Temporary pre-scan proton pump inhibitor cessation reduces gastric tracer uptake in patients undergoing rb-82 PET myocardial perfusion imaging
S Kouadria, R Estorque, C Orongan, S Lousada, N Fernandes, I Badreddine, T Hussain, L Menezes, G B CollinsAbstract
Background/Introduction
Rb-82 PET myocardial perfusion imaging (MPI) is widely used to investigate angina and coronary artery disease. However, gastric tracer uptake is the most frequent image artifact and can hinder accurate contouring of inferior myocardial segments and therefore left ventricular ejection fraction quantification. Also, unlike during SPECT MPI, the short half-life of Rb-82 precludes image reacquisition after interventions to reduce gastric uptake, such as eating, drinking, and delayed imaging. Proton pump inhibitors (PPIs) such as omeprazole are associated with high gastric uptake.
Purpose
This study aimed to assess the relationship between PPI use and gastric uptake and determine whether temporary pre-scan PPI cessation reduces gastric uptake during Rb-82 PET MPI.
Methods
A serial cross-sectional study was conducted. The pre-intervention audit included 9,588 Rb-82 PET scans (62% male, mean age 65.1 ± 12.0y). After introducing temporary pre-scan PPI cessation, a post-intervention reaudit of 612 scans (66% male, mean age 64.8 ± 11.3y) was performed. The primary outcome, ‘high gastric uptake’, was measured with rules-based text mining of reports and validated on a 200-scan subset by comparing text-mined classifications with blinded 4-point visual intensity scoring. Statistical tests included chi-square, t-tests, and odds ratios (OR).
Results
Visual validation of the scan images confirmed reliability of the text-mining method in identifying gastric uptake. Specifically, scans reporting gastric uptake (n = 39) had higher mean visual scores (3.87/4) than those that did not (n = 161, 2.96/4) (p<0.0001). Visual scores were higher in PPI users (3.39/4) than non-users (2.87/4) (p = 0.001). In the pre-intervention audit, 54.9% of patients (n = 5258) were taking a PPI. Gastric uptake was significantly more common in PPI users (15.1%; 794/5,258) than non-users (7.9%; 343/4,330) (OR 2.07, 95% CI 1.81–2.36; p < 0.001). This was a class effect, with no differences between omeprazole, lansoprazole, pantoprazole, or esomeprazole (p = 0.76). Male sex was an independent risk factor for gastric uptake both on (OR 1.61, p < 0.001) and off PPIs (OR 2.19, p < 0.001). In the post-intervention reaudit, adherence to PPI cessation was 37.8% (150/396) and associated with significantly lower rates of reported gastric uptake (24.0%) than non-adherence (33.3%) (OR 0.63; p = 0.048), yielding a NNT for pre-scan PPI cessation = 11.
Conclusion
PPI use and male sex are strong, independent predictors of significant gastric uptake on Rb-82 PET scans. Temporary PPI cessation before Rb-82 PET MPI reduces gastric uptake. This large serial cross-sectional study is the first to demonstrate that PPIs significantly increase gastric uptake on Rb-82 PET MPI, that temporary pre-scan PPI cessation reduces this artifact, and that the effect is consistent across all PPIs, with male sex emerging as a novel independent risk factor.