Clinical and Radiographic Results of Ankle Joint Preservation Surgery in Posterior Ankle Arthritis

Introduction

Joint preservation surgery for eccentric ankle arthritis has shown promising results, leading to its growing popularity in treating ankles with mild to moderate arthritis.^{1,12,14,16,17} To date, the majority of joint preservation modalities have primarily focused on correcting coronal plane deformities, specifically varus or valgus ankle arthritis.^{1,3,10-12,14,16} However, some ankles may demonstrate eccentric narrowing in the sagittal plane either at the anterior or posterior aspect of the tibiotalar joint space with no or minimal varus/valgus talar tilt, which can be referred to as anterior or posterior ankle arthritis, respectively (Figure 1). ⁶

OPEN IN VIEWER

Figure 1.

Weightbearing ankle standing radiographs of patients with posterior ankle arthritis. (A) Anteroposterior (AP) ankle view shows relatively well-preserved tibiotalar joint; however, lateral ankle view reveals an eccentric narrowing at the posterior aspect of the tibiotalar joint space (asterisk). (B) Although diffuse narrowing of the joint space, which suggests end-stage arthritis, is observed in the AP ankle view, the lateral ankle view reveals eccentric narrowing of the posterior aspect of the tibiotalar joint (arrowhead) whereas more than half of the joint space is preserved, suggesting the possibility of the joint preservation surgery. Weightbearing computed tomography (white boxes in A and B) confirmed eccentric narrowing of the posterior tibiotalar joint.

The fundamental principle of joint preservation surgery is shifting the weightbearing load from the degenerated area to the uninvolved area. ⁴ For example, valgus angulation of the tibial plafond or hindfoot results in uneven weightbearing load favoring the lateral side, and vice versa in patients with varus angulation. Therefore, correcting this angulation shifts the load toward the uninvolved side. On the other hand, little information is known on the measures to correct sagittal plane deformities from the perspective of the joint preservation method. Correcting sagittal plane deformity can be equally crucial for joint longevity because malalignment in the sagittal plane is known to significantly alter joint kinematics compared with that of the coronal plane, which can be compensated by the subtalar motion.^15,18,23

Recent studies have demonstrated that an anteriorly translated talus can be repositioned through a correction of distal tibial angle in the sagittal plane either in the setting of joint preservation surgery or total ankle arthroplasty.^2,20 However, no studies exist on the cause of posterior translation of the talus, its associated arthritis, and the measures to reposition the talus.

The authors of the current study have observed that the radiographs of ankles with posterior arthritis demonstrate plantarflexion of the talus, which can be seen in the flatfoot deformity. Therefore, we hypothesized that dorsiflexion of the talus through flatfoot reconstruction would restore joint space and improve patient outcomes in the setting of posterior ankle arthritis. The aim of this study was to investigate the radiographic and clinical results of joint preservation surgery in consecutive patients with posterior ankle arthritis.

Methods

Study Cohort

Between 2017 and 2018, a total of 16 ankles in 16 consecutive patients underwent joint preservation surgery for posterior ankle arthritis at our institution. Posterior ankle arthritis was defined as having both (1) eccentric narrowing of the joint space at the posterior aspect of the tibiotalar joint on weightbearing ankle lateral radiographs and (2) coronal talar tilt angle less than 4 degrees on weightbearing ankle anteroposterior (AP) radiographs. A cutoff of 4 degrees for the talar tilt angle was chosen based on findings from a previous weightbearing computed tomography (WBCT) study, in which ankles with a large talar tilt may be associated with the sagittal translation of the talus. ⁹ All patients complained of symptoms consistent with ankle arthritis, including pain and discomfort with ambulation, and were diagnosed clinically and radiographically by a senior surgeon. Patients aged ≥18 years, with a minimum of 2-year clinical and radiographic follow-up, were included in this analysis. Patients with (1) a history of fractures and dislocations around ankle joint, (2) septic arthritis, (3) rheumatoid arthritis, and/or (4) end-stage ankle arthritis were excluded. Ultimately, 10 ankles in 10 patients, with 3 male and 7 female, remained for analysis (Table 1). The study cohort’s mean age was 61.8 years (range, 53-68), and the mean body mass index was 26.7 (range, 20-31.6). The mean follow-up period of the study cohort was 2.3 years (range, 2-3.1). The hospital’s Institutional Review Board approved the study protocol, and informed consent was obtained from all patients under the Declaration of Helsinki.

