Variability of Late-Night Salivary Cortisol in Cushing Disease: A Prospective Study

Variability of Late-Night Salivary Cortisol in Cushing Disease: A Prospective Study Abstract Background The frequency of variable hormonogenesis in patients with Cushing disease (CD) but without cyclical symptoms is unclear. Aim To assess the frequency of variable hormonogenesis in patients presenting with CD. Methods Over a 6-month period, patients with confirmed or suspected CD provided late-night salivary samples for up to 42 consecutive nights. Results Of 19 patients confirmed to have CD, 16 provided at least 7 consecutive salivary samples, and 13 provided at least 21; these 16 patients are the subjects of this report. Twelve patients had at least three peak and two trough levels of late-night salivary cortisol (LNSC) but in only two patients were strict criteria for cyclical hormonogenesis fulfilled; variation was assessed as random in the others. Eight patients had de novo CD, and eight had recurrent/persistent disease. All patients with recurrent/persistent CD had two or more normal results, and in four of these patients, >50% of LNSC were normal. In six patients with de novo disease with at least one normal LNSC level, the maximum levels ranged from 1.55 to 15.5 times the upper limit of normal. Conclusions Extreme fluctuations of cortisol production, measured by sequential LNSC, are common in CD. In newly diagnosed disease, this may only occasionally impair diagnostic ability, whereas in most patients with recurrent/persistent disease after pituitary surgery, LNSC is frequently within the reference range, with potential to cause diagnostic problems. Cyclical Cushing syndrome (CS) is said to be an uncommon disorder, characterized by intermittent hypercortisolism accompanied by clinical symptoms and signs that typically regress during periods of eucortisolism (1–3). Documentation of three hypercortisolemic peaks and two troughs with normal cortisol levels is an accepted way of confirming the diagnosis (4), though another retrospective study accepted the demonstration of two clinical and/or biochemical hypercortisolemic peaks separated by a period of clinical and biochemical remission as defining cyclicity (5). In a review of cases reported in the literature up to 2007, Meinardi et al. (6) pointed out that many patients might have cyclical or variable hormonogensis with nonvariable clinical expression. Although the term “cyclical CS” has been used somewhat indiscriminately in the literature to include cases both with and without variable symptoms, variable hypercortisolism with nonvarying symptoms may be considered a different phenomenon from cyclical CS and could be of potentially greater significance in routine practice. The Endocrine Society guidelines suggest measurement of late-night salivary cortisol (LNSC) as the appropriate method of investigating for cyclical CS or variable hormonogenesis, as it is simple to perform repeatedly and, when samples are collected appropriately, accurate (7). We set out to investigate prospectively the phenomenon of variable hormonogenesis in consecutive patients referred to our center with suspected CS over a 6-month period. Patients and Methods From January to June 2014, all patients investigated for a suspected diagnosis of CS in the Endocrine Department of the Cleveland Clinic were invited to participate in the study. This included patients with de novo CS and those with possible recurrent/persistent disease after previous transsphenoidal surgery (TSS). The study was approved by the Cleveland Clinic Foundation Institutional Review Board. Patients unable to cooperate with providing salivary samples and those doing shift work that interfered with normal sleep patterns were excluded. Informed written consent was obtained from each patient before study procedures began. Investigation for CS was by conventional methods, including urinary free cortisol (UFC), dexamethasone suppression tests, serum and salivary cortisol, plasma adrenocorticotrophic hormone (ACTH), corticotrophin-releasing hormone (CRH) stimulation, pituitary magnetic resonance imaging (MRI), and inferior petrosal sinus sampling (IPSS) (Table 1). Table 1. Patient Characteristics and Diagnostic