Reliability of the radiopharmaceutical shunt flow study for the detection of a CSF shunt malfunction in the presence of stable ventricular size

Abstract

The authors sought to determine the reliability of a radiopharmaceutical (RP) shunt flow study for the detection of a CSF-diverting shunt malfunction in the presence of stable ventricular size. After the authors obtained IRB approval, all CSF RP shunt flow studies done between January 1, 2014, and January 1, 2019, in pediatric patients at Children's Hospital Los Angeles were identified. Included in the study were only those patients in whom an MRI or CT scan was done during the hospital admission for shunt malfunction and showed no increase in ventricular size compared with the most recent prior MRI or CT scan when the patient was asymptomatic. Data recorded for analysis were patient age and sex, etiology of the hydrocephalus, shunt distal site, nonprogrammable versus programmable valve, operative findings if the shunt was revised, and follow-up findings for a minimum of 90 days after admission. The RP shunt flow study consisted of tapping the reservoir and injecting technetium-99m DTPA according to a set protocol. The authors identified 146 RP flow studies performed in 119 patients meeting the above criteria. Four of the 146 RP studies (3%) were nondiagnostic secondary to technical failure and were excluded from statistical analysis. Of the 112 normal flow studies, operative intervention was not undertaken in 102 (91%). The 10 (9%) remaining normal studies were performed in patients who underwent operative intervention, in which 8 patients had a proximal obstruction, 1 had a distal obstruction, and 1 patient had no obstruction. Of the 30 patients with abnormal flow studies, symptoms of shunt malfunction subsided in 9 (30%) patients and these patients did not undergo operative intervention. Of the 21 (70%) operated patients, obstruction was proximal in 9 patients and distal in 5, and for 7 patients the shunt tubing was either fractured or disconnected. Regression analysis indicated a significant association between the flow study interpretation and the odds for shunt revision (OR 27, 95% CI 10-75, p < 0.0001). No other clinical variables were significant. The sensitivity of a shunt flow study alone for detection of shunt malfunction in cases with stable ventricular size was the same as a shunt flow study plus an MRI or CT (70% vs 70%), but performing a shunt flow in addition to MRI or CT did increase the specificity from 92% to 100% and the accuracy from 87% to 94%. RP shunt flow studies were of definite value in deciding whether to operatively intervene in patients with symptoms of shunt malfunction in whom no change in ventricular size was detected on current MRI or CT scans compared to scans obtained when the patients were asymptomatic.