COMPLICATIONS OF HALLUX VALGUS SURGERY. MANAGEMENT OF THE FIRST SHORT METATARSAL AND THE FAILED RESECTION ARTHROPLASTY
July 14th, 1997
Verônica F. Vianna, MD and Mark S. Myerson, MD
Failure of hallux valgus surgery can be a complicated and frustrating process for both the patient and surgeon. In addition to problems of the soft tissues such as scar contracture and neuromas, bone loss presents one of the more challenging issues in reconstructive forefoot surgery.
In this chapter bone loss of the first ray is covered by evaluating avascular necrosis, infection, the failed Keller and silastic arthroplasty, as well as the short first metatarsal.
In each situation the presentation varies, however the goal is to reestablish the length and function of the first ray. Although arthrodesis of the hallux metatarsophalangeal joint is usually necessary, steps should always be taken to preserve motion at the metatarsophalangeal joint.
Avascular necrosis of the first metatarsal after distal first metatarsal osteotomy is an uncommon complication.4,10,11,26,33 Although some avascularity occurs as a result of interruption of the blood supply to the metatarsal head after a distal osteotomy, this does not present a clinical problem provided the blood supply around the metatarsal head remains intact, particularly in its dorsal and lateral aspects.18,29 The presence of simple bone cysts in the absence of collapse and fragmentation of the metatarsal head should be differentiated from true avascular necrosis. These small cysts which for example may be present after chevron osteotomy are rarely associated with clinical symptoms of arthritis and typically are not progressive.31
When either complete loss of the metatarsal head or partial collapse of the metatarsal head occurs patients will present with pain and swelling of the metatarsophalangeal joint associated with varying degrees of instability, shortening of the metatarsal and transfer of weight-bearing to the second metatarsal.
Rest with limitation of activity, an orthotic support, use of antiinflammatory agents and intraarticular steroid injection may improve symptoms. Steroid injection must be used judiciously since the effect of intraarticular steroid in the setting of established avascular necrosis is not clear. If nonoperative methods fail to relieve the symptoms, a metatarsophalangeal arthrodesis is usually indicated.18,19 Other methods of joint decompression with synovectomy and drilling of the metatarsal head may suffice for very early stage avascular necrosis, but if any fragmentation or collapse is already present this will not work.
When the arthrodesis is carried out the avascular portion of the metatarsal should be excised. Regardless of the extent of this bone loss, some authors recommend an "in situ" arthrodesis and encourage the patient to leave the toe somewhat shortened.19 An alternative would be arthrodesis with interpositional bone graft to restore length of the first ray and therefore the ability of the hallux to bear weight.29 The choice of procedure is sometimes difficult, particularly if bone loss is minimal. An in situ arthrodesis is technically easier to perform, the fusion rate is more reliable and the morbidity from the bone graft is avoided. However, adequate bleeding bone surfaces may be difficult to obtain in the setting of avascular necrosis without excessive shortening of the metatarsal. One should therefore have the option available for an interposition graft depending on the extent of avascular necrosis present or anticipated.
The Short First Metatarsal
Shortening occurs after most metatarsal osteotomies.19 However, there are certain procedures that cause more shortening than others including the Mitchell bunionectomy, closing wedge osteotomy of the base of the first metatarsal, and the Wilson type osteotomy.27 These complications should be anticipated and a metatarsal osteotomy which is known to cause shortening should not be selected for a patient in whom the first metatarsal is already short or the second metatarsal relatively long.. The main problem associated with shortening of the first ray is a transfer of weight to the second metatarsal ultimately causing callosity and pain.. It is difficult to determine the amount of shortening that will produce metatarsalgia, since it is probably related to a number of factors including the initial length of the first and second metatarsals, whether any accompanying dorsiflexion is associated with the shortening, the stability of the metatarsophalangeal joint, and the presence of hypermobility of the first ray.