and Confirmatory Tests/Features Patient Number  Sex/Age (y)  Diagnostic Tests That Establish Diagnosis  Confirmed by  De novo         1  F/57  UFC, LNSC, ACTH, Dex/CRH, MRI  Path, postop AI   2  F/46  UFC, ACTH, MRI  Path, postop AI   3  F/62  UFC, LNSC, ACTH, CRH, MRI  Path, postop AI   4  F/48  UFC, ACTH, Dex, IPSS  Path, postop AI   5  F/33  UFC, LNSC, ACTH, Dex/CRH, IPSS  Path, postop AI   6  F/49  UFC, LNSC, ACTH, IPSS  Path, postop AI   7  F/36  UFC, LNSC, ACTH, Dex/CRH, MRI  Path, postop AI   8  M/50  UFC, LNSC, ACTH, Dex/CRH, IPSS,  Postop AI        Prior Confirmation and Subsequent Treatment  Recurrent/persistent         9  F/27  UFC, LNSC, ACTH  Prior Path, no postop AI, radiation, bilateral adrenalectomy   10  F/46  UFC, LNSC, ACTH, MRI  Prior Path, no postop AI, repeat TSS, Path, no postop AI, radiation, bilateral adrenalectomy   11  F/38  UFC, LNSC, ACTH  Prior Path and postop AI, repeat TSS, Path, postop AI   12  F/50  LNSC, Dex, ACTH  Prior postop AI, radiation, cabergoline   13  F/59  LNSC, ACTH  Prior Path, no postop AI, radiation   14  F/56  LNSC, ACTH, Dex  Prior Path, no postop AI, radiation   15  F/56  UFC, ACTH, Dex  Prior Path, postop AI, radiation   16  F/44  UFC, ACTH  Prior TSS × 2, Path on each postop AI after first, not second TSS, investigational medications in clinical trial  Patient Number  Sex/Age (y)  Diagnostic Tests That Establish Diagnosis  Confirmed by  De novo         1  F/57  UFC, LNSC, ACTH, Dex/CRH, MRI  Path, postop AI   2  F/46  UFC, ACTH, MRI  Path, postop AI   3  F/62  UFC, LNSC, ACTH, CRH, MRI  Path, postop AI   4  F/48  UFC, ACTH, Dex, IPSS  Path, postop AI   5  F/33  UFC, LNSC, ACTH, Dex/CRH, IPSS  Path, postop AI   6  F/49  UFC, LNSC, ACTH, IPSS  Path, postop AI   7  F/36  UFC, LNSC, ACTH, Dex/CRH, MRI  Path, postop AI   8  M/50  UFC, LNSC, ACTH, Dex/CRH, IPSS,  Postop AI        Prior Confirmation and Subsequent Treatment  Recurrent/persistent         9  F/27  UFC, LNSC, ACTH  Prior Path, no postop AI, radiation, bilateral adrenalectomy   10  F/46  UFC, LNSC, ACTH, MRI  Prior Path, no postop AI, repeat TSS, Path, no postop AI, radiation, bilateral adrenalectomy   11  F/38  UFC, LNSC, ACTH  Prior Path and postop AI, repeat TSS, Path, postop AI   12  F/50  LNSC, Dex, ACTH  Prior postop AI, radiation, cabergoline   13  F/59  LNSC, ACTH  Prior Path, no postop AI, radiation   14  F/56  LNSC, ACTH, Dex  Prior Path, no postop AI, radiation   15  F/56  UFC, ACTH, Dex  Prior Path, postop AI, radiation   16  F/44  UFC, ACTH  Prior TSS × 2, Path on each postop AI after first, not second TSS, investigational medications in clinical trial  Abbreviations: Dex, dexamethasone suppression test; Dex/CRH, combined dexamethasone CRH test; F, female; M, male; MRI, pituitary MRI; Path, demonstration of pituitary adenoma tissue with positive immunohistochemical staining for ACTH; postop AI, development of postoperative adrenal insufficiency. View Large LNSC Once the diagnosis of CS had been established, patients were asked to collect late-night salivary samples, between 11 pm and midnight or as late as possible before going to sleep if they normally retired before 11 pm, for up to 42 consecutive nights. They were given 10-mL Sarstedt salivettes (Sarstedt, Newton, NC) for sample collection. Food was not to be eaten in the 2 hours before sample collection, and they were to avoid brushing teeth, smoking, and drinking alcohol or coffee during that time. Ten minutes before starting the sample collection, they were to rinse the mouth with water to remove any food remnants. Patients were also advised to avoid use of any skin creams and to wash their hands prior to collection. Patients stored these samples in their fridge or freezer and mailed them to our department at weekly intervals. Samples were subsequently stored at −70°C until analysis in our laboratory. If a patient occasionally missed a day, the data were still included. Analysis for salivary cortisol was by liquid chromatography (LC) and tandem mass spectrometry (MS) in our laboratory. Each subject’s samples were analyzed in a single assay run. Saliva (200 μL) was mixed with internal standard (50 μL; d4-cortisol 5 ng/mL in 1:1 methanol/water) and vortex mixed, followed by addition of 1 mL