Comparing the Upper Instrumented Vertebrae Tilt Angle vs Screw Angle in the Development of Proximal Junction Kyphosis After Adult Spinal Deformity Surgery: Which Matters More?

Discussion

The current study analyzed the impact of UIV tilt angle and UIV screw angle on PJK/F in ASD patients with a UIV in the lower thoracic or lumbar region. When controlling for several confounding variables associated with PJK/F, lordotically angled UIVs were significantly associated with increased odds of PJK/F. Furthermore, cranially directed UIV screws were associated with increased odds of PJK/F. Comparing the 2, it appeared that the UIV screw angle was the stronger predictor of PJK/F than the UIV tilt angle. Conversely, neither variable demonstrated a significant relationship with the remaining types of mechanical complications or with PROMs.

UIV tilt angle and the development of mechanical complications have been less rigorously studied than pedicle screw placement/angle. A recent study of 190 patients undergoing ASD surgery found that PJK occurred more frequently in patients with a higher postoperative UIV slope, but this association was not significant on regression analysis.11 However, the authors did not control for the UIV screw angle, which showed a significant association with PJK in previous literature. Lafage et al12 found that posterior construct inclination, or more lordotically angled UIV endplates, were present more often in patients who developed PJK. However, neither of these studies accounted for screw angle when assessing UIV tilt. Our data suggest that more lordotically angled UIVs in the thoracolumbar spine are associated with increased odds of PJK/F, which was present when controlling for other factors, including the UIV screw angle. However, UIV tilt angle proved to be the less important variable in predicting PJK/F than UIV screw angle.

UIV screw angle was more strongly associated with the development of PJK/F than the UIV tilt angle in the current analysis. Specifically, cranially directed UIV screw angles were associated with increased odds of the development of PJK/F. Several potential reasons exist as to why UIV screw angle was the more important factor to UIV tilt angle. A screw that compromises the integrity of the unfused facet joint or violates the superior endplate may accelerate the degeneration process and lead to the development of PJK/F. In contrast, a perfectly placed screw, parallel to the superior endplate or slightly caudally directed, in an otherwise lordotic UIV, may have a lower chance of failure. Our findings are in line with at least 2 previous studies. Harris et al5 found that UIV pedicle screws angled ≥3° cranially had significantly greater odds of developing PJK and PJF, compared with those with screws angled <3° cranially. Similarly, Jung et al13 found that straightforward screws at the thoracolumbar junction were protective against PJK (straightforward screws are functionally caudally angled due to the natural lordotic curvature of the thoracolumbar junction). Although we produced similar findings, our study offered a larger cohort (117 vs 96 and 83 patients) and provided insight into the importance of the relationship between UIV screw and tilt angles. Importantly, our study focused on ASD fusions to the lower thoracic or upper lumbar spine, which represents a different population that fusions to the upper thoracic spine. Furthermore, our study found that UIV screw angle was neither associated with any additional varieties of mechanical complications nor was it predictive of differences in PROMs, an assessment that has not been conducted in the existing literature.

While the current study cannot explain causation, several potential theories may explain our findings. In terms of the UIV angle, a lordotic UIV tilt angle may predispose one to PJK/F due to excessive spinal realignment during ASD surgery, causing undue force on the construct. Conversely, a UIV angle in line with the horizontal may allow for more flexibility at the top end of the construct and may reduce exogenous forces that predispose one to adjacent segment degeneration. According to the Roussouly classification,14 if the lower arc of spinal kyphosis is too pronounced, the unfused spine may have to compensate with more kyphosis, thus precipitating PJK/F. Regarding UIV screw angle, more cranially directed UIV screw angles may be associated with increased odds of PJK/F secondary to multiple factors. First, as constructs settle and patients become more active after surgery, cranially directed screws may disrupt the suprajacent disc space/endplate. Even if the screws do not violate the disc space, as cranially directed screws approach the endplate, they may begin to erode the bone abutting the endplate, thus decreasing cortical purchase and predisposing to mechanical complications. It is also possible that caudally directed screws are placed in a different vector than the vector in which screws are most biomechanically predisposed to pullout. By being placed in a more downward direction, not in line with the easiest method of loosening and failure, caudally directed screws may protect against a more horizontal pullout force. Finally, the more cranially directed screws may provide a greater pullout strength, as has been suggested in prior biomechanical studies,3,4 and may in turn make the construct more stiff than is optimal and may hasten adjacent segment degeneration.

