An Alternative Method for Adrenal Venous Sampling in Cases in Which Right Adrenal Vein Sampling is Difficult


 PurposeCatheterization of the right adrenal vein (rt.AdV) to obtain blood samples can often be difficult. The aim of the present study was to investigate whether blood sampling from the inferior vena cava (IVC) at its juncture with the rt.AdV can be an alternative to sampling of blood directly from the rt.AdV.Materials and MethodsThis study included 44 patients diagnosed with primary aldosteronism (PA) in whom AVS with adrenocorticotropic hormone (ACTH) was performed, resulting in a diagnosis of idiopathic hyperaldosteronism (IHA) (n=24), and patients diagnosed with unilateral aldosterone-producing adenoma (APA) (n=20; rt.APA=8, lt.APA=12). In addition to regular blood sampling, blood was also sampled from the IVC, as the substitute rt.AdV [S-rt.AdV]. Diagnostic performance with the conventional lateralized index (LI) and the modified LI using the S-rt.AdV was compared to examine the utility of the modified LI.ResultsThe modified LI of the rt.APA (0.4±0.4) was significantly lower than those of the IHA (1.4±0.7) (p<0.001) and the lt.APA (3.5±2.0) (p<0.001). The modified LI of the lt.APA was significantly higher than those of the IHA (p<0.001) and rt.APA (p<0.001). Sensitivity and specificity to diagnose rt.APA and lt.APA using the modified LI with threshold values of 0.7 and 2.2, respectively, they were 87% and 75%, respectively, and 94% and 94%, respectively.ConclusionThe modified LI has the potential to be an alternative method for rt.AdV sampling in cases in which rt.AdV sampling is difficult. Obtaining the modified LI is extremely simple, which might complement conventional AVS.


Introduction
Primary aldosteronism (PA) is recognized as the most common cause of secondary hypertension. Current diagnostic approaches have shown that the prevalence of PA in patients with hypertension is around 10%. This gure is much higher than previously estimated, and PA is believed to account for 17%-23% of cases of refractory hypertension (1,2). When selecting the treatment strategy for PA, the localization of hormone-producing lesions is crucial, because a unilateral aldosterone-producing adenoma (UAPA) can be treated by laparoscopic adrenalectomy or open surgery, which may result in a cure or signi cant improvement of blood pressure, thus decreasing the incidence of cardiovascular and cerebrovascular complications (1,(3)(4)(5). However, drug treatment is chosen for idiopathic hyperaldosteronism (IHA), which is caused by bilateral adrenal glands.
Various noninvasive examinations, such as computed tomography (CT), magnetic resonance imaging (MRI), and iodocholesterol (NP-59) scintigraphy performed under dexamethasone suppression, have been used to distinguish UAPA from IHA. However, CT/MRI misdiagnosed the subtyping of PA in 37.8% of patients (6). On nuclear imaging, the reported diagnostic accuracy of the criterion for diagnosing the laterality of aldosterone secretion was only 47% (7). These imaging examinations were unreliable to con rm surgically curable types and led to inappropriate management of patients with PA.
Currently, adrenal venous sampling (AVS) serves as the gold standard for subtyping of PA, with the highest diagnostic accuracy of any test (1,8). However, AVS has a number of problems. These include its greater invasiveness compared with other imaging modalities, the basic requirement for hospital admission, and the potential for bleeding, infection, vascular damage, and other complications associated with catheterization. However, the most important factor is the technical success rate of AVS is widely different between facilities. The reported success rates of AVS range from 55% to 98%, with technical success rates varying widely depending on the operator's skill to success of rt.AdV sampling (1,9,10). Right AdV catheterization is often di cult due to unfavorable anatomy of the vein, including small vein size, short length, caudal direction for the transfemoral approach, and rare drainage to the accessory hepatic vein. Even in patients with successful right AdV catheterization, the catheter might get dislodged with respiration, or the tip of the catheter might become lodged in the vascular wall, making blood sampling di cult. All of these factors limit the widespread use of this approach (1,3,(9)(10)(11)(12)(13)(14). Given the wide range of success rates for right AVS (9,10,15), it is necessary to achieve a more consistent AVS success rate and eliminate the technical failures associated with right AVS cannulation.
Identi cation of the con uence of the right AdV with the inferior vena cava (IVC) by preoperative contrastenhanced thin-slice computed tomography (CT) is an essential and feasible procedure (16).
Catheterization of the right AdV is often di cult, but if preoperative CT is used as a reference, the catheter tip can very easily be guided downstream to its con uence with the IVC.
The aim of the present study was to investigate whether blood sampling from the IVC at its juncture with the right AdV can be an alternative method to sampling of blood directly from the right AdV.

