Pedicle screw fixation has been widely recognized as the most reliable anchor for the spinal fusion procedure used with various kinds of spinal pathologies. During this decade, the cortical bone trajectory (CBT) technique has been gaining popularity as an alternative to the traditional trajectory (TT) technique for lumbar pedicle screw insertion.1,2 Contrary to the TT, which engages more cancellous bone than cortical bone along the anatomical axis of the pedicle, CBT follows a craniolaterally directed path through the pedicle and can maximize contact with denser bone within the vertebra. Because cortical bone is less affected by osteoporotic change than cancellous bone, CBT achieves rigid screw fixation even in osteoporotic vertebrae, becoming a promising surgical strategy for spinal disease treatment in an aging society. In addition, the divergent trajectory of CBT allows for minimum invasiveness for screw insertion, such as reduced muscle dissection and retraction,3 preservation of the neurovascular supply to the fused segment, and avoidance of iatrogenic injury to the superior adjacent facet joints.4 These factors have been reported to contribute to reducing the development of adjacent segment disease, which is an inevitable long-term complication in spinal fusion.5
The main advantage of the CBT technique is that one limited midline incision allows for the entire surgical procedure, including decompression, interbody fusion, and screw placement, and there are numerous clinical reports demonstrating the minimally invasive nature of this technique.6,7 However, controversy exists regarding the achievement of successful bone fusion with the use of CBT, with some reports of lower fusion rates.8 Biomechanical drawbacks of CBT with a short and divergent trajectory, such as improper load distribution in the vertebral body and inferior resistance against torsional motion, are considered to be associated with this phenomenon.2,9
In recent years, “long CBT” directed more anteriorly in the vertebral body has been recommended as a countermeasure.10–12 Unlike the original CBT, which uses screws of 25–30 mm in length, long CBT, which uses screws of 40–45 mm in length, extends deeper into the vertebral column and improves screw fixation (Fig. 1). Long CBT is expected to facilitate both maximum cortical purchase and effective load sharing within the vertebra, thus promoting bone fusion and reducing screw loosening. However, to the best of our knowledge, there has been no report on the clinical significance of the screw length and screw insertion depth using the CBT technique. It is also not clear what screw length and depth are desirable for achieving successful bone fusion, given that a longer trajectory could increase a risk of cortical breach. The aim of the present study was to investigate the influence of the screw insertion depth in the vertebra on lumbar spinal fusion.
The original and long CBT techniques. Screw insertion by the original CBT using screws of 25–30 mm in length (left), and long CBT using screws of 40–45 mm in length (right).
Methods
This study was a retrospective radiological evaluation of prospectively collected patients who underwent spinal fusion using the CBT technique approved by our institutional ethics committee (no. 12-10). The study participants consisted of 101 consecutive patients with L4 degenerative spondylolisthesis who underwent single-level posterior lumbar interbody fusion (PLIF) at L4–5 performed by four board-certificated spine surgeons in our hospital during the period from January 2015 to September 2018. One patient with rheumatoid arthritis, one with a neuromuscular disorder, one who required reoperation due to screw misplacement, and one with postoperative infection were excluded from the study. Of the 97 remaining patients, 96 who were followed up for more than 2 years after surgery were enrolled (follow-up rate 99%, mean follow-up 32.9 ± 10.8 months). There were 41 male and 55 female patients, with a mean age of 68.6 ± 8.7 years (range 45–84 years).
Surgical Procedure
Through a midline skin incision (approximately 5 cm), the paraspinal muscles were dissected to expose the lateral borders of the pars interarticularis. During the procedure, intraoperative fluoroscopy was used to confirm the correctness of the entry point position and screw path of CBT. According to the previously described method,10 the trajectory started from the pars interarticularis, passed the inferior border of the pedicle, and ended in a more anterior region in the vertebral body, compared with the original CBT. The entry point was created in a position that avoided violation of an unfused adjacent facet joint, and the standard trajectory was directed 20°–25° cranially in the sagittal plane and 10° laterally in the axial plane. After screw path creation, posterior decompression with facetectomy, interbody procedures, and insertion of two interbody cages were performed before screw insertion because the medial position of the screw heads would hinder these processes. Last, the screw path was tapped to the size of the planned screw to prevent cortical fissures during screw insertion. The standard screw size was 5.5–6 mm in diameter and 35–45 mm in length.10 The screw tip was placed close to the superior endplate without bicortical penetration. In all cases, only local bone was used for bone grafting. In all patients a lumbosacral orthosis was applied for 3 months postoperatively.
