Secondary Logo

Outcomes after Interposition Arthroplasty for Treatment of Hallux Rigidus

Kennedy, John, G; Chow, Francis, Y; Dines, Joshua; Gardner, Michael; Bohne, Walther, H

Clinical Orthopaedics and Related Research: April 2006 - Volume 445 - Issue - p 210-215
doi: 10.1097/01.blo.0000201166.82690.23

Interposition arthroplasty reportedly improves outcomes after traditional salvage procedures for advanced hallux rigidus. We hypothesized this procedure can provide pain relief and satisfactory function with few complications. We examined 18 patients with severe articular cartilage loss who received 21 interposition arthroplasties. The patients a mean age was 56 years. They had a mean followup of 38 months. All patients had substantial loss of articular cartilage when examined intraoperatively. Patients were evaluated using the American Orthopaedic Foot and Ankle Society and Short Form-36 scores. All 18 patients had pain relief, and 17 of 18 patients said they would have the procedure again. The mean postoperative increase in range of motion of the first meta-tarsophalangeal joint was 37°. The mean American Orthopaedic Foot and Ankle Society and Short Form-36 scores were 78.4 and 96.3, respectively. The complication rate was 6%. Results of our study indicate that interposition arthroplasty relieves pain and restores motion in patients with advanced hallux rigidus and may offer a reliable option to fusing the joint.

Level of Evidence: Therapeutic study, Level IV (case series). See the Guidelines for Authors for a complete description of levels of evidence.

From the Hospital for Special Surgery, New York, New York.

Received: June 29, 2004

Revised: January 30, 2005; July 18, 2005; November 8, 2005

Accepted: November 10, 2005

Each author certifies that he or she has no commercial associations (eg, consultancies, stock ownership, equity interest, patent/licensing arrangements, etc.) that might pose a conflict of interest in connection with the submitted article.

Each author certifies that his or her institution has approved the human protocol for this investigation and that all investigations were conducted in conformity with ethical principles of research, and that informed consent was obtained.

Correspondence to: J. G. Kennedy, MD, FRCS (Orth), Hospital for Special Surgery, 523 East 72nd Street, Suite 514, New York, NY 10021. Phone: 212-606-1104; Fax: 212-717-1016; E-mail:

Hallux rigidus is a common osteoarthritic condition of the foot. Several surgical procedures have been described to treat this condition when conservative measures have failed.21,26 In general, cheilectomy10,15,16 and osteotomy18 aim to preserve the metatarsophalangeal joint and are used when there is minimal evidence of joint destruction. When the joint cannot be salvaged, excision arthroplasty, replacement arthroplasty, and arthrodesis offer symptomatic relief to most patients.2,6,13,17,20,24,28 However, these procedures have limitations, particularly for more active patients.2,13,20 Transfer metatarsalgia, malposition deformity of the hallux, and weakness in the push-off phase of gait have been described after excision arthroplasty.11 Arthrodesis also has been associated with significant complications, including failure of fixation,22 malposition,5 non-union,4 shortening, and transfer metatarsalgia.5 Prosthetic joint replacement with silicone implants have been associated with high failure rates and silicone debris-induced synovitis.8,23

To address these limitations, interposition arthroplasty was described by Hamilton et al,9 and a modified tendon interpositional arthroplasty was described by Cosentino.3 An interpositional arthroplasty maintains joint motion by using capsular interposition in a joint that is beyond salvage by cheilectomy or osteotomy. However, the excellent outcomes reported in initial reports have been challenged.14 The conflict with the reported outcomes may be because of patient selection, differences in methods of outcome analysis, or a combination.

We hypothesized that interposition arthroplasty would provide a surgical solution to hallux rigidus and provide a pain-free and functional joint with fewer limitations and complications than traditional treatments.

Back to Top | Article Outline


We retrospectively evaluated 21 consecutive patients treated with interposition arthroplasty of the first metatarsophalangeal joint from 1996 to 2000. Eighteen of the 21 patients (21 feet) were available for assessment at a minimum of 30 months followup. The mean followup was 38 months (range, 30-89 months). Three patients had bilateral procedures. Three patients were not available for followup. Two of these three patients had moved, and one patient had died. Eleven of the 18 patients were women. The mean age at the time of surgery was 56 years (range, 24-72 years). All patients were diagnosed with hallux rigidus. Diagnosis was based on clinical and radiographic criteria. All patients complained of pain in their first metatarsophalangeal joints and limited range of motion (ROM). The severity of the disease was graded by radiographic appearance using Hattrup and Johnson's classification.10 Three feet were classified as having radiographic Grade II hallux rigidus and 18 feet were classified as having Grade III hallux rigidus.

