Clinical and Radiological Changes of Ankle in Knee Osteoarthritis With Varus After Total Knee Arthroplasty: A Systematic Review

Background: Arthritis with severe varus deformity remains a challenge in total knee arthroplasty (TKA). Until recently, surgeons aimed at a neutral lower limb alignment when performing a TKA. However, the impact of TKA on the ankle joint has been ignored. Therefore, we conducted a systematic review to assess the clinical and radiological changes in the ankle joint after TKA on knees with severe varus deformity. Methods: A systematic search was conducted in four English (PubMed, Embase, Cochrane Library, and Web of Science) and four Chinese (CBM, VIP, CNKI, and Wan Fang Database) databases. Screening of literature and extraction of data were independently performed by two researchers. The modified methodological index for non-randomized studies (MINORS) was used to assess the quality. Results: A total of eight studies were eligible, namely, four prospective and four retrospective studies. TKA resulted in a negative clinical effect in the ankle joint in patients with ankle osteoarthritis. Seven studies reported changes in the mechanical tibiofemoral angle, and four studies reported radiological changes in the hindfoot. The mean score of the MINORS was 9.8 out of eight (9–11). Conclusion: As a result of the correction of the knee osteoarthritis with severe varus deformity following mechanically aligned TKA, the radiological malalignment of the ankle joint was improved. However, some patients experience increased ankle pain after undergoing TKA, especially, if there was a residual knee varus deformity, a stiff hindfoot with varus deformity, or ankle arthritis.


INTRODUCTION
Concomitant ankle osteoarthritis (OA) is frequently found in patients who undergo TKA, which is a classic surgery for elderly patients, who have experienced knee osteoarthritis (KOA), with multiple joint degenerative changes (1)(2)(3). Furthermore, malalignment of the knee joint caused by KOA could induce ankle tilt that would further aggravate ankle OA (4). Nonetheless, a few patients experienced increased or newly developed ankle pain after TKA (5). Surgeons lack a comprehensive consideration of the ankle condition of the patient when performing TKA.
Most of the patients with knee arthritis who undergo TKA have varying degrees of knee joint deformity, which may be associated with hindfoot deformity (6)(7)(8). Previous studies reported an association between varus alignment of the knee joint and valgus alignment of the hindfoot in patients with KOA (7,8). Because the mild deformity of the knee joint can be compensated by the subtalar joint (8), the knee joint alignment modification after TKA can affect the hindfoot alignment. Previous studies have demonstrated that improvements in the hindfoot alignment have been observed in patients who experience OA after TKA (7,9,10). However, hindfoot valgus may show little improvement and persist after TKA (7,9).
Patients with a hindfoot deformity present a particular challenge when undergoing TKA. Clarifying the relationship between hindfoot alignment and TKA is useful to change the current state. Thus, this systematic review attempted to assess prospective and retrospective cohort studies to determine the relationship between the radiological changes and clinical symptoms of the ankle joint following acute correction of the lower alignment using TKA for severe varus knee.

Literature Search
A systematic search was conducted in four English (PubMed, Embase, Cochrane Library, and Web of Science) and four Chinese (CBM, VIP, CNKI, and Wan Fang Database) databases from inception to December 2020 for relevant non-randomized controlled trials (nRCTs). The search was based on the following search terms: ("Arthroplasty, Replacement, Knee" OR "Total knee arthroplasty" OR "Knee replacement" OR "TKA") AND ("varus knee") AND ("ankle alignment" OR "ankle deformity" OR "hindfoot alignment" OR "hindfoot deformity"). Additional records were obtained by screening the references. Detailed inclusion and exclusion criteria were formulated to review the results of the search by two reviewers (Feng ZH W and Ma M); the abstract was screened first before reading the full text.

Literature Inclusion and Exclusion
The inclusion criteria were as follows: (1) nRCT, prospective or retrospective studies, and case control studies; (2) patients who underwent primary TKA as research participants; (3) with mechanical tibiofemoral angle (MTA), hindfoot alignment angle (HA), American Orthopedic Foot and Ankle Score (AOFAS), talar tilt (Tt), medial ankle joint space (MAJS), and medial ankle clear space (MACS) as indicators of the postoperative results; and (4) all retained studies should have the same hip knee ankle (HKA) goal, i.e., neutral mechanical alignment (MA). If not, the study should be excluded. The exclusion criteria were as follows: (1) diagnosis of a disease other than primary KOA; (2) KOA with valgus malalignment or rheumatoid arthritis; (3) history of femoral or tibial fractures in the past; (4) incomplete or missing research data; and (5) research data from animal experiments or theoretical analysis. Repeated publications, letters, case reports, comments, conference abstracts, or books were also excluded.

