Analyzing parsicle screws as a viable alternative to pars screws and pedicle screws for C2 posterior instrumentation fixation

Abstract

Introduction

This study examined the possible clinical utility of “parsicle screws” in securing C2 instrumentation.

Methods

Ten patients’ C2 vertebrae were virtually reconstructed using computer-aided design software. Pedicle, pars, and parsicle screws were virtually placed in the vertebrae.

Results

In addition to establishing the trajectory and theoretical safety of parsicle screws, this study determined that parsicle screws were significantly longer than pars screws (p = 0.005).

Conclusion

The additional length of parsicle screws may improve construct stability. As such, parsicle screws should be examined as an alternative to pars screws in patients unable to receive C2 pedicle screws.

1. Introduction

Posterior instrumentation of the C2 vertebra is typically secured with either pedicle screws or pars screws. Pedicle screws are preferred as they allow for the placement of longer screws,¹ a characteristic which correlates with pullout strength.2, 3, 4 That said, approximately 34% of patients are considered to be at high risk for C2 pedicle screws and require the use of shorter pars screws to avoid breaching the vertebral arch or spinal canal medially.⁵ Anatomical constraints which prevent the use of C2 pedicle screws include the presence of infection on the posterior portion of the cervical spine, destroyed pedicles, marked osteoporosis, abnormally small pedicles, and major anomalies of vertebral arteries.⁶ The majority of these contraindications are rare, however, high-riding vertebral arteries (HRVAs) and narrow C2 pedicles are observed in approximately 14.5% and 9.5% of patients respectively.⁷ In order to maximize screw length in patients where pedicle screws cannot be safely placed, finding an alternative screw trajectory has garnered interest.

The use of a hybrid or “parsicle” screw has recently been explored as a way to increase screw length in patients unable to receive traditional C2 pedicle screws.⁸ As the name suggests, parsicle screws have a trajectory between that of pars and pedicle screws. This hybrid trajectory theoretically avoids vital anatomical structures medially and laterally while allowing for the placement of longer screws than the trajectory of pars screw would (Fig. 1, Fig. 2). Despite the potential benefits, the literature examining the feasibility of parsicle screws is sparse.

Fig. 1.

Posterior 3D rendering of a patient's C2 vertebra with virtual pedicle, parsicle, and pars screw shown

Fig. 2.

Superior 3D wire mesh rendering of a patient's C2 vertebra with pedicle, parsicle, and pars screw shown.

The purpose of this study is to define the trajectory of a parsicle screw and to determine whether there is a statistically significant difference between pars, pedicle and parsicle screw lengths. To test this, the authors generated 3D computer models of patients' C2 vertebrae using pre-operative CT scans. Advances in 3D computer-aided design (CAD) technology offer spine surgeons a more precise method of measuring optimal screw trajectory and screw length. 3D modeling also provides the opportunity to directly compare all three screw types within an individual patient's vertebra. In this study, we modeled ten patients' C2 vertebrae with pedicle, pars, and parsicle screws using CAD technology in order to define the trajectory of a parsicle screw and to perform a comparative analysis of screw lengths based on screw type.

2. Methods

A trauma radiological database was queried for cervical CT scans. Patients were considered if they were older than 18 years, had no C2 fracture, and had no prior instrumentation at C2. DICOM images were obtained from these scans and imported into a CAD software program that generated 3D models of the entire CT scan. The C2 vertebrae were isolated by excluding all bony structures above and below.Virtual screws were then placed based on starting points previously described in the literature.^9,10 The starting point for pars and parsicle screws was the midpoint of the pars in the X-axis and 3–4 mm above the facet joint (or 3–4 mm from the inferior border of the inferior articular process) in the Z-axis, while the starting point for the pedicle screw was 2 mm lateral to the point that bisects the midpoint of the lamina in the Z-axis and midpoint of the pars in the X-axis. Screw lengths and angular trajectories were recorded for pars, pedicle and parsicle screws, including whether the screw breached the vertebral foramen or central canal. Ten models with a total of 20 screws for each screw type were analyzed. A one-way ANOVA was conducted to determine if there was a significant difference in screw length based on screw type. Following ANOVA and Levene's test to determine homogeneity of variance, post-hoc analysis was performed. Cohen's f statistic was calculated to determine effect size.

