Wall Street Is Wary Of NAMs — And For Good Reason: What Happens To Yesterday's Drugs?

By Zaher Nahle, Ph.D., MPA

Better human-relevant evidence could change more than how tomorrow’s medicines are developed. It could change what we think we know — and what markets think they know — about the drugs already in the portfolio.

In part 1, I argued that the transition toward NAMs is creating something more consequential than a new collection of laboratory tools. It is beginning to alter the architecture through which evidence about medicines is generated and evaluated.^1–3 The transition will be gradual, and human-relevant methods should not be presumed superior merely because they are new. Their value must ultimately rest on predictive performance against the reference point that matters most: human biology and outcomes.

For investors, however, that transition raises a second question. Most discussion of NAMs concerns the prospective pipeline — whether better human-relevant evidence can identify toxicity earlier, eliminate weak candidates before expensive clinical trials, rescue compounds penalized by species-specific findings, or improve dose and patient selection. But what happens if increasingly predictive methods also change our understanding of medicines that have already been developed, approved, and valued?

That is where NAMs become a Wall Street story.

Rewriting The Investment Case

The valuation of a pharmaceutical asset rests on a chain of assumptions: efficacy, toxicity, mechanism, therapeutic window, probability of clinical success, addressable patient population, competitive differentiation, dosing, treatment duration, and expected commercial adoption. Better evidence can alter virtually every one of them.

The effect need not be negative. A NAM might identify a human toxicity that conventional testing missed, but it might also demonstrate that a toxicity observed in an animal has limited relevance to humans, potentially rescuing a candidate that otherwise would have been abandoned. Human-derived evidence might reveal that a medicine believed to benefit a broad population works particularly well in a biologically identifiable subgroup, shrinking the apparent market while strengthening the therapeutic proposition within the patients who actually benefit.

Better evidence can refine dose, explain mechanism, identify susceptibility to adverse events, distinguish responders from nonresponders, support new indications, or challenge an assumed competitive advantage. Some assets may become less valuable. Others may become substantially more valuable. Many will simply become better understood. A drug does not have to be withdrawn for its economics to change: if better evidence changes its estimated responder population, safety profile, dosing assumptions, treatment duration, or competitive positioning, it changes the assumptions underlying its valuation. NAMs are therefore not simply about replacing animal studies; they may change the information on which drugs are developed, regulated, and valued.

The Molecule Did Not Change. The Evidence Did.

Pharmacogenomics provides a useful preview of what happens when biological resolution increases after a medicine has already been developed. Abacavir remains the same molecule, but identification of the association between HLA-B*57:01 and serious hypersensitivity transformed prescribing: FDA states that patients positive for the allele should not receive the drug.⁴ The molecule did not change; the resolution of the evidence surrounding it did.

The principle extends further. Pharmacogenomic information can identify patients at elevated risk of adverse events, change dosing, and distinguish populations more or less likely to respond.⁴ Molecular biomarkers have similarly transformed oncology by dividing diseases once treated as relatively homogeneous entities into biologically distinct populations, changing both who receives a therapy and the size and character of its addressable market.

Post-approval evidence provides another illustration. FDA’s accelerated approval pathway permits earlier approval for serious conditions based on surrogate or intermediate endpoints reasonably likely to predict clinical benefit, with subsequent studies required to verify that benefit.⁵ Some indications subsequently achieve traditional approval; others are withdrawn when the expected benefit is not confirmed or evidentiary requirements are otherwise not met.^6,7 The economically meaningful result of better evidence, therefore, need not be wholesale withdrawal of a drug. It can be a redrawing of its therapeutic boundaries — and that is a more realistic way to think about the eventual consequences of NAMs.

The Question Wall Street Is Not Yet Asking

Most discussion of NAMs concerns tomorrow’s portfolio: Can toxicity be identified earlier? Can weak candidates be eliminated before expensive clinical trials? Can promising compounds be rescued when an animal finding proves irrelevant to humans? Can better doses or patient populations be identified before entering the clinic?

A more disruptive question concerns the portfolio already generating revenue: What might increasingly predictive human-relevant methods tell us about the drugs already on the market?

There is currently no indication that FDA intends to conduct a DESI-like retrospective NAM review of thousands of approved drugs. I would argue, however, that in the long run some form of systematic reexamination should be considered as these methods mature and demonstrate their predictive value. If human-relevant methods become demonstrably better at answering particular questions than models on which earlier regulatory decisions were based, it becomes difficult to justify applying that knowledge only to tomorrow’s medicines while never asking what it tells us about medicines already in use.