Data Extraction
We performed data extraction of all the included documents with predesigned tables, and all the studies were independently retrieved and assessed by two reviewers (ZH W Feng and M Ma). The extracted data should include the following: (1) title, first author, publication time, study design, clinical indicators, and radiological indicators; (2) patient characteristics such as number of patients, age, sex, number of patients who underwent TKA, and follow-up time; and (3) study details such as clinical results, radiological results, and conclusions. The extracted information was cross-checked to ensure accuracy. When there were disagreements, the decision was made by a third person.

Quality Assessment
The methodological quality of nRCTs was assessed using the modified MINORS (11). Eight questions for the nRCTs were used to score the relevant aspects of each study: a clear research purpose (I), continuous follow-up of patients (II), prospective research (III), suitable research purpose (IV), unbiased assessment of the research outcome (V), an appropriate follow-up period for the study (VI), the lost-to-follow-up rate was < 5% (VII), and perform a prospective calculation of the sample size (VIII).
The score criteria are as follows: not reported scored as 0, reported but inadequate scored as 1, and reported and adequate scored as 2. The MINORS score was assessed separately by the two reviewers (ZH W Feng and M Ma). Then, the categories were determined according to the study by Ekhtiari et al. (12): "0-4 points" were categorized as very low; "5-8 points" as low; "9-12 points" as good; and "13-16 points" as excellent.

Study Selection
A total of 733 studies were included in this study. After removing 277 duplicates, the titles, abstracts, and full articles were screened, and 425 studies were further excluded. A total of eight eligible articles were considered for further analysis. The results of the study selection process are presented in Figure 1.

Study Characteristics
A total of 913 patients and 1,157 knees were included in the eligible studies. All included studies were observational studies, and their publication years ranged from 2012 to 2019. The followup duration was extended from 6 months to 3 years. Tables 1, 2 show the characteristics of the included study.

Quality Assessment
We used the MINORS criteria to assess the methodological quality of the nRCTs. The categories were determined according to the research of Ekhtiari et al. (12). The quality evaluation of the included research literature is shown in Table 3.

Research Details
After removing duplicate studies and applying the inclusion and exclusion criteria, eight studies were eligible for further analysis (Figure 1). The eligible studies included a total of 913 patients     and 1,157 knees, which included four prospective cohort studies (13,16,17,19) and four retrospective studies (5,14,15,18). Seven studies (5,(13)(14)(15)(16)(17)19) only studied the varus KOA, and one study (18) investigated both varus and valgus KOA. The characteristics of the eligible studies were extracted by two researchers ( Table 1). All the clinical and radiological outcomes of the knee and hindfoot after TKA are shown in Table 2. Four studies (5,14,17,19) have reported the specific clinical outcomes of the knee and hindfoot after TKA. Okamoto et al. (5) reported that patients with hindfoot deformity after TKA would have a significant improvement in Knee Society Score, Knee Society Functional Score, and AOFAS. Chang et al. (14) reported that patients with ankle OA experienced increased ankle pain and poor clinical prognosis after TKA. Meanwhile, Palanisami et al. (17) reported that the Oxford Knee Score and AOFAS of patients with foot deformities after TKA significantly improved. Kim et al. (19) also reported that a persistent ankle varus deformity could be attributed to increased ankle pain after TKA.
Seven studies (5,(13)(14)(15)(16)(17)19) compared the change in MTA pre-and postoperative, which found significant postoperative improvement of MTA after TKA. Five studies (13,14,16,17,19) included a radiological analysis of HA before and after TKA, and found significant improvement in hindfoot deformity. Cho et al. (13) also found that there was a weak negative correlation between the preoperative HA and MTA (−0.484, p < 0.001), and a very weak correlation between the postoperative MTA and postoperative HA at 6 weeks (−0.147, p = 0.040). Six studies (5, 14-16, 18, 19) included a radiological analysis of pre-and postoperative Tt, which found significant improvement. Lee et al. (18) found that the incidence of ankle arthritis would obviously increase when the preoperative Tt was closer to the ankle medial or when the angle of correction was greater after TKA. Lastly, three studies (14,18,19) included a radiological analysis of MAJS and MACS pre-and postoperatively, which showed a significant difference.

