Pedicle Screw–Aortic Interaction: A Single Centre Experience and Proposal for a Morphology Based Management Algorithm Authors: Federica Campana, Stefano Ancetti, Chiara Cini, Paolo Spath, Alessia Pini, Enrico Gallitto, Alessandro Gasbarrini, Mauro Gargiulo Published: 2026 ABSTRACT AbstractObjectivePedicle screw (PS) fixation is widely used in thoracolumbar spine surgery. Although major vascular injury is uncommon, PS–aortic interactions may lead to severe complications, and optimal management remains poorly defined. This study proposes a morphology based algorithm to guide multidisciplinary management of PS–aortic interactions and reports a single centre experience.MethodsAll consecutive patients referred to a vascular surgery unit between 2020 and 2025 for PS interaction with the descending thoracic aorta were reviewed retrospectively. Pre-operative assessment was performed using contrast enhanced computed tomography angiography (CTA). Screw–aorta interactions were classified as near contact (NC), aortic contact (AC), aortic indentation (AI), or aortic penetration (AP). Management followed a multidisciplinary morphology based algorithm: NC and AC were managed with screw removal under serial intra-operative angiographic control, whereas AI and AP underwent prophylactic thoracic endovascular aortic repair (TEVAR).ResultsEight patients (median age 25 years [interquartile range 18, 38]) were included. Indications for spinal fixation were scoliosis (n = 6) and trauma (n = 2). CTA demonstrated NC in two patients, AC in three, and AI in three; no AP cases were identified. All patients underwent a multidisciplinary vascular–spine procedure with surgical left femoral access in the lateral decubitus position. In NC and AC cases, posterior screw removal was followed by completion angiography confirming the absence of contrast extravasation; therefore, rescue TEVAR was not required. All AI cases underwent prophylactic TEVAR immediately before screw removal. Technical success was achieved in all cases. There were no major adverse aortic events, neurological complications, or 30 day deaths. At a median follow up of 26.9 months (interquartile range 7, 42), all patients were alive and asymptomatic, with no aortic related complications.ConclusionA morphology based algorithm prioritising aortic wall deformation rather than proximity alone may provide a feasible and potentially safe strategy for managing PS–aortic interactions, supporting standardised multidisciplinary decision making in this rare but high risk condition.Keywords: Aortic injury, Endovascular repair, Screw malposition, Spinal instrumentation, TEVAR, Thoracic aorta Introduction Pedicle screw (PS) fixation is a cornerstone of modern thoracic and thoracolumbar spine surgery, providing reliable three column fixation across a wide range of pathological conditions, such as deformity correction, trauma, and degenerative disease. The close anatomic proximity between the thoracic spine and the descending thoracic aorta exposes patients to a rare but clinically relevant risk of iatrogenic aortic injury in the event of screw malposition1,2 that is reported in approximately 4% of cases, whereas aortic abutment represents an exceedingly rare event, reported in less than 0.07%.3 Nevertheless, the true incidence is probably underestimated, as many cases remain asymptomatic and are only detected incidentally during post-operative imaging or delayed evaluation. The descending thoracic aorta is particularly vulnerable due to its close relationship with the anterior vertebral cortex, a risk further increased in spinal deformity where anatomic relationships are altered.2,4 Clinical presentation is frequently asymptomatic, with complications potentially occurring months or years after the index procedure.5, 6, 7 Despite this relevant clinical issue, there is no standardised description of the anatomic relationship between malpositioned pedicle screws and the aortic wall, with the spectrum of screw–aorta interactions remaining poorly characterised. Moreover, management strategies reported in literature remain highly heterogeneous and left to case by case selection. Some authors advocate conservative management or isolated screw removal, whereas others recommend prophylactic or therapeutic thoracic endovascular aortic repair (TEVAR), particularly in the presence of wall deformation or penetration.6,8 Importantly, there is no consensus regarding the indications for vascular surgeon involvement, or the optimal timing and modality of intervention. This study proposes a morphology based classification and management algorithm and reports a single centre experience. Methods Study design and patient selection This retrospective, single centre study included all consecutive patients referred to the vascular surgery unit of a tertiary referral hospital between January 2020 and December 2025 for evaluation of a malpositioned pedicle screw adjacent to the descending thoracic aorta. Patients were considered eligible if they had undergone posterior spinal fixation with thoracic or thoracolumbar pedicle screws and demonstrated potential proximity or contact between a screw tip and the thoracic aorta on post-operative computed tomography angiography (CTA). Referral to vascular surgery was initiated by the orthopaedic spine team based on post-operative imaging or intra-operative suspicion. The study was conducted in accordance with the Declaration of Helsinki. According to institutional policy