Osteoarthritis of the first metatarsophalangeal joint, referred to as hallux rigidus, inhibits weight transfer and adequate first metatarsophalangeal flexion force to aid toe-off during gait.1 Loss of first metatarsophalangeal joint motion can interfere with activities that require extensive toe motion, such as running and jumping, and may affect the choice of footwear. Arthrodesis has remained the gold standard procedure for the treatment of end-stage hallux rigidus. Driven by patient requests for reliable pain relief while preserving a mobile first metatarsophalangeal joint, several prosthetic designs have been developed.
Prosthetic designs and surgical techniques have advanced from early silicone implants that showed high rates of wear, osteolysis, and foreign body reactions.2 However, although some current first metatarsophalangeal total joint prosthetic designs have shown good short-term clinical outcomes, they have been associated with a high rate of aseptic loosening at midterm follow-up.3,4 Additional complications include soft tissue instability of the joint, transfer metatarsalgia, and substantial bone loss.5 Hemiarthroplasty by means of resurfacing of the damaged articular surface of the first metatarsal head has shown promising results.6–8 The advantage of hemiarthroplasty over total arthroplasty is the decreased amount of bone resection necessary, thus maintaining sufficient bone stock for fusion if needed.7 The current retrospective study evaluated the authors' experience with hemiarthroplasty of the first metatarsophalangeal joint using a metatarsal head resurfacing prosthesis and compared the results with findings reported in the literature.
The objectives of this study were to (1) determine the degree of postoperative pain relief; (2) evaluate functional outcomes and patient satisfaction; (3) determine the rate of secondary interventions; and (4) analyze radiographic alignment, osseous changes in the proximal phalangeal base, and implant loosening.
Materials and Methods
The ethics committee at the study institution approved this retrospective study, and all participants provided written informed consent.
The authors identified all patients who underwent first metatarsophalangeal hemiarthroplasty at their study institution from January 2011 to February 2016. Included in the study were all patients who underwent hemiarthroplasty of the first metatarsophalangeal joint with a metatarsal head resurfacing implant (HemiCAP; Arthrosurface Inc, Franklin, Massachusetts). The indication for hemiarthroplasty of the first metatarsophalangeal joint was a painful joint with advanced arthritic changes and no response to conservative measures but without degenerative sesamoid arthritis. The authors excluded all patients who had undergone an additional procedure at the time of the first metatarsophalangeal hemiarthroplasty implantation. Data collected included the following demographic factors: sex, age, body mass index, surgical indications, and previous hallux surgery. The score on the hallux metatarsophalangeal–interphalangeal scale of the American Orthopaedic Foot and Ankle Society (AOFAS) was recorded preoperatively and at last follow-up. Patients rated their pain on a visual analog scale ranging from 0 (no pain) to 10 (maximal pain) points before surgery and at final follow-up. For all patients, passive flexion, extension, and total range of motion were measured with a standard goniometer preoperatively, intraoperatively after implantation of the prosthesis and closure of the joint capsule, at 6-week follow-up, at 3-month follow-up, and at final follow-up. All patients rated their postoperative level of athletic activity compared with their preoperative level, and all patients indicated their level of satisfaction with the procedure.
Radiographic assessment included standard dorsoplantar, oblique, and lateral weight-bearing radiographs obtained preoperatively and postoperatively to determine the hallux valgus angle, intermetatarsal 1–2 angle, and dorsiflexion angle of the hallux. The hallux valgus angle was defined as the intersection of the longitudinal bisection of the first metatarsal and the first proximal phalanx. The intermetatarsal 1–2 angle was defined as the intersection of the longitudinal bisection of the first and second metatarsal shafts. The dorsiflexion angle, with reference points placed at the midpoint of the proximal and distal aspects of the diaphyses of the proximal phalanx and first metatarsal, was measured on lateral radiographs. Degenerative changes in the first metatarsophalangeal joint were graded (grades I–III) as described by Hattrup and Johnson.9 This grading system, developed for radiographic grading of hallux rigidus, is based on increasing osteophyte production, joint space narrowing, and subchondral sclerosis.6 The bone–implant interface was classified as normal or lucent, based on the most recent postoperative standard weight-bearing radiographs. Osseous changes of the base of the proximal phalanx also were noted on postoperative radiographs.
