FAIXXX10.1177/1071100717737481Foot & Ankle InternationalDa Cunha et al 737481research-article2017 Case Report Foot & Ankle International® 2018, Vol. 39(2) 242 –249 Distraction Osteogenesis and Fusion © The Author(s) 2017 Reprints and permissions: sagepub.com/journalsPermissions.nav for Failed First Metatarsophalangeal DOI:https://doi.org/10.1177/1071100717737481 10.1177/1071100717737481 Joint Replacement: Case Series journals.sagepub.com/home/fai Rachael J. Da Cunha, MD, FRCSC1, Sydney C. Karnovsky, BA1, Austin T. Fragomen, MD2, and Mark C. Drakos, MD1 Level of Evidence: Level V, expert opinion Keywords: failed MTP joint arthroplasty, distraction osteogenesis, first MT joint fusion Introduction arthroplasty and arthrodesis over a 6-year period, 24% of the arthroplasties failed, with 4 converted to an arthrode- Hallux rigidus is a progressive osteoarthritic condition, 8 6 sis to alleviate pain. Failed implants, once removed, cre- and operative intervention is often required. It can affect ate significant bone loss, which can make the gait, lead to a decreased range of motion, particularly dor- 1 5,6 reconstruction operation difficult. Because of the bone siflexion, and can cause stiffness and pain. First metatar- loss, the revision arthrodesis usually necessitates bone sophalangeal (MTP) arthrodesis is the traditional standard 2 1 grafting in order to restore first ray length. Depending on treatment for end-stage hallux rigidus and has been the amount of bone loss, local autologous bone graft may repeatedly shown to be the most consistent and successful 10 be an option, but in cases with significant loss, grafts operative technique. In 1952, first MTP arthroplasty from other areas might be necessary as well, increasing emerged as a new treatment option for end-stage hallux 5 3,6 the potential rate of failure of the operation. Commonly, rigidus. surgeons use tricortical iliac crest autograft wedges to First MTP arthroplasty involves prosthetic replace- manage defects. However, this can have associated mor- ment of the first MTP joint with either a unipolar or bipo- bidity in terms of pain and even fracture.12 lar implant and has the potential to improve joint motion 3 These salvage arthrodesis procedures often have a long and reduce pain. However, arthroplasty involves addi- time to union and increased rates of nonunion and malunion tional risks, including malposition, implant fracture, compared to primary first MTP arthrodesis procedures.2 To stress fracture, arthrofibrosis, and synovitis, all of which 3 avoid the complications associated with iliac crest bone graft can lead to failure. Additionally, failed arthroplasty often harvest, tricortical allograft wedges have also been used. leads to a significant amount of first ray shortening, which 1 However, the nonunion rate associated with allograft in foot can make revision surgery a challenge. There is little lit- arthrodesis cases has been reported to range from 9% to 23% erature following the long-term results of hemiarthro- 7,8 6 and is thus a less favorable option. To improve the chance plasty and total joint arthroplasty, and even less exists of healing, many surgeons elect to acutely shorten the first describing outcomes of salvage arthrodesis in cases where arthroplasty fails.5 Brage and Ball reported that many of 2 the hemiarthroplasty implants would fail. Delman et al 1Department of Orthopedic Surgery, Foot and Ankle, Hospital for also reported that arthroplasty procedures to treat hallux Special Surgery, New York, NY, USA 2 rigidus would fail, opining that the compressive and shear Department of Orthopedic Surgery, Limb Lengthening and Complex stresses placed on the implant surfaces put them at high Reconstruction Service, Hospital for Special Surgery, New York, NY, USA risk of loosening.3 Furthermore, unlike the knee and hip, there is little surface area available to establish a strong Corresponding Author: bone-implant interface. This, combined with higher Rachael J. Da Cunha, MD, FRCSC, Clinical Fellow, Department of Orthopedic Surgery, Hospital for Special Surgery, 535 East 70th Street, stresses, particularly during walking and stair-climbing, New York, NY 10021, USA. leads to higher failure rates. In a study comparing Email: [email protected] Da Cunha et al 243 Table 1. Patient Data. packed into and around the fusion site. Satisfactory align- ment was obtained and the joint was pinned in position with Patient 1 Patient 2 Patient 3 1 to 2 K-wires. Age 55 67 48 The first ray was confirmed to be short by 10 to 25 mm Gender Female Female Male and indicated for lengthening. The Biomet mini rail exter- Type of implant Unipolar Unipolar Bipolar nal fixation system (Warsaw, IN) was utilized to construct a Defect size, mm 10 24 24 dorsally mounted compression-distraction frame. Three- Infection None None None millimeter self-drilling, self-tapping half pins were placed Implant loose Yes No Yes percutaneously with fluoroscopic assistance. The most Bacteria isolated None None None proximal pin in the proximal first