Non-surgical management
Most papers on surgery for chronic instability comment that "the patients had full non-surgical treatment before being considered for surgery" and this is considered good practice (Karlsson and Lansinger 1993, Trevino et al 1994).
However only one published series examined the effect of functional rehabilitation on chronic instability. Karlsson (1991a) found that 50% of patients with chronic instability benefited from rehabilitation. Patients with mechanical instability were less likely to benefit than those with purely functional instability.
Surgery
Functional ankle instability often resolves after arthroscopic debridement of synovitis
Arthroscopy
The studies of Schafer and Hintermann (1996), Kibler (1996), Ogilvie-Harris et al (1997), Okuda (2005) and Ferkel (2007) found a high incidence of intra-articular pathology:
- 25-60% synovitis and impingement lesions
- 25% spurs
- 6-60% osteochondral lesions
- 6% major deltoid tears
About a quarter of intra-articular lesions were not identified pre-operatively.
Laing et al (2004) found that arthroscopic treatment of intra-articular lesions avoided the need for surgery in 50% of patients with demonstrated mechanical instability.
It is our practice to offer an examination under anaesthesia, stress radiographs and arthroscopy to all patients with persisting functional instability after a functional rehabilitation programme. This allows us to treat intra-articular pathology and identify those with mechanical instability who can be offered a stabilisation procedure. It also identifies the small group with combined ankle and subtalar instability for whom a Sammarco tenodesis rather than a modified Brostrom procedure is appropriate. Approximately 40% of our patients have had only an arthroscopic procedure. Patients with mechanically stable ankles and other intra-articular problems have generally had good results from arthroscopic surgery. Patients with unstable ankles will be offered a stabilisation procedure. We prefer to perform this about 6 weeks after the arthroscopy rather than at the same sitting because of the amount of swelling that tends to accompany an ankle arthroscopy.
Stabilisation
There are two main approaches to surgical stabilisation:
- Reattachment of the existing ligaments to the bone – the Brostrom procedure and its modifications, especially that described by Gould, in which the inferior extensor retinaculum is used to reinforce the repair. A capsular shift procedure has also been described by Zwipp.
- Use of a graft, usually all or part of the peroneus brevis but other autografts, allografts and artificial implants have been described. Few of these procedures are routed through the normal ligament attachment sites, and hence cannot reproduce normal ankle kinematics. However, Bahr (1997) and Sammarco (1999) described similar anatomical reconstructions in which a peroneus brevis graft was fixed to the normal ligament attachment sites through bone tunnels (Bahr) or suture anchors (Sammarco).
Biomechanical studies (Liu and Baker 1994, Hollis et al 1995, Bahr 1997, Schmidt 2004, Fujii 2006) show that, if anything, the Brostrom-Gould procedure confers more stability and more normal ankle kinematics than any of the non-anatomical tendon grafts. The Bahr anatomical tendon graft also produced kinematics close to normal.
There have been two randomised controlled trials comparing the Brostrom to non-anatomical reconstructions. Both had significant methodological deficiencies. Neither showed any difference in functional outcome between the trial groups.
Longitudinal studies of single procedures report:
- satisfactory clinical results in 85% with the Brostrom procedure at a mean of 6 years (Karlsson et al 1988b) – failures were all due to recurrent instability, mainly in patients with generalised joint laxity or longstanding instability pre-operatively. Satisfactory results were maintained at 26 years in 18/21 naval personnel (Bell 2006). The presence of intra-articular damage at pre-stabilisation arthroscopy did not reduce the prevalence of good results (Ferkel 2007).
- clinical failure in over 50% of the traditional Evans and Watson-Jones tendon grafts (non-anatomical) due to recurrent instability, pain and ankle ankle arthritis after 10-20 years (van der Rijt and Evans 1984, Karlsson et al 1988a, Hoy and Henderson 1994, Nimon et al 2001)
- good stability but a high rate of complications in the semi-anatomical Chrisman-Snook tendon graft at 4-27 years (Snook et al 1985)
- good short-term stability with little ankle stiffness, pain or complications in the Sammarco anatomical tendon graft (Sammarco and Idosuyi 1999)
- deVries (2005) reported satisfactory clinical results in 2/3 of 40 patients 20-30y after anatomical repair of the ATFL with plantaris, although 3 patients had OA.
Because it is simple, restores near-normal ankle kinematics and has a fuller long-term clinical outcome evidence base than the anatomical tendon grafts, we advise that the Brostrom procedure is to be preferred in the average patient. For the more complex case we offer the Sammarco anatomical tenodesis, accepting that further long-term results may influence this advice. Indications for the Sammarco procedure include:
- insufficient ligamentous tissue for a Brostrom (so our Brostrom patients are always asked to give consent for a Sammarco procedure if necessary)
- neuromuscular patients with pes cavus
- chronic severe instability
- very high-demand patients where stability is preferred over flexibility
- generalised joint laxity (though we strongly prefer to avoid surgery in these patients if possible
Post-op management
Standard post-operative regimes after ankle ligament reconstruction involve 4-6 weeks in a below-knee cast, sometimes with a period of non-weightbearing. Karlsson (1995, 1999) reported trials comparing casting with walker boots and ankle braces, finding no deterioration in outcome with these simpler and more comfortable supports. Wearing a plaster cast is not necessary after an ankle ligament reconstruction and delays functional recovery. We manage our patients in functional braces from the beginning, except for an initial 48 hours splintage in a backslab to allow post-op swelling to settle.