All surgical procedures were performed by the senior author (A.S.). The patient was placed in the supine position on the operating table, with a tourniquet placed on the thigh. The tourniquet was adjusted to 300 mm Hg. Preoperatively, all patients received a second-generation cephalosporin antibiotic. The first metatarsophalangeal joint was approached through a dorsomedial skin incision. The extensor hallucis tendon was retracted laterally during the entire procedure. The lateral and medial collateral ligaments and the plantar plate were released to allow full exposure of the joint. A drill guide was used to insert a Kirschner wire into the center of the first metatarsal, in line with the midline axis of the shaft. Fluoroscopy was performed to verify that the wire was positioned adequately in the sagittal and coronal planes. A step drill was inserted over the guidewire, and the metatarsal head was drilled until the proximal shoulder of the drill was flush with the plantar articular surface of the metatarsal head. The hole was tapped, and a taper post was inserted over the guidewire until the line on the driver was flush with the plantar articular surface of the metatarsal head. The articular geometry of the metatarsal head was checked with mapping measuring guides, and the final size of the implant was determined. The appropriate implant was chosen, and the metatarsal head was resurfaced to match the shape of the implant. After the capsule was closed, range of motion of the first metatarsophalangeal joint was documented.
Patients were allowed to ambulate while wearing a stiff-soled postoperative shoe under full weight bearing for 6 weeks. They were instructed to perform passive range of motion exercises of the great toe.
Statistical analysis was performed with R software (R Foundation for Statistical Computing, Vienna, Austria). Differences in visual analog scale score, AOFAS fore-foot score, radiographic alignment, and range of motion were determined with the Wilcoxon signed rank test. Statistical significance was defined as P<.05.
During the study period, 15 patients underwent hemiarthroplasty of the first metatarsophalangeal joint with the Hemi-CAP implant. Two patients underwent concomitant first tarsometatarsal arthrodesis and were excluded from the study. One patient was lost to follow-up because of an oncologic disease unrelated to the hemiarthroplasty. The remaining 12 patients (6 women, 6 men) underwent 12 hemiarthroplasty procedures of the first metatarsophalangeal joint. Mean age was 58.8 years (range, 46–82 years), mean body mass index was 24.9±2.8 kg/m2, and mean follow-up was 22.3±19.8 months.
Indications for hemiarthroplasty were hallux rigidus with symptomatic arthrosis for 11 patients and osteonecrosis of the metatarsal head for 1 patient. Of these patients, 3 had previously undergone hallux valgus surgery and 2 had undergone cheilectomy. All previous surgical procedures were performed through a dorsomedial skin incision.
For 11 patients, a 15-mm HemiCAP DF implant with a dorsal flange (1.5×3.5 mm, 7 patients; 2.5×3.5 mm, 3 patients; 2.5×4.5 mm, 1 patient) was used to resurface the metatarsal head. For 1 patient, a 12-mm classic (1.5×2.0 mm) HemiCAP implant without dorsal flange was used.
Mean forefoot AOFAS score improved significantly from 47.3±14.7 preoperatively to 71.8±15.2 at final follow-up (P=.033). Mean visual analog scale pain score decreased significantly from 7.0±2.3 preoperatively to 3.5±3.3 at final follow-up (P=.024). Patients rated satisfaction on the postoperative questionnaire as follows: very satisfied, 2; satisfied, 5; not satisfied, 3; and discomfort worse postoperatively than preoperatively, 2.
