Insufficient evaluation of S2 alar iliac screw malposition with the intraoperative inlet view: utility of the obturator inlet and iliac oblique views

Article information

Asian Spine J. 2025;.asj.2025.0344

Publication date (electronic) : 2025 November 18

doi : https://doi.org/10.31616/asj.2025.0344

Shun Okuwaki, Toru Funayama^*, Yohei Yanagisawa^*, Takahiro Sunami, Takane Nakagawa, Yosuke Ogata, Kotaro Sakashita, Hisanori Gamada, Kousei Miura, Hiroshi Noguchi, Hiroshi Takahashi, Masao Koda

Department of Orthopaedic Surgery, Institute of Medicine, University of Tsukuba, Tsukuba, Japan

Corresponding author: Shun Okuwaki, Department of Orthopaedic Surgery, Faculty of Medicine, University of Tsukuba, 1-1-1, Tennodai, Tsukuba, Ibaraki, 305-8575, Japan, Tel: +81-29-853-3900, Fax: +81-29-853-3904, E-mail: shunokuwaki@tsukuba-seikei.jp

*These authors contributed equally to this work.

Received 2025 June 10; Revised 2025 July 16; Accepted 2025 July 20.

Abstract

Introduction

The S2 alar iliac (S2AI) screw [1,2] is commonly used as a caudal anchor during spinal deformity correction surgery due to its biomechanical strength and minimal soft tissue dissection [3,4]. While it provides strong fixation, its optimal insertion path is limited, and malpositioning can lead to severe neurovascular complications. Anterior deviation may damage the internal iliac artery and its pelvic branches, while caudal deviation in the sciatic notch can injure the sciatic nerve and the superior gluteal artery and nerve [5,6]. These complications can be life-threatening or result in reoperation [7].

Visualizing safe insertion corridors with conventional imaging is challenging due to the complex three-dimensional (3D) sacral anatomy and anatomical variations [8]. The fluoroscopic teardrop view indicates the correct S2AI screw trajectory [9,10]; it is an overlapping image of the anterior and posterior pelvic columns, illustrating the corridor from the posterosuperior iliac spine (PSIS) to the anteroinferior iliac spine (AIIS). When the screw is positioned within this teardrop, it can be considered correctly placed; however, there are instances when this view cannot be confirmed due to limitations of the surgical table, the patient’s body size, or the fluoroscopy angle. Additionally, it is unsuitable for real-time insertion [11].

The inlet and outlet views can serve as alternatives or be used in combination with the teardrop view. These views, obtained at approximately 45° of cephalad tilt and 45° of caudad tilt, respectively [12], can detect anterior and caudal deviations, allowing for safe real-time monitoring of screw insertion [4,13,14].

Recently, the obturator inlet (OI) and iliac oblique (IO) views, commonly used in pelvic fracture surgery [11,15], have been proposed as safer options for S2AI screw placement. They enable accurate, real-time assessment of the screw axis and vertical plane relative to the ilium. However, no studies have compared the safety and accuracy of the OI and IO views with those of the inlet and outlet views for S2AI screw insertion. Furthermore, whether deviations, particularly anterior deviations, can be assessed using only the inlet view remains unexamined.

This study aimed to (1) examine the accuracy of S2AI screw insertion using inlet and outlet views; (2) investigate whether postoperative computed tomography (CT) can detect deviations in the inlet view through postoperative CT reconstruction images and whether deviations can be identified in the OI and IO views; and (3) verify whether anterior deviations can be detected in the OI view when they are not visible in the inlet view using model bones.

Materials and Methods

Study 2: image-based observational study

Study 2-1: using the inlet view to detect cases of anterior deviation

Cases of deviation identified in Study 1 were analyzed using 3D reconstruction software (Synapse Vincent; Fujifilm Medical, Tokyo, Japan) on postoperative CT images. The 3D CT images were converted into synthetic radiographic views.

The inclination angle of the sacral slope relative to the vertical direction of the prone radiograph was defined as the prone SS (Fig. 1A). The preoperative prone SS angle served as an indicator of pelvic tilt, and 3D fluoroscopic images were reconstructed to create an indicator of the tilt angle required for fluoroscopy. In cases where a prone radiograph was taken before surgery, the SS was measured in the prone position. In cases without a preoperative prone radiograph, the prone position SS immediately after surgery was used as a reference.