Once shortening of the first metatarsal and metatarsalgia occurs, and the patient is not comfortable with conservative management, surgical correction may be required. Initial treatment should be directed toward alleviating the pain of weight-bearing and redistributing the weight from the second metatarsal by the use of an orthosis with a Morton’s extension and a metatarsal pad. However, surgical correction of this condition with second metatarsal osteotomy is difficult and the results are often less than optimal.19
Lengthening of the metatarsal may be considered only if the metatarsophalangeal joint is congruent without arthritis. The lengthening of the first metatarsal is difficult to achieve technically.19 It generally causes a contracture of the soft tissues about the hallux metatarsophalangeal joint which will likely limit motion even when a gradual distraction using a mini external fixator is performed.2,17 Our experience with lengthening of the first metatarsal through distraction osteogenesis has been frustrating. Although we have been able to achieve 10 to 16 mm of length reliably, the period of immobilization, frequent pin problems, and hallux stiffness make this less appealing. Lengthening is a complex process and each patient should be carefully selected for this procedure.
If the second metatarsal is significantly longer than the first and third, shortening of the second metatarsal to bring it back into a normal weight-bearing alignment can be beneficial. Occasionally a plantar flexion osteotomy combined with slight lengthening of the first metatarsal can be used to increase its weight-bearing ability.19
When shortening is caused by a delayed or nonunion of a metatarsal osteotomy then a single stage lengthening with interposition bone graft will work. This depends on the mobility of the hallux metatarsophalangeal joint, and the amount of bone loss present. We have successfully lengthened the metatarsal as a single stage procedure with structural bone graft. The maximum recommended single stage lengthening in our experience has been 15 mm.
Rarely, arthrodesis of the metatarsophalangeal joint may be required to restore the weight-bearing function of the medial ray. This is however less desirable since motion is lost and this procedure should only be selected if the metatarsophalangeal joint is dysfunctional or arthritic.
Failed silastic implant arthroplasty. Although some early studies demonstrated good results after hallux implant arthroplasty,15 there are many concerns regarding the longterm outcome and frequency of complications after this procedure, and later studies with longer followup show less uniform results.25,30 The reasons for failure have been multiple including: implant fracture, silicone synovitis, stress fracture, transfer metatarsalgia, wound infection, and bone resorption at the implant-bone interface. 13,15,16,28,30
If failure is established but the only complaint is pain and swelling about the metatarsophalangeal joint then removal of the prosthesis and synovectomy may be all that is required. This operation has the advantage of simplicity. The hallux metatarsophalangeal alignment in the transverse plane is maintained and the acute inflammation subsides.14 Although metatarsophalangeal motion, toe-off strength and overall function of the hallux are compromised, such limited function may be adequate for selected patients.
If the hallux is already short, deformity is present, or there is pain under the second metatarsal an attempt of salvage may be made with a more extensive procedure.1,14,19 The timing of this procedure is important. Once bone resorption, lysis and cyst formation occur with pain, the implant should be removed. Active silastic synovitis will cause further bone loss and delaying implant removal is not ideal. At times, simple removal of the prosthesis with curettage may be sufficient to alleviate pain, although strength and function are impaired. If arthrodesis is performed, a decision will have to be made whether or not to stage the procedure. It may be preferable to pack the metatarsal and phalangeal cavities with cancellous bone graft, delaying the arthrodesis until the graft has been incorporated, improving the bone strength and likelihood of fusion. The arthrodesis is then performed either "in situ" or with a tricortical interpositon bone graft, depending on the length of the hallux.
Failed resection arthroplasty. Excisional arthroplasty as a treatment for hallux valgus was described by Keller in 1904.12 Postoperative metatarsalgia is probably the most frequent reported complication that is associated with this procedure.21 Other complications include : excessive shortening of the hallux, cock-up deformity, recurrent hallux valgus, narrowing of the metatarsophalangeal joint space with time, impaired control and function of the hallux, weakness of toe-off and reduced motion of the interphalangeal joint. It has been shown that inability of the hallux to bear weight correlates directly with the development of lateral metatarsalgia.21 Excessive resection (more than one third of the phalanx) leads to a flail toe with a decreased capacity of weight-bearing and resultant metatarsalgia. Failure of the procedure leaves only limited options for salvage, which depend largely on the extent of the excisional arthroplasty that was done.
Successful arthrodesis is more difficult to achieve when a previous resection of bone has been done.32 Attempted salvage by arthrodesis of a failed result of resection arthroplasty has been reported in the literature with a 82% rate of successful fusion.6
When the hallux is short, toe-off is weakened and pain exists under the second metatarsal, an arthrodesis of the metatarsophalangeal joint may be performed. Occasionally, sufficient bone is present and arthrodesis can be accomplished without interpositioning bone graft, but because shortening usually occurs, an interposition structural bone graft should be used to improve function of the hallux.