After failed nonoperative treatment, the patient selection for interposition arthroplasty was based on clinical, radiographic, and preoperative criteria as defined by the operating surgeon. Patients who were active recreationally and who were opposed to joint fusion were included. Recreationally active was defined as a patient who walked more than 5 miles a week. Radiographic criteria for inclusion included Grade II or Grade III degenerative changes on either side of the joint on the lateral and the antero-posterior views. The final criterion could only be met at the time of surgery when the articular damage could be properly assessed. Interposition arthroplasty was used after a cheilectomy if the condition of the remaining articular surfaces reflected the initial radiographic appearance with gross articular loss. Interposition arthroplasty was deferred if the condition of the remaining cartilage was intact on both joint surfaces after cheilectomy. The final decision to proceed with interposition arthroplasty was made after the joint was inspected at the time of surgery. Exclusion criteria for interposition arthroplasty included patients who were not interested in restoration of joint function and who wanted pain relief alone. These patients often had almost no joint motion at presentation and were offered a surgical arthrodesis. Patients who had a short first ray compared with the second or third ray also were excluded. This protected against surgically induced transfer metatarsalgia. Elite athletes, defined as anyone who was a professional athlete or who pursued a competitive sporting activity five or more times a week, also were excluded from having the procedure. These patients were advised to have a cheilectomy as a temporizing procedure, or a joint fusion, in the belief that the interposed soft tissue would be subject to excess stress and would wear rapidly.

All surgeries were completed by the senior authors (WB, JD) in a standard fashion through a dorsomedial incision. After release of the extensor hallucis brevis, the metatarsophalangeal joint capsule was incised longitudinally and reflected from the bone to expose the joint. The capsule was reflected from the entire circumference of the base of the proximal phalanx, leaving a pedicle of capsule attached dorsally and laterally around the head of the metatarsal. The dorsal osteophyte and 25% of the dorsal aspect of the metatarsal head were resected. Any prominent medial and lateral eminences, if present, were removed. The proximal phalanx then was resected at its base. Approximately 10% of the base was resected to allow decompression of the joint and retain the attachment of the flexor hallucis brevis tendon. The dorsal joint capsule and the extensor hallucis brevis tendon were mobilized as a flap to interpose the space resulting from the bone resection. The interposed soft tissue was anchored to the flexor hallucis brevis and the intersesamoid ligament on the plantar aspect with three number 2 polyester sutures (Figs 1, 2). After 2 weeks of immobilization wearing a bulky dressing, patients were encouraged to begin weightbearing. Passive and active ROM exercises were commenced at 4 weeks. Patients were allowed to walk wearing a regular shoe 6 weeks postoperatively. Patients were followed up at 2 weeks, 4 weeks, and 6 weeks postoperatively.

Fig 1

Fig 1

Fig 2

Fig 2

All patients were assessed with the Short Form-36 Health Survey (SF-36)27 and the American Orthopaedic Foot and Ankle Society (AOFAS) clinical rating system for the hallux.12 The SF-36 Health Survey is a self administered, patient-based scoring system. It consists of 36 questions with eight scales leading to two summary measures describing physical and mental health, respectively. This summary outcomes measure has been validated to have internal and external consistency and reliability and has increasingly been used to assess orthopaedic conditions.7,25 The AOFAS clinical rating system is a 100-point physician-based scale with scoring based on pain (40 points), function (45 points), and alignment (15 points). Patients were asked to grade their pain as no pain, mild, moderate, or severe. Preoperative and postoperative ROM were measured with a standard goniometer (Howmedica, Rutherford, NJ). Preoperative ROM was measured by the operating surgeon, and postoperative measurements were made by the two principle investigators (JK, FC) who were not the primary surgeons to reduce bias.

Two-tailed p values were used to establish significant differences. The relationship of outcome to age, gender, and disease grade was examined by multiple regression analysis. A p value < 0.05 was considered significant. We used the Spearman correlation coefficient to find associations between the SF-36 and AOFAS scores. Correlation was expressed as the coefficient of determination and assigned r2. All data were analyzed using statistical software (SPSS 12.0, SPSS Inc, Chicago, IL).

