The impact of acquisition time of planar cardiac 123I-MIBG imaging on the late heart to mediastinum ratio

Purpose The aim of this study was to investigate whether performing the late cardiac 123I-metaiodobenzylguanidine (MIBG) scan earlier than 4 h post-injection (p.i.) has relevant impact on the late heart to mediastinum ratio (H/M ratio) in patients with heart failure (HF). Methods Forty-nine patients with HF (median left ventricular ejection fraction of 31 %, 51 % ischaemic HF) referred for cardiac 123I-MIBG scintigraphy were scanned at 15 min (early) p.i. and at 1, 2, 3 and 4 h (late) p.i. of 123I-MIBG. Late H/M ratios were calculated and evaluated using a linear mixed model with the mean late H/M ratio at 4 h p.i. as a reference. A difference in late H/M ratios of more than 0.10 between the different acquisition times in comparison with the late H/M ratio at 4 h p.i. was considered as clinically relevant. Results Statistically significant mean differences were observed between the late H/M ratios at 1, 2 and 3 h p.i. compared with the late H/M ratio at 4 h p.i. (0.09, 0.05 and 0.02, respectively). However, the mean differences did not exceed the cut-off value of 0.10. On an individual patient level, compared to the late H/M ratio at 4 h p.i., the late H/M ratios at 1, 2 and 3 h p.i. differed more than 0.10 in 24 (50 %), 9 (19 %) and 2 (4 %) patients, respectively. Conclusion Variation in acquisition time of 123I-MIBG between 2 and 4 h p.i. does not lead to a clinically significant change in the late H/M ratio. An earlier acquisition time seems to be justified and may warrant a more time-efficient cardiac 123I-MIBG imaging protocol.


Introduction
Heart failure (HF) is an increasing healthcare problem in the Western world [1]. Among other features, the pathophysiology of HF is characterized by a deregulated sympathetic nervous system [2]. 123 I-Metaiodobenzylguanidine (MIBG) imaging is able to visualize cardiac sympathetic innervation by providing (semi-)quantitative information on the myocardial sympathetic activity [3]. Several studies have reported that cardiac 123 I-MIBG uptake provides unique information for predicting prognosis and evaluating therapeutic effects in patients with HF [4][5][6][7]. The semi-quantitative parameter that is most used to analyse cardiac 123 I-MIBG uptake is the heart to mediastinum ratio (H/M ratio) of the late planar image. The reported acquisition time for this so-called late scan varies considerably, but it is recommended 4 h after injection of 123 I-MIBG [8]. A shorter time interval between the injection of 123 I-MIBG and the late scan reduces waiting time and may facilitate clinical use. However, only a few studies have investigated the impact of variation in acquisition time after 123 I-MIBG administration on the late H/M ratio [9,10]. Therefore, the aim of this study was to investigate whether performing the late scan earlier than 4 h post-injection (p.i.) has relevant impact on the late H/M ratio in patients with HF.

Materials and methods
Patients Fifty patients with HF from two centres (Leiden, The Netherlands and Pisa, Italy) clinically referred for cardiac 123 I-MIBG scintigraphy between 2012 and 2015 were included. The inclusion criteria were symptoms of HF, New York Heart Association functional class II-IV and age between 18 and 74 years. Patients with severe shortness of breath who were unable to lie flat for several times under a gamma camera were excluded from this study. All patients were treated according to the European Society of Cardiology guidelines for the diagnosis and management of HF and therefore received optimal pharmacological therapy, revascularization and implantable defibrillator with or without cardiac resynchronization therapy (CRT) device when appropriate [11]. The study complied with the Declaration of Helsinki and was approved by the Medical Ethics Committee of the Leiden University Medical Center. Written informed consent was obtained from all patients.

I-MIBG data acquisition
After thyroid blockade according to the local protocol and a resting period of 30 min, 185 MBq of 123 I-MIBG (AdreView, GE Healthcare, Princeton, NJ, USA) was administrated as an intravenous bolus. Afterwards, 123 I-MIBG planar imaging was performed in the supine position five times p.i.; the early acquisition was performed 15 min p.i. and the late acquisitions were performed at 1, 2, 3 and 4 h p.i. Planar images were acquired in the anterior view for 10 min with dual-head gamma cameras, a Siemens e.cam (Pisa, Italy) or a Toshiba GCA-7200/PI (Leiden, The Netherlands) and stored in a 256×256 matrix. All camera heads were provided with low-energy high-resolution collimators and the images were obtained with a 15 % energy window centred at the 159 keV photopeak of 123 I.