3. Results

The presence of vertebral foramen and central canal breaches was assessed for each screw type to determine which allowed for a safe trajectory. 90% (18/20) of pedicle screws, 100% (20/20) of pars screws, and 95% (19/20) of parsicle screws could be placed without breaching the vertebral foramen or central canal.

The average trajectory and range of trajectories for each screw type were calculated (Table 1). Screws that breached the C2 vertebrae were excluded. The average mediolateral angle was 18.2° (range: 5–25°) for parsicle screws, 31.4° (range: 20–40°) for pedicle screws, and 0.05° (range: 0–1°) for pars screws. The average cephalad angle was 40.0° (range: 22–58°) for parsicle screws, 23.2° (range: 13–33°) for pedicle screws and 44.2° (range: 25–60°) for pars screws.

Table 1.

Average trajectory for pedicle, pars, and parsicle screws.

	Mediolateral Angle (degrees)			Cephalad Angle (degrees)
	Pedicle	Pars	Parsicle	Pedicle	Pars	Parsicle
Patient 1R	40	0	18	15	35	27
Patient 1L	36	0	21	15	35	30
Patient 2R	30	0	23	30	50	55
Patient 2L	35	0	25	30	55	52
Patient 3R	24	1	19	30	50	50
Patient 3L	31	0	17	30	60	58
Patient 4R	37	0	23	16	49	47
Patient 4L	32	0	15	21	45	46
Patient 5R	35	0	17	25	55	38
Patient 5L	32	0	19	31	47	47
Patient 6R	20	0	5	32	49	44
Patient 6L	26	0	12	33	51	47
Patient 7R	34	0	20	15	28	25
Patient 7L	34	0	Breach	13	38	Breach
Patient 8R	29	0	18	20	47	32
Patient 8L	28	0	21	24	47	40
Patient 9R	Breach	0	18	Breach	25	22
Patient 9L	Breach	0	21	Breach	27	26
Patient 10R	30	0	15	20	45	35
Patient 10L	32	0	18	17	45	39
Average	31	0	18	23	44	40

Table 1: The mediolateral and cephalad angle for pedicle, pars, and parsicle screws are provided in the above table. “Breach” implies that the screw was not able to be safely placed in that patient without penetrating the vertebral arch or spinal canal medially.

Average screw lengths for pedicle, pars, and parsicle screws are displayed in Table 2. A statistically significant difference in average screw lengths was identified using one-way ANOVA (F (2,54) = 5.81, p < 0.05, η² = 0.177). Post-hoc analysis was conducted to determine pairs of screw types with statistically significant differences in screw length. Based on Levene's test, homogeneity of variance can be assumed for post-hoc analysis (F = 0.384, p = 0.682). Post-hoc Tukey HSD analysis indicated a significant difference between pars and parsicle screws (Q = 4.64, p = 0.005) and no statistically significant difference between pars and pedicle screws (Q = 3.20, p = 0.0702) or between pedicle and parsicle screws (Q = 1.44, p = 0.568). Cohen's f statistic between the difference in pars and parsicle screw length was 0.454.

Table 2.

Average screw length for pedicle, pars, and parsicle screws (mm).

	Pedicle	Pars	Parsicle
Patient 1R	28	34	34
Patient 1L	32	32	34
Patient 2R	30	34	34
Patient 2L	30	32	36
Patient 3R	30	28	30
Patient 3L	30	30	30
Patient 4R	30	22	24
Patient 4L	28	24	24
Patient 5R	26	22	28
Patient 5L	26	28	30
Patient 6R	32	30	32
Patient 6L	30	28	32
Patient 7R	34	28	38
Patient 7L	34	28	Breach
Patient 8R	32	22	34
Patient 8L	34	26	34
Patient 9R	Breach	32	32
Patient 9L	Breach	30	32
Patient 10R	34	28	34
Patient 10L	34	28	34
Average	31	28	32

Table 2: Screw lengths for pedicle, pars, and parsicle screws are provided in the above table. “Breach” implies that the screw was not able to be safely placed in that patient without penetrating the vertebral arch or spinal canal medially.

4. Discussion

The first objective of this study was to determine whether parsicle screws can be safely inserted in the C2 vertebra. It was determined that pedicle, parsicle, and pars screws could be placed without breaching the vertebral foramen and central canal at least 90% of the time. It's important to note that this study did not examine the safety of placing parsicle screws in patients with anatomical variants.