Fig. 1

Reconstruction for computed tomography (CT). (A, B) When reconstructing the postoperative CT (sacral slope [SS]=36°), the pelvic tilt is reproduced by correcting the cranial–caudal tilt using sacral tilt (SS=27°) obtained from the preoperative or interoperative supine X-ray as an indicator and setting the zero position. (C) A three-dimensional image on an imaginary fluoroscopic frontal view.

During image reconstruction, the SS of the postoperative supine CT was calibrated with the preoperative prone SS, which represents the virtual angle of 0° of frontal fluoroscopy in the intraoperative prone position (Fig. 1B, C). The standard was established at the rotation angle where the S1 spinous process is centered on the sacrum.

The inlet view was defined as a tilting caudal view in which the anterior cortex of the S1 body overlaps with the anterior cortex of the S2 body [21,22].

We measured the required inclination angle and assessed whether deviation could be evaluated using the inlet view in CT-reconstructed images when deviation was present.

Study 2-2: teardrop view in cases of deviation

A “teardrop sign” must be visible to confirm the radiographic view. The outer edges of the “teardrop sign” are formed by the cortical layers of the iliac bone. The teardrop view indicates the direction of screw insertion, which should be placed directly in the center of the teardrop sign [9,10]. We measured the required inclination angle for the teardrop view in CT-reconstructed images in cases of deviation.

Study 2-3: detection of deviation in the OI and IO view

The OI view provides a tangential perspective of both the inner and outer cortices of the ilium at the PSIS–AIIS corridor, enabling visualization of the SI joint and guiding mediolateral screw trajectories. An accurate inlet view should be obtained from the frontal pelvic view to create the OI view. This view is then rotated to generate an axial view of the pelvis on the side being observed. An accurate OI view is characterized by the overlapping outer plates of the ilium [11].

The IO view offers an en face view of the instrumented iliac wing, highlighting the sciatic notch, hip joint, and AIIS. The IO view guides craniocaudal screw placement, which should not aim cranially beyond the superior AIIS. To achieve an optimal IO view, it should be rotated toward the instrumented hemipelvis until the sciatic notch is visible. The AIIS is clearly outlined in the OI view, ensuring that the screw is directed between the proximal and distal AIIS [11]. We measured the required inclination angle for each case and evaluated whether deviation could be assessed using OI and IO views in CT-reconstructed images in cases with deviation.

Results

Study 1: retrospective cross-sectional study

This study included 101 patients with 202 S2AI screws (72 females and 29 males), with a mean age of 70.4±8.5 years (range, 46–85 years). The cohort comprised 48 cases of degenerative scoliosis, 20 cases of kyphosis following osteoporotic vertebral fractures, eight cases of revision after lumbar surgery, eight cases of pyogenic spondylitis, and six cases of lumbar spinal stenosis (Table 1).

Table 1

Patient demographic data

S2AI screw deviation was noted in 12 cases (11.9%) and involved 13 screws (6.4%), with five cases (5.0%) and six screws (3.0%) displaying anterior deviation, and seven cases (6.9%) and seven screws (3.5%) exhibiting posterior deviation.

No significant differences in age, sex, height, or body weight were observed between the groups. SS was greater in the following order: standing, prone, and spinal positions; however, there were no significant differences between the groups. The transitional vertebrae were also assessed for lumbarization and sacralization, with no significant differences found between the groups (Table 1). In cases of deviation, no neurological deficits or visceral or vascular complications were associated with S2AI screw placement.

Study 2: image-based observational study

Study 2-1: using the inlet view to detect cases of anterior deviation

Of the 12 cases and 13 screws with deviations listed in Table 2, we evaluated five cases and six screws exhibiting anterior deviations. Two cases of anterior deviation involved transitional vertebrae showing lumbarization (40.0%). In the other three cases and four screws, anterior deviation was noted in the inlet view for one case. In two cases and two screws, deviation could not be identified in the inlet view (Fig. 2). In all seven cases of posterior deviation, the deviation was undetectable in the inlet view.

