The “V” Sign

Abstract

Background:

Sacral (S1) pedicle screw misplacement in posterior percutaneous fixation (PPF) can be related to anatomical variability and a lack of reliable radiographic landmarks. This study highlights a reproducible anatomical landmark (the “V” sign) for the safe localization of the S1 pedicle entry point under fluoroscopy.

Methods:

Human cadavers (n = 14) were dissected for the anatomical description of the “V” landmark and its relationship with the entry point of the S1 pedicle screw. The “V” landmark was defined medially by the lateral border of the superior articulating process of S1 and laterally by the posterior projection of the sacral ala. The mean distance was measured between the bottom point of the “V” landmark and the anatomical entry point to the S1 pedicle (V-S1 entry point distance). A similar measurement was conducted on computed tomography (CT) scans of 135 patients who underwent PPF using the “V” sign as a landmark for S1 pedicle screw placement (270 screws). These were retrospectively evaluated for appropriateness of S1 screw entry points and for proper S1 screw alignment and breaches.

Results:

In the 14 cadavers, irrespective of the laterality and sex, the V-S1 entry point distance averaged 11.7 mm. On the medial-lateral axis, all entry points converged within 2 mm of a vertical line intersecting the base of the “V.” Additionally, the CT scan analysis (135 patients, 270 screws) revealed an optimal entry point for 100% of the screws and a 3.3% (n = 9 screws) breach rate. Six of the 9 identified breaches were minor, and only 1 (0.4% of the 270 screws) warranted revision.

Conclusions:

The “V” sign serves as a reliable anatomical and radiographic landmark for identifying the S1 pedicle entry point under fluoroscopic guidance. This landmark can help surgeons overcome the radiographic ambiguity of the sacral anatomy and ultimately reduces the rate of S1 pedicle screw misplacement.

Level of evidence:

Diagnostic Level IV. See Instructions for Authors for a complete description of levels of evidence.

S1 pedicle screws are commonly used during various lumbosacral fixation procedures. In open techniques, the ideal entry point for the S1 pedicle screw has been described as the inferolateral portion of the superior S1 facet. However, with adoption of minimally invasive spine surgery (MIS), the optimal entry point for percutaneous S1 pedicle screw placement remains speculative^1-4.

The confluence of the S1 articular process and sacral ala is V-shaped and has been described in open surgery¹. We are designating this confluence of osseous landmarks under fluoroscopy as the “V” sign (Fig. 1).

Fig. 1.

The left image is an anteroposterior radiograph of the spine highlighting the “V” sign on the right side (yellow curve) along with the entry point (yellow dot) for S1 pedicle screw insertion. The upper right image is a drawing (by C. Tannoury) of the posterior view of a male’s sacrum, including the “V” sign (red dotted lines), the inferior entry point (marked by an “X”), and the medial-lateral axis (green lines). The blue line represents the distance between the “V” sign and the entry point. The lower right image is a photograph of a cadaveric specimen. On the left side, the entry point is marked by a lead dot (yellow arrow) inferior to the “V” metallic wire; the L5 inferior facet was dissected and removed to highlight the S1 superior articular process. On the right side, the L5 inferior facet was preserved, the “V” metallic wire is highlighted by red dotted lines, and the entry point is marked by an “X” (medial-lateral axis; green lines).

Investigations of conventional fluoroscopy-guided S1 pedicle screw placement have shown a misplacement rate exceeding 11%, even reaching 20%^5-7. Advanced navigation and robotic technologies offer superior accuracy but remain limited by increased costs and complexity and lower availability^8,9.

Misplacement of S1 pedicle screws under fluoroscopic guidance in MIS can be the result of an inaccurate choice of the entry point, and may lead to iatrogenic injuries to adjacent neurovascular structures and sometimes to the need for revision surgery^10-12.

The primary purpose of this cadaveric morphometric study was to generate a quantitative description of the optimal S1 pedicle entry point in relation to a reliable surrounding osseous anatomical landmark designated as the “V” landmark. Also, with the hope that the fluoroscopic “V” sign can serve as a reliable feature to help locate the optimal entry for percutaneous S1 pedicle screw placement, the validity of the fluoroscopic “V” sign was investigated retrospectively using computed tomography (CT) scans in a cohort of patients who underwent MIS lumbosacral fixation using that landmark. In that investigation, we assessed the accuracy of S1 pedicle screw entry point placement and the overall S1 pedicle screw breach and misplacement rates.