methyl-tert butyl ether. The mixture stood at room temperature for an hour, vortex mixed for 30 seconds, and centrifuged at 2000g for 3 minutes. The organic layer (900 μL) was transferred into a glass tube and dried under a flow of nitrogen at 35°C. The residue was reconstituted with 1:1 methanol/water (200 μL) and injected (25 μL). All reagents used were of LC-MS grade. This method was performed on a Prelude LC system (Thermo Fisher Scientific, Waltham, MA) coupled to a TSQ Quantiva MS with water and methanol as mobile phase using an Accucore C18 analytical column (Thermo Fisher Scientific). Analytical measurement range was determined to be 28 to 1562 ng/dL (0.8 to 43.1 nmol/L) with accuracy between 95% and 118%. No substantial carryover was observed up to 4748 ng/dL (131 nmol/L). The total coefficient of variation was <14.8% at levels of 68.9 to 2160 ng/dL (1.9 to 59.6 nmol/L). Comparison (n = 33) with an independent LC-tandem MS method offered by a separate clinical laboratory gave a Deming regression correlation coefficient of 0.925, slope of 1.04, an intercept of −0.006, a standard error of estimate of 0.86, and a mean bias of 3.6%. The reference interval was established by measuring LNSC in 41 healthy subjects aged 19 to 71 years. Normal LNSC in this assay was <101 ng/dL (2.8 nmol/L), with a lower limit of quantitation of 29 ng/dL (0.8 nmol/L). Cyclic patterns were evaluated using a robust, nonparametric method described by Ahdesmaki et al. (8). A periodogram was fit to each patient’s cortisol measures, and the magnitude of the maximum spectral estimator was tested to see if the observed pattern exhibited more periodicity than would be expected from a random pattern based on a permutation test. Analyses were performed using the GeneCycle package within R software (version 3.1; SAS Institute, Cary, NC), and a P value <0.05 was considered statistically significant. Table 1 summarizes the criteria on which the diagnosis of Cushing disease (CD), de novo or recurrent/persistent, was based in each case. In seven of the eight de novo cases, a pituitary adenoma that stained positively for ACTH but negatively for other anterior pituitary hormones was demonstrated, and all of these patients developed postoperative adrenal insufficiency requiring hydrocortisone replacement for variable lengths of time. In seven of the eight recurrent/persistent cases, confirmation of pituitary adenoma tissue, staining positively for ACTH, was available in previous records. The one patient in whom this was lacking had had postoperative adrenal insufficiency. Four of these patients did not have postoperative adrenal insufficiency, thereby blurring the distinction between persistent or recurrent CD. One patient had two transsphenoidal operations prior to participating in this study; she had adrenal insufficiency after the first and remained in remission for 9 years before having a second surgery, after which she did not develop adrenal insufficiency. Results Over the 6-month period, 27 patients signed consent to participate. Patients were not routinely asked whether they experienced fluctuation in their symptoms. In only one case was mention made of the possibility of cyclical CS, and this patient was eventually excluded from the study due to use of steroid cream during the collection period (discussed later). No patient with CD diagnosed during this period refused to participate. CD was confirmed in 19 patients (17 women and 2 men; age ranging from 27 to 62 years) by conventional investigations independent of the tests in this study. The diagnoses in the other eight patients were adrenal carcinoma (one), adrenal adenoma (two), initially undisclosed use of steroid cream (one), and obesity with fatigue (four). Of the 19 with confirmed CD, 16 (15 women and 1 man) submitted at least 7 consecutive LNSC samples, and 13 patients submitted >21 samples. Eight of these patients had de novo CD and subsequently underwent TSS at Cleveland Clinic, and eight had recurrent/persistent CD having had previous TSS. The results from these 16 patients are presented. Individual LNSC levels during the study are shown in Figs. 1 (de novo) and 2 (recurrent/persistent). On simple visual inspection, 12 of the 16 patients exhibited