Catheterization During Adrenal Vein Sampling for Primary Aldosteronism: Failure to Use (1–24) ACTH May Increase Apparent Failure Rate

Abstract

“Successful” adrenal vein catheterization in primary aldosteronism (PA) is often defined by a ratio of >3:1 of cortisol in the adrenal vein vs the inferior vena cava. Non‐use of corticotropin (ACTH) during sampling may increase the apparent failure rate of adrenal vein catheterization due to lower cortisol levels. A retrospective study was performed on all patients with confirmed unilateral PA between June 2005 and August 2011. Adrenal vein sampling (AVS) included simultaneous bilateral baseline samples with repeat sampling 15 minutes after intravenous infusion of 250 μg of Cortrosyn (ACTH‐S). Successful catheter placement was judged as adrenal cortisol:IVC cortisol of >3:1, applied to both baseline and ACTH‐S samples and lateralization of aldosteronism was judged as normalized aldosterone/cortisol (A/C) ratio >3 times the contralateral A/C ratio. In ACTH‐S samples, 94% of right‐sided catheterizations were biochemically successful with 100% success on the left. Among baseline samples, only 47% of right‐ and 44% of left‐sided samples met the 3:1 cortisol criteria. However, 95% of apparent “failed” baseline cortisol sets still showed lateralization of A/C ratios that matched the ultimate pathology. Non–ACTH‐stimulated samples may be incorrectly judged as failed catheter placement when a 3:1 ratio is used. ACTH‐stimulated sampling is the preferred means to confirm catheterization during AVS.

Primary aldosteronism (PA) is increasingly recognized as a common form of secondary hypertension.1, 2, 3 Cases of lateralized aldosterone excess may be treated with adrenalectomy. However, selection of patients for adrenalectomy depends on the ability to confirm unilateral adrenal aldosterone excess, which, in turn, is demonstrated through a combination of anatomical imaging and adrenal vein sampling (AVS).4 AVS is a technically demanding procedure, and successful adrenal vein catheterization is a prerequisite for correct interpretation of hormone levels.5

There is uncertainty as to whether there is an “optimal” method for AVS.6 Most centers now agree that some measure of catheter positioning should be evaluated prior to comparing the aldosterone levels.7 However, not all investigators agree as to the criteria for demonstration of catheter positioning.8 In some series, adrenal venous catheter placement is often defined as an adrenal_cortisol/IVC_cortisol >2‐3:19, 10 but there is debate as to whether this ratio should apply to samples drawn with and without prior (1–24) corticotropin (ACTH) stimulation.11 In fact, not all authors are convinced of the usefulness of ACTH infusion during AVS.12 We hypothesized that non‐use of ACTH during AVS might result in a higher apparent catheterization failure rate due to the lack of an exaggerated adrenal venous cortisol rise with subsequent adrenal/IVC cortisol ratios <3:1. It may be that a different definition of successful catheterization should be applied if (1–24) ACTH is not going to be used, in order to avoid falsely excluding otherwise valid patient data.

In order to determine if a 3:1 cortisol/IVC ratio is a true prerequisite to accurate interpretation of lateralization indices in AVS and whether ACTH stimulation changes the interpretation of catheter placement, we conducted a retrospective review of our PA‐AVS database.

Methods

This analysis was approved by the local ethics review board. The study was conducted in the practice of a single Aldosteronism Interest Group operating an endocrine hypertension specialty clinic with aggressive screening for PA in all referred patients.

A retrospective review of all patients with PA having had AVS between June 2005 and July 2011 was performed. We included all patients who had lateralizing AVS findings and subsequent unilateral adrenalectomy with surgical pathology (n=32).

The study cohort comprised 32 patients (18 men, 14 women), with a mean age of 45.5 years. Mean blood pressure (BP) at presentation was 147/92 mm Hg while taking an average of 3.4 antihypertensive medications. A total of 91% of patients had a mean adrenal mass by computed tomographic (CT) imaging of 16.2 mm. The mean initial K+ was 2.91 mmol/L and 84% of patients were hypokalemic (<3.5 mmol/L) prior to diagnosis and AVS. The mean diagnostic aldosterone‐renin ratio (ARR) was 7868 pmol/L/ng/mL/hour (normal <550). Surgical pathology revealed a solitary adrenal adenoma in 69% of patients, an adenoma with surrounding hyperplasia in 19%, and pure hyperplasia in 12%. All patients were seen within 3 weeks of discharge from the hospital. Postoperatively, hypokalemia resolved in all patients who had prior low potassium levels. The ARR normalized in all patients, and 34% had outright “cure” of hypertension (BP <140/90 mm Hg without medications). The other 66% of patients had BP fully controlled to <140/90 mm Hg using an average of 1.4 antihypertensive medications per patient. In the 19 patients who had a repeat ARR after surgery, the mean result was 150 and all patients were <280, indicating biochemical endocrinologic cure of PA.

