How to Read FDA Briefing Documents: The Investor's 2-Hour Guide

It’s 4:30 PM on a Tuesday. The FDA just posted 150 pages of briefing documents for a drug with a $2 billion market opportunity. The Advisory Committee (AdCom) meeting is in 48 hours. You have 2 hours to decide whether to hold or sell.

Most investors read the company’s press release summary and stop there. That’s a mistake. The FDA briefing docs contain the review division’s unfiltered opinion, before the AdCom votes, before the company spins, before the market prices anything in. Here’s exactly what to read.

Where to Find Briefing Docs and When They Post

Briefing documents are posted on the FDA’s advisory committee calendar at fda.gov/advisory-committees, typically 2 business days before the AdCom meeting at approximately 4:00-5:00 PM ET. Set a calendar alert. This is one of the few truly asymmetric information windows in biotech investing, the documents are public, the market hasn’t read them yet, and the company’s spin hasn’t caught up.

If no briefing docs appear by 6:00 PM ET the day before the AdCom, the FDA may have canceled the meeting (rare but it happens, typically means the review issues were resolved or the application was withdrawn).

The 5 Sections to Read (and the 10 You Can Skip)

Open the PDF. You do not need to read all 150 pages. Here are the 5 sections that matter and the order to read them:

1. Executive Summary (pages 1-5). The FDA’s bottom line in their own words. Read this first. If the summary says “the applicant has demonstrated efficacy”, positive. If it says “the applicant has not demonstrated…” or “the clinical relevance is unclear”, that’s your signal. The executive summary sets the tone for the entire document.

2. Clinical Efficacy section. The review division’s independent analysis of the efficacy data. Look for these specific phrases: “marginally significant,” “not statistically convincing,” “single-arm study,” “post-hoc analysis,” and “limited follow-up.” Each one signals the FDA has concerns about the data quality that may not appear in the company’s press release.

3. Safety section. The adverse event tables. Compare serious adverse event (SAE) rates between the treatment and control arms. A 5% or greater imbalance is a flag. Look specifically for deaths, treatment-related discontinuations, and any adverse event with an incidence rate above 10% in the treatment arm. The FDA’s safety reviewer is paid to find problems, they will.

4. Clinical/Statistical review questions (last 5-10 pages). The questions the FDA is asking the AdCom panel to vote on. The wording of the questions tells you what the FDA is worried about. “Discuss the adequacy of the single-arm trial design” means the FDA thinks the trial design is inadequate. “Discuss whether the observed benefit outweighs the safety risk” means the FDA thinks the benefit-risk balance is unfavorable.

5. Benefit-risk summary (if separate). The final integrated assessment. If this section is longer than 2 pages, the FDA has reservations. A short, confident benefit-risk summary is bullish. A long one that weighs every possible concern is bearish.

Skip the chemistry and manufacturing (CMC) section, the pharmacokinetics, the toxicology, the labeling proposals, and the appendices. You can come back to these if there’s a specific manufacturing concern, but for the initial read, they’re noise.

The FDA Tone Decoder

FDA reviewers write in careful, diplomatic language. Here’s what they actually mean:

• “The applicant’s analysis suggests…” = The FDA disagrees with the analysis.
• “The clinical relevance of this endpoint is unclear” = The FDA thinks the endpoint is garbage.
• “A randomized controlled trial would provide more interpretable data” = The single-arm study is not enough for approval.
• “The safety database is limited to N patients” = The sample size is too small for the FDA’s comfort.
• “The applicant’s proposed labeling overstates the clinical benefit” = The FDA is going to restrict the label.

When the Replimune RP1 briefing docs dropped on July 28, 2026, the FDA called the IGNYTE trial “not interpretable” and questioned whether the single-arm design could support approval. The stock dropped 32% that day. The AdCom still voted 10-3 in favor four days later, but the briefing docs set the floor for the discussion.

Compare to the Company’s Spin

Companies always issue a press release on briefing doc day. Read it side by side with the FDA document. Highlight every sentence where the company’s characterization differs from the FDA’s. The gap between the two is where the trade lives.

Example: the company says “The FDA briefing documents highlight the strong efficacy data.” The FDA’s executive summary says “Efficacy was demonstrated in a single-arm trial; the durability of response requires confirmation in a randomized setting.” The company spun confidence. The FDA expressed caution. The gap tells you to reduce your position size.

What the Language Actually Predicts

Briefing documents set the floor for the discussion. They do not predict the AdCom vote.

The RP1 melanoma briefing docs in July 2026 were among the most negative in recent memory, the FDA called the pivotal trial “not interpretable.” The panel voted 10-3 to recommend approval anyway. Why? Because AdCom panels are external experts who weigh unmet medical need alongside regulatory mechanics. Patients dying of melanoma with no options changes the calculus.

The pattern: negative briefing docs + high unmet need + a drug that clearly helps some patients = the AdCom might still vote yes. Negative briefing docs + marginal efficacy + an existing standard of care = high probability of a negative vote and eventual Complete Response Letter (CRL).

The 30-Minute Checklist

Here’s the 10-point scan I run on every set of briefing documents:

1. Read the executive summary. Positive or negative tone?
2. Count the number of “not” and “unclear” modifiers in the efficacy section.
3. Check SAE rates, treatment arm vs. control arm, any imbalance above 5%.
4. Note the number of deaths in each arm.
5. Read the AdCom vote questions. What does the FDA want the panel to weigh?
6. Check the benefit-risk summary length (over 2 pages = bearish).
7. Look for “single-arm” or “non-randomized”, these are efficacy concerns.
8. Check patient numbers: under 100 patients = the FDA will note the limited database.
9. Compare the company’s press release characterization to the FDA’s language.
10. Set a reminder to watch the AdCom webcast live, the Q&A often reveals more than the vote.

Common mistake: Reading the briefing docs and concluding “the FDA hates this drug, sell everything.” The FDA is supposed to be skeptical. They’re the regulator. A well-argued set of concerns in briefing docs is not a CRL, it’s the FDA doing its job. The question is whether the concerns are resolvable with additional discussion, additional data, or label restrictions. If they’re not, that’s when you sell.

For how to trade around the AdCom event itself, see how to trade FDA catalysts. For the mechanics of how AdComs work and what the vote means, read what is an FDA AdCom. For the full RP1 briefing doc breakdown that taught me most of what’s in this guide, see RP1 FDA briefing docs analysis.