Materials and Methods

Anatomical Study

All study specimens were obtained from the Department of Anatomy (Anatomical Gift Program) at Boston University School of Medicine. Fourteen human embalmed cadaveric specimens with presumably intact lumbar and sacral rachises were included. Two spine specialists performed the experimental procedure, each assisted by a spine surgery fellow. The specimens were placed in prone positions, and an anatomy specialist supervised the dissection, which was carried subperiosteally to expose the posterior osseous surfaces of the sacrum and the lumbosacral junction. Generous L5-S1 laminectomy and partial facetectomy were performed, allowing adequate visualization of the superior, medial, and inferior walls of the S1 pedicle. The ideal anatomical entry points to the S1 pedicles were identified under direct visualization. Then a small pilot hole was created using an awl at the posterior osseous surface of the sacrum, indicating the ideal entry point, and was filled with a lead bead for identification on radiographs and fluoroscopy. Similarly, the “V” landmark was outlined using a “V”-shaped remodeled metal wire (Figs. 1 and 2). Under direct anatomical evaluation, the “anatomical” distances between the ideal entry point and the base of the “V” landmark on both axes, cephalad-caudad and medial-lateral, were measured in millimeters by each spine specialist, to minimize human error, using a graded surgical ruler. Circular metallic wires were wrapped around the S1 osseous pedicle to radiographically underline its contour in relation to the ideal entry point (Figs. 3 and 4). Then, 7 × 45-mm pedicle screws (Expedium Spine System; DePuy Synthes) were placed through the S1 pedicles (Fig. 4).

Fig. 2.

Sacral region of a cadaveric specimen showing anatomical landmarks and the entry point, which is marked by an “X.” SAP = superior articular process.

Fig. 3.

“V” signs marked by metallic wire and entry points marked by lead under fluoroscopy.

Fig. 4.

Before (left) and after (right) insertion of S1 screws through a circular metallic wire loop using the “V” sign as an anatomical landmark and the lead dot as the entry point. The metallic “V” wire was installed for descriptive purposes and to highlight the “V” sign; it is not required intraoperatively. The circular metallic loop was wrapped around the pedicle to highlight the full perimeter of the S1 pedicle and demonstrates proper screw trajectory.

In the anatomy laboratory, images were made using C-arm fluoroscopy with the cadaveric specimens placed prone on radiolucent tables, and corresponding radiographs (in the anteroposterior and lateral planes) of the lumbosacral spine were obtained (Figs. 3 and 4). Similar to clinical practice and to reduce angular variability, the C-arm tube was systematically positioned collinearly with each cadaveric specimen’s S1 superior end plate, thereby projecting the end plate as a flat surface on radiographic imaging (Fig. 5). The spatial relationship between the metal markers, delineating the “V” landmark, and the optimal S1 pedicle entry point was analyzed.

Fig. 5.

Intraoperative fluoroscopic image highlighting the S1 screw entry points according to the “V” sign. The V-S1 entry point distance was measured between the bottom of the “V” and the tip of the awl.

CT Study in Retrospective Patient Cohort

To highlight the clinical relevance of this guided procedure, CT scans of patients who had undergone posterior percutaneous fixation (PPF) with S1 screws using the “V” landmark as a guiding sign were retrospectively studied; the validity of this landmark was determined by evaluating the accuracy of both the S1 screw entry points and the overall S1 screw placement and breaches. The CT scans (abdomen-pelvis) had been obtained because of problems unrelated to the spine or to investigate new or worsening residual back and/or leg pain.

After obtaining institutional review board approval, we conducted a retrospective review of the records of 806 consecutive patients who had undergone spinal PPF involving the S1 level between 2008 and 2017. Of these 806 patients, 135 (with a total of 270 S1 pedicle screws) had available postoperative CT scans of the lumbosacral spine, which were evaluated by 2 blinded spine specialists using the Gertzbein-Robbins classification system (GRS) for S1 pedicle screw breach¹³. According to the GRS, pedicle screws with a grade of A or B (0 to <2-mm breach) are considered to be accurately placed, whereas grade C indicates a mild-to-moderate breach (2 to <4 mm) and grades D and E, a major breach (4 to 6 mm and >6 mm, respectively). A major breach can be associated with severe complications^9,10. Interobserver disagreement, if encountered, was settled by a third reviewer. A kappa coefficient was calculated to evaluate the interobserver agreement¹⁴. The 2 investigators also measured the radiographic V-S1 entry point distance using the intraoperative fluoroscopy (Fig. 5). Each data point was calculated based on the average of both readers’ measurements to account for inter-rater variations. To account for magnification, each patient’s radiographic measurement was matched with their respective V-S1 entry point distance measured on the postoperative CT scan and a magnification rate was calculated based on the 2 measurements.