metatarsal was placed first, followed by the most distal pin in the proximal pha- lanx establishing the axis of compression and distraction. metatarsal defect to achieve fusion without addressing the The remaining 4 half-pins were placed using the frame as a loss of length of the ray. We present a series of failed first targeting device. All pins were placed bicortically. Overall, MTP arthroplasties requiring revision-reconstruction opera- there were 2 pins in the proximal phalanx, 2 pins in the dis- tions using a novel application of the Ilizarov technique, in tal first metatarsal, and 2 pins in the proximal first metatar- which distraction osteogenesis was used in combination sal. The pins were oriented to allow compression across the with primary arthrodesis to restore stability and length to the first MPT joint and distraction through the proximal meta- first ray. In our experience, this technique leads to good tarsal. Once satisfactory position of all pins was confirmed, results and minimized morbidity to the iliac crest. the frame was removed, and the mid-diaphysis of the first metatarsal was exposed through a small dorsomedial inci- Methods sion. Multiple drill holes were made at the desired osteot- omy site in the middle of the first metatarsal, and the A retrospective chart review was conducted from a prospec- osteotomy was completed using an osteotome. The external tively collected registry to identify fusion after joint arthro- fixator was reattached, and compression was applied across plasty surgery. We identified all cases of a failed first MTP the first MTP joint fusion site. The K-wire was kept in some joint arthroplasty that were revised to a first MTP joint cases across the joint to aid in maintaining the joint reduc- arthrodesis with first metatarsal lengthening between 2011 tion and preventing apex plantar deformity during and 2014. We excluded patients with concomitant foot compression. pathology including nonunion. Four patients met the inclu- sion criteria. One patient was excluded because of preoper- Postoperative Protocol ative complex regional pain syndrome (CRPS). Thus, 3 patients were included in this case series. Baseline patient The patient was initially instructed to remain nonweight- demographics and intraoperative findings are summarized bearing with the operative foot in a postoperative splint. (Table 1). Empiric oral antibiotics were prescribed for infection pre- vention prophylaxis, a practice we no longer advocate.4 Chemical deep vein thrombosis prophylaxis was also pre- Operative Technique scribed in each case. On postoperative day 5, distraction of The anesthetic protocol was similar for all patients, consist- the osteotomy was initiated at a rate of 0.5 mm per day, split ing of a peripheral popliteal nerve block under ultrasound into two 0.25-mm turns. The technical challenge was to guidance and spinal anesthesia. Antibiotics were held until lengthen the metatarsal while keeping the Meary angle con- tissue samples were obtained for culture. A dorsal incision stant with appropriate dorsiflexion of the first MTP arthrod- was made, centered over the first MTP joint. In 2 of the 3 esis. The frame was parallel to the first metatarsal on both cases, the implants were loose. When the implant was the AP and lateral view in order to maintain the position of removed, all devitalized or necrotic tissue was removed, first MTP fusion (Figure 1B and C). The patient was including bone. Bone defects measured, on average, 19 mm assessed at 2-week intervals postoperatively to evaluate (range 10-24 mm). The Wright Medical (Memphis, TN) progress of the distraction regenerate bone. Additional conical and cup reamer set was used to contour the joint compression (1-mm) was applied with the external fixator surfaces for arthrodesis. A 3.5-mm Kirschner wire (K-wire) across the fusion site at these visits. Distraction was contin- was used to penetrate each joint surface several times to ued at this rate until the metatarsal length was restored to improve bony integration at the fusion interface. Cancellous within 1 mm of the contralateral side. Adequate distraction bone autograft was harvested from the calcaneus. Between and restoration of metatarsal length was typically achieved 3 and 7 cc of bone graft was harvested. The autograft was at 4 to 6 weeks postoperation. The patient was then permit- used to fill any defects in the metatarsal head as well as ted partial heel weightbearing only in a postoperative rigid 244 Foot & Ankle International 39(2) Table 2. Results. Time NWB Time to to Partial Heel Time to Full Final Union, Weightbearing, Weightbearing, Follow-up, Pain at Final Patient Age mo Reoperation Complications wk mo mo Follow-up Final Function 1 55 5 No None 3 5 6 No Return to all activities 2 67 4.5 No Superficial pin 2 4.5 9 No Return to all site infection activities, including jogging 3 48 3.5 Yes Premature 3 3.5 8 Mild Returned to consolidation work as requiring a laborer revision with rocker osteotomy bottom shoe Abbreviation: NWB, nonweightbearing.