¹²³I-MIBG heart-to-mediastinum ratio is influenced by high-energy photon penetration of collimator septa from liver and lung activity.

Abstract

The 123I-metaiodobenzylguanidine (123I-MIBG) late heart-to-mediastinum ratio (H/M) is a well-established prognostic parameter in patients with chronic heart failure (CHF). However, 123I presents imaging problems owing to high-energy photon emission leading to penetration of collimator septa and subsequent reduction in image quality. Most likely this affects the H/M ratio and may subsequently lead to incorrect patient risk classification. In this prospective study we assessed the intrapatient variation in late H/M ratio between low-energy high-resolution (LEHR) and medium-energy (ME) collimators in patients with CHF. Fifty-three patients with CHF (87% male, age 63±8.3 years, left ventricular ejection fraction 29±7.8) referred for 123I-MIBG scintigraphy were enrolled in the study. In each patient, after the administration of 185 MBq I-MIBG, early (15 min after injection) and late (4 h after injection) planar anterior thoracic images were acquired with both LEHR and ME collimators. Early and late H/M ratios were calculated on the basis of the mean count densities from the manually drawn regions of interest (ROIs) over the left ventricle and a predefined fixed ROI placed in the upper mediastinum. Additional ROIs were drawn over the liver and lungs. Liver/lung to myocardium and liver/lung to mediastinal ratios were calculated to estimate the effect of collimator septa penetration from liver and lung activity on the myocardial and mediastinal ROIs. The mean LEHR collimator-derived parameters were lower compared with those from the ME collimator (late H/M 1.41±0.18 vs. 1.80±0.41, P<0.001). Moreover, Bland-Altman analysis showed that with increasing late H/M ratios the difference between the ratios from the two collimator types increased (R2=0.73, P=0.001). Multivariate regression analysis showed that almost 90% of the variation in the difference between ME and LEHR late H/M ratios could be explained by scatter from the liver in both the mediastinal and myocardial ROIs (R2=0.90, P=0.001). Independent predictors for the difference in the late H/M between ME and LEHR were the liver-to-heart ratio and the liver-to-mediastinum ratio assessed by ME (standardized coefficient of -1.69 and 1.16, respectively) and LEHR (standardized coefficient of 1.24 and -0.90, respectively) (P<0.001 for all). Intrapatient comparison in H/M between the ME and LEHR collimators in patients with CHF showed that with increasing H/M the difference between the ratios increased in favour of the ME collimator. These differences could be explained by septal penetration of high-energy photons from both the liver and the lung in the mediastinum and myocardium, being lowest when using the ME collimator. These results strengthen the importance of the recommendation to use ME collimators in semiquantitative 123I-MIBG studies.