Presenter(s)
V. Sai Shreya1, V. Shankar2, H. S. Vangipuram3, S. Ghosh4, S. Cholayil5, and D. Arjundas6; 1Dept of General Medicine, Sri Rama Chandra Medical College, Porur, Chennai, India, 2Division of Radiosurgery, Apollo Cancer Centers, Chennai, India, 3Department of Undergradute Education, All India Institute of Medical Sciences, Ansari Nagar, New Delhi, India, 4Dept. of Neurosurgery, Apollo Proton Cancer Center, Chennai, India, 5Apollo Cancer Centers, Chennai, India, 6Chief Neurologist, Mercury Hospital, Chennai, India
Purpose/Objective(s):
Outcome prediction after stereotactic radiosurgery (SRS) for trigeminal neuralgia (TN) remains imprecise. Conventional predictors — including BNI class, neurovascular contact, and dose parameters — lack biological specificity and fail to capture intrinsic nerve pathology. Diffusion-tensor imaging (DTI) enables detection of subtle microstructural injury. To translate these multidimensional metrics into a clinically actionable biomarker, we developed the Trigeminal Response Index (TN-RI) — a composite DTI-based score reflecting demyelination, axonal imbalance, and diffusional heterogeneity.Materials/Methods:
TN-RI = (1 – FA) + RD + MD + (1 – AD_norm)
where AD_norm = AD(affected)/AD(contralateral) to minimize inter-subject variability in axonal diffusivity.
Patients were stratified into minimal/mixed (<2.0), moderate (2.0–2.5), and severe disruption (>2.5). Clinical outcomes were assessed at 12 months and longitudinally over a median 5-year follow-up. Associations between TN-RI and meaningful pain relief (BNI I–IIIb) were evaluated using Spearman correlation, logistic regression, and ROC analysis. Optimal thresholds were identified using Youden’s J statistic. Continuous variables were compared using Mann–Whitney U testing. FA–RD–MD heat-map modeling was performed to visualize the spectrum of microstructural injury.
Results: TN-RI demonstrated strong predictive performance at both early and long-term follow-up. Patients with TN-RI >2.5 achieved high rates of durable pain relief (~90%), sustained through 5 years. Intermediate outcomes were observed in the 2.0–2.5 range, while TN-RI <2.0 was associated with significantly reduced long-term durability (~60%) and correlated with minimal or mixed tractographic pathology. Logistic regression confirmed TN-RI as an independent predictor of favorable outcome. ROC analysis showed high discriminative accuracy. Heat-map modeling revealed a continuous biological gradient consistent with progressive microstructural compromise.
Conclusion: