Outcomes of Resection and Joint-Preserving Arthroplasty for Forefoot Deformities for Rheumatoid Arthritis

Introduction

Foot problems are common with rheumatoid arthritis (RA),^19,28 frequently causing pain during weightbearing activities. ¹⁰ The typical rheumatoid forefoot deformity is characterized by hallux valgus and subluxation or dislocation of the lesser toes at the metatarsophalangeal (MTP) joints. ³⁸ Patients with RA may undergo joint surgery to improve physical function and quality of life. ²⁵

An increasing number of RA patient-required foot arthroplasties have been reported in a Japanese cohort study. ³⁰ Conventionally, resection arthroplasty was widely performed to treat forefoot deformities associated with RA.^{4,8,15,24,27,40} However, this procedure sacrifices the function of the MTP joint. The clinically beneficial effects of disease-modifying antirheumatic drugs (DMARDs), including methotrexate and/or biologic DMARDs (b-DMARDs), have contributed to the increase in the number of patients who achieve clinical remission. ¹² Advanced disease control for RA involves prevention of joint destruction and joint remodeling even in joints damaged by RA. ¹⁸ For the rheumatoid forefoot surgery, other alternative operative procedures that can preserve the MTP joint have become reasonable for patients with mild to moderate disease. Several joint-preserving procedures have been reported to yield good clinical results.^{11,13,14,33,34,36,42} We recently reported on correctional osteotomies of the first metatarsal and shortening oblique osteotomy (SOO) of the lesser metatarsals to preserve joint function as an alternative to joint-sacrificing surgeries. ³⁴ In the current study, we investigated the clinical and radiographic outcomes of resection and joint-preserving arthroplasty for forefoot deformities in patients with RA.

Methods

This retrospective study was approved by the ethics committee of our institution, and informed consent was obtained from all patients. Thirty-four feet of 29 women with RA underwent surgery for treatment of forefoot deformities in our institute from July 2010 to December 2014. All patients met the American Rheumatism Association 1987 revised criteria for RA. ¹ The joint-sacrificing procedure (arthrodesis or resection arthroplasty) was indicated for all rheumatoid forefoot deformities in our institute until 2011, and then the joint-preserving procedure was indicated for painful plantar callosities, deformity of the forefoot with subluxation or dislocation of 1 or more MTP joints, and mild (Larsen grade I or II) or moderate (Larsen grade III) destruction of the MTP joints from 2012. At present, resection arthroplasty of the metatarsal head still tends to be indicated for relatively older patients (over 65 years) with severe dislocation of the MTP joints, destruction of the metatarsal head, and with poor disease control due to comorbidity. Sixteen feet of 14 women (average age, 67.1 years; range, 53-82) underwent resection arthroplasty of the metatarsal head (resection group), and 18 feet of 15 women (average age, 61.3 years; range, 40-73) underwent an MTP joint-preserving procedure with correctional osteotomy of the first metatarsal and SOO for the lesser toe (joint preservation group). The SOO of the lesser metatarsal was done for second, third, fourth, and fifth lesser metatarsal in 14, 16, 11, and 9 feet, respectively. The minimum follow-up period was 24 months. We have already reported outcomes of a joint-preserving procedure with a shorter follow-up (minimum 6-month follow-up) in 17 patients included in this study. ³⁴ Two patients underwent resection arthroplasty on 1 foot and joint-preserving arthroplasty on the contralateral foot. In these 2 patients, the same evaluations were used for disease activity and clinical score relating to activities of daily living (ADL) in both feet. On the other hand, clinical score excluding ADL and radiographic outcomes were assessed in each foot separately. The average disease duration in the resection and joint preservation groups was 23.6 (range, 10-54) years and 19.1 (range, 4-47) years, respectively, and the average follow-up period was 37.3 (range, 24-71) months and 33.2 (range, 24-61) months, respectively. Twenty-four patients had been treated with conventional synthetic DMARDs (cs-DMARDs) preoperatively (methotrexate, n = 18; salazosulfapyridine, n = 6; bucillamine, n = 3; prednisolone, n = 17; tacrolimus, n = 6; shiosol, n = 1), and 6 patients had been treated with b-DMARDs preoperatively (infliximab, n = 1; etanercept, n =5). One patient was not taking any medication at the time of surgery. In 1 patient, cs-DMARDs was switched to b-DMARDs postoperatively and all patients except this patient had been treated with the same medications between pre- and postoperatively. The mean Disease Activity Score 28 (DAS28)–CRP ⁴¹ in the resection and joint preservation groups was 2.7 (range, 1.3-4.1) and 2.7 (range, 1.5-3.8), respectively, and the mean modified Health Assessment Questionnaire (mHAQ) ³⁵ score was 0.56 (range, 0-1.25) and 0.36 (range, 0-1.25), respectively (Table 1).

