| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Other Original Articles |
Departments of Endocrinology, Diabetes and Metabolism (S.T., V.V., N.T.) and Radiology (J.K.), Evangelismos Hospital; 2nd Department of Internal Medicine, Research Institute and Diabetes Center, Athens University and Hellenic National Diabetes Center (C.T., P.T., S.A.R.); 10676 Athens, Greece
Address all correspondence and requests for reprints to: S. Tsagarakis, M.D., Ph.D., Department of Endocrinology, Diabetes and Metabolism, Evangelismos Hospital, 10676 Athens, Greece. E-mail: . stsagara{at}otenet.gr
Abstract
The role of desmopressin, alone or in combination with CRH, in the differential diagnosis between Cushing’s disease (CD) and ectopic ACTH secretion (EAS) still remains uncertain. Based on existing data, the desmopressin test is regarded as an alternative to the CRH stimulation test and, when given in combination with CRH, it has been suggested to completely discriminate between patients with CD and EAS. However, assessment of these tests has been limited in only a small number of patients with EAS. Desmopressin is a relatively specific V2 vasopressin receptor (V2R) agonist. Although expression of V3 vasopressin receptor (V3R) is common in tumors with EAS, the expression of V2R has not been extensively investigated. In the present study, we report our findings of the desmopressin and the combined CRH-desmopressin test in a series of patients with CD and EAS; also, the expression of V2R and V3R was investigated in tumors with EAS by a RT-PCR method.
We assessed a cohort of 31 patients with ACTH-dependent Cushing’s syndrome, including 26 patients with CD and five cases with histologically confirmed EAS. To avoid bias of predetermined criteria, univariate curves of the receiver operating characteristics (ROC) were constructed by plotting the sensitivity against 1-specificity at each level of the percent cortisol (F) and ACTH responses to these tests. Following desmopressin administration there was an overlap of the percent F and ACTH responses among patients with CD and EAS, and the area under the ROC curve for both these responses was not significantly different than that occurring by chance. This was also true for the percent F response following the combined CRH-desmopressin test. However, the area under the ROC curve for the percent ACTH rise following the combined test was significantly different; the point of the ROC curve closest to 1 corresponded to a percent ACTH rise of 218% (88% sensitivity and 80% specificity). Expression of V2R and V3R mRNA was investigated in four of the five excised tumors with EAS and revealed the presence of the V2R in all, whereas the V3R mRNA was expressed in three of these cases.
In conclusion, in this series the desmopressin test produced a significant overlap of responses between CD and patients with EAS and, therefore, is of limited value in the differential diagnosis of the ACTH-dependent Cushing’s syndrome. This is most probably due to the expression of the V2R in tumors with EAS. Moreover, following the combined CRH-desmopressin test only the ACTH but not the F responses were diagnostically useful, but still far from completely discriminating patients with CD and EAS.
IT HAS LONG been known that vasopressin contributes to the regulation of the hypothalamo-pituitary adrenal axis. Although it is a weak ACTH secretagogue, vasopressin has a significant synergistic effect with CRH, thus potentiating several-fold pituitary ACTH secretion (1). In the past, several investigators have used the administration of vasopressin analogs (lysine- or arginine-vasopressin), alone or in combination with CRH, to distinguish between the forms of ACTH-dependent Cushing’s syndrome (CS) (2, 3, 4, 5). The majority of patients with pituitary-dependent CS had positive ACTH and cortisol (F) responses to vasopressin analogs, whereas no such responses were observed in most of the few patients with ectopic ACTH secretion (EAS) tested. However, because administration of these vasopressin analogs was commonly associated with several side effects due to their smooth-muscle constricting actions, these tests have not gained wide acceptance for their use in the investigation of patients with CS. Desmopressin is a long-acting vasopressin analog with selective V2 agonist activity and, hence, devoid of the V1 receptor-mediated pressor side effects. Malebri et al. (6) were the first who administered desmopressin in patients with various forms of CS; desmopressin administration produced a significant rise of F secretion in the majority of patients with Cushing’s disease (CD) whereas patients with other forms of CS were unresponsive. Similar results were also reported in subsequent studies (7, 8, 9). More recently, Newell-Price et al. (10) introduced a combined test using CRH plus desmopressin for the differential diagnosis of ACTH-dependent CS. Interestingly, in this series the combined administration resulted in nonoverlapping F responses between the patients with CD and EAS, thus achieving a complete discrimination between these entities.