Statistical Analysis

For statistical analysis, each S1 pedicle, regardless of its laterality, was considered an independent data point for the distance measurement between the bottom of the “V” landmark and the S1 entry point (V-S1 entry point distance). The mean distance in the specimens was calculated. This measurement was also compared between the right and left sides of the sacrum using a paired Student t test (Microsoft Excel version 16.47.1).

Source of Funding

This work did not receive any outside financial support.

Results

Of the 14 dissected cadavers, 2 had evidence of pre-investigation extensive unilateral osteotomies that resulted in distortion of the “V” sign and therefore precluded its use as a radiographic reference. Hence, the total number of S1 pedicle screw entry points examined in the study was 26, with a single screw used in those 2 cadavers and bilateral S1 screws used in the remaining 12. Of the 14 studied cadavers, 9 were male, and the mean age of the donors was 86.6 years (range, 67 to 99 years).

The average anatomical V-S1 entry point distance (and standard deviation) was 12.1 ± 1.89 mm (range, 10 to 16 mm) on the left side and 11.3 ± 1.67 mm (range, 10 to 14 mm) on the right (Table I). The medial-lateral position of the entry points in reference to the vertical projection from the base of the “V” was within 0 to 2 mm in all specimens (Fig. 1). These measurements did not differ significantly between the right and left sides, as shown via a 2-tailed, paired Student t test (p = 0.55). Taking all points together, the average V-S1 entry point distance was 11.7 mm. There was no significant difference between male and female specimens, as shown by a Student t test (p = 0.11).

TABLE I.

Distance Between “V” Sign and S1 Entry Point in Cadaver Study

Sex	Age (yr)	V-S1 Entry Point Vertical Distance* (mm)		Medial-Lateral Offset Versus Vertical Line‡ (mm)
		Left Side	Right Side†
F	97	11 ± 0	10 ± 0	<1
F	92	12 ± 1	11 ± 1	<1
F	79	11 ± 1	10 ± 0	<1
F	99	10 ± 0	10 ± 0	<1
F	72	11 ± 0	14 ± 0	<1
M	90	12 ± 1	10 ± 0	<1
M	88	11 ± 1	12 ± 2	<1
M	97	14 ± 1	Already damaged	<1
M	79	15 ± 0	Already damaged	<1
M	84	11 ± 1	13 ± 1	<1
M	90	11 ± 0	10 ± 0	<1
M	92	11 ± 0	10 ± 0	<1
M	67	16 ± 0	12 ± 1	<1
M	87	12 ± 1	13 ± 1	<1

^*The values are given as the mean and standard deviation.

^†Already damaged = cadaver with S1 area damaged prior to the study.

^‡The medial-lateral offset of the S1 entry point versus a vertical line from the base of the “V”.

The patient cohort’s mean age was 58.7 years (range, 21 to 88 years), and 85 (63%) were female. Under fluoroscopy, the vertical distance between the metal markers highlighting the “V” base and the optimal S1 entry points averaged 12.9 ± 0.9 mm (versus an average anatomical value of 11.7 mm). The medial-lateral interval variability in reference to the vertical projection from the base of the “V” sign was 0 to 2.5 mm (versus anatomical values of <2 mm). The average V-S1 entry point distance in the retrospective CT cohort was 10.8 mm. Similar to the cadaveric cohort, there was no significant difference in entry point distance between males and females (p = 0.08) or between the right and left sides (p = 0.60). Additionally, there was no significant difference in entry point distance between patients with and those without an S1 pedicle screw breach (p = 0.20). Comparing the retrospective cohort’s individual fluoroscopic and CT measurements revealed a 119% magnification (range, 107% to 128%).

A pedicle screw breach can occur with a suboptimal entry point or improper screw angulation despite an optimal entry point. The latter was noted in our patient cohort in which we retrospectively reviewed CT scans—i.e., all entry points were noted to be optimal and situated within the confinements of the S1 pedicle wall projections in the axial, sagittal, and coronal planes but 9 screws (3.3%) were found to have radiographic breaches using the GRS. In 6 pedicles (2.2%), the breach was classified as minor (<3 mm; GRS grade C)^13,15. Of the remaining 3 pedicles, 2 (0.7%) had a grade-D breach and 1 (0.4%), grade-E. Analysis of the interobserver agreement for the GRS grading yielded a kappa of 0.86 (almost perfect agreement)¹⁴ (Table II). Furthermore, we examined the orientation of the 9 breaching screws and found 5 to be medial, 2 inferior, and 2 inferomedial (Tables III and IV).

TABLE II.

Interobserver Grading Agreement^* in Retrospective CT Study

	Grader 1
	A	B	C	D	E
Grader 2
A	250	1	0	0	0
B	1	8	0	0	0
C	0	1	5	1	0
D	0	0	0	2	0
E	0	0	0	0	1

^*Kappa score = 0.86.

