The primary purpose of this study was to introduce a new method for measuring patellar height and assess the validity by comparing with the four commonly used methods. The results indicated that a high inter- and intra-observer reliability was found for AP and jAP indices. Pearson’s correlation analysis showed that the AP and jAP indices were strongly correlated with the four classical counterparts. The new method proposed has been demonstrated of capable to evaluate patellar height both preoperatively and postoperatively for patients undergoing TKA due to primary osteoarthritis.
All four of classical methods mentioned in this article have intrinstic methodological difficulties, especially in patients who performed TKA resulted from severe osteoarthritis, which may lead to an increase in inter- and intra-observer variability [25-28], and evenly the measurement may be unavailable. Preoperatively, severe osteoarthritis with destruction of the articular surface and osteophytes formation, especially combined with lower extremity malalignment, usually leads to measurement errors to some extent, which are the common shortcomings of these four methods of using a marginal reference . For IS and mIS indices, it is imperative to standardize the tibial attachment of the patellar tendon in order to avoid a significant variation due to the obscure morphology and the pathological overgrowth of the tibial tubercle [26, 29]. Besides, the variants of patellar shape can bring about a greatly different result for IS index, especially in a patella being characterized by a long, non-articulating inferior pole (Cyrano patella) [12, 28, 29]. For BP index, one problem is the determination of the tibial plateau line after TKA. A joint line is constructed on the tibial articular surface, which is difficult to determine each time postoperatively due to the variation of polyethylene insert. As described by Rogers et al , the mean values of tibial insert thickness had to be used and compensate for the shape of the polyethylene insert, whereas the use of the mean value is often not an accurate substitute for individual prosthetic height in clinical measurements, in which requires knowing the thickness of the insert. On the other hand, the tibial component is usually placed with a posterior slope, but the vertical line (Fig. 1D) emanating from the tibial plateau line is significantly shortened evenly unavailable if the vertical line cuts the patella when a tibial component positioning with an excessive posterior inclination, although PTS showed a poor correlation with BP in this study. For CD index, preoperatively, it is located around a point outside the joint line, both in anterior and superior direction, the tibial reference point is difficult to determine particularly in severe arthritic knee due to osteophyte attachments. Postoperatively, however, the tibial component become a part of the articular surface, which is more precisely formed by the insert. This shows a shift of the reference point in the anteroposterior and superoinferior direction owing to thickness of bone resection and implant design. Thus, the comparability of the pre- and postoperative reference point is questionable [17, 25]. In view of this, it is necessary to find an accurate measurement method, in which the reference point is relatively constant and comparable before and after surgery.
The tibial landmark of the AP index, formed by the tibial shaft axis and its perpendicular line passing through the tip of the fibular head, intrinsically owns better positional stability before and after TKA. Compared with the tibial plateau, the fibular head is independent of TKA, as well as osteophytes or the destruction of tibiofemoral articular surface. And the recognition of the styloid process of fibular head is accurately acquired on the standard lateral radiograph [20, 30]. In this study, the tibial shaft axis was determined by selecting two mid-diaphyseal points at 70- and 110-mm distal to the tibial plateau down the tibial shaft, in which could avoid an increase of the measurement variability, as described by previous studies [22, 23]. As an intramedullary reference, the tibial shaft axis can theoretically provide a constant landmark compared to marginal counterparts, especially during tibial rotating subsequent to knee flexion. In present work, no significant difference was found between pre-operation and post-operative two weeks for AP and IS indices, consistently with the short-term results of IS index in previous reports [3, 31]. Hence, we speculate that this result is reasonable, because the T1 point on the tibial shaft axis, which located at the level of the tip of the fibular head, maintains a relatively constant distance from the tibiofemoral surface affected by the surgical operation. It may attribute to the constant characteristics of the T1 point, in the both anteroposterior and superoinferior direction, which is highly similar to the attachment point of the patellar tendon on the tibia in the IS index.
