Dose Optimization for Rubidium PET Imaging in Patients With Known or Suspected Ischemic Heart Disease

Background Selection of the appropriate administered activity for each patient's body habitus is very important to obtain diagnostic image quality. Current SPECT imaging guidelines suggest "…an effort to tailor the administered activity to the patient's habitus and imaging equipment should be made… [however] strong evidence supporting one particular weight-based dosing scheme does not exist." [Henzlova JNC 2016]. An increase in body weight leads to higher fractions of attenuated and scattered photons, resulting in lower quality PET images for a given injected activity [Cherry 2004, Ghanem JNMT 2011]. Weight-based tracer dosing is commonly recommended as a solution in whole-body PET imaging with F-18-FDG [Masuda JNM 2009, Boellaard EJNMMI 2010]. In contrast, Rb-82 PET imaging has traditionally been performed using a single dose (e.g. 40 mCi) administered for all patients [Tout NMC 2012] but this is known to result in lower count-density and image quality in larger patients. This effect can be mitigated to some degree by administration of Rb-82 activity as a proportion of body weight while maintaining accuracy for the detection of disease [Kaster JNC 2012].

The most recent European Association of Nuclear Medicine (EANM) guidelines recommend Rb-82 dosing for 3D PET imaging at 10 MBq/kg [Sciagrà EJNMMI 2020] although the American Society of Nuclear Cardiology (ASNC) still accepts the use of a single dose ranging from 740 to 1110 MBq (20-40 mCi) depending on the PET-CT device sensitivity [Dilsizian JNC 2016]. The ASNC lower limit of 740 MBq may not allow adequate dose reduction in very small or pediatric patients, and conversely the upper limit of 1480 MBq may not allow adequate image quality in the largest patients. For whole body FDG PET, weight-based dosing as a linear function of patient weight (MBq/kg) still does not result in uniform image quality across all patients [Nagaki JNMT 2011]. Recent oncology PET studies have suggested the dose of F-18-FDG be administered as a quadratic function of weight [de Groot EJNMMI Res 2013] and demonstrated that uniform quality of PET images can be maintained across a wide range of patient weights [Musarudin IJNM 2019].

Our center has for many years used weight-based dosing as a linear function of body weight (9-10 MBq/kg) to reduce variations of image quality depending on body habitus, and to reduce detector saturation during the tracer first-pass for accurate blood flow quantification [Renaud JNM 2017a]. Despite this approach, larger patients still suffer from reduced counts and image quality [Renaud JNM 2017b].

Objective To determine whether Rb-82 activity administered as a squared function of patient weight (quadratic dosing) can standardize PET myocardial perfusion image quality over a wide range of body weights.

Primary Hypothesis

1. Rb-82 PET perfusion image quality is consistent across a wide range of patient body sizes when using quadratic dosing of Rb-82. Secondary Hypothesis
2. Administered activity of Rb-82 is consistently accurate over a wide range of injected doses prescribed from 100 to 3500 MBq.

Patient Population

Sequential patients referred for dipyridamole stress Rb-82 PET perfusion imaging at the University of Ottawa Heart Institute. Patients will be divided into 4 weight groups to determine if there are significance differences in image quality or accuracy of injected Rb-82 activity between patients with:

1. 30kg ≤ Weight < 70kg ii. 70kg ≤ Weight < 110kg iii. 110kg ≤ Weight < 150kg iv. 150kg ≤ Weight < 190kg Because patients referred to uOHI generally fall within the lower 3 groups, initial subjects will be identified in the highest weight group, and then those nearest in time within the 3 lower weight groups to avoid bias over time.

Image Analysis Methods ECG-gated stress PET images will be identified from patients referred for Rb-82 MPI on a Siemens Vision 600 PET-CT scanner. Myocardium signal will be measured as the maximum activity in the left ventricle (LVMAX) at end-diastole (ED). Corresponding background signal and noise will be measured as the left atrium blood cavity mean and standard deviation (BLMEAN and BLSD). Image quality will be assessed as myocardium signal-to-noise ratio (SNR = LVMAX / BLSD) and myocardium-to-blood contrast-to-noise ratio (CNR = (LVMAX - BLMEAN) / BLSD).

Statistical Analysis Twelve (12) patients will be recruited in each of the 4 weight groups (3 in each 10 kg interval) to uniformly sample the full range of patient weights from 30 to 190 kg. Based on the previous oncology PET literature [de Groot EJNMMI Res 2013] image quality is not expected to change as a function of weight, i.e. SNR and CNR will be proportional to weight0 (no weight-dependence) with quadratic dosing of Rb-82. Two operators will perform the PET image analysis as described above. Measurements of LVMAX, BLMEAN, BLSD, SNR and CNR will be compared between operators using Bland-Altman and Box-plot analyses. The mean values between operators will be used in the final analyses of weight-based effects. SNR and CNR will be fit to power functions of patient weight Beta, and the Beta coefficients will be compared against the expected value of zero. If the primary hypothesis is true, then the Beta coefficients will not be significantly different from zero (P>0.05) indicating that image quality is not significantly affected by patient weight. N=12 subjects per group is sufficient to detect an effect-size equal to the within-group standard deviation (α=0.05, β=0.03) using single-factor ANOVA.