Fluoroscopy-guided intrathecal lumbar drain placement in severe post-polio spinal deformity: A case report

Abstract

Spontaneous cerebrospinal fluid (CSF) rhinorrhoea undergoing endoscopic endonasal repair often requires intraoperative intrathecal lumbar drain (ITLD) placement for fluorescein-assisted fistula localisation and postoperative CSF diversion. Severe spinal deformity can render conventional landmark and ultrasound-guided neuraxial access impossible. We describe a 45-year-old female with post-polio thoracolumbar scoliosis and vertebral rotation undergoing endoscopic repair of a cribriform plate CSF leak, in whom both landmark and ultrasound-guided ITLD placement failed because of absent acoustic windows. Fluoroscopy revealed an apparent lateral image on the standard antero-posterior projection, reflecting near 90-degree axial vertebral rotation.A patient-specific 30-degree oblique C-arm angulation restored a functional antero-posterior-equivalent view, enabling successful catheter placement after fine-needle confirmation. Fluorescein-guided endoscopic repair was subsequently completed uneventfully, with no neurological or CSF-leak-related complications. This case highlights a reproducible, stepwise rescue algorithm-landmark, ultrasound, and patient-tailored fluoroscopic oblique angulation for safe neuraxial access in severely deformed spine, applicable across diverse difficult neuraxial access scenarios.

INTRODUCTION

Spontaneous cerebrospinal fluid (CSF) rhinorrhoea from skull-base defects carries a cumulative meningitis risk of 10–25% per year without repair.[1] The cribriform plate is the single most common site of such defects, accounting for approximately 65% of cases in a recent tertiary-centre series, although spontaneous (non-traumatic) leaks overall represent fewer than 4% of all CSF leaks.[2] Endoscopic endonasal closure achieves success rates exceeding 90%,[3] and an intrathecal lumbar drain (ITLD) facilitates intraoperative fluorescein localisation of the fistula and postoperative CSF diversion.[4] In post-polio syndrome, scoliosis and vertebral rotation distort the posterior bony anatomy relied upon by landmark and ultrasound-guided neuraxial techniques.[5] We report a case in which both approaches failed and a sequential fluoroscopic strategy with a patient-tailored oblique C-arm projection achieved successful ITLD placement.

CASE REPORT

A 45-year-old female (weight 52 kg, American Society of Anesthesiologists physical status III) presented with a 1-year history of right-sided position-dependent clear nasal discharge confirmed as CSF by beta-2 transferrin assay. Computed tomography (CT) cisternography and magnetic resonance imaging (MRI) demonstrated a right cribriform plate defect. She had para-polio with thoracolumbar scoliosis and vertebral rotation, was Mallampati class II, and had documented mild baseline lower-limb weakness. Written consent was obtained including for fluoroscopic exposure.

Anaesthesia was induced with target-controlled infusion (TCI) propofol (Schnider model, effect-site concentration [Ce] 3.5–4.0 µg/mL), fentanyl 2 µg/kg, and rocuronium 0.6 mg/kg.[6] A reinforced metallic endotracheal tube (7.0 mm internal diameter) was selected for kinkresistance during supine-to-lateral-to-prone repositioning; an unchanged end-tidal carbon dioxide waveform confirmed airway integrity throughout. Propofol TCI (Ce 3–3.5 μg/mL) was employed for maintenance with pressure-controlled volume guaranteed ventilation.

Following securing of the airway, ultrasound-guided ITLD placement in the right lateral decubitus position was abandoned. Systematic interrogation of the L1–L2, L2–L3, and L3–L4 lumbar levels in both paramedian sagittal-oblique and transverse planes revealed no identifiable acoustic window at any levelas rotational scoliosis abolished acoustic windows at all lumbar levels. The patient was repositioned prone on a Wilson frame. The initial antero-posterior (AP) C-arm projection produced an apparent lateral image before and after the catheter was placed [Figure 1] due to extreme axial rotation of thoraco-lumbar spine, approximately 90 degrees within the AP beams. The C-arm was angulated to a patient-specific 30-degree right oblique projection before and after placement of the catheter [Figure 2], aligning the interlaminar spaces and revealing the target entry point. Under real-time fluoroscopy, a 23G Quincke needle via a paramedian approach yielded free flow of CSF, confirming intrathecal placement before large-bore instrumentation. Intrathecal placement was achieved within 2–3 combined needle passes of the 23G finder needle and 14G introducer after C-arm angulation (exact pass count could not be recorded). The total fluoroscopic screening time for the entire lumbar drain procedure was approximately 30 minutes, comprising 10–12 discrete pulsed fluoroscopic exposures rather than continuous screening. A 14G catheter introducer followed along the identical trajectory, the catheter was threaded 3–4 cm cephalad, and its radiopaque tip confirmed within the thecal sac. Drainage was set to 5–10 mL/hour (maximum 150 mL/day).

Figure 1.

Standard antero-posterior (AP) C-arm projection of the lumbar spine (a) Severe post-polio rotational scoliosis of lumbar vertebrae approximately 90 degrees within the AP beam, producing an apparent lateral image (b) AP view of lumbar vertebrae with the intrathecal catheter (radiopaque) is seen in situ within the thecal sac

Figure 2.