Patient Population
This study was conducted in accordance with the Declaration of Helsinki and approved by the Ethics Committee of A liated Hospital of Kawasaki Medical School (Kurashiki, Japan; ID No. 3288). As it was a retrospective study, and the requirement for informed consent was waived by the Ethics Committee of A liated Hospital of Kawasaki Medical School (Kurashiki, Japan; ID No. 3288). This study initially included 58 consecutive patients who underwent AVS with adrenocorticotropic hormone (ACTH) for local diagnosis of the lesion because of suspected primary aldosteronism (PA) with hypertension and an aldosterone-renin ratio (ARR)>200 (average 992.3; range 200.0-5233.3) from January 2010 to February 2019. The following patients were excluded because AVS could not be technically completed or the nal local diagnosis of the hormone-producing lesion was unclear: patients who could not complete blood sampling from the right AdV (n=9); patients who did not undergo surgery, despite being diagnosed with unilateral aldosterone-producing adenoma (UAPA) by AVS (n=2); patients whose ARR did not decrease to within the normal range (<200) after surgery (n=2); and patients whose postoperative course could not be followed-up (n=1). Finally, this study included 44 patients (21 men, 23 women; mean age 56.0 years, range 28-73 years) with technical success of AVS, and all patients who were diagnosed with UAPA by AVS underwent surgery, and the ARR decreased to the normal range. The resulting diagnosis was idiopathic hyperaldosteronism (IHA) (n=24) and unilateral aldosterone-producing adenoma (UAPA) (n=20) including right APA (n=8) and left APA (n=12).

AVS Procedure
AVS was always conducted by 2 or 3 senior radiology residents or a radiologist with supervision by one of four interventional radiologists, after the administration of anticoagulant therapy with 3000 units of heparin. Under local anesthesia, 4-Fr sheaths were percutaneously inserted in the femoral veins of both legs. Before and after (between 15 and 30 minutes) bolus injection of 0.20 mg of the adrenocorticotropic hormone (ACTH) cortrosyn (cosyntropin), at least 3 mL of venous blood were sampled from the following four sites: the lowest point of the inferior vena cava (IVC) and the left AdV, the right AdV, and the IVC at one vertebral level above its con uence with the right AdV (the substitute right AdV [S-rt. AdV]). For the Srt. AdV, the tip of the catheter was positioned on the same side of the IVC as the right AdV. Figure 1 shows how the positions of the S-rt. AdV were determined. For cannulation of the left AdV, a 4-Fr catheter of Simons-type (Terumo, Japan) and a 1.7-Fr microcatheter (Boston Scienti c, United States) were mainly used, and sampling from the main trunk of the left AdV was conducted via a 4-Fr sheath (Medikit, Japan) inserted into the left femoral vein. For cannulation of the right AdV, a 4-Fr Cobra-type (Medikit, Japan) or Hook-type (Medikit, Japan) catheter inserted into the right femoral vein via a 4-Fr sheath was used. The same microcatheter was used for the right AdV if blood could not be withdrawn from the 4-Fr catheter.
Digital subtraction angiography (DSA) was performed using a Allura Xper FD20 (Philips Healthcare, Netherland) and a In nix 8000v (Canon Medical Systems, Japan). For patients in whom it was di cult to determine whether the catheter had entered the right AdV, unenhanced or enhanced cone-beam CT was performed to con rm the locations of the catheters in the right AdV.

Data Analysis
To evaluate successful catheterization of an AdV, the selectivity index (SI) was calculated as follows: SI = plasma cortisol concentrations of the AdV / plasma cortisol concentrations of the lowest point of the IVC. On the basis of guidelines in previous reports (17)(18)(19), an SI 5.0 was considered to indicate successful catheterization of each AdV.
The aldosterone/cortisol (A/C) ratio of each AdV was calculated to evaluate the hormone-producing local diagnostic ability for PA. The conventional lateralized index (LI) (A/C ratio larger) / (A/C ratio smaller) was calculated. On the basis of guidelines in previous reports (17)(18)(19), an LI 4.0 was diagnosed as IHA, and an LI 4.0 was diagnosed as UAPA. Our own modi ed lateralized index (modi ed LI) (Left AdV A/C ratio) / (S-rt.AdV A/C ratio) was also calculated, the cutoff value with the best diagnostic performance was determined, and the diagnostic performances of the conventional LI and modi ed LI were compared.

