Role of T2 pedicle subtraction osteotomy in rigid cervicothoracic sagittal imbalance correction: A case series with technical details Authors: Mohsen Rostami, Bayard R. Wilson, Daniel I. Wolfson, Sadegh Bagherzadeh, Christina A. Rogers, Vincent C. Traynelis, Ricardo B. V. Fontes Published: 2026 ABSTRACT ABSTRACTBackground:Correction of rigid cervicothoracic sagittal imbalance is technically demanding, especially in patients with prior cervical fusion or ankylosing spondylitis. Three-column osteotomies at C6–T1 can achieve deformity correction but carry risks of vertebral artery and cervical nerve root injury. T2 pedicle subtraction osteotomy (PSO) may provide comparable correction with a more favorable neurovascular risk profile. We evaluated radiographic outcomes, durability of correction, and safety of T2 PSO for rigid cervicothoracic deformity.Materials and Methods:Nineteen consecutive patients who underwent T2 PSO (2017–2025) were retrospectively reviewed. Radiographic parameters, including C2–C7 sagittal vertical axis (SVA), C2–S1 SVA, T1 slope (TS), TS–CL mismatch, and C6–T4 cervicothoracic angle (CTA), were assessed preoperatively, immediately postoperatively, and at last follow-up. Complications and reoperations were recorded.Results:Mean C2–C7 SVA improved from 65.5 mm to 40.1 mm postoperatively (−25.4 mm, P < 0.001) and remained improved at 46.5 mm at final follow-up (P = 0.005). C2–S1 SVA decreased by 30.8 mm (P = 0.008). TS improved from 47.3° to 31.2° (−16.1°, P < 0.001), TS–CL mismatch by 12.9° (P < 0.001), and CTA from 16.7° to 5.2° (−11.5°, P = 0.002). Radiographic gains were largely maintained at a mean 11.6-month follow-up. Complications occurred in 47.4%; one transient neurologic deficit occurred, with no permanent spinal cord, nerve root, or vascular injuries.Conclusion:T2 PSO provides substantial and durable correction of cervicothoracic sagittal imbalance with a favorable neurovascular safety profile, particularly in patients with prior cervical fusion or ankylosing spondylitis. INTRODUCTION Fixed cervicothoracic sagittal deformities pose significant treatment challenges, particularly since they are often diagnosed in patients with a prior cervical fusion, or with ankylosing spondylitis.[12] They often lead to difficulties in maintaining an upright posture and a forward gaze, and can result in debilitating neck and lower back pain.[34] Early efforts to surgically address rigid cervicothoracic deformities focused largely on cases of ankylosing spondylitis, using opening wedge osteotomies, but the high incidence of neurological complications associated with those techniques limited their widespread application.[5] Over time, the prevailing surgical approach shifted from performing opening wedge osteotomies to performing three-column osteotomies (3COs), thought to carry a lower risk profile. Today, the most common 3COs performed for rigid cervicothoracic kyphosis are pedicle subtraction osteotomies (PSOs), most often performed at C6, C7[6] or T1 vertebral levels.[78] While these techniques are effective in achieving radiographic success, their risk profile is significant. At C7 and T1, there is a potential for injury to the lower cervical nerve roots and the sympathetic plexus, which can result in triceps and intrinsic hand muscle weakness, or in a postoperative Horner syndrome.[9] At C6, damage to the hand intrinsic roots can be avoided, but other clinically meaningful nerve roots remain at risk, and the presence of the vertebral artery within the transverse foramen at that level presents its own set of surgical challenges.[10] Pedicle subtraction osteotomy (PSO) at the T2 level shares many of the advantages of the same procedure at more cranial levels without the risk of cervical nerve root or vertebral artery injury. In this study, we describe our institutional experience with this approach using a case series of 15 patients with cervicothoracic sagittal imbalance treated with a T2 PSO, the largest such series to date. MATERIALS AND METHODS This is an IRB-approved (ID Number: 25010604), retrospective study involving adult patients who underwent T2 PSO for severe cervicothoracic sagittal deformity between November 2017 and October 2025. Study design, data extraction method This study is a retrospective case series focusing on patients who underwent T2 PSO for the correction of