Radiological Evaluation
Screw loosening and bone fusion were assessed by radiographs or CT at the last follow-up. CT was performed at 6 months, 1 year, and 2 years postoperatively and every year thereafter until adequate fusion was confirmed. Screw loosening was defined as a radiolucent zone greater than 1 mm around the screw. Bone fusion was defined as bone bridge formation between the vertebral bodies and absence of the following signs of nonfusion: motion of the fused segment exceeding 3°, clear zone around the intervertebral cages, and screw loosening.13 These evaluations were performed twice by a single observer, and to evaluate the reproducibility of the findings 30 of 96 randomly selected patients were also assessed by two independent observers. The following patient factors that may contribute to the incidence of screw loosening and bone fusion were analyzed: 1) age, 2) sex, 3) body mass index (BMI), 4) bone mineral density (BMD), 5) intervertebral mobility, 6) screw diameter, 7) screw length, 8) depth of the screw in the vertebral body (%depth), 9) facetectomy (partial or total), 10) crosslink connector (with or without), and 11) cage material (titanium or titanium-coated polyether ether ketone). BMD was estimated as Hounsfield units (HU) of the vertebral body by setting on an oval-shaped range inside the cortical shell on axial slices of preoperative CT according to the method of Schreiber et al.14 HU from L4 and L5 vertebrae were averaged to determine HU for each patient. All CT examinations were performed using a helical 64-channel CT scanner (Aquilion CXL; Toshiba Medical), which was calibrated weekly, with a slice thickness of 0.5 mm and tube voltage of 120 kV, under automatic exposure control to determine HU values accurately without the use of phantoms. The intervertebral mobility was defined as sagittal translation (%) and angulation (°) of the L4–5 segment on flexion and extension radiographs. The %depth was defined as the ratio of the screw length within the vertebral body to the anteroposterior length of the vertebral body in the axial plane using a postoperative CT image obtained immediately after surgery (Fig. 2).12 In cases with lateral cortical breach outside the vertebral body, the %depth was calculated without including the screw length outside the vertebral body. The %depth values of the four L4 and L5 screws per case were averaged to give a mean %depth for each patient.
Evaluation of screw insertion depth. The %depth was defined as the ratio of the screw length within the vertebral body (A) to the anteroposterior length of the vertebral body (B) in the axial plane.
Statistical Analysis
All results are shown as the mean ± standard deviation. The kappa coefficient was used for the assessment of screw loosening and bone fusion to evaluate intra- and interobserver reliability. After testing for the normal distribution of continuous variables using Shapiro-Wilk tests, we used nonparametric analysis (Mann-Whitney U-test) to analyze the differences between groups. The chi-square test and Fisher exact tests were used for nominal variables. Multivariate logistic regression analysis was performed using variables with p < 0.20 in univariate analysis to determine the independent factors predicting screw loosening and bone fusion. Then, receiver operating characteristic (ROC) curve analysis was performed and the area under the curve (AUC) was assessed to determine the best cutoff value of the identified factor for predicting radiological outcomes. The sensitivity and specificity of the best cutoff value were calculated. IBM SPSS Statistics version 26.0 (IBM Corp.) was used for all analyses, with the level of significance set as p < 0.05.
