Surgical treatment of complex spinal cord lipomas.

Purpose: This paper shows the long-term benefits of total/near-total resection of complex spinal cord lipomas and meticulous reconstruction of the neural placode, and specifically, its advantage over partial resection, and over non-surgical treatment for the subset of children with asymptomatic virgin lipomas.

Methods: The technique of total resection and placode reconstruction, together with technical nuances, are described in detail. We added 77 patients with complex lipomas to our original lipoma series published in 2009 and 2010, to a total of 315 patients who had had total or near-total resection and followed for a span of 20 years. Long-term outcome is measured by overall progression-free survival (PFS) with the Kaplan-Meier analysis, and by subgroup Cox proportional recurrence hazard analysis for the influence on outcome of 4 predictor variables of lipoma type, presence of symptoms, prior surgery, and post-operative cord-sac ratio. These results are compared to an age-matched, lesion-matched series of 116 patients who underwent partial lipoma resection over 11 years. The results for total resection is also compared to two large published series of asymptomatic lipomas followed without surgery over 9 to 10 years, to determine whether prophylactic total resection confers better long-term protection over conservative treatment for children with asymptomatic lipomas..

Results: The PFS after total resection for all lipoma types and clinical subgroups is 88.1 % over 20 years versus 34.6 % for partial resection at 10.5 years (p < 0.0001). Culling only the asymptomatic patients with virgin (previously unoperated) lipomas, the PFS for prophylactic total resection for this subgroup rose to 98.8 % over 20 years, versus 67 % at 9 years for one group of non-surgical treatment and 60 % at 10 years for another group of conservative treatment. Our own as well as other published results of partial resection also compare poorly to non-surgical treatment for the subset of asymptomatic virgin lipomas. Multivariate subgroup analyses show that cord-sac ratio is the only independent variable that predicts outcome, with a 96.9 % PFS for ratio < 30 % (loosest sac), 86.2 % for ratio between 30 and 50 %, and 78.3 % for ratio > 50 % (tightest sac), and a threefold increase in recurrence hazard for high ratios (p = 0.0009). Pre-operative patient profiling using multiple correspondence analysis shows the ideal patient for total resection is a child less than 2 years old with a virgin asymptomatic lipoma, who, with a PFS of 99.2 %, is virtually cured by total resection.

Conclusion: Total/near-total resection of complex lipomas and complete reconstruction of the neural placode achieves far better long-term protection against symptomatic recurrence than partial resection for all lesions; and for the subset of asymptomatic virgin lipomas, also better than non-surgical treatment. Partial resection in many cases produces worse outcome than conservative treatment for asymptomatic lesions.