Keloid Evidence

Recurrence

How often do keloids come back?

Where the keloid is matters far more than most pages say. Below is every figure we could trace, split by body site and treatment. Each one shows how long patients were actually watched.

Why every number here has a time attached

Keloids usually come back one to two years after treatment. Studies that tracked the timing report typical waits of 16.7 and 23 months. Yet the average combination study only watches patients for 11 months. A study that reports a low rate at one year has missed about half the returns still to come. A figure without its time window is not roughly right — it is wrong.

Of the 14 figures on this page, 2 come from studies that stopped watching within 12 months and 8 never reported how long they watched at all. Both are marked.

The difference one centimetre makes

These three figures come from the same study, the same radiotherapy dose and the same minimum follow-up. The only thing that changes is where on the body the keloid was.

  • 5.7% Earlobe (lobule) 2 of 35 keloids came back everyone watched for at least 18 months
  • 38.5% Auricle excluding earlobe (helix / concha) 5 of 13 keloids came back everyone watched for at least 18 months
  • 39% Anterior chest 32 of 82 keloids came back everyone watched for at least 18 months

This is why ear keloids do well is misleading. It is true of the earlobe and false for the rest of the ear. The helix and concha sit barely a centimetre away — and recurred at nearly seven times the rate.

By body site

Site & treatmentCame backWatched for
Earlobe (lobule) Excision with adjuvant radiotherapy · 15 Gy / 3 fractions PMID 31840001 5.7% at least 18 months
Auricle excluding earlobe (helix / concha) Excision with adjuvant radiotherapy · 15 Gy / 3 fractions PMID 31840001 38.5% at least 18 months
Anterior chest Excision with adjuvant radiotherapy · 15 Gy / 3 fractions PMID 31840001 39% at least 18 months
Earlobe (lobule) Surgical excision · topical imiquimod post-excision The review's own verdict: very low certainty, not recommended as a treatment. PMID 31587416 5.4%95% CI 0-21.7 not reported
Trunk (general) Surgical excision · topical imiquimod post-excision The review pooled 'other areas, predominantly the trunk' — and its verdict: very low certainty, not recommended. PMID 31587416 76.8%95% CI 36.1-100 not reported
Earlobe (lobule) Surgical excision · core extirpation or excision + postoperative lobular compression + scar ointments PMID 28913244 10% not reported

Not broken down by site

These studies pooled every body site together. Given the spread above, treat a single number covering all sites as a rough average of very different situations.

TreatmentCame backWatched for
Surgical excision alone PMID 38463702 45–100% not reported
Excision with adjuvant radiotherapy brachytherapy (pooled) PMID 41401628 14% not reported
Excision with adjuvant radiotherapy X-ray (pooled) PMID 41401628 18% not reported
Excision with adjuvant radiotherapy electron beam (pooled) PMID 41401628 16% not reported
Intralesional 5-fluorouracil monotherapy PMID 38807454 16% average 27 weeks under-observed
Triamcinolone + 5-fluorouracil combination combination PMID 37432467 23.3% average 11 months under-observed
Shave-down surgery (KILE) 608 keloids A different operation from standard removal: the keloid is shaved down from inside its own edge. PMID 38463702 13%95% CI 9-16 average 19.2 months
Excision with adjuvant radiotherapy modern site-adjusted doses (2013 onward): 18 Gy chest/shoulder-blade/above-pubic, 8 Gy earlobe, 15 Gy other · 494 keloids PMID 31840001 9.3% not reported

Surgery on its own

Surgery on its own is the treatment most likely to end with the keloid coming back. Reviews report between 45% and 100%. The American Academy of Dermatology puts it at nearly 100%. That is why surgery is almost always paired with something else.

Evidence argues against Systematic review / meta-analysis Low certainty Primary sources not traced

We could not trace this range back to its original studies. So we publish it as "reported by reviews", not as a primary finding. We include it because pages that call surgery the permanent option have the risk backwards.

Reviews stating the range: PMID 38463702 · American Academy of Dermatology — Keloids: diagnosis and treatment

Keep reading

Do the popular keloid remedies actually work? · Keloid or piercing bump?

Where these figures come from