Results
Cortical breach was observed in 9 of 384 screws (2.3%), while there was no incidence of neurovascular injuries using this technique. Screw loosening was observed in 3.1% of patients (3 of 96) and bone fusion was achieved in 91.7% of patients (88 of 96) at the last follow-up. There were no cases requiring revision surgery due to screw loosening or pseudoarthrosis. The assessment of screw loosening (intraobserver reliability 0.79, interobserver reliability 0.65) and bone fusion (intraobserver reliability 0.78, interobserver reliability 0.71) showed good agreement. There was no significant difference in the parameters between groups with [loosening (+)] and without [loosening (−)] loosening (p > 0.20) (Table 1). In a comparison of the groups with [fusion (+)] and without [fusion (−)] bone fusion, there were no significant differences in age, sex, BMD, or intervertebral mobility (Table 2). BMI in the fusion (+) group was significantly higher than that in the fusion (−) group (24.1 ± 3.3 vs 26.0 ± 3.4 kg/m2, respectively; p = 0.036). The screw diameter and length in the fusion (+) group were larger and longer, respectively, than those in the fusion (−) group, although the differences were not significant (screw diameter 5.9 ± 0.2 vs 5.8 ± 0.3 mm, p = 0.44; screw length 39.3 ± 2.5 vs 37.3 ± 4.0 mm, p = 0.21, respectively). Also, the %depth was significantly greater in the fusion (+) group than in the fusion (−) group (50.3% ± 8.2% vs 37.0% ± 9.5%, respectively; p = 0.001). Other parameters, such as facetectomy, crosslink connector, and cage material, were not different between the two groups. Multivariate logistic regression analysis using variables with p < 0.20 in univariate analysis revealed that %depth (OR 0.82, 95% CI 0.73–0.93, p < 0.001, R2 = 0.303) was a significant independent predictor of bone fusion (Table 3). ROC curve analysis identified %depth > 39.2% as a predictor of bone fusion (p = 0.001, AUC 0.851, 95% CI 0.711–0.991, sensitivity 90.9%, specificity 75.0%) (Fig. 3). The bone fusion rate in patients with %depth > 39.2% was 97.5% (78 of 80 patients), whereas it was 62.5% (10 of 16 patients) in patients with %depth ≤ 39.2% (Fig. 4).
Univariate analysis with screw loosening
| Loosening | p Value | ||
|---|---|---|---|
| (+) | (−) | ||
| No. of patients | 3 | 93 | |
| Age (yrs) | 73.7 ± 5.7 | 68.4 ± 8.7 | 0.27 |
| Sex (M/F) | 1:2 | 40:53 | > 0.99 |
| BMI (kg/m2) | 24.9 ± 3.7 | 24.3 ± 3.4 | 0.75 |
| BMD (HU) | 139.5 ± 11.1 | 138.7 ± 51.2 | 0.77 |
| Intervertebral mobility | |||
| Translation (%) | 7.7 ± 2.3 | 7.8 ± 4.1 | 0.83 |
| Angulation (°) | 8.4 ± 4.8 | 8.0 ± 4.3 | 0.25 |
| Screw diameter (mm) | 5.8 ± 0.3 | 5.9 ± 0.2 | 0.33 |
| Screw length (mm) | 37.5 ± 5.0 | 39.2 ± 2.6 | 0.52 |
| %depth (%) | 44.2 ± 4.5 | 49.4 ± 9.1 | 0.26 |
| Facetectomy (partial/total) | 0:3 | 22:71 | > 0.99 |
| Crosslink connector (w/ vs w/o) | 0:3 | 41:52 | 0.26 |
| Cage material (T/TP) | 2:1 | 74:19 | 0.51 |
T = titanium; TP = titanium-coated polyether ether ketone.
Univariate analysis with bone fusion
| Fusion | p Value | ||
|---|---|---|---|
| (+) | (−) | ||
| No. of patients | 88 | 8 | |
| Age (yrs) | 68.6 ± 8.9 | 68.4 ± 5.0 | 0.80 |
| Sex (M/F) | 36:52 | 5:3 | 0.28 |
| BMI (kg/m2) | 24.1 ± 3.3 | 26.0 ± 3.4 | 0.036 |
| BMD (HU) | 138.5 ± 52.2 | 141.8 ± 24.5 | 0.59 |
| Intervertebral mobility | |||
| Translation (%) | 7.8 ± 4.0 | 7.4 ± 5.2 | 0.64 |
| Angulation (°) | 8.0 ± 4.2 | 8.2 ± 4.7 | 0.72 |
| Screw diameter (mm) | 5.9 ± 0.2 | 5.8 ± 0.3 | 0.44 |
| Screw length (mm) | 39.3 ± 2.5 | 37.3 ± 4.0 | 0.21 |
| %depth (%) | 50.3 ± 8.2 | 37.0 ± 9.5 | 0.001 |
| Facetectomy (partial/total) | 20:68 | 2:6 | > 0.99 |
| Crosslink connector (w/ vs w/o) | 38:50 | 3:5 | > 0.99 |
| Cage material (T/TP) | 70:18 | 6:2 | 0.67 |
Multivariate logistic regression analysis with bone fusion
| p Value | OR | 95% CI | |
|---|---|---|---|
| BMI | 0.15 | 1.41 | 0.98–1.85 |
| %depth | <0.001 | 0.82 | 0.73–0.93 |
The ROC of %depth for predicting bone fusion. The AUC was 0.851.