Back to Top | Article Outline


All 18 patients had less pain than they experienced preoperatively.

Sixteen of 18 patients stated they had little or no pain. Two patients stated they had moderate pain on exercise, but that pain was less than preoperatively. The mean pre-operative ROM in the first metatarsophalangeal joint was 27° (range, 5°-35°). The mean postoperative ROM in the first metatarsophalangeal joint was 64° (range, 25°-70°). All 18 patients were satisfied with the position and appearance of their first toe.

The mean postoperative AOFAS score was 78.4 points (range, 49-100 points). The mean postoperative SF-36 score was 96.3 points (range, 64.91-115.7 points). When split into the two outcome measure scores, the mean physical outcome score was 47.3 (range, 25.2-60.3), and the mean mental outcome measure score was 49.1 (range, 37.6-61.5). The SF-36 correlated (r2 = 0.62) with the AOFAS hallux score. Gender, age at the time of surgery, and radiographic grade had no influence on outcome.

There were few reported complications. Restricted motion of the first metatarsophalangeal joint less than 20° occurred in two patients postoperatively. One of these patients required manipulation under anesthesia to improve motion. This occurred 2 months after the index procedure, and the patient was satisfied with the outcome. At the latest followup, the patient has a painless joint with 32° motion. The patient who reported stiffness postoperatively continues to have a painless joint at the latest followup. However, the arc of motion is restricted to 30° when passively moved and was restricted to 20° when the patient was asked to actively range the joint. The patient was not satisfied with the final ROM, and did not want manipulation or any other procedure to address the limitation in motion. Transfer metatarsalgia occurred in one patient who subsequently had a stress fracture of the second metatarsal shaft develop. This was secondary to a relatively short first ray after resection of the joint and an excess of mobility of the first tarsometatarsal joint. This produced an excess of force through the overloaded second metatarsal. The stress fracture healed without surgical intervention within 10 weeks.

Back to Top | Article Outline


Before the introduction of interposition arthroplasty, surgical treatment of moderate to severe hallux rigidus was confined to excision arthroplasty, prosthetic replacement, or arthrodesis. These methods of treatment have substantial functional limitations and complication rates.2,4,5,8,11,13,20,22,23 The goal of any surgical intervention for an arthritic joint should be establishment of a pain-free and functional joint. We found interposition arthroplasty can satisfy these criteria. The procedure also had a low rate of complications compared with rates in historical reports for traditional surgical treatments (Table 1).5,9,14,15,23,29



Our study has several limitations. Patients did not complete an AOFAS score or SF-36 score preoperatively; therefore, conclusions cannot be made regarding improvement in postoperative scores. All 18 patients reported improvements in pain compared with preoperatively. This is an objective measurement that cannot be biased by the investigators and validates the subjective postoperative AOFAS and SF-36 scores. Inherent bias by nonblinded data collection also may confound the results reported. This is a problem in many retrospective studies but should not diminish the comparative impact of the investigation when judged against other nonblinded retrospective studies.9,14

Studies detailing differences from the technique described by Hamilton et al,9 have yielded satisfactory outcomes.1,19,23 However, these studies were limited by small sample size and lack a standardized or validated outcome score.1,19,23 In the absence of standardized outcome scoring, meaningful comparison could not be made with other investigations. Data supporting the use of the interposition arthroplasty are sparse and inconclusive. We found interposition arthroplasty can be a useful procedure for treatment of hallux rigidus by using standardized and validated outcome scales. These outcome scores have shown content validity in evaluating outcomes in hallux rigidus by the correlation shown between the AOFAS and the SF-36 scores, underscoring the value of this technique.

Lau and Daniels14 found interposition arthroplasty did not perform well in comparison with cheilectomy using the AOFAS and SF-36 outcome analyses. In one group of 19 patients, cheilectomy was used whereas the second group of 11 patients had interposition arthroplasties. Their technique of interposition arthroplasty was similar to that described by Hamilton et al.9 Meaningful data analysis is confounded by the comparison between two methods of treatment in patients who were not matched in severity of disease. The group of 11 patients treated with interposition arthroplasty seemed to have more severe disease, judging from the radiologic grading. However, there was no mention of the degree of joint destruction at the time of surgery. The authors did conclude that comparison between the two treatment modalities was difficult based on the difference in degree of disease at presentation. They suggested cheilectomy is the preferred method of treatment for moderate hallux rigidus, and interposition arthroplasty should be regarded as a salvage procedure for severe degenerative changes.14 Our indications for the interposition arthroplasty agree with those of Lau and Daniels.14 However, we do not think the data presented should implicate the interposition arthroplasty as a salvage procedure alone. The majority of patients treated with interposition arthroplasty had good to excellent AOFAS scores.

