Correction of other deformities
A lot of patients with metatarsalgia also have other forefoot problems such as hallux valgus, hallux rigidus or lesser toe deformities. Correcting these will often be enough to treat the sub-metatarsal pain.
In papers on hallux valgus surgery which gave data on metatarsalgia before and after first ray straightening, 50% of patients who had hallux valgus correction got rid of metatarsalgia, although 12% got it for the first time after surgery. Of the 8 hallux valgus procedures reported, only the Wilson osteotomy failed to relieve more metatarsalgia than it caused.
The only series of lesser toe correction to document the prevalence of sub-metatarsal pain before and after surgery is that of Lehman and Smith (1995): 29 patients had pre-operative metatarsalgia and 2 afterwards (although the post-operative status is not entirely clear). In our own series (Dhukaram 2002, pre-operative data for metatarsalgia not published), 65 patients had MTP joint pain pre-operatively and 11 afterward, although residual MTP joint pain was the commonest cause of dissatisfaction.
Forefoot pain is twice as common in people with reduced ankle dorsiflexion as in the general population (di Giovanni 2002). Maskill (2010) reported 43 gastrocnemius lengthenings for isolated foot pain, although most of these were primarily fo planatr fasciitis. Of the six who mainly had forefoot pain, mean visual analogue pain score reduced from 7.5-2.2/10. It is difficult to know how much this small, heterogenous study can be generalised, but further data would be valuable.
Surgery for other specific conditions
Patients with severe rheumatoid forefoot disease would generally be offered a forefoot reconstruction, consisting of a 1st MTP fusion and Stainsby procedures to the lesser rays.
Patients who have failed conservative treatment of interdigital neuralgia may be offered an interdigital neurectomy.
Patients with pes cavus may need metatarsal, tarsal or calcaneal osteotomies, arthrodeses or tendon transfers in addition to toe straightening.
Occasionally a patient with a tight Achilles tendon may be offered a percutaneous Achilles lengthening.
Lesser metatarsal surgery
Relatively few need isolated metatarsal surgery and these should be considered individually after adequate imaging as detailed. There are many techniques, most reported only in case series, with few comparative studies and no RCTs.
The Weil osteotomy is a technique for shortening a lesser metatarsal. A near-horizontal cut is made through the head and neck. The near-horizontal orientation, large surface area and ease of fixation minimise the risk of mal-union or non-union and allow early weightbearing.
A longitudinal incision in the adjacent interspace is deepened bluntly to the extensor apparatus, which is mobilised from the joint capsule. The dorsal digital nerves are at risk. A longitudinal capsulotomy is made and the metatarsal head delivered into the wound.
The saw cut begins just below the upper edge of the articular surface. In the second metatarsal it can be kept almost horizontal with respect to the ground; in the other rays it inevitably sloes downward increasingly. The tendency for the head to displace downwards can be counteracted by taking an extra sliver out of the cut. The head is displaced proximally according to the pre-operative plan and fixed with one or two small screws. The remaining “beak” on the upper edge of the metatarsal is excised (alternatively, the beak can be excised back to the planned level of shortening and the head displaced to lie flush with it – this often makes it easier to get the head in the right place.)
Garg et al (2008) described a modification of the Weil osteotomy in which a segment of the metatarsal head/neck is excised obliquely instead of sliding. In a series of 71 case (see below the results and complication rates were similar to the standard Weil.
Garcia-Fernandez (2011) reported 59 Weil osteotomies fixed with a standard screw and 38 left without fixation. There was no difference in clinical results or the risk of recurrent metatarsalgia or floating toe.
No protection is necessary unless the bone is very soft or the fixation unstable. The patient can fully weightbear immediately. The MTP joints can become very stiff and it is important for thepatient to mobilise their toes early and to practise flexing them to minimise the risk of non-functioning “floating” toes.
