Eslam Pour Aidin, Lazennec Jean Yves, Patel Kunj P, Anjaria Manan P, Beaulé Paul E, Schwarzkopf Ran
Department of Orthopaedic Surgery, Yale University, New Haven, CT, USA.
Department of Orthopaedic and Trauma Surgery, Pitié-Salpétrière Hospital Assistance Publique-Hopitaux de Paris, UPMC, Paris, France.
Clin Orthop Relat Res. 2022 Apr 1;480(4):818-828. doi: 10.1097/CORR.0000000000002106.
Many THA simulation models rely on a limited set of preoperative static radiographs to replicate sagittal pelvic tilt during functional positions and to recommend an implant orientation that minimizes the risk of prosthetic impingement. However, possible random changes in pelvic or lower extremity angular motions and the effect of coronal and axial pelvic tilt are not included in these preoperative models.
QUESTIONS/PURPOSES: (1) Can prosthetic impingement occur if the pelvic tilt or lower extremity alignment randomly varies up to ± 5° from what is measured on a single preoperative static radiographic image? (2) Do changes in coronal and axial pelvic tilt or lower extremity alignment angles have a similar effect on the risk of prosthetic impingement?
A de-identified pelvis and lower-body CT image of a male patient without previous THA or lower extremity surgery was used to import the pelvis, femur, and tibia into a verified MATLAB computer model. The motions of standing, pivoting, sitting, sit-to-stand, squatting, and bending forward were simulated. THA implant components included a full hemispherical acetabular cup without an elevated rim, polyethylene liner without an elevated rim, femoral head (diameter: 28 mm, 32 mm, 36 mm, or 40 mm), and a triple-taper cementless stem with three different neck shaft angles (127°, 132°, or 135°) with a trapezoidal neck were used in this model. A static model (cup anatomical abduction 40°, cup anatomical anteversion 20°, stem anatomical anteversion 10°) with a predefined range of sagittal pelvic tilt and hip alignment (0° coronal or axial tilt, without random ± 5° change) was used to simulate each motion. We then randomly varied pelvic tilt in three different pelvic planes and hip alignments (flexion, extension, abduction, adduction, rotation) up to ± 5° and assessed the same motions without changing the implant's anatomical orientation. Prosthetic impingement as the endpoint was defined as mechanical abutment between the prosthetic neck and polyethylene liner. Multiple logistic regression was used to investigate the effect of variation in pelvic tilt and hip alignment (predictors) on prosthetic impingement (primary outcome).
The static-based model without the random variation did not result in any prosthetic impingement under any conditions. However, with up to ± 5° of random variation in the pelvic tilt and hip alignment angles, prosthetic impingement occurred in pivoting (18 possible combinations), sit-to-stand (106 possible combinations), and squatting (one possible combination) when a 28-mm or a 32-mm head was used. Variation in sagittal tilt (odds ratio 4.09 [95% CI 3.11 to 5.37]; p < 0.001), axial tilt (OR 3.87 [95% CI 2.96 to 5.07]; p < 0.001), and coronal tilt (OR 2.39 [95% CI 2.03 to 2.83]; p < 0.001) affected the risk of prosthetic impingement. Variation in hip flexion had a strong impact on the risk of prosthetic impingement (OR 4.11 [95% CI 3.38 to 4.99]; p < 0.001).
The combined effect of 2° to 3° of change in multiple pelvic tilt or hip alignment angles relative to what is measured on a single static radiographic image can result in prosthetic impingement. Relying on a few preoperative static radiographic images to minimize the risk of prosthetic impingement, without including femoral implant orientation, axial and coronal pelvic tilt, and random angular variation in pelvis and lower extremity alignment, may not be adequate and may fail to predict prosthetic impingement-free ROM.
Determining a safe zone for THA implant positioning with respect to impingement may require a dynamic computer simulation model to fully capture the range of possible impingement conditions. Future work should concentrate on devising simple and easily available methods for dynamic motion analysis instead of using a few static radiographs for preoperative planning.
许多全髋关节置换（THA）模拟模型依赖于一组有限的术前静态X线片，以在功能位复制骨盆矢状面倾斜，并推荐一种能将假体撞击风险降至最低的植入物方向。然而，这些术前模型未考虑骨盆或下肢角运动可能的随机变化以及骨盆冠状面和轴面倾斜的影响。
问题/目的：（1）如果骨盆倾斜或下肢对线相对于术前单张静态X线影像测量值随机变化高达±5°，是否会发生假体撞击？（2）骨盆冠状面和轴面倾斜或下肢对线角度的变化对假体撞击风险有类似影响吗？
使用一名未曾接受过THA或下肢手术的男性患者的匿名骨盆和下半身CT图像，将骨盆、股骨和胫骨导入经过验证的MATLAB计算机模型。模拟了站立、 pivoting、坐、从坐到站、蹲和向前弯腰等动作。THA植入部件包括无高边的全半球形髋臼杯、无高边的聚乙烯衬垫、股骨头（直径：28mm、32mm、36mm或40mm），以及该模型中使用的具有三种不同颈干角（127°、132°或135°）且颈部为梯形的三锥度非骨水泥柄。使用一个具有预定义矢状骨盆倾斜和髋关节对线范围（冠状面或轴面倾斜0°，无随机±5°变化）的静态模型（髋臼杯解剖外展40°，髋臼杯解剖前倾角20°，柄解剖前倾角10°）来模拟每个动作。然后，我们在三个不同的骨盆平面和髋关节对线（屈曲、伸展、外展、内收、旋转）中随机改变骨盆倾斜高达±5°，并在不改变植入物解剖方向的情况下评估相同动作。将假体撞击作为终点定义为假体颈部与聚乙烯衬垫之间的机械抵靠。使用多元逻辑回归研究骨盆倾斜和髋关节对线变化（预测因素）对假体撞击（主要结果）的影响。
无随机变化的基于静态的模型在任何情况下均未导致任何假体撞击。然而，当使用28mm或32mm股骨头时，骨盆倾斜和髋关节对线角度随机变化高达±5°时，在pivoting（18种可能组合）、从坐到站（106种可能组合）和蹲（1种可能组合）动作中发生了假体撞击。矢状面倾斜变化（优势比4.09 [95%可信区间3.11至5.37]；p < 0.001）、轴面倾斜（OR = 3.87 [95%可信区间2.96至5.07]；p < 0.001）和冠状面倾斜（OR = 2.39 [95%可信区间2.03至2.83]；p < 0.001）影响假体撞击风险。髋关节屈曲变化对假体撞击风险有强烈影响（OR = 4.11 [95%可信区间3.38至4.99]；p < 0.001）。
相对于术前单张静态X线影像测量值，多个骨盆倾斜或髋关节对线角度变化2°至3°的综合影响可能导致假体撞击。仅依靠几张术前静态X线影像来将假体撞击风险降至最低，而不考虑股骨植入物方向、骨盆轴面和冠状面倾斜以及骨盆和下肢对线的随机角度变化，可能并不充分，且可能无法预测无假体撞击的活动范围。
确定THA植入物相对于撞击的安全定位区域可能需要动态计算机模拟模型，以充分捕捉可能的撞击情况范围。未来的工作应集中于设计简单且易于获得的动态运动分析方法，而不是使用几张静态X线片进行术前规划。