OPEN IN VIEWER

Table 1.

Patients’ Demographic and Clinical Parameters. ^a

Characteristic	Resection Group (n = 16)	Joint Preservation Group (n = 18)	P Value
Age, y	67.1 (53-82)	61.3 (40-73)	.277
Disease duration, y	23.6 (10-54)	19.1 (4-47)	.805
Follow-up period, mo.	37.3 (24-71)	33.2 (24-61)	.215
cs-DMARDs, No.
MTX	7	13	.163
Other	11	10	.497
PSL, No.	11	9	.315
b-DMARDs, No.	2	5	.405
DAS28-CRP	2.69 (1.37-4.12)	2.71 (1.45-3.76)	.587
mHAQ	0.56 (0-1.25)	0.36 (0-1.25)	.559

Abbreviations: cs-DMARDs, conventional synthetic disease-modifying antirheumatic drugs; MTX, methotrexate; PSL, prednisolone; b-DMARDs, biologic disease-modifying antirheumatic drugs; DAS28, Disease Activity Score 28; mHAQ, modified Health Assessment Questionnaire.

a

All values are expressed as mean (range) unless otherwise indicated.

Assessment

Clinical assessment

Preoperative and postoperative clinical evaluation included assessment of Japanese Society of Surgery of the Foot (JSSF) score using the standard rating system for the RA foot and ankle scale for pain (0-30 points), deformity (0-25 points), range of motion (ROM) (0-15 points), walking ability (0-20 points), and ADL (0-10 points).^31,32 The presence of symptomatic plantar callosities and residual toe deformities was examined postoperatively.

Radiographic assessment

The classification of Larsen et al ²¹ was used to assess the severity of destruction of the MTP joint. Grade I or grade II indicated mild destruction, grade III moderate destruction, and grade IV or grade V severe destruction. Preoperative and postoperative radiographic assessments included measurement of the hallux valgus angle (HVA), intermetatarsal angle of the first and second metatarsals (M1M2A), and intermetatarsal angle of the first and fifth metatarsals (M1M5A) on anteroposterior weightbearing radiographs using Miller’s method. ²⁹

Operative technique

The resection arthroplasty was performed using the method described by Hoffmann, ¹⁵ which involved resection of the second to fifth metatarsal heads. For the hallux, the medial MTP joint capsule was opened with a Y-shaped incision, and the flap was reflected distally. After the resection of the metatarsal head, the capsule was closed by advancing the Y-shaped incision into a V-shape, reefing to the metatarsal shaft through a drill hole while holding the phalangeal bones at the corrected position (Figure 1). We did not use longitudinally positioned Kirschner wires (K-wires). Manipulation of the proximal interphalangeal (PIP) joint was done in all feet. The extensor plication and dermodesis at the PIP joints were performed in 6 toes of 2 feet. In 7 toes of 3 feet, Z-lengthening of the extensor digitorum longus (EDL) tendon was performed with release of the extensor digitorum brevis (EDB) tendon.

OPEN IN VIEWER

Figure 1.

Radiographic findings before (A) and after (B) resection arthroplasty for severe destruction of the metatarsophalangeal (MTP) joints. Radiographic findings before (C) and after (D) resection arthroplasty for moderate destruction of the MTP joints.

In the joint preservation group, we modified the SOO procedure for correcting lesser toe deformities described by Hanyu et al ¹¹ (Figure 2). The metatarsal bone was fixed with a screw instead of a longitudinal K-wire from the distal phalanx to the metatarsal shaft (Figure 3). ³⁴ In the hallux, hallux valgus deformities were corrected using the modified Mann method (Figure 2B). A modified Mann method was applied in combination with elevation of the abductor hallucis tendon, which was dislocated to the plantar side of the hallux and removal of the inflamed connective tissue caused by bunions. Proximal osteotomy of the metatarsal bone was performed for shortening and varus correction. ⁵ Locking plates were used for fixation of the osteotomy site. For the big toe, a modified Mann method with plate (DARCO; Wright Medical Technology, Memphis, TN) fixation was performed in 7 feet. For the lesser toe, cortex screws (Modular Hand System; DePuy Synthes, Zuchwil, Switzerland) were used in 16 cases, and cannulated cancellous screws (DARTFIRE; Wright Medical Technology) were used in 2 cases to fix the osteotomy site according to our previously described procedure. ³⁴ Manipulation of the PIP joint was done in all feet. The extensor plication and dermodesis at the PIP joints were performed in 9 toes of 5 feet. In 39 toes of 12 feet, Z-lengthening of the EDL tendon was performed with release of the EDB tendon.