cervicothoracic sagittal imbalance. Demographic data, including age, sex, American Society of Anesthesiologists score,[11] body mass index, prior surgical history, and follow-up duration, were collected. Surgical data were also extracted, including indications for surgery, uppermost instrumented vertebra (UIV), lowermost instrumented vertebra (LIV), procedure duration, estimated blood loss (EBL), and complications. Radiological parameters were collected using perioperative and outpatient follow-up X-rays and included the C2–C7 sagittal vertical axis (SVA), C7-S1 SVA, C2-S1 SVA, occipital-C2 angle, chin-brow vertical angle (CBVA), C2–C7 cervical lordosis (CL), T1 slope (TS), C6-T4 cervicothoracic angle (CTA), TS-CL mismatch (TS-CL), T5-T12 kyphosis, sacral slope (SS), and pelvic tilt (PT). Surgical description All procedures were performed by a deformity-focused, fellowship-trained senior spine surgeon with assistance from senior neurosurgical trainees at a tertiary academic center. All cases included the use of intraoperative neuromonitoring, including somatosensory evoked potentials and motor evoked potentials. Patients were positioned prone on a Jackson open frame table (Mizuho OSI, Tokyo, Japan) with arms tucked, and with the head secured in a Mayfield clamp attached to an electric mechanical arm (Levo®; Mizuho OSI, Tokyo, Japan). Bony elements were exposed in standard fashion using subperiosteal dissection, and all instrumentation was placed using a free-hand technique with fluoroscopy for verification. Pedicle screws were placed at the T1 vertebra in each case, and lateral mass screws were placed in the subaxial cervical spine. In cases extending to C2, pars screws were placed at that level. In each case, pedicle screws were placed at least three levels caudal to the osteotomy site, with one exception in which it stopped at T4. After confirming screw position with fluoroscopy, the osteotomy was started by first resecting the transverse processes at the T2 level and removing bilateral T2 rib heads, ensuring to protect the parietal pleura. These steps were completed prior to any canal decompression to minimize the number of steps performed with the canal and thoracic spinal cord exposed. Anterolateral retractors were placed (“spoon” or “Cobra” retractors) to protect structures anterolateral to the thoracic spine, and then posterior column osteotomies were begun, starting with an en bloc removal of the pars, lamina, and inferior articular processes of T1 and T2, resulting in a wide decompression to allow for dural buckling after closure of the osteotomy. After confirming generous pedicle-to-pedicle decompressions of bilateral T1 and T2 nerve roots, the T2 pedicles were then resected using osteotomes and Leksell rongeurs down to their base flush with the T2 vertebral body. A wedge-shaped osteotomy was then cut using straight osteotomes such that its apex was just posterior to the anterior vertebral cortex at T2 [Figure 1]. In nearly every case, the T1-2 disc was included in the resected wedge. A temporary rod was applied to span the osteotomy once one side of the wedge had been resected (~60% of the vertebral body) for ease of workflow. Following completion of the osteotomy on the contralateral side, manipulation of the Levo mechanical arm to allowed for correction of kyphosis by changing the position of the head in space. Further correction was achieved by compressing the T1 and T3 pedicle screws over the temporary rod, thereby reducing the stress on the fixation points during closure [Figure 2]. MEPs and SSEPs were monitored throughout the case and verified after each corrective maneuver; in no single case did the osteotomy have to be reopened. Schematic illustrations of spinal alignment before and after T2 pedicle subtraction osteotomy (left and middle), and a magnified anatomical view (right) depicting the spinal osteotomy through resection of posterior and anterior elements, enabling spinal realignment Radiographic demonstration of cervicothoracic deformity correction. (a and b) Preoperative and postoperative lateral radiographs of a patient with a stooped-forward posture due to distal junctional failure and pseudoarthrosis including fractured implants following prior posterior cervical surgeries, demonstrating marked improvement in sagittal alignment after a posterior-based T2 pedicle subtraction