Principle of Local Diagnosis of PA by the Modi ed LI
To explain the modi ed LI in simple terms, assume that a hormone-producing lesion produces 10 units of hormones and a normal adrenal gland produces 1 unit. Taking no account of hormone dilution, if there is a hormone-producing lesion in the right adrenal gland, then the 1 unit secreted normally and the 10 units of oversecreted hormone produced by the lesion in that gland, making a total of 11 units, will be detected in the blood sampled from the S-rt. AdV (Fig.2a). If there are hormone-producing lesions in both adrenal glands, the 10 units oversecreted by the hormone-producing lesion in the left adrenal gland and the 10 units oversecreted by the hormone-producing lesion in the right adrenal gland, making a total of 20 units, will be detected (Fig.2b). If there is a hormone-producing lesion in the left adrenal gland, the 10 units oversecreted by the hormone-producing lesion in the left adrenal gland and the 1 unit secreted by the right adrenal gland, a total of 11 units, will be detected (Fig.2c). Substituting these values into the formula for calculating the modi ed LI [modi ed LI = (Lt. AdV A/C ratio) / (S-rt.AdV A/C ratio) gives the following results. If there is a hormone-producing lesion in the right adrenal gland (Rt. UAPA), the modi ed LI =1/11 = 0.1 (Fig.2a). If there are hormone-producing lesions in both adrenal glands (IHA), the modi ed LI = 10/20 = 0.5 (Fig.2b). If there is a hormone-producing lesion in the left adrenal gland (Lt.UAPA), the modi ed LI = 10/11 = 0.9 (Fig.2c). The results are thus different in these three cases. In fact the hormone is diluted in the bloodstream, but the location of the hormone-producing lesion(s) will nevertheless make some difference.

Statistical Analysis
A nonparametric Mann-Whitney U test was used to compare and identify signi cant differences in the mean conventional LI and the modi ed LI between UAPA and IHA. Receiver operating characteristic (ROC) curve analysis was performed for the comparison of UAPA and IHA using the conventional LI and the modi ed LI. The area under the ROC curve (AUC) was calculated for each variable, and the optimal threshold for each variable was determined from its respective ROC analysis by evaluating the sensitivity, speci city, and accuracy at different cutoff points on the ROC curve. The cutoff values of the modi ed LI values were determined so that the sum of sensitivity and speci city was maximized. All statistical analyses were performed using SPSS software (v. 17.0J for Windows; Chicago, IL). All tests were twosided, and P <.05 was considered signi cant.
On ROC curve analysis for the diagnostic performance of the conventional LI, the area under the curve (AUC) was 0.90 in UAPA, whereas for the modi ed LI, the AUC was 0.92 in right APA and 0.81 in left APA (Fig.5). Sensitivity and speci city to diagnose UAPA using the conventional LI were 95% and 83%, respectively, with a threshold value of 1.9, and to diagnose right APA and left APA using the modi ed LI, they were 87% and 75%, respectively, and 94% and 94%, respectively, with threshold values of 0.7 and 2.2, respectively (Table 1).
In 5 of the 44 cases in which catheterization of the right AdV was deemed technically successful, the SI was <5.0, suggesting unsuccessful catheterization. For the left AdV, however, the SI was >5.0 in all cases, indicating successful catheterization. For the 5 cases of suspected unsuccessful right AdV catheterization, the localization of the hormone-producing lesion according to the conventional LI was mistaken in 3. Localization by the modi ed LI was mistaken in only 1 case.

