Treating the Canvas, Not the Spot: Why AK Therapy Went Field-Wide

The model most people arrive with

A patient notices a rough, scaly patch on the scalp or the temple that will not moisturise away. A clinician confirms it is an actinic keratosis, applies liquid nitrogen for a few seconds, and the patch crusts and falls off over the following fortnight. Problem solved.

Except that eight months later there are three more, in the same general region, and the cycle repeats. Eventually somebody proposes a cream to be used across the entire forehead and scalp for several weeks, and the patient — who came in about one rough spot — wonders why the treatment has expanded so dramatically.

It has expanded because the mental model was wrong from the start. The visible actinic keratosis was never the problem. It was the part of the problem that reached the surface.

Field cancerisation, in plain terms

Decades of ultraviolet exposure do not damage skin one discrete spot at a time. They deposit mutations across an entire exposed region — a bald scalp, a forehead, the tops of the ears, a forearm — producing a mosaic of keratinocyte clones carrying varying degrees of genetic damage. That region is what dermatology means by a field.

Most of those altered clones sit below the threshold of visibility. They are not raised, not scaly, not palpable, and not findable by the most careful examination. A visible actinic keratosis is simply a clone that grew far enough to become detectable — the tip of something considerably larger.

Once that is clear, the recurrence pattern explains itself. Cryotherapy destroys the lesion it is aimed at. It does nothing at all to the surrounding field, which continues developing on its own schedule and produces the next visible lesion a few months later a centimetre away. Patients experience this as treatment failure or as bad luck. It is neither: it is a spot-directed treatment doing exactly what it can do, applied to a problem that is not spot-shaped.

This matters beyond cosmetics, because a proportion of actinic keratoses progress to cutaneous squamous cell carcinoma. Treating the field is not tidiness. It is reducing the number of clones available to make that transition. The ↗actinic keratosis pathway sets out how that escalation is actually managed.

When freezing is enough, and when it is not

Cryotherapy has not been superseded, and framing this as old-versus-new would be wrong.

Spot treatment is appropriate when there are a small number of discrete, well-defined lesions on skin that is otherwise not diffusely damaged. It is fast, it is done in the appointment, it needs no compliance, and its downtime is a scab.

Field therapy is indicated when the pattern says field: multiple lesions in a continuous area, a scalp or forehead that feels generally rough rather than focally rough, lesions that keep reappearing in the same territory, or a history of squamous cell carcinoma in that region.

The tell is the pattern, not the count. Three lesions scattered across otherwise clear skin is a different situation from three lesions clustered in one sun-battered zone, even though the number is identical. And there is a diagnostic step before either: rough, scaly lesions in sun-damaged skin are not all actinic keratoses, and distinguishing them matters — ↗telling an AK from a seborrhoeic keratosis changes the treatment entirely, and the two frequently coexist on the same scalp.

The four agents, and the trial that ranked them

For years the choice between field treatments was made on habit and availability, because no head-to-head evidence existed. In 2019 the New England Journal of Medicine published the trial that settled a great deal of it.

Dutch investigators enrolled 624 patients with five or more actinic keratoses on the head within one continuous area of 25 to 100 cm², and randomised them to 5% fluorouracil cream, 5% imiquimod cream, methyl aminolevulinate photodynamic therapy, or 0.015% ingenol mebutate gel. The primary outcome was a reduction of at least 75% in lesion count, sustained to 12 months after treatment ended.

The cumulative probability of remaining free from treatment failure at twelve months:

Fluorouracil 74.7%. Imiquimod 53.9%. Photodynamic therapy 37.7%. Ingenol mebutate 28.9%.

Against fluorouracil, the hazard ratio for treatment failure was 2.03 with imiquimod, 2.73 with PDT and 3.33 with ingenol mebutate — all statistically significant. Old, cheap, generic 5-fluorouracil beat everything, including the in-clinic procedure.

Three honest qualifications. This was head-region disease in a Dutch population, and results do not automatically transfer to a forearm or a different skin type. PDT protocols vary considerably by photosensitiser, light source and incubation time, and a different protocol may perform differently from the one tested. And efficacy is not the only axis: the most effective agent here is also among the most unpleasant to use, and a course completed beats a superior course abandoned in week two.

Then there is what happened to the fourth arm. Ingenol mebutate performed worst in this trial, and separately, in 2020, the European Medicines Agency suspended its marketing authorisation as a precaution after a three-year study in 484 patients found skin malignancy in 3.3% of those treated with ingenol mebutate against 0.4% on imiquimod; an eight-week vehicle-controlled trial in 1,262 patients also showed more skin tumours on the drug than on vehicle. The agency subsequently concluded the risks outweighed the benefits and the product was withdrawn from the EU market.

