The Press Release Is Not Their Product
Why Pharma Commercial Strategy Built on Competitive Corporate Messaging Keeps Losing to Strategy Built on the Regulatory Record
Every Commercial team in pharma runs on intelligence. The question is what that intelligence is made of.
For most organizations, the answer is uncomfortable. The competitive picture is assembled from topline press releases, conference abstracts, investor decks, analyst notes, and expert network calls. All of it is downstream of what a competitor chose to say on a day of its choosing, framed the way its investor relations team wanted it framed.
None of it is what the FDA actually decided.
The gap between those two things is where billions of dollars are lost.
The Evidence That Messaging Is a Biased Signal
Let’s look at the hard data: PR spin is a documented reality. A cohort study examined 70 randomized controlled trials and found that nearly half (47%) of press releases contained "spin" – meaning they actively exaggerated the benefits or downplayed the harms of a new treatment or under-reported harm (Yavchitz and colleagues, PLoS Medicine, 2012).
Even worse? The spin actually works. In a separate study (Boutron and colleagues, Journal of Clinical Oncology, 2014), clinicians were randomized to read cancer trial abstracts with or without spin. Same trial, same data, two framings. Those who read the spun version rated the treatments as more beneficial than they actually were.
The takeaway: If PR spin can fool experienced oncologists, it can definitely fool a commercial team building a competitive launch plan. The chain runs like this: Abstracts introduce spin. Press releases amplify it. And commercial teams build launch plans, indication priorities, and multibillion-dollar bids on the far end of that chain.
Four Times the Narrative Cost Billions
1. Aduhelm: The adcomm said no before the market said yes
Bernstein analysts projected peak sales of around $10 billion for Biogen. Cantor Fitzgerald projected $8.2 billion by 2028. Fierce Pharma described the initial label as shockingly broad. The narrative was that this would be one of the largest launches in industry history.
The regulatory record said something different, and it said it first. In November 2020, the FDA's Peripheral and Central Nervous System Drugs Advisory Committee decided that the pivotal study did not provide strong evidence of effectiveness. Two Phase III trials, ENGAGE and EMERGE, had been halted for futility in March 2019. The approval that followed in June 2021 was accelerated, not full.
Full year 2021 sales were $3 million. One month after approval, the label was narrowed to the mild cognitive impairment and mild dementia population actually studied. In 2022, CMS restricted coverage to patients enrolled in qualifying trials. Biogen discontinued the program in 2024.
The market bought the narrative of a $10 billion mega-blockbuster. The regulatory record quietly predicted a commercial zero.
2. CheckMate-026: A protocol decision that shaped a decade
In 2016, Bristol Myers Squibb led the immuno-oncology market, and Merck was the challenger. One visible difference between their first-line lung cancer programs was the biomarker threshold.
CheckMate-026 enrolled patients with PD-L1 expression of 1% or greater, with the primary efficacy analysis in the 423 patients at 5% or greater. KEYNOTE-024 enrolled only at 50% or greater. CheckMate-026 missed its mark: median progression-free survival of 4.2 months against 5.9 for chemotherapy, with hazard ratio 1.15 (95% CI 0.91 to 1.45; New England Journal of Medicine, 2017). BMS shares closed down about 16% on August 5, 2016, and fell roughly 23% across that month. Merck won first-line approval in October 2016.
Keytruda recorded $29.5 billion in 2024 revenue, up 18%, against Merck's total of $64.2 billion.
The enrollment threshold was not the only factor. PD-L1 assays and cut points differed, chance played a part, the trial arms were imbalanced in ways the authors noted, and BMS later recovered ground with combination strategies in CheckMate-227 and CheckMate-9LA. However, the design decision was material, it was consequential, and it was visible in the protocol long before it was visible in BMS’ revenue.
3. Rova-T: $5.8 billion in, roughly $4 billion written off
AbbVie acquired Stemcentrx in 2016 for approximately $5.8 billion upfront, with up to $4.0 billion in milestones. Contemporaneous forecasts put peak sales near $5 billion a year.
The Phase II TRINITY trial returned a 16% objective response rate. In December 2018, the Phase III TAHOE trial was halted after an independent monitoring committee found shorter overall survival on Rova-T than on topotecan (6.3 months against 8.6, hazard ratio 1.46). MERU was closed at interim analysis in August 2019.
AbbVie's Form 8-K of January 4, 2019, put the estimated losses at approximately $4 billion.
4. NKTR-214: The number that moved with the denominator
Bristol Myers Squibb paid Nektar $1.85 billion upfront in February 2018 for exclusive rights across nine tumor types.
The excitement came from small early cohorts. In first-line kidney cancer, the objective response rate was 46% in 13 patients at SITC in 2017 and 71% in 14 patients at the ASCO update in 2018. In the mature cohort of 49 patients, it was 34.7% (Journal for ImmunoTherapy of Cancer, 2022). The Phase III program failed in melanoma, kidney, and urothelial cancer, and the collaboration was terminated in April 2022.