Table 2

Deviation patients’ demographic data

Fig. 2

Anterior deviation cases. An accurate inlet view reconstructed from computed tomography images in five cases with anterior deviation. The caudal tilting angle of each case is shown. (A–C) Anterior deviation can be detected (arrow). (D, E) Anterior deviation cannot be detected (arrow dotted line).

For the five cases of anterior deviation, the average inclination angle required for an accurate inlet view, in reference to the supine SS, was 41.7° (range, 27.8°–54.3°). Of the three cases where the accurate inlet view detected anterior deviation, the deviation was difficult to identify in two cases due to an insufficient inlet view angle of 30° (Fig. 3).

Fig. 3

Impact of insufficient inlet view angulation. In three cases where proper inlet views demonstrated anterior deviation (A–C), inadequate inlet views with insufficient caudal angulation of 30° failed to detect the same anterior deviation in two cases (A, B). Dotted arrows indicate missed displacement).

Study 2-2: teardrop view in cases of deviation

In the 12 cases and 13 screws with deviation, the average inclination required for the teardrop view was 13.2°±7.5° (range, 3.1°–27.0°) on the cranial side and 31.5°±4.8° (range, 25.0°–39.3°) for rotation, with maximum values of 27° on the cranial side and 39.3° for rotation (Table 2, Fig. 4). Deviations were detected in all cases.

Fig. 4

Teardrop view on postoperative computed tomography (CT) reconstruction. Postoperative CT-reconstructed teardrop views demonstrated deviation in all cases (arrows indicate areas of displacement).

Study 2-3: detection of deviations in the OI and IO view

The average tilt necessary for the OI view was 23.4°±7.4° caudally (range, 12.1°–33.7°), and 25.1°±6.2° of rotation (range, 14.0°–39.4°), with maximum values of 33.7° caudally and 39.4° of rotation. In all 12 cases of deviation, both anterior and posterior deviations were detected by the OI view. Notably, the OI view was able to identify anterior deviation in two cases where the inlet view did not. A representative case is shown in Fig. 5. The average rotation angle required for the IO view was 36.2°±10.2° (range, 13.0°–59.0°). The IO view successfully detected caudal deviation in all cases with deviation.

Fig. 5

Representative case with anterior deviation. (A) Inlet view on computed tomography (CT) image reconstruction, showing the angle of overlap between S1 and S2 anterior walls (arrowhead). Anterior deviation of the right S2 alar iliac (S2AI) screw is not detected (arrow). (B) Obturator inlet view on CT reconstruction. The inlet view was rotated to the angle of overlap of the iliac outer cortex (arrowhead). Anterior deviation of the right S2AI screw can be detected (arrow). (C) Iliac oblique view on CT reconstruction. The right S2AI screw was evaluated without caudal deviation. (D) Three-dimensional CT image. The right S2AI screw deviated anteriorly from the sacroiliac joint (arrow).

Study 3: experimental replication using model bones

In Study 2, we identified two cases of anterior deviation that could not be evaluated using the inlet view but were assessable through the OI view, along with two additional cases that could be evaluated using the OI view. These four cases exhibited anterior displacement above the arcuate line of the pelvic brim, originating from the sacral alar region. Consequently, we reproduced the same level of deviation using a model bone.

We utilized an 8.5×90 mm screw to achieve deviation of the SI joint above the arcuate line in all four cases. The deviation was assessed using fluoroscopy, with the fluoroscope tilted 45° caudally relative to the SS angle of 30° in the prone position. While the accurate inlet view did not allow for the evaluation of anterior deviation, the OI view revealed that the screw was indeed anteriorly deviated (Fig. 6).

Fig. 6

Anterior deviation cases in model bone. (A) The accurate inlet view did not allow the anterior deviation to be evaluated (arrow dotted line). (B) The screw deviates anteriorly (arrow) when rotated from the obturator inlet view. (C) View of the pelvic cavity from 30° at the cranial side, (D) Frontal pelvic view, (E) View from 30° at the caudal side.

