Radiation Dose Reduction During Adrenal Vein Sampling Using a New Angiographic Noise Reduction Imaging Technology

Purpose: To compare the patient radiation doses during angiographic selective adrenal vein sampling (AVS) before and after a noise reduction imaging technology upgrade. Methods: In this retrospective single-center-study, cumulative air kerma (AK), cumulative dose area product (DAP), uoroscopy time and contrast agent dosage were recorded from 70 patients during AVS. 35 procedures were performed before and 35 after a noise reduction imaging processing technology upgrade. Mean values were calculated and compared using an unpaired student’s t-test. DSA image quality was assessed independently by two blinded readers using a four-point Likert scale (1=poor; 4=excellent) and compared using Wilcoxon signed-rank test. Results: After the technology upgrade we observed a signicant reduction of 35% in AK (1.7±0.7 vs. 1.1±0.7 Gy, p=0.01) and a signicant reduction of 28% in DAP (235.1±113 vs. 170.1±94 Gy*cm 2 , p=0.01) in comparison to procedures before the upgrade. There were no signicant differences between the number of exposure frames (143±86 vs. 132±61frames, p=0.53), uoroscopy time (42±23 vs. 36±18 min, p=0.22), or the amount of contrast medium used (179.5±84 vs. 198.1±109 ml, p=0.41). There was also no signicant difference regarding image quality (3 (2-4) vs. 3 (2-4), p=0.67). Conclusion: The angiographic noise reduction imaging technology upgrade signicantly decreases the radiation dose during adrenal vein sampling without increasing time of uoroscopy or contrast volume and without compromising image quality.


Introduction
Primary aldosteronism (PA) is the most frequent secondary form of hypertension and is characterized by an autonomous adrenocortical over secretion of aldosterone (1). PA has a prevalence of up to 11% in hypertensive patient populations (2,3). The aldosterone excess either results from bilateral idiopathic adrenal hyperplasia (IHA) or unilateral aldosterone-producing adenoma (APA) (4).
The appreciation that primary aldosteronism is far more common than previously recognized and the essential distinction between APA and IHA results in an increasing requirement for adrenal vein sampling (5).
Angiographic selective adrenal vein sampling (AVS) allows identi cation of patients with primary aldosteronism by direct sampling of hormones produced by each gland (8). Patients with APA can bene t from adrenalectomy, while IHA-patients require conservative drug treatment (4). Removing the identi ed gland producing excess adrenal hormone may cure secondary hypertension, thereby signi cantly improving patient's prognosis regarding cardiovascular complications and renal function (9).
Unfortunately, diagnostic angiographic procedure time is relatively long because AVS is a technically demanding procedure: AVS may be prolonged because of anatomic variations of adrenal vein ostia, particularly of the right adrenal vein (10). A prolonged search for venous ostia can result in long uoroscopy times up to 184 minutes with high dose area products (DAP) up to 3181 Gy*cm 2 (11). Different techniques of radiation dose reduction for angiographic procedures have been developed over the past years (12)(13)(14)(15). Particularly the potential of noise reduction by image-processing have been successfully introduced and recently advanced further (16). Real-time post-processing imaging algorithms allow reduced radiation doses, while maintaining adequate image quality (17). However, the degree of dose reduction varies and depends on the examined body region. Speci c post-processing imaging algorithms offer tailor-made settings depending on the particular eld of intervention, such as noise reduction, motion compensation and improved edge imaging. Motion compensation is extremely useful during thoracic or upper abdominal interventions in order to reduce breathing artifacts. Algorithms for stable interventions such as cerebral or peripheral angiography mainly bene t from noise reduction and improved edge imaging. Published dose reductions range from 43% during cardiac interventions up to 83% during iliac angiography (16-18). Furthermore, signi cantly reduced radiation doses up to 59% during bronchial artery embolization as well as up to 57% during intrahepatic interventions such as transjugular intrahepatic portosystemic shunts have been reported (19,20,21).
Up to now, the dose reduction potential of the AlluraClarity upgrade has not been assessed in patients undergoing AVS. The aim of our study was therefore to compare radiation doses in patients undergoing AVS between AlluraClarity and the precursor technology.