osteotomy and instrumented fusion. (c and d) Another case showing correction of a fixed cervicothoracic deformity achieved through a posterior T2 osteotomy Statistical analysis Statistical analyses were performed using GraphPad Prism. The Kolmogorov–Smirnov test was used to assess the normality of the data. For nonnormally distributed data, non-parametric tests, including the Wilcoxon test and Kolmogorov–Smirnov test, were employed to compare outcome measures. All statistical tests were two-sided, and a P < 0.05 was considered statistically significant. RESULTS Demographic and clinical characteristics As summarized in Table 1, 19 patients (6 males and 13 females) with cervicothoracic junctional kyphosis were included in the study. The mean age of the participants was 62.4 years. All patients had a history of prior spine surgery, and 57.9% had a history of smoking, although none were active smokers at the time of surgery. The mean follow-up duration was 11.6 months. Table 1Demographic and surgical characteristics and complication profile of the 19 patients included in the studyVariablen (%)Age (year), mean±SD62.4±10.01Male: female6:13ASA3 (0.48)BMI (kg/m2)28 (6.2)History of smoking11 (57.9)History of previous spine surgery18 (100)Follow-up duration (month)11.6 (9.9)Surgery time (min)559 (108)Estimated blood loss (mL)1458 (653)Surgery indication Upper thoracic DJF8 (42.1) Upper thoracic PJF3 (15.8) C7 DJF2 (10.5) Cervicothoracic rigid kyphosis6 (31.6)PSO level Only T218 (94.7) T10 + T21 (5.3)UIV selection C215 (78.9) C31 (5.2) C41 (5.2) C51 (5.2) C61 (5.2)LIV selection T42 (10.5) T57 (36.8) T61 (5.2) T104 (21) T121 (5.2) Pelvis4 (21)Complications Any complication9 (47.4) SSI4 PJF1 DJF4 Postoperative spinal root deficit (temporary/permanent)1/0 PTE2 Posterior fossa ICH1Reoperation Total number of reoperations9 (47.4) Due to DJF4 Due to SSI3 Due to PJF1 Due rod fracture1BMI - Body mass index; DJF - Distal junctional failure; PJF - Proximal junctional failure; UIV - Upper instrumented vertebra; LIV - Lower instrumented vertebra; SSI - Surgical site infection; SD - Standard deviation; ASA - American Society of Anesthesiologists; PTE - Pulmonary thromboendarterectomy; ICH - Intracerebral hemorrhage; PSO - Pedicle subtraction osteotomy Demographic and surgical characteristics and complication profile of the 19 patients included in the study BMI - Body mass index; DJF - Distal junctional failure; PJF - Proximal junctional failure; UIV - Upper instrumented vertebra; LIV - Lower instrumented vertebra; SSI - Surgical site infection; SD - Standard deviation; ASA - American Society of Anesthesiologists; PTE - Pulmonary thromboendarterectomy; ICH - Intracerebral hemorrhage; PSO - Pedicle subtraction osteotomy Surgical data assessment Surgical data are presented in Table 1. All patients underwent T2 PSO, and one patient underwent an additional PSO at T10. The most frequently selected UIV was C2, representing 78.9% of cases, followed by C3, C4, C5, and C6, each accounting for 5.2%. The most common lower instrumented vertebra was T5 (36.8%), followed by T10 and Pelvis (21%). The mean surgical duration was 559 ± 108 min, and the mean EBL was 1458 ± 653 mL. Complications and reoperations Complications and reoperation rates are also detailed in Table 1. In total, 47.4% experienced at least one medical or surgical complication. Notably, one patient experienced a transient postoperative neurological deficit, which self-resolved, and another suffered a delayed cerebellar hemorrhage in the postoperative period requiring take-back for decompressive craniectomy. One patient developed proximal junctional failure (PJF), while four patients experienced distal junctional failure (DJF). Surgical site infections (SSIs) occurred in four cases. Overall, reoperations were required in 36.8% of patients, primarily for DJF (4 cases), followed by PJF, Rod fracture, and SSI (each 5.2%). Radiological outcomes Radiological measurements taken preoperatively, immediately postoperatively, and at the last follow-up are summarized in Table 2. For better comparison, we included patients with full radiological outcomes for radiographic analysis. Significant improvements were observed in several parameters, as illustrated in Figure 3. Table 2Comparison of radiographic outcomes of the patients at preoperative, immediate postoperative, and at last