OPEN IN VIEWER

Figure 2.

Radiographic findings before (A) and after (B) shortening oblique osteotomy for the lesser metatarsal bone with screw fixation for moderate destruction of the metatarsophalangeal (MTP) joints. Correction osteotomy and plate fixation were performed for the first metatarsal bone. Radiographic findings before (C) and after (D) shortening oblique osteotomy for the lesser metatarsal bone with screw fixation for severe destruction of the MTP joints.

OPEN IN VIEWER

Figure 3.

Operative procedure for shortening oblique osteotomy for lesser metatarsal bone. An osteotomy is done first (A), and then the distal fragment is slid up at the osteotomy site and a drill hole for a 2.0-mm or 2.5-mm diameter screw is made along with the guidewire inserted from the center of the distal fragment at the level of the metatarsal neck (B). Bones are then fixed by selected length of a cannulated cancellous screw (C).

Postoperative management

In the resection group, a bulky compressive bandage was applied with the toe held in plantarflexion for 10 to 14 days. Patients were allowed to bear full weight and begin passive and active ROM exercises on the day after surgery. After removal of the bandage, the patients wore an arch support (custom-made orthosis to support the longitudinal arch) for 2 months. In the joint preservation group, a bulky compression bandage was applied with all toes held in plantarflexion for 7 to 10 days. Passive and active ROM exercises were started at 5 to 7 days. Heel bearing gait was permitted at 7 days if possible, and partial weightbearing with an arch support was permitted at 2 weeks. Full weightbearing with an arch support was permitted 3 weeks after surgery. The patients wore an arch support for 2 months. ³⁴

Results

Preoperative radiographic destruction of the MTP joint (Larsen grades II, III, IV, and V) was present in 0, 29, 39, and 12 joints in the resection group and 13, 67, 9, and 1 joints in the joint preservation group. The mean JSSF score improved significantly from 61.3 (range, 45.0-84.0) to 83.9 (range, 52.0-99.0) points in the resection group (P < .001). In the joint preservation group, the mean JSSF score improved significantly from 62.2 (range, 46.0-81.0) to 90.8 (range, 70.0-98.0) points (P < .001) (Table 2). In the subanalysis of ADL in the resection group, standing on the toe was easy in 6 feet, difficult in 7 feet, and impossible in 3 feet. In the joint preservation group, standing on toes was easy in 13 feet and difficult in 5 feet. No patients in this group were unable to stand on their toes. In the resection group, preoperative and postoperative radiographic evaluation showed that the HVA, M1M2A, and M1M5A improved significantly from 43.5 (range, 25.0-59.0) to 20.1 (range, 5.0-36.0) (P < .001), 11.1 (range, 0.0-19.0) to 9.3 (range, 1.0-18.0) (P = .020), and 32.6 (range, 23.0-45.0) to 28.5 (range, 21.0-36.0) (P = .040), respectively. In the joint preservation group, preoperative and postoperative radiographic evaluation showed that the HVA, M1M2A, and M1M5A also improved significantly from 30.3 (range, 4.0-56.0) to 23.7 (range, 4.0-44.0) (P = .020), 11.8 (range, 6.0-21.0) to 10.1 (range, 4.0-20.0) (P = .007), and 34.2 (range, 24.0-46.0) to 30.8 (range, 19.0-45.0) (P = .017), respectively (Table 2). In the resection group, recurrence of callosities and claw toe deformity was observed in 6 and 3 feet, respectively. In the joint-preserving group, recurrence of callosities and hammer toe deformity was observed in 1 foot each.

OPEN IN VIEWER

Table 2.

