Von Willebrand factor and fibrinolytic parameters during the desmopressin test in patients with Cushing's disease

Abstract

AIMS

Desmopressin, a vasopressin analogue, is used for various clinical purposes, including haemostasis and, in recent times, the diagnostic work-up of patients with Cushing's syndrome, a condition associated with a known prothrombotic profile. We decided to evaluate whether and to what extent a diagnostic dose of desmopressin induces significant changes in endothelial parameters in patients with Cushing's disease (CD) and obese and normal weight controls.

METHODS

Twelve patients with CD, 10 obese and five normal weight controls were studied. Von Willebrand antigen (VWF : Ag), tissue plasminogen activator (t-PA) and plasminogen activator inhibitor type 1 (PAI-1) were measured at baseline and up to 4 h after 10 µg desmopressin i.v.

RESULTS

Desmopressin 10 µg transiently increased VWF : Ag and t-PA and decreased PAI-1 in all subjects. The magnitude of the VWF : Ag and t-PA increases after desmopressin was comparable in the three groups (VWF : Ag peak-to-basal ratio 1.9 ± 0.17, 1.5 ± 0.11 and 1.8 ± 0.13 and t-PA peak-to-basal ratio 1.6 ± 0.18, 1.6 ± 0.20 and 1.8 ± 0.24 for CD, obese and controls, respectively, all NS). The PAI-1 decrease observed in patients with CD was comparable with obese (0.7 ± 0.07 and 0.6 ± 0.09, NS) and controls (0.7 ± 0.07 vs. 0.4 ± 0.09, P = 0.08).

CONCLUSIONS

Administration of desmopressin to patients with CD for diagnostic purposes induces a transitory increase in VWF : Ag counterbalanced by a decrease in PAI-1 and increase in t-PA. The magnitude of these changes is largely comparable with that observed in obese and normal weight controls. Our data show that testing with desmopressin does not induce disease-specific changes in endothelial markers in patients with CD.

WHAT THIS STUDY ADDS

• Administration of desmopressin to patients with Cushing's disease induces changes in endothelial cell markers comparable with those observed in obese and normal weight subjects. It follows, that desmopressin testing does not induce disease-specific untoward changes in coagulatory markers in patients with endogenous hypercortisolism and its use in this context appears safe.

WHAT IS ALREADY KNOWN ABOUT THIS SUBJECT

• Desmopressin is a known haemostatic agent and is also being used, albeit at lower doses, during the diagnostic work-up of Cushing's syndrome, a condition characterized by excess cortisol concentrations and frequent thromboembolic events. No study has yet evaluated whether admininistration of desmopressin for diagnostic purposes induces significant, adverse changes in endothelial cell markers in these patients.

Introduction

Desmopressin, a long-acting synthetic derivate of vasopressin, has various clinical uses, e.g. to contain diuresis in nocturnal enuresis and diabetes insipidus, as a haemostatic agent in bleeding disorders, such as von Willebrand's disease and mild haemophilia A, and, most recently, in the diagnostic work-up of endogenous Cushing's syndrome, i.e. cortisol excess due to a pituitary, adrenal or extrapituitary tumour. The former uses arise from the interaction of desmopressin with the V2 vasopressin receptor subtype whereas interaction with the V3 or V1B (AVPR1B) receptor subtype underlies the latter use. Upon desmopressin administration both receptors are expected to be activated, but little is known about any overlapping action.

Considerable interest in desmopressin testing safety arises from the knowledge that patients with Cushing's syndrome display a procoagulatory profile, characterized by variable increases in von Willebrand factor (VWF), Factor VIII (FVIII), plasminogen activator inhibitor-1 (PAI-1) and thrombin activity [1, 2], and, indeed, these patients are known to develop deep vein thrombosis and pulmonary embolism. These same factors mediate the haemostatic activity of desmopressin as the drug is known to increase VWF and FVIII concentrations [3]. Desmopressin also evokes a milder release of tissue plasminogen activator (t-PA) and decreases plasminogen activator inhibitor (PAI-1), the main inhibitor of fibrinolysis [3], thereby favouring fibrinolysis and preventing excessive blood clot formation.