follow-up imaging (17 patients with complete radiographic follow-up)Preoperative, mean±SDPostoperative, mean±SDLast follow-up, mean±SDPost versus pre, mean difference (change %) P Last follow-up versus pre, mean difference (change %) P CBVA (°)5±10.32.6±6.8−0.3±8.5−2.4 (−48)0.312−5.3 (−106)0.046C2–C7 SVA (mm)65.5±24.540.1±14.846.5±14.8−25.4 (−25.5)<0.001−19 (−29)0.005C7–S1 SVA (mm)43.5±59.437.1±47.858.1±62.9−6.4 (−14.7)0.47814.6 (+33.6)0.326C2–S1 SVA (mm)104.1±75.473.3±55.596.8±75.1−30.8 (−29.6)0.008−7.2 (−6.9)0.368PT (°)23.8±12.223.8±9.926±11.60 (0)0.9772.2 (+9.2)0.121SS (°)30.6±11.530.5±9.128.1±10.2−0.1 (−0.3)0.977−2.5 (−8.2)0.155T5–T12 kyphosis (°)47.4±17.443.2±14.743.4±16.5−4.2 (−8.8)0.240−4 (−8.4)0.429C2–C7 lordosis (°)11.2±12.17.9±6.111.6±8.4−3.3 (−29.5)0.2100.4 (+3.6)0.999TS (°)47.3±15.231.2±7.731.8±8.3−16.1 (−34.1)<0.001−15.5 (−32.8)<0.001TS–CL (°)36.1±13.823.2±10.320.2±11.2−12.9 (−35.7)<0.001−15.9 (−44.1)<0.001C6–T4 CTA (°)16.7±13.95.2±10.14±10.5−11.5 (−68.9)0.002−12.7 (−76.1)<0.001OC2 (°)27±10.923.8±13.227.9±7.7−3.2 (−11.9)0.4280.9 (+3.3)0.773CBVA - Chin brow vertical angle; SVA - Sagittal vertical axis; PT - Pelvic tilt; SS - Sacral slope; CTA - Cervicothoracic angle; OC2 - Occipit-C2 angle; TK - Thoracic kyphosis; SD - Standard deviation; CL - Cervical lordosis; TS - T1 slope Comparison of radiographic outcomes of the patients at preoperative, immediate postoperative, and at last follow-up imaging (17 patients with complete radiographic follow-up) CBVA - Chin brow vertical angle; SVA - Sagittal vertical axis; PT - Pelvic tilt; SS - Sacral slope; CTA - Cervicothoracic angle; OC2 - Occipit-C2 angle; TK - Thoracic kyphosis; SD - Standard deviation; CL - Cervical lordosis; TS - T1 slope Longitudinal changes in cervical sagittal parameters across three timepoints: Preoperative, immediate postoperative, and last follow-up images. Blue lines represent sagittal vertical axis (SVA) measurements: C2–S1 SVA (square markers) and C2–C7 SVA (circle markers), plotted on the right Y-axis (in mm). Red lines represent angular parameters: T1 slope (upward triangle), T1 slope-cervical lordosis mismatch (downward triangle), and C6-T4 cervicothoracic angle (C6–T4) (diamond markers), chin-brow vertical angle (circle marker), plotted on the left Y-axis (in degrees). Asterisk (*) indicates a statistically significant difference between immediate postoperative and preoperative values. Solid circle (•) indicates a statistically significant difference between the last follow-up and preoperative values The C2–C7 SVA decreased from a preoperative mean of 65.5 mm to 40.1 mm immediately postoperatively (P < 0.001) and then increased slightly to 46.5 mm by the time of last follow-up (P = 0.005) but remained significant in comparison to preoperative measure. The C2–S1 SVA decreased from a preoperative mean of 104.1 mm to 73.3 mm postoperatively (P = 0.008) but returned to a near-preoperative value of 96.8 mm at final follow-up (P = 0.368). The CTA improved significantly, decreasing from a preoperative mean of 16.7° to 5.2° immediately postoperatively (P < 0.001), and further to 4° at the last follow-up (P < 0.001). The TS decreased from 47.3° preoperatively to 31.2° immediately after surgery (P < 0.001), and to 31.8° at the final follow-up (P < 0.001). TS–CL mismatch also showed significant improvement when compared to baseline, with P = 0.002 and < 0.001 at the immediate and last follow-up, respectively. CBVA change was not significant in the early postoperative image (P = 0.312), but in the last follow-up, it reached significance level with a mean decrease of 5.3 (106% improvement) compared to baseline (P = 0.046). No other measured radiological parameters showed statistically significant changes following surgery. DISCUSSION Our series of 19 patients treated with T2 PSO for rigid cervicothoracic kyphosis suggests that this technique can be effectively employed to achieve radiographic correction with relatively lower risk of neurovascular complications. During surgical planning for correction of cervicothoracic kyphosis, selecting the appropriate level at which to perform a 3CO typically involves a familiar trade-off: A more caudal osteotomy, such as at T1, increases the risk for intrinsic hand weakness and Horner’s syndrome, while a more rostral osteotomy – such as at C6– increases the risk for injury to clinically significant cervical nerve roots and the vertebral artery and affords poorer biomechanical fixation above the osteotomy in lateral mass screws. While not without risks inherent to large deformity surgery, a T2 PSO represents a third option that may minimize overall risk and is particularly well-suited for two sub-groups of patients with cervicothoracic sagittal imbalance. The first is any patient with rigid cervical kyphosis due to prior fusion surgery. For these patients, a T2 PSO usually allows for the osteotomy to be performed in a segment of spine without a laminectomy defect or (sometimes) instrumentation. These factors alone can improve surgical efficiency and reduce complications related to operating over a previously decompressed spinal canal. The second is any patient with ankylosing spondylitis, since AS patients often exhibit concurrent thoracolumbar sagittal imbalance, and a T2 PSO is sufficiently caudal in many cases to address both cervicothoracic and thoracolumbar sagittal deformities simultaneously, offering a comprehensive correction of sagittal balance. In our series of 19 patients with cervicothoracic kyphosis, use of a T2 PSO achieved an average of-25.4 mm (−25.5%) correction in the C2–C7 SVA and −30.8 mm (−29.6%) correction in the C2-S1 SVA, significantly improving cervicothoracic sagittal alignment. This extent of correction is similar to that achievable via multi-level cervical posterior column osteotomies, such as those described by Lau et al.,[12] though another smaller case series described an even greater degree of correction (38.5 mm) using a similar technique.[13] Compared to a cervical PSO, a T2 PSO provides greater translational correction in SVA,[8] a phenomenon attributed to two key biomechanical factors. First, as the osteotomy level moves caudally along the spine, the available pedicle size and vertebral height increase, enabling more extensive bone resection and, consequently, greater angular correction.[14] Second, a more caudal application of the correction maneuver naturally produces greater translational impact on SVA due to an increase in the size of the lever arm being moved,[15] a principle which applies to the entire spine, in that lumbar PSO provides a greater degree of SVA correction than thoracic PSO.[16] Our cohort also experienced a significant reduction in their TS (16.1° [34.1%]), and TS-CL mismatch (12.9° [35.7%]), and while their CL decreased by 3.3°, this change did not reach statistical significance. Previous studies have established a positive correlation between TS and CL, showing that an increase in TS requires a proportional increase in CL to maintain sagittal balance.[17] Thus, reducing the TS-CL mismatch via a reduction of the TS is critical in deformity correction, especially when a prior fusion or ankylosis prohibits straightforward corrections to CL.[18] In contrast, osteotomies performed in the cervical region (i.e., above T1) do not affect TS; instead, they achieve TS-CL mismatch correction by increasing CL. Traditionally, CL in the subaxial spine has been predominantly corrected via circumferential (i.e., anterior-posterior) approaches; an additional advantage of correcting TS-CL mismatch via T2 PSO is that it allows for definitive correction via a single position and approach. A significant number of patients in our cohort had previously undergone circumferential surgery and subsequently developed DJF; their subjective impression, shared by the senior author, is that recovery from the T2 PSO was easier than from the combined anterior-posterior approach.[1920] In our cohort, the (CTA, C6-T4, encompassing both lower CL and upper thoracic kyphosis) improved by 11.5° (68.9%), achieving a postoperative mean of 5.2°, similar to the normal range observed in asymptomatic individuals (5° ±8°).[21] Further studies will be necessary to determine the impact of CTA correction specifically on postoperative pain and/or disability in the treated population. CBVA, by contrast, failed to see a significant change following T2 PSO in our cohort in early postoperative image, however in the last follow-up, it showed 5.3° (106%) correction. This is likely because the mean preoperative CBVA for our cohort (5 ± 10.3°) falls within the normal range for CBVA, suggesting that our patients were still able to compensate for subaxial and cervicothoracic kyphosis with hyperextension at the O-C2 or lumbar segments. Another plausible explanation is the presence of a Type II error,[22] since the high standard deviation in preoperative CBVA reflected considerable variability among patients, and the limited sample size further constrained our ability to detect a statistically significant effect.