Study design and patient selection
Our retrospective single-center-study was approved by the local institutional review board "Aerztekammer Hamburg". Informed consent of all study participants had been obtained. All procedures were performed in accordance with the standards of the act for healing professions of Hamburg, Germany with the principles of the 1964 Declaration of Helsinki and its following amendments. All clinical imaging data were anonymized. The anonymization of patient data in the research process ensured data protection in accordance with the European General Dara Protection Regulation.
Between June 2012 and August 2018, a total of 91 AVS procedures have been performed at our department on patients with suspected primary aldosteronism using an Allura FD20 angiographic system (Philips Healthcare, Best, Netherlands).
In June 2014 the angiographic image-processing technology received an upgrade from "Allura Xper" to "AlluraClarity" (Philips Healthcare, Best, The Netherlands).
Our study group was recruited from 44 patients undergoing AVS between June 2014 and August 2018 after the AlluraClarity upgrade. Nine of these 44 patients (20.5%) had to be excluded due to incomplete unilateral left adrenal vein sampling. The remaining 35 patients with successful sampling of both left and right AVS during a single AVS procedure resulted in our study group.
Our control group was recruited from 47 patients undergoing AVS between June 2012 and May 2013 before the AlluraClarity upgrade. Twelve of these 47 patients (25.5%) had to be excluded due to incomplete unilateral left adrenal vein sampling. The remaining 35 patients with successful sampling of both left and right adrenal veins during a single AVS procedure resulted in our control group (Table 1).

Adrenal vein sampling (AVS)
AVS is a standardized procedure at our institution in accordance to the Endocrine Society guidelines and expert consensus statement of the American Heart Association (6, 7). Aldosterone release was stimulated before and during AVS with synacthen, a synthetic signal peptide of adrenocorticotropic hormone (ACTH) in a 50 µg/h infusion. The ACTH infusion prior the AVS provides the advantage of constant adrenal stimulation which levels circadian deviation (6).
Sampling was performed in all cases by board certi ed radiologists with 8-20 years of experience in interventional radiology. Details of the procedure are described in detail elsewhere (5). First, blood samples (5 ml) from the inferior vena cava (IVC) were collected as a reference. Second, the left adrenal vein ostium was carefully probed selectively with 5F-Cobra or 5F-Aachen-I catheters (Radifocus, Terumo, Japan). Third, the right adrenal vein ostium was carefully probed selectively with a 5F-Mikaelsson catheter (Impress, Merit Medical, USA) or 5F/4F-Sidewinder or 5F-Cobra catheters (Radifocus, Terumo, Japan). Venograms were performed on both left and right adrenal veins using 3-5 ml contrast agent (Imeron 300, Bracco, Italy) to document successful probing of adrenal vein ostia. Blood samples (5 ml) were collected from each adrenal vein ostium. Intraprocedural cortisol measurements con rmed successful adrenal vein sampling. Finally, IVC blood samples (5 ml) were collected as a second and nal reference.

Imaging systems
The imaging technology AlluraClarity upgrade improves noise reduction by both optimized hardware and real-time image processing-algorithms, tailor-made for different body areas by adjusted acquisition parameters (16). Compared to the precursor technology hardware changes include additional ltering of 1.0 mm aluminum und 0.1 copper and a reduced tube current for uoroscopy and digital subtraction angiography (DSA). Further, during uoroscopy the tube voltage is decreased from 80 to 70 kV. Different uoroscopy settings ranged from low to medium to high dose include increasing frames per second (fps) and decreasing levels of ltration (Table 2). After the technology upgrade, the pulse width was halved from 7.0 to 3.5 ms resulting in shortened pulses during uoroscopy. The focal spot size was decreased from 7.0 to 3.5 mm during DSA to improve spatial resolution. Apart from hardware changes, the new technology uses different background post-processing algorithms during image acquisition. These algorithms include temporal and spatial noise reduction techniques as well as motion compensation and improved edge imaging (17).

