Biotech Clinical Trial Phases: I, II, III Explained

Key Takeaways

• Biotech clinical trial phases filter candidates at increasing cost: only about 7.9 percent of drugs entering Phase 1 are eventually approved according to the BIO/Informa 2011-2020 study of 9,704 programs.
• The Phase 2 to Phase 3 transition is the most lethal filter at 28.9 percent success rate; Phase 3 to NDA/BLA approval is the strongest at 57.8 percent, so the stock reaction to Phase 2 success is often the largest single binary event.
• A common mistake is treating Phase 3 success as approval, the FDA still rejects roughly 9 percent of NDA/BLA submissions, and a Complete Response Letter can add 6 to 18 months and material cost.
• Disease-area differences are large: oncology's Phase 1-to-approval rate is roughly 3.4 percent versus above 20 percent for some hematology programs, so applying an average probability misprices biotech risk systematically.

What It Is

A clinical trial is a controlled human study of a drug candidate. The FDA organizes these studies into four phases. Phase 1 tests basic safety in a small group. Phase 2 tests whether the drug actually does something useful at a tolerable dose. Phase 3 confirms the result in a much larger population that looks like the real patient pool. Phase 4 happens after approval and tracks long-term safety.

Only a small fraction of compounds that enter Phase 1 ever reach the market. The BIO/Informa study of 9,704 drug programs from 2011 to 2020 found an overall 7.9% probability of approval from Phase 1. Wong, Siah, and Lo (2019), working with 21,143 compounds, reported a similar 13.8% success rate under broader inclusion rules, with oncology far below at roughly 3.4%.

The Intuition

Drug development is an expensive elimination tournament. Most ideas fail, and the industry would rather fail them cheaply than late. Each phase is a filter designed to kill bad candidates before more money is spent on them.

Phase 1 asks a simple question. Can a human tolerate this molecule without serious harm? If yes, Phase 2 asks whether the molecule does what the biology suggested it should, at a dose the body can accept. If yes, Phase 3 asks whether the effect holds up in hundreds or thousands of real patients, against a placebo or the current standard of care. A candidate that clears all three still has to pass regulatory review, which rejects roughly 10 percent of applications on quality, manufacturing, or data issues.

For investors, the staging matters because risk drops sharply with each clearance. A Phase 2 success is worth more than a Phase 1 success, and a Phase 3 success is worth more again. This is the source of the "phase transition" valuation framework used across the industry.

How It Works

Each phase has a standard size, duration, and purpose.

Phase 1: 20-100 healthy volunteers (or patients for oncology)
         12-18 months
         Goal: safety, tolerability, pharmacokinetics

Phase 2: 100-300 patients with the target disease
         ~2 years
         Goal: efficacy signal, dose finding

Phase 3: 300-3,000 patients, often multi-country
         ~3 years
         Goal: confirmatory efficacy, safety at scale

Phase 4: Post-marketing surveillance, thousands to millions
         Ongoing
         Goal: long-term safety, rare adverse events

Industry-wide phase transition probabilities from BIO/Informa 2011-2020:

Phase 1 -> Phase 2:    52.0%
Phase 2 -> Phase 3:    28.9%
Phase 3 -> NDA/BLA:    57.8%
NDA/BLA -> Approval:   90.6%
Overall Phase 1 -> Approval: ~7.9%

DiMasi, Grabowski, and Hansen (2016) put the capitalized cost of a single approved drug at roughly $2.6 billion in 2013 dollars, because every approval has to carry the cost of all the failures along the way.

Worked Example

A biotech has a Phase 2 readout due next quarter for a neurology drug. The company's pipeline has one asset, and analysts expect peak sales of $2 billion if the drug reaches market. Risk-adjusted valuation at Phase 2 uses the remaining probability of approval.

Using BIO/Informa neurology-specific rates, which sit near the bottom of the dataset at roughly 5-6% from Phase 1, the remaining probability from an ongoing Phase 2 is approximately:

P(Phase 2 -> Approval) = 28.9% x 57.8% x 90.6% = ~15.1%

Apply that probability to the net present value of peak sales, and you get a risk-adjusted NPV. Before the readout, the stock trades on that expected value. After the readout, the probability jumps (to roughly 52% if Phase 2 succeeds) or collapses to near zero (if it fails). This binary revaluation is why biotech stocks can halve or double on a single press release.

