Scenario Two: The dynamic nature of development risk
Our second scenario looks at development risk and what the findings of the P3i report mean for CI teams when applied using our four-stage process.
PTRS, the probability of technical and regulatory success from Phase I to approval, fell from 9.6% to 8.2%, and nearly all of the decline comes from Phase I. Phase I transition success dropped from 49.9% to 41.6%, while later phases held steady or improved slightly. Approval success rose from 94.0% to 97.0%. The bar is higher at the door, and the odds improve once a drug is through.
The bar is higher at the door, and the odds improve once a drug is through.
Figure 9: Phase transition success rates
Across the Top 25 – each represented by a dot in figure 10 below - companies with a higher share of Phase I discontinuations tend to have higher Phase III success.
Figure 10: Early fails and late-stage successes
Phase I now accounts for 38% of disclosed failures, up from 32%. Approval-stage failures, the costliest point to fail, fell from 1.8% to 0.8%. The shift also responds to cost inflation. When each new drug costs more than $100m to bring into the pipeline, an early read on whether to stop has real value.
Phase I timelines are lengthening the most, and failures are moving to Phase I. Together, these suggest that Phase I is increasingly designed for go/no-go calls.
Figure 11: Discontinuations by phase of development
For our scenario, we return to the team covering solid tumor indications. Phase I success rates have dropped significantly, and a higher percentage of failures is disclosed at Phase I. The same four steps show how that applies to our CI team.
Phase I is increasingly designed for go/no-go calls.
In oncology, the Phase I drop is steeper than the industry's, from 49.8% to 35.8%. It holds in both hematological indications, from 52.8% to 37.4%, and solid tumors, from 49.8% to 36.6%.
Figure 12: Oncology: Top 20 phase transition success rates (2020-2025)
Figure 13: Oncology: Top 20 Phase I transition success by type
At the indication level, samples are smaller and noisier. Directionally speaking, we found that colorectal Phase I transition success fell the most, from 61.9% to 37.5%. Bladder, stomach, NSCLC and liver cancer also declined. Pancreatic and prostate cancer moved the other way.
We need to treat those last two with care due to small sample size, but the useful questions are whether something in the science changed, such as a new target class or better patient selection, or whether a handful of assets shifted the rate. Treat each as a question to investigate.
Company-level analysis can separate a strategic signal in one therapy area from Phase I discipline across the whole portfolio. Comparing each company's Phase I transition success in solid tumors with its success across all indications shows the difference.
In our sample data, blinded Company B’s solid tumor rate fell from 61.5% to 45.9%, while its overall rate barely moved, from 46.4% to 44.7%. Company C shows a similar pattern: solid tumors fell from 46.9% to 32.8% and overall from 44.9% to 41.6%. One reading is that these companies are raising the bar specifically in solid tumors. The other companies in the example set display company-wide Phase I discipline, which carries no specific solid tumor signal.
If a specific competitor is becoming more selective in solid tumors, assets that reach Phase II likely carry higher conviction and deserve closer attention from a CI team. It could also point to greater conviction elsewhere in the portfolio. Either way, the signal can stay hidden when the analysis stops at the industry average.
Figure 14: Phase I transition success rates – solid tumors vs. all indications
Norstella is in a unique position to conduct these analyses at scale. If you’re looking to set your CI analyses apart and need support, we welcome the opportunity to collaborate.
Get in touch here
A company's attrition also reflects its portfolio mix. PTRS ranges from 6.2% to 10.4% by modality and therapy area, against 8.2% overall. Biologics sit at 10.0% and small molecules at 6.7%, because biologics track pathophysiology and biomarkers more closely. Immunology is at 10.4%, oncology at 7.3% and CNS at 6.2%, where classification is still largely based on symptoms.
A competitor with small-molecule assets in oncology could look riskier than a biologics-heavy immunology player, even if both execute equally well. Set the base rate first, then judge a competitor's attrition against its own modality and therapy area mix.
Figure 15: PTRS rates by modality and therapy area
Three considerations help translate development risk benchmarks into competitive analysis:
A falling PTRS may indicate concentration. A tougher Phase I bar means competitors may be backing fewer, higher-conviction programs.
Map each competitor's discontinuation behavior. Assess historical patterns: how often does the competitor drop assets early versus late? A competitor that historically discontinues assets late presents a different planning risk from one that cuts programs quickly.
Compare like with like. Adjust for modality and therapy area mix when interpreting a competitor's attrition trends.
A benchmark is a starting point. The value comes from taking it down to the indication, trending competitors against it and asking why the outperformers outperform.
Benchmarking the home company should be done in the same exercise. Both examples in this report test assumptions before leadership does.
This report covered two of the four P3i stages. Market access and commercial return carry their own findings. Download the full Pipeline-to-Patient Productivity Index to see how the industry performs from pipeline to patient.
Join our experts, Kelly Chamberlain and Adam Shaffner as they dive into the detail in our on-demand webinar.
Watch now
