efficiency and program prioritization. As Head of Portfolio Strategy & Analytics, he consolidated several portfolio management groups, strengthening governance and strategic alignment across the company. Cho joined Intelligencia AI’s Scientific Advisory Board, where he continues to champion data-driven approaches to drug development and portfolio management.
From Serial to Parallel Development
As a consequence of recent global policy changes, Cho sees a shift to parallel indication development. This can be illustrated looking at the example of cancer medications. Previously, oncology programs generally followed a familiar path: start in a late line of therapy – often third or fourth line – prove benefit in the sickest, hardest-to-treat patients, and then gradually move earlier in the line of treatments: second line, first line, and finally adjuvant. That evolution could easily span five or more years.
A similar pattern played out across immunology and other therapeutic areas. Companies would typically launch in the indication with the highest confidence of success and then expand into adjacent diseases over many years as evidence accumulated.
In response to global pressures, companies are contemplating parallel development: multiple phase 2B or phase 3 programs starting in close succession, sometimes across different lines of therapy or multiple indications at once.
That shift comes with real consequences. Running several large trials in parallel increases the absolute risk of failure and ties up budget and resources in a small number of assets. Inevitably, that constrains the number of distinct programs a company can support at the clinical stage.
Given that historical PTRS from a phase 1 asset to approval is well under 10 percent, the long-term effect could be fewer new medicines overall, even if individual assets are developed more aggressively.
Orphan and Rare Diseases Move to Center Stage
Many countries afford additional protections and more favorable reimbursement policies for orphan and rare diseases.
This recognition and shift have important strategic implications: more companies are likely to consider clustering development around rare and orphan diseases for a given mechanism of action and could delay or even avoid pursuing broader, non-orphan indications with that same asset. In some cases, a second-generation molecule or different modality might be pursued to go after more prevalent diseases, while the original asset remains focused on rare conditions providing more favorable reimbursement conditions.
For patients with rare diseases, this could be a positive: investigational medicines could arrive earlier, and diseases that once struggled to attract investment may now get more attention. At the same time, the pivot toward parallel indication expansion and orphan-heavy portfolios again points to the potential for fewer truly new mechanisms entering development for more common diseases.
Why PTRS and Structured Risk Assessment Are Essential
Against this backdrop, the quality of risk assessment becomes a central strategic lever. Earlier in his career, Cho brought together a cross-functional team to evaluate if there were better ways to calculate PTRS based on the totality of clinical and nonclinical evidence. These methods combined external benchmarks for similar programs, a careful evaluation of the target product profile, and a structured review of the totality of clinical and nonclinical evidence to adjust PTRS up or down.
What is new today is not the concept of PTRS, but the complexity of the questions it is being asked to answer. Instead of judging a single indication in isolation, companies now need to evaluate entire asset-level integrated development plans: which mix of indications, which order of entry, which lines of therapy, and which trial designs will best balance time to launch, probability of success, patient benefit, and return on investment under the constraints of recently introduced legislation.
Cho sees data-driven PTRS models—especially those that use artificial intelligence (AI) to make sense of the vast amounts of data generated by the industry—as particularly valuable for scenario planning. Rather than relying solely on static historical tables, teams can simulate alternative development paths: what happens to overall PTRS and value if they start with a high-risk “dream” indication versus a more conservative one; how parallel versus serial development changes the shape of risk; what different global launch sequences might mean once MFN and published maximum fair prices are taken into account.
Crucially, he does not frame AI as a replacement for portfolio teams. In his view, most organizations will, at least for some time, run AI-driven risk models and traditional human-led assessments in parallel, comparing outputs and interrogating differences. That process, he argues, is essential to building trust in the underlying data, features, and algorithms, especially when model outputs are used to justify uncomfortable choices about which programs to accelerate and which to shelve.
Governance in a Higher-Stakes World
For heads of portfolio strategy who still think in terms of serial indication expansion, Cho’s message is direct: that mindset is no longer sufficient. Parallel development, difficult as it may feel, is likely to be unavoidable for all assets but especially for small molecule and RNA-based medicines if companies want to fully realize their potential. PTRS-based risk modeling at both asset and portfolio level becomes a core governance tool to decide where to place those bigger, more concentrated bets.
He also anticipates more pointed discussions in governance forums. Leaders will need to challenge proposed indication lists and sequencing, ask whether the plan truly reflects the full opportunity set for a mechanism, and bring pricing and launch-sequence considerations into the conversation earlier than they might have in the past. That, in turn, raises the bar for the quality and transparency of the data and models supporting those decisions.
In short, the new policy environment does not simply lower prices. It reshapes how and where companies innovate, which patients get early access, what kinds of medicines are prioritized, and how risk is measured and managed. For Stephen, the organizations that will navigate this environment best are those that treat PTRS as a living, data-rich discipline—one that combines human judgment, structured evidence review, and increasingly sophisticated AI-enabled tools—and use it to design integrated development plans that work for both patients and portfolios.
For more tailored, data-rich insights, let’s talk.