That does not mean rerunning thousands of approved drugs through organoids or computer models. Such an exercise would be scientifically crude and economically wasteful. A rational approach would be risk based and question specific, deploying mature, validated methods where they can resolve clinically meaningful uncertainty around human-specific toxicity, mechanism, susceptible populations, dose, or differential response. The objective would not be to recreate DESI, but to apply better knowledge where better knowledge can materially improve decisions.

There is an ethical dimension as well. If a method eventually becomes reliable enough that regulators accept it to help determine whether tomorrow’s medicine can proceed toward human use, it is reasonable to ask whether information generated by that method should also matter when millions of people are already taking yesterday’s medicine. The answer will depend on the method and clinical context, but as predictive performance improves, the question becomes harder to dismiss.

Better Evidence Cuts Both Ways

The comparison with DESI naturally directs attention toward assets that could lose value, but that is only half the investment thesis. Better prediction can create value as readily as it destroys it. For drug developers, the economic opportunity is not simply the cost of an animal experiment avoided; it is better capital allocation. Killing a weak candidate before Phase 2 or Phase 3 preserves capital for stronger programs. Rescuing a compound wrongly penalized by species-specific toxicity can recover an asset that the old system might have discarded. Identifying a high-response subgroup can turn an apparently modest drug into a differentiated precision therapy.

This creates asymmetric exposure across the industry. Companies heavily dependent on mature assets supported by older evidentiary assumptions may face a different risk profile from companies with younger pipelines built around human genetics, translational biomarkers, and human-relevant models. Organizations that integrate NAMs early may also accumulate something more valuable than regulatory experience: comparative data sets showing where new and conventional methods agree, where they diverge, and, critically, which better predicts human outcomes. If those data sets improve subsequent decisions, the informational advantage can compound.

The same logic applies to companies building the evidentiary infrastructure. Organoid platforms, microphysiological systems, human tissue technologies, computational models, high-dimensional data sets, and predictive AI should not ultimately be valued by how many animal experiments they eliminate. Their strategic value will depend on whether they generate information that improves decisions and predicts human outcomes.

As with other platform transitions, early enthusiasm will not make every technology a winner. Validation will matter. Context of use will matter. Regulatory acceptance will matter. Reproducibility, scalability, and integration into pharmaceutical workflows will matter. As comparative evidence accumulates, a market shakeout could follow: some platforms may prove narrow, others redundant, while a smaller group could become embedded in the development infrastructure. For venture investors, the opportunity is therefore not simply to identify companies labeled “NAM,” but to identify the platforms capable of becoming part of the evidentiary infrastructure of drug development.

Where Investors Should Look For The Signal

The earliest investment signal may not be the number of animal studies eliminated. It may be the accumulating comparative evidence generated during the transition. Investors should watch programs in which:

• sponsors generate NAMs and conventional evidence in parallel
• circumstances where regulators conclude that additional animal studies add little information
• cases where human-derived evidence changes toxicity predictions, candidate selection, dose, or patient stratification
• compounds that appear problematic under one evidentiary system are viable under another.

Above all, they should watch what subsequently happens in humans.

This is also where an early-mover advantage could emerge. Sponsors accumulating validated human-relevant evidence before it becomes standard practice may improve candidate selection and capital efficiency while simultaneously creating proprietary data sets that competitors cannot quickly reproduce. Investors capable of distinguishing regulatory acceptance from genuine predictive performance may likewise recognize changes in asset quality before those changes appear in conventional clinical or financial metrics.

The critical inflection point will come when particular NAMs cease to be viewed primarily as alternatives and begin to be viewed as better predictors for defined decisions. If that happens, adoption can move from experimentation to infrastructure, and the investment question changes from “Will regulators accept this?” to something potentially much more valuable, like “Who owns the best evidence?”

History Is A Guide, Not A Forecast

The historical analogy has limits. There is no current mandate for a DESI-like retrospective NAM review. Evidence supporting approved medicines does not become invalid because a new organoid, microphysiological system, or computational model becomes available. NAMs vary enormously in maturity and reliability, and “human-relevant” should never become shorthand for “correct.” Every model has limitations.

Nor was DESI simply a technological substitution. It followed a statutory change requiring substantial evidence of effectiveness. Today’s transition is occurring through a more complicated combination of scientific advancement, legislative change, regulatory modernization, and accumulating evidence about the strengths and limitations of different models. But the deeper historical lesson does not require the two periods to be identical.

