Initial post-injection imaging appearance of the reservoir site is indicative of successful access during nuclear medicine (NM) CSF shunt scintigraphy procedures.

Objectives To determine the significance of the appearance of the radiotracer at the injection site during initial NM CSF shunt imaging as a predictor of successful shunt access.

Methods With IRB approval, three NM physicians performed a retrospective review of all consecutive CSF shunt studies performed in pediatric patients at our institution in 2015. Antecedent NM CSF shunt studies in these same patients were also assessed and included in the review. The appearance of the reservoir site immediately after radiotracer injection was classified into one of three configurations: “figure of 8”, “comet tail”, or “round”. The presence or absence of early systemic distribution of the tracer in the thyroid, heart, or stomach on the 5-minute static images was noted and separately evaluated as a predictor of successful CSF shunt access.

Results A total of 101 NM CSF shunt studies were evaluated in 39 patients. 20 patients had one study, 7 patients had 2 studies, 12 patients had 3 or more studies. In 97% of these studies, a ventriculoperitoneal shunt was present while in 3% a ventriculoatrial shunt was present. In 88 studies (87%), either the “figure of 8” or “comet tail” configuration was identified and, when present, had 93% sensitivity, 78% specificity, 92% accuracy, and 98% positive predictive value (PPV) as an indicator of successful shunt access. The round configuration was identified in 13 studies, with 54% PPV for extravasation at the injection site. This low PPV was due to the presence of confounding variables such as injection into the valve, or obstruction of the diversionary shunt at the proximal limb, at the valve, or at both levels. Lack of early systemic distribution of the radiotracer had 96% sensitivity, 78% specificity, 94% accuracy, and 98% PPV as a predictor for successful shunt access. Early systemic distribution of the radiotracer had a 78% sensitivity, 96% specificity and 64% PPV for extravasation at the injection site, being confounded by reflux of radiotracer into the ventricles with resultant absorption into the systemic circulation, and expected early systemic activity in cases with VA shunt. A round configuration in conjunction with the presence of early systemic tracer activity had 100% specificity and 100% PPV for extravasation at the injection site which results in a non-diagnostic study.

Conclusions The configuration of the radiotracer at the injection site on initial post-injection imaging is highly predictive of successful CSF shunt access. Specificity for predicting successful shunt access increases when this data is used in combination with an assessment of the presence or absence of early systemic radiotracer activity.