Characterization of spinal findings in children and adults with neurofibromatosis type 1 enrolled in a natural history study using magnetic resonance imaging.

To characterize spinal abnormalities in patients with neurofibromatosis type 1 (NF1) using magnetic resonance imaging (MRI). NF1 patients with at least one spine MRI were selected from participants prospectively enrolled in the National Cancer Institute NF1 Natural History Study. Data were analyzed retrospectively. Ninety-seven patients (38 females, median age 14.2 years, standard deviation [SD] 7.6) had baseline imaging of the spine, and 26 patients (27 %) had one follow-up spine MRI (follow up time 2.5 years, SD 1.1, range 0.7-4.7). Seventy-eight patients (80 %) had spinal neurofibromas, with rising frequency from 70 % in patients younger than 10 years to 80 % in patients aged 10-18 years to 89 % in individuals older than 18 years of age. At baseline, 33/97 patients (34 %) had MRI changes consistent with spinal cord compression that was most prevalent at the cervical (43 %) and lumbar spine region (40 %). Seven of nine patients with progression of their spinal neurofibromas developed cord compression. Paraspinal plexiform neurofibromas (PNs) were present in 77/97 patients (79 %), of which 68 patients (88 %) had concomitant spinal neurofibromas. Spinal curvature abnormality was present in 50/97 patients (51 %, 20 females, median age 14.6 years, SD 7.6). Patients with paraspinal PNs had six-fold higher odds of developing spinal curvature abnormalities compared to patients without PN (OR = 5.9, 95 % CI 1.81 to 19.44, p = 0.0033). A total of 58/97 patients (60 %, median age 16.1 years, SD 7.8, range 4.8-48.2 years) presented with neurologic abnormalities that progressed in 12/26 patients (46 %). Substantial spinal neurofibroma and paraspinal PN burden was observed in our study population, which represents a selective group of patients with specifically more severe tumor involvement than the general NF1 population. Occurrence and progression of spinal neurofibromas on repeat evaluations highlight the need for longitudinal clinical monitoring in patients with known spinal disease.