Discussion

Previous studies have reported S2AI screw deviation rates of 1.5%–2.7% using O-arm navigation [23,24], 3.7% with 3D-printed guides [25], and 5.1%–15% with fluoroscopy alone [27,28]. In the present study, we found a deviation rate of 6.4% using fluoroscopy alone, which was confirmed in the frontal, inlet, and outlet views, as well as in the teardrop view, where possible. To date, no clinical study has accurately reported the incidence of neurovascular injuries resulting from S2AI screw malpositioning, with only one case of L5 nerve root injury documented in the literature [7]. Although anterior deviations are theoretically associated with the risk of injury to the internal iliac vessels or pelvic nerves, no such complications were observed in our study. Careful fluoroscopic evaluation is essential to prevent potentially serious complications.

The inlet view is crucial for assessing pelvic morphology, with a recommended caudal tilt angle of 23°–33° in the general population [22,28]. In this study, the average required tilt angle was 41.7°, surpassing the tilt angle reported in previous studies. This discrepancy may be attributed to the inclusion of patients with spinal deformities and posterior pelvic tilt, unlike previous studies that focused on healthy individuals [29,30]. Additionally, the prone SS is smaller than the supine SS, which further increases the required tilt angle.

The inlet view is commonly used to assess pelvic morphology and guide S2AI screw placement, particularly for detecting pelvic cavity protrusion [22,28]. In this study, the inlet view identified anterior deviation in five cases but failed in two, missing two additional cases of anterior deviation due to an insufficient tilt angle. This suggests that the inlet view is inadequate for evaluating anterior deviation, highlighting the need to revise current evaluation methods for S2AI screw placement.

Although anterior deviation can sometimes be detected in the inlet view, it may be overlooked when the deviation lies above the pelvic arcuate line, which serves as an anatomical boundary separating the lesser and greater pelvis with a variable slope (pelvic brim). Since the inlet view is observed from the caudal side, deviations above the arcuate line may cause the screw to appear intraosseous due to the slope difference. With the correct insertion points of the S2AI screw, a deviation to this superior position may occur if the angulation to the caudal side is too small.

S2AI screws may deviate from the sacral alar by skiving against the hard cortex of the SI joint [23]. One contributing factor is an insufficient caudal insertion angle. While a caudal tilt of 20°–29° is generally recommended [18], this angle must be increased when the SS in the prone position is large. An inadequate caudal angle for screw insertion raises the risk of anterior deviation above the arcuate line. In this series, the average prone SS was 27° (range, 21°–30°) across four cases, exceeding the overall average of 21° and indicating an insufficient tilt angle of the screw to the caudal side. Evaluating anterior deviation in the inlet view was challenging due to the larger-than-average angle detected in these cases.

Another contributing factor is the presence of a transitional vertebra. In cases where S1 becomes lumbarized, determining the entry point of the S2AI screw using the foramen as a landmark becomes difficult. If the entry point is more cranial than usual, the deviation will be superior, increasing the likelihood of deviation above the arcuate line (Figs. 5D, 6C–E). Compared to the 10 cases (9.9%) where S1 was lumbarized, anterior deviation was more frequent in these instances, further contributing to deviations above the arcuate line.

This study had several limitations. First, the number of cases in which the teardrop view could be evaluated was not recorded, preventing an estimation of the deviation rate based solely on the inlet view. Nonetheless, this is the first study to report that the inlet view is inadequate for assessing anterior deviation, and surgeons should be made aware of this issue. Second, it was not possible to verify findings through actual operations, and the impact of body size and subcutaneous fat thickness could not be assessed. Third, due to the small sample size, we could not determine whether all anterior and posterior deviations could be evaluated using the OI view. Additionally, it is important to note that the OI view is not a perfect tool; it cannot detect inferior deviations due to its inherent limitations. Therefore, it is essential to confirm screw evaluation using two or more axial views, as inferior deviations can be identified using the IO view.

Conclusions

The inlet view was an inadequate imaging method for evaluating the anterior deviation of S2AI screws. All cases in this series were assessed using the OI view. Specifically, the inlet view was unsuitable for evaluating screws that deviated above the arcuate line of the pelvic brim. When assessing S2AI screws using fluoroscopy alone, both the OI and IO views proved useful for evaluating deviations.

Article information Continued