and given the retrospective design of the study and the analysis of anonymised data, formal ethics committee approval was not required. Data were acquired through institutional medical and radiological records. Imaging assessment All patients underwent CTA with thin slice acquisition (≤ 1 mm) and multiplanar reconstructions aligned along the screw trajectory. Minimum screw–aorta distance was measured orthogonally from the closest point of the screw to the nearest point of the aortic wall. Aortic wall deformation was defined as any loss of the native circular or elliptical contour, and indentation depth as the maximum inward displacement of the aortic wall. Patients were referred to the vascular team when CTA demonstrated a screw–aorta distance ≤ 5 mm, measured from any part of the screw. Imaging was independently blindly reviewed by a vascular (F.C.) and a spine surgeon (C.C), and interactions were classified (near contact, aortic contact, aortic indentation, aortic penetration) to guide management. In case of disagreement, classification was defined by consensus with the additional involvement of senior vascular and spinal surgeons (M.G., A.G.). Morphology based classification Screw–aorta interaction was categorised according to a pre-defined morphology based classification designed to reflect the spectrum of anatomic involvement and corresponding procedural risk. Classification relied on CTA with multiplanar and oblique reconstructions optimised along the screw trajectory (Fig. 1). Near contact (NC) is the measurable distance between the screw and the aortic wall, without direct contact. Aortic contact (AC) is direct contact between the screw and the aortic wall without deformation of the aortic contour. Aortic indentation (AI) is deformation of the aortic wall caused by the screw without evidence of luminal penetration. Aortic penetration (AP) is intraluminal extension of the screw with interruption of the aortic wall, with or without contrast extravasation, pseudoaneurysm, intraluminal defect, or aortic rupture. When imaging artefacts prevented confident discrimination between AC and AI, the interaction was conservatively classified as AI for procedure planning. Morphology based classification of pedicle screw–aortic interaction. (1) Near contact: measurable screw–aorta distance without contact; (2) Aortic contact: direct contact without contour deformation; (3) Aortic indentation: aortic wall deformation without intraluminal breach; (4) Aortic penetration: intraluminal screw extension or contrast extravasation. Management and surgical strategy Based on the proposed morphological classification, a standardised management pathway was adopted (Fig. 2). All procedures were performed in a hybrid operating room equipped with a fixed imaging system (Allura Clarity, Philips, Hamburg, Germany) within a multidisciplinary team including vascular and spine surgeons. Patients were positioned in the right lateral decubitus, and surgical exposure of the left common femoral artery was routinely performed to obtain endovascular access (Fig. 3). A stiff guidewire (Lunderquist, Cook Medical, Bloomington, IN, USA) and large calibre sheath were advanced into the ascending aorta to allow intra-operative angiography and immediate TEVAR if required (Fig. 4). Morphology based management algorithm for pedicle screw–aortic interaction. CTA = Computed tomography angiography; PS = pedicle screw; TEVAR = thoracic endovascular aortic repair. Patient positioning and surgical exposure. (A) Lateral decubitus positioning with appropriate stabilisation to allow optimal exposure of both the common femoral artery for vascular access and the spinal column for pedicle screw removal. (B) Surgical exposure of the spine for pedicle screw removal. (C) Skin preparation and surgical field draping, highlighting lateral decubitus and anatomic landmarks. Intra-operative imaging. (A) Thoracic spinal instrumentation in proximity to the descending thoracic aorta. (B) Deployment of a thoracic endograft across the affected aortic segment (T6) to exclude the area at risk. (C) Completion angiography after endograft deployment and posterior screw removal, confirming correct stent graft positioning and absence of contrast extravasation. In NC and AC cases, posterior screw removal was performed while repeated intra-operative angiograms were performed to detect any aortic injury. In the absence of extravasation, spinal fixation was completed; otherwise, immediate TEVAR was performed. In AI, prophylactic TEVAR was performed before screw removal, followed by spinal revision. Completion angiography was performed to confirm absence of extravasation and correct endograft positioning. In aortic penetration cases treatment should be performed in an emergency setting, with immediate thoracic endovascular aortic repair before screw removal. A contrast enhanced computed tomography scan was routinely performed in all patients 24–48 hours post-operatively. Study outcomes and definitions Primary outcomes were intra-operative major adverse aortic events (MAAE) and technical success. MAAE included aortic rupture, uncontrolled bleeding requiring unplanned TEVAR or open repair, haemodynamic collapse, or intra-operative death. Technical success was defined as complete screw removal without angiographic evidence of bleeding and without unplanned conversion or emergency aortic intervention. In TEVAR cases, this additionally required correct endograft deployment, patency, and