The main findings of the present study were twofold. Firstly, our data showed that increased SP-ET index was associated with the risk of developing patellar cartilage lesions. Secondly, the SP-ET index might be a better indicator to define the vertical position of the patella relative to femoral trochlear than the Insall–Salvati ratio in patellar cartilage lesions patients. These findings supported our hypothesis and indicated that the SP-ET index may be a useful complement parameter to define the vertical position of the patella relative to the femoral trochlear, and a higher position of patella might play a role in the pathogenesis of patellar cartilage lesions.

The previous researches have analyzed the influence of patellofemoral malalignment on patellar cartilage lesions or patellar cartilage defects [8, 11, 13, 15, 20, 26]. Overall, patellofemoral malalignment and structural abnormalities of the patellofemoral joint led to pathologic pressure on the patellofemoral joint might be an important pathogenesis of patellar cartilage lesions [20, 27]. Some studies found that a higher position of patella was an important influencing factor of patellar cartilage lesions [11, 15, 20], and in addition to the standard measurements of patellar height, some other measurements were analyzed, such as modified Insall–Salvati index [15, 20, 27], patellotrochlear index, and patellophyseal index [20, 27]. The assessment of the sagittal position of the patella is still worth discussing due to the diversity of examination techniques, including X-ray, CT, and MRI [5, 8]. Take the Insall–Salvati index, for example, which is an indicator of the height of the patellar and has traditionally been measured by conventional radiography [23]. But MRI is currently being used more clinically to assess the relationship between the patellofemoral joint. The advantages of MRI over conventional radiography lie in its capability of multiplanar, high-resolution imaging of chondral, and soft tissue lesions [1, 5, 8]. There have been some studies based on MRI measurements, but they are still controversial [20, 28]. Therefore, how to use MRI to accurately assess the sagittal position of the patella still needs to be discussed.

The Insall–Salvati ratio was defined as the length of the patellar tendon divided by the length of the patella, which was a validated and widely used index for evaluating the position of patella [23, 29]. The Insall–Salvati ratio can be measured on both lateral radiograph and sagittal MR image of the knee. A higher position of patella might result in failure of the patella to engage the trochlear groove early during knee flexion [17], and decrease the contact area between the patellar articular surface and trochlea [10, 16]. Previous studies have shown that patella alta was associated with patellofemoral joint pain, instability, chondromalacia, and osteoarthritis [1, 10, 30, 31]. In this study, though the Insall–Salvati ratio was significantly higher in the patellar cartilage lesions group than that in the control group, there were just 24.2% (24/99) of patients in the patellar cartilage lesions group showed a pathologic Insall–Salvati ratio of > 1.2, which coincided with the results of Ambra et al. [11]. Moreover, Ali et al. [28] found that the Insall–Salvati ratio did not correlate with the severity of patellotrochlear articular cartilage defected. In our opinion, the Insall–Salvati ratio in most patients with patellar cartilage lesions was in the normal range, which might reduce the instruction of this parameter in practical clinical work. Therefore, it was of great clinical impact to find more effective and direct evaluation indicators.

In the present study, the SP-ET index was used to define the position of the patella relative to the femoral trochlear, and SP-ET index showed a moderate correlation with the Insall–Salvati ratio. Ali et al. [28] and Mehl et al. [20] have used the patellophyseal index to assess patellar height in their studies, which was calculated as the distance from the most superior point of the patellar cartilage to the most superior point of the femoral cartilage divided by the length of patellar articular cartilage, and there were some similarities between the SP-ET index and patellophyseal index. However, the measurement of the SP-ET index only needed to be done at one plane (the sagittal plane with the greatest longitudinal diameter of the patella), and without the effect of axial displacement of patella. Therefore, we thought that the SP-ET index had a good practicability, and the intra- and inter-reader reliabilities of this parameter were also good in our study.

Mehl et al. [20] suggested that there was no significant difference in the patellophyseal index between the patellar cartilage defect group and control group. But different from the study of Mehl et al. [20], the study of Ali et al. [28] have found that there was a statistically significant difference in the patellophyseal index between normal group and severe patellar cartilage defect group. In the present study, we found that the SP-ET index in patellar cartilage lesions group was significantly higher than that in control group, and the AUC of the SP-ET index was significantly higher than that of the Insall–Salvati ratio. The difference in the results might be related to the different severity of the included participants in these studies, and the strict control of the flexion angle of the knee in our study. Because in our practical measurement process, it was found that the flexion angle of the knee had a great impact on the value of the SP-ET index. According to the results of these studies, we thought that the correlation of the SP-ET index and patellophyseal index with the occurrence and severity of patellar cartilage lesions/patellar cartilage defect required prospective and biomechanical studies to explore further.

Our study has several limitations. First, this was a retrospective study, and the selection of participants in the patellar cartilage lesions group was based on MR images without histologic or pathologic evidence. Second, the measurement errors might be caused by the formation of patella spurs, only patients with grade 1–2B patellar cartilage lesions were included in the patient group, therefore, patients with severe patellar cartilage lesions were excluded in this study. Third, the flexion angle of the knee has a great influence on the SP-ET index, so the measurement of the SP-ET index needs to control the flexion angle of the knee strictly. Fourth, our results showed that the specificity of SP-ET index was lower, which will increase the false negative rate, and may lead to an incorrect assessment of the sagittal position of the patella. Moreover, the comprehensive assessment of the influence of patellar and trochlear morphology on patellar cartilage lesions is also our next step. In addition, there were no strict age and gender matching between the two groups of participants. Future research efforts should focus on the relationship between the SP-ET index and patellofemoral osteoarthritis using a prospective longitudinal study, and the role of the SP-ET index in different grades of patellar cartilage lesions.