Midfoot arthritis, consisting of the tarsometatarsal (TMT) and the midtarsal joints, continues to present a challenging problem to orthopedists as a significant cause of foot pain, deformity, and progressive disability. Whereas traditional operative management included fusion of both the TMT joints of the medial and the middle column and naviculocuneiform joints, selective fusion of the involved joints only is currently gaining in popularity to preserve unaffected joints and the native anatomy.1
The etiology of midfoot arthritis varies widely, with posttraumatic arthritis and degenerative arthritis cited as the most common causes.2,3 There are approximately 55,000 midfoot injuries per year, primarily affecting patients in their third and fourth decades and frequently associated with athletic injuries and motor vehicle collisions.2,4 Patients with symptomatic degenerative arthritis frequently present in their sixth or seventh decades, report no history of trauma, and are predominantly women.5 Additional causes of midfoot pain and arthritis include inflammatory arthropathies, diabetic neuroarthropathy, and more recently postulated first-ray hypermobility and medial column instability.4,6 The focus of this work is patients who develop primary osteoarthritis (OA) of the midfoot without a history of trauma, neuropathy, or inflammatory arthritis.
Irrespective of the etiology, midfoot arthritis presents with increasing foot pain, difficulty ambulating, swelling, and limitations in activities of daily living or recreational activities. Patients may also complain of increasing deformity and difficulty with foot wear.2,4 Deformities associated with TMT joint arthritis include midfoot collapse and loss of the longitudinal arch with forefoot varus, forefoot abduction, and hindfoot valgus.3,7 Radiographically, patients with midfoot arthritis demonstrate joint space narrowing of the TMT joints, dorsal osteophytes, and a low arch.8
A recent study has demonstrated that mechanical stress and increased plantar loading may contribute to foot pain and symptoms in patients with midfoot OA. Biomechanically, patients with pes planovalgus and midfoot OA exhibit higher medial midfoot pressures and mechanical overloading of the midfoot.9 Further research has suggested that excessive range of motion of the medial column causes a failure of the windlass mechanism and overloads the midfoot more laterally, leading to the speculation that the loss of stability of the medial column contributes to midfoot collapse, a planovalgus foot, and transfer metatarsalgia.6,10,11 In addition, patients with midfoot OA have both clinically and radiographically flatter feet and greater loading of the midfoot while walking.12
Patients with midfoot OA, with or without a history of trauma, commonly report restrictions in activities of daily living, restrictions of walking, deformity about the midfoot, and difficulty with foot wear.5 Initial treatment of symptomatic midfoot OA consists of anti-inflammatory medication, activity modification, change in foot wear (stiff-soled shoes), orthoses, and injections.4 Full-length orthoses have been used to limit motion at the TMT joints, restore the arch height, and provide pain relief.8 Patients who fail nonoperative treatment are candidates for operative intervention, with fusion being the mainstay of treatment.
Surgical management of midfoot OA most commonly entails midfoot fusion, as the TMT joints can be fused with minimal functional impairment.4 Goals of fusion include anatomic realignment, creating a stable plantigrade foot, and clinical and radiographic evidence of fusion.4,13
INDICATIONS AND CONTRAINDICATIONS
Evaluation of patients with midfoot arthritis can be challenging. Before surgical intervention is considered, other causes of midfoot pain should be ruled out by a proper physical examination including posterior tibialis tendoinitis, anterior tibialis tendonitis, Charcot arthropathy, and inflammatory arthritis. The surgeon should determine whether there are any signs of TA tendonitis, such as swelling around the tendon or pain extending around the base of the first metatarsal or with resisted dorsiflexion of the ankle, as tendonitis here would replicate dorsal foot pain and may be confused with midfoot arthritis. However, weakness in dorsiflexion should be appreciated with tenosynovitis or attritional rupture of the anterior tibialis tendon.14
In addition, Charcot arthropathy should also be considered and ruled out before surgical intervention, especially in patients with a history of diabetes mellitus. Significant redness and swelling are not consistent with primary OA, and significant deformity would be more consistent with neuroarthropathy. Imaging should be evaluated for instability and osseous destructive changes commonly seen with neuroarthropathy.