TABLE III.

S1 Screw Accuracy Data in Retrospective CT Study

Screw Trajectory		Screw Breach		Screw Breach Orientation
		Minor (<3 mm)	Major (>3 mm)	Medial	Inferior	Inferomedial
Accurate	261 (96.7%)	—	—	—	—	—
Misplaced	9 (3.3%)	6 (2.2%)	3 (1.1%)	5 (1.9%)	2 (0.74%)	2 (0.74%)
Total	270	6	3	5	2	2

TABLE IV.

Patient Characteristics in Retrospective CT Study (N = 135)

Demographics
Age* (yr)	58.7 ± 12.6
Sex (no. [%])
Female	85 (63.0%)
Male	50 (37.0%)
BMI* (kg/m2)	30.6 ± 6.1
Indication for primary surgery
Degenerative scoliosis	24 (17.8%)
Degenerative spondylolisthesis	44 (32.6%)
Degenerative disc disease without deformities	6 (4.4%)
Spondylolisthesis	18 (13.3%)
Stenosis with degenerative arthritis	21 (15.6%)
Pathological compression fracture	2 (1.5%)
Indication for revision surgery (no. [%])
Adjacent segment disease	15 (11.1%)
Pseudarthrosis	4 (3%)
Rod breakage	1 (0.7%)

^*The values are given as the mean and standard deviation.

Moreover, among all studied patients (135), only 1 (1.0%) developed postoperative neurologic symptoms (radiculopathy due to a right S1 screw medially breaching; grade E) and warranted surgical revision. The surgery resulted in complete symptom resolution. No other neurologic, vascular, or visceral injuries were encountered due to S1 screw placement.

Discussion

When performing fluoroscopically guided percutaneous S1 screw placement, the C-arm must be tilted caudally in line with the sacral slope. Such tilt exaggerates the depth of the anatomical “V” landmark, a confluence between the L5-S1 facet joint (superomedial) and the sacral ala (lateral) (Fig. 1). The radiographic/fluoroscopic reflection of the osseous cortex of the anatomical “V” landmark is referred to as the radiographic “V” sign (Figs. 1 and 3).

Unlike the established anatomical and precise radiographic landmarks used for percutaneous placement of lumbar pedicle screws, specific challenges are imposed at the S1 level by its unique vertebral shape, spatial interference from the overlapping posterior iliac bones, and the critical adjacent neurovascular structures^1,12,16. The S1 pedicles are larger than those in the lumbar spine, but the cortices are thinner and the cancellous bone is less dense, limiting sacral pedicle screw cortical purchase and pullout strength¹. Consequently, the sacral osseous anatomy and osteological characteristics predispose the S1 level to an increased rate of screw misplacement^6,17. Ahmad and Wang examined the accuracy of fluoroscopically guided pedicle screw placement in the lumbar spine and lumbosacral junction and found the perforation rate to be highest at S1 (20.6%)⁶. Similarly, meta-analysis of pooled data from the lumbar and thoracic levels demonstrated a significantly higher rate of pedicle screw breaches at S1¹⁸. The wider S1 body along with the steep angles of the pedicles and the “trefoil” canal shape characterizing the S1 level were hypothesized as risk factors⁶. Accordingly, Ahmad and Wang recommended prudence with percutaneous placement of S1 pedicle screws⁶. Our study investigated the “V” sign and its exclusive application in percutaneous S1 screw insertion. Thus we could not establish its use as an anatomical landmark for PPF with screws at other sacropelvic levels, including S2 or alar screws.

As described previously, conventional open S1 pedicle screw placement starts with identifying the entry point at the confluence of the S1 articular process and the sacral ala and then follows a medial (about 30° to 35°) and caudal (15° to 25°) trajectory to provide strong cortical purchase in the sacral promontory¹.

A study indicated that CT-guided surgery could be valuable when the posterior lumbosacral anatomy is distorted¹⁹. Nonetheless, compared with the ubiquitous C-arm technology, the navigation modalities are associated with increased cost, radiation exposure, and surgical time, and lack of widespread institutional availability^20,21.

Wiesner et al. analyzed the screw angle and pedicle angle of 11 misplaced S1 pedicle screws. They found that only 1 violation could be explained by improper angulation and concluded that most misplacements result from an inappropriate entry point for the screw²². Therefore, it is of great importance to reliably define the optimal entry point for S1 pedicle screw placement in terms of the surrounding anatomical (for open conventional techniques) and radiographic (for fluoroscopy-guided MIS) landmarks.

