Hallux Rigidus

The pathophysiology of hallux rigidus is similar to that of degenerative arthritis in any joint. Overuse, injury, or abnormal joint mechanics lead to abnormal stresses on the articular cartilage.

In an in-vitro study, Ahn et al used a magnetic tracking system to monitor the three-dimensional movement of the proximal phalanx while the toe position was changed from a neutral position to full extension.[6] The contact distribution shifted dorsally with increasing degrees of extension. These data are consistent with the observation that chondral erosions associated with hallux rigidus and degenerative arthritis initially affect the dorsal articular surface of the MT.[7]

Articular degenerative changes are associated with dehydration of the cartilage, which, in turn, is more susceptible to injury resulting from shear and compressive forces. The subchondral bone shares these stresses, which subsequently lead to increased subchondral bone density, formation of periarticular osteophytes, and, in severe cases, cystic changes. The osteophytes limit first MTP joint motion and further compromise the normal mechanics of this joint. This effect can accelerate the degenerative process. In severe cases, the articular cartilage is completely denuded.

The true etiology of hallux rigidus is not known. Most commonly, hallux rigidus is thought to be caused by chronic overuse or unfavorable use of the first MTP joint. Multiple theories have been proposed for the underlying etiology.

Some authors have associated hallux rigidus with athletic activities involving running[8] ; in this case, the disorder possibly results from repetitive hyperextension of the first MTP joint with chronic gradual attenuation of the plantar plate and subsequent instability. Hallux rigidus has also been seen as a long-term sequela of acute injuries to the great toe MTP joint (eg, turf toe). Several authors have suggested traumatic injury to the articular cartilage—either acute trauma (as in turf toe) or chronic, repetitive, minor injury—as the underlying mechanism.

Clanton et al found hallux valgus and early hallux rigidus to be long-term sequelae. After more than 5 years of follow-up, Clanton and Seifert found that among 20 athletes with previous turf toe injury, half suffered from persistent symptoms. The long-term effects of turf toe require further study.

In 1933, Kingreen reported that osteochondritis dissecans led to development of hallux rigidus. Goodfellow proposed that the development of an osteochondrosis in childhood creates a defect and secondary slow-remodeling collapse, leading to abnormal motion in the forefoot.[9] McMaster reported on seven adolescent patients who had an articular defect of approximately 5 mm located directly beneath the dorsal lip of the proximal phalanx; this defect was associated with symptoms of hallux rigidus.[10]

In 1938, Lambrinudi proposed the so-called metatarsus primus elevatus.[11] Theoretically, an abnormally elevated first MT causes excessive flexion of the great toe during gait and subsequent development of flexion contracture at the first MTP joint. These abnormal mechanics cause hallux rigidus.

Others, such as Jack, in 1940, postulated that with the elevated first MT, increased overload of the second MT occurs, with compensatory contracture of the flexor hallucis brevis (FHB).[12] This contracture pulls the proximal phalanx inferiorly, driving its dorsal rim into the MT head and leading to localized degenerative changes in the articular cartilage. Hypermobility of the first ray leading to flexor spasm and impingement of the proximal phalanx on the MT head is another proposed theory.[13]

Yet other researchers, such as Jansed, in 1920, implicated flatfoot. All of these theories are without true scientific data.

In 1986, Mann first theorized that a flat first MT head restricts, to a relative extent, the medial and lateral motion of the first MTP joint, creating increased stress in the sagittal plane, and that this restriction of motion accelerates the degenerative process. Others have proposed that flattening of the head is a secondary result.[14]

Some authors have proposed that the disease may develop somewhat differently in adolescents than in adults. The observation that the first MTP joint returns to normal under anesthesia in adolescents suggests that anatomic anomalies and spasm may be contributing factors.

Bingold et al suggested that the disease proceeds in stages from adolescence through adulthood.[13] Vilaseca et al found that a distal physis of the first MT head is present in 75% of children's feet and is visible in children aged 2-11 years.[15] They also found that the first MT is longer than the second in children who have had a longer persistence of this distal physis. Therefore, individual anatomic variations may play a role in causing functional changes in the MTP motion and position during gait.[16]

An abnormally long first MT (index-plus foot) increases the first MTP joint stress during toe-off, as proposed by Nilsonne in 1930.[17] This predisposes an individual to hallux rigidus. Nilsonne et al[18] suggested that the excessively long toe requires a longer shoe, which in turn requires constant contraction of the great toe flexors to grip the shoe while the person is walking. This gripping can lead to inflammation and secondary spasm, therefore limiting motion at the MTP joint at the great toe.