So far, the administration of desmopressin alone or in combination with CRH has been limited to only a small number of patients with EAS, and the use of these tests in discriminating between CD and EAS is not firmly established (11). Moreover, despite the demonstration that a number of tumors with EAS express V3 vasopressin receptor (V3R) (12), the expression of the V2 vasopressin receptor (V2R), for which desmopressin has a relatively selective activity, has not been extensively investigated in these tumors. Here, we report our findings of the desmopressin and the combined desmopressin-CRH test in a series of patients with ACTH-dependent CS. We also analyzed the expression of both V2R and V3R subtypes in the tumors with EAS by a RT-PCR method.
Patients and Methods
A series of 31 patients with ACTH-dependent CS were retrospectively analyzed. These included 26 patients with CD (Table 1
) and five cases with histologically confirmed EAS (Table 2). The diagnosis of CD was histologically confirmed in 14 patients following transsphenoidal surgery. In the patients with negative or unavailable histology the diagnosis was confirmed by a curative outcome following transsphenoidal surgery (n = 6). In unoperated patients (n = 2) or when no tumor was found at surgery and the patient was not cured (n = 4) the diagnosis of CD was based on stimulated central/peripheral ACTH gradients higher than 2, obtained during bilateral inferior petrosal sinus sampling. In the patients with EAS the diagnosis of an ectopic ACTH-producing tumor was histologically and immunohistochemically confirmed; these included three bronchial carcinoid tumors and two medullary thyroid carcinomas. Four of these patients had undergone bilateral inferior petrosal sampling and had basal and stimulated with a combination of CRH-desmopressin (13) ACTH gradients of less than 2 (Table 2).
|
|
Hormone measurements
Serum F was assayed using a direct RIA featuring an iodine-125 radioligand and F antibody-coated tubes (Coat-A-Count, Cortisol RIA; Diagnostic Products, Los Angeles, CA). The reported sensitivity of the assay used is 6 nmol/liter, with the lower detection limit for routine use set at 28 nmol/liter. Intra- and interassay coefficients of variation for F concentrations of 28, 140, and 750 nmol/liter were 6%, 2.5%, 4% and 6%, 6%, and 4.5%, respectively. Plasma ACTH was measured by a highly specific immunoradiometric assay (Nichols Institute Diagnostics, San Juan Capistrano, CA). Inter- and intra-assay coefficients of variation were 8% and 4%, respectively.
Calculation of responses
For all tests, the responses were calculated as a percent rise in the mean circulating ACTH (at 15 and 30 min) and F (at 30 and 45 min) values above the mean basal values (at -15 and 0 min). Following desmopressin administration a percent ACTH and F rise of more than 50% and more than 20%, respectively, was considered as indicating a positive response, as described previously for the CRH test (14). However, to avoid bias of predetermined criteria, receiver operating characteristics (ROC) curves were also constructed by plotting the sensitivity against 1-specificity at each level of the percent F and ACTH responses to these tests.
Statistics
Variables were expressed as the mean ± SEM. Statistical analysis to identify differences between the tests in patients with CD was carried out with ANOVA for repeated measures, followed by the Tukey’s test for multiple comparisons. ROC curves were constructed to examine the diagnostic test performance—that is the ability to discriminate between patients with CD and EAS. Sensitivity against 1-specificity was plotted at each level, and the area under the curve was computed by the nonparametric Wilcoxon statistic (15). Area under the curve represents the probability of correctly identifying CD and EAS patients. A value of 0.5 means that the test is no better than chance. The level of significance was set at 0.05 for all statistical tests. The statistical analysis was performed using the SPSS for Windows statistical package (version 10.0; SPSS, Inc., Chicago, IL).