Mean passive range of motion of the first metatarsophalangeal joint was 32°±10° preoperatively; 93°±18° intraoperatively after implantation of the prosthesis and closure of the joint capsule; 30°±15° at 6-week follow-up; 38°±16° at 3-month follow-up; and 38°±19° at final follow-up (P=.26). Athletic activity level after hemiarthroplasty compared with preoperatively remained the same in 7 patients, decreased in 4, and increased in 1.
Preoperatively, 7 patients had grade II symptomatic arthrosis and 4 patients had grade III symptomatic arthrosis, as described by Hattrup and Johnson.9
As measured on standard dorsoplantar radiographs with patients bearing weight, mean preoperative and postoperative values were as follows: hallux valgus angle, 8°±8° and 9°±5°, respectively (P=.375); intermetatarsal 1–2 angle, 8°±5° and 7°±5°, respectively (P=.33); and dorsiflexion angle, 10°±4° and 12°±5°, respectively (P=.05). For all patients, the metatarsal bone–implant interface was classified as normal at last follow-up. Dysplastic changes of the base of the proximal phalanx were noted for 6 of 11 patients (12 total cases) at final follow-up (Figures 1–2).
A 77-year-old patient 4 years after implantation of a first metatarsophalangeal classic HemiCAP (Arthrosurface Inc, Franklin, Massachusetts) hemiarthroplasty prosthesis without dorsal flange. The patient was very satisfied with the postoperative functional result and had no pain. Standard weight-bearing dorsoplantar (A), oblique (B), and lateral (C) postoperative radiographs of the foot showed significant osseous dysplastic changes of the proximal phalangeal base caused by the resurfacing implant of the metatarsal head.
A 49-year-old patient 1 year after implantation of a first metatarsophalangeal HemiCAP DF (Arthrosurface Inc, Franklin, Massachusetts) hemiarthroplasty prosthesis with dorsal flange. Standard weight-bearing dorsoplantar (A) and lateral (B) radiographs of the foot.
No complications, including infection or delayed wound healing, were seen during the postoperative period. During follow-up, 5 patients underwent therapeutic infiltration of the first metatarsophalangeal joint with 20 mg of triamcinolone for pain and loss of range of motion. These 1-time infiltrations were performed at 2, 4, 5, 7, and 10 months postoperatively and resulted in decreased pain and increased joint motion. In addition, 1 patient underwent another open adhesiolysis of the first metatarsophalangeal joint 9 months postoperatively. Another patient underwent interposition arthrodesis because of hallux varus malalignment and pain 53 months postoperatively. To fill the small osseous metatarsal defect and prevent shortening of the first ray, 5 mL of demineralized bone matrix was used (Grafton DBM Putty; Argon Dental, Bingen, Germany).
The main finding of this study was that resurfacing of the metatarsal head by the HemiCAP implant significantly decreased pain. However, this small patient cohort also showed a small gain in range of motion (6°), a high rate of patient dissatisfaction (5 of 12), and a high rate of secondary interventions.