Radiation dose measurement and documentation
The dose area product (DAP) de nes the amount of dose absorption of an irradiated area in Gy*cm 2 , measured by ionization chambers placed nearby X-ray collimators (23). The DAP is independent of the distance from the radiation source.
Kerma, the kinetic energy released in matter can be calculated in the air at the interventional reference point (IRP) in units of Gray (Gy) depending on the DAP, the collimation of a eld, the tube voltage, the tube current, and the source to image-receptor distance. The IRP is located along the X-ray beam at a distance of 15 cm above the isocenter in direction from the X-ray tube towards the detector. Therefore, the air kerma (AK) can be used as a surrogate for the patient's skin dose (22

Discussion
Our study demonstrates that the AlluraClarity technology upgrade enables a radiation dose reduction up to one third in patients during AVS compared to the precursor technology Allura Xper without compromising image quality.
It is conceivable that the main reason for a signi cant reduction of radiation dose after the upgrade using AlluraClarity results from the additional lters of aluminum and copper during DSA and a reduced focal spot size during uoroscopy. By changing system parameters AlluraClarity renders over 500 different acquisition settings with respect to the particular area of interest. Various studies have already demonstrated the potential in radiation dose reduction using AlluraClarity during intracranial angiography, cardiac angiography, iliac angiography or abdominal angiography (16)(17)(18)(19). Apparently, procedural settings during cerebral interventions focus more on the pixel shift feature during steady conditions, as thoracic and especially abdominal settings emphasis more on the motion control feature in order to reduce breathing artifacts (20).
Previous studies of abdominal interventions such as transarterial chemoembolization (TACE) and transjugular intrahepatic portosystemic shunt (TIPS) using AlluraClarity abdominal settings reported dose reduction of > 50% (19,21). We observed a lower dose reduction of only 28% using the identical AlluraClarity abdominal settings during AVS.
The most important reason for less dose reduction during AVS might be due to its different anatomic region and the complexity of intervention. The most challenging part of AVS remains the reliable cannulation of the right adrenal vein (RAV) (10). Identifying the RAV-ostium from the IVC with 1-2 mm diameters requires increased contrast of the vessel edges and therefore increased radiation dose because of increased uoroscopy time during the search of the RAV-ostium.
Furthermore, AlluraClarity provides three different levels of uoroscopy-setups at the operator's console ranging from low and medium to high dose settings. These changes can be applied differently for each uoroscopy-run and might change due to individual preferences from operator to operator (Table 2). Unfortunately, these speci c individual sub-settings are currently not monitored by the vendors dose reports and could not be tracked retrospectively. The default settings contain the low dose pro le, which can be increased individually.
The relative increase in contrast agent used by 10% for additional DSA-or uoroscopy runs may be caused by the veri cation of correct catheter placement before and after sampling (  (11).
The main limitation of our study is the variety of different operators involved over 4 years due to its retrospective nature. Although AVS is to be considered as a highly standardized procedure, operators at different skill levels use their personal work ow of intervention during AVS. Evidently, identical and matched operators should have performed the interventions for both groups of patients before and after the technology upgrade. For this study we used our standard AVS-protocol. Busser et al. could already demonstrate potential dose reduction and reduced DSA-runs via intraprocedural CBCT image registration guidance improving the detection of adrenal vein ostia (14). However, this new feature of image registration is currently not practiced at our angiography suite and therefore has not been included in our routine AVS-protocol yet. Further studies should also include CBCT image registration for further radiation dose reductions.
The image quality assessment is a further limitation of the study. Image quality and vessel contrast depend not only from the angiographic system, but also from the applied ow rate during injection and amount of injected contrast agent for each series. We believe that these incongruences regarding procedure work ow apply to both groups and do not result in a systemic bias. Although the image quality assessment had been adopted from well-established previous studies, it remains a subjective tool.
In summary, we have shown that the new angiographic noise reduction imaging technology signi cantly decreases the radiation dose during adrenal vein sampling without compromising image quality or increasing uoroscopy time or contrast volume.