absence of type I endoleak. Secondary outcomes included neurological complications, 30 day re-intervention, access site complications, length of hospital stay, and follow up imaging findings. Statistical analysis Given the descriptive nature of the study and the limited sample size, no formal hypothesis testing was planned. Data were analysed using descriptive statistics to characterise patient demographics, radiological findings, management strategies, and clinical outcomes. Continuous variables are reported as median and interquartile range (IQR), reflecting non-normal data distribution, while categorical variables are presented as absolute numbers and percentages. Statistical analyses were performed using IBM SPSS Statistics, version 28.0 (IBM Corp., Armonk, NY, USA). Results Patient characteristics and morphology of screw–aorta interaction Eight patients were included (seven women; median age 25 [interquartile range, IQR, 18, 38] years). Indications for spinal fixation were scoliosis in six patients and trauma in two. The affected vertebral levels ranged from T4 to T12. The median interval between index spinal surgery and vascular referral was five months [IQR 1, 20] (Table 1). Table 1Baseline characteristics, interaction pattern and treatment details.PatientSexAge – ySpinal indicationVertebral levelTime to presentation – moScrew–aorta interactionTreatmentTEVAR deviceSheath size1F54ScoliosisT432Aortic indentationScrew removal + TEVARGORE CTAG TGM 282810E20 Fr2F25ScoliosisT99Aortic indentationScrew removal + TEVARGORE CTAG TGM262610E20 Fr3F20TraumaT66Near contact (3 mm)Screw removal--4M30TraumaT64Aortic contactScrew removal--5F25ScoliosisT936Aortic indentationScrew removal + TEVARGORE CTAG TGM262610E20 Fr6F17ScoliosisT51Near contact (4 mm)Screw removal--7F47ScoliosisT60.6Aortic contactScrew removal--8F16ScoliosisT50.1Aortic contactScrew removal--25 (18, 38)∗5 (1, 20)∗For patients diagnosed with near contact, the measured minimum distance between the screw and the aortic wall is reported. GORE CTAG = GORE Conformable Thoracic Stent Graft with Active Control System.∗Data reported as median and interquartile range. Baseline characteristics, interaction pattern and treatment details. For patients diagnosed with near contact, the measured minimum distance between the screw and the aortic wall is reported. GORE CTAG = GORE Conformable Thoracic Stent Graft with Active Control System. Data reported as median and interquartile range. Two patients had an asymptomatic pedicle screw–aorta NC detected one and six months after the index spinal procedure, respectively. The diagnosis was confirmed by CTA following suspicious post-operative radiographic findings. Three patients presented with AC. Two had undergone spinal fixation for traumatic injury, one of whom was paraplegic because of the initial trauma, while the other patient had been treated for thoracolumbar scoliosis. All patients in this group were asymptomatic at presentation. In patient 8, the malpositioned pedicle screw was already suspected in the immediate post-operative period. A post-operative CTA confirmed screw–aorta contact without deformation, and the patient was referred to the vascular team to schedule a combined elective procedure. Three patients presented with AI. Patient 1 presented 32 months after thoracolumbar scoliosis surgery with back pain; the involved screw was located at the T4 level, with an aortic indentation measuring 4 mm on CTA. Patient 2 presented nine months after thoracolumbar scoliosis surgery with asymptomatic aortic indentation at the T9 level (3 mm). Patient 5 presented 36 months after thoracolumbar scoliosis surgery with thoracic back pain; in this case, the involved level was T9, with an aortic indentation depth was 5 mm. No AP cases were identified. The two assessors who blindly reviewed pre-operative images agreed on the morphology based classification in seven of the eight patients; the single discordant case was resolved by adjudication of the senior surgeons. Surgical management and outcomes All patients were treated according to the morphology based management strategy previously described. In patients classified as NC or AC (five cases), after left common femoral artery surgical access and placement of a Lunderquist guidewire and a large calibre introducer, the malpositioned pedicle screw was removed through a posterior surgical approach by the spine surgeons. Aortic angiography was performed through a catheter positioned in the descending thoracic aorta using iodinated contrast medium and showed no contrast extravasation or evidence of aortic injury in any patient. Consequently, immediate thoracic endovascular aortic repair (TEVAR) was not required in any of these cases, and spinal stabilisation was completed with correct repositioning of the pedicle screw. Patients classified as AI (three cases) underwent prophylactic TEVAR before screw removal. After femoral access, 10 cm long thoracic endografts (Gore CTAG, Gore, W.L. Gore & Associates, Flagstaff, AZ, USA) were successfully deployed with adequate proximal and distal landing zones, achieving complete coverage of the affected aortic segment. Thoracic endografts were oversized by approximately 10–15% relative to the healthy proximal aortic diameter. The proximal sealing zone was distal to the left subclavian artery (LSA) in all cases (Ishimaru zone 3– 4), and no LSA coverage or revascularisation was required. Subsequent posterior screw removal and spinal