RNA preparation and RT-PCR
Total RNA was isolated separately from four of the five excised tumors (two medullary thyroid carcinomas and two bronchial carcinoids) using the TriPure isolation reagent (Roche Molecular Biochemicals, Mannheim, Germany). The integrity of RNA samples was determined on agarose gels (1.2%) and spectophotometrically, using the absorption ratio at 260/280 nm. The RNA quantification was done by measuring the absorption at 260 nm. For the removal of any residual DNA contamination, RNA samples were preincubated with DNase (Promega Corp., Madison, WI) at 37 C for 25 min and then at 65 C for 10 min before the reverse transcription (RT) and PCR method. Four micrograms of total RNA from the tumors were reverse transcribed, using random hexanucleotide (200 pmol; Biolabs, Hertfordshire, UK), Moloney murine leukemia virus reverse transcriptase (400 U; Promega Corp.), RNasin inhibitor (20 U; Promega Corp.), and dNTPs (0.4 mM) at 37 C for 90 min. This was followed by the addition of 400 U Moloney murine leukemia virus to the reaction, and the incubation was continued for another 60 min. The RT reaction was terminated with the addition of 3 U RNase H (Life Technologies, Inc., Paisley, UK) for 30 min at 37 C. The PCR reaction (35 cycles) was subsequently performed in a final volume of 25 µl, using 20 pmoles of each primer, and 5 U Taq polymerase (Promega Corp.) at 94 C for 30 sec, 62 C for 30 sec, and 72 C for 90 sec. PCR reactions were also performed in RNA samples before RT to exclude DNA contamination.
Primer design. Three sets of specific primers were used to amplify PCR products from the V2, V3, and CRH receptor (CRHR) mRNA transcripts. Oligonucleotides 5'-AGCAGCTGTGGCCAAGACTGTAG-3' (1025–1047) and 3'-CAAGCTTCGCAGCTCTGAGGACAC-5' (1252–1229) were used to amplify a 229-bp PCR product from V2 receptor mRNA transcript (GenBank accession no. NM_000054). Oligonucleotides 5'-GCTGGGCTCCCTTCTTCAGTGTCC-3' (1085–1108) and 3'-AGGGTTAGGATGAGGCTGAGGCTG-5' (1398–1374) were used to amplify a 313-bp PCR product from V3 receptor mRNA transcript (GenBank accession no. D31833). Oligonucleotides 5'-CGCGTGAATTACTCCGAGTGCCAG-3' (323–346) and 3'-CCCTCGCCGAACATCCAGAAGAAG-3 (666–643) were used to amplify a 344-bp PCR product from CRHR mRNA transcript (GenBank accession no. L23332).
Enzyme restriction. The V2R, V3R, and CRHR RT-PCR products were digested at 37 C with HpaII, Sau3AI, and AvaI for 1 h, and the products were run on a 1.3% agarose gel and stained with ethidium bromide. HpaII was predicted to cut the V2R PCR product at position 1128 into two fragments (104 and 126 bp). Sau3AI was predicted to cut V3R PCR product at position 1214 into two fragments (180 and 132 bp). AvaI was predicted to cut CRHR PCR product at position 571 into two fragments (248 and 96 bp).
Results
The percent F and ACTH responses to desmopressin in each patient are shown in Fig. 1. In the patients with CD, 19 of 26 (73%) and 21 of 26 (80%) had a percent F and ACTH response of more than 20% and more than 50%, respectively. Also, three of five patients with EAS had F and ACTH responses above these levels. ROC curve analysis of the desmopressin test showed that both the percent F and ACTH rise had an area under the curve not significantly different than that occurring by chance (Fig. 2).
|
|
). The percent F and ACTH responses to the combined CRH-desmopressin in each patient with CD and EAS are shown in Fig. 4. ROC curve analysis showed that the percent F rise following the combined CRH-desmopressin test had an area under the curve that was not significantly different than that occurring by chance. However, the area under the ROC curve for the percent ACTH rise was significantly different (Fig. 5). The point closest to 1 corresponded to a percent ACTH rise of 218% (sensitivity, 88%; specificity, 80%). In this series, similar calculations for the CRH test showed that the points closest to 1 corresponded to a percent F of more than 20% and a percent ACTH response of more than 35%, and both these cutoff levels had 73% sensitivity and 100% specificity in discriminating between patients with CD and EAS (data not shown).