OPEN IN VIEWER

Table 1.

Demographic Data of the Patients. ^a

No.	Sex	Age (y)	BMI	Side	Concurrent Procedures	Preop. Stage b	Postop. Stage b	FU (mon)
1	F	62	28.8	L	SA, DMCS, GR	II	I	37
2	F	59	27.0	L	SA, SMO, SE, GR	III	I	30
3	F	64	31.6	L	SA, DMCS, GR	III	II	29
4	M	65	24.0	L	SA, SMO, SE, GR	III	III	28
5	M	68	27.1	L	TA, SMO, GR	III	II	28
6	F	64	20.0	L	SA, SMO, DMCS, SE, GR	IV	II	24
7	F	61	24.0	L	SA, DMCS, DCMO, GR	IV	I	26
8	F	64	30.3	L	SA, DMCS, SE, GR	III	III	24
9	M	53	26.4	R	SA, SMO, DMCS, SE, GR	II	II	28
10	F	58	27.3	R	SA, SMO, DMCS, MDCO, SE, GR	IV	II	24

Abbreviations: BMI, body mass index; DCMO, distal chevron osteotomy for hallux valgus correction; DMCS, dynamic medial column stabilization using flexor hallucis tendon transfer; F, female; FU, follow-up; GR, gastrocnemius recession; L, left; M, male; MDCO, medial displacement calcaneal osteotomy; R, right; SA, subtalar arthrodesis; SE, spur excision; SMO, supramalleolar osteotomy; TA, triple arthrodesis.

a

All patients included in the study cohort did not have a history of ankle fracture or syndesmotic injury.

b

Stage represents radiographic osteoarthritis stage in the sagittal plane as described in Figure 3.

Radiographic Analysis

All patients completed a preoperative and minimum 2-year postoperative radiographic evaluation, including standard weightbearing ankle AP and lateral radiographs, foot AP view, and hindfoot alignment view. On the ankle AP radiographs, the talar tilt angle, medial distal tibial angle (MDTA), and talar center migration (TCM) were measured.^6,22,27 The TCM was defined as the distance between the tibia axis and the center of the talus, as previously described by Yi et al. ²⁷ On the lateral radiographs, the anterior distal tibial angle (ADTA), lateral TCM (LTCM), and talo–first metatarsal angle (Meary angle) were assessed.^9,14,27 The ADTA was measured by connecting the most anterior and posterior points of the distal tibia, ignoring osteophytes. The LTCM was defined as the distance between the center of circles encompassing the tibial plafond and talar dome in the sagittal plane. On foot AP radiographs, the talonavicular coverage angle was measured. ²⁸ On the hindfoot alignment view, the hindfoot moment arm and hindfoot alignment angle were assessed (Figure 2). ¹⁹ All radiographic measurements were performed using a Picture Archiving Communication System.

OPEN IN VIEWER

Figure 2.

Demonstration of radiographic measurements. (A) Talar tilt angle (positive value indicates varus tilt), medial distal tibial angle (MDTA), and talar center migration (TCM; a positive value indicates medial translation) were measured in weightbearing anteroposterior ankle view. (B) Talonavicular coverage angle (TNC) was measured in weightbearing anteroposterior foot view. (C) Hindfoot alignment angle (HAA; a positive value indicates varus angulation) and hindfoot moment arm (HMA; a positive value when the calcaneus is medial to the tibial axis) were measured in hindfoot alignment view. (D) Anterior distal tibial angle (ADTA), Meary angle (lateral talo–first metatarsal angle; a positive value indicates planus), lateral talar center migration (LTCM; a positive value indicates anterior translation, measured as described in the box with an asterisk) were measured in weightbearing lateral ankle view.