Planar 123 I-MIBG image analysis
All planar 123 I-MIBG images were transferred to the Leiden University Medical Center for a centralized analysis. The images were semi-quantitatively analysed by calculating the H/M ratio using dedicated post-processing software on a Syngo-MI workstation (Siemens Medical Solutions, Malvern, PA, USA). Manually a polygonal region of interest (ROI) was drawn over the myocardium including the left cavity and a rectangular ROI was placed upon the upper half of the mediastinum using the following anatomical landmarks: the lung apex as upper border, the upper cardiac border and the medial contours of the lungs. The H/M ratio was calculated by dividing the mean count density (i.e. counts/pixel) within the cardiac ROI by the mean count density in the mediastinal ROI. The washout rate was expressed as a percentage of decrease in myocardial counts between early and late planar images with correction for background and physical decay of 123 I. For this calculation the following formula was used: WRBKG corrected=[(He-Me)-(Hl-Ml x 123 I decay correction)]/(He-Me) × 100 [8]. In this formula H represents the mean count density in the cardiac ROI and M represents the mean count density in the mediastinal ROI in the early (e) and (l) late images.

Statistical analysis
Normally distributed variables were expressed as mean ± standard deviation and non-normally distributed variables as median and interquartile range (IQR). Categorical variables were given as absolute values and percentages. Patient characteristics and 123 I-MIBG scintigraphy results of subgroup analyses were compared with the independent sample t test, the Mann-Whitney U test or the chisquare test when appropriate. The late H/M ratios at 1, 2 and 3 h p.i. were compared with the late H/M ratio at 4 h p.i. with a linear mixed model with a given fixed factor time. A mean difference in late H/M ratio of more than 0.10 between the (mean) different H/M ratios at 1, 2 and 3 h p.i. and the (mean) late H/M at 4 h p.i. was considered clinically relevant. This was assessed by evaluating whether the 95 % confidence intervals (CI) for the real differences in means of the late H/M ratios at 1, 2 and 3 h p.i., respectively, compared with the late H/M ratio at 4 h p.i. included the value 0.1. In the subgroup analysis, the influence of left ventricular ejection fraction on the differences in late H/M ratios between the ischaemic and nonischaemic HF groups at 1, 3 and 4 h p.i. was assessed by performing multivariate logistic regression analyses including H/M ratio and left ventricular ejection fraction. The washout rates between 15 min and 1, 2 and 3 h p.i. were compared with the washout rate between 15 min and 4 h p.i. with a linear mixed model with a given fixed factor time. A paired t test power analysis demonstrated that a sample size of 49 patients achieved 90 % power to reject the null hypothesis that the test and standard are equivalent, thereby assuming an absolute difference of more than 0.10 with an estimated standard deviation of 0.20 with a significance level (alpha) of 0.01670. A twosided p value <0.05 was considered statistically significant. All analyses were performed with SPSS software (Version 22.0, SPSS IBM Corp., Armonk, NY, USA).

Patients
The study population comprised 49 patients with HF, because 1 patient withdrew his informed consent after the examination. The patient characteristics of the study population are depicted in Table 1. Seventy-six per cent (37/49) of the total study population was male with a mean age of 64±9 years. The aetiology of HF was ischaemic in 51 % (25/49) of the patients. Patients with ischaemic HF had a statistically significantly lower left ventricular ejection fraction compared with patients with non-ischaemic HF, 28 % (IQR 23-35) versus 33 % (IQR 29-40) (p=0.04), respectively, and antiplatelet or oral anticoagulant therapy was more frequent at 96 % (24/25) versus 75 % (18/24) (p=0.04), respectively. ACE angiotensin-converting enzyme inhibitor, AT-II angiotensin II, BMI body mass index, CRT-D cardiac resynchronization therapy defibrillator, CVD cardiovascular disease, HF heart failure, ICD implantable cardioverter defibrillator, IQR interquartile range, NT-proBNP N-terminal pro-brain natriuretic peptide, NYHA New York Heart Association, OAC oral anticoagulant a Self-reported history of diabetes mellitus and/or treatment with antidiabetics b Presence of coronary artery disease in first-degree family members at age <55 years in men and <65 years in women c Self-reported history of hypercholesterolaemia and/or treatment with lipid-lowering drugs d Self-reported history of hypertension and/or use of antihypertensive medication or a systolic blood pressure ≥140 mmHg and/or diastolic blood pressure ≥90 mmHg  (Fig. 1a). An example of a patient with only small differences between the late H/M ratios at 1, 2 and 3 h p.i.  Figure 3 shows the mean differences in late H/M ratios at 1, 2 and 3 h p.i. compared with the late H/M ratio at 4 h p.i. in both subgroups. Compared to the late H/M ratio at 4 h p.i., the mean late H/M ratios at 1 and 2 h p.i. were statistically significantly different in the ischaemic HF group (both p < 0.001) as well as in the non-ischaemic HF group (p < 0.001 and p = 0.005, respectively). In the non-ischaemic HF group only the mean late H/M ratio at 1 h p.i. exceeded the predefined clinically relevant 0.10 difference with the late H/M ratio 4 h p.i.
The mean washout rate between the early scan and the late scans at 1, 2, 3 and 4 h p.i. increased from 11±12 % to 21± 16 %, 29±17 % and 34±17 %. The mean washout rates between the early scan and the late scans at 1, 2 and 3 h p.i. differed all statistically significantly from the washout rate between the early and 4-h p.i. scan (all p <0.001). There was no statistically significant difference between the washout rates in the ischaemic HF and the non-ischaemic HF subgroups.

