Is Prolactin Measurement of Value during Inferior Petrosal Sinus Sampling in Patients with ACTH-dependent Cushing’s Syndrome?

Abstract

Inferior petrosal sinus sampling (IPSS) is considered the gold standard test to distinguish between Cushing’s disease (CD) and ectopic ACTH syndrome (EAS). Anomalous venous drainage, abnormal venous anatomy and lack of expertise can lead to false-negative IPSS results and thereby misclassification of patients with ACTH-dependent Cushing’s syndrome. Prolactin measurement during IPSS can improve diagnostic accuracy and decrease false negative results. A baseline prolactin inferior petrosal sinus to peripheral (IPS/P) ratio (ipsilateral to the dominant post-CRH ACTH IPS/P ratio) of 1.8 or more suggests successful catheterization during IPSS. Prolactin-normalized ACTH IPS/P ratios can then be used to differentiate between a pituitary and ectopic source of ACTH. Values ≤ 0.7 are suggestive of EAS and those ≥ 1.3 are indicative of CD but the implication of values between 0.7 and 1.3 remains unclear and needs further investigation. Larger prospective studies are also needed for further evaluation of the role of contralateral prolactin IPS/P ratios, post-CRH prolactin values and prolactin-adjusted ACTH inter-sinus ratios for tumor localization in CD.

Inferior petrosal sinus sampling (IPSS) is considered the most reliable test to differentiate between a pituitary and an ectopic source of adrenocorticotropin hormone (ACTH)-dependent Cushing’s syndrome (CS). A petrosal sinus to peripheral ACTH gradient of ≥ 2.0 at baseline or ≥ 3.0 after corticotropin releasing hormone (CRH) administration in the setting of sustained hypercortisolemia suggests a pituitary source of ACTH (1–2). Although earlier studies on IPSS using CRH stimulation reported a diagnostic sensitivity and specificity close to 100% (2), with increasing experience, false negative rates of 1–10% have been noted. These false negative IPSS results have been attributed to anomalous venous drainage, abnormal venous anatomy (e.g. hypoplastic or plexiform petrosal sinuses), lack of expertise and/or technical problems (3–5).

Transsphenoidal surgery with resection of the ACTH-secreting pituitary adenoma remains the treatment of choice in patients with Cushing’s disease (CD). Given the considerable reliance placed on IPSS as a diagnostic test, false negative IPSS results can lead to misclassification of patients, resulting in unnecessary exposure to radiation from imaging studies and side-effects of medical control of hypercortisolemia or even bilateral adrenalectomy. Improved diagnostic accuracy of IPSS is therefore important to obtain the correct diagnosis and minimize morbidity in these patients.

One way to evaluate the success of catheterization is to review the IPSS venous angiography to demonstrate retrograde flow of contrast dye into a normal contralateral cavernous sinus prior to sampling (2). Studies also have looked at the use of a reference hormone to document successful catheter placement, as has been successfully used in adrenal vein sampling during the evaluation of primary hyperaldosteronism (6). We and others have evaluated the role of prolactin as a reference hormone during IPSS in patients with ACTH-dependent CS (7–11).

Why Prolactin?

Prior studies investigated the measurement of various pituitary hormones, including growth hormone (GH), thyroid-stimulating hormone (TSH), α-subunit and prolactin, to assess catheter placement during IPSS (12–14). Because the hypercortisolism of CS can suppress GH and TSH levels (13), their use to assess successful catheterization during IPSS may not be advisable in this setting. Alpha-subunit also may be suppressed in EAS (12–13) and is not available at all institutions.

Prolactin is an abundant anterior pituitary hormone, as lactotrophs comprise 10 to 25% of the hormone-secreting anterior pituitary cells. The distribution of lactotrophs in the anterior pituitary is different from the normal corticotrophs – posteromedial region versus central ‘mucoid wedge’ area – suggesting that prolactin would be a reliable reference hormone (15). Additionally, prolactin assays are widely available. For these reasons, most of the recent studies have focused on prolactin as a possible marker of successful catheterization during IPSS.

Prolactin as a marker of pituitary venous effluent

McNally et al. studied the use of thyrotropin releasing hormone (TRH)-stimulated TSH and prolactin measurement as a reliable method for correcting for the dilution from non-pituitary venous blood during IPSS (7). Using uncorrected basal and/or post-CRH ACTH values, two of the eight patients were misdiagnosed with a non-pituitary source of ACTH. Correcting the absolute ACTH levels for TRH-stimulated TSH and prolactin inter-sinus ratios led to the correct diagnosis of CD in all eight patients. This study was limited by the assumption that there was uniform secretion of both TSH and prolactin into both petrosal sinuses. Moreover, TRH is no longer available for clinical diagnostic purposes in the United States.