The patients in our study had few complications, unlike patients in other studies who had alternative methods of treatment and where incidences of transfer metatarsalgia and malposition were reported to be as much as 27% and 38%, respectively.4,5,11,13,30 There was no evidence of malposition in our patients, and only one patient had transfer metatarsalgia. The low rates of complication in our patients are not solely determined by the surgery, but also by the selection criteria of patients suitable for surgery. Selection criteria are critical for patients before interposition arthroplasty. Patients with longer second and third rays in comparison with the first metatarsal are more likely to have transfer metatarsalgia develop than patients with longer first rays.9 Excluding these patients from having interposition arthroplasties has reduced this complication. These patients can be treated with fusion of the joint. By resecting just 10% of the base of the proximal phalanx, the correct tension of the soft tissues about the joint is maintained. This prevents the malposition that can occur with the traditional Keller's excision arthroplasty, where 25% of the base is resected.29

Selecting recreational athletes rather than elite athletes also could have contributed to the successful outcomes in the active patients. No data are available regarding outcomes after any surgical treatment for hallux rigidus in the elite athlete; however, we were concerned that a soft tissue interposition in this cohort would have been compromised by the level of stress and wear on the graft. Patients who were more sedentary also were excluded from the study population. As such, results of our study suggest interposition arthroplasty can be used in an active population that anticipates continued activity postoperatively. This activity level can be substantiated by the physical component score of the SF-36. The physical component score has four subscales that include physical function, physical role, general health, and bodily pain. These domains most accurately reflect the activity of the patient and are the most responsive to surgical intervention.27 The patients in our study reported vitality similar to that of the general population of similar age measured by the physical component score subscale score. The data suggest this population was active, but was not at an elite level of activity that would be reflected by at least two standard deviations above the norm.27

We identified two patients with joint motion less than 20°. We initially hypothesized inadequate resection of less than 10% of the base of the proximal phalanx or of the dorsal or lateral osteophytes may have been responsible for the postoperative lack of motion. However, there were no clear radiologic differences between these two patients and the remainder of the cohort with respect to the amount of bone resected. When the two patients were examined, there was a reduced arc of motion of the central metatarsophalangeal joints and the first metatarsophalangeal joint on the surgically treated foot and on the contralateral side. Whether this occurred because of surgery, or whether it was a contributing factor to postoperative stiffness is not clear. It may be that there is a subset of patients with restricted motion in the metatarsophalangeal joints, and these patients are more prone to have postoperative stiffness.

Interposition arthroplasty of the first metatarsophalangeal joint is indicated for treatment of advanced hallux rigidus with joint destruction. The original interpositional arthroplasty as described by Hamilton et al9 has reproducible outcomes evaluated by contemporary outcome scoring systems. Increased motion of the first metatarsophalangeal joint and excellent pain relief are indications for interposition arthroplasty to be used in patients with advanced hallux rigidus.