There are no RCTs comparing the Weil osteotomy with any other procedure. Trnka (1999) reported a retrospective comparison between 15 patients who had 22 Helal osteotomies and 15 patients who had 25 Weil osteotomies, all for MTPJ dislocations. The Helal procedures were a historic series and had mean follow-up of 26m compared with 15m for the Weils. The groups were otherwise comparable.The mean AOFAS lesser ray score in the Weil patients was 85/100 and in the Helals 78 (p=0.02). The Weils also had less pain, fewer transfer lesions and fewer recurrent MTP dislocations.
Davies and Saxby (1999) reported 50 Weil osteotomies in 42 patients at short-term follow-up of 6-16m. 21 also had hallux valgus correction. 77% were painfree, but 2 still had moderately severe pain. One transfer lesion developed, and there was recurrent subluxation of one MTP joint.
O’Kane and Kilmartin (2002) reported another short-term series of 19 procedures. They used the line between the apices of the 1st and 4th metatarsal heads as a reference to determine the amount of shortening, which averaged 5.2mm. Mean AOFAS lesser ray score improved from 44 to 89. They noted 8 floating toes and one transfer lesion. They found no relationship between the amount of shortening and either non-functioning or stiffness of the toes.
Vandeputte (2000) 59 Weil osteotomies in 37 feet and 32 patients. 33 of these joints were dislocated pre-operatively. They operated on 33 2nd metatarsals, 22 3rd metatarsals and four 4th metatarsals. Patients rated 32/37 feet good or excellent and the mean AOFAS score increased from 59 to 81. There were four feet with transfer lesions and 3 with persistent pain. 36% of the toes were stiff and 22% defunctioned, and 5/33 MTP joint dislocations recurred. Improvements in pain were correlated with reduction in sub-metatarsal pressure.
Hofstaetter (2006) reported 7yr results in 53 toes (25 feet, 24 patients). Some also had hallux valgus corrections and/or hammertoe straightening. The mean AOFAS score improved from 48 to 83, with some improvement continuing between 1 and 7 years. One patient had continuing pain, another developed a painful prominent screw tip with a sub-metatarsal callus which resolved on removal of the screw. Of 25 feet with pre-operative dislocations only 3 still had dislocated MTP joints at 7 years. 68% of toes did not contact the ground on standing, but this did not affect the clinical outcome.
Garg et al (2008) described a modification of the Weil osteotomy in which a segment of the metatarsal head/neck is excised obliquely instead of sliding. 71 osteotomies in 48 patients were followed up for 6-26 months. 2/3 of patients had the surgery mainly for metatarsalgia; the others were for toe deformities, dislocation and instability.Mean AOFAS score at follow-up was 87.6/100 and 85% of patients were satisfied (25% with some reservations).27% had floating toes, 19% transfer lesions, 15% infection and 10% wound healing problems.
Khurana (2011) reported 61 patients with a mean follow-up of 31 months. Mean AOFAS score improved from 42/100 to 79 and Foot Function Index from 0.52 to 0.78. Six patients had recurrent metatarsalgia and calluses, and this was related to residual plantar prominence of the metatarsal heads. Twelve patients had (asymptomatic) floating toes and 13 stiff toes. Khurana emphasised the use of metatarsal skyline views to assess plantar metatarsal head prominence and used this intra-operatively in some patients.
Perez-Munoz (2012) described 314 Weil osteotomies in 76 patients. Where >4mm shortening was indicated on pre-op planning they used an excisional modification similar to Garg (2008). The median AOFAS score was 90/100 and 80% of patients were satisfied. Recurrent metatarsalgia developed in 22% of patients, four of whom needed revision. Stiffness occurred in 60% of feet but floating toes in only 4%.
After Weil osteotomy, it is common for the toe not to touch the ground in the standing position (“floating toe”). Most series report an incidence of about 20%, although Hofstaetter (2006) found that 2/3 of their patients had floating toes. Migues (2004) highlighted the problem of floating toes in their report of 70 operations. They found that the amount of metatarsal shortening did not predict the development of a floating toe, but it was commoner (50%) when a PIP joint fusion was also carried out. Trnka (2001) showed, in a cadaver model, that the osteotomy plane is always inclined plantarwards, and this causes the lumbrical muscles to change from flexors of the MTP joint to extensors. Boyer and deOrio (2004) avoided floating toes in patients who also had a flexor-extensor transfer and trans-articular pin fixation, but 11/13 MTP joints were moderately or severely stiff. Gregg (2007) reported a combination of Weil osteotomy and plantar plate repair. It is not entirely clear how many patients had floating toes, probably 3/35 (8.6%).
