MR Fingerprinting Predicts Gait Improvement After Lumbar Cerebrospinal Fluid Drainage for Normal Pressure Hydrocephalus.

BACKGROUND

Normal pressure hydrocephalus (NPH) results in gait disturbance, but clinical response to extended lumbar drainage (ELD) of cerebrospinal fluid varies. Noninvasive predictors of response are needed.

PURPOSE: To determine whether MR fingerprinting (MRF) findings have potential as a quantifiable biomarker for predicting clinical response to ELD in patients with NPH.

STUDY TYPE: Prospective cohort.

POPULATION: Twenty individuals with NPH (mean age 77.7 years; 11 male sex).

FIELD STRENGTH/SEQUENCE: 3T; 2D T1,T1/T2 MRF, 3D T1.

ASSESSMENT: Patients with NPH referred for ELD between August 2020 and February 2023 were recruited and underwent brain MRI and 10 m walking test before and after ELD. Clinical response was defined as improvement in the walking test plus a positive neurosurgical assessment. Quantitative T 1 and T 2 values were obtained using MRF. Whole-brain T 1 and T 2 times were obtained for cerebral white matter (WM) and gray matter (GM). Brain volumetry was derived using a 3DT1-weighted sequence.

STATISTICAL TESTS: Linear discriminants to predict responder status were derived from logistic regression against T 1 and T 2 times; supplementary discriminants from visual inspection (manual delineation) were created as well. Fisher's exact test was applied to resulting contingency tables with significance at p < 0.05. Receiver operating characteristic curves were generated to assess predictors.

RESULTS

There were 11 responders and 9 nonresponders. Pretreatment MRF demonstrated an ability to distinguish responders from nonresponders (areas under the curve: WM T 1, 0.83 ± 0.21; GM T 1, 0.76 ± 0.23; WM T 2, 0.81 ± 0.22; GM T 2, 0.81 ± 0.22). Via logistic regression, GM measures significantly predicted responder status (OR = 14.0 [1.54, 130]) but WM measures did not (p = 0.37). Visual inspection yielded significance for GM (OR incalculable due to zero-valued contingency cell) and WM (OR = 18.7 [1.56, 222]).

DATA

CONCLUSION

Elevated whole-brain T 1 and T 2 relaxation times may serve as a noninvasive predictive biomarker of response to ELD in patients with NPH.

EVIDENCE LEVEL: 2 TECHNICAL EFFICACY: Stage 2.