In a study involving 110 patients with hallux rigidus, Coughlin et al examined possible associations between the disorder and various physical, health, and lifestyle factors.[19] The authors saw no association between hallux rigidus and pes planus, first MT length, metatarsus primus elevatus, first-ray hypermobility, hallux valgus, footwear, occupation, obesity, or metatarsus adductus. However, they did see an association between hallux rigidus and hallux valgus interphalangeus (mean, 18°), family history (in bilateral cases of hallux rigidus), and trauma (in unilateral cases of the condition). No specific distinction was made between adolescent and adult patients.

Bejarano-Pineda et al, in a study of 809 patients (870 feet) with end-stage ankle arthritis, found the prevalence of radiographic hallux rigidus in this population to be significantly higher than that seen in patients who had no documented foot or ankle comorbidities; the prevalence also rose with increasing age.[20]  

Hallux rigidus is the second most common disorder of the first MTP joint, after hallux valgus.

Coughlin and Shurnas reported findings in 110 patients who had undergone surgery for hallux rigidus.[19] The authors noted that on final evaluation, about 80% of the patients showed bilateral involvement. In the bilateral cases, 98% of the patients had a positive family history. Although 62% of the patients in Coughlin and Shurnas's report were women, other investigators have reported a slight male predominance.

Nonsurgical measures can often be used to successfully treat patients with varying degrees of severity of hallux rigidus. For patients in whom the condition is refractory to nonoperative treatment methods, the operative options depend on the severity of the DJD.

Waizy et al analyzed long-term clinical outcome and patient satisfaction in 60 patients (20 grade 1, 35 grade 2, 5 grade 3) with symptomatic hallux rigidus who received joint-preserving operative care.[21]  At follow-up, mean dorsiflexion increased to about 21.7º in grade 1 patients, 23.7º in grade 2, and 26.3º in grade 3. At first follow-up, 100% of grade 1 patients had only occasional pain or no pain at all, compared with 63.3% of grade 2 and 75% of grade 3. At second follow-up, 77.8% of grade 1 and 73.9% of grade 2 patients had no pain. Four patients had persistent hypoesthesia of the medial side of the great toe, and three had delayed wound healing. No patients required revisions or further surgical procedures.

Tagoe et al followed patients who underwent total sesamoidectomy for hallux rigidus/limitus (N = 33; 36 procedures) for 2-4 years.[22] They reported high levels of clinical improvement and patient satisfaction after the procedure, with no significant functional impairment or malalignment and no instances of pain on metatarsal compression or of transfer metatarsalgia. The authors concluded that for symptomatic patients in whom joint replacement/fusion is not indicated, total sesamoidectomy may be beneficial as an interim procedure for joints with a moderate degree of arthrosis (grade 2-3).

Cöster et al used data from 296 patients registered in Swefoot (the Swedish national registry of foot and ankle surgery) to assess patient-reported outcomes of Youngswick osteotomy (n = 115) and cheilectomy (n = 181) for moderate hallux rigidus at 1 year.[23]  Improvements in the SEFAS (SElf-reported Foot and ANKLE Score) were noted with both procedures: 12 points with Youngswick osteotomy and 10 points with cheilectomy. However, patients who underwent osteotomy reported a higher level of satisfaction than those who underwent cheilectomy (84% vs 70%).

Cheilectomy

Mann et al reported on 20 patients who were treated with cheilectomy and monitored for an average of 67.6 months.[24]  They stated that patient satisfaction with the procedure was uniform and that they had an 85% success rate, with an average DF of 30° after surgery.

Mann and Clanton reported on 34 cheilectomies, with 74% improvement in motion, an average of 20° improvement in joint motion, and 90% pain relief, with an average resultant DF of 48°.[14]  Over an average follow-up period of 56 months, they reported no complications.

Easley et al reported on 75 feet treated with cheilectomy, with an average follow-up of 63 months.[25]  The average DF improved by 20°, but nine of 21 patients with recurrent dorsal formation were symptomatic.

Mulier et al followed up 22 athletes who had undergone cheilectomy for grade 1 or 2 hallux rigidus and noted that functional results were excellent in 14 of the athletes, good in seven, and fair in one at 5-year follow-up.[26]  Seven patients had radiographic signs of progression at follow-up.