Overall about 90% of patients reported have been satisfied with the results of Weil osteotomy. About 40% of MTP joints are stiff and 25% of toes floating. Transfer lesions occurred in 4%, residual pain in 15% and recurrent subluxation or dislocation of the MTP joint in 13% of those that were dislocated pre-operatively. It may be possible to reduce the incidence of floating toes by formal stabilisation, probably at the cost of more stiffness.
Many authors describe pre-operative planning according to the method of Maestro (2003). Bevernage (2008) showed that despite careful pre-operative planning it was difficult to achieve the pre-operative plan and the amount of shortening achieved did not correlate with outcome. Perhaps this is related to Kaipel’s (2011) finding that maximum peak pressure and maximum force did not correlate with 1st or 3rd metatarsal length. Indeed, Chauhan (2010) compared three conceptually different ways of measuring relative metatarsal length and found that the results depended strongly on the method used. It seems that the concepts underlying metatarsal shortening to treat metatarsalgia have not been well supported by empirical studies. Khurana’s (2011) emphasis on plantar prominence may be a useful line of enquiry, especially given Dreeben’s (1989) finding that that the mean ground-metatarsal head vertical distance was 7.1 mm in symptomatic rays and 8.3 mm in asymptomatic rays. More critical, better designed studies would be useful in advising patients on the benefits and risks of surgery for metatarsalgia.
Distal wedge osteotomy
This operation was originally described by Wolff. A dorsally based wedge is excised from the metatarsal neck and the osteotomy closed by cracking the intact inferior cortex. Some dorsal translation may be possible.
Dreeben (1989) reported 45 operated feet, with 29 asymptomatic controls for radiological and pressure measurements. The osteotomies were done without internal fixation. Sixty-seven percent were painfree at review. On skyline radiographs they found that the mean ground-metatarsl head distance was 7.1mm in symptomatic toes and 8.3mm in asymptomatic toes. Residual pain characteristically occurred in toes where the metatarsal head was elevated <3mm, and transfer lesions in toes which were elevated >4.5mm - not a large margin for error. Mean submetatarsal pressure was 13.4psi in normal toes, 22psi in symptomatic 2nd toes and 18.3psi in symptomatic 3rd toes. Post-operatively the mean pressure reduction was 15psi, but in failed toes only 2psi.
Leventen and Pearson (1990) reported 21 feet. 16/21 were painfree at review and 3 others usefully improved. Mean VAS pain score fell from 7.7/10 to 1.3. Ten patients could wear any shoes. There were 6 transfer lesions and one non-union.
Vertical chevron osteotomy
A vertically-oriented chevron osteotomy with the apex just behind the metatarsal head allows vertical translation with some stability expected from impaction. Kitaoka (1998) reported 19 patients who had 22 osteotomies, examined 2-7y post-op. In this series, the osteotomies were stabilised with K-wires for 3 weeks. 15 had no pain and 17 no interference with activities of daily living. However, 14 had some restrictions in shoewear and 3 continued to use insoles. There were 4 persistent painful keratoses and two had been revisd for transfer lesions.
Snyder compared the Weil and vertical chevron osteotomies in cadaver feet. The Weils were shortened 5mm and the chevrons elevated 3mm. The chevron osteotomy produced significant reduction in mean pressure under the 2nd and, to a lesser extent, 3rd metatarsal heads. The Weil, if anything produced a small increse in pressure. This would fit with Dreeben's (1989) findings that elevation of the metatarsal head was closely correlated with reduction in pressure. Vandeputte (2000), however, found that the Weil did reduce sub-metatarsal pressure in real patients with pre-operative pain and calluses.
