Anterolateral Approach Might Be Superior to Posterolateral Approach in Controlling Lower Limb Discrepancy for Special Total Hip Arthroplasty Patients

Objectives: Variable approaches for total hip arthroplasty (THA) are promoted but investigations of associations between the approaches and the postoperative lower limb discrepancy (LLD) are limited. We intended to compare the ecacy between the anterolateral approach (ALA) and posterolateral approach (PLA) in controlling of postoperative LLD. Methods: Two retrospective cohorts were established including patients undergone unilateral primary THA. Preoperative and latest follow-up radiographs were compared and evaluated. Two hundred and sixty patients (140 females and 126 males; aged 46.70 ± 13.62 years) were recruited with around 20 months follow-up postoperatively. Results: Univariate analysis identied lower postoperative LLD, lower postoperative acetabular offset and lower hospital costs in the ALA group than in the PLA group (p(cid:0)0.01), however, the femoral and global offset and length of stay (LOS) are comparable between (p > 0.05). Multivariable liner regression revealed that PLA, preoperative LLD, developmental dysplasia of the hip and aseptic avascular necrosis are major contributors to the postoperative LLD (p<0.05). Conclusions: This study illustrates that the ALA might be superior to the PLA in controlling the postoperative LLD among some of the THA patients. Both ALA and PLA are comparable in regard to femoral and offset and LOS.


Introduction
Total hip arthroplasty (THA) is a valid method for various hip diseases such hip osteoarthritis (HOA), aseptic avascular necrosis (AVN) of the femoral head, developmental dysplasia of the hip (DDH), and femoral neck fracture (FNF) [1,2]. The clinical e cacy of THA has been signi cantly improved with the developments of relevant technologies with the 15-year survival rate of the prosthesis over 95% [3]. Large quantity of THAs are necessary in the setting of the aging population [4]. Successful THA can provide adequate pain relief, restore hip function, improve quality of life, and obtain satisfactory long-term outcomes [5].
Postoperative lower limb discrepancy (LLD) and improper offset occur occasionally, which might result in the adverse events, instability, impingement, or even dislocation of the prosthesis [6,7]. The LLD also remains the most common cause of patient dissatisfaction and postoperative gait disorders [8]. The improper offset also unbalances the arm of the abduction muscle to weaken the strength, increases the pressure of the hip joint, and accelerates the wear of prosthesis [9].
Anterolateral approach (ALA) and posterolateral approach (PLA) are the two commonly applied THA approaches [10,11]. The ALA is performed thought the gluteus medius to increase the hip load and abduction with reproduceable results and to decrease the dislocation rates [12,13]. The PLA is an approach with an excellent exposure view by damaging joint capsule with shorter duration, lower intraoperative fracture but higher risk of dislocation, compared with direct anterior approaches [14,15]. To the best of our knowledge, the e cacy of either ALA or PLA to control the postoperative LLD and to restore the offset has not been su ciently investigated. Therefore, the present study aimed to investigate the possible association of surgical approaches (ALA or PLA) with postoperative LLD and offset in primary unilateral THAs.

Participations
The present study was carried out in Peking University Shenzhen Hospital located in South China between October 2018 and July 2019. The study protocol was approved by the Ethics Committee of Peking University Shenzhen Hospital (No. 2019038) and all patients gave their inform and consent.
Patients undergone the unilateral primary THA via the ALA or the PLA were retrospectively included. Since radiographic evaluation is probably unreliable for FNF patients, only patients who were diagnosed as HOA, DDH, in ammatory arthritis, and AVN were recruited. Patients who failed to understand or signed consent, lost follow-up or missed information, or obtained unquali ed hip radiographs were excluded.
The exclusion criteria were illustrated in the owchart ( Figure. 1). All participations were invited for postoperative clinical and radiological follow-ups. All operations were conducted by the experienced and skilled surgeons specialized in arthroplasty with over 10-year experience of both the ALA and PLA approaches (over 200 arthroplasty operations annually).

