November 07, 2017
8 min read

A 16-year-old female patient with progressively worsening back pain and deformity

You've successfully added to your alerts. You will receive an email when new content is published.

Click Here to Manage Email Alerts

We were unable to process your request. Please try again later. If you continue to have this issue please contact

A 16-year-old female high school soccer player with no known medical comorbidities (2 years post-menarche) presented to the clinic for evaluation of diffuse back pain. She reported being diagnosed with adolescent idiopathic scoliosis at age 14 and placed in a brace. For the two years prior to presentation, she was largely non-compliant with the bracing regimen. She noted wearing the brace for “several hours a day” over the past few months.

On physical examination, the patient was 5 feet 8 inches tall and weighed 123 pounds. She had mild shoulder asymmetry with the left shoulder sitting slightly lower than the right. Her pelvis was level and symmetric. She had a normal gait pattern. There was no tenderness to palpation about the spinous processes. There were no limitations with lumbar flexion, extension or lateral bending to either side. An Adam’s forward bend test was positive for a right-sided rotational thoracolumbar prominence. She exhibited 5/5 strength throughout the bilateral upper and lower extremities with 2+ symmetric patellar and Achilles reflexes. There was a down-going Babinski and negative Hoffman’s sign bilaterally. A negative straight leg raise was observed bilaterally.

The patient had standing anteroposterior (AP) and lateral scoliosis films as well as left and right bending films. The radiographs showed a right thoracic curve from T4-T10 measuring 45° and a left lumbar curve from T10-L3 measuring 55° (Figure 1). The bending radiographs showed a flexible thoracic curve (corrected to 23.1°) and lumbar curve (corrected to 3.3°) (Figure 2). There was generalized hypokyphosis and flattening of the lumbar lordosis on the lateral radiographs. Her sagittal alignment was otherwise well-maintained with sagittal vertical axis (SVA) of -7.75 mm. Given the progression of the thoracic and thoracolumbar (TL) curves in the setting of worsening back pain and recent brace wear, the recommendation was made to perform posterior spinal fusion from T4-L3 to prevent further curve progression. The decision to proceed with surgery was made following extensive discussion with the patient and family members.

Preoperative AP and lateral plain radiographs reveal a right thoracic curve
Figure 1. Preoperative AP (a) and lateral (b) plain radiographs reveal a right thoracic curve from T4-T10 that measures 45.7° and a left lumbar curve from T10-L3 that measures 54.8°, as well as TK.
Figure 2. Lateral bending radiographs reveal correction of the lumbar scoliosis and partial correction of the thoracic and lumbar scoliosis.

Source: Howard S. An, MD

What is your diagnosis?

See answer on next page.


Adolescent idiopathic scoliosis

Preoperative planning

Preoperative radiographic measurements were performed using Surgimap Spine version 2.05 or later (Nemaris Inc.). Quantitative assessments included curve magnitudes of the structural curve (Cobb angles), sagittal curves, thoracic and lumbar modifiers (Lenke classification for adolescent idiopathic scoliosis or AIS), coronal offset, number of levels to be fused, neutral/stable vertebrae, apical vertebra of the main thoracic (MT) curve, SVA, cervical lordosis (CL), thoracic kyphosis T4-12 (TK), lumbar lordosis L1-S1 (LL), sacral slope (SS), pelvic tilt (PT), pelvic incidence (PI) and the best fit radius curve in the sagittal plane encompassing the MT and TL curves (Figure 3). Rotation of the apical vertebra was also assessed using the Perdriolle method. On an AP radiograph, the edges of a nomogram were directly aligned with the innermost points of the lateral walls of the vertebral body, with a rotation angle being read from a vertical line drawn through the convex pedicle. The patient subsequently underwent supine push-prone in the coronal plane to determine the flexibility of each curve. Subsequently, quantitative comparison of the axial rotation of the apical vertebra between the standing AP radiograph and the push-prone radiograph was assessed to determine a flexibility index. Based on the rigidity of the spinal deformity determined from the supine push-prone and lateral bending radiographs, a concave cobalt chromium (CoCr) 6.0-mm rod was selected (for the patient’s right side); specialized rod designs can be used for more rigid deformities. The convex rod or rail was then determined by “stepping down” to a less stiff rod/rail, ie, one made of titanium (Ti; the patient’s left side). The rod selections were then cut to length. Next, the CoCr concave rod/rail was “over-bent” to approximately 20° greater than the resultant goal TK (67°) as determined by using the patient’s own spinopelvic measurements (Figure 4). This was performed, as described by Cidambi and colleagues, to allow for the rod flattening that is expected to occur during correction based on the rigidity of the deformity, rod/rail geometry and the metal component (Ti vs. CoCr). Target kyphosis was determined mathematically by initially determining the PI.

Preoperative planning with Surgimap Spine
Figure 3. Preoperative planning with Surgimap Spine using AP (a) and lateral (b) standing full-length plain radiographs.

Pertinent surgical technique

Templating of the CoCr and Ti rods
Figure 4. Templating of the CoCr and Ti rods includes measuring the length and approximate curvature of the rods.