Findling et al. retrospectively measured prolactin levels in three index patients with no central to peripheral ACTH gradient during IPSS and negative imaging studies for a possible ectopic source of ACTH (8). Results from these three patients were then compared to those from archived IPSS samples in 44 patients with proven CD and five patients with proven EAS. The authors concluded that a baseline (before CRH administration) inferior petrosal sinus to peripheral (IPS/P) prolactin ratio of ≥ 1.8 ipsilateral to the peak ACTH IPS/P ratio indicated successful catheterization. Furthermore, after “normalizing” the peak ACTH IPS/P ratio by dividing it by this baseline ipsilateral prolactin IPS/P ratio, values greater than 0.8 suggested CD whereas those less than 0.6 implied EAS. All three index cases were later found to have surgically proven CD. The baseline prolactin IPS/P ratio was <1.8 in all three cases indicating unsuccessful catheterization and normalized ACTH/prolactin ratios correctly diagnosed them as CD.

In another series, Mulligan et al. studied 35 patients with CD and one with EAS. One patient with surgically proven CD (right-sided ACTH staining pituitary adenoma) had false negative IPSS ACTH results. The highest pre-CRH IPS/P prolactin ratio was 1.0 on the right and 2.2 on the left suggestive of unsuccessful catheterization of the right inferior petrosal sinus. Another proven CD patient with equivocal ACTH IPSS results (all ACTH IPS/P ratios <2.0 except one pre-CRH ratio of 2.2) had baseline IPS/P prolactin ratios between 1.1 and 1.7 suggestive of unsuccessful catheterization (9).

To further validate the role of prolactin during IPSS, we conducted a retrospective analysis of prolactin levels in basal and CRH-stimulated IPSS samples in 29 patients with ACTH-dependent CS (17 proven CD, 8 proven EAS, 4 occult cases with biochemical testing suggestive of EAS) (10). Measurement of prolactin levels was not found to be useful in patients with an ACTH IPS/P ratio consistent with CD. All but one patient with surgically proven CD were accurately diagnosed without prolactin measurement. One CD patient with false negative IPSS result had abnormal venous anatomy (small right petrosal vein) and a baseline ipsilateral IPS/P prolactin ratio of 1.0 indicating unsuccessful catheterization. The prolactin-normalized ACTH IPS/P ratio of 1.3 correctly diagnosed CD. One patient with surgically proven EAS was found to have falsely positive IPSS results with both absolute ACTH and prolactin-normalized ACTH IPS/P ratios. This was thought to be due to unsuppressed normal corticotropes in the setting of cyclic hypercortisolism. Confirmation of sustained hypercortisolism therefore remains a prerequisite for IPSS.

Two patients with occult CS (biochemical testing suggestive of EAS) and one with proven EAS in our study (10) had prolactin normalized ACTH IPS/P ratios in the 0.8–1.3 range and would be incorrectly diagnosed with CD using the cut-off of 0.8 proposed by Findling and colleagues (8). Our data suggested that a normalized ACTH/prolactin IPS/P ratio of no more than 0.7 is consistent with EAS and a ratio of 1.3 or higher implies CD with values between 0.7 and 1.3 being indeterminate.

Recently, Grant et al. retrospectively analyzed prolactin levels in IPSS samples from 83 patients with ACTH-dependent CS (72 proven CD, 10 EAS, 1 indeterminate) (11). Five of the 72 proven CD patients had false negative IPSS results. Prolactin normalized ACTH IPS/P ratios correctly diagnosed CD in all five patients using a cut-off of 0.8. One patient with no central to peripheral ACTH gradient on IPSS had a prolactin normalized ACTH IPS/P ratio greater than 0.8 suggestive of a pituitary source of ACTH. However, the patient remained hypercortisolemic after three unsuccessful transsphenoidal surgeries and radiotherapy treatment.

Taken together these data suggest that prolactin measurement during IPSS helps to identify unsuccessful catheterization during IPSS, particularly in those with negative results. However, the criteria for interpretation of prolactin normalized ACTH IPS/P ratios need further study. Although a prolactin normalized ACTH IPS/P ratio of < 0.7 suggests EAS and one of > 1.3 indicates CD, the implication of values between 0.7 and 1.3 is not fully clear.