A radiographic classification system was used to grade the arthritis of the tibiotalar joint in the sagittal plane and to compare pre- vs postoperative results: stage I, early sclerosis and osteophyte formation without joint narrowing; stage II, narrowing of the joint space; stage III, obliteration of tibiotalar joint space <50%; stage IV, obliteration of tibiotalar joint space ≥50% (Figure 3).

OPEN IN VIEWER

Figure 3.

Radiographic arthritis stage in the sagittal plane (stage I to IV).

Additionally, WBCT analysis was performed in order to evaluate the joint space before and after surgery. WBCT was performed using a cone-beam CT scanner (Planmed Verity Extremity; Planmed Oy, Helsinki, Finland). To obtain the scan, the designated foot was placed on a dedicated platform, while the patient was positioned in a natural plantigrade stance with the ankle in the neutral position. Axial, sagittal, and coronal plane images of the ankle joint were obtained (tube voltage, 96 kB; tube current, 7.5 mAs; CTDIvol, 4.3 mGy; matrix size, 160 × 160 × 130 pixels; pixel size, 0.4 mm; slice interval, 0.4 mm).^8,9 All images were reconstructed with a thickness of 2 mm. The assessment was carried out by determining whether the joint space remained in the same stage (0) or improved (1), comparing pre- and postoperative images. An improvement in joint space was confirmed when the joint space in the narrowest area increased more than 2 times after surgery.

Radiographic measurements were performed on 2 separate occasions by 2 independent foot and ankle fellowship-trained investigators in a blinded manner. Each set of measures was obtained with a minimum interval of 6 weeks to minimize recall bias. Intraobserver and interobserver reliability were determined by calculating the intraclass correlation coefficients. Cohen kappa and linear weighted kappa scores were used to assess the reliability of the WBCT joint space analysis and the radiographic classification between the observers, respectively.

Results

Radiographic Results

A comparison of pre- and postoperative radiographic measurements is tabulated in Table 2. Preoperatively, the median LTCM and MDTA were −3.3 mm (IQR 0.9) and 87.5 degrees (IQR 3), indicating posterior translation of the talus and varus distal tibial alignment. The median Meary angle, talonavicular coverage angle, hindfoot alignment angle, and hindfoot moment arm were 32.8 degrees (IQR 15.1), 23.8 degrees (IQR 14.3), −8.5 degrees (IQR 5.5), and −10.5 mm (IQR 19), indicating lower medial longitudinal arch, forefoot abduction, and valgus hindfoot alignment.

OPEN IN VIEWER

Table 2.

Comparison of Pre- and Postoperative Radiographic Parameters.

	Preoperative			Postoperative			P Value
	Mean	Median	IQR	Mean	Median	IQR
Ankle AP view
TT	0.2	0	0.8	−0.8	0	1.5	.219
MDTA	87.5	87.5	3	92.8	92.3	4.6	<.05
TCM	−0.2	−2.5	6	−1.3	−0.5	3.5	.445
Ankle lateral view
ADTA	83.2	83	5	82.7	82.5	5	.738
Meary a	34.7	32.8	15.1	16.5	14.8	11.8	<.001
LTCM	−3.4	−3.3	0.9	−0.3	−0.3	1.4	<.001
Foot AP view
TNC	22.6	23.8	14.3	10.9	11.0	6.9	<.001
Hindfoot alignment view
HAA	−8.6	−8.5	5.5	−1.6	0	8.0	<.05
HMA	−10.3	−10.5	19.0	−6.8	0	7.0	.313

Abbreviations: AP, anteroposterior; ADTA, anterior distal tibial angle; HAA, mindfoot alignment angle; HMA, hindfoot moment arm; LTCM, lateral talar center migration; MAD, mechanical axis deviation; MDTA, medial distal tibial angle; TCM, talar center migration; TNC, talonavicular coverage angle; TT, talar tilt angle.

a

Meary, lateral talo–first metatarsal angle.