revision were performed during the same session. Final angiography confirmed correct endograft positioning, adequate proximal and distal seal, complete exclusion of the indentation site, and absence of intraluminal defects or contrast extravasation. Technical success was achieved in all patients (100%). There were no major adverse aortic events (MAAE), 30 day deaths, or neurological complications. There were no access site complications, re-interventions, or conversions to open surgery. The median length of hospital stay was three (IQR 2, 4) days. All patients treated with TEVAR were discharged on single antiplatelet therapy (acetylsalicylic acid) according to institutional protocol. Follow up outcomes In hospital thoraco-abdominal CTA in TEVAR treated patients demonstrated correct endograft positioning and complete exclusion of the indentation site, with no endoleaks, intraluminal defects, or contrast extravasation. In patients managed without TEVAR, post-operative CTA performed during the index hospitalisation showed preserved aortic wall integrity, with no evidence of contrast extravasation, aortic rupture, wall weakening, dissection, or other vascular abnormalities. There were no deaths at a median follow up of 27 months (IQR 7, 42; range 4 –63). In TEVAR treated patients, surveillance CTA at one year, with subsequent imaging individualised to limit cumulative radiation and contrast exposure, confirmed endograft patency, stable positioning, and absence of migration, thrombosis, or device related complications. Single antiplatelet therapy was continued lifelong. Patients managed without TEVAR underwent structured clinical and telephone follow up on a yearly basis. Available thoraco-abdominal imaging performed for orthopaedic surveillance or other clinical indications showed no evidence of delayed aortic wall abnormalities or vascular complications. However, as routine serial CTA was not systematically performed in this subgroup, the absence of subclinical late aortic changes could not be established. Discussion Despite the small sample size, this case series represents the largest single centre experiences to date on pedicle screw–aortic interaction. It is believed that it is the first series proposing and systematically applying a morphology based management algorithm within a standardised multidisciplinary pathway. The prevalence of malpositioned pedicle screws in proximity to major vascular structures remains unclear, as most available evidence derives from isolated case reports. Large surgical series have reported either no aortic injuries or only sporadic cases,9, 10, 11 and a review of 6 816 thoracolumbar pedicle screws found that only 0.22% encroached upon a major vascular structure.8 These data likely underestimate the true burden, given the potential for delayed presentation.4,8,12,13 Although technological advances such as computer assisted navigation, robotic guidance, and intra-operative three dimensional imaging have improved pedicle screw placement accuracy,3,14 most studies focus on intrapedicular precision and orthopaedic outcomes rather than the vascular consequences of anterior cortical breach. Consequently, structured vascular risk stratification in screw–aorta interaction remains insufficiently defined.15 Despite their low prevalence, pedicle screw–aortic interactions are clinically relevant and often affect young patients, raising concerns about long term vascular outcomes. Acute aortic perforation may cause haemorrhage or haemodynamic compromise, whereas injury may also present insidiously with pseudoaneurysm formation.1,16,2,17 This is supported by a bovine model showing that chronic screw–aorta contact may induce medial thinning.23 CTA remains the imaging modality of choice in suspected PS–aorta interaction, whereas prone position imaging has also been proposed to assess aortic mobility and potential wall adherence.25,26 Because absolute measurements may be affected by metal artefact or dynamic aortic displacement, aortic wall deformation was considered a more reliable marker of clinically relevant interaction. Moreover, intravascular ultrasound (IVUS) has also been proposed as an intra-operative support for real time screw–aorta relationship confirmation.29 Presentation is usually incidental and often asymptomatic. When present, symptoms are non-specific, such as back pain, abdominal discomfort, dyspnoea, or low grade fever.24 Vertebral levels T4–T8 may be particularly vulnerable because spinal and aortic relationships can change with position.12 Management remains controversial in the literature, with some authors18,19 advocating close radiological surveillance based on reassuring follow up experiences, whereas others20, 21, 22 recommend proactive removal due to the potential for late vascular complications.6,9,10,27,28 Unlike previous reports, which mainly distinguish between acute rupture and chronic contact,6,9,12,17 this algorithm prioritises aortic wall morphology by differentiating between AC and AI, the latter reflecting true aortic wall deformation and potential progression to vascular injury. This approach closely reflects current European Society for Vascular Surgery guidelines for management of descending thoracic and thoraco-abdominal aortic diseases,35 where management of blunt thoracic aortic injury is guided by lesion morphology rather than proximity alone. In the current practice, a referral threshold defined by a minimum screw–aorta distance ≤ 5 mm on CTA was