Although the incidence of infection after bunion surgery is low,8,11,21 treatment of an infected first metatarsophalangeal joint presents the surgeon with many challenges since it involves concomitant treatment of bone and soft-tissue infection, bone loss and instability.
Treatment of this condition includes open drainage, rest, antibiotic agents and debridement with removal of infected tissue creating a resection arthroplasty. This may lead to a short metatarsal and a claw hallux with decrease in weight-bearing function of the first ray and subsequent lateral metatarsalgia.
Reconstructive techniques that address the bone loss by maintenance of the first metatarsal length and stability of the first metatarsophalangeal joint are needed to restore proper weight distribution of the foot.
One method proposed to solve this problem follows the principles of salvage techniques in total knee arthroplasty complicated by infection. It consists of a two-stage arthrodesis performed to provide a stable first ray that can participate in weight-bearing and to eradicate infection.23 The initial stage consists of an aggressive debridement with removal of necrotic bone and the use of a spacer of antibiotic-impregnated bone cement placed in the joint to maintain soft-tissue tension and length. Typically we use 1 g of Vancomycin and 1 g of Tobramycin powder mixed with one packet of cement. Although K-wires may be used to stabilize the hallux further, these are not always necessary. If they are used, they are inserted retrograde across the hallux and anterograde through the cement while still doughy in texture. The use of an antibiotic impregnated cement spacer in the first stage of reconstruction results in a rapid reduction in the symptoms of inflammation at the joint. The fixation of the spacer with pins possibly enhances healing and helps decrease inflammation by adding rigidity. After 3 to 5 weeks, a healthy bed of granulation tissue forms around the spacer and the second stage is performed. We have reported previously on this procedure with the use of this technique in 5 patients with a mean followup of 28 months. No recurrent infection was reported. Four patients had successful arthrodesis and one went on to an asymptomatic nonunion.23
Arthrodesis of the metatarsophalangeal joint with interposition bone graft
The patient is positioned supine on the operating table. General, spinal or ankle block anesthesia may be used largely determined by the type of bone graft to be used.. When utilizing autograft structural iliac crest, either general or spinal anesthesia is used. With the use of allograft local anesthesia is usually sufficient and we prefer, whenever possible, the use of ankle block with or without sedation.24 A tourniquet is not used, and the visible bleeding is particularly useful during bone debridement. Furthermore the blanching of the skin during distraction is evident and skin slough can be prevented in this manner.
The metatarsophalangeal joint is approached preferably through a longitudinal dorsal incision, centered over the metatarsophalangeal joint. This approach is easier, and with dissection the medial digital nerves, which are often difficult to distinguish due to scarring from previous procedures, are avoided. The dorsal approach also facilitates the lengthening of the extensor hallucis longus, the section of the extensor hallucis brevis and a more extensive circumferential release around the proximal phalanx and the metatarsal head. Although prior incisions may have some relevance to the current incision selected, we have not observed any wound complication when selecting the dorsal midline incision. It might be necessary to excise a retracted and/or hypertrophic scar. By and large, the previous incisions may be ignored, particularly if they are more than 6 months old. Also, when planning the incision, it should be kept in mind that with surgery the hallux will be lengthened and, because of that, wound closure almost always constitutes a problem. Whenever wound slough occurs it is usually due to skin tension. In order to prevent this problem it may be necessary to use a skin Z-plasty.22
The extensor hallucis longus tendon can be reflected medially or laterally, depending upon the deformity of the metatarsophalangeal joint. Generally, however, this tendon is lengthened in a standard Z-step cut fashion technique leaving enough tendon length for an overlap of 1.0 to 1.5 cm of the tendons ends. The proximal cut to the tendon is medial and the distal cut is lateral and in this manner the distal cut is extended laterally to incise the extensor hallucis brevis. The capsule and extensor hallucis brevis are cut transversely over the joint and medial and lateral full thickness subperiosteal flaps are created. An aggressive debridement is carried out with removal of any prosthetic material, debris and synovium. All scar around the phalanx that would prevent the lengthening of the hallux is removed and inflammatory cysts and all necrotic bone are debrided using currettes, rongeurs and oval burrs. A larger burr (4 to 5 mm) is preferred since the smaller diameter burrs may cause cortical perforation. Debridement and burring are performed until punctate bleeding is present preserving as much bone as possible. A lamina spreader is then inserted and the joint distracted.. While it is in place a radiograph is obtained and the length and the shape of the graft as well as the position of the hallux are determined.