Discussion
In this study, the diagnostic performance of the modi ed LI to identify the location of hormone-producing lesions was not much inferior to the diagnostic performance of the conventional LI. One reason for this may be that, in many cases, the ACTH challenge greatly increases aldosterone secretion, such that its diagnostic performance is comparatively good even when blood is sampled from the IVC, where the aldosterone concentration is diluted, instead of from the right AdV. According to a review of 47 reports, the success rate of cannulating the right AdV in 384 patients was 74% (1,15), but with experience, the success rate increases to 90-96% (1,9,18). As shown by that result, the technical success rate of AdV sampling is thus highly dependent on the operator's skill. However, the demand for AdV sampling in recent years makes it an investigation that should be provided by more hospitals, and it is undesirable that the results should be affected by the operator's skill. In most cases, unsuccessful AdV sampling is due to either unsuccessful catheterization of the right AdV or an inability to suction blood from this vein despite its successful catheterization. Assessment using our modi ed LI may resolve this problem by making selection of the right AdV unnecessary, instead using the IVC where catheter placement is extremely simple. Although selecting and sampling from the right AdV are indisputably the best, when this cannot be achieved, sampling from the IVC at one vertebral level above its con uence with the right AdV (the S-rt. AdV) can be performed simply and reliably. Assessment using the modi ed LI may then provide data on which the treatment strategy can be based.
Even more interestingly, for the ve cases in which catheterization of the right AdV was judged to have been technically successful, but the SI data indicated unsuccessful catheterization, assessment using the conventional LI resulted in mistaken localization in 3/5 cases, whereas assessment using the modi ed LI resulted in mistaken localization in only 1/5 cases, and correctly diagnosed 2 cases in which the conventional LI provided the wrong localization. This result suggests that the S-rt. AdV should be used not only when right AdV catheterization is technically di cult, but also for patients in whom it has been successfully achieved, since this practice may assist with the localization of hormone-producing lesions in patients with postoperative SI <5.0. However, the modi ed LI also has its own problems. The most important of these is that the position of the catheter for sampling from the S-rt. AdV cannot be determined if the con uence of the right AdV has not been identi ed intraoperatively. This problem can be overcome by careful evaluation of the con uence of the right AdV on preoperative CT (20).
A limitation of this study was that the catheter tip for S-rt. AdV blood sampling was positioned after the con uence of the right AdV had been identi ed intraoperatively. In patients in whom the con uence of the AdV cannot be identi ed intraoperatively, catheter tip positioning must be entirely dependent on information obtained from preoperative CT, and this might lead to different results. However, careful identi cation of the right AdV branch with reference to preoperative CT should enable the catheter tip to be positioned with reasonable accuracy. The second limitation of the present study was that the procedures were performed by different operators, which may have caused technical differences. However, since all procedures were conducted under the supervision of an interventional radiologist, these technical differences were probably minor. A third limitation was the small number of subjects. Further studies of more cases are required.
In conclusion, the modi ed LI has the potential to be an alternative method for right AdV sampling in cases in which right AdV sampling is di cult. Obtaining the modi ed LI is extremely simple, and it might complement conventional AVS. The modi ed LI using blood sampled from the IVC at the juncture of the right AdV, which can be done easily in such patients, is a potentially useful clinical method.  Figure 1 Schematic diagram (A) and digital angiography (B,C) of the location of catheter placement in the S-rt. AdV. The catheter tip is placed at one vertebral level above the con uence of the IVC with the right AdV, and its orientation is adjusted so that it is pointing toward the right dorsal side in the same way as the right AdV.

Figure 2
Schematic diagram in case of hormone-producing lesion in the right adrenal gland (A), hormoneproducing lesions in both adrenal glands (B) and hormone-producing lesion in the left adrenal gland (C).
To explain the modi ed LI in simple terms, assume that a hormone-producing lesion produces 10 units of hormones and a normal adrenal gland produces 1 unit. Taking no account of hormone dilution. In case of (A), total of 11 units of hormone will be detected in the blood sampled from the S-rt. AdV by the 10 units of oversecreted by the lesion in right adrenal gland and the 1 unit secreted normal left adrenal gland. In case of (B), total of 20 units of hormone will be detected in the blood sampled from the S-rt. AdV by the 10 unit oversecreted by the lesion in both adrenal. In case of (C), total of 11 units of hormone will be detected in the blood sampled from the S-rt. AdV by the 1 unit secreted normal left adrenal gland and the 10 units of oversecreted by the lesion in right adrenal gland. Substituting these values into the formula for calculating the modi ed LI, in case of (A) (Rt. UAPA), the modi ed LI =1/11 = 0.1, in case of (B) (IHA), the modi ed LI = 10/20 = 0.5 and in case of (C) (Lt. UAPA), the modi ed LI = 10/11 = 0.9. The modi ed LI of right APA (0.4±0.4) was signi cantly lower than those of IHA (1.4±0.7) (p<0.001) and left APA (3.5±2.0) (p<0.001). The modi ed LI of left APA was signi cantly higher than those of IHA (p<0.001) and right APA (p<0.001)