Mackay et al evaluated 39 patients at a mean follow-up time of 3.8 years after cheilectomy.[27]  The patients had grade 1, 2, or 3 hallux rigidus as defined by the Regnauld classification. The investigators noted significant improvements in pain, activity level, tiptoe walking ability, and ROM among patients no matter which grade of hallux rigidus had been treated. Footwear selection improved significantly in patients with grade 1 or 2 hallux rigidus but not in patients with grade 3.

Nicolosi et al performed a retrospective study of 58 patients (mean age, 55.71 ± 9.51 years) to evaluate the long-term efficacy (mean follow-up, 7.14 years; range, 39 weeks to 14.87 years) of aggressive cheilectomy for addressing degenerative joint disease (DJD) of the first MTP joint.[28]  In all, 51 of the 58 patients had no limitations in their daily activities; only two subsequently required arthrodesis.

Proximal phalanx osteotomy

Thomas and Smith reviewed 24 great toes after proximal phalanx osteotomy and reported improvement in all patients.[29]  Recovery time was 2-12 months. The average increase in DF was 7°, and PF increased 3°. The increase in PF was attributed to aggressive physical therapy. Radiographically, the dorsal MTP joint space increased, and the length of the proximal phalanx decreased by an average of 4 mm. The resting position of the toe showed an average increase in elevation of 5.4 mm.

Citron and Neil reported on 10 toes at a 22-year follow-up.[30]  All patients experienced complete pain relief shortly after the osteotomy, and relief was permanent in five of the 10 toes. One patient required an MTP fusion. The researchers found that after the osteotomy, PF of the MTP joint was lost, but the arc of movement of the interphalangeal (IP) joint was shifted toward PF.

Arthroscopy

van Dijk et al performed a prospective study in 24 consecutive patients, 17 of whom were high-level athletes. In the dorsal impingement group, eight of 12 patients had a good or excellent result after a minimum follow-up of 2 years.[31]  One patient in this group had a persistent loss of sensitivity of the hallux.

Iqbal et al evaluated 15 patients who underwent arthroscopic cheilectomy for hallux rigidus and reported encouraging early results without the need for revision surgery.[32]

Metatarsal osteotomy

Despite the number of first-MT osteotomies for hallux rigidus that have been reported in the literature, long-term outcome data on these procedures have been sparse. In addition, the high rates of complications (eg, nonunion, malunion, intractable plantar ketosis under the first MT head, sesamoiditis, transfer lesions, stress fractures, dorsal contractures, arthrofibrosis, and progressive DJD), have made the results unpredictable. The associated high risks and the minimal potential benefits in most patients with hallux rigidus must be considered.[33, 34]

In a retrospective study that assessed the long-term outcomes of Youngswick osteotomy in 61 patients treated for grade II and grade III hallux rigidus, using the need for first MTP joint arthrodesis as an end point, Slullitel et al found the outcomes to be satisfactory in terms of function, pain, relief, and patient satisfaction, even in patients followed for longer than 13 years.[35]

Arthrodesis

Coughlin et al reported a 92% fusion rate with the use of congruent cup-shaped reamers and a dorsal plate (as performed on 58 feet).[36]  They described a 98% fusion rate, and 93% of their cases had good or excellent results. Plate removal was necessary in four cases, and they reported delayed union in one case and plate breakage in one case.

Chana et al reported the use of suture stabilization using size 00 chromic catgut postoperative casting to achieve stable arthrodesis in 87 of 87 feet. They reported a 10% incidence of pseudoarthrosis and four malunions.

In 1993, Curtis et al compared the biomechanical results of four methods of internal fixation and noted that bony preparation with power conical reamers and supplementary interfragmentary screw fixation had the most stable results.[37]

Coughlin et al reported on 16 feet that underwent arthrodesis after the failure of a Keller procedure.[38]  Using multiple intramedullary threaded Steinmann pins to fix the bone, the authors noted a 100% arthrodesis rate. They used interposition of an iliac crest bone graft in four feet with an excessively short hallux. In 92% of cases, intractable keratosis was relieved; the characteristic deformity of the hallux was improved.

Smith et al, using five threaded 0.062 Kirschner wires (K-wires) for fixation in 34 feet after the joint surfaces had been prepared with conical reamers, reported an arthrodesis rate of 97%.[39]

Hamilton et al reported on 37 feet after resection arthroplasty with the extensor hood and extensor brevis reattached to the flexor hallucis brevis (FHB) as a capsular interposition arthroplasty (with minimal bone resection).[40]  They found maintenance of PF strength in four of five cases, with an average of 50° DF.

Kennedy et al reported on a peg-in-socket method for the salvage of failed Keller excisional arthroplasty in the presence of poor bone stock.[41]

DeFrino et al evaluated plantar pressure distribution and gait patterns after first MTP arthrodesis.[42]  They reported the restoration of the weightbearing function of the first ray, with greater maximum force carried by the distal hallux at toe-off. The authors noted a significantly shorter step length, with some loss in ankle PF at toe-off on the fused side. They suggested a reduction in ankle torque and ankle power at pushoff. The authors also reported a high level of patient satisfaction with this procedure.

Arthroplasty

Excisional/interposition arthroplasty

Lau et al retrospectively reviewed 11 feet with grade 3 osteoarthritis that were treated with interposition arthroplasty and 24 feet with grade 2 osteoarthritis that were treated with cheilectomy.[43]  Approximately 2 years after treatment, results were compared with respect to postoperative motion, visual analogue pain scale scores, and Medical Outcomes Study 36-item short-form (SF-36) results. About 73% of patients who underwent interposition arthroplasty reported weakness, compared with 17% of the group that underwent cheilectomy. About 73% of patients in the former group reported satisfaction with interposition arthroplasty, and about 88% in the latter group were satisfied with cheilectomy.

Significant improvement of pain and function and an average DF of 50° can be achieved, as reported by Hamilton et al.[40]  Relative contraindications for interposition arthroplasty for the treatment of hallux rigidus include relatively short first MTs (Morton or Grecian foot) owing to the development of transfer lesions and, for high-level athletes and ballet dancers, because of the loss of the windlass mechanism.

Silicone arthroplasty

Rahman et al reviewed 78 feet after silicone hemiarthroplasty of the first MTP joint.[44]  At a mean follow-up of 4.5 years, 56 feet showed radiologic evidence that was suggestive of silicone granulomatous disease, with three cases of the disease being confirmed at histologic study. The investigators suggested that this operation be abandoned.

Shankar reported on 40 cases of hallux rigidus that were treated with silicone hemiarthroplasty and then followed up for an average of 110 months.[45]  The author stated that 36% of the study's patients were unhappy with the results of the procedure and that six implants had to be removed because of pain and fragmentation.

Metallic arthroplasty

Townley et al conducted a long-term retrospective study of 279 nonconstrained metallic resurfacing implants used for hemiarthroplasty.[46]  Among the cases, 95.3% had good or excellent results with a follow-up time of 10 months to 33 years.

Johnson et al reported results achieved with a stainless steel and polyethylene surface replacement prosthesis for the first MTP joint, fixed with methylmethacrylate.[47]  In a series of 21 joints, the satisfaction rate was 81% at an average follow-up of 43 months.

Lu et al reported findings from an 8-year follow-up study of 14 cases with a titanium total joint prosthesis.[48]  They reported a 31.8% complication rate and a 27.3% rate of revision.

Gheorgiu et al retrospectively reviewed data from 11 patients (12 feet) who underwent first MTP joint hemiarthroplasty with the HemiCAP (Arthrosurface, Franklin, MA) prosthesis (mean follow-up, 47 months; range, 36-48 months).[5]  Of the 12 feet, 41.7% were pain-free at follow-up, 25% had mild pain, 16.7% had moderate pain, and 16.7% had severe pain. In addition, 42% showed no evidence of radiologic subsidence, and 58% showed a mean subsidence of 2.71 mm (range, 1-6 mm). Most of the patients continued to have a limited ROM, reporting only reasonable levels of satisfaction.

The most typical presentation of hallux rigidus is pain on the top of the great toe in an active, middle-aged individual. In addition, patients may also complain of diffuse, lateral forefoot pain resulting from increased weightbearing on the lateral foot to offload the hallux. The pain is worse with certain activities and with certain shoes. Hallux rigidus has been associated with a history of athletic activities involving running or kicking. Whether running is a true cause of the disorder or a factor that aggravates the symptoms is not known.

Dysesthesia along the dorsomedial hallux occasionally occurs as a result of compression by shoes or stretching of the dorsomedial cutaneous nerve. Complaints of stiffness and motion loss are not common, except in adolescents, who often present with complaints of a rigid first metatarsophalangeal (MTP) joint. Dorsal pain caused by external pressure over a prominent osteophyte frequently accompanies other presentations. Diffuse arthritic pain usually occurs late and is associated with more severe degenerative changes.