DESI demonstrates that evidentiary systems have consequences. FDA’s retrospective review ultimately encompassed more than 3,000 products and more than 16,000 therapeutic claims. By 1984, final action had been completed on 3,443 products.⁸ When the standard governing what medicine claims to know changes substantially, the effects can propagate beyond the next experimental drug, altering therapeutic claims, regulatory decisions, medical practice, and ultimately the economic value attached to pharmaceutical products.

There is also a striking contemporary echo. In June 2026, FDA issued revised draft guidance on demonstrating substantial evidence of effectiveness, more than six decades after Congress made that concept central to drug approval.⁹ At the same time, the agency is constructing frameworks for NAMs and other modern approaches to generating and evaluating evidence.^1–3 This is not another 1962, but it is an evidentiary transition worth taking seriously.

The Next Evidentiary Standard

The evolution of drug regulation can be understood as a progression of increasingly demanding questions. The 1938 framework placed safety at the center of federal premarket review. The 1962 amendments added effectiveness backed by substantial evidence. Today’s emerging methods may allow more granular questions to be asked: In whom does a drug work? Through what biology? At what dose? With which human-specific risks? Can responders and nonresponders be distinguished? Can toxicity be predicted earlier and with greater human relevance?

Each transition increases the resolution of the evidentiary system, and increased resolution can make previously invisible distinctions visible. There is no reason to assume history will repeat itself in the same form. But there is equally little reason to assume that major improvements in our ability to interrogate human biology will affect only medicines that have yet to be invented.

If better evidence is eventually good enough to determine the fate of tomorrow’s drug, we will ultimately have to confront a more uncomfortable question: what does that evidence tell us about yesterday’s drug? For Wall Street, the real question is what happens when the evidentiary standard changes — and what tomorrow’s evidence reveals about yesterday’s drugs.

References

1. U.S. Food and Drug Administration, Center for Drug Evaluation and Research. General Considerations for the Use of New Approach Methodologies in Drug Development: Draft Guidance for Industry. March 2026; issued March 18, 2026. Docket No. FDA-2025-D-6131.
2. U.S. Food and Drug Administration. FDA Achieves Year 1 Goals in Reducing Animal Testing in Drug Development. FDA Press Announcement. April 20, 2026.
3. U.S. Food and Drug Administration, Oncology Center of Excellence. Oncology Pharmaceuticals: Streamlined Nonclinical Safety Studies for Biologics and Conjugated Products: Draft Guidance. May 2026. Docket No. FDA-2026-D-2839.
4. U.S. Food and Drug Administration. Table of Pharmacogenetic Associations. FDA Precision Medicine. Accessed September 3, 2026.
5. U.S. Food and Drug Administration. Accelerated Approval Program. FDA. Accessed September 3, 2026.
6. U.S. Food and Drug Administration, Oncology Center of Excellence. Verified Clinical Benefit: Cancer Accelerated Approvals. FDA. Accessed September 3, 2026.
7. U.S. Food and Drug Administration, Oncology Center of Excellence. Withdrawn: Cancer Accelerated Approvals. FDA. Accessed September 3, 2026.
8. U.S. Food and Drug Administration. A Brief History of the Center for Drug Evaluation and Research. FDA History.
9. U.S. Food and Drug Administration, Center for Drug Evaluation and Research, Center for Biologics Evaluation and Research, and Oncology Center of Excellence. Demonstrating Substantial Evidence of Effectiveness for Human Drug and Biological Products: Draft Guidance for Industry. June 2026; posted June 24, 2026. Docket No. FDA-2019-D-4964.

About The Author

Zaher Nahle, Ph.D., MPA, is a science policy expert, biomedical scientist, and senior advisor whose work has helped shape reforms at the U.S. Food and Drug Administration and National Institutes of Health. He serves as senior scientific advisor to the Center for a Humane Economy and is the founder of The Ivyctory Group, a market research and life sciences advisory firm. Nahle has held executive leadership positions at biomedical and medical research organizations, including as CEO, chief scientific officer, and vice president for research. Earlier in his career, he led independent research programs and served as a founding investigator of an NIH-funded data management and coordinating center. He is published in leading scientific journals, including Nature, and received competitive grants and distinctions from organizations, such as the American Heart Association, Qatar Foundation, and the U.S. Department of Defense, as well as an American Cancer Society Scholar Award designation. He earned a Ph.D. in physiology and biophysics through the Stony Brook University–Cold Spring Harbor Laboratory joint program, a certificate in public policy and management from Harvard Kennedy School, and an MPA from Harvard University, where he was a Mason Fellow.