Variables
The discharge notes of all patients were reviewed and relevant variables were recorded including their basic demographic information and radiographic parameters. Basic demographic information included the gender, age, height, weight, body mass index (BMI), marriage, hospital cost, length of stay (LOS), diagnosis, and American Society of Anesthesiologists (ASA) score [4]. Radiographic outcomes included the postoperative LLD, femoral offset (FO), acetabular offset (AO), and global offset (GO) [11]. LOS was de ned as the periods between the admitted date to discharge date [16]. The extraction differences, if existing, were resolved by discussing with a third party and the data were independently extracted by two investigators (GQL and SL).

Radiological assessment
The preoperative and postoperative plain radiographs were performed in the anteroposterior and lateral positions for both hip joints routinely for THA patients and archived in the Picture Archiving and Communication System (PACS) (Tomorrow Medical Network Science and Technology, Ningbo, Zhejiang, China). These historical radiographs retrieved from PACS were evaluated by the senior surgeons (AX and DLW) to con rm the quali cation elsewise a new one would be lmed. The LLD was measured between both the teardrop lines and the lesser trochanter base and the absolute value represented the distance of the THA side compared to the untreated side as described elsewhere [17]. The FO was de ned as the distance between the center of rotation (COR) and the femoral axis. The AO was de ned as the distance between the COR and the tear drop axis ( Figure. 2). The GO was measured as the distance between the femur longitudinal axis to the COR plus the distance from the COR to a vertical line of the medial edge of the ipsilateral teardrop point of the pelvis [2]. To reduce the bias of measurements, all radiographic variables were measured and recorded as continuous variables by two independent observers (GQL and SL) with a standardized process. Inter-observer and intra-observer reliabilities of the afore-mentioned radiological measurements were evaluated.

Statistical analysis
All statistical analysis was performed by the R software 4.0.2 (R Foundation for Statistical Computing, Vienna, Austria) [18]. Continuous variables were summarized by mean ± standard deviation (SD) while categorical variables were summarized by number and percentage. For the continuous variables and data with a normal distribution, Student-t test was applied. For the categorical variables, Chi-square test and Fisher's exact test were applied properly according to the expected values greater than ve or not. Univariate analysis was used to analyze the relationship of clinical outcomes and variables. Multivariate linear regression analysis was performed to examine the signi cant factors on postoperative LLD. All statistics were 2-tailed. With p < 0.05 as the statistical signi cance [19] and reported with 95% con dence interval (CI).

Results
According to the inclusion and exclusion criteria, 266 patients (140 females and 126 males) were included for the in-depth analyses and categorized into the ALA group (N = 66) and the PLA group (N = 200). All enrolled participations were aged 46.70 ± 13.60 (range, 22.00 to 90.00) years old with mean BMI 22.73 ± 3.28 (range, 14.53 to 35.16) kg/m 2 and followed up for an average of 20 months. Three cases with dislocation in PLA group were cured with the conservative treatments, while no dislocation case occurred in the ALA group. No other postoperative complications (e.g. periprosthetic fracture and infection) and readmission within the follow-up period were reported within either group. All patients reported satisfactory hip function and normal muscle strength at the last follow-up.
Regression curves of postoperative discrepancy and preoperative LLD between the two groups displayed in Figure. 3a. Obviously lower postoperative discrepancy in the ALA group than that in the PLA group (p < 0.01). The regression curves of postoperative discrepancy and diagnosis and BMI were displayed in Figure. 3b and Figure. 3c.
Regression curves of preoperative offset and postoperative offset (AO (a), FO (b), and GO (c)) ( Figure. 4), which implied that they were comparable between the two groups.