Illustrative cases of patients in the fusion (+) and (−) groups. A and B: Images obtained from a 71-year-old female patient. Screw insertion %depth of 61.0% with bone fusion. C and D: Images obtained from a 67-year-old male patient. Screw insertion %depth of 33.1% with a clear zone around the intervertebral cages.
Discussion
The goal of instrumentation with the pedicle screw system is to stabilize the spinal segments and achieve bone fusion. In this study, we investigated factors contributing to the radiological outcome of PLIF for degenerative lumbar spondylolisthesis using the CBT technique, with a focus on the depth of screw insertion. Multivariate regression analysis revealed that the screw insertion depth in the vertebra contributed to bone fusion, with a cutoff value of approximately 40% screw depth in the vertebral body.
With respect to the effect of the depth of pedicle screw insertion on screw fixation, some biomechanical studies reported that the pullout strength using TT tended to increase with longer screws, but only improved when the screws were placed deep enough to penetrate the anterior vertebral cortex.15,16 In contrast, longer screws following CBT significantly improved the fixation even if the screws did not engage the anterior cortex.12 This difference could be theoretically explained by the variations in bone density in the vertebral body. While TT goes to the central portion of the vertebral body with lower-density bone, CBT goes to the peripheral portions with higher-density bone, such as the superior vertebral endplate and lateral vertebral wall. A finite element analysis revealed the biomechanical superiority of a longer trajectory, whereby long CBT using 40-mm-long screws achieved a 35% higher pullout strength, 20% higher resistance against flexion-extension motion, and 36% higher resistance against rotational motion than the original CBT using 25-mm-long screws.12 However, what is particularly interesting in the present study is that the screw insertion depth in the vertebral body (%depth), not the actual screw length, was an independent factor contributing to bone fusion. This result may be associated with individual anatomical differences in the anteroposterior length of the vertebral body; that is, %depth varied with the anteroposterior vertebral length, even if the screw length within the vertebral body was the same. Deeper screw insertion in the vertebral body using longer screws could achieve both enhanced anchoring and better load sharing within the vertebrae, leading to successful bone fusion.
Osteoporosis has a marked impact on both bone fusion and screw fixation due to negative bone remodeling and the fragile characteristics of the bone.17,18 In terms of BMD evaluation, dual-energy x-ray absorptiometry is a 2D assessment method for which the accuracy could easily be impaired by lumbar degenerative changes, such as osteophyte formation and facet hypertrophy.19 On the contrary, HU measurement allows 3D assessment and provides site-specific bone information using data from CT conducted prior to instrumentation surgery.14 In the present study, however, contrary to our expectations, BMD assessed using HU was not identified as a significant factor contributing to bone fusion. One reason for the reduced effect of BMD was the lack of assessing the bone fusion process and early fusion status. We evaluated the bone fusion status only at the last follow-up and did not investigate the time to achieve fusion, which may have led to different conclusions. Another reason may be that CBT maximized contact with cortical bone, which ensured screw fixation regardless of osteoporotic changes.2
Some biomechanical studies have shown that CBT led to inferior rotational stability compared with TT because of its divergent spinal construct and the short length of screws.2,9 As technical countermeasures against torsional motion for the screw-rod construct, preservation of the facet joint and the addition of a crosslink connector were recommended.2 Nevertheless, these factors were not identified as contributors to bone fusion in the present study. Because a longer trajectory directed more anteriorly into the vertebral body than the original CBT was adopted in this case series and improved rotational stability,12 the biomechanical effects of the facet joint and crosslink connector were considered to become weak and nonsignificant.
