Evaluation of the Feasibility of Low Dose 18F-FDG PET in Pediatric Oncology: How Low Can It Be?

Objectives 18F-FDG administration for pediatric patients is generally computed in dose formulas using adult activities as reference. The typical administrated 18F-FDG dose in children and adolescents as recommended by North American Consensus Guidelines (NACG) is about 3.7-5.2 MBq/kg (7-10mCi at 70kg), while higher administered activities are commonly used if applying patient BSA and Webster’s formula. Of increasing importance has been whether or not 18F-FDG PET imaging can be performed at the lowest practical dose levels for pediatric patients in particular infants and young children. This study evaluates and demonstrates the feasibility of low dose FDG PET in pediatric oncology patients using both virtual low dose PET simulation and clinical validation PET/CT scans.

Methods Wholebody 18F-FDG (5.9±2.2 MBq/kg or 0.16±0.06 mCi/kg) PET/CT of 39 pediatric oncologic patients was acquired on a time-of-flight PET/CT system (Gemini TF 64) at 75±5 min post FDG injection using 3min/bed. Listmode data was rebinned and reconstructed in 15s, 30s, 60s, 90s, 120s and 180s per bed position to simulate corresponding lower dose PET (1/12, 1/6, 1/3, 1/2, 2/3 and the reference, respectively). Considering that the data rebin is based on discrete count subsets with time gaps between bed positions, 26 recons of each PET (12x15s, 6x30s, 3x60s, 2x90s, 2x120s and 1x180s volumes over all bed positions) were also performed and compared to verify variances and feasibility. VOIs were placed on lesions and normal anatomical tissues. Quantitative assessment (Body mass index (BMI), counts, count density, SUVmax and SNR) and qualitative assessment (blinded image reviews) were performed and compared. Supportive dedicated phantom experiments (hollow spheres varying with volumes and sphere-to-background ratios of Flangeless Jaszczak PET Phantoms) were performed to evaluate and validate the methodology of virtual low dose PET simulation. For validation, low dose FDG PET (~ 92.5MBq / 2.5mCi) of 5 research adolescents were performed to further validate the methodology.

Results While PET image quality appears to be influenced in a multi-factorial way by acquisition frame duration, dose administration level and patient BMI, all lesions are visible even on the 1/12th orginal dose or count density PET scans (15s/bed). A 30%-60% dose reduction from current standard of care dosing can be proposed to maintain equivalent quality and PET quantification with a SUV and SNR variability of <6% depending on BMIs. An optimized BMI-based instead of weight-dependent FDG administration is recommended (40±19MBq for BMI<18.5; 74±26MBq for BMI of 18.5-25.0; 107MBq±19MBq for BMI of 25-30 and 177±56MBq for BMI >30). A linear relationship between the lowest proposed FDG dose vs BMI was also found. Consistent SUVs from 15s, 30s, 60s, 120s and 180s/bed PETs with correlation to the local standard 90s/bed PET (R2 = 1.08, 0.99, 1.01, 1.00 and 0.98) were determined. No significant variance of CR, SUV and SNR were found in normal anatomical regions or 18F-FDG avid focal lesions in regard to different frame durations (p<0.01).

Conclusions This 5-yr study investigated and validated the feasibility of reducing 18F-FDG-PET dose by 30-60% to a substantially lower level than current standard of care practice without compromising qualitative and quantitative image quality. This enables a considerable radiation dose reduction in the pediatric patient population. A BMI-based approach (30 MBq to 222 MBq for BMI of <18.5 to >30) instead of a weight-dependent is recommended for 18F-FDG PET pediatric imaging at the lowest practical dose levels in particular in infants and smaller children. This justifies to challenge the current practice of either just fixed or weight dependent dosing.