Back to Top | Article Outline


1. Barca F. Tendon arthroplasty of the first metatarsophalangeal joint in hallux rigidus: preliminary communication. Foot Ankle Int. 1997;18:222-228.
2. Brage ME, Ball ST. Surgical options for salvage of end-stage hallux rigidus. Foot Ankle Clin. 2002;7:49-73.
3. Cosentino GL. The Cosentino modification for tendon interpositional arthroplasty. J Foot Ankle Surg. 1995;34:501-508.
4. Coughlin MJ. Arthrodesis of the first metatarsophalangeal joint. Orthop Rev. 1990;19:177-186.
5. Fitzgerald JA. A review of long-term results of arthrodesis of the first metatarsophalangeal joint. J Bone Joint Surg. 1969;51:488-493.
6. Flamme CH, Wulker N, Kuckerts K, Gosse F, Wirth CJ. Follow-up results 17 years after resection arthroplasty of the great toe. [Erratum in: Arch Orthop Trauma Surg. 1999;119:243] Arch Orthop Trauma Surg. 1998;117:457-460.
7. Gatchel RJ, Mayer T, Dersh J, Robinson R, Polatin P. The association of the SF-36 health status survey with 1-year socioeconomic outcomes in a chronically disabled spinal disorder population. Spine. 1999;24:2162-2170.
8. Gordon M, Bullough PG. Synovial and osseous inflammation in failed silicone rubber prostheses: a report of six cases. J Bone Joint Surg. 1982;64:574-590.
9. Hamilton WG, O'Malley MJ, Thompson FM, Kovatis PE. Capsular interposition arthroplasty for severe hallux rigidus. Foot Ankle Int. 1997;18:68-70.
10. Hattrup SJ, Johnson KA. Subjective results of hallux rigidus following treatment with cheilectomy. Clin Orthop Relat Res. 1988;226:182-191.
11. Henry AP, Waugh W, Wood H. The use of footprints in assessing the results of operations for hallux valgus: a comparison of Keller's operation and arthrodesis. J Bone Joint Surg. 1975;57:478-481.
12. Kitaoka HB, Alexander IJ, Adelaar RS, Nunley JA, Myerson MS, Sanders M. Clinical rating systems for the ankle-hindfoot, midfoot, hallux, and lesser toes. Foot Ankle Int. 1994;15:349-353.
13. Kitaoka HB, Patzer GL. Arthrodesis versus resection arthroplasty for failed hallux valgus operations. Clin Orthop Relat Res. 1998;347:208-214.
14. Lau JTC, Daniels TR. Outcomes following cheilectomy and interpositional arthroplasty in hallux rigidus. Foot Ankle Int. 2001;22: 462-470.
15. Mann RA, Clanton TO. Hallux rigidus: treatment by cheilectomy. J Bone Joint Surg. 1988;70:400-406.
16. Mann RA, Coughlin MJ, DuVries HL. Hallux rigidus: a review of the literature and a method of treatment. Clin Orthop Relat Res. 1979;142:57-63.
17. Mann RA, Oates JC. Arthrodesis of the first metatarsophalangeal joint. Foot Ankle. 1980;1:159-166.
18. Moberg E. A simple operation for hallux rigidus. Clin Orthop Relat Res. 1979;142:55-56.
19. Mroczek KJ, Miller SD. The modified oblique Keller procedure: a technique for dorsal approach interposition arthroplasty sparing the flexor tendons. Foot Ankle Int. 2003;24:521-522.
20. O'Doherty DP, Lowrie IG, Magnussen PA, Gregg PJ. The management of the painful first metatarsophalangeal joint in the older patient. J Bone Joint Surg. 1990;72:839-842.
21. Pfeffer GB. Cheilectomy. In: Kitaoka HB, ed. The Foot and Ankle.2nd ed. Philadelphia, PA: Lippincott Williams & Wilkins; 2002:119-134.
22. Riggs SA, Johnson EW Jr. McKeever arthrodesis for the painful hallux. Foot Ankle. 1983;3:248-253.
23. Shereff MJ, Jahss MH. Complications of silicone implant arthroplasty in the hallux. Foot Ankle. 1980;1:95-101.
24. Smith RW, Joanis TL, Maxwell PD. Great toe metatarsophalangeal joint arthrodesis: a user friendly technique. Foot Ankle. 1992;13: 367-377.
25. Taylor SJ, Taylor AE, Foy MA, Fogg AJ. Responsiveness of common outcome measures for patients with low back pain. Spine. 1999;24:1805-1812.
26. Thompson FM, Mann RA. Arthritis. In: Mann RA, Coughlin MJ, eds. Surgery of the Foot and Ankle.6th ed. St Louis, MO: Mosby; 1993:615-671.
27. Ware JE, Sherbourne CD. The MOS 36-Item Short-Form Health Survey (SF-36): I. conceptual framework and item selection. Med Care. 1992;30:473-483.
28. Wu KK. Arthrodesis of the metatarsophalangeal joint of the great toe with Herbert screws: a clinical analysis of 27 cases. J Foot Ankle Surg. 1993;32:47-52.
29. Wrighton JD. A ten year review of the Keller's operation at the Princess Elizabeth orthopaedic hospital, Exeter. Clin Orthop Relat Res. 1972;89:207-214.
30. Zembsch A, Trnka HJ, Ritschl P. Correction of hallux valgus: metatarsal osteotomy versus excision arthroplasty. Clin Orthop Relat Res. 2000;376:183-194.
© 2006 Lippincott Williams & Wilkins, Inc.