How far the evidence for XERF has come

The XERF reference page on this site read: “within the range we have confirmed, no published results of any clinical trial carried out with XERF have been confirmed”. At the time we wrote it, that was true. It is not true now.

It is not common for a clinic website to publish a piece overturning a sentence it wrote itself. But in this series we have set ourselves the rule of “stating the source together with its limits”, and that rule has to apply in the same way when the material turns in our favour. Leaving the old sentence standing after new material has appeared is also a statement at odds with the facts.

Let us be clear about the direction, though. This piece is not a piece saying “the evidence has arrived, so we are fine now”. It is a piece that copies across what the new material says and what it does not say, and by the end of it you will probably find it more cautious than you expected.

Read as numbers alone, that is not bad. What matters more, though, is which questions this study cannot answer.

One more thing to note. Medical Lasers, the journal that carried this paper, is indexed in the National Research Foundation of Korea's KCI and in KoreaMed, but not in PubMed/MEDLINE or SCIE. Whether a journal is indexed does not decide whether a paper is right or wrong. But the fact that it is not yet on the routes along which work is searched, cited and checked internationally is not something you can leave out when weighing the evidence.

There is nothing improper in this. It is normal for the manufacturer to carry out the early studies of a new medical device, and declaring the interest in the paper is normal handling too. But when you weigh the evidence, the next step is “has a researcher unconnected with that company reproduced it independently”, and XERF is not yet at that step.

The interesting one in substance is the simulation in Lasers Med Sci 2025. 2 MHz produced a broader and deeper thermal response within the fat layer, and 6.78 MHz heating localised along the fibrous septa. It is offered as the rationale for a design using two frequencies, but this study used only one type of tip (20 × 30 mm) — the effect of tip size is not addressed. That story is taken up separately in the next piece.

XERF passed the US FDA's 510(k) premarket notification on 11 August 2025 (submission number K251327). The “21 August” in Korean press coverage is the date of the company's announcement; the FDA decision date is 11 August.

But you have to know what kind of review a 510(k) is before you can read that fact accurately.

Two sentences copied straight out of the XERF 510(k) summary.

“No clinical tests were included as part of this submission.”
— that is, this submission contained no clinical trial.

“indicated for use in dermatologic and general surgical procedures for electrocoagulation and hemostasis”
— the indication cleared is electrocoagulation and hemostasis in dermatologic and general surgical procedures. Wrinkles · firmness · lifting are not in that sentence.

So what went through in the United States is “this device is as safe as an existing device” and not “this device tightens the face”. Those are two different sentences.

This part is not our opinion but the wording of the US Code of Federal Regulations (21 CFR 807.97). Here it is verbatim.

“…does not in any way denote official approval of the device. Any representation that creates an impression of official approval of a device because of complying with the premarket notification regulations is misleading and constitutes misbranding.”

— a 510(k) clearance does not in any way denote official approval of the device. A representation that creates the impression of official approval because the premarket notification rules were complied with is misleading and amounts to misbranding.

It is rare for a regulator to write this plainly about its own procedure. And yet the phrase “FDA approved device” is common in Korean advertising and in clinic material. For XERF too, we do not use that phrase.

The accurate form of words is this — “it passed the US FDA's 510(k) premarket notification, the indication cleared in that procedure is electrocoagulation and hemostasis, and no clinical trial was submitted.” It is long, but every part of it is true.

In case XERF should read as unusually thin on evidence, here are the corresponding sentences from other devices that went through the same procedure. All of them come from public summaries filed with the FDA.

The bottom row is the point of this table. It is not that “there is no clinical trial because it is a 510(k)”. When the claimed indication is as minimal as electrocoagulation · hemostasis, it goes through without clinical data; put a specific effect such as fat reduction into the indication and clinical data has to be submitted. Most aesthetic radiofrequency devices choose the former. So the fact that a device “passed the FDA” tells you nothing about a lifting effect — because that was never what was reviewed.

