Why Minoxidil Doesn’t Work for Everyone: Sulfotransferase, the Predictive Test, and What 2024–2025 Evidence Shows

The client has been on 5% topical minoxidil twice a day for nine months. She is the kind of patient who actually does the protocol. Her bottle is empty when she comes back in. Her photographs at month nine look exactly like her photographs at the consult. She is frustrated. Her dermatologist is frustrated. And there is a non-trivial chance that she is one of the half of all topical minoxidil users whose follicles cannot finish the chemistry that turns the molecule into a hair growth drug.

Minoxidil is a prodrug. That is not an opinion. It is the most clinically important pharmacologic fact about the most-prescribed hair loss treatment in dermatology, and it is the fact most rarely communicated at the chair. Topical minoxidil itself does almost nothing for follicles. The active molecule is minoxidil sulfate, produced inside the outer root sheath by an enzyme called sulfotransferase 1A1 (SULT1A1). If the follicle’s SULT1A1 activity is low, the prodrug arrives but the active drug is never made. The bottle empties. The hair does not change.

What the 2014–2025 peer-reviewed literature now shows is that follicular sulfotransferase activity is measurable, the assay is predictive, and there are at least two evidence-supported strategies for converting predicted non-responders into responders. There is also a 2024 paper on oral minoxidil that complicates the entire picture in a clinically interesting way. Here is what trichologists actually need to understand.

The Mechanism: Why Minoxidil Has to Be Activated

Topical minoxidil is a piperidinopyrimidine that, in its parent form, is a relatively mild vasodilator and an inefficient potassium channel opener. The active hair growth molecule is the sulfated metabolite, minoxidil sulfate, which is the species responsible for opening the ATP-sensitive potassium channels (K_ATP) on dermal papilla cells, prolonging anagen, and producing the hair shaft thickening and density increase that the marketing photographs document. The enzymatic step that produces minoxidil sulfate is sulfotransferase-mediated. The enzyme is expressed in the outer root sheath. The activity of that enzyme varies dramatically between individuals.

The follicle is, in effect, a small chemical factory that has been asked to manufacture its own drug. Some factories run at full output. Others run at a fraction. The same 5% solution that produces dramatic regrowth in one client produces nothing in the next, and the difference often is not adherence, formulation, or stage of disease. It is enzymology.

The Foundational Predictive Study

Roberts and colleagues established the clinical framework. Plucked hair follicles were assayed for sulfotransferase enzymatic activity using a colorimetric assessment, and the result was tested against subsequent response to topical 5% minoxidil over treatment. Using a 0.4 absorbance-unit cutoff to define low follicular SULT1A1 activity, the assay identified responders to topical minoxidil with a sensitivity of approximately 93% and a specificity of 83%. That is a level of predictive performance most clinical biomarkers do not achieve.

The importance of the result is in what it lets you do at the start of treatment instead of at month six. A test that flags a likely non-responder before the client buys the first bottle changes the conversation from “try this for six months and see” to “here is a measurement, here is what it means, and here is what we are going to do about it.” That is the difference between guessing and practicing.

The 2022 Review and the 2024 Confirmation

The 2022 review consolidated more than a decade of work from the Goren group and independent investigators across Brazil, Italy, and the United States. The conclusion is consistent: SULT1A1 follicular activity correlates with topical minoxidil response in androgenetic alopecia, the assay is reproducible across labs, and the enzymology hypothesis explains the response variability that decades of dosing studies and formulation tweaks could not.

A 2024 study published in QJM confirmed the relationship in a fresh cohort. Less than half of the patients tested showed a meaningful response to topical minoxidil at follow-up, and the responder group had baseline sulfotransferase activity averaging 0.78 absorbance units versus 0.31 in non-responders, a statistically significant separation (P < 0.001). The authors concluded that the sulfotransferase enzyme assay has clinical utility as a pre-treatment screen.

Tretinoin: The Cheap, Available, Evidence-Based Booster

This is the paper that turned the predictive assay into an actionable treatment lever. Twenty patients (ten male, ten female) with androgenetic alopecia were treated with topical tretinoin 0.1% for five days. Follicular SULT1A1 activity was measured before and after. Tretinoin increased SULT1A1 activity 1.8-fold in 75% of subjects, and 43% of patients initially predicted to be minoxidil non-responders converted to predicted responders after the brief priming course.

The implication is direct. Topical tretinoin is inexpensive, widely available by prescription, and routinely tolerated. Five days of priming is not a major ask. For a substantial fraction of clients who have been told minoxidil won’t work, the right next step is not abandoning the molecule. It is upregulating the enzyme that finishes the molecule’s job. Practitioners working alongside prescribing clinicians have a real conversation here: the patient may not be a non-responder by biology, only a non-responder by enzyme starting position.

The SULT1A1 Booster Trials

A 2024 prospective single-center study enrolled 101 patients who were already documented topical minoxidil non-responders and combined a SULT1A1 enzyme booster solution with their existing 5% minoxidil regimen for 90 days. The combination produced a meaningful increase in response rates in the previously non-responding cohort, building on an earlier randomized placebo-controlled trial published in 2021 in the Journal of Cosmetic Dermatology that had reported approximately 75% regrowth in patients using the booster plus daily minoxidil for 60 days versus roughly 33% on placebo adjuvant.

