Are dorsal column deficits inevitable in intramedullary spine tumor resection?

Objective: The rate of dorsal column deficit after intramedullary spine tumor resection remains unknown. In a cohort of patients undergoing posterior intramedullary spinal tumor resection, the authors sought to 1) report the rate of dorsal column deficits, 2) report the rate of new motor deficits, and 3) determine risk factors and recovery characteristics for both deficits.

Methods: A single-institution, retrospective cohort study of patients undergoing posterior intramedullary spinal cord tumor resection was performed from 2010 to 2020. Primary and metastatic cord tumors were included; cauda equina tumors and patients not undergoing midline myelotomy were excluded. Exposure variables included posterior midline myelotomy, tumor location within the spinal cord, spinal segment, tumor size, presence of cord edema on preoperative imaging, and complete/partial resection. The primary outcome was dorsal column deficit (new numbness/tingling, diminished fine touch, vibration, or 2-point discrimination, or balance/proprioceptive problems). Descriptive statistics were performed.

Results: Of 34 patients undergoing intramedullary tumor resection, the mean ± SD age was 44.4 ± 12.1 years and 55.9% were male. Histological analysis showed that 22 (64.7%) patients had ependymoma, 3 (8.8%) astrocytoma, 2 (5.9%) glioblastoma, 2 (5.9%) low-grade glioma, 2 (5.9%) lung adenocarcinoma, and 1 (2.9%) each of hemangioblastoma, lipoma, and necrosis possible neoplasm. Tumor locations were cervical in 17 (50.0%) patients, thoracic in 15 (44.1%), and thoracolumbar/conus in 2 (5.9%). Complete resection was achieved in 19 (55.9%) patients. At presentation, 26 (76.5%) patients had dorsal column deficits and 17 (50%) had motor deficits. Deficits worsened in 17/26 patients with preoperative motor deficits, remained the same in 9, and improved in 0. In 8 patients without preoperative dorsal column deficits, 7 had new dorsal column deficit and 1 had none. A postoperative dorsal column deficit was seen in 33/34 (97%) patients. At the most recent follow-up, 6/33 (18%) patients had improvement in dorsal column deficits and 27/33 (82%) had stable deficits.

Conclusions: Postoperative dorsal column deficits occurred in 97% patients who underwent midline myelotomy for intramedullary tumor resection. Dorsal column deficits improved in 18% yet remained stable in the remaining 82%. Motor deficits improved in only 12% and remained stable/worsened in 88%. These results highlight the high rate of dorsal column deficits in patients requiring midline myelotomy for resection of intramedullary tumors.