Discussion
The present study showed that the ALA might be superior to the PLA in controlling the short-term postoperative LLD among some of the unilateral primary THA patients. However, their postoperative FO, GO restoration, and LOS are comparable. Both approaches achieved satisfactory joint function within the 20-month follow-up.
Our most import nding is that the ALA might be superior to the PLA in controlling postoperative LLD. The univariate analyses revealed the postoperative LLD were signi cantly lower and more stable in the ALA group than in the PLA group over 20 months postoperatively. Similarly, multivariate liner regression analysis also revealed that THA approach is signi cantly associated with postoperative LLD. Notable relationships between discrepancy and approaches were re ected with a higher discrepancy in the PLA group (p = 0.001; β = 4.71; 95% CI, 1.78 to 7.64). These results are highly consistent with several earlier studies [20]. A previous study demonstrated that the accurate reconstruction of biomechanical parameters related to the limb discrepancy after THAs is critical to ensure a long-term success [21]. Nevertheless, discrepancies after THAs may necessitate subsequent interventions but current methods during the operation are either inaccurate or expensive [22]. Intelligent HIP smart tool, a newly introduced smart tool providing an accurate and real-time intraoperative leg length measurements, is properly useful for an improved measurements [22]. Special-designed prostheses [23], surgeon-controlled table, and advanced uoroscopic imaging techniques [24]. would improve the biomechanical and physiological hip reconstruction and subsequently enhance the radiographic and clinical outcomes.
Interestingly, both the ALA and PLA are comparable in restoring the postoperative offset. Previous literatures evidenced that the offset parameters (FO, AO, and GO) have signi cant impacts on the postoperative function rehabilitation and preoperative templating, and the proper intraoperative veri cation is helpful to achieve an accurate offset restoration, yielding a better abductor strength as well as a long-term successful prognosis [25,26]. Our data showed that postoperative AO, FO and GO was lower in the ALA group than that in the PLA group. These results indicate that both the ALA and the PLA are possibly proper for the unilateral primary THA in selected conditions regarding the offset restoration in line with the general investigations [26].
The DDH represents abnormalities of both acetabulum and femur, requiring special techniques to reconstruct its physiological biomechanical [27,28]. Our data with 81 DDH patients (including 49 young females) showed that the DDH patients tend to yield the more postoperative discrepancy and the DDH diagnosis is a signi cant contributor to the postoperative LLD. Previous literatures demonstrated that the discrepancy and offset might result from an improper position of the femoral stem and acetabular cup, and special thinner, therefore, shorter stems with more non-sprouted sleeves might be more suitable [29,30]. Arthroplasty surgeons are recommend paying more attention to the perioperative plan in consideration of high risk of postoperative LLD. External obturator footprint, robotic-assisted intervention, gait training, and 3D-printed individual templates provide personal plan to improve the THA outcomes for DDH patients [31].
Moreover, the ALA seems cause the less soft tissue damage than the PLA group for the primary THAs with the comparable safety and e cacy. Previous research showed that the joint degeneration is associated with the body weight of patients [32,33]. However, our data revealed that BMI is not a signi cant contributor to LLD (p = 0.126; β = 0.30, 95% CI, -0.08 mm to 0.67 mm) but an increased postoperative discrepancy is noticed in patients with a higher BMI ( Figure. 3). Therefore, controlling weight might still be helpful for obese patients before the elective surgeries. More studies are necessary to identify such a possible bene cial effect of the body weight controlling strategy for the ultimate success of THAs.
Several limitations should be noticed in this study. First, the retrospective nature of the study might not allow us to draw conclusive statements. Second, signi cant intergroup difference of patient age was observed with divergent sample size, which seems as an unbalanced group (PLA = 200 vs ALA = 66) of patients with what appears to be predominantly DDH (81/266, ALA = 22, PLA = 59). Third, the follow-up is relatively short, more trails with high-level evidence were needed to con rm the ndings from the represent study. Strengths of this study include the detailed comparison and analysis of radiographic data performed in a standard manner with a relatively large and homogenous patients operated by the identical highly experienced surgeon team from a single institution.

Conclusions
In conclusion, our study illustrates that the ALA might be superior to PLA in controlling short-term postoperative LLD among some of the unilateral primary THA. In regard to the FO and GO restoration and medical quality including LOS, both approaches are comparable. More investigations are required to provide convincing favorable evidence to further compare the ALA and PLA in primary THA. Availability of data and materials Data are available from the authors upon reasonable request and with permission of Department of bone and joint in Peking University Shenzhen Hospital.

Competing interests
All of the authors have disclosed potential con icts of interest, which may include receipt of payment, either direct or indirect, institutional support, or association with an entity in the biomedical eld which may be perceived to have potential con ict of interest with this work. No conflicts reported in regards to this manuscript.

Author contributions
Ao Xiong and Guoqing Li participated in the design and prepared the manuscript, they contribute equally to this work and they were listed as cooperated rst authors. Guoqing Li and Su Liu participated in the data collection and radio assessment; Chang Xu participated in execution and analysis of the work presented. Deli Wang and Hui Zeng participated in the design and revision, and reviewed the final manuscript, they contribute equally to this work and they were listed as corresponding authors.