Patients with a history of rheumatoid or other inflammatory arthropathies should be evaluated carefully as these patients commonly demonstrate a different pattern of foot arthritis and only a small majority demonstrate isolated midfoot arthritis.15 Matsumoto and colleagues, using cluster analysis of 542 feet, demonstrated 5 types of foot arthritis in patients with rheumatoid arthritis, which included normal (mild affect to whole foot), forefoot, midfoot, mid-hindfoot, and ankle, with each area demonstrating the predominant area of damage. In this series, only 20% of the patients with RA had midfoot-predominant arthritis. Distinguishing midfoot-predominant RA from primary OA of the midfoot can be challenging. Commonly, patients with RA have a collapsed planovalgus foot, and on imaging, there is a more erosive-type arthritis with minimal bone formation.
Frank contraindications to midfoot fusion include any active infection, poor soft-tissue envelope, or active wound or ulcer-type process on the foot, other than the midfoot, which may be excised at the time of surgery. These issues should be treated and stable for a period of time before elective fusion of the midfoot is further considered.
Midfoot fusion for primary OA is indicated after the failure of conservative management. At the initial evaluation, patients diagnosed with midfoot arthritis are offered nonsteroidal anti-inflammatory medication and rigid or semirigid custom orthotic devices. Ibuki et al16 demonstrated significant improvement in patients with midfoot arthritis using both types of custom insoles. Patients with significant dorsal osteophytes commonly complain of pain in tight-laced shoes. Patients with midfoot arthritis and gastrocnemius tightness are placed on a stretching program.
PHYSICAL EXAMINATION AND IMAGING
Evaluation of patients with midfoot arthritis requires a complete physical examination. Patients are evaluated in shorts to assess the full motor unit and to appropriately assess any gait abnormalities, including toe walking and heel walking. Patients with midfoot OA commonly have pain with any activity that requires toe-off and transmission of forces through the midfoot.1 They are viewed standing toward and away from the examiner to assess ankle, hindfoot, and midfoot alignment. The Silfverskiöld test is performed to evaluate gastrocnemius tightness and differentiate it from isolated Achilles tightness. The soft-tissue envelope is assessed for swelling, ulceration, or infection. Patients with midfoot OA may have obvious dorsal osteophytes. Palpation of each midfoot joint as well as isolated dorsiflexion and plantarflexion of each metatarsal head (the piano key test) may elicit pain and help the examiner identify as to which joints are primarily involved.
Imaging of the foot consists of weight-bearing anteroposterior (AP), lateral, and oblique views of the foot. The AP view is the best for evaluating the second TMT and the medial facet of the NC joint, whereas the oblique view provides the best visualization of the third TMT joint. The lateral view demonstrates any dorsal osteophytes and the arch height.
Preoperative weight-bearing images (AP, lateral, and oblique) are all that are needed before surgical intervention. Advanced imaging is not routinely performed for primary midfoot OA. Examination of patients and a review of radiography should be completed to determine as to which joints are causing pain and should be fused. Patients with known medical comorbidities should have appropriate medical risk stratification.
Isolated Second TMT Fusion
After the induction of anesthesia or a local block, the patient is positioned supine with a bump under the hip. The second TMT is identified with fluoroscopic imaging. A 4 to 5 cm dorsal incision is made directly over the top of the second TMT joint, and the dissection is taken to the periosteum, avoiding injury to the neurovascular bundle and extensor tendons. Imaging is used to confirm that the exposure is appropriate. The capsule and the periosteum are incised longitudinally, and the joint is exposed with subperiosteal elevation. Exposure can be assisted with the use of a small lamina spreader. Dorsal osteophytes are frequently present and can make exposure challenging. Movement of the second metatarsal head and fluoroscopy can commonly assist in the identification of the joint. Once the joint is identified, a small osteotome and curette are used to remove the remaining cartilage. Focus on the plantar aspect of the joint is critical as the cartilage remains here most commonly. Preparation should also include the medial aspect of the second metatarsal, where it articulates with the medial cuneiform in an attempt to improve the fusion mass by including this portion of the joint. However, care should be taken to avoid violating the first TMT joint. After cartilage removal, both sides are drilled with a 2.5 mm drill bit, again focusing plantarly. Any bone from osteophyte removal is added to the joint as a bone graft. A 2-hole plate is selected, and depending on the patient size, it will range from 17 to 20 mm in length. The plate is bent to accommodate the plantar curve of the joint and to prevent plantar gapping during compression. The plate should be used in compression mode to narrow the joint space and promote fusion. Alternatively, a plate with the ability to be compressed after screw placement may be used. A hole in the plate overlying the middle cuneiform is drilled and the screw is placed. The screw is slightly loosened and a screw for the second metatarsal is placed after confirming the plate position with fluoroscopy. During screw placement, it is important for the surgeon or the assistant to position the metatarsal heads in neutral to ensure that there is no dorsiflexion of the second TMT joint. The compression device is used with the surgeon or the assistant holding the metatarsal heads in alignment and preventing dorsiflexion of the second TMT. Forced dorsiflexion of the forefoot with palpation of the metatarsal heads should be performed to ensure that the second TMT is neither plantar nor dorsiflexed. After fluoroscopic confirmation, the wound is closed in layers: 3-0 Vicryl for capsule, 3-0 Vicryl for subcutaneous tissue, and 3-0 nylon for skin. The patient is placed in a plaster splint in a neutral position (Figs. 1, 2).