established in collaboration with spine surgeons, considering any part of the screw, not exclusively the tip. This cutoff was adopted as a pragmatic safety criterion in the absence of validated distance based guidelines and requires future external validation. This threshold is in line with Valič et al., who proposed pedicle screw removal whenever a malpositioned screw lies close to the thoracic aortic wall, particularly when the screw–aorta distance is ≤ 5 mm.4 Screw design features such as tip configuration, thread profile, and self-tapping vs. non-self-tapping geometry may influence vessel interaction and the risk of cortical breach,30, 31, 32 although current techniques mainly address pedicle accuracy rather than vascular risk. In this series, homogeneous screw morphology precluded analysis of design related effects, and further studies are needed to clarify this aspect. According to the algorithm, confirmed aortic indentation was managed with prophylactic TEVAR before screw removal, which is consistent with previous reports.2,5 The authors recognise that a defensible strategy is screw removal under serial angiographic control, reserving endograft deployment for cases of demonstrated bleeding. However, pre-emptive coverage was favoured because deploying the endograft before hardware manipulation provides immediate aortic protection against sudden, potentially catastrophic haemorrhage, and because wall deformation may reflect underlying injury not visible on angiography. This benefit must, however, be weighed against the burden of TEVAR in a predominantly young cohort, such as the risks of retrograde dissection, stent graft induced new entry (SINE), spinal cord ischaemia, the commitment to antiplatelet therapy, and the associated surveillance requirements.33,34 The authors selected short, conformable endografts (CTAG) to minimise aortic coverage and spinal cord ischaemia risk, and observed no SINE, retrograde dissection, or neurological complications; nonetheless, these risks remain central considerations. Prophylactic TEVAR is therefore regarded as a reasonable option for aortic indentation, but requires additional follow up data to support this approach. In the absence of wall deformation, isolated screw removal with completion angiography under serial angiographic control was sufficient, with immediate endovascular bailout available. Procedures were performed in a hybrid operating room with surgical femoral access, ensuring rapid vascular control if needed. No ambiguous case required defaulting to AI in this study; however, this precautionary rule could favour over treatment in larger cohorts and should be weighed against the morbidity of unnecessary TEVAR. Regardless of the morphological category, it is recommended that screw removal procedures be performed in the presence of a vascular surgeon, with prophylactic exposure and cannulation of the left common femoral artery, to ensure immediate endovascular control in case aortic injury becomes apparent during hardware manipulation. Post-operative surveillance remains undefined. Given the young age of this population, routine long term CTA raises concerns regarding cumulative radiation and contrast exposure. In the absence of evidence based protocols, follow up was individualised, balancing clinical vigilance with imaging minimisation. Further studies are required to clarify the natural history and optimise surveillance strategies. Limitations This study was limited by its retrospective, single centre design and the small sample size, reflecting the rarity of the condition. As patients were referred following imaging suspicion, the true incidence of pedicle screw–aortic interaction could not be determined. All assessments were based on CTA, which, despite standardised thin slice acquisition and multiplanar reconstructions, may be influenced by metal artefacts and physiological aortic motion. Classification relied on independent evaluation by two authors, with discordant cases resolved by senior author consensus. Furthermore, the sample was too small for stable reliability estimation, and external validation of reproducibility on a larger image set is required before the classification can be recommended for general clinical use. In addition, follow up was not uniform across treatment groups: patients undergoing TEVAR received structured imaging surveillance, whereas those managed without TEVAR were primarily followed clinically, with imaging performed selectively. Although midterm outcomes were reassuring, longer follow up is required, particularly in this young population, to assess durability and better define the natural history of these lesions. Conclusion Pedicle screw–aortic interaction is a rare but potentially life threatening condition requiring multidisciplinary management. A morphology based classification enables consistent risk stratification and tailored treatment. Isolated screw removal with vascular control appears potentially safe in the absence of aortic deformation, whereas prophylactic TEVAR provides controlled aortic protection when wall indentation is present. This standardised approach appears feasible, but larger studies with longer follow up are needed to validate this strategy and refine surveillance in this predominantly young population. Conflict of interests M.G. and E.G. are consultants for Cook Medical for fenestrated/branched endovascular aneurysm repair. The other authors declare that they have no competing interests.