XERF is a medical device approved by the MFDS. And yet this time too we did not find the marketing approval number. It is not on the manufacturer's official site, and the MFDS electronic civil petition window and the medical device information portal both block automated lookups.

The manufacturer publishes only one number, and it is not an approval number.

This structure is the reason the only number anyone displays is the advertising review number. The approval number carries no obligation to display and the advertising review number does, so the long number you see in the material is usually the latter. It is not an attempt to deceive; it is how the system is built — we have written about this at more length in how to read approval numbers.

One more thing. XERF's product classification is “general-purpose electrosurgical unit”. That is a broad category holding every device that cuts · coagulates with radiofrequency current, and the electrocautery unit in an operating theatre sits in the same box. The name says lifting device, but it is not a classification in which a “facial lifting effect” was recognised individually.

This is where the misunderstanding runs deepest. Look at the wording of the rules directly. This is the proviso to Article 9(2) of the Enforcement Rule of the Medical Devices Act.

“Provided that, in the case of a medical device whose structure · principle · performance · intended use and method of use are substantially equivalent to a medical device already approved … the materials under subparagraphs 5 through 7 need not be submitted.”

Subparagraph 6 here is the material on clinical trials.

That is, if it is substantially equivalent to a product already approved, no clinical trial material need be submitted. Article 25 of the Regulation on Approval · Notification · Review of Medical Devices grants the same range of exemption to class 3 · 4 devices.

The structure resembles the American 510(k) to a startling degree. Both countries ask “is it the same as the product before it” rather than asking “does it work” over again each time. This is the regulatory reason why, in aesthetic medical devices, an approval exists but a clinical paper does not so often.

So in a consultation the question “is this device MFDS approved” is only half useful. Approval tells you “is this a device legally on the market”; it does not tell you whether an effect has been confirmed. If you want the second, you have to ask separately: “is there a study done with this product”.

People often say “an advertisement that passed prior review”, but that is not what the present system is.

On 28 August 2020 the Constitutional Court held the former prior review of medical device advertising unconstitutional (breach of the ban on prior censorship). The Medical Devices Act was then amended and self-regulatory review took effect on 24 June 2021, with the registered self-regulatory body being the Korea Medical Devices Industry Association. Even the duty to display a review number rests not on statute but on the review body's operating rules.

A caution — these are two different systems. What is described here is medical device advertising review, which a manufacturer obtains when advertising its own device. Medical advertising review, which a clinic obtains when advertising its own services, is a separate system under the Medical Service Act with a different review body. When you look at a number written in some material, do not mix the two up.

We will write it honestly. It is not “because it is clinically proven”. The material set out above does not let us say that.

• The principle of the class is established. Handling collagen with heat has been studied for a long time, and across monopolar radiofrequency as a class there is material that is small in scale but consistent in direction.
• A tip range that can be used region by region is useful in practice. Sweeping both a narrow curved area and a broad flat one with the same tip is physically unreasonable.
• There is now material on this device itself. Thin is different from absent. And this material may grow in future, or be overturned.

We hold no grounds for saying anything beyond that. We do not say “the result is different because it is this device”.

This piece is about XERF, but the method itself works for any device.

1. “Is this number an approval number or an advertising review number?” — if it is long with several hyphens, it is usually the latter.
2. “Is there a study with this device, or is it a study of the class?” — answering “there are studies of the class but none of this device” is the more accurate answer.
3. “For what indication did the FDA clear it?” — for most aesthetic radiofrequency devices it is electrocoagulation · hemostasis.

For all three questions, an answer of “there is none” is not in itself the problem. The problem is the side that speaks of what does not exist as though it did.

The left column of this table has grown since a year ago. The right column is much as it was. The most accurate sentence we can offer is “a stage at which the material has begun to appear”.

Written and reviewed by Lee Chi-Hak, MD, medical director of Miso Clinic in Daegu, South Korea. Every study cited above is given together with its design, its size and the limitations the authors themselves recorded, and where we could not find data, we have said that we could not find any.