Clinical and Radiographic Outcomes of Resection and Joint Preservation Groups. ^a

Characteristic	Resection Group (n = 16)			Joint Preservation Group (n = 18)
	Preoperative	Postoperative	P Value	Preoperative	Postoperative	P Value
JSSF-RA scale (total)	61.3 (45-84)	83.9 (52-99)	.001 b	62.2 (46-81)	90.8 (70-98)	.001 b
General pain (30 points)	22.5 (20-30)	28.1 (20-30)	.003 b	20.0 (20)	28.9 (20-30)	.001 b
Deformity (25 points)	12.2 (5-17)	19.9 (12-25)	.001 b	14.7 (8-21)	22.0 (19-25)	.001 b
Motion (15 points)	10.1 (0-15)	12.0 (0-15)	.023 b	10.9 (0-15)	14.3 (13-15)	.004 b
Walking ability (20 points)	13.1 (10-20)	18.8 (10-20)	.003 b	11.4 (5-20)	18.9 (10-20)	.003 b
Activities of daily living (10 points)	3.3 (0-8)	5.4 (2-10)	.015 b	5.2 (2-9)	7.1 (3-10)	.012 b
HVA, deg	43.5 (25-59)	20.1 (5-36)	.001 b	30.3 (4-56)	23.7 (4-44)	.020 b
M1M2A, deg	11.1 (0-19)	9.3 (1-18)	.020 b	11.8 (6-21)	10.1 (4-20)	.007 b
M1M5A, deg	32.6 (23-45)	28.5 (21-36)	.004 b	34.2 (24-46)	30.8 (19-45)	.017 b

Abbreviations: HVA, hallux valgus angle; JSSF-RA scale, Japanese Society of Surgery of the Foot standard rating system for the RA foot and ankle scale; M1M2A, intermetatarsal angle of the first and second metatarsals; M1M5A, intermetatarsal angle of the first and fifth metatarsals.

a

All values are expressed as mean (range).

b

Statistically significant (P < .05).

Postoperative infection occurred in 1 foot each in the resection and joint preservation groups (Table 3). In 5 feet with recurrence of hallux valgus (HVA of more than 30 degrees), severe hallux valgus (HVA of greater than or equal to 47 degrees) was observed preoperatively despite the fact that the RA disease activity of these patients had reached remission at the final follow-up. Recurrence of claw toe deformity occurred in 3 feet in the resection group at the final follow-up, and these feet had mid-hindfoot deformities, including bony ankyloses. Bone union was achieved in all cases in the joint preservation group.

OPEN IN VIEWER

Table 3.

Number of Complications in the Resection and Joint Preservation Groups.

Complication	Resection Group (n = 16)	Joint Preservation Group (n = 18)
Callosities	6	1
Hallux valgus (HVA >30 degrees)	2	3
Hammer toe	0	1
Claw toe	3	0
Infection	1	1

Abbreviation: HVA, hallux valgus angle.

Discussion

Various modifications of resection arthroplasty have been performed with good clinical results in patients with RA.^{4,8,15,24,27,40} In our study, resection arthroplasty of the big and lesser metatarsal heads with the treatment of medial soft tissue for big toe (modified Lelièvre procedure) showed good clinical results. However, resection arthroplasty has some disadvantages: the metatarsal head with destruction may be allowed with resection of other heads even when 1 or 2 MTP joints are radiographically normal, and there is a risk of painful callosities with bony proliferation of the metatarsal stump with longer follow-up.^4,17,23 In addition, 1 study showed that with long-term follow-up, hallux valgus recurred in more than 50% of patients who underwent resection arthroplasty, ⁶ and another study showed that the recurrence of hallux valgus and toe deformity may cause patient dissatisfaction. ²⁶ Arthrodesis of the first MTP joint provides a low recurrence of the hallux valgus deformity but also has some disadvantages, including more degeneration of the interphalangeal joint,^2,39 excessive pressure on the first toe, ² and relatively high complication and reoperation rates. ¹⁶ Therefore, we indicated the arthrodesis for severe hallux valgus, and these patients have not been included in the current study.