Helal described an oblique osteotomy of the distal metatarsal shaft in 1967. The osteotomy runs from proximal dorsal to distal plantar. After making the osteotomy the metatarsal head is freed with an elevator and allowed to translate upwards and proximally. The osteotomy is not fixed; rather, the patient is encouraged to bear full weight early so that the metatarsal head will find its level, so "levelling the tread".
Helal and Greiss (1984) described 310 patients (508 feet) with a mean follow-up of 4.3y. 88.4% were pain-free and 92% had no calluses. Using Helal's score, 77% had a good result. However, less than half the patients had free choice of footwear. 15.5% had non-unions, although most were asymptomatic; 16% had wound healing problems and 12% had repeat osteotomies for malalignment.
Winson (1988) reported good results in less than half of 94 patients, however. 40% had residual metatarsal head prominence with painful callosities, and 14% had required revision osteotomy.
As noted above, Trnka reported a retrospective comparison between the Helal and Weil osteotomies for MTP dislocation. Helal believed that dislocation did not need to be formally reduced as it would correct with metatarsal shortening. In Trnka's series, the Helal osteotomy patients had, on average, one year longer follow-up but were otherwise comparable. The results were significantly better in the Weil group.
Shortening in the shaft
A number of different methods have been described to shorten the lesser metatarsals through the shaft. Some of these techniques also allow elevation of the metatarsal head.
Giannestras (1958) described a "step-down osteotomy" in which a Z-osteotomy twice as long as the required shortening was made. Opposing blocks were then taken from each side of the osteotomy to shorten the metatarsal and provide a Z-shape whihc would be reasonably stable and easy to fix (with chromic catgut in his series). Of 40 procedures, 33 were rated excellent. There were 4 transfer lesions, and 3 procedures were considered failures - in 2 cases because they had not been shortened enough.
Spence (1990) simply excised 5mm from the proximal shaft. 89% of their 54 patients had satisfactory results. 18% had transfer lesions and 7% of keratoses recurred. Results were poorer in patients who had simultaneous hallux valgus correction.
Coughlin and Mann described an oblique sliding osteotomy of the metatarsal shaft in their textbook of foot and ankle surgery. A clinical series was reported by Kennedy (2006) with follow-up of 2-6y. 32 patients had 42 osteotomies, and 27 also had hallux valgus correction. Adjustment of length and metatarsal height was judged intra-operatively and the osteotomy fixed with trans-osseous wire loops. Mean shortening was 3.4mm (1-5mm). Mean AOFAS lesser ray score improved from 50 to 82. 31/32 patients were "relieved" and satisfied. The osteotomy was slow to unite with a mean 10 weeks and a maximum of 15 weeks. Our experience is that this is a very fiddly osteotomy to control and stabilise.
Lauf (1996) described a variation of the long oblique osteotomy with a short near-transverse arm to increse stability, and screw fixation. A small amount could be removed from the transverse arm to allow shortening. 30 patients with 40 osteotomies were followed up for 12-18 months. There were two delayed unions, no transfer lesions or recurrent keratoses. However, it appears that only 12 of the 30 patients were fully followed-up, of whom 10 rated their satisfaction 8/10 or better and 11 would have the operation again.
Galluch (2007) described a techinque in which a segment was removed from the midshaft of the metatarsal and the osteotomy fixed with a compression plate. One osteotomy out of 126 failed to unite. All but one of the 95 patients in the series had other procedures which might improve their metatarsalgia, such as hallux valgus corrections, toe straightening and Achilles lengthenings, so the authors chose not to report the clinical results, arguing that the effect of the metatarsal osteotomies could not be isolated.
There is not enough evidence to give clear evidence-based recommendations. In our practice we differentiate between procedures aimed at metatarsal shortening and those aimed at raising a metatarsal head. Generally, patients with long 2nd or 3rd metatarsals would be offered Weil osteotomies and those with a plantarflexed metatarsal a modified Weil or BRT osteotomy. A short 1st ray may be improved with a scarf osteotomy or 1st MTP fusion with or without an intercalated bone graft.