Our results showed that the incidence of PJK was 18.6%, and stood in the lower range of the literature for similar patients [3, 8, 22]. Previous studies have found that greater correction of LL resulted in higher incidence of PJK and they attributed surgical overcorrection of sagittal deformities as a risk factor [23-25]. Kim et al. [23] identified that excessive lordosis and larger sagittal balance correction led to PJK, requiring revision surgery. Lafage et al. [24] recently refocused the attention of PJK prevention onto spinal alignment, by providing age-adjusted sagittal alignment parameters for adults. They also reported that with increased overcorrection beyond age-adjusted sagittal alignment there is a corresponding increase in PJK severity. PJK is more likely to occur after sagittal alignment (like LL) overcorrection, which is accepted by most researchers. Our clinical experience suggests that LL overcorrection is more likely to occur in patients with a low PI. Up to now, there are no reports on the influence of PI value on PJK in Lenke 5 AIS patients.
Pelvic incidence remains relatively constant during childhood and it determines the pelvic orientation and the size of the lordosis, which is closely correlated with it [16]. In short, the greater the PI, the greater has to be SS, PT, or both. The PI value varies widely among individuals and it is related to the capacity to compensate for sagittal alignment. Our results showed that there was no significant difference in the incidence of PJK between the high and low PI group patients, but the mechanism of PJK was different among high PI and low PI patients. Chinese children and adolescents had lower PI, PT, and SS compared with the Caucasian population [17-19]. In our previous study [20], the mean value of PI of Chinese adolescents was about 45°, so we chose the value of 45° as the critical value of high PI and low PI.
For the patients with low PI, The PJK patients have a larger postoperative LL and worse PI-LL than the Non-PJK patients. However, for patients with high PI, PJK patients have a similar PI-LL to Non-PJK patients. Roussouly et al. advocated that the ability to retrovert the pelvis is limited by a patient’s PI, and patients with a low PI have a small capacity to compensate for their sagittal imbalance through pelvic retroversion [26]. In the circumstance of low PI, if the surgical procedure introduces more lumbar lordosis than the patient’s PI can accommodate, then the thoracic spine would begin to join into compensation mechanism with additional thoracic kyphosis (PJA↑), which may result in PJK (Figure 1). Patients with low PI has poor pelvic compensation to abnormal sagittal alignment. In such low PI cases, the overcorrection of lumbar lordosis after surgery would easily exceed pelvic compensation. Therefore, postoperative LL increase was more likely to be compensated by increased proximal kyphosis above the fusion levels. For the reasons mentioned above, it is reasonable to assume that low PI patients require strictly matched postoperative LL.
In contrast, patients with high PI had larger PT, SS, LL. These results are consistent with previously published articles [10, 27]. In this study, we found no statistically significant differences in features on preoperative radiological findings between the PJK and Non-PJK subgroups with high PI. The PI-LL were all within 10° in the high PI group before and after surgery, which means PI and LL match well. Even though LL did not meet the ideal curvature at the time of surgery, the remaining lumbar segments below LIV could compensate. It is well known that AIS patients generally have a flat back. In particular, posterior column osteotomy may further reduce TK. To keep well-matched between PI and LL, TK and LL, patients will have a compensatory increase in TK after surgery. Patients increased their proximal kyphosis above UIV to compensate for the postoperative decrease in instrumented TK and to balance head over the pelvis, which introduces PJK (Figure 2). The closer the UIV selection is to the cranial level, the fewer remaining thoracic segments that can compensate for TK reduction. The risk of PJK may increase in this situation. Our current findings revealed that the UIV in the PJK group was closer to the cranial level, which was consistent with the above compensatory mechanism. In practice, there remains no clear consensus regarding the optimal LIV and UIV in individual patients. The selection of fusion levels continues to vary widely based on patient, regional, and surgeon variables. Therefore, more prospective studies or multi-center studies are needed to verify our results.
For Lenke 5C AIS patients, we suggest that the surgical procedure should be designed according to the PI to minimize the incidence of PJK: the conventional posterior column osteotomy will potentially increase LL. For patients with low PI, overcorrection of LL should be strictly avoided during surgery. We recommend that we should not bend too much lordosis in the lumbar level and begin bending the lordosis below L3 or L4. For patients with high PI, the selection of UIV should not be too cranial (usually below T6). If UIV selection was inevitable to enter the cranial level (above T6), we suggest to increase the density of the screws in the thoracic level and try to bend more TK to correct the flat-back deformity in such case.
Several limitations still exist. Firstly, it was a retrospective study and had a relatively small sample size. Further analysis, such as a randomized trial with a larger number of patients, will be necessary to confirm our findings. Secondly, data collection for pre- and postoperative clinical information was incomplete in some areas such as health-related quality of life scores. Therefore, we were unable to measure the clinical impact of radiographical PJK.