Exposure to the posterior spine was performed. Spinous processes from T4-L3 were identified and pedicle screws were placed with intraoperative neuromonitoring guidance. Initially, the 5.5-mm Ti rod was placed on the convex side of the thoracolumbar deformity in the standard fashion, obtaining partial correction of the coronal deformity. Cricket reducers (K2M Inc.) were then advanced, but not fully tightened. This action provides a cantilever push against the apex of the curve, but allows rotation to occur around the convex rod. The convex rod acts as the axis of rotation allowing the spinal deformity to “derotate,” thereby decreasing the rib hump deformity. Next, the overbent 5.5-mm CoCr concave rod was placed using Cricket reducers, slowly reducing the coronal deformity while simultaneously derotating the spinal deformity by pulling up the concave, rotated side of the spine in a slowly advancing manner working from outside-in towards the apex (Figure 5). To avoid point loading, the reducers were tightened differentially, back and forth, which created a zipper effect to slowly correct the spinal curvature in all three planes. In summary, for this patient, a stiffer metal, such as CoCr, was used for the concave rod, and we stepped down a grade for the convex rod (Ti). The stiffness/Young’s modulus value for CoCr of 240 GPa is more than two-times the corresponding value for Ti of 115 GPa. Thus, we attempted to under-bend the convex or softer metal rod and over-bend the concave or stiffer metal rod to achieve and maintain correction in all three planes.


At 3 months after surgery, the patient’s back pain is slowly improving. She has been walking daily and states that she feels straighter when she stands at this point. Postoperative full-length (36”) AP and lateral standing scoliosis radiographs reveal improvements in the coronal and sagittal alignment to date (Figure 6).

Special reducers
Figure 5. Special reducers were then advanced, but not fully tightened (a). This action provides a cantilever push against the apex of the curve with rotation occurring around the axis of this rod. Next, the concave rod/rail was placed in standard fashion (b). The reducers were then advanced working from “outside-in” toward the apex. To avoid point loading, the reducers were tightened differentially, back and forth, to create a “zipper effect.”


Sagittal balance is an important aspect of the upright adult posture and adult spinal deformity. Although there have been fewer papers examining the impact of sagittal alignment in adolescents, there is no doubt that treatment for AIS can have a profound impact on standing sagittal alignment as an adult. The relationships between the overall sagittal balance, the spinopelvic measurements and radiographic measurements in the lumbar spine (PI, PT, SS and LL) have been widely described. A greater understanding of spinal sagittal balance in AIS has increased the surgical objective among scoliosis surgeons. Traditionally, the prevention of flatback syndrome, which is a common consequence of scoliosis surgery that results in a poor sagittal profile, has been focused on spinal fusions that extend into the lumbar spine. However, little discussion has been focused on what the thoracic fusion’s sagittal profile effect would be on an unfused lumbar spine, as is often the scenario following posterior spinal fusion for AIS. As demonstrated in this patient, the TK was mildly improved to allow for an improved LL relationship to the PI. The entirety of the lumbar spine was not involved in the fusion construct, yet improvement in the TK will hopefully allow for improvement of the LL, resulting in improved sagittal balance as described by Legaye and colleagues.

Postoperative AP and lateral standing full-length scoliosis plain radiographs
Figure 6. Postoperative AP (a) and lateral (b) standing full-length scoliosis plain radiographs revealed significant improvement in coronal scoliosis curves and early mild improvements in the sagittal alignment.

Corrective strategies have been trialed to address TK in AIS patients: in situ rod bending, rod rotation, rod cantilever, Ponte osteotomies, different implant densities, direct vertebral rotation and compression/distraction of rod segments. We utilized a combination of differential rod bending and the use of different metals to apply dual rod correction of the AIS deformity. Using our described method, the material properties of the concave and convex rods were critical in performing the differential rod bending technique to achieve our goal of ideal correction in the coronal, axial and sagittal planes. Both steel and Ti rods have been observed to flatten during insertion and/or rod derotation maneuvers. Using two steel rods, Cidambi and colleagues found that over-correcting the concave rod by about 20° resulted in a high degree of correction in the coronal and axial planes without loss of sagittal alignment. The described differential bending technique includes an initial cantilever push against the apex of the curve with a flattened convex ‘softer’ rod that acts as the axis of rotation. Subsequently, we apply rotation around this axis of rotation (the convex rod) via the special reducers on the concave, stiffer rod. Thus, translation coupled with rotation causes the apex of the curve to rotate up to the concave rod while simultaneously, the stiffer, concave rod flattens partially as correction is obtained and the softer rod flexes to match the stiffer, concave rod into the ideal sagittal alignment. There is still much to learn regarding the role of spinopelvic parameters and the pathogenesis of the thoracic spine in AIS. Further studies are necessary to determine the importance of the sagittal plane to determine overall functional outcome in this patient population.

Key points

Disclosures: DeWald reports he has a research grant from K2M for this technique using differential metals. An, Hirsch, Iyer and Louie report no relevant financial disclosures.