The goal of every arthroplasty is to significantly decrease pain and increase the function of the specific joint. The current study showed a significant decrease in pain from a visual analog scale score of 7.0 (on a scale of 0 to 10) preoperatively to 3.5 at final follow-up. Further, the study found a significant improvement in forefoot AOFAS score from 47.3 preoperatively to 71.8 at final follow-up. These observations are in accordance with published data on HemiCAP hemiarthroplasty of the first metatarsophalangeal joint.6,7,10,11
Nawoczenski et al12 evaluated 33 volunteers who had no history of foot or ankle pathology and found mean dorsiflexion of the first metatarsophalangeal joint during gait of 42°. Therefore, the goal of first metatarsophalangeal arthroplasty must be to establish the range of motion required for normal gait patterns. The current study found mean passive range of motion of the first metatarsophalangeal joint of 38° at final follow-up, improving by 6° from preoperatively. This range of motion is not as favorable as the mean postoperative range of motion of 45° to 66° at final follow-up after implantation of the Hemi-CAP reported by other studies.7,10,11
Dorsiflexion contracture may not disturb gait but probably alters loading patterns similar to arthrodesis.5 The remaining range of motion may not be sufficient to avoid long-term interphalangeal joint degeneration; therefore, first metatarsophalangeal hemiarthroplasty may not avoid one of the primary complications observed with first metatarsophalangeal arthrodesis.5 Studies with longer follow-up are needed to clarify whether the interphalangeal joint undergoes degenerative changes after HemiCAP implantation similar to the changes observed after arthrodesis. In the current cohort, 5 of 12 patients were not satisfied with the postoperative result. In sharp contrast, Carpenter et al6 and Kline and Hasselman7 reported a 100% satisfaction rate at mean follow-up of 27 months. According to Kline and Hasselman,7 all patients reported excellent satisfaction at 60-month follow-up, with 11.5% reporting that they were very satisfied and 88.5% reporting that they were satisfied.
Why did the patients in the current cohort show results inferior to those of other studies? A possible explanation is the use of less aggressive soft tissue release during implantation of the prosthesis. However, the current study recorded mean passive range of motion of 93° intraoperatively, which is in accordance with the range of dorsiflexion of 75º to 90º reported by Kline and Hasselman.7 Another reason for the difference in outcomes may be the less aggressive postoperative regimen used in the current study. The current patients wore a stiff-soled postoperative shoe for 6 weeks to allow adequate osseointegration of the implant and normal wound healing. Kline and Hasselman7 encouraged patients to walk as soon as possible postoperatively and recommended the use of regular shoes at the first postoperative visit at 2 weeks. In contrast, the current authors instructed patients to perform aggressive passive range of motion exercises of the first metatarsophalangeal joint immediately postoperatively.
In the current series, 5 patients underwent therapeutic infiltration because of loss of range of motion and pain during the follow-up period, 1 underwent additional adhesiolysis, and 1 underwent interpositional arthrodesis because of hallux varus malalignment and pain. No revision was required for implant loosening. This high number of secondary interventions is in contrast to previous reports. Of the 30 cases of hemiarthroplasty reported by Kline and Hasselman,7 4 required revision 3 years postoperatively because of phalangeal pathology. Of these patients, 3 underwent conversion to Silastic implants and 1 underwent arthrodesis.7 Circi et al10 reported that, after hemiarthroplasty, 3 of 12 patients underwent arthrodesis or periosteal interposition arthroplasty following implant removal during the follow-up period because of resting pain and painful range of motion. The HemiCAP prosthesis was designed to allow immediate rigid fixation and thereby prevent subsidence of the implant.13
During radiologic follow-up, none of the implants in the current cohort showed a lucent line in the bone–implant interface. Even for 1 case of osteonecrosis of the metatarsal head after previous distal metatarsal osteotomy for hallux valgus correction, the tapered screw fixation component of the implant bypassed the area of osteonecrosis and provided secure fixation of the implant.
At final radiographic follow-up, osseous dysplastic changes of the phalangeal side of the first metatarsophalangeal joint were observed for 6 of 11 patients (12 total cases), similar to the osseous changes that occur in the acetabulum after hemiarthroplasty of the hip. This finding may be in concordance with the radiographic changes observed by Kline and Hasselman.7 However, no percentage of osseous changes of the proximal phalangeal base after resurfacing of the metatarsal head has been reported in the literature.
The current study had some limitations. The number of patients was small, and the study design was retrospective. A further limitation was the relatively short follow-up period. The lack of a matched control group precludes direct comparison of the results with arthrodesis for treating symptomatic osteoarthritis of the first metatarsophalangeal joint. However, the authors assessed function, subjective outcomes, and radiologic results, which are the most important elements of success in joint replacement surgery. Further randomized studies are necessary to determine whether the range of motion that is maintained compared with arthrodesis with similar pain relief and clinical function justifies the additional costs of first metatarsophalangeal hemiarthroplasty.