No study has yet evaluated whether and to what extent administration of desmopressin to patients with Cushing's disease (CD) activates the coagulatory cascade and induces disease-specific changes in haemostatic factors. In this context, it is worth recalling that the dosage of desmopressin used by endocrinologists for diagnostic purposes is roughly half the one used by haematologists in bleeding disorders (10 µg and 18–20 µg, respectively). We decided to investigate the effects of a diagnostic dose of desmopressin on endothelial haemostatic markers in patients with Cushing's disease and, in comparison, in obese and normal weight controls.

Methods

Twelve patients with CD (two men, 10 women, age 47.6 ± 3.5 years, range 30–75 years, body mass index [BMI] 29 ± 1.84 kg m⁻², range 22–45 kg m⁻²) were enrolled in the study. The diagnosis of CD was established according to standard diagnostic criteria [4] and confirmed by trans-sphenoidal surgery. Ten subjects with simple obesity (BMI ≥ 30 kg m⁻², range 31–75 kg m⁻²) and five normal weight healthy subjects (BMI range 20–23.5 kg m⁻²) volunteered to serve as controls (Table 1). No patient/subject was taking drugs known to interfere with haemostasis and no patient suffered from epilepsy, ischaemic heart disease or uncontrolled hypertension. Thyroid function was within the normal range in all subjects. The study was approved by our Institution's Ethics Committee and informed consent was obtained from all subjects prior to testing.

Table 1.

Clinical and demographic data of study population

	Cushing's disease	Obese patients	Normal weight subjects
Number of patients	12 (2 M–10 F)	10 (1 M–9 F)	5 (1 M–4 F)
Age (years)	47 ± 3.55 (30–75)	45.5 ± 3.98 (25–68)	37 ± 5.49 (16–47)
BMI (kg m−2)	29.1 ± 1.69 (22–45)	47.5 ± 5.23* (31–75)	24.6 ± 2.25 (20–31)
Baseline VWF (U dl−1)	148.0 ± 19.05	154.7 ± 32.34	116.8 ± 13.57
Baseline PAI-1 (U ml−1)	12.9 ± 2.99†	38.9 ± 4.19*	7.6 ± 2.13
Baseline t-PA (ng ml−1)	8.6 ± 2.01	12.7 ± 0.85*	6.4 ± 0.80

Mean ± SEM, range in parentheses

^*P < 0.016 obese patients vs. Cushing's syndrome patients and normal weight subjects

^†P = 0.06 vs. normal weight subjects. P < 0.016 was used after applying Bonferroni's correction for multiple comparisons.

Study protocol

After an overnight fast, desmopressin (10 µg Minirin/desmopressin, Ferring Pharmaceuticals Ltd, Malmo, Sweden) was injected i.v and serial blood samples for VWF : Ag, PAI-1 and t-PA estimation obtained from an indwelling catheter. The catheter had been inserted in a forearm vein 30 min before desmopressin injection and samples were collected before and 15, 30, 45, 60, 90, 120 and 240 min after drug administration. Sampling was extended to 4 h compared with the 2 h endocrine protocol [5] as the endothelial marker response to desmopressin, usually expressed as a peak or nadir ratiovs. basal, is commonly evaluated within this time frame [6].

Assays

VWF antigen (VWF : Ag) was assayed using an enzyme-linked immunoadsorbent assay (ELISA), as previously described [7]. The normal range for VWF : Ag is 50–150 U dl⁻¹, sensitivity 0.1 U dl⁻¹. t-PA and PAI-1 were tested by commercially available ELISA, as reported previously [1]. All samples were assayed in the same batch and intra-assay coefficients of variation were 8%, 10% and 7% for VWF : Ag, t-PA and PAI-1, respectively.