At physical examination, the presence of a tender dorsal osteophyte at the first MTP joint (see the images below) usually confirms the diagnosis of hallux rigidus. First MTP dorsiflexion (DF) is limited by periarticular osteophytes, and pain often occurs at maximum DF. Dorsal pain with maximum plantarflexion (PF) is common and likely represents irritation of the extensor hallucis longus (EHL) as it passes over the dorsal osteophyte. Pain and crepitus that occur throughout the entire range of motion (ROM) indicate late-stage degenerative arthritis of the first MTP joint. The patient's gait may be slightly antalgic, and limitation of MTP joint DF at toe-off may be noted.

In 1988, Hattrup and Johnson described the following radiographic classification system for hallux rigidus:

• Grade 1 - Mild changes with a maintained joint space and minimal spurring
• Grade 2 - Moderate changes, joint-space narrowing, bony proliferation on the metatarsal (MT) head, and phalanx and subchondral sclerosis or cysts
• Grade 3 - Severe changes with significant joint-space narrowing, extensive bony proliferation, and loose bodies or a dorsal ossicle

In 1999, Coughlin and Shurnas proposed a classification system based on ROM, as well as on radiographic and examination findings, as follows[19] :

• Grade 0 - DF of 40-60° (20% loss of normal motion), normal radiographic results, and no pain
• Grade 1 - DF of 30-40°, dorsal osteophytes, and minimal to no other joint changes
• Grade 2 - DF of 10-30°, mild flattening of the MTP joint, mild-to-moderate joint narrowing or sclerosis, and dorsal, lateral, or medial osteophytes
• Grade 3 - DF of less than 10°, often less than 10° PF, severe radiographic changes with hypertrophied cysts or erosions or with irregular sesamoids, constant moderate to severe pain, and pain at the extremes of the ROM
• Grade 4 - Stiff joint, radiographs showing loose bodies or osteochondritis dissecans (OCD), and pain throughout the entire ROM

In patients with hallux rigidus, radiographs show a variable degree of degenerative changes. Early changes typically include dorsal and marginal osteophytes with well-maintained joint space. More severe degenerative changes are seen in more advanced cases. These changes include narrowing of the joint space, marginal osteophytes, sclerosis, joint irregularities, cystic degeneration, degenerative involvement of the sesamoid and/or the first metatarsal (MT) head joint, and sesamoid enlargement. (See the images below.)

Hallux valgus deformity can accompany hallux rigidus, though it is not common. Hallux valgus interphalangeus is more commonly seen.

Magnetic resonance imaging (MRI) may be useful in assessing for early-stage disease.

Nonsurgical techniques can often be used to successfully treat patients with varying degrees of severity of hallux rigidus. However, when the condition is refractory to nonoperative treatment methods, there are a number of procedures that can be employed as treatment. The choice of operation depends on the degree of involvement, the range-of-motion (ROM) limitations, the individual's activity level, and the surgeon's and patient's preference. Options include the following:

• Joint-sparing procedures, such as cheilectomy, with or without proximal phalanx osteotomy (Moberg procedure)
• Metatarsal (MT) osteotomy
• Joint arthroplasty
• Arthrodesis

Dorsal cheilectomy is indicated in patients with mild-to-moderate arthritic changes with less than 50% involvement of the joint surface.[49] A proximal phalangeal osteotomy can be added in patients in whom sufficient dorsiflexion (DF) is not obtained (if plantarflexion [PF] is sufficient).

Excisional arthroplasty, or the Keller procedure, is associated with a number of potential complications and is not generally recommended.[50] Capsular interposition arthroplasty, however, can provide good pain relief in select individuals with advanced degenerative disease.

Metatarsophalangeal (MTP) arthrodesis is an excellent procedure that is indicated in most cases of advanced hallux rigidus. The benefits of MT osteotomies are theoretical, and these procedures are not recommended for the treatment of uncomplicated hallux rigidus.