Previous studies have proved that the patella baja is a common complication of TKA [2, 9], which can be classified as a true patella baja caused by the shortening of the patellar tendon and the pseudo patella baja caused by the elevating of the joint line [12, 32]. The incidence of the pseudo patella baja following TKA had been published in 34% - 65% . As described by Grelsamer , CD and BP indices using the tibial reference at joint line can be applied to determine the true patella baja and pseudo patella baja, whereas the IS and mIS indices can be used to identify patellar tendon shortening. Generally, surgeons must combined both types of methods to distinguish the true and pseudo patella baja. In this study, the jAP index was found to have a strong correlation with CD and BP indices. In addition, when comparing the pre- and post-operative values of the jAP, CD and BP indices, all three indices showed a synergistic decreasing trend of patella. The small difference in the pre- and postoperative values can certainly be explained in large part by the elevation of the joint line postoperatively. In this study, the jAP index also presented an excellent interobserver reproducibility and intraobserver repeatability. The potential explanation is that the determination of the T2 reference point is independent of the implants design. Meanwhile, the tibial shaft axis usually passes through the anterior-middle segment of the tibiofemoral joint line on the lateral radiographs, the measurement “a” of jAP index is less likely to be affected by an excessive posterior inclination than BP and CD indices. In fact, we believe that the theoretical advantage of jAP index is that the T2 point will shift correspondingly along the tibial shaft axis with the change of the joint line position postoperatively, with less error in the anteroposterior direction. Therefore, once true patella baja is diagnosed on a lateral knee radiograph of postoperative TKA, both the AP and jAP indices will be abnormally low. When pseudo patella baja occurs, the AP index remains normal, but the jAP index will decrease, correspondingly JLH will increase and can accurately reflect the change of joint line height when compared with the preoperative JLH value (Fig. 3).
Of the widely accepted surgical goals of TKA, restoration of the joint line is crucial. Accurately restoring the natural joint line is a challenge and one of the most important factors in achieving satisfied clinical outcomes [19, 20, 33]. Nevertheless, accurate restoration of the joint line depends on the anatomical landmarks of the knee [19, 20]. Typically, in complicated or revision TKA, these landmarks are commonly obscured even missing, resulting in a difficult in restoration [19, 21]. Pereira et al  described a series of procedures to determine the joint line position using the following bony landmarks: tibial tubercle, proximal tibio-fibular joint, femoral epicondyles, and femoral metaphyseal flares. Laskin  used the fibular styloid and medial epicondyle as reference points to measure the joint line during TKA. However, there remains a lack of consensus on an optimal measurement method. Moreover, several clinical retrospective studies and biomechanics evaluations had proved that postoperative joint line elevation of more than 5 to 8 mm is closely related to poor clinical outcomes, such as anterior knee pain, reduction of movement range, knee instability, and accelerated polyethylene wear [18, 19, 33]. Therefore, there should be a robust method to accurately determine the level of bone resection, thus to reproduce the position of the natural joint line. As a vital element of the new method, the measurement of JLH provides feasibility for quantitative analysis of the alteration of the joint line on the lateral radiograph before and after TKA. In this study, the JLH averaged 10.89 mm before TKA, and 14.06 mm after TKA, with an average elevation of 3.17 mm (paired Student’s test P <0.001) (Table 1).The JLH was measured directly on the tibial shaft axis, which is strongly associated with biomechanics and component positioning of prosthetic knee . Furthermore, because the tibial shaft axis is determined by connecting two mid-diaphyseal points at 70- and 110-mm distal to the tibial plateau down the tibial shaft, it is less prone to intra- and interobserver variation (ICCs: 0.80-0.95). In the lateral radiographs, compared with the reference points located outside the axis, we speculate that the reference points located on the tibial axis are more direct for guiding the restoration of the joint line and further evaluation of the knee instability, especially in flexion instability.
There are several limitations of this study. Firstly, the retrospective nature is clear limitation of the study, and the research data were extracted from the hospital information system. Secondly, a normal reference range was not provided for the new indices considering the included cases suffering from severe osteoarthritis. Thirdly, although the osteophytes were excluded out during measurement preoperatively and removed intraoperatively, they could remain a source of error. Finally, the patellar articular surface in the postoperative radiographs was occasionally not completely assessable, since it was partially embedded in the femoral component and therefore had to be estimated. Despite limitations, there are many strengths like use of the same prosthesis, a single senior joint surgeon being the main operating surgeon, in an effort to minimize variation.