Cumulative Dose and Relapse: The Isotretinoin Number That Predicts Round Two

The mirror clears first

Most people finishing isotretinoin describe the same experience. Somewhere around month three the skin goes quiet — not improving, done — and the remaining months feel like an administrative formality attached to a problem that has already been solved. This is the point at which courses get abandoned, shortened by mutual agreement, or quietly tapered because the side effects no longer feel proportionate to a face that looks fine.

It is also the point at which the treatment is least finished, because isotretinoin isn't dosed in months. It's dosed in total milligrams per kilogram of body weight accumulated across the whole course, and the daily dose and the duration are simply two ways of arriving at that total. A person on a lower daily dose needs proportionally longer. Two people who both took capsules for five months may have received substantially different treatments.

The reason the endpoint isn't visual comes down to what the drug does to sebaceous glands. Isotretinoin shrinks them and suppresses sebum output, and both the depth of that suppression and how long it persists after the drug is withdrawn scale with exposure. Clearance of visible lesions happens once sebum falls below the level that sustains them — which occurs well before the glands have been suppressed durably enough to stay that way. Clear skin is a leading indicator. Stopping when it appears is stopping on the signal rather than the result, and it is the single most common way a course under-delivers.

Where 120 to 150 came from, and how solid it is

Ask about the target and you'll be given a range: a cumulative dose of 120 to 150 mg/kg, quoted with the confidence of a physical constant. It appears in textbooks, in patient leaflets, in the arithmetic that prescribers do at the first appointment. It deserves considerably more scepticism than it gets.

A systematic review published in the Journal of Cutaneous Medicine and Surgery in 2016 went looking for the evidence underneath it and reported an uncomfortable finding: the threshold derives from historical parameters of treatment duration rather than from studies designed to test it, and only two studies had ever explicitly evaluated cumulative dosing of 120 to 150 mg/kg for inducing remission — both graded low. The reviewers also noted that definitions of clearance, relapse and treatment endpoint varied so widely across the literature that comparing outcomes between studies is genuinely difficult. Their conclusion was that the optimal cumulative dose appears to vary with disease severity, which is a considerably more useful statement than a fixed range and a considerably less quotable one.

This matters because the number is doing real work in clinics. It determines when a course is declared complete, and it is treated as a ceiling as often as a floor. The current American Academy of Dermatology acne guideline, published in 2024, strongly recommends oral isotretinoin for acne that is severe, causing psychosocial burden or scarring, or failing standard oral or topical therapy — a recommendation about when to use the drug, arrived at through a formal evidence assessment. The cumulative-dose convention has never had that kind of scrutiny applied to it and has been carried forward largely by inertia.

The dose-response signal is real, and it points upward

The most useful study on this question is a prospective, observational one from the University of North Carolina published in JAMA Dermatology in 2013. It enrolled 180 patients with acne resistant to other treatments, followed 116 of them at twelve months, and split them by cumulative dose at a threshold well above the conventional target: under 220 mg/kg versus at or above it.

The results are the clearest dose-response data available. Overall, about a third of patients relapsed at twelve months. In the lower-dose group the relapse rate was 47.4 percent; in the higher-dose group it was 26.9 percent, a statistically significant difference. Nearly all patients, in both groups, reported that their acne was improved.

The cost side is equally clear and worth taking seriously. Retinoid dermatitis was significantly more common at higher cumulative doses — 53.8 percent versus 31.6 percent — and essentially every patient in both groups developed cheilitis and xerosis during treatment. No other adverse effect differed significantly between the groups.

So the trade is legible: several additional months of a drier, more fragile face in exchange for a materially lower chance of needing to treat acne again a year later. Whether that's worth it is a genuine preference question rather than a clinical one, and it should be asked out loud rather than resolved by default. What it is not is a free choice — stopping early has a price, and the price is quantified.

The near-universality of the side effects also explains why the supporting routine during a course is not optional. Cheilitis affects effectively everyone, and it responds to occlusion rather than treatment; an ointment-based lip repair like Aquaphor's Lip Repair applied far more often than feels reasonable is the standard answer. For the retinoid dermatitis around the nose and mouth, a ↗panthenol-led repair balm such as Cicaplast Baume B5+ is what most people end up using, and body xerosis — which is what actually drives people to ask about dose reduction — is best pre-empted with a heavy, ceramide- or oat-based body cream like Lipikar Eczema Cream from the first week rather than the first crack.