Silicone-implant arthroplasty probably has no place in the treatment of hallux rigidus. It is not clear that current metallic hemiarthroplasty and total arthroplasty have significant advantages over capsular interposition arthroplasty, and the techniques have many reported complications.[51]

The use of a polyvinyl alcohol (PVA) hydrogel implant to treat hallux rigidus has received substantial attention. Early studies suggested results similar to those of arthrodesis.[52]  A limited number of patients who underwent this joint-sparing procedure (< 10%) required conversion to arthrodesis within 5 years.[53]  Additionally, successful later conversion to arthrodesis is not prohibitively difficult.[54]  High levels of patient satisfaction can be achieved.[55]

Chrea et al found that whereas the addition of a PVA implant to cheilectomy with Moberg osteotomy was safe and useful, the same procedure without the implant might have equivalent or better results.[56]  Shimozono et al reported a high failure rate with the PVA hydrogel implant in patients with hallux rigidus.[57]  A number of adverse events have been associated with synthetic cartilage implants, including implant subsidence, fragmentation, infection, bone erosion, and foreign body reaction; because of the voluntary nature of the reporting, such events may be underreported.[58]

An absolute contraindication for operative treatment is poor peripheral circulation. Active infection should be considered a relative contraindication.

Pharmacologic options include analgesics and nonsteroidal anti-inflammatory drugs (NSAIDs).

Mechanical methods to limit the motion of the first MTP joint are helpful. The use of in-shoe orthotics with medial stiffness, stiff-soled shoes with a rocker bottom, shoes with a wide toe box, low-heeled shoes, and shoe modifications (eg, a steel shank placed along the entire medial side) may be beneficial. Activity modifications include avoiding extremes of DF of the great toe, such as those caused by kneeling or squatting with the toes in an extended position.

Solan et al reported on the use of manipulation and injection of the joint, but this was noted to be helpful only in grade 1 or 2 hallux rigidus.[59] Patients with grade 1 changes obtained symptomatic relief for a median of 6 months. Little symptomatic relief was seen in patients with grade 3 arthritis.

Surgical options include joint-sparing procedures, such as cheilectomy, with or without proximal phalanx osteotomy (the Moberg procedure), as well as MT osteotomy, joint arthroplasty, and arthrodesis.[60, 61]

Cheilectomy

Cheilectomy involves the excision of all irregular bony spurs that limit motion.[49] These spurs are primarily dorsal, but excision of impinging medial and lateral spurs, if present, is usually part of the procedure as well. Most authors recommend dorsal cheilectomy for patients with mild to moderate radiographic changes, particularly young or active patients who have less than 50% articular cartilage loss, as seen at surgery.

The procedure can be performed through a  5- to 6-cm dorsal midline or medial incision centered over the first MTP joint. The dorsomedial cutaneous nerve is protected. A longitudinal capsulotomy exposes the joint, and all excessive synovial tissue is debrided.

The joint is explored for any loose bodies and joint abnormalities. Often, the articular surface of the MT head shows erosions down to bone in the dorsal half. Occasionally, erosive changes of the proximal phalanx surface also are present. An osteotome or saw is used to resect the large dorsal osteophyte on the MT head and, sometimes, the dorsal base of the proximal phalanx. As much as one third of the dorsal joint surface of the MT head is excised. Usually, most of the eroded surface is excised. (See the image below.)

Medial and lateral bone spurs are excised flush with the joint, and the ROM is checked. The goal is to achieve at least 60º of MTP DF intraoperatively. A Freer elevator is used to free any plantar adhesions between the MT head and the sesamoids. In addition, the plantar capsule is released, if necessary, to achieve improved ROM.

Postoperatively, the patient may bear weight as tolerated in a postoperative shoe. Passive motion exercises are started in 1 week if the wound is dry.

Proximal phalanx osteotomy

Bonney et al first described proximal phalanx osteotomy for the treatment of hallux rigidus in 1952.[18] In 1958, Kessel et al reported on the use of the procedure in 10 adolescents, and in 1979, Moberg discussed its use in eight patients.[62, 63] Moberg popularized the proximal phalanx osteotomy, and his name is commonly associated with this operation.

The Moberg procedure involves increasing the DF by translating the arc of motion from plantar to dorsal, thereby increasing the functional ROM of the MTP joint to a more dorsal position because of a more dorsiflexed position of the toe. This position decreases pushoff stress on the hallux and provides decompression of the dorsal joint space.

A prerequisite to the procedure is the presence of adequate PF. Proximal phalanx osteotomy is always accompanied by dorsal joint cheilectomy/decompression. Reported complications are relatively few but have included nonunion or malunion, tendon injury, neuritis or neuroma, progressive DJD, and decreased pushoff power. Most reports of this procedure show functional improvement with low complication rates. They also show that this procedure can be added to dorsal cheilectomy when sufficient DF is not attained with cheilectomy alone in cases of mild-to-moderate hallux rigidus (particularly in running athletes), provided that preoperative PF is adequate.