Baseline versus CRH-stimulated prolactin values during IPSS

Most studies have used the baseline (before CRH) prolactin values to assess successful catheterization during IPSS. A potential disadvantage of using only baseline prolactin values to evaluate the adequacy of catheterization is the possibility of catheter displacement at the time of post-CRH sampling. Many studies have documented an increase in prolactin levels after administration of ovine CRH in patients with ACTH-dependent CS. Loli et al. reported a CRH-induced increase in PRL levels during IPSS in 11 of 22 patients with CD even though none of the pituitary tumors stained positive for prolactin on immunocytochemistry. Findling et al. reported a greater than 10% increase in prolactin in the dominant petrosal sinus after CRH in 21 of 44 (48%) CD patients (8). In our study, six EAS (75%) and 10 CD (59%) patients had greater than 10% increase in prolactin levels after CRH. On the other hand, six (35%) patients with CD and two (25%) with EAS had a greater than 10% decrease in prolactin levels after CRH administration. Moreover, use of post-CRH prolactin values did not improve diagnostic accuracy of IPSS and led to some indeterminate results (10). Given variable changes in prolactin levels after CRH stimulation, regardless of the type of CS, the use of post-CRH prolactin values to assess successful catheterization may be unreliable. Intermittent fluoroscopy during IPSS may be a more useful way to prevent errors related to catheter displacement during post-CRH sampling.

Ipsilateral versus bilateral prolactin measurement

Most studies used the prolactin value on the IPS side with the dominant ACTH ratio to evaluate successful catheterization. However, abnormal drainage on the contralateral side can give falsely lower ACTH values and may mask an otherwise dominant ACTH IPS/P result. Useful information regarding unsuccessful catheterization can therefore be obtained from both sides in IPSS.

In the series by Mulligan et al, one of the 35 proven CD patients had false negative IPSS ACTH results. Using the dominant (left) side, the baseline prolactin IPS/P ratio of 2.2 was suggestive of successful catheterization and the corresponding normalized ACTH/prolactin ratio of 0.77 was in the indeterminate zone. On the other hand, the contralateral (right) prolactin IPS/P ratio of 1.0 was indicative of unsuccessful catheterization and the normalized ratio of 1.3 correctly diagnosed CD (9). In our study, two patients with proven EAS and negative ACTH IPSS results had contralateral prolactin IPS/P values suggestive of unsuccessful catheterization (10). Resulting normalized ACTH/prolactin IPS/P ratios (0.8, 1.0) were in the indeterminate zone.

Given lack of sufficient evidence, the routine use of contralateral prolactin values for correction of ACTH IPSS results is not recommended. Future studies should evaluate the role of contralateral prolactin IPS/P ratios in improving diagnostic accuracy of IPSS.

Role of Venography

Review of the IPSS venogram is an essential part of the interpretation of IPSS, especially in patients without a central to peripheral ACTH gradient in whom abnormal venous anatomy would raise the concern for a false negative IPSS result. However, not all cases with abnormal venous anatomy are associated with false negative results. In a series of 501 patients with surgically proven CD who underwent IPSS, four patients with false negative IPSS results had a hypoplastic or plexiform inferior petrosal sinus. However, despite the presence of unilateral or bilateral atrophic inferior petrosal sinuses in 25 of 100 consecutive control patients with proven CD, none had a false negative IPSS result (4). Thus, it is recommended to evaluate the venous anatomy only in the case of negative results.

IPSS venogram results are not always concordant with the baseline prolactin IPS/P ratios. In our study (10), eight of 11 patients with abnormal venous anatomy or drainage had a baseline prolactin IPS/P ratio ≥ 1.8 consistent with successful catheterization while three had a prolactin IPS/P ratio < 1.8, including one CD patient with a false negative result. All patients with a peak IPS/P ACTH ratio suggestive of EAS and normal venous anatomy had a baseline prolactin IPS/P ratio ≥ 1.8 consistent with successful catheterization. Therefore, in our series, prolactin measurement provided additional useful information only in those cases with negative IPSS results who also had an abnormal venous anatomy.

On the other hand, Mulligan et al. (9) reported two CD patients with equivocal or false negative IPSS results and normal venous anatomy who had prolactin IPS/P ratios < 1.8 suggestive of unsuccessful catheterization. All three index cases with false negative IPSS results in the series by Findling et al. (8) were documented to have successful catheterization based on contrast reflux into the contralateral cavernous sinus on venogram but had prolactin IPS/P ratios suggestive of unsuccessful catheter placement. Therefore, although prolactin measurement in cases with negative IPSS results and abnormal venous anatomy is clearly beneficial, its value in cases with apparently normal venous anatomy and drainage needs further investigation.

