Masayoshi ishii, atsunori ohnishi, akira yamagishi, tetsuo ohwada: freehand screw insertion technique without image guidance for the cortical bone trajectory screw in posterior lumbar interbody fusion: what affects screw misplacement: https://thejns.Org/doi/abs/10.3171/2021.Summary: from august 2014 to august 2020, a total of 426 patients underwent single-level plif with cbt screw fixation at our institution.To enable statistical risk analysis of a homogeneous population, we set criteria that reduced the enrolled number of patients by excluding the cases with the following features: 1) patients with spinal surgery site other than the l4¿5 level (n = 153 patients); 2) patients undergoing revision surgery (n = 15 patients); 3) patients with isthmic spondylolisthesis (n = 13 patients); 4) patients who underwent surgery without the use of screws 5.5 mm in diameter (n = 41 patients); and 5) patients whose operation was performed by a left-handed surgeon (n = 22 patients).After the exclusion of these 244 patients, 182 patients were enrolled in the study, comprising 78 men and 104 women with a mean age of 70 years (range 37¿88 years).Cortical bone trajectory (cbt) screw insertion using a freehand technique is considered less feasible than guided techniques, due to the lack of readily identifiable visual landmarks.However, in posterior lumbar interbody fusion (plif), after resection of the posterior anatomy, the pedicles themselves, into which implantation is performed, are palpable from the spinal canal and neural foramen.With the help of pedicle wall probing, the authors have placed cbt screws using a freehand technique without image guidance in plif.This technique has advantages of no radiation exposure and no requirement for expensive devices, but the disadvantage of reduced accuracy in screw placement.To address the problem of symptomatic breaches with this freehand technique, variables related to unacceptable screw positioning and need for revisions were investigated.From 2014 to 2020, 182 of 426 patients with single-level plif were enrolled according to the combined criteria of l4¿5 level, excluding cases of revision and isthmic spondylolisthesis; using screws 5.5 mm in diameter; and operated by right-handed surgeons.We studied the number of misplaced screws found and replaced during initial surgeries.Using multiplanar reconstruction ct postoperatively, 692 screw positions on images were classified using previously reported grading criteria.Details of pedicle breaches requiring revisions were studied.We conducted a statistical analysis of the relationship between unacceptable (perforations > 2 mm) misplacements and four variables: level, laterality, spinal deformity, and experiences of surgeons.A freehand technique can be feasible for cbt screw insertion in plif, balancing the risks of 3.3% unacceptable misplacements and 2.2% revisions with the benefits of no radiation exposure and no need for expensive devices.Pedicle palpation in l4 is the key to safety, even though it requires deeper and more difficult probing.In the initial surgeries and revisions, 75% of revised screws were observed in l4, and unacceptable screw positions were more likely to be found in l4 than in l5.Preoperative planning was conducted based on images from multiplanar ct.Screws 5.5 mm in diameter in both l4 and l5, with lengths of 40¿45 mm in l4 and 35¿40 mm in l5, were routinely selected to yield the maximum purchase and total involvement inside the pedicles.Positional relationships between the surfaces of the pars interarticularis and pedicles, which form screw pathways, were meticulously checked to clarify virtual trajectories.No fluoroscopy was used during surgery.Under general anesthesia, each patient was placed in a prone position on a hall four-point frame.Surgical level was confirmed by lateral radiography of the 18-gauge injection needle buried in the l4 spinous process before skin incision.Operators, who were all right-handed in this study, stood on the left side of the patient and performed all pro cedures from the left side.A midskin incision was made at the l4¿5 level.Soft tissue was dissected minimally and laterally to the inferomedial aspect of l3¿4 facets and the medial aspects of l4¿5 facets, and then l4¿5 bilateral laminotomy with or without total facetectomy was performed in response to pathological conditions.Following disc curettage, two carbon fiber¿reinforced polyetheret herketone