A treatment for precancerous lesions associated with more cancer, licensed and widely prescribed for years before the signal emerged. It is worth carrying that as a general lesson rather than a footnote: being on the market is not the same as being right, and the correction came from post-marketing surveillance rather than from the approval process.

Tirbanibulin, a newer topical, is the current tolerability play — a five-day course rather than weeks, with a substantially lighter reaction. The trade-off is a shorter evidence record and far less head-to-head data than the agents above. It is a reasonable option, particularly for someone who will not complete a longer course, and it should be chosen with that trade-off stated rather than presented as an equivalent.

The reaction is the treatment, and nobody prepares you for it

Everything above is background to this, because the commonest reason field therapy fails has nothing to do with the drug: the patient stops it.

Fluorouracil applied across a field produces, over roughly two to four weeks depending on the regimen and site, a sequence that looks genuinely alarming: redness, then inflammation, then erosion, crusting, weeping and real discomfort. Faces look burned. Scalps look infected. People cancel plans.

Here is the concept that reframes all of it: the reaction maps the field. The drug produces its response where abnormal cells are. Areas you did not know were damaged light up; areas that are healthy stay comparatively quiet. Patients frequently report that the reaction appeared in places they had no lesions — and that is not a side effect, that is the entire reason for using a field agent instead of liquid nitrogen. You are seeing the subclinical disease become visible on its way out.

So the two sentences that should be said at every prescription, and often are not: you will look considerably worse before you look better, and the worst point is around weeks two to three. Someone who understands that finishes the course. Someone who does not stops at day ten convinced something has gone wrong, having done enough to be uncomfortable and not enough to clear the field.

Practical preparation. Photograph the area before you start, because you will lose your baseline. Plan the calendar — this has among the highest social downtime of anything in dermatology, and it is typically dispensed with a leaflet rather than a conversation, so schedule it into a quiet few weeks rather than before an event. Keep the rest of the routine bland: ↗acids and vitamin C during a fluorouracil course is a question with a clear answer, and the answer is not now. Use plain emollients — Vanicream Moisturizing Skin Cream for general dryness, CeraVe Healing Ointment or Avène Cicalfate+ Restorative Protective Cream where skin is genuinely eroded and crusted.

And know the line between expected and wrong. Expected: redness, erosion, crusting, soreness, itch, looking dreadful. Call the prescriber for spreading pain out of proportion, purulent discharge, fever, or anything that suggests infection rather than inflammation — that is a genuine complication and it is managed, not endured. Do not simply stop the course without speaking to someone; there is usually a middle option between continuing exactly as instructed and abandoning it.

After clearance, sunscreen stops being hygiene

Field therapy clears the field it treats. It does nothing about the ultraviolet exposure that produced it, and the field will rebuild.

The evidence for what to do about that is unusually good and unusually old. In 1993 the New England Journal of Medicine published a randomised controlled trial in 588 Australians aged 40 and over, comparing daily use of a broad-spectrum sunscreen against the identical base cream without the filters, applied to the head, neck, forearms and hands across a single summer.

The base-cream group gained a mean of 1.0 actinic keratosis per person. The sunscreen group lost 0.6. The sunscreen group had fewer new lesions, with a rate ratio of 0.62, and more remissions of existing ones, with an odds ratio of 1.53. And there was a dose-response relationship: the amount of cream people actually used predicted both how few new lesions they developed and how many old ones regressed.

Read that carefully, because it is stronger than the usual sunscreen claim. Existing lesions went away in the sunscreen group. Over one summer. That is not prevention in the abstract-future sense; it is a measurable therapeutic effect on disease already present.

Which is why 'keep using sunscreen' after field therapy is not a closing pleasantry. It is the maintenance phase of the treatment, and the dose-response finding means quantity and consistency are the active variables — the sun safety tool covers how much and how often, which is where most people's sunscreen use quietly fails. Hats and shade do the same job with better compliance on a bald scalp, which is the highest-risk site in this entire article.

The other half of maintenance is surveillance. An actinic keratosis is a marker that the field exists and that squamous cell carcinoma risk is elevated, so the follow-up interval matters more than it does for someone without that history — the skin cancer risk tool is a reasonable way to see where you sit, and the derm visit prep tool helps make a short review appointment count. Bring a list of what is new, what has changed, and what has not healed. That last category is the one worth escalating.