Incision Options

Axillary vs Inframammary Incision

UNE Plastic Surgery · Dr. Kim Uigeon, board-certified plastic surgeon · Published 2026-09-21 · Updated 2026-09-21

1. The real difference between the two incisions

For reference, the periareolar incision once common is rarely used in Korea now, because it passes through the milk ducts, where bacterial exposure is considered relatively greater.

2. The inframammary scar’s location and shape

The IMF scar forms along the fold under the breast. Standing or clothed, the fold hides it well. Honestly, though, there are angles where it does not show and angles where it does: lying down, sitting up, or raising the arm so the breast is stretched reveals the line. A lean body may show it a little more, as less flesh covers it, and where the natural fold is faint, deciding where to place the scar is itself tricky. A common misunderstanding: an IMF incision does not put the scar at the original fold. Because the fold drops once the implant is in, the incision is placed at where the new fold will settle after surgery. Getting this wrong leaves the scar floating on the flesh below the breast, which is the most visible outcome. Scar care is the same regardless of incision: at UNE the inner tissue is closed in two layers and the skin sealed with Dermabond, so no stitches are removed; from a month after surgery, a silicone sheet or scar ointment is used consistently for over 3 months, with sun protection — sunlight on a scar leaves pigment. Redness begins to fade around 6 months and lightens over 1–2 years.

3. The actual visibility of the axillary scar

The axillary scar’s advantage is clear: nothing is left on the breast. Those who often wear revealing clothes or dresses value this greatly. But there is a condition: the incision must sit in the deepest underarm crease, inside the hair-bearing skin, to be fully hidden with the arm down — even slightly outside and the mark shows. What happens when the arm is raised high? It can show — but consider how often full arm-raising actually occurs in daily life. One more point: the inner underarm fat layer is threaded with vessels, nerves, and lymphatics. At UNE, entry hugs the lateral side of the pectoralis rather than going into that fat, because disturbing it can leave lasting arm numbness or swelling. And making the incision as small as possible is not better: too small, and the retractor keeps pushing and pulling the skin so the edges are crushed — worse for the scar — and with no room for instruments, the inner tissue is scraped more. The visible line gets shorter while the internal injury gets larger. So the point where scar length meets surgical accuracy is 3.0–3.5 cm.

4. The papers compared by incision

Studies compare by incision, but the numbers alone mislead — the conditions must be read too.

The table makes IMF look advantaged, but these numbers cannot be taken at face value. Start with 6.4% vs 0.5% contracture (Jacobson et al., Aesthetic Surgery Journal, 2012):^[1] across 336 breasts, contracture occurred in a total of 6 cases — the rate is those 6 split among three incisions, so a difference of a few cases swings the percentage greatly; the mean follow-up was about 13 months, and it was a retrospective chart review (the same study put periareolar at 2.4%). Operating time and pain also carry conditions (Chen et al., Aesthetic Plastic Surgery, 2024):^[2] the axillary group was significantly younger and had fewer births than the IMF group — a limitation the team noted — while every other measure showed no difference, complications were zero on both sides, and satisfaction was equal. In other words, the accurate reading is not “axillary is dangerous” but “axillary is a more hands-on operation.”

5. Technique decides the result more than the incision

This is the point most worth emphasizing. Tebbetts organized 28 years, 1977–2005, of 690 axillary surgeries — subglandular in 84, submuscular in 294, dual plane in 312 (Tebbetts, Plastic and Reconstructive Surgery, 2006).^[3] The conclusion mattered: with an optimized endoscopic technique and instruments, the axillary incision can reach the same level as other incisions — in aesthetic result, recovery, reoperation rate, and complication rate. What makes the paper striking is that it splits the 28 years at one point: after introducing the endoscope in 1992 and eliminating blunt, unseen dissection by pushing with instruments, the results changed. Baker III–IV contracture fell from 4.2% to 1.3%, the overall reoperation rate from 9.3% to 3.3%, and inner-arm lymphedema and tightness from 10.3% to 2.2%; infection and seroma stayed at 0% throughout the 28 years. The same axillary incision became almost a different operation depending on how it was dissected. In other words, a number like the 6.4% above may be less the nature of the incision than the hand that works it.

A study that randomized the endoscope

One surgeon’s era comparison can mix in other factors, so a randomized comparison is needed — and in 2026 the result arrived. It randomized 118 axillary patients into two methods: one without an endoscope, the other using an endoscope for the final muscle release and hemostasis, followed for 2 years (Boliglowa et al., Plastic and Reconstructive Surgery, 2026).^[4] One surgeon did all of them, and the two groups matched on age, BMI, and implant size (median 330 cc).

• Animation deformity (grade 1): 33.9% without the endoscope, 13.6% with it — the one item with a statistically significant difference. Whether you can see how far the muscle was released in dual plane was the fork.
• Contracture and double bubble: 4 vs 1 each — numbers fell but were not statistically significant, as one group of 59 swings on a case or two.
• Hematoma, seroma, infection: zero on both sides.
• Operating time: 12 minutes longer with the endoscope (46 vs 58 minutes).

In this study the implant was inserted with a Keller Funnel so it did not touch the skin, and any contracture was reopened through the original axillary incision to remove the capsule and exchange — the scar was recorded as barely visible at 2 years, though this requires the endoscopic equipment. Combining the two studies gives one conclusion: axillary outcomes are set not by the incision position but by how precisely the dissection and hemostasis are done under the endoscope. At UNE the entire process from incision to finish is done endoscopically, and the implant is placed no-touch with a Keller Funnel following contamination-minimizing steps.^[5] Because the axillary route enters from a distance, the order of finding and controlling a vessel before cutting matters especially — pooled bleeding turns into inflammation, which leads to contracture — and with today’s endoscope resolution, fine vessels are often seen better than by eye. The main reason an operation runs long is bleeding; once it is controlled, the time drops sharply, and for a first surgery the operating-time difference between the two incisions is small.

6. The difference when revision is considered

For a first surgery alone, scar preference can decide it. The issue is later. Even after a first surgery through the axilla, some revisions can still be done that way: changing only the implant size, exchanging a ruptured implant, or widening the pocket a little are possible through the axillary incision — reusing the existing incision rather than opening anew, so no scar is added to the breast. But cases needing a capsulectomy, such as bottoming out or symmastia, are different: surgery that removes the whole capsule or re-sutures the pocket lives on precision and must be done through the IMF — which does add a scar to the breast. A first surgery done through the IMF handles most revisions at the same site, though a large size change moves the fold and may need a new incision line. Some cases must go IMF from the start: ptosis, and a case where the fold must be lowered a lot but the original fold is distinct (the latter risks a double bubble, so IMF access is needed); tuberous breast is also done through the IMF. Conversely, for a first surgery with no ptosis, wanting no scar on the breast at all, the axilla is a good choice — not a preference but a well-grounded one. Five things to actually check:

• How much scar you can accept — a breast scar or an underarm one.
• How complex the dissection is — a deep dual plane favors the IMF.
• Whether there is revision potential — prior surgery and complications change the approach.
• Whether there is ptosis — if a lift is added, the IMF is the base.
• Your usual postures — how often you raise your arms or lie at an angle.

None of these has a fixed answer, but revision potential is weighed more heavily than the rest, because it is about a new scar later.

7. How the scar changes over time

Regardless of incision, the scar’s course is similar; knowing it in advance prevents alarm.

The key is that it stands out most at 3–6 months. Many worry “why is it worse?” then, but it is a normal course — this is the peak, and it improves after. There is individual variation, though: some progress to a raised hypertrophic scar, influenced by constitution, care, and incision site. Care starts a month after surgery once the suture line has stabilized — a silicone sheet or scar ointment; applying too early is an irritant. Sun protection and not smoking matter more than expected. Begin massage only after confirming the timing, and if the scar rises thickly or redness has not settled by 6 months, mention it at a follow-up; a hypertrophic scar needs a separate treatment such as injection or laser. If you know you scar readily, say so at consultation so a prevention plan can be made together.

8. Why the periareolar incision is rarely used

A third option is the periareolar incision, cut along the lower border of the areola. Its advantage is clear: the scar resembles the areola color and is hard to see, hidden at the border. But it is rarely used in Korea, because it must cross the gland and milk ducts — ducts hold bacteria, and passing an implant along that path raises contamination risk. In the Jacobson study above, the periareolar contracture rate was 2.4%, between axillary (6.4%) and IMF (0.5%) — a good choice on the scar alone, but paying the price elsewhere.

Summary

• The incision length is the same — both 3.0–3.5 cm. What differs is where the scar sits and the distance from there to the implant.
• Read the numbers with their conditions. The 6.4% vs 0.5% contracture is a rate from 6 total cases, and the operating-time comparison had different ages between groups. Take the direction as reference, not as settled.
• Conditions decide the incision. Ptosis, a short and distinct fold risking a double bubble, and tuberous breast are done through the IMF. Size change, rupture exchange, and pocket widening are possible through the existing axillary incision.

Incision is not a matter of superiority — it is a matter of which is predictable for your breast.

References

1. Jacobson JM, Gatti ME, Schaffner AD, Hill LM, Spear SL. Effect of incision choice on outcomes in primary breast augmentation. Aesthet Surg J. 2012;32(4):456–462. doi:10.1177/1090820X12444267. (183 patients, 336 breasts; retrospective; mean follow-up 392.6 days.)
2. Chen Z, Qiu Z, Tong J, et al. Surgical trauma comparison of inframammary fold versus endoscopic transaxillary approaches in breast augmentation: a 7-year cohort study. Aesthetic Plast Surg. 2024. doi:10.1007/s00266-024-04211-x. (194 patients; single institution; retrospective.)
3. Tebbetts JB. Axillary endoscopic breast augmentation: processes derived from a 28-year experience to optimize outcomes. Plast Reconstr Surg. 2006;118(7 Suppl):53S–80S. doi:10.1097/01.prs.0000246109.33009.28. (690 patients, 1977–2005; single surgeon; reoperation 9.3%→3.3% after the endoscope.)
4. Boliglowa D, Pontillo M, Velotti N, Rocco N. Transaxillary breast augmentation: a randomized controlled trial. Plast Reconstr Surg. 2026;158:43e. (118 randomized; animation deformity 33.9% vs 13.6%.)
5. Adams WP, Culbertson EJ, Deva AK, et al. Macrotextured breast implants with defined steps to minimize bacterial contamination around the device: experience in 42,000 implants. Plast Reconstr Surg. 2017;140(3):427–431. doi:10.1097/PRS.0000000000003575.