Second and Third TMT Fusion
Similar to an isolated second TMT fusion, the patient undergoes LMA/general anesthetic with a local block. The patient is again placed supine with a bump under the hip. Using fluoroscopy, the interval between the second and the third TMT joints is identified. A 5 cm skin incision is made in line with this mark, and the dissection is taken down to the periosteum, with this interval protecting the neurovascular bundle, which is medial to this approach. After verifying the location, subperiosteal dissection should be performed with exposure of both joints. Dorsal osteophytes are removed, and cartilage from both sides of the joint is removed with a small osteotome and curette. The use of a small lamina spreader assists in the exposure of both joints. Similar to the second TMT fusion in isolation, it is critical to remove cartilage on the plantar aspect of both the 2/3 TMTs. After cartilage removal, both sides of the joints are drilled with a 2.5 cm drill bit. Any remaining bone is placed into each joint as bone graft. Two 2-hole plates are placed and used in compression mode, ranging in size from 17 to 20 mm, again depending on the patient size. The plates are bent to accommodate the shape of the midfoot and prevent dorsiflexion and plantar gapping of the TMT joints. Fixation of the second TMT is performed first. A screw is first placed in the center of the middle cuneiform and in the center of the plate on the second metatarsal. Again, the surgeon must hold the metatarsal head while placing the screws to prevent dorsiflexion of the TMT joint. These steps are repeated for the third TMT joint. After both plates are placed, they are compressed. It is imperative to note the location of the metatarsal heads during compression to ensure that no sagittal plane malalignment occurs. After compression, the foot is maximally dorsiflexed and the metatarsal heads are palpated to confirm that there is no forefoot imbalance. Wounds are closed in layers with 3-0 Vicryl and 3-0 nylon for the skin. The patient is placed in a plaster splint in the neutral position and is kept non–weight-bearing for a total of 8 weeks (Figs. 3, 4).
The Second and the Third TMT and the Medial Facet of the Naviculocuneiform Joint Fusion
Positioning is the same as described previously. The second and the third TMT joints are identified. The lateral incision is over the third TMT to provide sufficient skin bridge from the incision over the medial facet of the NC joint. Similar to the previously described approach, a 5 cm incision is made over the third TMT and sharply taken down to the level of the periosteum and the capsule. The joints are exposed from lateral to medial. Exposure of the second TMT is beneath the full-thickness skin flap, and the incision is routinely extended if there is a need to reduce the tension on the skin. Once the joints are exposed, preparation and cartilage removal are the same as described previously. After the second and the third TMTs are debrided, the medial facet of the naviculocuneiform is identified using fluoroscopy. A medial incision is centered over the medial third of this medial facet. A sharp dissection is taken down to the level of the capsule. Care is taken to protect both the TA tendon and its tendon sheath. The medial facet is exposed subperiosteally while avoiding the middle facet. Cartilage is removed using curettes and osteotomes. After all 3 joints are prepared, two 2-hole plates are used to fix the second and the third TMT in neutral position, whereas a 3- or 4-hole T-type plate is used for NC joint fixation. Because the middle and the lateral facets of the naviculocuneiform joint are not included in the fusion and can act as a block to the compression of the medial facet and hold the joint distracted, we routinely add bone graft to the medial facet, either calcaneal autograft or demineralized bone matrix. Second and third TMTs are fixed as described previously. Fixation of the NC joint first requires plate preparation. A T-type compression or a 4-hole compression plate is frequently selected for this joint. First, 2 screws are placed in the navicular starting in the medial aspect of the second screw more dorsally. Depending on the plate selection, 1 or 2 screws are then placed in the cuneiform. The foot is held in neutral position, avoiding dorsiflexion of the joint to prevent lateral overload while the medial cuneiform screws are placed. Each of the 3 plates is compressed while maintaining the appropriate position of the foot. The foot is checked in forced dorsiflexion to determine whether there is any sagittal imbalance at the metatarsal heads. Wounds are closed with 3-0 Vicryl and 3-0 nylon, and the patient is placed in a short leg plaster splint in the neutral position.