AVS Protocol

Prior to AVS, all mineralocorticoid receptor antagonists or potassium‐sparing diuretics were stopped for 6 weeks and calcium channel blockers or α‐blockers were used where possible for BP control. Oral potassium supplements were given to normalize serum potassium levels prior to AVS. Following our previously published protocol,10 AVS was performed with baseline, simultaneous bilateral adrenal, and IVC sampling of cortisol and aldosterone, with a second set of matching samples drawn following an intravenous infusion of 250 μg (1–24) ACTH over 15 minutes. Collected samples were immediately labelled and transferred to the central laboratory for processing. All AVS procedures were performed by a single, experienced operator. There were no complications with any AVS procedure.

Unilateral disease was confirmed by histologic findings in addition to postoperative normalization of the ARR and objective clinical response (normalization of hypokalemia if present, and of BP to <140/90 mm Hg on the same or fewer medications).

The AVS data were then examined in retrospect. During the entire study period, successful catheter placement was judged as an adrenal_cortisol: IVC_cortisol of >3:1, a criterion applied to both baseline and stimulated (ACTH‐S) samples. Lateralization was defined as a normalized adrenal aldosterone/cortisol (A/C) ratio >3:1 compared with the contralateral adrenal A/C ratio. Again, a similar definition of lateralization was used for both baseline and ACTH‐S samples.

The datasets were then examined to see whether, in surgically confirmed lateralized cases, the lateralization indices were concordant in both baseline and stimulated samples and, by inference, whether the catheter placement index of >3:1 was truly a prerequisite for accurate interpretation of the lateralization in each case.

Standard descriptive statistics were used to define the population and the AVS success rates. As the data were not normally distributed, Spearman's correlation was used to describe the relationship between baseline cortisol and cortisol ratios achieved. Chi‐square analysis was used to determine the significance of the difference between proportions of datasets judged successful by differing cortisol ratios. All computations were performed with AnalyseIt (2011; Analyze‐it Software, Leeds, UK) statistical computing software.

Results

Table 1 summarizes the AVS results in each of the 32 patients with surgically, pathologically, clinically, and frequently biochemically confirmed lateralized PA. Note that using the 3:1 cortisol ratio placement criterion, 94% of the right adrenal vein catheterization attempts were judged to be biochemically confirmatory of successful placement in the ACTH‐S samples. Similarly, the left‐sided ACTH‐S samples showed apparent 100% successful placement. Virtually all ACTH‐S cortisol ratios are well above the cut‐off used to determine success. However, in the baseline samples, using the same criterion, only 44% and 47% of the right and left samples were judged “successful.” Thus, without the (1–24) ACTH infusion, 63% of AVS attempts would be considered “unsuccessful” for having “failed” to place either the right or left or both adrenal venous catheters. However, of the 17 baseline right adrenal cortisol ratios <3:1, 88% exceeded the 3:1 ratio after (1–24) ACTH infusion. Of the 18 baseline left adrenal cortisol ratios <3:1, 100% exceeded the 3:1 ratio after (1–24) ACTH infusion.

Table 1.

Patient Data: Baseline and (1–24) Corticotropin‐Stimulated Cortisol Ratios With Biochemically Defined Successful Placement and Lateralization