DISCUSSION
Traditionally, obtaining a neutral lower limb alignment after a TKA was perceived as the ideal goal (20)(21)(22)(23). However, since very few individuals have such anatomy, it implies a significant modification for most. Freeman et al. (24) first introduced the concept of right-angled femoral and tibial bone cuts in TKA-MA. Subsequently, Scuderi et al. (25), raised the importance of balancing the resulting medial-lateral and flexion-extension joint gaps. MA technique gradually became the gold standard in TKA. However, the recent studies found that MA-TKA generates disappointing efficacies (26,27), probably due to MA-TKA produces a non-physiological prosthetic knee by changing the native anatomy, physiological ligament balance, and kinematics (28)(29)(30)(31). Stephen Howell developed the kinematic alignment (KA) technique (28,30). KA-TKA aims to generate a more physiological prosthetic knee joint by restoring the inherent knee joint anatomy and physiological soft tissue balance of the individual. Several studies have suggested that KA-TKA can accurately locate the position of the prosthesis (32), and also restore the native anatomy of the knee (32,33) and physiological laxity (34,35). The ultimate goal of TKA is to restore the anatomy and kinematics of native knees and provide a forgotten joint.
Recent studies have found that the alignment of the ankle joint and hindfoot could be affected by the malalignment of the lower limb (4,5,8,36,37). Moreover, the hindfoot deformity is closely associated with the lower limb alignment (38), and the clinical symptoms and radiographic outcome changes of the ankle joint and hindfoot in patients who underwent TKA would be influenced by changes in lower limb alignment (5). Chang et al. (14) found that there was less radiographic alignment change in the ankle and hindfoot when patients had ankle OA and increased ankle pain after TKA; and patients with a stiff hindfoot who underwent TKA experienced increased ankle pain, probably because of the inability to reorient the ankle after TKA. Nevertheless, different findings have been reported by these studies, but there have been no comprehensive reviews of this issue in the literature so far.
In this systematic review, we found that TKA may affect the clinical and radiological outcomes of patients with hindfoot deformity before TKA. In addition, the HA before TKA had a weak negative correlation with MTA (13). Clinically, four studies (5,14,17,19) detected an obvious improvement in AOFAS at the hindfoot after TKA. However, increased ankle pain after TKA has also been reported in the patients with ankle OA or persistent knee deformity (14,19). When patients have a residual deformity and pain of the hindfoot 6 weeks after TKA, they must receive active treatment because further improvement cannot be expected (13). Furthermore, five studies (13,14,16,17,19) reported an obvious radiological improvement in hindfoot alignment after TKA. Kim et al. (19) also reported a relationship between residual varus deformity and ankle pain after TKA. Figure 2 shows the flowchart for the general treatment of patients with knee varus and ankle stiffness before and after TKA. Before TKA, the clinicians should perform clinical and radiological examinations of the ankle/hindfoot. If the patient has clinical or imaging problems with the ankle/hindfoot, the clinicians should inform the patient about the results and provide some treatment before and after TKA. If the patient does not have ankle/hindfoot problems before TKA, the clinician should tell them about possible problems after TKA. Patients should be informed and treated when discomfort occurs.
This study has several limitations. First, eligible papers according to the search criteria used were not identified completely. Second, the eligible studies found for this systematic review had a relatively small number and a relatively high heterogeneous group, which could not be studied by meta-analysis. Third, the follow-up time in two studies (16,17) was <2 years, which limited the ability to draw a long-term conclusion.
In conclusion, an improvement in the clinical function and radiological alignment of the hindfoot can be achieved following TKA. However, when patients had concomitant ankle OA with hindfoot stiffness, there is an increased ankle pain and a worse clinical outcome after TKA.
Therefore, the comprehensive preoperative evaluation of surgeons of the ankles of patients who complain of pain preand postoperatively and the correction of alignment during TKA should be given careful attention.

DATA AVAILABILITY STATEMENT
The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

AUTHOR CONTRIBUTIONS
ZF and MM worked on literature search, study review, and manuscript draft. YW, CY, and ZL prepared the tables and figures. YX worked on manuscript review, process supervision, and draft revision. All authors have read and approved the final manuscript.