The technique involves a medial longitudinal or a dorsal incision over the base of the proximal phalanx, with protection of the dorsal and plantar medial cutaneous nerves and the flexor and extensor tendons. A dorsal closing-wedge greenstick osteotomy is performed with a saw or osteotome just distal to the MTP joint. The plantar cortex is maintained, and 2-6 mm of dorsal cortex is cut, depending on the degree of hallux PF and joint stiffness. The goal is to attain 20-30° of DF relative to the first MT.

The osteotomy can be stabilized with a Kirschner wire (K-wire), a small screw, or a staple to allow early DF at 1-2 weeks and PF at 3-4 weeks. Full weightbearing is usually allowed.

Arthroscopy

Although not commonly performed, arthroscopic removal of the dorsal osteophytes in hallux rigidus has been reported. Arthroscopy of the great-toe MTP joint requires the availability of small arthroscopic equipment (1.9 mm) and the use of digital equipment. The role of arthroscopy in the treatment of various grades of hallux rigidus and its benefits compared with those of open cheilectomy have not been defined. Fast rehabilitation and prompt return to work are the advantages of this arthroscopic procedure.

Metatarsal osteotomy

MT osteotomies are based on the premise that metatarsus elevatus or a long first MT are the underlying causes of hallux rigidus. Therefore, a number of different osteotomies to cause PF and/or to shorten the first MT have been described.

Most authors define indications for osteotomy in lower grades of hallux rigidus. Proximal and distal osteotomies have been described. Various authors, including Suppan in 1921, Hohman in 1924, Waterman in 1927, Lambrinudi in 1938, Youngswick in 1982, Pittman and Burns in 1984, and Davies in 1989, have described different methods of MT osteotomy.[11] These include wedge-, oblique-, or crescentic-type osteotomies, performed at the base or neck to result in PF and/or shortening of the first MT. Waterman's dorsal closing-wedge osteotomy with removal of the dorsal osteophytes in effect rotates the articular surface dorsally for a more functional arc of motion.

Arthrodesis

Arthrodesis eliminates all motion across the first MTP joint; it is a reliable and reproducible procedure with a high percentage of good results.[64] This procedure is indicated in severe cases of hallux rigidus and as a salvage procedure when other surgical methods fail. Preoperative considerations are whether enough bone stock is available for an in situ fusion and whether an interposition bone block is necessary for a shortened first ray.[16]

The procedure involves a dorsomedial or medial exposure of the first MTP joint, as well as the removal of the articular cartilage to create opposable, cancellous surfaces and stabilization using one of a number of different devices, including pins, screws, or plates. The medial exposure has the benefit of avoiding a scar contracture involving the dorsomedial cutaneous nerve.

A number of different techniques have been described for preparation of bone ends and for fixation of the toe. Conical reamers can be used to allow continued good apposition throughout the range of positions of the toe. The use of these reamers requires good circumferential exposure of the joint. The position of the toe is the most important aspect of this operation. The most commonly recommended position is that in which the toe is placed in approximately 15° of DF. However, the optimal position depends on the hallux declination angle, which has been described as 15-30°.

To check the position intraoperatively, a simulated weightbearing position is created by using a hard, flat surface on the plantar aspect of the foot, and the ankle is brought into neutral. The great toe is positioned in approximation to the surface. The degree of valgus is usually a few degrees and is somewhat dependent on the position of the second toe, allowing a small separation between the great and second toes.[65, 66]

Techniques used to stabilize the prepared surfaces can be stabilized range from simple suturing to the use of single or multiple smooth or threaded pins, lag and cross screws, variable pitch screws, and a variety of different plates. Precontoured plates with built-in DF and valgus simplify positioning and provide a strong and stable construct. (See the image below.) Shape-memory staples have been used as well.[67]  

Complications of arthrodesis include nonunion, malunion, hardware failure, and potential development of arthritis at the IP and tarsometatarsal (TMT) joints. Scar contracture and development of neuromas of the dorsomedial cutaneous nerve can be minimized using the medial exposure. Excessive DF is compensated for by IP joint flexion, resulting in painful dorsal keratosis over the IP joint, functional offloading of the hallux, and resultant metatarsalgia. On the other hand, excessive PF is compensated for by hyperextension at the IP joint and an increased load under the hallux MP and IP joints, with resultant painful calluses.