The HemiCAP offered good pain relief and improved AOFAS forefoot score in this study cohort. However, the authors found the results unpredictable. Because of the relatively high dissatisfaction rate and the high rate of secondary interventions, the authors have limited the indication for first metatarsophalangeal hemiarthroplasty to patients who are reluctant to undergo first metatarsophalangeal arthrodesis.
- Granberry WM, Noble PC, Bishop JO, Tullos HS. Use of a hinged silicone prosthesis for replacement arthroplasty of the first metatarsophalangeal joint. J Bone Joint Surg Am. 1991; 73(10):1453–1459. doi:10.2106/00004623-199173100-00002 [CrossRef]
- Freed JB. The increasing recognition of medullary lysis, cortical osteophytic proliferation, and fragmentation of implanted silicone polymer implants. J Foot Ankle Surg. 1993; 32(2):171–179.
- Chee YH, Clement N, Ahmed I, Thomson CE, Gibson JN. Functional outcomes following ceramic total joint replacement for hallux rigidus. Foot Ankle Surg. 2011; 17(1):8–12. doi:10.1016/j.fas.2009.11.005 [CrossRef]
- Dawson-Bowling S, Adimonye A, Cohen A, Cottam H, Ritchie J, Fordyce M. MOJE ceramic metatarsophalangeal arthroplasty: disappointing clinical results at two to eight years. Foot Ankle Int. 2012; 33(7):560–564. doi:10.3113/FAI.2012.0560 [CrossRef]
- Horisberger M, Haeni D, Henninger HB, Valderrabano V, Barg A. Total arthroplasty of the metatarsophalangeal joint of the hallux. Foot Ankle Int. 2016; 37(7):755–765. doi:10.1177/1071100716637901 [CrossRef]
- Carpenter B, Smith J, Motley T, Garrett A. Surgical treatment of hallux rigidus using a metatarsal head resurfacing implant: mid-term follow-up. J Foot Ankle Surg. 2010; 49(4):321–325. doi:10.1053/j.jfas.2010.04.007 [CrossRef]
- Kline AJ, Hasselman CT. Metatarsal head resurfacing for advanced hallux rigidus. Foot Ankle Int. 2013; 34(5):716–725. doi:10.1177/1071100713478930 [CrossRef]
- Hasselman CT, Shields N. Resurfacing of the first metatarsal head in the treatment of hallux rigidus. Tech Foot Ankle Surg. 2008; 7(1):31–40. doi:10.1097/BTF.0b013e318165c356 [CrossRef]
- Hattrup SJ, Johnson KA. Subjective results of hallux rigidus following treatment with cheilectomy. Clin Orthop Relat Res. 1988; 226:182–191.
- Circi E, Tuzuner T, Sukur E, Baris A, Kanay E. Metatarsal head resurfacing arthroplasty in the treatment of hallux rigidus: is it a reliable treatment option?Musculoskelet Surg. 2016; 100(2):139–144. doi:10.1007/s12306-016-0410-7 [CrossRef]
- Aslan H, Citak M, Bas EG, Duman E, Aydin E, Ates Y. Early results of HemiCAP resurfacing implant. Acta Orthop Traumatol Turc. 2012; 46(1):17–21. doi:10.3944/AOTT.2012.2610 [CrossRef]
- Nawoczenski DA, Baumhauer JF, Umberger BR. Relationship between clinical measurements and motion of the first metatarsophalangeal joint during gait. J Bone Joint Surg Am. 1999; 81(3):370–376. doi:10.2106/00004623-199903000-00009 [CrossRef]
- Kline AJ, Hasselman CT. Resurfacing of the metatarsal head to treat advanced hallux rigidus. Foot Ankle Clin. 2015; 20(3):451–463. doi:10.1016/j.fcl.2015.04.007 [CrossRef]