|
|
|
Using specific primers for the V2 receptor, a single band of the expected size (
229 bp) was observed after the RT-PCR in all four studied tumors (patients 1, 2, 3, and 5) (Fig. 6). By contrast, using specific primers for the V3R, a single band of the expected size (313 bp) was detected in the tumors from patients 2, 3, and 5 but not from patient 1. With specific primers for the CRHR, a single band of the expected size (344 bp) was observed only in patient 3 (Fig. 6). PCR without prior RT detected no message of the V2R, V3R, or the CRHR in the tumors examined. The specificity of the PCR signal for the V2R, V3R, and CRHR was confirmed by enzyme restriction with HpaII, XhoI, and AvaI, respectively. As predicted, HpaII cut the V2R PCR product at one site into two smaller fragments (104 and 126 bp), Hho I cut the V3R PCR product at one site to two smaller fragments (180 and 139 bp), and AvaI cut the CRHR PCR product into two fragments (248 and 96 bp) (data not shown).
|
This study provides evidence that the expression of V2R in tumors with EAS limits the diagnostic use of desmopressin in the differential diagnosis of ACTH-dependent CS. Thus, in contrast to most previous studies, we found a fair amount of overlap between the responses of 5 histologically confirmed cases of EAS and 26 cases of CD. In the majority of CD patients the diagnosis was confirmed histologically and/or by surgical outcome; in only six patients the diagnosis was based on bilateral inferior petrosal sinus sampling gradients. All six patients responded to desmopressin and the combined CRH-desmopressin administration with a substantial rise in ACTH and F concentrations, and, therefore, inclusion of these cases does not alter the conclusions drawn by this study.
Several previous studies, based on findings showing positive F and ACTH responses to desmopressin in the majority of patients with CD but not in the patients with EAS, have suggested that the desmopressin test may be useful in the differential diagnosis of ACTH-dependent CS. However, a caveat of these studies is that only few patients with EAS have been investigated, including also some nonhistologically confirmed cases. Moreover, different cutoff criteria were used to define a positive response. Thus, in the study of Malebri et al. (6), by using the criteria adopted for the interpretation of CRH test, desmopressin administration produced a significant rise of F secretion in 15 of 16 patients with CD whereas 1 patient with proven EAS and 2 patients with suspected EAS were unresponsive. Colombo et al. (8), using similar criteria, demonstrated a positive F and ACTH response in 14 of 17 of patients with CD; no response was observed in the single patient with nonhistologically confrmed EAS studied. Sakai et al. (7), using a high percent ACTH rise threshold of 120%, reported a positive ACTH response in all 10 patients with CD, whereas all 3 patients with EAS were unresponsive to desmopressin. In a previous study (9) we also did not observe a positive response to desmopressin in three patients, including one patient with still suspected EAS and the two patients with histologically confirmed EAS presented also in this study. In the present series, however, desmopressin administration in three additional consecutive patients with histologically confirmed EAS produced F and ACTH responses indistinguishable from those observed in the patients with CD. ROC curve analysis, performed to avoid bias of predetermined criteria, also confirmed the inability of the desmopressin to differentiate between patients with CD and EAS in this series. Similar findings have been recently reported by Terzolo et al. (16), in a group of 19 patients with CD and 5 patients with EAS. Newell-Price et al. (10) have also reported positive F and ACTH responses to desmopressin in one of five and three of five patients with EAS, respectively. Interestingly, the single patient in the study by Newell-Price et al. (10) with a robust response to desmopressin had ectopic ACTH production originating from a medullary thyroid carcinoma. Notably, two of the responders in our study also had medullary thyroid carcinomas, suggesting that a higher rate of desmopressin responsiveness may be observed in cases of this tumor type. However, two recently reported cases (17, 18) and one case of this series of ACTH-producing bronchial carcinoid tumors also had positive F and ACTH responses to desmopressin. It seems, therefore, that an increasing number of patients with various tumor types leading to EAS present with false positive responses to desmopressin, which greatly diminishes the value of the desmopressin test as an adjunctive tool in the differential diagnosis of ACTH-dependent CS.
