The efficacy and deficiency of contemporary treatment for spinal cord arteriovenous shunts.

Contemporary treatments for spinal cord arteriovenous shunts are only based on clinicians' treatment experiences and expertise due to its rarity. We reviewed the clinical course of the largest multicentred cohort to evaluate the efficacy and deficiency of contemporary interventional treatments for spinal cord arteriovenous shunts. The clinical features, treatment results and clinical outcomes of 463 patients with spinal cord arteriovenous shunts were retrospectively assessed. The main outcome was the neurological deterioration that was evaluated based on the modified Aminoff and Logue scale. According to post-treatment digital subtraction angiography, complete obliteration was defined as disappearance of the intradural lesion, whereas partial obliteration was defined as any residual intradural lesion remaining visible and was further categorized as shunt-reduction obliteration (the nidus or shunt points were reduced) or palliative obliteration (only obliterated aneurysms or feeders). Cure rate was 40.6% for the whole cohort, 58.5% after microsurgery, and 26.4% after embolization. The curative resection was associated with non-metameric lesions, lesions with a maximum diameter <3 cm and lesions without anterior sulcal artery supply. The curative embolization was associated with fistula-type lesions, non-metameric lesions, and main drainage diameter <1.5 mm. The permanent treatment-related neurological deficits rate was 11.2% for the whole cohort, 16.1% after microsurgery, and 5.6% after embolization. The pretreatment clinical deterioration rate was 32.5%/year, which decreased to 9.3%/year after clinical interventions. Following partial treatment, the long-term acute and gradual deterioration rates were 5.3%/year and 3.6%/year, respectively. The acute deteriorations were associated with metameric lesions, craniocervical lesions, lesions with a maximum diameter ≥2 cm and residual aneurysm. Residual aneurysm was the only predictor of acute deterioration for non-metameric spinal cord arteriovenous shunts. The gradual deteriorations were associated with palliative obliteration, absence of pretreatment acute deterioration and intact main drainage. Although clinical risks of spinal cord arteriovenous shunts were reduced following clinical interventions, contemporary treatments for spinal cord arteriovenous shunt remains associated with considerable risks and incomplete efficacy. Individualized treatment plans should be adopted according to the angio-architectural features and major clinical risks of specific lesions. There is a higher opportunity for complete obliteration for lesions with simple angio-architecture. However, for most of spinal cord arteriovenous shunts with complex vascular anatomy, partial treatment is the only choice. For these patients, palliative obliteration targeting the aneurysms is recommended for reducing haemorrhagic risk, whereas shunt-reduction obliteration is necessary for non-haemorrhagic myelopathy. Contemporary treatment is ineffective in reducing haemorrhagic risk of incurable metameric spinal cord arteriovenous shunts.