Statistical analysis

Differences within groups were evaluated with the Wilcoxon test while those between groups were established by the Mann-Whitney test (Statview software, Abacus Concepts, Berkeley, USA). Values are given as mean ± standard error (SEM), differences are reported as median and 95% confidence intervals in square brackets [8]. P < 0.05 was considered statistically significant for paired comparisons and P < 0.016 for intergroup comparisons according to Bonferroni's correction. With the current patient sample, we expected to detect a difference of at least 25% in any variable at 0.05 type I error and 0.20 type II error (MedCalc Software, Mariakerke, Belgium).

Results

von Willebrand factor (VWF)

Mean baseline VWF : Ag appeared somewhat higher in patients with CD and obese subjects compared with controls but this failed to achieve statistical significance (Table 1). In detail, 42% of patients with CD and 20% of obese subjects presented with VWF : Ag concentrations higher than normal (i.e. >150 U dl⁻¹) whereas concentrations were normal in all normal weight subjects. No significant differences were observed between the three groups divided by sex, smoking, hypertension and blood group. A significant increase in VWF : Ag concentrations was observed within 60 min of desmopressin administration in patients with CD as well as in obese and normal weight subjects (Figure 1). This increase was followed by a progressive decrease over the next 2–4 h. Mean VWF : Ag peak was only slightly higher in patients with CD compared with obese and normal weight controls (347.3 ± 48.57, 248.6 ± 32.26 and 287.4 ± 40.27 U dl⁻¹, respectively, all comparisons NS; median differences: CD vs. obese 104 U dl⁻¹[−40, 222]; CD vs. normal 78 U dl⁻¹[−69, 296]). The peak-to-basal VWF : Ag ratio was comparable in the three groups (1.9 ± 0.17, 1.5 ± 0.11 and 1.8 ± 0.13 for CD, obese and normal weight subjects, respectively, all comparisons NS, Figure 1; median differences: CD vs. obese 0.33 [−0.06, 0.74]; CD vs. normal 0.01 [−0.51, 0.53]).

Figure 1.

Changes in endothelial markers after desmopressin administration to patients with Cushing's disease (filled circles, full line), obese subjects (empty circles, full line), healthy controls (filled circles, dotted line): von Willebrand antigen (VWF : Ag) in the upper panel, plasminogen activator inhibitor-1 (PAI-1) in the lower panel and tissue plasminogen activator (t-PA) in the middle panel

Tissue plasminogen activator (t-PA)

Baseline t-PA values were significantly higher in obese patients compared with patients with CD and normal weight subjects (Table 1). t-PA increased rapidly after desmopressin administration in all subjects, peaked around 30 min, then declined to pre-test values in patients with CD and normal weight subjects within 4 h (mean peak 17.7 ± 3.79 and 16.2 ± 1.56 ng ml⁻¹, respectively, NS; median difference 1.25 [−7.7, 9.5]). A more sustained increase in t-PA was apparent in obese subjects although the peak was comparable with CD (median difference 7.6 [−4, 25]). The peak-to-baseline ratio was comparable in the three groups (1.6 ± 0.18, 1.6 ± 0.20 and 1.8 ± 0.24 for CD, obese and normal weight subjects, respectively, NS; Figure 1; median differences: CD vs. obese – 0.25 [−2.15, 0.43]; CD vs. normal – 0.06 [−0.17, 0.39]).