We select the type of bone graft, ie autograft or allograft based on the patient’s decision with appropriate informed consent. Even though allografts undergo biological changes that are qualitatively similar, but usually quantitatively and temporally inferior to autografts, these changes are usually compatible with success of the graft. Furthermore, the loss of biological potential is often well justified by averting morbidity at the donor site.9 We have reviewed retrospectively the results of large structural and cancellous allografts and autografts and found no significant differences with respect to either the rate or success of arthrodesis (Toomey SD, Myerson MS, Trnka HJ: Comparison of the results of large structural/cancellous allograft/autograft bone grafts in foot and ankle surgery. To be presented at the 28th Annual Winter Meeting of the American Orthopaedic Foot and Ankle Society, New Orleans, 1998)
There is usually a need for structural graft. When a choice is made for autograft we most often use tricortical iliac crest. However, if at the same surgical procedure the lesser toe metatarsal heads are being resected they might be used as the graft, but it constitutes basically cancellous and not structural graft.
In terms of allograft we prefer the versatility of the femoral head. It is easy to contour and its rounded surface adapts well to the dorsal surface of the distal metatarsal. It is a unicortical graft, but seems to work well as a structural graft. The graft is contoured using a microsagittal saw. In addition to the structural bone graft, cancellous bone graft is packed into the void created by the bone debridement.
The graft must be fashioned to fit the defect and positioned in place while the surgeon manually and longitudinally distracts the hallux. When shaping the graft one must be creative since the defect rarely is symmetric. A smooth lamina spreader (one with the teeth removed previously) promoting over distraction might help when inserting the graft. Sometimes it is useful to plantarflex the hallux, position the graft and then bring it up while distraction is applied. With the graft in place the amount of lengthening, position of the hallux, the condition of the skin and the circulation to the toe must be checked.
A decision now must be made regarding the type of fixation. It will vary according to the size and the shape of bone graft and the quality of the bone. The main options are: threaded Steinmann pins or larger K-wires, cancellous screws, a dorsal plate or a combination of these. When using threaded Steinmann pins, at least 2 but if possible 3 pins are inserted in a retrograde fashion from the base of the proximal phalanx out the distal end of the toe. Once the graft is positioned in place the pins are inserted across the graft into the cortex of the metatarsal shaft. While the pins are being inserted it is important that the surgeon manually hold the hallux in the desired position. Sometimes it is hard to accomplish with this fixation a good position of the toe without having the pins coming out through the plantar aspect of the metatarsal, as the hallux is positioned in a slight dorsiflexion. To go back and forward repeatedly trying to find a better position for the pins might compromise the graft. If the graft is not too long the use of a 4.0 mm fully threaded cancellous screws in a crossed fashion may be sufficient. The difficulty is using screws in a revision procedure relies on the fact that both the phalangeal and the metatarsal metaphyseal flares are usually absent leaving no place to accommodate the screws. Occasionally a dorsal plate combined with Steinmann pins or with interfragmentary screws may be used. Biomechanically the dorsal surface is not ideal to position a plate at the metatarsophalangeal joint. The dorsal surface is the compression side of the joint and ideally the plate should be positioned at its tension side. Biomechanical studies comparing the strength and rigidity of different methods of internal fixation for arthrodesis of the first metatarsophalangeal joint confirmed that the use of a single interfragmentary screw produces a more stable construct than a plate on the compression side of the joint.7 There have also been reports of plate breakage and a high rate of plate removal due to subcutaneous bulk what makes this less useful. These problems may be minimized by the use of a small vitallium plate. The increase strength of the vitallium and the absence of screw holes at the fusion site add strength to the plate construct as well as the low profile nature of it seems to make it less bothersome to patients after surgery. Coughlin and Abdo5 have shown 98% of successful fusion rate in 58 cases of primary metatarsophalangeal joint fusion using a vitallium small plate, but no information is available on more extensive revision procedures where a longer plate is required.
The position of the hallux it is not easy to establish when performing a bone graft interposition fusion. The ideal position of the first metatarsophalangeal joint arthrodesis must take into account the sagittal and transverse planes as well as rotation in the frontal plane. The desired position of the axis of the proximal phalanx in relation to the axis of the first metatarsal shaft is 20 to 30° of dorsiflexion - although ideally the hallux should remain parallel with the floor, 15 to 20° of valgus and neutral rotation.20 A clinically useful method of assessing sagittal plane position is to obtain a simulated loadbearing position in the operating room and determine the relationship of the hallux to the "floor". Pressure is applied to the sole of the foot on a flat surface, the interphalangeal joint is extended and the head of the proximal phalanx should clear the surface by 5 to 10 mm. The dorsal aspect of the hallux should be parallel to the flat surface.3
After surgery, patients are kept nonweight-bearing for 4 to 6 weeks. Then a cast with a pedestal heel weight-bear is used for more 4 to 6 weeks or until fusion is determined. The parameters to determine the presence of fusion are absence of warmth, swelling and pain at the arthrodesis site as well as radiographic parameters, being the latter the least reliable.
1. Alexander IJ: Arthrodesis of the metatarsophalangeal and interphalangeal joints of the hallux. In M Myerson (ed.): Current Therapy in Foot and Ankle Surgery, St. Louis, Mosby-Year Book Inc, 1993, p 81
2. Amillo S, Gil-Albarova J, Pampliega T: Lengthening of the first metatarsal bone. A case of congenital shortening. Acta Orthop Scand 62:295, 1991
3. Conti SF, Dhawan S: Arthrodesis of the first metatarsophalangeal and interphalangeal joints of the foot. Foot Ankle Clin 1:33, 1996
4. Corless JR: A modification of the Mitchell procedure [abstract]. J Bone Joint Surg 58B:138, 1976
5. Coughlin MJ, Abdo RV: Arthrodesis of the first metatarsophalangeal joint with Vitallium plate fixation. Foot Ankle Int 15:18, 1994
6. Coughlin MJ, Mann RA: Arthrodesis of the first metatarsophalangeal joint as salvage for the failed Keller procedure. J Bone Joint Surg 69A:68, 1987
7. Curtis MJ, Myerson M, Jinnah RH, et al: Arthrodesis of the first metatarsophalangeal joint: a biomechanical study of internal fixation techniques. Foot Ankle 14:395, 1993
8. Fitzgerald JA: A review of long-term results of arthrodesis of the first metatarso-phalangeal joint. J Bone Joint Surg 51B:488, 1969
9. Friedlaender GE: Bone grafts. The basic science rationale for clinical applications. J Bone Joint Surg 69A:786, 1987
10. Green MA, Dorris MF, Baessler TP, et al: Avascular necrosis following distal chevron osteotomy of the first metatarsal. J Foot Ankle Surg 32:617, 1993
11. Hattrup SJ, Johnson KA: Chevron osteotomy: analysis of factors in patients' dissatisfaction. Foot Ankle 5:327, 1985
12. Keller WL: The surgical treatment of bunions and hallux valgus. NY Med J 80:741, 1904
13. Kitaoka HB, Cracchiolo A, III: Stress fracture of the lateral metatarsals following double-stem silicone implant arthroplasty of the hallux metatarsophalangeal joint. Clin Orthop 239:211, 1989
14. Kitaoka HB, Holiday AD, Jr., Chao EYS, et al: Salvage of failed first metatarsophalangeal joint implant arthroplasty by implant removal and synovectomy: clinical and biomechanical evaluation. Foot Ankle 13:243, 1992
15. Laird L: Silastic joint arthroplasty of the great toe. A review of 228 implants using the double-stemmed implant. Clin Orthop 255:268, 1990
16. Lemon RA, Engber WD, McBeath AA: A complication of Silastic hemiarthroplasty in bunion surgery. Foot Ankle 4:262, 1984
17. Magnan B, Bragantini A, Regis D, et al: Metatarsal lengthening by callotasis during the growth phase. J Bone Joint Surg 77B:602, 1995
18. Mann RA: Avascular necrosis of the metatarsal head following a chevron osteotomy. Foot Ankle 3:125, 1982
19. Mann RA, Coughlin MJ, eds: Surgery of the Foot and Ankle, ed. 6. St. Louis, Mosby-Year Book Inc, 1993
20. Mann RA, Oates JC: Arthrodesis of the first metatarsophalangeal joint. Foot Ankle 1:159, 1980
21. Mann RA, Rudicel S, Graves SC: Repair of hallux valgus with a distal soft-tissue procedure and proximal metatarsal osteotomy. A long-term follow-up. J Bone Joint Surg 74A:124, 1992