The 2021 trial established proof of concept. The 2024 trial extended it to a real-world non-responder cohort. The mechanism is the same in both: the booster does not replace minoxidil. It restores the enzymatic step the follicle was failing to perform. This is a design philosophy worth pausing on. Most failed hair loss protocols try to swap the molecule. The sulfotransferase literature suggests the better question is often whether the existing molecule is actually being activated.

The 2020 Genetic Variant Study

The Ramos paper extended the conversation from enzyme activity to underlying genetics. Specific SULT1A1 polymorphisms were associated with response to oral minoxidil in female pattern hair loss. The G allele at the studied locus was associated with response in roughly 84% of carriers, while only about 16% of A-allele carriers responded. The result frames the response variability as partly heritable and explains why the same dose, applied identically, lands so differently across patients.

The clinical translation is that what looks like adherence variability or formulation variability in the chair is sometimes population-level pharmacogenetics. There are women whose biology will produce minoxidil sulfate efficiently and women whose biology will not, and the difference can be read in their hair follicles or, ultimately, in their genome. The fact that the assay is now commercially available as a six-strand plucked-hair test is what moves this from an academic finding to a tool a practitioner can actually use.

The 2024 Oral Minoxidil Paradox

This is the paper that scrambles the rule. Jimenez-Cauhe and colleagues at the Vallé Hair Clinic group in Madrid measured plucked-follicle SULT1A1 activity in patients on low-dose oral minoxidil and tracked clinical response. The relationship was the opposite of what the topical literature predicts. Patients who responded best to oral minoxidil had lower follicular SULT1A1 activity, not higher. The negative association was statistically significant.

The proposed explanation is biology, not contradiction. Oral minoxidil undergoes extensive first-pass sulfonation in the liver before circulating systemically. By the time the drug reaches the scalp, it has already been substantially activated upstream, and minoxidil sulfate (and unmetabolized minoxidil that the follicle can still convert) reaches the hair follicle through circulation. In that scenario, follicles with very high local SULT1A1 may metabolize the drug too quickly, narrowing the time the active species sits at the K_ATP channel. In follicles with lower SULT1A1, the active drug persists longer in its therapeutic window. The pharmacology inverts.

This is a small study and a hypothesis-generating one, not a guideline-changer. But it has two important practical implications for trichologists. First, the plucked-hair sulfotransferase test is not interchangeable across topical and oral routes. A high SULT1A1 result that predicts a topical responder may also predict an oral non-responder, or at least a different therapeutic window. Second, when a client switches from topical to oral, the same biology that failed to activate the topical may actually favor the oral. That is a real conversation, and it cuts against the assumption that a non-responder is a non-responder.

What the Caveats Look Like

The honest version of the story includes the limits. The plucked-hair assay was developed and published largely by a single research group, although independent replication has now appeared in Brazilian, Egyptian, and European cohorts. Sample sizes in most trials are modest. Multiple sulfotransferase isoforms beyond SULT1A1 contribute to minoxidil sulfation, and the colorimetric assay measures aggregate activity rather than the specific isoform. The commercial test sold to consumers as a minoxidil response assay is not FDA-cleared as a diagnostic, and its performance characteristics are based on the published literature rather than independent regulatory review.

There is also confounding from minoxidil’s downstream pharmacology that the enzymatic story does not fully capture. Vascular response, K_ATP channel density, prostaglandin pathway activity, and the contribution of microneedling-mediated upregulation of Wnt/β-catenin all sit on top of the sulfotransferase step. A patient can have adequate follicular SULT1A1 activity and still respond poorly because of any of those downstream factors. Conversely, the 2024 oral minoxidil paradox suggests the enzymatic story may not even point in the same direction depending on route.

None of that invalidates the predictive utility of the assay for topical minoxidil. It contextualizes it. The test is best understood as a screen that materially shifts pre-test probability, not a deterministic readout of who will and will not respond.

What This Means at the Chair

The clinical conversation changes once you internalize the prodrug fact. A few practical translations:

Set expectations differently at the start of topical therapy. Tell the client at consult that topical minoxidil is a prodrug, that the follicle has to do a final activation step, and that approximately half the population produces enough of the activating enzyme to respond. The client who hears that at month one is a different client at month six than the one who heard a generic “just give it time.”

Treat a true six-month non-response as a branching point, not a dead end. The two evidence-supported levers are tretinoin priming and a SULT1A1 booster adjuvant. Both target the enzyme. Both are layered on top of, not in place of, minoxidil. If the practice does not have access to a prescribing clinician, the conversation is still the right one to have so the client can carry it to dermatology.

Reframe a topical non-response as a possible oral candidate. The 2024 Jimenez-Cauhe data suggests that low follicular SULT1A1 activity may actually favor oral minoxidil. This is not yet enough evidence to write a guideline on, but it is more than enough to keep oral minoxidil on the table after a documented topical failure rather than abandoning the molecule entirely.

Take the client’s genetics seriously without overclaiming. The Ramos pharmacogenetic data is real but small. The right framing for the client is that minoxidil response is partly heritable, that the variability they may have noticed in their family is consistent with the data, and that a measurable assay is available. The wrong framing is to promise that a positive genetic result will guarantee a clinical outcome.

Stop selling minoxidil as a generic “put it on the scalp” protocol. Once the prodrug fact is on the table, the conversation shifts from “use this consistently” to “use this consistently and we will measure whether the follicle is finishing the chemistry.” That is the conversation that distinguishes a serious trichology practice from a generic referral.