Plasminogen activator inhibitor type 1 (PAI-1)

Not surprisingly, baseline PAI-1 values were significantly higher in obese patients in comparison with patients with CD and normal weight subjects whereas the difference between patients with CD and normal weight subjects was barely significant (Table 1). A decrease in PAI-1 was observed within the first hour after desmopressin injection (nadir 4.0 ± 1.10 for CD, 18.7 ± 5.42 for obese and 3.0 ± 1.12 U ml⁻¹ for normal weight controls, P < 0.05 vs. baseline) followed by gradual rise; PAI-1 concentrations in obese patients started out higher than in the other two groups, fell steeply and remained lower than pre-test values even 4 h after desmopressin bolus. On balance, patients with CD showed mean nadir-to-baseline ratios (Figure 1) comparable with obese patients (0.67 ± 0.07 vs. 0.58 ± 0.09, NS; median difference: 0.14 [−0.18, 0.29]) and barely higher than normal weight controls (0.67 ± 0.07 vs. 0.40 ± 0.09, P = 0.08; median difference 0.26 [0.03, 0.51]).

Discussion

Desmopressin, the long-acting vasopressin analogue, is employed in a variety of clinical settings but little is known on the possible overlapping actions of this drug. In particular, desmopressin administration has recently proven useful in the diagnostic work-up of Cushing's syndrome as it evokes marked ACTH and cortisol responses in patients with corticotroph tumours [9] and thus enables the differential diagnosis with pseudoCushing states [5]. Testing with desmopressin during inferior petrosal sinus sampling is also useful to distinguish Cushing's disease from ectopic ACTH secretion [10] and this latter use is an interesting alternative to constraints, both economic and distribution-wise, to corticotropin-releasing hormone, the classic stimulant for inferior petrosal sinus sampling. Lastly, desmopressin is a promising agent to establish remission or predict relapse after trans-sphenoidal surgery for Cushing's disease [11].

On the other hand, desmopressin has long been used as a haemostatic agent in patients with bleeding disorders as it triggers the release of VWF from endothelial cells and increases FVIII and, thus, activates the coagulatory cascade [3]. Concurrently, desmopressin promotes fibrinolysis [3], possibly to counteract excessive blood clot formation, by stimulating the release of t-PA, a profibrinolytic peptide, and reducing PAI-1 concentrations, the main inhibitor of clot lysis. No data are yet available, however, on these endothelial markers during desmopressin testing in patients with Cushing's disease and this is certainly worth evaluating given the known prothrombotic profile of these patients. Indeed, patients with Cushing's disease present with elevated VWF, FVII and PAI-1 concentrations [1, 2] and an increased incidence of deep vein thrombosis and pulmonary embolism [12]. It should be recalled that the dosage and route of administration of desmopressin used for diagnostic or haemostatic purposes differ considerably as endocrinologists administer a 10 µg i.v. bolus whereas haematologists inject 0.3 µg kg⁻¹ bodyweight (corresponding to 18–20 µg for normal weight individuals) either as an i.v. bolus or subcutaneously over 30 min. Desmopressin administration, even at lower doses, could alter pro-aggregating and fibrinolytic factors and could therefore raise some concerns in this patient population.

Our study shows that the increase in VWF : Ag induced by a diagnostic dose of desmopressin in patients with Cushing's disease largely overlaps the one observed in normal weight controls and obese subjects. Further, as in controls and obese subjects, the increase in VWF : Ag is counterbalanced by increased fibrinolysis, as indicated by the decrease in PAI-1 and increase in t-PA. In this context, it is worth recalling that desmopressin has been used since 1977 in the management of VWD without major adverse effects [6] and, further, has been employed also in normal individuals to prevent excessive blood loss during cardiac, orthopaedic and other major surgery without significant increases in the occurrence of thromboembolic events [13]. It thus appears, by and large, safe.

In conclusion, injection of a diagnostic dose of desmopressin induces transient changes in endothelial markers, i.e. VWF : Ag, t-PA and PAI-1. The magnitude of these changes in patients with Cushing's disease is comparable with the one recorded in obese and normal weight controls. Thus, desmopressin testing does not appear to induce a disease-specific thrombotic pattern in patients with Cushing's disease, although careful surveillance remains mandatory in these patients given their inherent prothrombotic profile.

Competing Interests

There are no competing interests to declare.