What 'relapse' actually means in these studies

Before those percentages induce despair, read the definition. In the JAMA Dermatology study, relapse was defined as treatment with a prescription topical or oral acne medication after the course. Retrial — a second course of isotretinoin — was counted separately, and it occurred in under two percent of patients.

That gap is the most reassuring number in the entire literature and almost nobody quotes it. A third of patients "relapsed" in the technical sense; a very small fraction needed the drug again. What the majority experienced was acne that returned in a milder, topically manageable form — which is a completely different outcome from the disease coming back as it was, and a completely different emotional experience from what the word implies.

It also reframes the point of the cumulative dose. The realistic goal isn't a permanently acne-proof face; it's converting severe, scarring, treatment-resistant acne into something a topical retinoid can hold. Judged against that goal, the outcomes are very good even in the lower-dose group. Judged against "never again," they will always disappoint, and the disappointment is a measurement error rather than a treatment failure.

Low-dose regimens, and who they're actually for

Low daily dosing and intermittent or pulsed regimens have become popular, particularly for mild-to-moderate acne that has failed conventional treatment, and particularly among adults who want the benefit without a face that peels through a winter.

The evidence is mixed and honestly reported. A systematic review and meta-analysis in Dermatologic Therapy in 2022 pooled the available comparisons in adults with mild-to-moderate acne and found that conventional dosing improved the odds of prolonged remission relative to low daily dosing, while low daily dosing outperformed pulsed regimens on achieving a response. The authors graded the overall quality of evidence as low to moderate — a caveat that applies to the conclusion in both directions, not just the one you'd prefer.

The defensible reading is that low-dose isotretinoin is a legitimate tool for a specific job: moderate acne where tolerability is the binding constraint and where a slightly higher chance of recurrence is an acceptable trade. It is not a lower-risk route to the same endpoint. If the goal is durable remission after severe nodulocystic acne, the accumulated evidence points toward reaching a proper total. The same logic explains why ↗low-dose isotretinoin used for rosacea is a different proposition entirely — different target, different endpoint, no remission threshold to hit.

One practical instruction follows from all of this, and it takes ten seconds. Ask your prescriber what your cumulative total is, in mg/kg, at each appointment — and what total you're aiming for. It is a number they are already tracking and rarely volunteer. Knowing it converts "how much longer?" from a mood into arithmetic, and our Derm Visit Prep tool is a good place to park the question so it doesn't get lost behind the blood test conversation.

The routine that starts the day after the last capsule

The end of a course is the moment of maximum complacency and the moment relapse risk begins. Sebum output climbs back over the following months, and whether that translates into lesions depends heavily on what's applied to the skin in the meantime.

Maintenance with a topical retinoid is the standard, evidence-supported move, and ↗adapalene is the usual choice because it is the best tolerated of the retinoids on skin that has spent months being systematically dried out. Starting something like PanOxyl's adapalene gel a few weeks after the last capsule — once the barrier has recovered enough to take it — keeps follicular keratinisation normalised through exactly the window where microcomedones would otherwise start re-forming. Two or three nights a week is usually enough; this is prophylaxis, not treatment.

Post-inflammatory marks are the other job. They fade slowly on their own and faster with help, and ↗azelaic acid is a reasonable, non-irritating option on recently-recovered skin — The Ordinary's azelaic suspension is the accessible version. Sun protection matters here more than the average person expects, because months of skin that has been thinned and inflamed does not need UV added to the pigment equation.

Measurement is the last piece, and it's worth doing properly because the whole failure mode of this drug is a course judged by impression. Photograph in consistent lighting from the start of the course through the year afterwards; the Photo Tracker exists for that comparison, and the Acne Severity check gives you a lesion count rather than a feeling, which is what a prescriber can actually act on. ↗The acne pathway sets out where isotretinoin sits relative to everything before and after it.

And the summary opinion, plainly: if you're on a course and the skin has cleared, that is not evidence you're finished. It's evidence the drug is working, which is the reason to keep taking it — and the number that decides whether you're back here in a year is one your prescriber can tell you this week.