[2324] Importantly, after a mean follow-up of 11.6 months, most of the improvements in radiologic parameters remained stable, and while longer-term follow-up data are necessary to determine the extended durability of this technique, results at 1 year are reassuring. Key alignment markers such as the TS, TS–CL mismatch, and C6–T4 (CTA) were all maintained at last follow-up, reflecting the anatomic configuration of the cervicothoracic junction, which plays a pivotal role in maintaining horizontal gaze and relieving stress on spinopelvic compensatory mechanisms. Global alignment, with or without DJF, however, remains a known significant challenge with a relative worsening in the C2-S1 SVA at the last follow-up. While a majority of the measured cervical and cervicothoracic parameters demonstrated significant improvement following T2 PSO in our cohort, neither of the measured spinopelvic parameters reflecting lumbar and global compensation (SS and PT) saw significant change. A failure to demonstrate significant changes in SS or PT following cervicothoracic correction may reflect additional lumbosacral pathology or under-correction of cervicothoracic alignment (still necessitating lumbosacral compensation). However, in our case, the most likely explanation is that these parameters – much like with CBVA – were within normal ranges for our cohort preoperatively, and so should not be expected to change substantially, even following a successful and significant cervicothoracic correction. Our study revealed that 47.4% of patients experienced at least one medical or surgical complication. Wound-related complications – among the most common – are an unfortunate reality in this cohort of patients with prior posterior surgery and existing posterior de-vascularized scar tissue. DJF is another well-documented mechanical complication in posterior-based cervical osteotomies. In their series, Lau et al.[25] identified DJF as their most common mechanical complication. In our series, DJF occurred in 21% of patients, all of whom required reoperation. While PJK – a relatively common consequence of thoracolumbar fusion – can often be observed and managed conservatively, DJF rarely can. All of the patients in our series with DJF were managed conservatively initially, but all failed very quickly, underscoring the importance of a solid base for cervicothoracic constructs. The most serious complication of this series, a cerebellar hemorrhage 24 h after the end of surgery, required a craniectomy and decompression. Despite there being no obvious durotomy during surgery, the patient’s spinal wound was explored for evidence of CSF leak; none was found. He would eventually make a significant recovery, and at his most recent follow-up (8 months), he was alert, oriented, and ambulatory but required assistance with activities of daily living. Finally, another patient developed a transient radicular C6-7 deficit, which resolved by the 3rd month postoperatively. In their smaller case series of four patients treated with T2 PSO, Yuk et al.[13] similarly reported no neurological complications. EBL and operative time in our series were comparable to the literature.[26] Four patients in our cohort died during the course of the study. However, all four deaths were due to causes unrelated to spine surgery and over 1 year after surgery – sudden cardiac arrest, glioblastoma multiforme, kidney failure, and COPD. While unfortunate, this mortality rate reflected a baseline level of medical complexity inherent to our cohort, and likely to be present in a significant proportion of patients undergoing these types of complex spinal deformity corrections, especially with longer follow-up. Despite these complications, the overall safety profile of the T2 PSO in our cohort compares favorably to other cervicothoracic osteotomies, particularly with regard to vertebral artery and hand intrinsic root injury.[27] Osteotomies at C7 or T1 levels carry a notable risk of injury to the hand intrinsic roots; postoperative hand dysfunction rates have been reported to be as high as 22% and 20% at 1 and 2 years postoperatively, respectively.