The findings of the present study have practical implications because there has been no consensus on the optimal screw trajectory for achieving successful radiological outcomes. Lee et al. performed a prospective randomized trial of CBT and TT in single-level PLIF for lumbar degenerative disease.6 The study by Lee et al. showed comparable fusion rates at 2 years postoperatively (CBT vs TT: 94.3% vs 94.5%, p > 0.99). Similarly, Sakaura et al. conducted a retrospective study to compare CBT and TT in single-level PLIF for degenerative lumbar spondylolisthesis over a 3-year postoperative follow-up. The fusion rate using CBT was lower than that using TT, although the difference was not significant (CBT vs TT: 88.4% vs 96.3%, p = 0.052).8 Another study showed satisfactory outcomes when CBT was compared with TT, such as regarding the fusion rate (100% vs 100%, respectively), loss of correction (2.6% vs 4.2%, respectively), and screw loosening (4.8% vs 7.1%, respectively).20 The differences in the literature regarding the fusion rates using CBT compared with TT may be due to the fact that the details of the screw path using CBT vary among surgeons. The present study is the first to elucidate optimal CBT based on radiological outcomes, and a bone fusion rate of 97.5% with screw insertion at a depth greater than 40% in the vertebral body is comparable to results reported for previous studies.
There are some limitations that should be acknowledged in this study. First, the long CBT is technically demanding. Because of the novelty of the technique and the narrow bone corridors to achieve both cortical bone contact and a deeper trajectory into the vertebral column, this technique necessitates high-level surgical skill and intraoperative fluoroscopy support to enhance accuracy. Some innovative methods, including 3D navigation systems, robotic guidance,21 and the patient-specific template guide system,22 have recently been introduced to improve screw placement accuracy and reduce intraoperative radiation exposure. When surgeons attempt a longer trajectory, the trajectory may become close to the anatomical pedicular axis in the sagittal plane, i.e., a less steep trajectory. However, this trajectory is undesirable for the following two reasons: 1) loss of appropriate contact with cortical bone concentrated between the pars interarticularis and inferior part of the pedicle,10 and 2) risk of cranial facet violation due to a more cranial entry point of the trajectory.4 We recommend that the ideal trajectory should start from the pars interarticularis, passing the inferior border of the pedicle, and end around the halfway point of the vertebral endplate. Second, this study included only patients with L4 degenerative spondylolisthesis who underwent L4–5 PLIF. Although this is the condition for which the PLIF procedure is most frequently performed, different factors may be involved in radiological outcomes depending on spinal levels and pathologies. Last, we investigated only radiological outcomes. Further clinical studies are needed to demonstrate the clinical superiority of long CBT.
Conclusions
To the best of our knowledge, this is the first study to demonstrate the significance of the screw insertion depth using the CBT technique. The cutoff value of the screw insertion depth for achieving bone fusion was 39.2% in the vertebral body.
Disclosures
Keitaro Matsukawa is a consultant for Medacta.
Author Contributions
Conception and design: Matsukawa. Acquisition of data: Matsukawa, Yanai. Analysis and interpretation of data: Matsukawa. Drafting the article: Matsukawa. Critically revising the article: Matsukawa. Reviewed submitted version of manuscript: Yato. Approved the final version of the manuscript on behalf of all authors: Matsukawa. Statistical analysis: Matsukawa, Yanai. Administrative/technical/material support: Yanai, Fujiyoshi, Kato. Study supervision: Yato.
References
- 1↑
Santoni BG, Hynes RA, McGilvray KC, et al. Cortical bone trajectory for lumbar pedicle screws. Spine J. 2009;9(5):366–373.
- 2↑
Matsukawa K, Yato Y, Imabayashi H, et al. Biomechanical evaluation of the fixation strength of lumbar pedicle screws using cortical bone trajectory: a finite element study. J Neurosurg Spine. 2015;23(4):471–478.
- 3↑
Hung CW, Wu MF, Hong RT, et al. Comparison of multifidus muscle atrophy after posterior lumbar interbody fusion with conventional and cortical bone trajectory. Clin Neurol Neurosurg. 2016;145:41–45.
- 4↑
Matsukawa K, Kato T, Yato Y, et al. Incidence and risk factors of adjacent cranial facet joint violation following pedicle screw insertion using cortical bone trajectory technique. Spine (Phila Pa 1976). 2016;41(14):E851–E856.
- 5↑
Sakaura H, Ikegami D, Fujimori T, et al. Early cephalad adjacent segment degeneration after posterior lumbar interbody fusion: a comparative study between cortical bone trajectory screw fixation and traditional trajectory screw fixation. J Neurosurg Spine. 2019;32(2):155–159.
- 6↑
Lee GW, Son JH, Ahn MW, et al. The comparison of pedicle screw and cortical screw in posterior lumbar interbody fusion: a prospective randomized noninferiority trial. Spine J. 2015;15(7):1519–1526.
