Diffusion tensor imaging (DTI) and tractography have increasingly been integrated into stereotactic radiosurgery (SRS) workflows to enhance visualization of eloquent neural pathways and to assess treatment-related microstructural changes. The specific roles of DTI in tractography-guided treatment planning, dosimetric optimization, clinical outcomes, and prognostication across intracranial SRS indications have not been comprehensively characterized. This study aimed to systematically review the literature and perform meta-analyses where appropriate to quantify feasibility, dosimetric impact, clinical outcomes, and potential prognostic value of DTI in intracranial SRS.
A systematic review was conducted according to PRISMA 2020 guidelines and prospectively registered in PROSPERO (CRD420261349539). The authors searched MEDLINE/PubMed, Scopus, Web of Science, and the Cochrane Library without date restriction for studies evaluating diffusion MRI or tractography in intracranial SRS. Two independent reviewers screened studies, extracted data, and applied risk-of-bias assessments using the Newcastle-Ottawa Scale where applicable. Random-effects meta-analysis was performed when three or more studies reported comparable outcomes.
A total of 42 studies met inclusion criteria, representing 1,464 patients. The included literature split evenly between studies focused on tractography-guided planning (21 studies) and studies evaluating DTI-derived quantitative metrics (21 studies), which included voxel-wise and region-of-interest (ROI) analyses.
Feasibility of tractography reconstruction was explicitly reported in 22 studies. Across those reports the overall tractography reconstruction failure rate was low, at 1.2%, indicating that tractography generation for SRS planning is achievable in the vast majority of reported cases.
Meta-analysis of comparative clinical studies assessed the incidence of motor deficits with tractography-guided planning versus conventional planning. The pooled analysis suggested a lower risk of motor deficits when tractography was used during planning, with a reported risk ratio of 0.31 (95% CI 0.14–0.67) and no observed heterogeneity (I2 = 0%). This finding supports a clinical advantage in preserving motor function when tractography information is incorporated into SRS planning decisions.
Dosimetric studies comparing tractography-guided planning to conventional approaches were pooled in a meta-analysis. Results indicated a reduced maximum dose to the corticospinal tract with tractography-guided planning (mean difference −3.35 Gy; 95% CI −4.83 to −1.87), with low heterogeneity (I2 = 6%). Importantly, this dosimetric improvement was reported without significant loss of target coverage, implying that tractography-guided optimization may spare critical white-matter tracts while maintaining therapeutic dose to the lesion.
Among studies assessing DTI-derived biomarkers, pooled data in the trigeminal neuralgia subgroup indicated that responders demonstrated greater fractional anisotropy (FA) reduction than non-responders. The reported mean difference in FA reduction was −13.70% (95% CI −25.09 to −2.30) with no heterogeneity (I2 = 0%). These results suggest that quantitative diffusion metrics may have prognostic value for treatment response in selected SRS indications, notably trigeminal neuralgia.
The assembled evidence from 42 studies and 1,464 patients indicates that integration of DTI and tractography into intracranial SRS planning is feasible and associated with both dosimetric and clinical advantages. Tractography-guided planning was linked to lower maximum dose to the corticospinal tract and a reduced risk of post-procedure motor deficits, without compromising target coverage in the studies analyzed. Additionally, DTI-derived diffusivity metrics, such as FA, may offer prognostic information in select conditions like trigeminal neuralgia.
The authors conclude that current evidence is consistent with the utility of tractography-guided planning for eloquent tract preservation and supports further consideration of quantitative diffusion metrics for earlier evaluation of treatment response in targeted indications.
The review authors state that this work was a systematic review and meta-analysis of previously published literature and did not involve new interactions with patients or animals; institutional review board approval and informed consent were therefore not required. The review was registered in PROSPERO and conducted per PRISMA 2020. The authors declared no competing interests and reported that generative AI was used only for language refinement and editorial assistance; no AI tools were used for data analysis or interpretation.
Notes: The details presented here reflect the data and summary results as reported in the source abstract and metadata. Additional study-level methods, numerical results beyond pooled estimates, and individual study characteristics were not provided in full in the source abstract.