Remplissage in shoulder instability — the missing middle between Bankart and Latarjet.

The operation in plain language.

Remplissage is the French word for "filling." The operation, described by Eugene Wolf and colleagues in 2004, does exactly that: it fills the divot on the back of the humeral head left by a dislocation — the Hill-Sachs lesion — with soft tissue.¹ The infraspinatus tendon and the underlying posterior capsule are sutured into the bony defect, converting a pocket the joint can lever into during external rotation into a flat, non-engaging surface.

It is an arthroscopic operation, performed through the same posterior portal used for the diagnostic arthroscopy. Two suture anchors are typically placed at the deepest point of the Hill-Sachs. The sutures pass through both the capsule and the infraspinatus, and when tied they pull the tendon down into the bony defect like upholstery into a button. In almost every modern instability case where remplissage is performed, it is added to an arthroscopic Bankart repair on the front of the shoulder — the two operations are done in the same anaesthetic, through the same setup, and the combined operation takes around 90 minutes.

The shift that made it useful: the glenoid track.

Remplissage existed for years before the shoulder world worked out what it was actually for. The shift came from a series of Japanese and American papers that reframed the engagement problem geometrically. Yamamoto and Itoi, in 2007, defined the glenoid track — the contact zone the humeral head sweeps across the glenoid during maximum abduction and external rotation.² If the Hill-Sachs lesion sits entirely within that contact zone, the lesion never engages the rim. If the lesion extends medial to the contact zone, the rim catches the lesion in abduction and external rotation and the shoulder levers out.

Di Giacomo, Itoi and Burkhart formalised the terminology in 2014: on-track lesions stay safely within the contact path; off-track lesions engage and cause recurrent instability.³ Critically, the off-track determination depends on both the size and position of the Hill-Sachs and the amount of glenoid bone loss — the two interact. A modest Hill-Sachs in a glenoid with significant bone loss can be off-track. A deep medial Hill-Sachs in a near-normal glenoid can also be off-track. That interaction is what is called bipolar bone loss.

This reframing did two things. It explained why some Bankart repairs failed catastrophically — the labrum had been reattached, but the off-track Hill-Sachs was still levering the shoulder out at the back. And it gave surgeons a way to predict, from a CT scan or MRI, which shoulders needed more than a Bankart.

Video: the on-track vs off-track concept in visual form. The 3D mechanics are easier to see than to read.

What Denard, Burkhart and the Oregon group have argued.

Patrick Denard's group at the Oregon Shoulder Institute, alongside Stephen Burkhart in San Antonio and Alexandre Lädermann in Geneva, have been the most consistent voices arguing that remplissage is the right answer in a specific patient group that is currently being over-treated with Latarjet.

Their argument, condensed:

1. The Latarjet has a high complication profile. Major and minor complications are reported in the 15–30% range in published series, including coracoid graft non-union or lysis, screw irritation, neurovascular injury (the musculocutaneous nerve, occasionally the axillary), and revision arthritis. The recurrence rate is low — typically 1–5% — but every Latarjet patient carries a non-trivial chance of needing a second operation that has nothing to do with their original instability.
2. Bankart-plus-remplissage has a much lower complication profile. No graft. No screws. No structures placed near the brachial plexus. The recurrence rate in well-selected patients sits in the same 2–5% range as Latarjet, in the published meta-analyses.
3. The motion cost is modest. The early concern about remplissage was that tenodesing the infraspinatus into the back of the humeral head would lock the shoulder into internal rotation. The clinical data have not borne that out: external rotation loss is typically 4–10 degrees, much of which returns by twelve months, and the loss is at the extremes of motion that most non-throwers never use.
4. There is a sweet spot of patients being treated wrongly. The patient with an off-track Hill-Sachs but glenoid bone loss under 15–20% — the so-called "subcritical" range — is the patient the Denard argument targets. In many practices this patient is offered Latarjet by default. In the Denard framework this patient is the natural remplissage indication.

Technique pearls.

The technique has been refined considerably since Wolf's original description. The Denard group and others have published technique notes that converge on a few principles:

• Do the remplissage first. Place the posterior anchors and pass the sutures before the Bankart is fully repaired on the front. This keeps the joint open and the working space generous; tying down the remplissage at the end, once the Bankart is repaired, prevents the posterior tenodesis from constricting the joint during anterior work.
• Two anchors, deepest point. One anchor at the superior aspect of the Hill-Sachs and one at the inferior aspect, both placed at the deepest part of the defect. Anchors placed too shallow create a partial fill that may not eliminate engagement; anchors placed at the edges fail to reduce the tendon into the bone.
• Pass through both infraspinatus and capsule. The construct is a capsulotenodesis, not just a tenodesis. Passing through both layers buys a more durable repair and uses the capsule's intrinsic healing.
• Tie low and from the cannula. Knot stack on the deep surface of the tendon, not exposed in the joint, to avoid mechanical irritation.
• Visualise from anterior, work from posterior. The scope sits in the standard anterior portal looking at the back of the joint; the working instruments come through an enlarged posterior portal placed slightly more lateral than the standard diagnostic portal.

The evidence base.

The single most important piece of evidence is the MacDonald randomised controlled trial, published in the Journal of Shoulder and Elbow Surgery in 2021.⁴ One hundred and eight patients with anterior shoulder instability and a Hill-Sachs defect were randomised to arthroscopic Bankart alone or arthroscopic Bankart plus remplissage. At two-year follow-up the recurrent instability rate was 18% in the Bankart-alone group and 2% in the Bankart-plus-remplissage group. External rotation loss in the remplissage group was small and not clinically significant. Patient-reported outcome scores were similar between groups. The number-needed-to-treat is, broadly, six — for every six Bankart-plus-remplissage operations performed, one further dislocation is prevented compared with Bankart alone.