Postoperatively, the median LTCM significantly improved to −0.3 mm (IQR 1.4), suggesting the restoration of sagittal tibiotalar alignment. In addition, there were significant differences observed in the postoperative MDTA, Meary angle, talonavicular coverage angle, and hindfoot alignment angle.

Preoperatively, the arthritis stage in the sagittal plane were: 2 (20%) in stage II, 5 (50%) in stage III, 3 (30%) in stage IV. Postoperatively, this changed to 3 (30%) in stage I, 5 (50%) in stage II, 2 (20%) in stage III (Figures 4 and 5). The arthritis stage improved in 7 (70%) patients, whereas 3 (30%) remain unchanged in the same stage (Figure 6). However, WBCT analysis revealed that joint space was restored in all patients (Figure 7).

OPEN IN VIEWER

Figure 4.

Weightbearing anteroposterior and lateral radiographs of a 61-year-old woman with posterior ankle arthritis who underwent subtalar arthrodesis, dynamic medial column stabilization (flexor hallucis longus transfer to first metatarsal base), distal chevron metatarsal osteotomy (for hallux valgus correction), and gastrocnemius recession: (A) before, (B) 1 year after, and (C) 26 months after realignment surgery. The radiographic arthritis stage improved from stage IV to stage I.

OPEN IN VIEWER

Figure 5.

Weightbearing anteroposterior and lateral radiographs of a 62-year-old woman with posterior ankle arthritis who underwent subtalar arthrodesis, dynamic medial column stabilization (flexor hallucis longus transfer to first metatarsal base), and gastrocnemius recession. (A) Preoperative ankle standing radiographs show mild varus talar tilt in the anteroposterior view and eccentric narrowing of the posterior tibiotalar joint in the lateral view. (B) Postoperative radiographs at 37 months postoperation show the correction of the talar tilt angle in the anteroposterior view and the restoration of joint space in the sagittal plane.

OPEN IN VIEWER

Figure 6.

Changes in the radiographic arthritis stage in the sagittal plane. Seven ankles exhibited improvements, whereas 3 ankles (case numbers 4, 8, and 9) remained in the same stage.

OPEN IN VIEWER

Figure 7.

Weightbearing radiographs of 3 ankles that remained in the same radiographic arthritis stage despite their clinical improvements. Note that the conformity of the joint and posterior translation of the talus improved postoperatively (B, D, and F) compared with preoperative images (A, C, and E). Weightbearing computed tomography analysis (images in the box for each image) revealed the joint space restoration. (Procedures performed in A and B: subtalar arthrodesis [SA], supramalleolar osteotomy [SMO], dynamic medial column stabilization [DMCS, flexor hallucis longus transfer to first metatarsal base], spur excision [SE], and gastrocnemius recession [GE]; C and D: SA, SMO, SE, and gastrocnemius recession [GR]; E and F: SA, MCS, SE, and GR.)

On reliability analysis of the radiographic measurements, the intraobserver reliability ranged from 0.86 to 0.99 and the interobserver reliability, from 0.86 to 0.96 (Table 3). The weighted kappa coefficients of interobserver reliability for the radiographic arthritis stage was 0.85, and the Cohen kappa value of WBCT joint space analysis was 0.91 (95% confidence interval 0.64-0.98)

OPEN IN VIEWER

Table 3.

Interobserver and Intraobserver Reliability of the Radiographic Parameters.

