FDA drug studies have traditionally required two pivotal randomized controlled trials. This makes the chance of a false positive very tiny. (I also thought it might be because statute requires approval after well-controlled trialS, with an S).
We now have a packet of 5 FDA documents on scientific evaluations based on plausibility and drug approvals based on single trials.
- See the Feb 23 rare disease press release here.
- See the Feb23 rare disease draft guidance document here. (Comment to ~ Apr 23).
- See also a 2025 NEJM article by FDA on the general "plausible evidence" theme and how it might be applied in rare diseases here.
- See the brand-new NEJM article on 1 trial approvals, here. (Prasad & Makary Feb 18).
- See a trade journal regarding the new 1-trial NEJM article here.
Relevance to Diagnostics:[Chat GPT]
This evidentiary shift also has relevance for molecular diagnostics, where randomized controlled trials are often neither feasible nor ethically appropriate. In genomic testing, clinical utility frequently rests on strong biologic rationale, risk stratification accuracy, and concordance with outcomes — not on forcing artificial randomization. For example, it would be ethically untenable today to randomize women with a very low recurrence score on Oncotype DX to receive chemotherapy, or to deny chemotherapy to women with a very high score, simply to generate RCT purity.Instead, validation relied on retrospective-prospective analyses, biologic coherence, and concordance with clinical outcomes. In that sense, molecular diagnostics have long operated within a “plausible mechanism plus confirmatory evidence” paradigm. The FDA’s broader embrace of mechanism-anchored inference and one-trial flexibility (NEJM 2025; NEJM 2026) brings drug regulation somewhat closer to the evidentiary norms that diagnostics have navigated for years — particularly in oncology, where risk stratification tools guide treatment decisions without mandating ethically problematic randomization.
FDA’s Two NEJM Articles: From “Plausible Mechanism” to a One-Trial Default?
Over the past three months, FDA leadership has published two closely related but strategically distinct articles in The New England Journal of Medicine.
First came the December 2025 article which was a general introductoin to the “Plausible Mechanism Pathway” (NEJM 2025).
Then, in February 2026, came a second article announcing that one pivotal trial will now be the FDA’s default standard for approval (NEJM 2026).
These are not isolated commentaries. Together, they represent a meaningful reframing of how FDA interprets “substantial evidence.”
Below, I unpack how the two pieces relate — and address the statutory question.
Part I: The 2025 NEJM “Plausible Mechanism” Framework
The December 2025 NEJM article describes a regulatory pathway for ultra-rare, individualized genetic therapies (NEJM 2025). The Agency formalized the framework in draft guidance issued February 23, 2026 (FDA Guidance 02-23), accompanied by a press release the same day (FDA Press 02-23).
The core elements of the framework are:
Identification of a specific genetic, molecular, or cellular abnormality
Therapy directly targeting that abnormality
Well-characterized natural history
Confirmation of target engagement
Clinical improvement inconsistent with expected disease course
The conceptual shift is clear. Rather than centering randomized replication, the framework emphasizes:
Mechanistic coherence + biologic targeting + natural history contrast.
Randomized controlled trials may not be feasible in ultra-small populations. In those settings, FDA signals that strong biologic plausibility and dramatic clinical effect may suffice (NEJM 2025; FDA Guidance 02-23).
The framework explicitly allows:
Small sample sizes
Patients as their own controls
External natural history controls
Modular “platform” approvals (e.g., adding genome editing variants later)
This is regulatory flexibility, but it is also a philosophical repositioning of evidentiary hierarchy.
Part II: The 2026 NEJM Announcement — “One Pivotal Trial” as Default
The February 2026 NEJM article broadens the discussion from ultra-rare therapies to drug approvals generally (NEJM 2026).
FDA leadership announces that:
One adequate and well-controlled trial, combined with confirmatory evidence, will now be the Agency’s default standard (NEJM 2026).
Historically, FDA relied on two pivotal trials to reduce the risk of false positive findings — the classic “one must be lucky twice” rationale described in the article (NEJM 2026). That standard emerged in the 1960s in an era of limited biologic understanding.
The 2026 article argues that modern drug development now incorporates:
Mechanistic science
Biomarker alignment
Bayesian interpretation
Magnitude of effect
Robust trial design features
Postmarket data collection
Two trials are reframed as one component of credibility — not the anchor of credibility (NEJM 2026).
Notably, the word “plausible” does not appear in the 2026 article. Instead, the discussion shifts to “biologic inference,” “confirmatory evidence,” and “interlocking facets of credibility.”
The philosophy is continuous with the 2025 article — but the rhetoric is more generalizable.