Hurley and colleagues published a meta-analysis pooling the available comparative evidence and reached the same conclusion: adding remplissage to Bankart reduces recurrence with minimal motion penalty.⁵ Garcia and colleagues compared Bankart-plus-remplissage with bone-block augmentation (Latarjet and iliac crest) in patients with bipolar bone loss and found the remplissage construct held its own across most outcome metrics with a substantially lower complication burden.⁶

The modern decision algorithm.

The framework I use in clinic, drawn from the Denard / Burkhart / Lädermann literature and the broader consensus, looks like this:

• On-track Hill-Sachs, glenoid bone loss under 15% → arthroscopic Bankart alone. The labrum is the problem; fix the labrum.
• Off-track Hill-Sachs, glenoid bone loss under 15–20% → arthroscopic Bankart plus remplissage. This is the Denard sweet spot. The bone loss is subcritical; the humeral side is the engaging problem; remplissage addresses it without the cost of a bone-block.
• Glenoid bone loss above 20%, with or without an off-track Hill-Sachs → Latarjet or another bone-block solution. The glenoid is now the problem; remplissage cannot rebuild glenoid bone.
• Critical glenoid bone loss above 30%, or revision Latarjet failure → distal tibial allograft or iliac crest autograft reconstruction.

This algorithm is not a substitute for the consultation. The decision is shaped by the patient's age, hand dominance, sport, occupation, the number of prior dislocations, the energy of the index injury, the presence of generalised laxity, and what they need their shoulder to do. The 28-year-old roof tiler with an off-track Hill-Sachs and 12% glenoid bone loss is a clean remplissage indication; the 19-year-old division-one baseball pitcher with the same imaging is not, because the external rotation cost is too consequential for his throwing.

When remplissage is not the answer.

• Critical glenoid bone loss (>20–25%). The remplissage cannot rebuild the front of the glenoid. The shoulder still has nowhere safe to sit in abduction.
• Posterior instability. Remplissage addresses anterior instability by filling a Hill-Sachs on the posterolateral humeral head. Posterior instability with a reverse Hill-Sachs on the anteromedial humeral head is a different problem with a different operation.
• Overhead throwing athletes. The 5–10 degrees of external rotation that remplissage can cost is the 5–10 degrees a baseball pitcher or javelin thrower lives in. The decision in this group has to be made carefully and is sometimes a careful Bankart and a careful conversation about retirement risk.
• Multidirectional instability with capsular laxity. The dominant problem is the capsule, not the Hill-Sachs. Remplissage is an adjunct, not the primary operation.

The deeper point.

For two decades the instability conversation in most clinics has been binary: Bankart for the simple cases, Latarjet for the complex ones. The Denard / Burkhart / Lädermann line of argument is that there has always been a third category in between — the off-track but subcritical-bone-loss patient — and that this group has been routed into Latarjet by default because remplissage was viewed as a niche bailout. The MacDonald RCT moved that argument from opinion to evidence.

The reason this matters in clinic is concrete. A Latarjet is a bigger operation than a Bankart-plus-remplissage. It is performed open. It involves a coracoid osteotomy, screws across the brachial plexus, and a non-trivial reoperation rate independent of its low recurrence rate. For the right patient it is the right operation. For the patient who could have been managed with two posterior anchors and a soft-tissue tenodesis, it is too much surgery.

The point of operating on a shoulder instability is to do exactly enough surgery to stop the shoulder dislocating, without leaving the patient with a different problem in five years. Remplissage, used in the right patient, often is exactly enough.

---

Related reading on this site

• Shoulder instability — the condition page
• Arthroscopic Bankart repair — procedure detail
• Shoulder stabilisation and Latarjet — procedure detail
• Shoulder instability: the age-at-first-dislocation rule

References

1. Wolf EM, Pollack M. Hill-Sachs "remplissage": An arthroscopic solution for the engaging Hill-Sachs lesion. Arthroscopy. 2004;20(S1):e14–e15 (technical note); subsequently published in detail: Purchase RJ, Wolf EM, Hobgood ER, Pollock ME, Smalley CC. Hill-Sachs "remplissage": an arthroscopic solution for the engaging Hill-Sachs lesion. Arthroscopy. 2008;24(6):723–726.
2. Yamamoto N, Itoi E, Abe H, et al. Contact between the glenoid and the humeral head in abduction, external rotation, and horizontal extension: a new concept of glenoid track. J Shoulder Elbow Surg. 2007;16(5):649–656.
3. Di Giacomo G, Itoi E, Burkhart SS. Evolving concept of bipolar bone loss and the Hill-Sachs lesion: from "engaging/non-engaging" lesion to "on-track/off-track" lesion. Arthroscopy. 2014;30(1):90–98.
4. MacDonald PB, McRae S, Old J, et al. Arthroscopic Bankart repair with and without arthroscopic infraspinatus remplissage in anterior shoulder instability with a Hill-Sachs defect: a randomized controlled trial. J Shoulder Elbow Surg. 2021;30(7):1518–1528.
5. Hurley ET, Toale JP, Davey MS, et al. Remplissage for anterior shoulder instability with Hill-Sachs lesions: a systematic review and meta-analysis. J Shoulder Elbow Surg. 2020;29(12):2487–2494.
6. Garcia GH, Liu JN, Wong A, Cordasco F, Dines DM, Dines JS, Gulotta LV, Warren R. Arthroscopic Bankart Repair With Remplissage in Comparison to Bone Block Augmentation for Anterior Shoulder Instability With Bipolar Bone Loss: A Systematic Review. Arthroscopy. 2019;35(8):2495–2503.