Studies examining the outcomes of midfoot fusion have demonstrated good results and significant functional improvement for both posttraumatic and degenerative OA.5,7,13,17 Mann et al5 reported a satisfaction rate of 93% (37 of 41) among patients treated with midfoot fusion for either degenerative or posttraumatic arthritis. Nemec et al13 reported a 92% fusion rate among 104 feet and improvement in the preoperative AOFAS score from 32 to a postoperative mean of 79 among the 74 patients with available outcome data. Jung and colleagues studied patients with atraumatic OA. Sixty-seven feet underwent midfoot fusion with an improvement in the AOFAS score from 34.1 to 83.9 postoperatively, with the most notable improvements occurring in pain, gait abnormality, and alignment components.7 In this study, however, patients were stratified on the basis of the foot deformity and also underwent correction for rockerbottom deformity, pes planovalgus, and hallux valgus at the time of surgery.
Complications after midfoot arthrodesis include nonunion, sesamoid pain, hardware irritation, prominent metatarsal heads, stress fracture of the metatarsals, wound complications, complex regional pain syndrome, development of adjacent joint OA, and neuroma.4,5,7,13
The incidence of nonunion after midfoot arthrodesis varies from 2% to 10% for both traumatic and atraumatic OA.5,7 Consistent with previous literature, Mann et al5 reported a union rate of 98%, Jung et al7 93%, and Nemec et al13 92% in their studies. Nonunion typically presents with persistent pain at the fusion site and frank instability of an attempted fused joint may be appreciated on examination. Radiographs may reveal hardware failure or broken hardware. If both physical examination and/or plain films are concerning for nonunion, a CT may be obtained to evaluate bony union. When nonunion is diagnosed, it is important to evaluate for other underlying etiologies such as infection or neuropathy before returning to the operating room for revision.
Jung et al7 reported a 39% complication rate (26 of 67 feet) with sesamoid pain being the most common complication, with the majority reporting tibial seasamoiditis from excessive plantarflexion of the first metatarsal. Authors also reported 4 patients with neuralgia involving the superficial peroneal or the sural nerve and 1 case of pulmonary embolism.
Hardware-related complications are frequently reported, with 6 of 67 (9%) feet in Jung et al7 study undergoing hardware removal and Nemec et al13 reporting a 25% hardware removal rate. Before hardware removal, symptom management can be attempted with foot-wear modification and NSAID usage. Hardware removal is not considered at our institution until at least 1 year after surgery.
As discussed in the surgical techniques section, metatarsal head alignment is critical. Elevation of the second metatarsal head intraoperatively leads to an overload of the third metatarsal, and metatarsalgia could result in third metatarsal stress fracture. Mann et al5 reported a stress fracture of the second metatarsal in 3 patients, all of which healed with nonoperative treatment. In addition, as mentioned above, plantar flexion of the first metatarsal may result in tibial sesamoiditis from first metatarsal overload.7
Lastly, soft tissues should be carefully assessed preoperatively and skin flaps meticulously created intraoperatively in an effort to avoid wound complications. Older patients with less robust soft-tissue envelops can frequently present with some mild erythema. This may be related to the underlying hematoma or superficial infection. At the first postoperative visit, sutures should not be removed if there is any concern of wound issues. Occasionally, a large hematoma is noted and requires return to the operating room for debridement and identification and cauterization of the small vessel. Superficial infections can often be treated with a short course of oral antibiotics and elevation. If frank pus or refractory wound issues occur, one should strongly consider operative debridement with intraoperative cultures. To maintain alignment, the hardware is frequently left in place and may require suppressive antibiotics until fusion is noted.
Postoperatively, all patients with midfoot fusion are treated with the same protocol irrespective of which joints are fused. After being placed in a splint in the operating room, patients are evaluated in the office approximately 2 to 3 weeks later. Sutures are removed if the skin appears to be well healed. The threshold is low to keep sutures an additional 10 to 14 days if any signs of wound-healing issues are present. Patients are placed into a short leg cast in neutral ankle position and remain non–weight-bearing until 8 weeks postoperatively. At the 8-week postoperative visit, if patients have no pain, minimal swelling, and satisfactory imaging (maintained alignment, no hardware complications, and early signs of healing), patients are transitioned to progressive weight-bearing in a boot. Progressive weight-bearing begins with 25 to 50 pounds of weight-bearing with crutches and advances 25 pounds per week until full weight-bearing. At 12 weeks, if patients demonstrate both clinical and radiographic signs of successful fusion, they are progressed to full weight-bearing and instructed to transition out of the boot and resume normal activities as swelling and pain of the foot allow. Frequently, patients benefit from physical therapy to focus on gait training and balance, but it is not required for all patients. Patients who participate in high-impact exercise are advised to wait a minimum of 4 to 6 months before resumption of these activities.
