Keloids by location: why where it sits predicts how it behaves

Two people are told the same statistic on the same afternoon. One has a keloid on an earlobe from a piercing. The other has one across the upper chest that started from acne. Both are told that keloids commonly recur after excision, and both are given a number that comes from the general literature.

That number is close to useless for both of them, in opposite directions, and the reason is that keloids are not one problem occurring in different places. They are a response to a wound that behaves very differently depending on what surrounds that wound.

The original element in this piece is a site by site behavior comparison assembled from studies that each look at a single anatomical region, with the assembly method stated openly, because the site specific literature exists and nobody puts it in one place for patients. The general reviews average across every location, which produces a figure that describes no actual patient. The single site studies are far more informative and are scattered across different journals and surgical specialties.

The mechanical reason location matters. A keloid is scar tissue that has grown beyond the boundary of the original wound and does not regress. The best supported explanation for why some wounds do this and others do not centers substantially on mechanical tension. Tissue that is repeatedly stretched during ordinary movement keeps signalling the wound to keep producing collagen, and the signal does not shut off at the point where a normal scar would mature and settle.

Now think about where on the body tension is highest. The front of the chest over the sternum is stretched with every deep breath and every shoulder movement, in multiple directions. The shoulder and upper back move constantly through large ranges. The jaw and the front of the neck stretch with speech, chewing and head turning.

And then think about the earlobe. It is a small flap of soft tissue with almost no muscle attachment, essentially no resting tension, and very little movement. It is the lowest tension site on the body that people routinely wound with a needle.

That single variable, tension, predicts most of what follows, and it explains the pattern of where keloids appear in the first place. Chest, shoulders, upper back, jawline and earlobes are the common locations, and the earlobe is the outlier in that list: it is common because it is pierced constantly, not because it is mechanically hostile. The other sites are common for the opposite reason.

What the site specific literature shows. Read the studies that stick to one region and a consistent picture emerges.

Ear keloid work is comparatively optimistic. Surgeons have developed and reported region specific techniques, such as the evaluation of stellate excision for keloids in the upper third of the ear, and the fact that refinements of excision technique are being studied and reported for this site at all reflects a setting where surgery plus a reasonable adjuvant is a realistic path rather than a last resort. Low tension tissue lets the wound behave more like a normal wound.

Chest keloid work is the opposite. A systematic review and meta analysis specifically of chest keloids treated by surgical excision and adjuvant radiotherapy exists as its own body of literature precisely because the chest was recognized as a distinct and harder problem. Studies pooling all body sites were producing figures that did not match what surgeons saw on sternums. That is an unusual and telling thing: the field split the anatomy out because the average was misleading.

The broader treatment evidence points the same way. A network meta analysis of keloid excision and adjuvant treatments confirms what every keloid surgeon says, which is that excision alone is the wrong operation almost everywhere. And the work on timing of adjuvant radiotherapy after excision shows how tightly the outcome is bound to the protocol around the operation rather than to the operation itself. In high tension sites, the adjuvant is not an optional extra. It is the treatment, and the surgery is the setup.

How to use this in your own consultation. Three practical consequences.

First, ask for a site specific number, and treat a general one as a warning sign. If you have a chest keloid and you are quoted a recurrence figure drawn from a mixed series dominated by earlobes, you are being given a misleadingly good number. Ask directly: what is your recurrence rate for this location, with this plan.

Second, understand why the plan differs by site. When you are offered excision plus radiotherapy for a sternal keloid but excision plus steroid injections for an earlobe, that is not inconsistency and it is not upselling. It is the tension environment driving the aggressiveness of the adjuvant.

Third, weigh prevention by site before you create a wound. The tension argument applies to elective wounds too. A piercing in a low tension site in a keloid prone person is a materially different risk from a surgical incision across the sternum, which is the reasoning behind preventing keloids before surgery and behind the caution around tattoos and piercings in keloid prone skin. The question is not only whether you form keloids. It is where you are about to place the wound.

How this comparison was assembled, and its limits. It is drawn from studies that each restrict themselves to a single anatomical region or that explicitly analyze location, read against the general treatment meta analyses. It is a synthesis, not a pooled statistical analysis, and it is not a substitute for one.

What the studies do not tell you. There is no single published study that compares recurrence across all the common keloid sites in one population, with one protocol and one follow up length. Without that, direct numerical comparison between sites is not strictly valid: the ear studies and the chest studies use different techniques, different adjuvant protocols and different follow up durations, and follow up duration matters enormously in a condition where recurrence can appear years later. Some of the difference between reported figures is genuine biology and some is study design, and nobody can currently say how much of each.

Definitions vary too. What counts as recurrence differs between papers, with some counting any regrowth and others only regrowth beyond the original margin, which alone can shift a reported rate substantially.

The takeaway. Location is not a detail on your chart. It is the strongest available predictor of how your keloid will behave, it explains why two people get completely different treatment plans for what looks like the same diagnosis, and it is the right thing to ask about when the recurrence conversation starts. A number that was not measured on your body part is not a number about you.