Measurement	Reliability (95% Confidence Interval)
	Interobserver	Intraobserver
Ankle anterior view
TT	0.96 (0.85-0.99)	0.99 (0.96-0.99)
MDTA	0.93 (0.72-0.98)	0.92 (0.64-0.98)
TCM	0.95 (0.76-0.99)	0.98 (0.88-0.99)
Ankle lateral view
ADTA	0.86 (0.33-0.97)	0.92 (0.62-0.98)
Meary	0.96 (0.83-0.99)	0.98 (0.93-0.99)
LTCM	0.87 (0.48-0.97)	0.86 (0.46-0.96)
Foot AP view
TNC	0.95 (0.80-0.99)	0.98 (0.93-0.99)
Hindfoot alignment view
HAA	0.88 (0.44-0.97)	0.91 (0.64-0.98)
HMA	0.91 (0.48-0.98)	0.97 (0.87-0.99)

Abbreviations: AP, anteroposterior; ADTA, anterior distal tibial angle; HAA, mindfoot alignment angle; HMA, hindfoot moment arm; LTCM, lateral talar center migration; MDTA, medial distal tibial angle; TCM, talar center migration; TNC, talonavicular coverage angle; TT, talar tilt angle.

Discussion

This study reviewed the clinical and radiographic results of joint preservation surgery in patients with posterior ankle arthritis. Radiographic analysis revealed plantarflexion of the talus, lower medial longitudinal arch, and valgus hindfoot alignment in the study cohort, all of which suggest its association with flatfoot deformity. Depending on the severity of the deformity, subtalar arthrodesis was performed in most patients to reposition the talus and restore medial longitudinal arch height. In patients demonstrating varus distal tibial plafond and medial translation of the talus, an additional SMO was performed. Clinical scores improved significantly in all patients. The radiographic stage improved in 7 patients (70%) and remained unchanged in 3 patients (30%). Radiographic parameters indicated anterior translation and dorsiflexion of the talus following flatfoot reconstruction. None of the patients required conversion to joint-sacrificing surgery at a minimum 2-year follow-up.

In the management of ankle arthritis, the results of joint preservation surgery have been encouraging with early to intermediate clinical outcomes.^1,12-14,16 However, previous joint preservation techniques have primarily aimed to correct the deformity in the coronal plane, which has been categorized as either varus or valgus malalignment. Arthritis appreciated in the sagittal plane has been considered relatively unremarkable, often described as a subsequent finding that follows coronal plane deformity. However, as seen in the current study, some ankles demonstrate eccentric arthritis in the sagittal plane without notable deformity in the coronal plane. As demonstrated by several cadaveric studies, sagittal plane malalignment places more kinematic impact on the ankle joint than that of the coronal plane, which can be compensated for by motion of the subtalar joint.^15,18,23 Therefore, we believe the restoration of the tibiotalar relationship at its earlier stage is equally important and crucial for joint preservation.

In the present study, intraoperative observations indicated that articular cartilage was preserved in the anterior aspect of the joint, suggesting that posterior ankle arthritis is a type of eccentric arthritis in which joint preservation can be attempted. However, it often appears as end-stage ankle arthritis on AP ankle radiographs because the narrowed posterior aspect of the tibiotalar joint is visualized. Therefore, lateral radiographs should be carefully evaluated to explore the possibility of joint preservation surgery in these ankles for asymmetric joint space narrowing in the sagittal plane.

Radiographic findings in the current study revealed that posterior ankle arthritis having posterior translation and plantarflexion of the talus, as evidenced by the LTCM (median −3.3 mm) and Meary angle (median 32.8 degrees). As well documented in the literature, the ankle joint is relatively stable in the sagittal plane from its bony conformity and posterior projection of the distal tibia. Besides, the talus is wider anteriorly and narrower posteriorly within the mortise. As such, the posterior translation of the talus seems unlikely to occur. The authors assume that plantarflexion of the talus places a narrower posterior portion of the talar dome within the mortise, which results in susceptibility of the talus to posterior translation. Consequently, a concentrated weightbearing load on the posterior aspect of the tibiotalar joint may lead to eccentric narrowing of the joint space and further posterior translation of the talus.

Based on the principles of joint preservation surgery, we postulated that dorsiflexion and anterior translation of the talus in the setting of flatfoot reconstruction was an appropriate method to preserve ankles with posterior arthritis. Flatfoot correction with subtalar arthrodesis and medial column stabilization using FHL transfer to the first metatarsal base was performed. Although corrective osteotomies such as lateral column lengthening can yield radiographic improvements in flatfoot deformity, its applications were felt to not be appropriate in correcting the severe degree of deformities or arthritic changes seen in the current study cohort (Figure 8).