Role of prolactin in lateralization during IPSS in Cushing’s disease

Although the utility of IPSS in differentiating between a pituitary and ectopic source of ACTH is well-established, its role in localizing pituitary microadenomas remains controversial. An inter-sinus ACTH ratio of 1.4 or greater before or after CRH administration has been proposed for ipsilateral localization of pituitary microadenomas (17). In a systematic review looking at 19 studies and 313 cases, the diagnostic accuracy of IPSS in localizing pituitary adenomas was reported at 78% (ranging from 50–100% in individual studies) using findings at pituitary surgery as gold standard (18). Although the results of IPSS lateralization may guide the neurosurgeon as to where to begin the initial examination of the pituitary gland, full exploration is necessary to avoid missing an adenoma as the source of ACTH.

In an attempt to improve localization, a few studies have looked at the measurement of other pituitary hormones including TSH, α-subunit and prolactin during IPSS to correct for dilution by non-pituitary blood draining into the inferior petrosal sinus (7, 12, 19, 20). In an early study by Zovickian and colleagues, measurement of petrosal sinus TSH, prolactin and α-subunit levels failed to aid in localization of ACTH-secreting pituitary microadenomas and led to erroneous lateralization in four of 11 cases (12). McNally et al. reported correct localization of unilateral pituitary adenomas in four of five cases using TRH-stimulated TSH and prolactin values (7). In a series of nine patients with CD and unilateral adenomas, prolactin measurement during IPSS failed to improve upon the localization of pituitary microadenomas using an inter-sinus ACTH gradient. However, the authors compared the inter-sinus ratio of prolactin to an inter-sinus ratio of ACTH in predicting the side of the adenoma rather than using the prolactin levels to correct ACTH values (19).

Recently, Mulligan et al. compared inter-sinus ACTH gradients before and after normalization with prolactin levels with surgical findings and pathology in 28 patients with surgically proven CD (20). Using an inter-sinus ratio criterion of 1.4 or more, tumor location was correctly identified with uncorrected ACTH levels in 15 of 28 (54%) patients and with prolactin-adjusted ACTH values in 21 of 28 (75%) patients. Combination of data from MRI of the pituitary gland and prolactin-adjusted ACTH inter-sinus ratios led to further improvement in localization of pituitary microadenomas (82%). Although these results seem promising, they are based on a small number of patients. Prospective larger studies are needed before routine measurement of prolactin for tumor localization in CS patients with positive IPSS results can be recommended.

Summary

Measurement of prolactin levels during IPSS can improve the diagnostic accuracy and decrease false negative results when ACTH IPS/P ratio suggests EAS. However, it is not required in cases with positive IPSS results. Reviewing the venogram in cases with negative IPSS results is vital to identify anomalous venous anatomy or drainage. We recommend storing baseline blood samples obtained during IPSS and measuring prolactin levels in bilateral inferior petrosal sinus and peripheral samples if all ACTH IPS/P ratios before and after CRH stimulation suggest an ectopic source of ACTH (Fig 1). Baseline ipsilateral prolactin IPS/P ratio of 1.8 or more suggests successful catheterization during IPSS. Prolactin-normalized ACTH IPS/P ratios can then be used to differentiate between a pituitary and ectopic source of ACTH with values ≤ 0.7 suggestive of EAS while those ≥ 1.3 indicative of CD. However, there is a large zone of uncertainty in the interpretation of these normalized ratios and larger prospective studies at less specialized centers are needed to better identify optimal diagnostic thresholds. The role of contralateral prolactin IPS/P ratios, post-CRH prolactin values and prolactin-adjusted ACTH inter-sinus ratios for tumor localization in CD also remains unclear and needs further investigation.

Figure 1.

Suggested algorithm for evaluation of ACTH-dependent Cushing's syndrome. BL, Baseline; CS, Cushing's syndrome; HD DST, high-dose dexamethasone suppression test; PRL, prolactin; IPSS, inferior petrosal sinus sampling; IPS/P, inferior petrosal sinus to peripheral; TSS, transsphenoidal surgery. Adapted from “Prolactin as a marker of successful catheterization during IPSS in patients with ACTH-dependent Cushing’s syndrome” by Sharma ST, Raff H and Nieman LK, 2011, J Clin Endocrinol Metab; 96(12):3687-94 (http://jcem.endojournals.org/content/96/12/3687.long)