Identification of Preoperative Language Tracts for Intrinsic Frontotemporal Pathologies: A Pilot Reconstruction Algorithm in a Middle-Income Country.

OBJECT: Diffusion tensor imaging (DTI) tractography provides useful information that can be used to optimize surgical planning and help avoid injury during subcortical dissection of eloquent tracts. The objective is to provide a safe, timely, and affordable algorithm for pre-operative DTI language reconstruction for intrinsic frontotemporal pathologies.

METHODS: The authors reviewed a prospectively acquired database of pre-operative DTI reconstruction for resection of left frontotemporal lesions over three years at Hospital de San José and Hospital Infantil Universitario San José, Fundación Universitaria de Ciencias de la Salud (FUCS), Bogota, Colombia. Pre- and post-operative clinical and radiographic features were determined from retrospective chart review. A comprehensive review of the structural and functional anatomy of the language tracts was performed. Separate reconstruction of both ventral (semantic) and dorsal (phonological) stream pathways is described: arcuate fasciculus, superior longitudinal fasciculus (SLF), inferior frontooccipital fasciculus (IFOF), uncinate fasciculus, and inferior longitudinal fasciculus (ILF).

RESULTS: Between January 2015 and January 2018, a total of 44 tumor cases were found to be resected with preoperative fiber tracking planning and neuronavigation guided surgery. Ten patients (7F, 3M) aged 28-65 years-old underwent resection of an intrinsic frontotemporal lesion utilizing pre-operative DTI tractography reconstruction of language tracts. 8 (80%) cases were high-grade gliomas, and 2 (20%) were cavernous malformations. In 5 (50%) cases the lesion was in the frontal lobe and 5 (50%) in the temporal lobe. The extent of resection was classified as gross total resection (100%), subtotal resection (>90%) or partial resection (<90%). Gross total resection was achieved in 5 (50%) cases, subtotal resection was achieved in 4 (40%) cases, and partial resection in the remaining case (10%). Compromised tracts included SLF in 7 (70%), ILF in 4 (40%), the arcuate fasciculus in 3 (30%), and uncinate fasciculus in 1 (10%). Language function was unchanged or improved in 90% of patients. New-onset postoperative language decline occurred in 1 patient, who recovered transient phonemic paraphasias one month after resection. The mean follow-up time was seven months (range 4 to 12 months). Residual tumors were treated with radiation and/or with chemotherapy as indicated in an outpatient setting.

CONCLUSION: We present a safe and efficacious pre-operative DTI language reconstruction algorithm that could be used as a feasible treatment strategy in a challenging subset of tumors in low- to middle-income countries.