POSSIBLE CONCERNS AND THE FUTURE OF THE TECHNIQUES
This is just 1 technique for addressing primary OA of the midfoot. All the hardware is placed dorsally, and there is some concern regarding dorsal irritation requiring future removal of the hardware. It is not clear at this point whether the use of these types of plates provides any superior clinical outcomes to warrant their increased cost as compared with techniques such as screw fixation alone or traditional nonlocked plate and screw constructs. Future studies will hopefully identify as to which patients benefit from bone grafting, advanced plating technologies, and/or the addition of biologics to reduce the number of complications and failures.
1. Patel A, Rao S, Nawoczenski D, et al.. Midfoot arthritis. J Am Acad Orthop Surg. 2010;18:417–425.
2. Rao S, Nawoczenski DA, Baumhauer JF. Midfoot arthritis: nonoperative options and decision making for fusion. Tech Foot Ankle Surg. 2008;7:188–195.
3. Zonno AJ, Myerson MS. Surgical correction of midfoot arthritis with and without deformity. Foot Ankle Clin. 2011;16:35–47.
4. Verhoeven N, Vandeputte G. Midfoot arthritis: diagnosis and treatment. Foot Ankle Surg. 2012;18:255–262.
5. Mann RA, Prieskorn D, Sobel M. Mid-tarsal and tarsometatarsal arthrodesis for primary degenerative osteoarthrosis or osteoarthrosis after trauma. J Bone Joint Surg Am. 1996;78:1376–1385.
6. Cowie S, Parsons S, Scammell B, et al.. Hypermobility of the first ray in patients with planovalgus feet and tarsometatarsal osteoarthritis. Foot Ankle Surg. 2012;18:237–240.
7. Jung HG, Myerson MS, Schon LC. Spectrum of operative treatments and clinical outcomes for atraumatic osteoarthritis of the tarsometatarsal joints. Foot Ankle Int. 2007;28:482–489.
8. Rao S, Baumhauer JF, Tome J, et al.. Orthoses alter in vivo segmental foot kinematics during walking in patients with midfoot arthritis. Arch Phys Med Rehabil. 2010;91:608–614.
9. Rao S, Baumhauer JF, Nawoczenski DA. Is barefoot regional plantar loading related to self-reported foot pain in patients with midfoot osteoarthritis. Osteoarthritis Cartilage. 2011;19:1019–1025.
10. Beek CV, Greisberg J. Mobility of the first ray: review article. Foot Ankle Int. 2011;32:917–922.
11. Bolgla LA, Malone TR. Plantar fasciitis and the windlass mechanism: a biomechanical link to clinical practice. J Athl Train. 2004;39:77.
12. Menz HB, Munteanu SE, Zammit GV, et al.. Foot structure and function in older people with radiographic osteoarthritis of the medial midfoot. Osteoarthritis Cartilage. 2010;18:317–322.
13. Nemec SA, Habbu RA, Anderson JG, et al.. Outcomes following midfoot arthrodesis for primary arthritis. Foot Ankle Int. 2011;32:355–361.
14. Coughlin MJ, Saltzman CL, Anderson RB. Disorders of Tendons. In: Mann's Surgery of the Foot and Ankle. 9th ed. Philadelphia, PA: Elsevier Saunders; pp 1188–1291.
15. Matsumoto T, Nakamura I, Miura A, et al.. Radiologic patterning of joint damage to the foot in rheumatoid arthritis. Arthritis Care Res. 2014;66:499–507.
16. Ibuki A, Cornoiu A, Clarke A, et al.. The effect of orthotic treatment on midfoot osteoarthritis assessed using specifically designed patient evaluation questionnaires. Prosthet Orthot Int. 2010;34:461–471.
17. Johnson JE, Johnson KA. Dowel arthrodesis for degenerative arthritis of the tarsometatarsal (Lisfranc) joints. Foot Ankle. 1986;6:243–253.
Keywords:Copyright © 2016 Wolters Kluwer Health, Inc. All rights reserved
midfoot; tarsometatarsal; arthrodesis; fusion; osteoarthritis