Patient	Left‐Baseline Adrenal/IVC Cortisol Ratio	Right‐Baseline Adrenal/IVC Cortisol Ratio	Placement Fail=Ratio <3:1 on 1 or Both Sides	Lateralization Normalized Aldosterone (A/C) >3:1	Left‐Stimulated Adrenal/IVC Cortisol Ratio	Right‐Stimulated Adrenal/IVC Cortisol Ratio	Placement Fail=Ratio <3:1 on 1 or Both Sides	Lateralization Normalized Aldosterone (A/C) >3:1	Lateralization Concordance Baseline and Stimulation Samples
1	6.3	1.1	F	Yes	123.2	1.0	F	Yes	NA
2	56.9	139.9		Yes	53.3	37.0		Yes	Yes
3	2.1	2.7	F	Yes	26.5	44.8		Yes	Yes
4	2.6	3.1	F	Yes	31.3	35.1		Yes	Yes
5	1.7	2.4	F	Yes	47.9	46.4		Yes	Yes
6	3.9	4.7		No	92.0	31.2		Yes	No
7	4.1	17.1		Yes	54.2	133.8		Yes	Yes
8	2.0	1.5	F	Yes	33.9	22.0		Yes	Yes
9	2.4	5.4	F	Yes	26.0	43.4		Yes	Yes
10	5.3	11.4		Yes	30.1	30.1		Yes	Yes
11	2.2	2.5	F	Yes	33.6	39.5		Yes	Yes
12	13.2	3.0		Yes	86.2	2.3	F	No	No
13	1.6	2.4	F	Yes	36.8	64.9		Yes	Yes
14	1.9	1.6	F	Yes	51.1	27.1		Yes	Yes
15	2.2	2.6	F	Yes	11.8	36.1		Yes	Yes
16	3.1	3.3		Yes	66.7	71.6		Yes	Yes
17	2.4	4.4	F	Yes	19.6	38.4		Yes	Yes
18	4.5	8.4		Yes	4.7	8.4		Yes	Yes
19	2.0	1.8	F	Yes	28.1	43.2		Yes	Yes
20	2.2	2.1	F	Yes	61.8	64.2		Yes	Yes
21	2.9	2.6	F	No	23.4	44.7		Yes	No
22	1.6	2.4	F	Yes	22.9	42.4		Yes	Yes
23	39.7	26.0		Yes	23.7	23.0		Yes	Yes
24	1.8	1.5	F	Yes	10.3	53.9		Yes	Yes
25	1.8	1.7	F	Yes	33.8	49.7		Yes	Yes
26	14.3	30.4		Yes	11.7	16.0		Yes	Yes
27	1.4	1.6	F	Yes	32.5	22.6		Yes	Yes
28	38.9	39.5		Yes	28.1	44.0		Yes	Yes
29	2.1	2.6	F	Yes	54.9	64.9		No	No
30	54.1	57.0		Yes	25.6	47.0		Yes	Yes
31	3.7	5.0		Yes	30.2	36.3		Yes	Yes
32	12.4	11.7		Yes	25.5	43.2		Yes	Yes
Mean	2.8	2.4			38.8	40.8

Abbreviations: AVC, adrenal vein catheterization, NA, not applicable.

Despite apparent “catheterization failure” on 63% of baseline samples, 95% of these “failed” baseline samples showed lateralization indices (>3:1) that ultimately still matched the proven, lateralized pathology. Similarly, 89% of the failed baseline samples showed lateralization indices that were concordant with the successful ACTH‐S sample results, suggesting that the lateralization indices were indeed correct, even if the placement cortisol ratios were possibly indeterminate.

We next examined whether the baseline IVC cortisol might predict the outcome of the sampling procedure in terms of whether it would ultimately be judged as successful or failed placement. These results are shown in the Figure1. This analysis showed that baseline IVC cortisol does not seem to correlate with the adrenal_cortisol:IVC_cortisol ratio. Table 2 displays the proportion of samples deemed “successful placement” depending on whether a 2:1, 3:1, or 4:1 ratio is used. There is a highly significant difference in the proportions of non–ACTH‐stimulated AVS that are defined as successful when the criterion for success is 2:1 or 4:1 compared with 3:1.

Figure 1.

Correlations between baseline IVC cortisol level and cortisol placement ratios for the right (upper panel) and left (lower panel) adrenal veins.

Table 2.

Effect of Biochemically Judged “Successful” Catheter Placement Depending on Criterion of Success: Samples Collected Prior to (1–24) Corticotropin Infusion

Judged Successful at >2:1, %				Judged Successful at >3:1, %			Judged Successful at >4:1, %
Right	Left	Both	P Value vs 3:1	Right	Left	Both	Right	Left	Both	P Value vs 3:1
78	75	69	.003	44	47	38	41	31	28	.001

Discussion

This analysis demonstrated that without the use of (1–24) ACTH infusions, many cortisol ratios in AVS would appear to be indicative of failed venous catheter placement when a 3:1 ratio is used to define success.

However, using our approach of collecting both unstimulated and (1‐24) ACTH‐stimulated samples, we have been able to show that many of these so‐called unsuccessful catheter placements are probably actually quite successful in the hands of an experienced operator, as shown by the very high success rates following ACTH infusion, without having moved the venous catheters in the interim. Additionally, there is a high degree of agreement in diagnostic lateralization indices seen in both baseline and ACTH‐S samples, despite the apparent high number of indeterminate or failed baseline sample placement data, again suggesting that the baseline samples were actually drawn from correctly positioned catheters.

This suggests that (1–24) ACTH stimulation is a valuable part of AVS because of its ability to confirm correct venous catheter placement. PA clinics with apparently high rates of unsuccessful AVS despite good operator competence might therefore be able to see an instant improvement in AVS catheterization success by simply adding a (1–24) ACTH infusion to their protocol. Similarly, centers that do not employ (1–24) ACTH during sampling may wish to examine their data and choose a much lower cut‐off than 3:1 for definition of success. Our post‐hoc analysis of differing placement ratios (confirmed by pathology) would suggest that a ratio of 2:1 may be more appropriate when (1–24) ACTH is not used.

