The Cross Bracing Protocol: can a torn ACL heal without surgery?

Until recently, the advice after a complete ACL rupture was straightforward: the ligament does not heal, so you either have it reconstructed or you learn to live without it. That advice is now being questioned, and the question is coming from Australia.

The Cross Bracing Protocol is a non-surgical method of managing an acute ACL rupture, developed by Dr Tom Cross in Sydney. It has had a great deal of media attention, and patients are increasingly asking about it in the rooms. This article sets out what the protocol involves, what the published evidence actually shows, and who it may and may not suit.

What the protocol involves

The principle is that a torn ACL is held in a position where its two ends sit close together, giving the ligament the mechanical conditions it needs to heal. That is achieved by immobilising the knee in deep flexion.

• Weeks 0–4: the knee is locked in a brace at 90 degrees of flexion. Movement is not permitted.
• Weeks 4–12: the permitted range of movement is increased in stages until the brace comes off at 12 weeks.
• Throughout: supervised physiotherapy, with progression governed by set goals rather than by the calendar.

Timing is the critical constraint. The protocol has to begin within a few weeks of the injury, while the torn ends are still capable of bridging. A patient who presents three months after the event has missed the window.

What the evidence shows

The primary study is a prospective cohort of 80 patients with acute ACL rupture, published in the British Journal of Sports Medicine in 2023. The results are genuinely striking, and they deserve to be reported accurately rather than as a headline.

At three months, 90 per cent of patients had MRI evidence of ACL healing. Those patients divided into two groups that behaved quite differently:

• 50 per cent had a continuous ligament of normal or increased thickness, the best healing grade.
• 40 per cent had a continuous but thinned or elongated ligament.
• 10 per cent showed no healing at all.

The distinction between those groups mattered a great deal at twelve months. Patients in the best healing grade returned to their pre-injury sport at a rate of 92 per cent; those in the poorer grades returned at 64 per cent. Every patient in the best grade had normal knee laxity at three months, compared with 40 per cent of the rest.

Eleven of the 80 patients, 14 per cent, re-injured their ACL.

That re-injury figure is the number that tends to get left out of the summaries. It is not necessarily worse than the re-rupture rate after reconstruction in a young pivoting athlete, but it is not zero, and any patient considering this pathway should have it put to them clearly.

The honest limitations

The authors of the study are careful about this, and so should anyone discussing it be.

• It is an observational cohort, not a randomised trial. There was no comparison group managed surgically or with standard rehabilitation.
• Follow-up was twelve months. ACL graft failures and instability problems often declare themselves later than that.
• Nearly half the patients had an associated meniscal injury. How those injuries fare in the long term under this protocol is not yet established.
• Four weeks of immobilisation at 90 degrees is not a trivial intervention. Stiffness, muscle wasting and the risk of venous thromboembolism all have to be managed actively.

What the study establishes is that ACL healing after complete rupture is possible and is more common than orthopaedic teaching has assumed. What it does not yet establish is how those healed ligaments perform at five and ten years, or how they compare with a reconstructed knee in a matched patient.

Who it may suit

On the evidence available, the patients most likely to be considered are those who present within the first two to three weeks of injury, who have a tear pattern on MRI that looks amenable to healing, who do not have a displaced meniscal tear or another injury requiring surgery, and who are in a position to commit to twelve weeks of bracing and closely supervised rehabilitation.

It is much less likely to suit a patient whose knee is locked, who has a bucket-handle meniscal tear or a multiligament injury, who presents late, or whose circumstances make three months in a brace impractical.

How Dr Mooney approaches the question

The protocol is a legitimate option that deserves to be discussed, not dismissed. It is also not a replacement for reconstruction in every patient, and the marketing around it has at times been more confident than the data supports.

Where a patient raises it, the discussion covers the timing of their presentation, what the MRI shows about the tear pattern and any associated injuries, what their sport and work demand of the knee, and what the realistic outcomes look like on each pathway, including the possibility that the protocol is attempted, does not produce a stable knee, and reconstruction follows. Choosing the protocol first does not remove reconstruction as an option later.

If you have had a recent knee injury and want that conversation, the practical thing is to be seen early. The window for this decision is measured in weeks.