cages filled with local bone graft were laterally inserted into the disc space, supplemented by medial grafting of several local bone blocks.According to the previously described technique based on morphometric research,1¿4 an entry hole was created on the intersection point 1¿3 mm medial to the lateral notch of the pars interarticularis and 1 mm caudal to the inferior edge of the transverse process.However, intraoperative adjustment by fluoroscopy or navigation for optimal location is usually required because of individual variations in the relationship between anatomical landmarks and pedicle shape.In our freehand technique without image guidance, the optimal position of the entry hole located on the inferomedial border of the pedicle wall was explored and adjusted by imagining the virtual shape of the pedicles with the help of both medial l4 pedicle wall probing from the spinal canal and inferior l4 pedicle wall probing from the neural foramen.Ball-tip hooks were used to touch the wall.A pilot hole was then drilled on the cortical surface with a 2-mm bar directed 10°¿15° laterally in the axial plane and 20°¿25° cranially in the sagittal plane.The hole was extended to the opposite upper lateral cortical wall of the vertebra using a straight probe, axial helical ct scans 1 mm thick of l4¿5 were performed on an 80-line multislice ct scanner (aquillion one; toshiba), and coronal and sagittal multiplanar reconstruction (mpr)¿ct (mpr-ct) images were created before and after surgery, routinely within a week, with the exception of urgent postoperative cases presenting with neurological deterioration.From the preoperative axial ct, the rotational angle formed between the line bisecting the l4 spinous process and the perpendicular line at the midpoint of bilateral l5 superior articular processes was measured.The cobb angle at l4¿5 on coronal ct was also measured as an indicator of coronal deformity (fig.2b).Spinal deformity was definedas a cobb angle or rotational angle > 5°.From postoperative mpr-ct images, the extent of pedicle perforations by screws was assessed using the grading system of gertzbein et al.And rao et al.(grade 0, no violation; grade 1, <(> <<)> 2 mm; grade 2, 2¿4 mm; grade 3, > 4 mm)5,6 (figs.3a¿d and 4a¿d).The slice with the largest deviation was chosen for grading.Grade 2 and grade 3 were defined as clinically unacceptable screw positions.7 all measurements were performed by one of the authors (m.I.), who was not enrolled as a surgeon in this study due to being left-handed.Four screws (2 in right l4, 1 in left l4, 1 in right l5) were found to be unacceptably misplaced on postoperative radiographs in the operation room.All were immediately re-implanted into correct positions before exiting.The final positions of screws were confirmed later on ct images.These misplaced screws caused no residual symptoms.The numbers of grade 0, 1, 2, and 3 positioned misplaced screws on postoperative ct images were 134, 24, 4, and 2 in the right l4; 127, 27, 8, and 3 in the left l4; 170, 8, 1, and 2 in the right l5; and 170, 9, 3, and 0 in the left l5, respectively.The total rate of unacceptable screws was 3.3% (table 1).Two screws in the right l4, 1 in the left l4, and another in the right l5 pedicle required revision surgeries, because they caused severe leg pain or motor weakness the next day.The rate of symptomatic screws in unacceptable positions was 17% (4/23 screws).All screws were in grade 3 and breached the inferomedial pedicle walls into the spinal canal.All revision surgeries were performed on either the day of or the day after ct examinations.After replacement of screws, symptoms resolved immediately.The revision rate was 2.2% (4/182 surgeries).There were 329 screws implanted into l4 pedicles, 363 into l5 pedicles, 345 on the right side, and 347 on the left side.In cases with spinal deformity (rotation or cobb angle > 5°), patients received 172 screws.Twenty-three cases presented with coronal deformities (10.0° ± 4.1°) only, while 4 cases showed rotational deformities (8.0° ± 1.0°) only.Seventeen cases showed both deformities (cobb angle, 10.8° ± 4.8°; rotation, 7.7° ± 2.1°).Experienced surgeons inserted 468 screws, while inexperienced surgeons inserted 224 screws.Reported events: 1.4 screws were found to be unacceptably misplaced on postoperative radiographs and re-implanted immediately into correct positions.2.23 screws of grade 1 and grade 2 are misplaced.
|