OPEN IN VIEWER

Figure 8.

Sagittal images of weightbearing computed tomography (WBCT) in all patients (case number 1-10). WBCT images revealed a higher degree of deformity than simple radiographs, such as bony impingement in the sinus tarsi (circle in each image) or anterior translation of the talus relative to the calcaneus (case numbers 2, 3, 4, 5, 6, and 8).

The LTCM was used to assess the talar position relative to the tibial plafond in the sagittal plane. In previous studies, radiographic parameters such as the tibiotalar ratio and lateral talar station have been used to assess the talar position in the sagittal plane.^24-26 However, these parameters are limited in their ability to measure small changes in the position of the talus relative to the tibial plafond and can be influenced by the change of the distal tibia following SMO. In contrast, the LTCM yielded high interobserver reliability, and its significant increase after surgery reflected the procedure’s effectiveness in restoring the tibiotalar relationship.

Previous studies in the management of anteriorly translated talus have demonstrated that correction of anteriorly opened distal tibial angle results in the posterior realignment of the talus. ²⁰ However, as seen in this study, deformity of the distal tibial angle and its correction may not be related to the development and management of posterior ankle arthritis. The median pre- and postoperative anterior distal tibial angle in this study was 83 and 82.5 degrees, respectively, indicating that the slope of the tibial plafond in the sagittal plane does not seem to be a causative factor. These findings suggest that posterior opening or anterior closing distal tibial osteotomies may not be applicable to treating posterior ankle arthritis.

In the current study cohort, the coronal plane realignment with SMO was combined in 6 of 10 patients who exhibited varus tibial plafond and medial translation of the talus. Although its relative role in the restoration of the talar position in the sagittal plane is unclear from this study, the authors believe that untreated proximal varus malalignment can exacerbate valgus tilting forces in the ankle joint, especially in the setting of hindfoot arthrodesis in flatfoot reconstruction.

Although the radiographic stage improved in 7 ankles, including all preoperative stage IV ankles (n = 3), 3 ankles with preoperative stage II or III remained in the same stage postoperatively. However, WBCT analysis revealed restored joint space following realignment surgery. In combination with improved clinical results for talar position relative to the tibial plafond, these findings suggest promising results for joint preservation surgery for posterior ankle arthritis.

This study has several limitations. First, this was a retrospective study with a relatively small number of patients. Despite this limitation, results from the current study demonstrate the possibility of joint preservation in an ankle with posterior arthritis, which shows diffuse joint space narrowing on AP radiographs. Second, although ankle arthritis involves 3-dimensional deformities, pre- and postoperative axial rotation of the talus was not investigated. We acknowledge the inherent limitation of using 2-dimensional plain films as this can have some rotational bias. Third, inhomogeneity of the treatment and different morphologies in the cohort makes it difficult to know the cause and result relationship or the outcome of specific operative techniques. However, all patients in this study had a plantarflexed and posteriorly translated talus, and the study results imply that repositioning the talus may restore the joint. Lastly, follow-up was a minimum of 2 years, and the current cohort should be followed longer-term because subtalar arthrodesis may alter the kinematics of the tibiotalar joint. ⁵ This would provide valuable information in determining the effectiveness of ankle joint preservation surgery with hindfoot realigning arthrodesis for symptomatic posterior ankle arthritis.

In conclusion, the results of the current study demonstrated a significant correlation between posterior ankle arthritis and plantarflexion of the talus, which can be seen in the setting of a flatfoot deformity. At a minimum of 2 years and an average of 2.3 years postoperation, reconstruction of the flatfoot deformity using subtalar arthrodesis restored the tibiotalar relationship in the sagittal plane and resulted in substantial clinical improvement. Because AP ankle radiographs with posterior arthritis may often demonstrate diffuse narrowing of the joint space, lateral radiographs or WBCT scans should be carefully examined to explore the possibility of joint preservation surgery.

Supplemental Material