The mechanism underlying the desmopressin responsiveness of EAS tumors is not well established. De Keyzer et al. (12) demonstrated that the V3R are commonly expressed in tumors presenting with EAS. Desmopressin, however, is a long-acting vasopressin analog (19) that has a high relative affinity for the V2R but a relatively low affinity for the V3R receptors. In contrast to the nonspecific vasopressin analogs, such as Lysine or Arginine-vasopressin, desmopressin has a modest or no ACTH-releasing activity when given in normal human volunteers, indicating that the V3R that predominate in the normal pituitary are insufficient to stimulate ACTH secretion (9, 20, 21, 22). In the present series, receptor expression studies in four of the five excised tumors with EAS demonstrated the presence of V2R mRNA in all the tumors examined, whereas the V3R mRNA was expressed in three of the four cases. It is of note that the patient with only V2R expression in the tumor had a substantial ACTH and F response to desmopressin (patient 1, in Fig. 6
). It is, therefore, suggested that the observed responsiveness of EAS tumors to desmopressin is most likely due to the ectopic expression of V2R by these tumors. Interestingly, two recently reported bronchial carcinoid tumors with positive in vivo responses to desmopressin also expressed the V2 subtype of vasopressin receptors (17, 18). V2R expression was also found in one patient categorized as nonresponder to the in vivo administration of desmopressin. However, in this particular patient, an erratic ACTH secretion may have obscured an in vivo response. Examination of more tumors from nonresponder patients to desmopressin would allow resolving the relationship between V2R expression and in vivo desmopressin responsiveness. It is finally of note that expression of CRHR mRNA in EAS tumors was only observed in a single case that is in concordance with the observed lack of CRH responsiveness during the in vivo administration of this agent.
Two recent studies raised interest in testing with peptide combinations (CRH plus vasopressin analogs), as a means to improve the sensitivity of the CRH test in establishing the diagnosis of CD. The observation that the combination of ovine CRH plus AVP resulted to positive ACTH and F responses in all patients with CD tested led Dickstein et al. (5) to propose that such a combined administration may provide a more reliable test for the differential diagnosis of ACTH-dependent CS. More recently, Newell-Price et al. (10) introduced a combined test, using human CRH and the better tolerated desmopressin in differentiating between the two forms of ACTH-dependent CS. In this latter study, all 17 patients with CD tested gave a F response, which did not overlap with the responses of the 5 patients with the EAS included in this study, thus achieving a complete discrimination between these entities. In good agreement, in our CD patients, stimulation with human CRH plus desmopressin produced significantly higher ACTH and F responses compared with those observed when CRH or desmopressin were given alone. However, F responses in our patients with histologically confirmed EAS overlapped with those observed in the patients with CD, and furthermore ROC curve analysis showed that the percent F rise following the combined test did not discriminate between patients with CD and EAS. A similar analysis for the percent ACTH response, however, showed that a cutoff of 218% had 88% sensitivity and 80% specificity in discriminating between CD and EAS patients. It is of note that at variance with our data Newell-Price et al. (10) found the percent ACTH rise less helpful beside the adoption of a much higher cutoff percent ACTH rise (350%). Application of such a higher cutoff in our series greatly diminishes the sensitivity of the procedure to 69% without improving its specificity. Differences in the assays used (ACTH RIA vs. immunoradiometric assay) and the unavoidable recruitment of only few patients with EAS may be partly responsible for the discrepancies between these studies. Another possible explanation for these discrepancies is that in our series we have included a larger number (8 of 26) of CRH unresponsive patients with CD. However, as the main value of the combined test is to detect such CRH unresponsive patients our population of CD patients is more appropriate for the assessment of the diagnostic use of the combined test. Certainly, in this study, the combined CRH-desmopressin test led to a higher sensitivity than the CRH test (88 vs. 73%) but at the cost of a lower specificity (80 vs. 100%), despite the adoption of a much higher percent ACTH cutoff rise for the combined test (218 vs. 35%).