(a) Fluoroscopic confirmation of the 23G Quincke spinal needle tip; (b) Radiopaque intrathecal lumbar drain catheter tip confirmed within the thecal sac (a) 23 G Quincke spinal needle within the intrathecal space in the 30-degree oblique fluoroscopic projection. Free flow of clear cerebrospinal fluid confirmed intrathecal placement, the mandatory safety checkpoint before 14G introducer needle insertion. (b) Radiopaque intrathecal lumbar drain catheter within the thecal sac, threaded 3–4 cm cephalad. The deformed vertebral architecture from post-polio scoliosis is evident in both panels

Repositioned supine, the patient received intrathecal fluorescein (0.1 mL of 10% solution in 10 mL autologous CSF) 20–30 minutes before endoscopy.[4] Blue-light endoscopy confirmed staining at the right cribriform plate defect. Multilayer repair used autologous fat, fascia lata underlay, a nasoseptal flap, and fibrin glue.[7] The operative time was 165 minutes; blood loss was under 100 mL. No new lower-limb neurological deficit was observed postoperatively. The catheter was removed at 48 hours, and the lumbar puncture site was monitored thereafter for postural headache and signs of a CSF leak, none of which developed. The patient was discharged on day 5.

DISCUSSION

Ultrasound is the established first-line modality for difficult neuraxial access[8]and was accordingly attempted first in this patient, but its efficacy depends on posterior acoustic windows, which may be entirely absent in severe rotational deformity.[5,8] Fluoroscopic guidance, in contrast, directly visualises bony anatomy and permits real-time trajectory modification; its use for intrathecal access nonetheless remains rarely described, despite growing recognition in the anaesthesia literature of the broader perioperative challenges posed by the spinal deformity and the value of protocol-driven care in such patients.[9]Achieving intrathecal access via fluoroscopy requires brief prone positioning. Although prone positioning can transiently raise intracranial venous pressure—particularly with abdominal compression, the risk of CSF rhinorrhoea was minimised in our case by keeping the duration short and ensuring the abdomen remained free.

The critical insight in the present case was interpreting the apparent lateral AP fluoroscopic image as a manifestation of extreme axial vertebral rotation and subsequently responding with C-arm angular compensation rather than patient repositioning. A 30-degree oblique projection recreated a functional AP-equivalent view and not an anatomical AP view, a principle from spinal interventional radiology underutilised in neuraxial anaesthesia.[10] The optimal oblique angle is patient-specific and determined empirically, similar to the contralateral oblique (CLO) technique used for interlaminar epidural access, in which the oblique angle is likewise individualised and determined by measured patient anatomy for optimal visualisation of the interlaminar window and spinolaminar line.[11]

For reproducibility, the target interlaminar space in our patient was identified by rotating the C-arm in the axial plane, while referencing the pedicle and inter-laminar window contra-lateral to the intended needle-entry side, and the 30-degree angle was selected empirically as the point of maximal, symmetrical opening of this window. The skin entry point and needle trajectory were marked under intermittent AP-oblique fluoroscopy, and the needle was advanced coaxially (“tunnel view”) towards the target. Midline orientation, rendered unreliable by vertebral rotation, was judged from the symmetry of the pedicle on the oblique projection rather than from surface landmarks or spinous-process alignment.

CT-guided placement, or pre-procedural CT-based virtual trajectory planning, represents an alternative strategy in such severely distorted anatomy and may reduce fluoroscopy time and needle redirections. However, it was not used in this case owing to the lack of intraoperative CT availability and the time-sensitive, combined-procedure setting.

A 23G spinal needle as a confirmatory finder before insertion of the 14G introducer needle is a sound safety precaution in any distorted spine.The intrathecal space may be displaced relative to the expected midline, and fine-needle confirmation before the larger introducer needle reduces neuraxial haematoma and inadvertent nerve root injury.[12] General anaesthesia with neuromuscular blockade facilitates the supine-to-lateral-to-prone sequence required for the combined procedure but abolishes patient-reported paraesthesia or radicular pain as a warning sign of nerve-root injury.Repositioning is admittedly easier to accomplish under light sedation. In our patient, that trade-off was accepted only after landmark and ultrasound techniques had failed, and were offset by real-time fluoroscopic visualisation and fine-needle confirmation as objective substitutes for subjective patient feedback.

Standard radiation safety precautions were followed throughout in our case, including lead aprons and thyroid shields for all personnel, maximal source-to-skin distance, pulsed rather than continuous fluoroscopy, and collimation to the region of interest, consistent with As-Low-As-Reasonably-Achievable (ALARA) principles. Nevertheless, the cumulative fluoroscopic exposure for this procedure (~30 minutes screening time, 10–12 discrete exposures) was substantially lower than that of a single CT scan of the spine and radiation exposure of this order is routinely accepted in interventional pain practice for fluoroscopy-guided epidural steroid injections and vertebroplasty, and is not considered a major additional risk in this context.

CONCLUSION

We describe a rescue algorithm for intrathecal access and catheter placement in severe post-polio spinal deformity: sequential attempts with landmark guidance, followed by ultrasound, and finally fluoroscopy with patient-tailored oblique angulation. Fine-needle intrathecal confirmation before large-bore introducer and catheter placement is a safety checkpoint in a severely deformed spine. As a single case report, this algorithm requires validation in larger case series to confirm its reproducibility and generalisability across differing patterns of spinal deformity.

Declaration of patient consent

Written informed consent was obtained from the patient for publication of this case report, including clinical information and fluoroscopic images. Appropriate measures have been taken to conceal patient identity.

Disclosure of use of artificial intelligence (AI)-assistive or generative tools

No AI writing or analytical tools were used in the preparation of this manuscript.

Author contributions

SMC: Conceptualisation, anaesthetic and procedural management, fluoroscopic technique design, manuscript writing (original draft), critical revision, final approval. SS: Anaesthetic management, data collection; manuscript drafting, final approval. KS: Literature review; manuscript drafting, final approval. PS: Literature review, manuscript drafting, critical revision, final approval.

Conflicts of interest

There are no conflicts of interest.

Funding Statement

Nil.