Postoperatively, the patient's foot is immobilized in a cast that extends beyond the toes or in a postoperative boot until union is seen on radiographs. Weightbearing is dependent on the security of fixation. Usually, patients who have undergone fixation augmented with plates can be allowed partial weightbearing when the wound is closed and dry at 2-3 weeks postoperatively. Full, unprotected weightbearing is allowed after union.

Arthroplasty

Excisional/interposition arthroplasty

Commonly known as the Keller technique, this procedure involves the excision of the base of the proximal phalanx. It is indicated in moderate to severe cases of hallux rigidus in individuals with low functional demands who desire the maintenance of joint motion. The original procedure involves the insertion of the medial capsule into the resected joint. It has been a popular procedure for many decades.[50, 68]

Complications are common, however, including hallux weakness, elevation, floppiness, a short hallux, and development of transfer lesions under the second MTP joint. For this reason, a number of modifications have been published. To avoid retraction of the extensor hallucis brevis (EHB), the thickened dorsal capsule is attached to the stumps of the EHB distal to the sesamoids. The surgeon should be careful not to excise too much of the base of the proximal phalanx and to excise only to the metadiaphyseal junction. Otherwise, a floppy toe may result. In addition, excessive resection further shortens the already shortened hallux.

A modified oblique cut of the proximal to spare the FHB has been described. Other modifications include the use of plantaris tendon as an interposition graft.

The procedure is usually performed via a medial incision. The capsule is exposed, the medial eminence is excised, and cheilectomy is performed. The soft tissues are released from the base of the proximal phalanx, including the capsule, the plantar plate, and FHB tendons. The EHB is tenotomized, and after the soft tissues are protected, the base of the proximal phalanx is excised with a saw or osteotome. The dorsal capsule is then mobilized and pulled into the resected joint space and sutured to the stumps of the FHB tendons. A 0.062 K-wire is placed across the IP and MTP joints for 3 weeks. Gentle ROM exercises are started once the K-wire has been removed.

Silicone arthroplasty

Silicone arthroplasty was originally designed to provide motion and length. A number of groups reported good results with the use of silicone arthroplasty, including, in 1993, Cracchiolo et al, who followed up 86 double-stem silicone implants in the first MTP joint for an average of 5.8 years and reported an 84% satisfaction rate.[69]

Many other studies, however, have shown that silicone lacks the structural durability and surface characteristics needed to withstand the severe shear and tension stresses generated by the repetitive motion associated with normal ambulatory activities. In addition, the friction of silicone on bone produced in hemiarthroplasty and at the bone-silicone interface in stemmed implants produces wear and fragmentation of the silicone. This can induce severe local synovitis and osteolysis, and systemic dissemination of silicone can result in granulomatous adenopathy after MTP arthroplasty.[70, 71, 72]

The use of grommets seems to improve the long-term results secondary to decreasing wear by protecting the silicone implant. Long-term reports on newer, more anatomic implant designs and the use in implants of better-grade silicone with better wear characteristics are not available. Given current knowledge and the long-term durability of alternative treatments, the use of silicone implants for the treatment of hallux rigidus is not advised.

Metallic arthroplasty

Total arthroplasty and hemiarthroplasty designs also have been available for the treatment of moderate-to-severe hallux rigidus. Several implants are available, including those used for hemiarthroplasties and full arthroplasties. They are made of cobalt chromium, titanium, or both, with a polyethylene bearing surface of ultrahigh molecular weight. Initial reports providing long-term results were mixed. Multinucleated foreign giant cells associated with fine particulate metallic wear debris have been reported in failed cases of titanium hemiarthroplasty with periarticular osteolysis.

The common view has been that the general use of metal implants for the treatment of hallux rigidus is probably not indicated, given the results demonstrated with arthrodesis, and that more research would be needed to define the long-term outcome of metallic implants and to compare their use with that of interposition arthroplasty and MTP joint fusion in severe hallux rigidus.

A systematic review and meta-analysis of 33 studies by de Bot et al compared metallic hemiarthroplasty with arthrodesis of the first MTP joint for treatment of hallux rigidus.[73] They found that whereas both procedures yielded excellent clinical outcomes and acceptable rates of complications and revisions, arthrodesis appeared to be superior in reducing pain, and metallic hemiarthroplasty is a reasonable alternative for patients who carry out activities requiring motion in the first MTP joint.