It should be noted that like Newell-Price et al. (10) we also used the human sequence CRH peptide. However, ovine CRH is a more potent stimulator in humans of ACTH and F release. Despite a quantitatively higher response to ovine CRH in the patients with CD (23), the diagnostic performance of both these peptides is comparable in differentiating CD from EAS (24, 25). However, it will be worthy investigating whether the use of the more potent ovine CRH in combination with desmopressin could have an even better discriminatory power than its human homologue.
In conclusion, despite the initial suggestions that the desmopressin test may be a good alternative in the differential diagnosis of ACTH-dependent CS, accumulating data suggest that this test is of limited value in the differential diagnosis of ACTH-dependent CS. This is most probably due to the expression of the relatively desmopressin selective V2R subtype in these tumors. Finally, although ACTH responses to the combined CRH-desmopressin test may be diagnostically useful, this test does not completely discriminate, as previously suggested, patients with CD from EAS.
Acknowledgments
Footnotes
Present address for C.S.: Intensive Care Unit, Evangelismos Hospital, 10676 Athens, Greece.
Abbreviations: CD, Cushing’s disease; CRHR, CRH receptor; CS, Cushing’s syndrome; EAS, ectopic ACTH secretion; F, cortisol; ROC, receiver operating characeristics; RT, reverse transcription; V2R, V2 vasopressin receptor; V3R, V3 vasopressin receptor.
Received August 7, 2001.
Accepted December 18, 2001.
References
This article has been cited by other articles:
 |
|
 |
|
  M. C. Machado, S. Valeria de Sa, M. L. Correa-Giannella, R. R. Giorgi, M. A. A. Pereira, V. A. S Cescato, D. Giannella-Neto, and L. R. Salgado Association between tumoral GH-releasing peptide receptor type 1a mRNA expression and in vivo response to GH-releasing peptide-6 in ACTH-dependent Cushing's syndrome patients Eur. J. Endocrinol., May 1, 2008; 158(5): 605 - 613. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
F Castinetti, I Morange, H Dufour, P Jaquet, B Conte-Devolx, N Girard, and T Brue Desmopressin test during petrosal sinus sampling: a valuable tool to discriminate pituitary or ectopic ACTH-dependent Cushing's syndrome Eur. J. Endocrinol., September 1, 2007; 157(3): 271 - 277. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Tsagarakis, D. Vassiliadi, I. S. Kaskarelis, J. Komninos, E. Souvatzoglou, and N. Thalassinos The Application of the Combined Corticotropin-Releasing Hormone plus Desmopressin Stimulation during Petrosal Sinus Sampling Is Both Sensitive and Specific in Differentiating Patients with Cushing's Disease from Patients with the Occult Ectopic Adrenocorticotropin Syndrome J. Clin. Endocrinol. Metab., June 1, 2007; 92(6): 2080 - 2086. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R M Testa, N Albiger, G Occhi, F Sanguin, M Scanarini, S Berlucchi, M P Gardiman, C Carollo, F Mantero, and C Scaroni The usefulness of combined biochemical tests in the diagnosis of Cushing's disease with negative pituitary magnetic resonance imaging Eur. J. Endocrinol., February 1, 2007; 156(2): 241 - 248. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
G. Arnaldi, A. Angeli, A. B. Atkinson, X. Bertagna, F. Cavagnini, G. P. Chrousos, G. A. Fava, J. W. Findling, R. C. Gaillard, A. B. Grossman, et al. Diagnosis and Complications of Cushing's Syndrome: A Consensus Statement J. Clin. Endocrinol. Metab., December 1, 2003; 88(12): 5593 - 5602. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
G. Arnaldi, T. Mancini, B. Kola, G. Appolloni, S. Freddi, C. Concettoni, I. Bearzi, A. Masini, M. Boscaro, and F. Mantero Cyclical Cushing's Syndrome in a Patient with a Bronchial Neuroendocrine Tumor (Typical Carcinoid) Expressing Ghrelin and Growth Hormone Secretagogue Receptors J. Clin. Endocrinol. Metab., December 1, 2003; 88(12): 5834 - 5840. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
