How Biotech Drugs Get from Lab to FDA Approval: The 8-Stage Pathway

You see a headline about a biotech stock doubling on FDA approval. You want in. But before you bet money on a drug you’ve never heard of, you need to understand the path it traveled to get there. Every biotech drug, from Keytruda to Casgevy, runs the same 8-stage gauntlet. Most don’t make it.

Here is the map I wish I had when I started.

The problem

Biotech press releases talk about Phase 2 readouts, PDUFA dates, and CRLs like everyone already knows what they mean. They don’t. And if you don’t know which stage a drug is in, you don’t know your odds of losing money.

The solution

Every drug travels through 8 stages. Each stage has a failure rate. That failure rate is your investment risk. Here they are, with real numbers.

Step 1: Target Discovery (2-3 years, $50-100M)

Researchers identify a biological target: a protein, gene, or pathway that drives a disease. They test thousands of compounds to find one that hits the target. At this stage, the drug exists only in test tubes and computer models. About 1 in 10,000 compounds that enter target discovery will ever become an approved drug.

Real example: The PD-1 protein that Keytruda blocks was discovered in 1992 at Kyoto University. The drug itself wasn’t approved until 2014. That’s 22 years from target to market.

Step 2: Preclinical Testing (1-2 years)

Before a drug touches a human, it goes through animal testing and lab studies called IND-enabling studies. The goal is to prove the drug is safe enough to test in people. Researchers study how the drug is absorbed, distributed, metabolized, and excreted (ADME). They look for toxicity signals in two animal species, usually a rodent and a non-rodent. About 70% of preclinical candidates advance to human testing.

Step 3: IND Filing (30 days)

The company files an Investigational New Drug (IND) application with the FDA. The IND includes all preclinical data, manufacturing details, and the proposed clinical trial protocol. The FDA has 30 days to review it. If the FDA doesn’t respond, the company can start Phase 1. If the FDA places a clinical hold, the company must resolve the FDA’s concerns before enrolling patients.

This is the first real regulatory gate. Most INDs clear it. When an IND is placed on clinical hold, it’s usually because of a safety signal in the preclinical data or a manufacturing problem.

Step 4: Phase 1 Clinical Trial (1-2 years, 20-100 people)

The first time the drug enters humans. Phase 1 trials test safety, not efficacy. The goal is to find the right dose and understand side effects. Participants are usually healthy volunteers, except in oncology where patients with the disease are enrolled.

The numbers: About 70% of Phase 1 drugs advance to Phase 2. The other 30% fail on safety or pharmacokinetics. A Phase 1 press release that says “well-tolerated” doesn’t mean the drug works. It means nobody died.

Real example: Moderna’s mRNA-1010 flu vaccine showed strong immune responses in Phase 1. But Phase 3 efficacy data is what moved the stock, not the safety data.

Step 5: Phase 2 Clinical Trial (2-3 years, 100-300 patients)

Here we finally test whether the drug actually works. Phase 2 looks for an efficacy signal, refines dosing, and collects more safety data. This is where most drugs die.

The numbers: Only about 33% of Phase 2 drugs reach Phase 3. That’s the single biggest cliff in drug development. A promising Phase 2 result is a hypothesis, not an approval signal. Two out of three drugs that look good in Phase 2 will fail to replicate in a larger trial.

Red flag: A company that reports “positive Phase 2 data” but the trial had 30 patients and no control arm. Small, uncontrolled Phase 2 trials are noise, not signal. If the press release emphasizes “encouraging trends” instead of concrete numbers, the data is soft.

Step 6: Phase 3 Clinical Trial (3-5 years, 300-3,000 patients)

The pivotal trial. Phase 3 is a randomized, controlled trial designed to prove the drug works better than placebo or standard of care. The primary endpoint is pre-specified. The statistical analysis plan is locked in. This is the data the FDA will use to decide approval.

The numbers: About 55-60% of Phase 3 drugs reach approval. Even at this stage, 40-45% fail, usually because the efficacy seen in Phase 2 didn’t hold up in a larger, more rigorous trial. Development costs for a single Phase 3 program can run $100-400 million.

Real example: Eisai and Biogen’s Leqembi showed a 27% slowing of cognitive decline on the CDR-SB scale in the Phase 3 CLARITY-AD trial. That result earned traditional FDA approval in July 2023. Without a clean Phase 3 win, the drug would have stayed in accelerated-approval limbo.

Step 7: NDA or BLA Filing (6-10 months review)

The company submits all clinical, preclinical, and manufacturing data to the FDA. Two filing types:

• NDA (New Drug Application): For small molecule drugs, pills and chemical compounds. 5 years of market exclusivity before generics can file.
• BLA (Biologics License Application): For biologics (monoclonal antibodies, gene therapies, cell therapies, vaccines). These are made from living cells. 12 years of exclusivity before biosimilars can file.

The FDA has 60 days to decide whether to accept the filing for review. If accepted, the review clock starts: 10 months for standard review, 6 months for priority review. The PDUFA date is the target decision date. On that date, the FDA issues an approval, a Complete Response Letter (CRL) rejecting the application, or occasionally a delay.

The numbers: 37% of NDAs and BLAs received a CRL from 2018 to 2022. That’s more than 1 in 3 applications rejected. A CRL can crater a stock 30-80% in a single day. You can search CRL outcomes at open.fda.gov.

Real example: Replimune’s RP1 BLA for advanced melanoma received two CRLs (July 2025, April 2026) before a third resubmission got an AdCom vote on July 30, 2026 and PDUFA on August 2, 2026. Three tries to cross the finish line.

For more on what happens at the PDUFA date, see our guide on FDA catalysts and what to watch. For how to read the trial data behind any filing, start with how to read a clinical trial press release.

Step 8: Approval and Post-Market (Phase 4)

The drug is on the market. But the FDA’s job isn’t done. Phase 4 trials monitor long-term safety in real-world patients. Companies can also file supplemental applications (sNDA/sBLA) to expand the label to new indications or patient populations. Supplemental applications have higher approval rates than initial filings because the safety profile is already established.

Real example: Vertex and CRISPR Therapeutics’ Casgevy was first approved in December 2023 for sickle cell disease. By mid-2026, only 39 patients had been infused out of roughly 60,000 addressable. Approval is not the same as commercial success.

Common mistakes

Mistaking Phase 2 for proof. A Phase 2 readout with 50 patients and no control arm tells you almost nothing about whether the drug works. Wait for Phase 3.

Buying the stock the day before the PDUFA date. You’re betting on a coin flip with a 37% failure rate. If the PDUFA is already priced in, the risk-reward is asymmetric in the wrong direction: a CRL drops you 40-60%, while an approval might only add 10-15%.

Ignoring manufacturing risk. Biologics are grown in living cells. Batch-to-batch variability is inherent. A CMC (chemistry, manufacturing, controls) CRL can derail a drug that works perfectly. Elevar Therapeutics received three CRLs for rivoceranib on manufacturing deficiencies alone, despite the drug being approved in China.

Final checklist

• Which stage is the drug in? (Phase 1 = 70% failure after this point, Phase 2 = 67% failure, Phase 3 = 40-45% failure)
• Has the FDA accepted the NDA/BLA for review? (60-day acceptance window)
• Is the review Standard (10 months) or Priority (6 months)?
• What’s the PDUFA date? Mark it two weeks early.
• Is the manufacturing facility ready? (FDA inspections, Form 483 observations)
• How much cash does the company have? Under 12 months of runway means dilution risk regardless of pipeline quality.

The 8-stage pathway is the same for every biotech drug. The only difference is how many make it through.