Non-inferiority of a non-contrast MRI follow up protocol for isolated optic pathway gliomas

Ezekiel Maloney | 2018

Advisor: Amanda Phipps

Research Area(s): Clinical Epidemiology, Public Health Practice



Pediatric optic pathway gliomas (OPGs) are typically indolent, but have variable clinical course. Treatment is dictated by symptoms and changes on contrast-enhanced magnetic resonance (CEMR) examinations. Gadolinium retention in children has motivated parsimonious use of gadolinium-based contrast agents (GBCAs).


1) Determine surveillance MR factors that motivate changes in tumor-directed therapies (TDTs), establish GBCA exposure burden for patients with OPG, and extrapolate cost-efficacy of a non-contrast follow-up protocol. 2) Perform a pilot series of blinded interpretations of follow-up MRI examination pairs for OPGs with attention to the tumor characteristics identified in Objective 1, and establish variability estimates to power a non-inferiority trial. 3) Build an appropriately powered and balanced case set and perform a non-inferiority, multi-reader-multi-case study of a non-contrast MRI protocol versus standard contrast-enhanced protocol for follow-up of isolated OPGs with attention to tumor characteristics identified in Objective 1.


An imaging database search identified patients with isolated OPG and ≥3 follow-up CEMRs. Medical records and imaging were reviewed for: 1) coincident changes on CEMR and TDT; 2) demographics and duration of follow-up; 3) motivations for intervention; 4) assessment of GBCAs utility; 5) healthcare utilization data. Cost impact was assessed in terms of Relative Value Unit (RVU) burden. Subsequently, two blinded readers reviewed 16 exam pairs from 8 individual patients and assessed change in tumor size between exams on both non-contrast and contrast-enhanced sequences. Lastly, a series of MRI brain exam pairs from OPG patients was built and was assessed in blinded fashion by 4 of 7 pediatric radiologists; readers assessed changes in tumor size on both non-contrast and contrast-enhanced versions of each exam pair with at least 1 week of memory washout between interpretations.


The 17 neurofibromatosis type 1 (NF1) and 21 non-NF1 patients included in analyses of objectives 1 and 2 underwent a median 16.9 and 24.3 cumulative CEMR exams over 7.7 and 8.1 years of follow-up, respectively. Eight patients (1 with NF1) had intervention based on CEMR findings alone. For these 8, increased tumor size was the only common feature, and was apparent on non-contrast T2 sequences. For the median patient, a non-contrast follow-up protocol could result in 15.9 (NF1) and 23.3 (non-NF1) fewer GBCA administrations, and a 39% reduction in yearly RVU burden. Based on power calculations from the pilot 2-reader assessments of OPG size changes and available MRI brain exams meeting inclusion criteria, 60 exam pairs were compiled from 42 patients with isolated OPG (19 with NF1), and the non-inferiority threshold was set at 12% for intra-reader agreement. Interim analyses from the multi-reader-multi-case study suggest that a non-contrast protocol might be non-inferior to a contrast-enhanced protocol for assessment of change in tumor size in routine follow up MRI exams.


Pediatric patients with isolated OPG undergo a large number of routine CEMR follow-up exams. Gadolinium may not be needed for these exams to inform management decisions. A non-contrast MRI follow-up protocol appears to be non-inferior to a contrast-enhanced protocol for assessment of changes in tumor size based on interim analysis. Secondary benefits of a non-contrast follow-up protocol include decreased cost and risk to the patient.