Rossi and colleagues have previously shown that apparent AVS failures increase as required cortisol ratios (termed selectivity indices) increase such that success rates may be as low as 40% when a ratio of 5:1 is set as a definition for confirmed placement.13 However, the same group has also questioned whether an adrenocorticotropin infusion might actually reduce the ability to detect lateralization.14 Our results from a slightly larger population did not seem to confirm this as we had a high degree of ultimate agreement between both pre‐ and post‐ACTH infusion results. Our results with ACTH infusion do tend to support other center's protocols, which stress the primacy of proving correct catheter placement prior to any attempt at interpretation of the lateralization since a surgical decision must ultimately be made with confidence.

It has also been suggested that an ACTH infusion may overcome any variability in results due to a physiologic fluctuation of ACTH during AVS.15 While this may have some physiologic basis, our concordant stimulated and nonstimulated results suggest that the impact of any such fluctuations may be wholly explained by apparent variance in successful catheterization when ACTH is not used. Again, this emphasizes the importance of carefully choosing a cortisol ratio to define success or, better still, using ACTH to ensure appropriate recognition of what is actually a successful AVS study.

The recently published results of the Adrenal Vein Sampling International Study (AVIS) demonstrated that there is a high degree of variability in protocols for AVS and only 55% of participating centers reported the routine use of a cosyntropin infusion.7 The same study highlighted the urgent need for an evidence‐based approach to AVS in order to standardize the approach to subtyping of PA. A uniform approach to AVS would clarify the optimal criteria for interpretation and subsequent decision making regarding surgical or medical management.

It is not known how many centers worldwide are able to offer AVS for the investigation of PA, despite the fact that recognition of PA, even in unselected hypertensive patients, has increased tremendously across the globe in the past decade.16, 17, 18 However, it is known that AVS is technically demanding and, as an invasive procedure, does carry a small risk of adrenal hemorrhage or vein rupture.19 This emphasizes the importance of knowing the steps that will maximize technical success rates for any new center that is embarking on having an AVS program. The present study suggests that cosyntropin infusion should be strongly considered as part of any AVS protocol and our results confirm those of a prior paper showing that ACTH infusion may help less experienced AVS operators to obtain clinically useful results by maximizing the recognition of successful sampling.20

Study Strengths

The strengths of our study include the comprehensive PA case confirmation by pathology, clinical, and biochemical measures, which ensures that the gold standard diagnosis is met when retrospectively assessing the steps in diagnosis. Our AVS success rates (by any published criteria) have become near 100% in the past 7 years since moving to a single operator. This has dramatically reduced the indeterminate cases that would otherwise confound this type of analysis. Lastly, our long‐standing protocol of double sampling pre‐ and post‐cosyntropin has served as a valuable internal control for preoperative diagnostic accuracy.

Where AVS fails or is deemed to be of uncertain diagnostic value, especially with discordant adrenal CT imaging, adrenal scintigraphy may prove to be a useful second‐line test of lateralization.21 There may be a role for repeat AVS attempts if a prior attempt is unsuccessful but consultation with interventional radiology is necessary to ensure that a repeat attempt could be feasible as opposed to failure due to scarred or obliterated adrenal veins. Ultimately, use of an aldosterone receptor antagonist is warranted where no clear surgical option exists. This has been shown to be highly successful in managing both PA and its complications.22

Study Limitations

There are some important limitations to the study, most notably that there are too few discordant or contradictory pairs to allow any conclusions as to how one should best resolve borderline or uncertain cases. As well, this lack of discordant pairs (with surgical outcomes) prevents a useful analysis of whether cosyntropin infusion changes the lateralization indices used to define unilateral vs bilateral disease. Some authors claim a decrease in diagnostic accuracy due to effects on lateralization ratios but a recent larger, multicenter study of this issue fails to show any adverse effect on final diagnosis.12, 20 In our experience, use of a 3:1 ratio for lateralization with ACTH infusion appears to have remarkable specificity for unilateral disease and therefore confidence for surgical decision making.

Conclusions

This study demonstrates that routine use of (1–24) ACTH should be strongly considered in order to minimize the chance of falsely discarding AVS results that might otherwise appear to have failed for lack of proper venous catheter positioning. If ACTH infusions are not used, a lower adrenal_cortisol:IVC_cortisol ratio should be used for a definition of successful catheter placement.

Disclosures

This paper was presented at the American Society of Hypertension 2012 scientific meeting in New York City and submission for publication was invited. The authors have nothing to disclose.

J Clin Hypertens (Greenwich). 2013;15:480–484. ©2013 Wiley Periodicals, Inc.23815536