Discussion
The present study evaluated the impact of variations in acquisition time of planar 123 I-MIBG scintigraphy on the late H/M ratio in patients with HF. There were statistically significant differences between the late H/M ratios at 1, 2 and 3 h p.i. compared with the late H/M ratio at 4 h p.i. However, compared with the late H/M ratio at 4 h p.i., the  [8]. This recommendation is based on several studies [13][14][15][16][17]. First, Nakajo et al. investigated the appropriate time for cardiac radioiodine-labelled MIBG imaging [14]. This study demonstrated in rats that the accumulation of cardiac 131 I-MIBG in adrenergic nerves reached a maximum plateau at 4 h p.i., while the accumulation of the tracer in other compartments rapidly decreased in a period of 6 h p.i. Second, the mechanism of myocardial uptake of 123 I-MIBG in the presynaptic sympathetic nerve endings was elucidated by DeGrado et al. [15]. This group demonstrated that 123 I-MIBG uptake was mainly between 1 and 7 min p.i. and depended primarily on both uptake-1 (the neuronal pathway via the norepinephrine transporter) as well as on uptake-2 (the extraneuronal pathway) activity. The clearance half-time of neuronal 123 I-MIBG uptake (via uptake-1) was 112 min, in contrast to the much lower clearance half-time of extraneuronal 123 I-MIBG uptake (uptake-2) of 22 min. Finally, Sisson et al. demonstrated that 123 I-MIBG patterns in the heart were consistent with the concept that 123 I-MIBG resides mostly in adrenergic neurons, mainly via the uptake-1 mechanism [16,17]. This supports the hypothesis that the late H/M ratio at 4 h p.i. mainly represents the uptake, storage and release of 123 I-MIBG in the myocardial vesicles at the nerve terminals, while accumulation in extraneuronal compartments is relatively low.
In the present study only small differences (≤0.10) were observed between the late H/M ratios at 1, 2 and 3 h p.i.    [9]. Slow efflux of 123 I-MIBG from the neuronal compartments can probably explain the small differences between late H/M ratios in the latter and the present study [13,18].
The present results show that, at a group as well as at an individual patient level, the change in the late H/M ratios between 2 and 4 h p.i. is limited and probably not clinically relevant. Since the late H/M ratio has been proven to provide important information in predicting prognosis in HF patients [4], our results may have clinical implications by allowing a more flexible and/or shorter interval between the injection of 123 I-MIBG and the acquisition time of the late scan for the determination of the late H/M ratio, especially in patients with HF.
With a shorter interval between the injection of 123 I-MIBG and the Blate^scan the washout rate will change. Previous studies have shown that the instant myocardial uptake of 123 I-MIBG uptake normalized for blood pool activity is close to zero after 3 h [10]. Evaluating the washout rate at least 3 h after radiotracer injection is therefore recommended. However, Arimoto et al. demonstrated in 42 patients with HF a good correlation of the early washout rate between 5 and 15 min and the washout rate between 15 min and 3 h (r=0.606, p<0.0001) [19].
Moreover, the early washout rate provided prognostic information as well. In addition, Henderson et al. demonstrated that a washout rate recorded between 15 and 85 min showed a significant difference in 16 patients with non-ischaemic cardiomyopathy compared with 14 healthy volunteers: 28±12 % and 6±8 %, respectively (p<0.001) [20]. Additional imaging was performed in a subset of patients (n =8) at 4 h p.i. and 123 I-MIBG retention appeared to be lower and showed greater disparity in the washout rate. In the present study, myocardial 123 I-MIBG washout increased over time from 11± 12 % at 1 h p.i. to 34±17 % at 4 h p.i. The mean washout rates between the early scan and the late scans at 1, 2 and 3 h p.i. differed all statistically significantly from the washout rate between the early and 4-h p.i. scan (all p <0.001).

Limitations
It should be addressed that although myocardial washout rates were corrected for 123 I decay and mediastinal background, the myocardial washout rate can still be influenced by 123 I-MIBG washout in surrounding lung and liver tissues [21]. Moreover, more research is needed before implementing the acquisition time at 2 h to avoid misclassification since the difference between the late H/M ratio at 2 h p.i. in comparison with 4 h p.i. at the patient level was not non-negligible.

Conclusion
Variation in acquisition time of 123 I-MIBG between 2 and 4 h p.i. does not lead to a clinically significant change in the late H/M ratio. With the knowledge that the late H/M ratio is the best validated and most used prognostic parameter for cardiac sympathetic innervation in patients with HF, an earlier acquisition time seems to be justified and may warrant a more timeefficient cardiac 123 I-MIBG imaging protocol.

Conflicts of interest None.
Ethical approval All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional and/or national research committee and with the 1964 Declaration of Helsinki and its later amendments or comparable ethical standards.
Informed consent Informed consent was obtained from all individual participants included in the study.
Open Access This article is distributed under the terms of the Creative Comm ons Attribution 4.0 International License (http:// creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made.
Open Access This article is distributed under the terms of the Creative Comm ons Attribution 4.0 International License (http:// creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made.