On the other hand, various modifications of joint-preserving procedures have also been performed with good clinical results.^{3,11,13,14,33,34,36,42} Joint-preserving procedures for the correction of RA-associated forefoot deformities has some advantages. It contributes to reconstruction of the transverse metatarsal arch, corrects the deformity, and preserves the joint function (especially the fulcrum of the flexor tendon). ³⁴ Fukushi et al ⁹ reported that joint-preserving procedures for rheumatoid forefoot deformities resulted in better clinical outcomes than resection arthroplasty with regard to the function of the hallux and the alignment of the lesser toes. However, the joint-preserving procedure also has some disadvantages: potential recurrence of hallux valgus deformity, ³⁶ forefoot stiffness,^22,33 or severe destruction of the metatarsal head is not a good indication for a joint-preserving procedure.^9,33,42 The SOO method described by Hanyu et al ¹¹ involved temporary fixation of the osteotomy sites by 1.2-mm K-wires from the distal phalanx to the metatarsal shaft. The K-wires are removed 3 weeks postoperatively. Using a similar postoperative protocol, Hirao et al ¹⁴ showed an MTP joint ROM limitation of less than 70 degrees and indicated that decreased ROM in the MTP joints should be a factor that decreases the postoperative functional score. In patients with inactive RA, decreased ROM of the MTP joints is associated with decreased walking velocity and stride length. ²⁰ However, screw fixation has several advantages over K-wire fixation. Our procedure did not involve fixation of the MTP joint and was not invasive to the phalanges. Early postoperative ROM exercises for the MTP joint can be started, thus contributing to improved postoperative ROM and function of the feet. Because the shortening of the metatarsal bone was performed mostly by sliding the distal fragment at the osteotomy site, the amount of shortening was less than that in previous reports. ³⁴ Increasing the amount of shortening contributes to imbalance between the flexor tendon and extensor tendon. Therefore, to minimize the amount of shortening of the metatarsal bone, we released the extensor digitorum brevis tendon and performed Z-lengthening of the extensor digitorum longus tendon, soft tissue release, and synovectomy at the MTP joints. More patients in the joint preservation than resection group found it easy to stand on the toe. One study showed that disease activity often persisted in the foot even when patients were judged to be in remission based on the 28-joint score. ⁷

In the both groups, we performed a synovectomy of the MTP joint and restarted the b-DMARD after the operative wound had healed. However, despite the disease activity being well controlled in both groups, severe hallux valgus (HVA of greater than or equal to 47 degrees) was observed in cases of recurrence. Some reports of patients without RA who underwent proximal metatarsal osteotomy for hallux valgus have suggested that a severe preoperative HVA can be a risk factor for recurrence of hallux valgus. ³⁷ As in all feet with recurrence of hallux valgus, arthrodesis of the MTP joint should be indicated for severe hallux valgus with an HVA of more than 50 degrees. Bony ankyloses in the mid-hindfoot can reportedly cause recurrence of dislocation of the lesser toe MTP joint. ¹⁴ Bony ankyloses were found in the mid-hindfoot in 3 feet with recurrence of claw toe deformity. Yano et al ⁴³ reported delayed union in 16.5% of patients who underwent SOO. In our procedure, the osteotomy site was secured with a single screw. The fixation force at the osteotomy site would not be lost before bone union.

There are some limitations in the current study. First, the number of patients was too small to compare the clinical and radiographical outcomes of the 2 procedures, because both procedures showed good clinical outcomes in the current study. Second, the indications for resection and joint-preserving procedures were not randomly divided. The indications were dependent upon the patients’ age, control of disease activity, level of ADL, and destruction of the MTP joints. In the present study, radiographic destruction of the MTP joints higher than Larsen grade IV were seen in 51 and 10 toes in the resection and joint-preservation groups, respectively. We performed resection arthroplasty in 29 toes with moderate destruction of the MTP joint and a joint-preserving procedure in 10 toes with severe destruction of the MTP joint. Therefore, this study has some degree of selection bias. Arthrodesis of the big toe MTP joint was indicated for severe hallux valgus (HVA of more than 50 degrees). Patients with arthrodesis of the big toe MTP joint were excluded to investigate the clinical scores of resection arthroplasty and joint-preserving procedures. In addition, this study included both patients who underwent bilateral surgery and those who underwent unilateral surgery. Seven and 4 patients in the resection and joint preservation groups, respectively, underwent toe arthroplasty on the contralateral foot during the follow-up. Subjective evaluations of patients were recently shown to be important tools for assessment of the postoperative outcomes of RA. Thus, studies involving subjective evaluation should be performed.

Conclusion

The resection arthroplasty and joint-preserving procedure showed satisfactory short-term clinical outcomes. However, whether both procedures can maintain the good clinical results without the recurrence of forefoot deformity would be better assessed with careful follow-up of the current set of patients.

Supplemental Material