Common Mistakes

1. Treating "Phase 3 success" as approval. The FDA still rejects about 9 percent of NDA/BLA applications, and a Complete Response Letter can add 6 to 18 months and millions in costs. A positive Phase 3 is a strong signal, not a guarantee.

2. Ignoring disease-area differences. Oncology has a Phase 1-to-approval rate near 3.4 percent. Rare disease, hematology, and infectious disease run materially higher. Applying an average probability to a specific indication will mis-price the risk in both directions.

3. Extrapolating a single trial to a probability. A positive open-label Phase 2 with 40 patients is not proof. Replication in a randomized, blinded Phase 3 is the real test. Many Phase 2 hits fail to replicate when the trial is larger and placebo-controlled.

4. Conflating FDA with EMA or PMDA. A drug approved in the US is not automatically approved in Europe or Japan. Each regulator reviews separately, may require local data, and can reach different conclusions. Global launch timelines often stagger by 6 to 24 months.

5. Assuming Breakthrough Therapy changes the probability of success. Expedited designations like Fast Track, Breakthrough, and Priority Review can shorten timelines and improve FDA engagement, but the underlying trial still has to hit its endpoints. The designation is about speed and support, not about the drug working.

Frequently Asked Questions

Q: What are biotech clinical trial phases in simple terms? Clinical trials run in three main phases. Phase 1 tests safety in 20 to 100 people. Phase 2 tests whether the drug actually works in 100 to 300 patients. Phase 3 confirms the result in hundreds to thousands of patients. Each phase is a filter that kills weak candidates before more money is spent on them.

Q: How do biotech clinical trial phases affect investment decisions? Each completed phase raises the probability of approval and justifies a higher risk-adjusted valuation. A Phase 2 success is worth more than a Phase 1 success because the drug has cleared another filter. Investors price the expected value before a readout and then revalue sharply when results arrive, which is why biotech stocks can move 50 to 80 percent on a single press release.

Q: What is a real-world example of biotech clinical trial phases? In the worked example, a Phase 2 neurology drug with $2 billion in peak sales potential has a remaining approval probability of about 15 percent, giving a meaningful but modest risk-adjusted NPV. A Phase 2 success would lift the probability to roughly 52 percent, nearly tripling the risk-adjusted value overnight.

Q: How can investors use biotech clinical trial phases data? Apply indication-specific phase transition rates, not the blended 7.9 percent all-in average. Oncology runs at roughly 3.4 percent from Phase 1; rare disease runs materially higher. Model the risk-adjusted NPV at each phase and watch how the probability steps up or collapses at each readout date on the company's pipeline calendar.

Q: How are biotech clinical trial phases different from FDA expedited designations? Breakthrough Therapy, Fast Track, and Priority Review speed the regulatory process and improve FDA engagement. They do not change the underlying trial efficacy or safety bar. A designated drug still has to hit its endpoints. The designation shortens the timeline after a successful trial but does not raise the probability of the trial succeeding.

Sources

1. FDA. "Step 3: Clinical Research." https://www.fda.gov/patients/drug-development-process/step-3-clinical-research
2. Wong, C.H., Siah, K.W., Lo, A.W. (2019). "Estimation of clinical trial success rates and related parameters." Biostatistics 20(2), 273-286. https://academic.oup.com/biostatistics/article/20/2/273/4817524
3. BIO, Informa Pharma Intelligence, QLS Advisors (2021). "Clinical Development Success Rates and Contributing Factors 2011-2020." https://go.bio.org/rs/490-EHZ-999/images/ClinicalDevelopmentSuccessRates2011_2020.pdf
4. DiMasi, J.A., Grabowski, H.G., Hansen, R.A. (2016). "Innovation in the pharmaceutical industry: New estimates of R&D costs." Journal of Health Economics 47, 20-33. https://www.sciencedirect.com/science/article/abs/pii/S0167629616000291

Disclaimer

This article is educational content only and is not financial advice. Nothing here is a recommendation to buy, sell, or hold any security. Consult a licensed advisor before making investment decisions.