BMS paid $1.85 billion based on the excitement of 14 hand-picked patients at ASCO. By the time the full mature cohort was published, the magic was gone – and the billions were lost.
Four Times the Regulatory Record Paid Off
The winning franchises did not win on better messaging. They won on a better reading of what regulators accept and what labels definitively say. For example:
Keytruda made label strategy their product strategy: biomarker selection in KEYNOTE-024, the first tissue-agnostic approval in FDA history (MSI-H and dMMR solid tumors, May 2017), then sustained expansion into adjuvant and perioperative settings.
Merck reached its 40th US indication in June 2024, securing $29.5 billion that year powered by label expansion.
Wegovy is the clearest case in the industry of a label change creating a market. Medicare Part D is statutorily barred from covering drugs used for weight loss. It is not barred from covering a drug indicated to reduce cardiovascular risk. SELECT delivered a 20% relative reduction in major adverse cardiovascular events (New England Journal of Medicine, 2023), FDA approved the cardiovascular indication in 2024, and CMS confirmed Part D plans could cover the product on that basis. Across the category, Ozempic, Wegovy, Mounjaro, and Zepbound together generated roughly $42 billion in 2024.
No press release did that. A label section did.
Trikafta, Tagrisso, and Enhertu all won on sequential, precedent-aware label strategy. Trikafta and Kaftrio brought in about $8.9 billion in 2023, expanding by age and by mutation to reach 272 CFTR mutations by December 2024. ADAURA opened the adjuvant EGFR setting for Tagrisso, approved December 18, 2020, and Tagrisso recorded $6.58 billion globally in 2024. Enhertu recorded $3.75 billion in combined global sales across AstraZeneca and Daiichi Sankyo in 2024.
What Intelligence Is Actually Sitting in the Regulatory Record
The usual objection is that primary regulatory documents are slow, dense, and not commercially useful. That has not been true for some time.
Want to know exactly what the FDA told your competitor behind closed doors? A quick search across FDA approval packages on the Basil Systems platform reveals the actual dialogue. You can see the exact moment the FDA tells a sponsor, "No, we do not agree," watches the sponsor ignore the advice, and then slaps them with a Complete Response Letter. You can follow FDA proactively coaching sponsors on design at pre-NDA meetings. You can even capture a full picture of the target population of a competitor before a single clinical trial is even announced publicly. That is your competitor’s exact playbook, their mistakes, and the agency's stated preferences, all in the public record. None of this will ever appear in a press release.
The same applies to the commercial questions. What specific indication language did a competitor secure, and what did the agency strike out? Which trial populations appear in Section 14, and which do not? What warnings constrain their promotional claims? What did FDA push back on in a comparable submission, and why?
Being Honest About the Limits
Primary regulatory data is the reliable spine of commercial intelligence. It is not the whole skeleton, and anyone selling it as such is doing the same thing this article is criticizing.
Documents lag. Approval packages and Complete Response Letters carry redactions. Labels tell you what a company can claim, not what payers negotiated, what the net price is, or what uptake looks like in the field. And regulatory documents, when read too confidently, produce their own errors, in both directions. In one example, AVEO argued a favorable survival trend for tivozanib that never firmed up, and an ODAC panel voted against it in 2013. Read as a permanent verdict on the molecule, that conclusion would have been wrong too: tivozanib was approved as Fotivda in March 2021 for relapsed or refractory advanced renal cell carcinoma on the strength of TIVO-3.
The point is not that the regulatory record replaces market research, payer intelligence, and field insight. The key insight for Commercial and Competitive teams is that regulatory intelligence should be the layer everything else is checked against, rather than the layer no one opens until submission.
Three Moves for Commercial and Competitive Intelligence Leaders
1. Audit where your competitive picture truly comes from. Take your last three competitive assessments and trace each material claim back to its source. If the trail ends at a press release, an abstract, or an analyst note rather than a label, an approval package, or a registry record, you have found your exposure.
2. Make the label the reference layer, not the appendix. The label governs what a competitor can claim, which patients they can reach, and what payers will fund. Track competitor label changes as they happen in real time, version against version and line by line, rather than reconstructing them after a launch surprises you.
3. Read your competitors' regulatory review documents before you design your own submission. If a comparable product had an indication narrowed, an endpoint challenged, or an estimand rejected, that is your precedent. Finding it during your own review cycle costs a cycle. Finding it beforehand costs an afternoon or less.
It costs roughly $4 billion to bring a drug to market, including the cost of failures. The overall Phase I-to-approval success rate is only 13.8% (Wong, Siah, and Lo, Biostatistics, 2019). Against those staggering odds, running your commercial strategy based on what competitors choose to announce in a press release is not a shortcut – it is an unpriced risk.
The regulatory record is specific, validated, and it is fully searchable with Basil. The only real question is whether your competitors are reading it before you do.
Regulatory document analysis conducted using the Basil Systems platform's Regulatory / Approval Search module, with full text search across FDA and EMA approval packages, review documents, labels, and correspondence, spanning over 700 million records.
Author: Sam Kay, VP of Pharma, Basil Systems