Media and Policy Context
BioPharma Dive summarized the 2026 policy shift as ending the “two-trial dogma,” emphasizing cost reduction and faster development timelines (2026 Biopharma Dive).
The FDA press release accompanying the rare disease guidance frames the policy as cutting red tape and accelerating cures for ultra-rare diseases (FDA Press 02-23).
The two narratives — innovation acceleration and evidentiary modernization — are clearly aligned.
How the Two NEJM Articles Fit Together
The 2025 article introduces the concept in a narrow setting.
The 2026 article extends the logic across the approval framework.
| December 2025 | February 2026 |
|---|---|
| Ultra-rare individualized therapies | All drug approvals |
| Explicit “plausible mechanism” | One-trial default |
| Natural history as control | Bayesian credibility framing |
| Platform expansion | Stronger postmarket monitoring |
The 2025 NEJM piece establishes the basis. The 2026 NEJM piece operationalizes it system-wide. However, the application in guidance is narrow, being this week's guidance on rare diseases.
Although not directly tied to the two NEJM articles, December and February, on Feb 23, the FDA released a press release and draft guidance about evaluation of drugs in rare diseases.
##
In addition to the two NEJM articles, FDA has now provided a concrete regulatory example in the February 23, 2026 draft guidance on individualized therapies for rare genetic conditions (FDA Guidance 02-23), accompanied by a press release emphasizing acceleration and regulatory flexibility (FDA Press 02-23). The Rare Disease guidance operationalizes many of the principles previewed in NEJM 2025 — particularly reliance on natural history data, use of externally controlled designs, and the possibility that a first-in-human study may also serve as the pivotal study. The guidance explicitly states that substantial evidence may be established through a single adequate and well-controlled clinical investigation with confirmatory evidence in settings where randomized trials are not feasible (FDA Guidance 02-23).
While the document is focused on ultra-rare, mutation-specific therapies, it functions as a tangible application of the broader evidentiary philosophy described in NEJM 2025 and extended in NEJM 2026. In that sense, the Rare Disease guidance is not merely adjacent to the NEJM discussion — it is an implementation example of the same doctrinal shift toward mechanism-anchored inference and one-trial flexibility.
The Statutory Question: What About “TrialS” (Plural)?
A common reaction is: didn’t the statute require two trials?
Section 505(d) of the FD&C Act historically required “adequate and well-controlled investigations.” For decades, this was interpreted as two pivotal trials.
However, the 1997 FDA Modernization Act explicitly allowed approval based on:
One adequate and well-controlled investigation plus confirmatory evidence.
The 2026 NEJM article references this statutory authority directly (NEJM 2026).
Thus:
FDA has had authority to approve on one trial since 1997.
The “two-trial rule” was a policy norm, not an absolute statutory requirement.
The 2026 article resets the default expectation, not the law.
The authors explicitly invoke behavioral economics: defaults matter (NEJM 2026).
What Is Actually Changing?
Two shifts are visible.
1. Statistical Philosophy
1960s paradigm:
Frequentist logic
Replication reduces type I error
Two trials as safeguard
2026 paradigm:
Bayesian interpretation
Mechanistic priors
Magnitude of effect
Totality of evidence
The argument is that two weak trials do not provide more credibility than one strong trial (NEJM 2026).
2. Regulatory Emphasis
The Plausible Mechanism guidance emphasizes flexibility in ultra-rare conditions (FDA Guidance 02-23; FDA Press 02-23).
The one-trial announcement emphasizes reducing development costs and speeding access (NEJM 2026; 2026 Biopharma Dive).
Both are framed as modernization, not relaxation.
Risks and Tensions
Supporters argue:
Precision medicine strengthens biologic priors.
Ultra-rare diseases cannot support conventional designs.
Stronger trial scrutiny plus postmarket monitoring can compensate for fewer trials.
Critics argue:
Mechanistic reasoning has historically misled medicine.
Surrogates can fail.
Postmarket correction is slower and politically difficult.
Reduced replication increases approval risk.
The tension is not new — but the formal resetting of the default is new.
Bottom Line
The December 2025 NEJM article introduces the Plausible Mechanism framework for individualized therapies (NEJM 2025).
The February 2026 NEJM article extends the logic and makes one adequate and well-controlled trial the default standard (NEJM 2026).
The statute has allowed one-trial approval since 1997. The change is cultural and operational — not legislative.
The word “plausible” disappears in the 2026 article because it has been absorbed into a broader doctrine of biologic and statistical credibility.
Whether this shift proves prudent will depend on implementation — particularly how rigorously FDA scrutinizes endpoints, controls, confirmatory evidence, and postmarket obligations.
What is clear is that the evidentiary center of gravity has moved.
And that matters.