22. Myerson MS, Fortin P, Girard P: Use of skin z-plasty for management of extension contracture in recurrent claw- and hammertoe deformity. Foot Ankle Int 15:209, 1994
23. Myerson MS, Miller S, Saxby T, et al: Staged arthrodesis for salvage of the septic hallux metatarsophalangeal joint. Clin Orthop 307:174, 1994
24. Myerson MS, Ruland CM, Allon SM: Regional anesthesia for foot and ankle surgery. Foot Ankle 13:282, 1992
25. Papagelopoulos PJ, Kitaoka HB, Ilstrup DM: Survivorship analysis of implant arthroplasty for the first metatarsophalangeal joint. Clin Orthop 302:164, 1994
26. Peterson DA, Zilberfarb JL, Greene MA, et al: Avascular necrosis of the first metatarsal head: incidence in distal osteotomy combined with lateral soft tissue release. Foot Ankle Int 15:59, 1994
27. Pring DJ, Coombs RRH, Closok JK: Chevron or Wilson osteotomy: a comparison and follow-up [abstr]. J Bone Joint Surg 67B:671, 1985
28. Sammarco GJ, Tabatowski K: Silicone lymphadenopathy associated with failed prosthesis of the hallux: a case report and literature review. Foot Ankle 13:273, 1992
29. Shereff MJ: Chevron osteotomy for hallux valgus. In M Myerson (ed.): Current Therapy in Foot and Ankle Surgery, St. Louis, Mosby-Year Book Inc, 1993, p 46
30. Shereff MJ, Jahss MH: Complications of silastic implant arthroplasty in the hallux. Foot Ankle 1:95, 1980
31. Thomas RL, Espinosa FJ, Richardson EG: Radiographic changes in the first metatarsal head after distal chevron osteotomy combined with lateral release through a plantar approach [see comments]. Foot Ankle Int 15:285, 1994
32. Wilson JN: Arthrodesis of the first metatarsophalangeal joint. J Bone Joint Surg 49B:98, 1967
33. Wu KK: Mitchell bunionectomy: an analysis of four hundred and thirty personal cases plus a review of the literature. J Foot Surg 26:277, 1987
Fig. 1. This patient sustained complete loss of the first metatarsal following infected avascular necrosis. Lengthening was performed in two stages , the first of which is demonstrated here with the mini external fixator.
Fig. 2. Nonunion of first metatarsal was associated with shortening(A-B). This was treated with a single stage lengthening using unicortical allograft with plate and screw fixation(C-D-E).
Fig. 3 This 42 years old male presented with pain, weakness and instability of the hallux two years following silastic arthroplasty (A). A unicortical allograft strut was inserted and secured with 2 threaded Steinmann pins (B-C). Despite nonunion of the distal bone graft site this patient is asymptomatic 3 years later (D-E).
Fig. 4. This patient is a 44 years old male who presented with acute osteomyelitis of the first metatarsal following a distal metatarsal osteotomy. This was treated during the first stage with antibiotic impregnated with bone cement and mini external fixator (A). At 6 weeks the cement was removed and a tricortical iliac crest autograft inserted using threaded Steinmann pins(B). Arthrodesis occurred at 4 months at which time the pins were removed. Shown here are the AP and lateral radiographs four years following arthrodesis(C-D).
Fig. 5 This patient presented with severe pain and instability of the hallux following silastic arthroplasty. Insufficient bone was present on the proximal phalanx to successfully perform a bone block arthrodesis(A). The revision of the metatarsal head resection was performed and metatarsal bone used to stabilize the hallux (B). Although an arthrodesis was not obtained the hallux metatarsophalangeal joint was stiffened sufficiently as well as lengthened to give this patient considerable functional improvement(C).
Fig. 6. In this patient the interposition graft was secured with a combination of two Steinmann pins and a dorsal plate (A-B)
Fig. 7. This patient developed avascular necrosis of the first metatarsal head following distal metatarsal osteotomy . She had previously undergone osteotomies of the fourth and fifth metatarsals with shortening. Note shortening associated with a cock-up deformity of the hallux(A-B). This was treated with a tricortical autogenous iliac crest graft shaped to fit the defect . In this patient a combination of screws and threaded K-wires was sufficient for stable arthrodesis (C).