In this cadaveric study, the optimal entry point for safe S1 pedicle screw placement was mapped on 14 cadavers, with 26 individual entry points, and the measurements were comparable with previous anthropometric studies of the dorsal sacrum²³. Based on anecdotal evidence derived from extensive experience with open surgery and MIS under fluoroscopic guidance, the senior authors had been locating the entry point for percutaneous S1 screw placement approximately 12 mm distal to the base of the “V” sign and within 1 mm of it on the medial-lateral axis. This cadaveric study and retrospective CT-based investigation were performed to validate our concept of using the “V” landmark as a reliable reference for S1 pedicle entry and to quantify the distance between the “V” sign and the optimal S1 pedicle entry point. Per our analysis, the “V” landmark was noted to be a consistent and reliable anatomical structure, unlikely to be altered by L5 spondylolisthesis or prior lumbosacral laminectomies. Interestingly, the nature of these pathological conditions would not affect the borders of the “V” sign. The anatomical entry point to the S1 pedicle was found to be approximately 11.7 mm distal to the base of the “V” landmark (Fig. 1). The difference between the anatomical (11.7 mm) and radiographic (12.9 mm) measurements is likely attributable to the x-ray magnification effect. This magnification effect is a well-recognized physical principle and in our study fell within the range reported in other orthopaedic studies (110% to 125%)²⁴. Although x-ray magnification can vary based on the patient’s body habitus, there is <2% added magnification per 10 units of increased body mass index (BMI)²⁵.

In our retrospective CT cohort, S1 pedicle screw placement using the “V” sign as a reference yielded 100% accuracy for identifying the optimal entry point and 96.7% accuracy for the screw trajectory. Nonetheless, breaches still occurred in 3.3% (9) of the 270 instrumented S1 pedicles. Six of the 9 breaches were minor (<3 mm) and 3 were major. Only 1 screw (0.37%) was clinically symptomatic (S1 radiculopathy) and warranted revision surgery for repositioning, with no long-term sequelae. Our results are reassuring: adopting the “V” sign as a reference for S1 pedicle screw placement yielded 100% accuracy in targeting the optimal entry point, 96.7% accuracy in the screw trajectory, and a low rate (1.1%) of high-grade breaches. Our overall S1 pedicle screw breach rate (3.3%) is lower than the rates in other reports on fluoroscopy-guided percutaneous instrumentation of S1 (5.2% to 20.6%)^6,22,26,27. Most CT scans (70.4%, n = 95) included in this study were performed for non-spine-related problems, leading to the possibility that the noted S1 breaches were mostly asymptomatic and possibly clinically irrelevant. This highlights the reliability of the fluoroscopic “V” landmark as a reference for optimal localization of the S1 pedicle entry point to achieve appropriate S1 screw placement.

The current study has numerous limitations. The number of participating donors at our institution limited the sample size of cadaveric specimens. While no differences were detected in entry points relative to the “V” landmark between sexes, this study may have been underpowered to detect sex-related differences. Additionally, in the cadaveric cohort, the lack of access to CT scans and the small sample size limited our ability to accurately analyze the rate of pedicle screw breaches when the entry point was determined relative to the “V” landmark sign. We sought to overcome this limitation by examining historical CT scan data of patients who underwent S1 pedicle screw instrumentation in our center by the same surgeons who performed the investigation in the cadaveric arm of the study. However, the lack of a control group treated with a “free-hand insertion” technique poses another limitation. Moreover, variations in intraoperative fluoroscopy angles can affect the estimate of the distance between the “V” and S1 pedicle entry points. Also, magnification based on patients’ BMI can occur during fluoroscopy and could therefore have affected the values of the reported measurements. Despite 2 spine specialists conducting independent measurements, human error could still have occurred, contributing to the discrepancy between anatomical and radiographic measurements. Finally, this study only investigated skeletally mature patients and cadavers, and V-S1 insertion measurements with respect to various ages and heights could not be established.

Conclusion

This cadaveric study highlights the “V” landmark as a reliable anatomical and radiographic reference for determining the optimal entry point for S1 pedicle screw insertion. The optimal entry point is anatomically located 11 to 12 mm distal to, and at the vertical line crossing the base of, the “V” landmark. This distance averages 12.9 mm on fluoroscopic images. This reliable “V” landmark provided great accuracy in locating the optimal S1 entry point and overall screw trajectory, with minimal breaches and revision surgery. Furthermore, it has the potential to circumvent the costs and challenges associated with advanced navigation technologies.

Note: The authors thank the Anatomical Gift Program at Boston University and the Department of Radiology at Boston Medical Center for their help and contribution in this study.