The second objective of this study was to define parsicle screws based on their trajectory. Results indicate that the average trajectory for a parsicle screw is 18° medially and 40° cranially. The mediolateral angle for a parsicle screw is approximately halfway between the 31° for a pedicle screw and 0° for a pars screw. In terms of cephalad angle, parsicle screws were similar to pars screws. The average cephalad angle for pedicle screws was 23° cranially while pars and parsicle screws had 44° and 40° respectively. Previous studies concur, recommending that parsicle screws be placed 15–20° medially and 30–45° cranially.⁸ As such, our results regarding screw trajectory validate the existing literature.

The third objective of this study was to calculate differences in screw length between screw types. Root area, the total surface area of a screw contacting bone, is positively correlated with pullout strength.¹¹ Longer screws increase root area and create more stable posterior constructs. As such, we aimed to see which screw types allow for the placement of the longest screws. One prior study reported the length of parsicle screws at 20–26 mm while the average parsicle screw length in this study was longer at approximately 31 mm.⁸ It could not be determined whether the 20–26 mm range in the referenced paper was an estimate or based on statistical analysis. If based on an estimate, our study adds statistically validated average parsicle screw length to the literature. If the 20–26 mm range was based on statistical analysis, the difference between this dataset and prior studies could be attributed to the use of 3D simulations to measure screw lengths opposed to measuring screw lengths in vivo. One drawback of 3D modeling compared to in vivo screw placement is the lack of tactile feedback and the inability to incorporate cortical thickness in the models. All screws in this study were placed iteratively using the longest screw possible without breaching the cortex ventrally, so this may overestimate some of the screw lengths. The lengths of the pedicle and pars screws in this dataset were also substantially longer than those reported in prior literature, providing credence to this theory.¹²

The significant difference in the average length of pars versus parsicle screws as well as the large effect size determined by Cohen's f statistic warrants further research with larger sample sizes in both patient populations and 3D simulations. To our knowledge, this is the first study examing whether parsicle screws allow for a significantly longer screw to be placed than pars screws. The fact that there was no significant difference between parsicle and pedicle screw length suggests that either should allow for superior pullout strength compared to pars screws. Further research is necessary to parse out differences between parsicle and pedicle screws to determine superiority or to better characterize the clinical indication for each. Interestingly, the difference in screw length between pars and pedicle screws approached statistical significance but did not reach it (p = 0.070) as previous literature has indicated.¹ It's likely that the small sample size of this study prevented the difference in these screws from reaching statistical significance. Table 3 summarizes the relevant characteristics of pedicle, parsicle, and pars screws from this study.

Table 3.

Summary of screw length and angles.

	Pedicle	Pars	Parsicle
Average Screw Length	31 mm	28 mm	32 mm
Average Mediolateral Angle	31°	0°	18°
Average Cephalad Angle	23°	44°	40°

Table 3: A summary of the average screw lengths, mediolateral angle, and cephalad angle for pedicle, pars, and parsicle screws.

This study has a few limitations. First, the small sample size makes it difficult to make firm conclusions on the safety risks each individual screw type poses. Larger studies must be completed to determine the safety profile for each screw type. Second, while 3D simulations come with many benefits, it's important to validate these results in vivo. Computer-generated images may allow for a longer screw to be placed than would be seen in a clinical setting.

5. Conclusion

Our results indicate the safe trajectory of a parsicle screw is a mediolateral angle approximately between a pedicle and pars screw and a cephalad angle similar to a pars screw. Additionally, based on computer-generated models of C2 vertebrae, our results show that parsicle screws allow longer screw placement than would be possible if pars screws were used.

Funding

This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

Author contributions

Eric Dilbone – Conceptualization, Data Curation, Investigation, Methodology, Project Administration, Resources, Software, Supervision, Validation, Writing – Review & Editing.

Rishabh Gupta – Formal Analysis, Investigation, Validation, Visualization, Writing – Original Draft, Writing – Review & Editing.

Byron Stephens – Conceptualization, Investigation, Methodology, Supervision, Project Administration, Supervision, Validation, Writing – Review & Editing.

Declaration of competing interest

Eric Dilbone - none, Rishabh Gupta – none, Byron Stephens – has received consultancy payments from Depuy-Synthes and Stryker Spine.