[13252829] Lau et al.[12] found that – of those patients affected by such injuries – only one-third regain their prior functionality. For reasons previously discussed, a T2 PSO carries a much lower risk of cervical plexus root injury and minimizes the likelihood of hand dysfunction.[25] Moreover, performing a T2 PSO allows for the possibility of sacrificing the T2 nerve root to facilitate osteotomy, an option that is unavailable at more cranial levels. In our cohort, no patient was found to have post-fixed brachial plexus anatomy based on intraoperative testing, and in three patients, the T2 nerve roots were ligated without any noticeable consequences to the patients. Finally, an osteotomy at T2 provides the surgeon with a generous pedicle at T1; having reliably available pedicles both above and below the osteotomy increases efficiency and biomechanical stability, as evidenced by the correction achieved in our cohort. For cases requiring significant correction, the T1 pedicle reduces the stresses placed on more cranial (and biomechanically inferior) lateral mass fixation.[30313233] Our study demonstrates that T2 PSO is a reproducible technique with consistent outcomes, particularly when performed within a high-volume academic center. Historically, this approach has been limited to a few specialized spine surgeons, raising concerns about its generalizability. However, our results – achieved across a team involving different senior trainees under the supervision of a deformity-focused attending – suggest that with appropriate training and adherence to technical principles, T2 PSO can be safely and effectively replicated. Disadvantages of a T2 PSO include a greater risk of pleural injury, an inability to correct CL, and uncertainty over LIV selection as it relates to DJF risk. While we did have a handful of parietal pleural violations in our series, there were no postoperative clinical and radiological pneumothoraxes. We also consistently utilized T5 as the LIV since our patients’ global alignment placed the apex of their thoracic kyphosis slightly lower than usual; with this LIV selection, our rate of DJF did not differ from previous reports of cervicothoracic 3CO’s.[343536] An osteotomy placed more cranially could – in theory – reduce the risk of DJF with a higher LIV, though this issue remains a matter of debate. This study has several limitations. Our series of 19 patients lacks the statistical power to broadly generalize our findings. This is unsurprising since we argue that this technique is underutilized; that our cohort of 15 patients is the largest published to date is perhaps the best evidence of this. Second, the retrospective design inherently introduces potential biases in data interpretation, particularly in patient selection, as this is a procedure preferentially performed in patients with an elevated TS. Third, these procedures are performed normally in combination with other, different single-column osteotomies and different fusion length which can introduce biases related to the complication pattern. Fourth, the relatively short follow-up period may limit the detection of long-term complications, such as adjacent-level degeneration, DJF, or pseudoarthrosis, which could emerge over time. Finally, one patient was lost to follow-up after surgery, so their long-term radiologic parameters could not be included in the analysis. CONCLUSION T2 PSO is an effective surgical technique to address cervicothoracic sagittal imbalance and offers a favorable neurovascular safety profile than more cranial 3CO’s. As compared to more cranial osteotomies, a T2 PSO affords the surgeon reliable pedicle fixation above the osteotomy site, as well as a more substantial correction in C2–C7 SVA. As such, it is particularly well-suited for patients with fixed cervical deformity from a prior fusion surgery and allows for significant correction while circumventing two of the most dreaded complications of more cranial PSOs: Vertebral artery injury and postoperative hand intrinsic weakness. Wider adoption of this technique should only serve to improve its safety and reliability, and we hope to see greater application of it in the deformity landscape moving forward. Conflicts of interest There are no conflicts of interest.