- 7↑
Takenaka S, Mukai Y, Tateishi K, et al. Clinical outcomes after posterior lumbar interbody fusion: comparison of cortical bone trajectory and conventional pedicle screw insertion. Clin Spine Surg. 2017;30(10):E1411–E1418.
- 8↑
Sakaura H, Miwa T, Yamashita T, et al. Posterior lumbar interbody fusion with cortical bone trajectory screw fixation versus posterior lumbar interbody fusion using traditional pedicle screw fixation for degenerative lumbar spondylolisthesis: a comparative study. J Neurosurg Spine. 2016;25(5):591–595.
- 9↑
Perez-Orribo L, Kalb S, Reyes PM, et al. Biomechanics of lumbar cortical screw-rod fixation versus pedicle screw-rod fixation with and without interbody support. Spine (Phila Pa 1976). 2013;38(8):635–641.
- 10↑
Matsukawa K, Taguchi E, Yato Y, et al. Evaluation of the fixation strength of pedicle screws using cortical bone trajectory: what is the ideal trajectory for optimal fixation?. Spine (Phila Pa 1976). 2015;40(15):E873–E878.
- 11
McKinley TO, McLain RF, Yerby SA, et al. Characteristics of pedicle screw loading. Effect of surgical technique on intravertebral and intrapedicular bending moments. Spine (Phila Pa 1976). 1999;24(1):18–25.
- 12↑
Matsukawa K, Yato Y, Imabayashi H, et al. Biomechanical evaluation of fixation strength among different sizes of pedicle screws using the cortical bone trajectory: what is the ideal screw size for optimal fixation?. Acta Neurochir (Wien). 2016;158(3):465–471.
- 13↑
Ito Z, Matsuyama Y, Sakai Y, et al. Bone union rate with autologous iliac bone versus local bone graft in posterior lumbar interbody fusion. Spine (Phila Pa 1976). 2010;35(21):E1101–E1105.
- 14↑
Schreiber JJ, Anderson PA, Rosas HG, et al. Hounsfield units for assessing bone mineral density and strength: a tool for osteoporosis management. J Bone Joint Surg Am. 2011;93(11):1057–1063.
- 15↑
Zindrick MR, Wiltse LL, Widell EH, et al. A biomechanical study of intrapeduncular screw fixation in the lumbosacral spine. Clin Orthop Relat Res. 1986;(203):99–112.
- 16↑
Karami KJ, Buckenmeyer LE, Kiapour AM, et al. Biomechanical evaluation of the pedicle screw insertion depth effect on screw stability under cyclic loading and subsequent pullout. J Spinal Disord Tech. 2015;28(3):E133–E139.
- 17↑
Park SB, Yang HJ, Kim CH, Chung CK. Difference in spinal fusion process in osteopenic and nonosteopenic living rat models using serial microcomputed tomography. J Korean Neurosurg Soc. 2017;60(3):348–354.
- 18↑
Weiser L, Huber G, Sellenschloh K, et al. Insufficient stability of pedicle screws in osteoporotic vertebrae: biomechanical correlation of bone mineral density and pedicle screw fixation strength. Eur Spine J. 2017;26(11):2891–2897.
- 19↑
Bühler DW, Berlemann U, Oxland TR, Nolte LP. Moments and forces during pedicle screw insertion. In vitro and in vivo measurements. Spine (Phila Pa 1976). 1998;23(11):1220–1228.
- 20↑
Elmekaty M, Kotani Y, Mehy EE, et al. Clinical and radiological comparison between three different minimally invasive surgical fusion techniques for single-level lumbar isthmic and degenerative spondylolisthesis: minimally invasive surgical posterolateral fusion versus minimally invasive surgical transforaminal lumbar interbody fusion versus midline lumbar fusion. Asian Spine J. 2018;12(5):870–879.
- 21↑
Buza JA III, Good CR, Lehman RA Jr, et al. Robotic-assisted cortical bone trajectory (CBT) screws using the Mazor X Stealth Edition (MXSE) system: workflow and technical tips for safe and efficient use. J Robot Surg. 2021;15(1):13–23.
- 22↑
Kaito T, Matsukawa K, Abe Y, et al. Cortical pedicle screw placement in lumbar spinal surgery with a patient-matched targeting guide: a cadaveric study. J Orthop Sci. 2018;23(6):865–869.