The development of, and therefore the prevention of, PJK/F itself is likely multifactorial. Other factors have been studied that may contribute to the development of PJK/F such as bone density, the use of adjuncts such as cement, hooks, or tethers, and selection of the UIV.3,4,7,15,16 These factors should also be carefully considered alongside UIV parameters. As UIV tilt angle and UIV screw angle are modifiable factors under the surgeon’s control, taken together they may be a useful tool to help mitigate the development of this complication. The surgeon should select a UIV level that is neutral when possible and ensure screws placed at the UIV are neutrally or caudally directed in order to minimize the risk of developing PJK/F. Nevertheless, the potential clinical implications should be exercised with caution, given that the current study was not designed to establish causality between these parameters and PJK/F. Future prospective studies should further investigate the causal relationships by including a control group to rigorously assess the impact of UIV tilt angle and UIV screw angle on the occurrence of PJK/F while controlling for potential confounding factors, thereby providing more robust evidence for clinical decision-making.

This study recognizes its limitations. First, a retrospective analysis of prospective data within a single institution has inherent limitations as some information might be lost. Prospective multicenter studies are encouraged to control for bias and increase generalizability. In addition, while we used Wald statistics and degree of freedom to compare the impact of tilt angle and screw angle on patient outcomes, the significant association does not imply causation due to the inherent limitations of a retrospective study with a small sample size. Retrospective studies pose significant limitations of the statistical analysis, making it challenging to confidently determine the association between UIV tilt angle and UIV screw angle with outcomes. Second, UIV tilt angle can be modified by patient position, and minor variation may occur during positioning for imaging, which may in turn affect study results. Third, factors such as screw length, screw-entry location, or endplate violation may affect the development of PJK/F and were not considered in this study. Fourth, a limited sample size may reduce the generalizability of the conclusions drawn, and further studies including larger cohorts should validate these findings. More than half of the total cohort were lost to follow-up and subsequently excluded from the analysis, which may further limit the generalizability of our analysis. However, given that these patients were from the same institution, it is likely that their exclusion would not substantially impact the overall findings and conclusions of our study. Fifth, the choice of the UIV between T7-L2 may have increased the heterogeneity of our patients sample due to the inherent difference in thoracic kyphosis, inflection point, and lumbar lordosis. Sixth, while we adopted the radiographic diagnosis of PJK,17 a consensus regarding the diagnosis of PJK remains poorly established. A normal reciprocal change in kyphosis angle due to the correction of hypolordosis of the lumbar spine may sometimes be mistaken with PJK, which may have overestimated our findings. A growing body of literature indicates that the classic definition of PJK may be inappropriately low.18 Seventh, since prior literature has shown that pelvic instrumentation may not directly impact the risk of PJK/F,19 we opted to include patients with and without pelvic extension of the arthrodesis. On the other hand, one of the major drawbacks of spinal deformity studies is that mechanical complications are multifactorial processes, and not all cofactors can be consistently accounted for. Even though we have accounted for multiple covariates, confounding bias may still exist due to the inherent limitation of retrospective studies and the multifactorial nature of mechanical complications. Finally, the mechanical complication rates were higher than previously reported in the literature, which may be due to the registry including operations from more than a decade ago, when techniques and knowledge of ASD surgery were evolving. It is likely that these complication rates do not reflect the current practice at our institution among the 4 fellowship-trained spine surgeons doing ASD surgery. However, we still found meaning in reporting our institution’s data to investigate the impact of UIV tilt angle and UIV screw angle on patient outcomes. Subsequently, it is important to mention that our current practice consists of optimizing bone mineral density preoperatively through anabolic prescriptions prior to proceeding with the surgery to mitigate the risk imposed by osteoporosis on postoperative outcomes, which we have controlled for in our regression analysis. Furthermore, ways to prevent PJK/F are currently being used, including surgical cements and adjunct devices, such as hooks and tethers. We also routinely incorporate spinal alignment principles as described by Roussouly, which has reinforced the importance of low lordosis and different spinal shapes dictated by pelvic incidence and sacral slope.