at least three peaks and two troughs. LNSC level was in the normal range [<101 ng/dL, (2.8 nmol/L)] on at least one occasion in 14 patients. This was the case in all eight patients with recurrent/persistent CD and in six of the eight with de novo CD. Applying the criteria of Ahdesmaki et al. (8), only 2 of the 16 patients exhibited fluctuations that were deemed cyclical: patients 5 (de novo) and 13 (recurrent/persistent), with the interval between peaks appearing to be ∼4 days with a little variation. Figure 1. View largeDownload slide Sequential LNSC levels in individual patients with de novo CD. The patient numbers at the top left-hand corner of each line graph correspond to the patient numbers in Tables 1 and 2. The fractions at the top right-hand corner of each line graph represent the fraction of the total number of tests that were elevated. The red lines indicate upper limit of normal for LNSC. Note that the scales for days on the abscissae and salivary cortisol on the ordinates vary from patient to patient for the purpose of formatting. To convert ng/dL to nmol/L, divide by 36.24. Figure 1. View largeDownload slide Sequential LNSC levels in individual patients with de novo CD. The patient numbers at the top left-hand corner of each line graph correspond to the patient numbers in Tables 1 and 2. The fractions at the top right-hand corner of each line graph represent the fraction of the total number of tests that were elevated. The red lines indicate upper limit of normal for LNSC. Note that the scales for days on the abscissae and salivary cortisol on the ordinates vary from patient to patient for the purpose of formatting. To convert ng/dL to nmol/L, divide by 36.24. Figure 2. View largeDownload slide Sequential LNSC levels in individual patients with recurrent/persistent CD. The patient numbers at the top left-hand corner of each line graph correspond to the patient numbers in Tables 1 and 2. The fractions at the top right-hand corner of each line graph represent the fraction of the total number of tests that were elevated. The red lines indicate upper limit of normal for LNSC. Note that the scales for days on the abscissae and salivary cortisol on the ordinates are vary from patient to patient for the purpose of formatting. To convert ng/dL to nmol/L, divide by 36.24. Figure 2. View largeDownload slide Sequential LNSC levels in individual patients with recurrent/persistent CD. The patient numbers at the top left-hand corner of each line graph correspond to the patient numbers in Tables 1 and 2. The fractions at the top right-hand corner of each line graph represent the fraction of the total number of tests that were elevated. The red lines indicate upper limit of normal for LNSC. Note that the scales for days on the abscissae and salivary cortisol on the ordinates are vary from patient to patient for the purpose of formatting. To convert ng/dL to nmol/L, divide by 36.24. In five of the eight patients with recurrent/persistent CD, the lowest LNSC level was at or below the limit of detection of the assay (29 ng/dL). This was the case in only one patient with de novo CD. Table 2 shows the lowest and highest LNSC levels in each patient and the number of LNSC values that were in the normal range out of each patient’s total number. Also, in five patients with recurrent/persistent CD, approximately one in three LNSC measurements were in the normal range, and four patients had normal levels >50% of the time. Table 2. Summary of LNSC Collections in Patients With De Novo (Patients 1 Through 8) and Recurrent/Persistent (Patients 9 Through 16) CD Patient Number  Number of LNSC Samples  Three Peaks/Two Troughs, Yes/No  Number in Normal Range  Lowest/Highest ng/dL (nmol/L)  Cyclical, a Yes/No  1  12  No  0  134/653 (3.4/18.1)  No  2  36  Yes  0  110/857 (3.0/23.6)  No  3  11  No  2  66/155 (1.8/4.3)  No  4  42  Yes  3  83/493 (2.3/13.6)  No  5  31  Yes  2  86/784 (2.4/21.6)  Yes  6  43  No  25  <30/635 (<0.8/17.5)  No  7  42  Yes  1  88/1557 (2.4/43.0)  No  8  40  Yes  6  40/1403 (1.1/38.7)  No  9  40  Yes  35  <30/146 (<0.8/4.0)  No  10  39  Yes  2  81/331 (2.2/9.1)  No  11  8  No  4  46/160 (1.3/4.4)  No  12  24  Yes  7  <30/1276 (<0.8/338)  No  13  41  Yes  25  30/189 (0.8/5.2)  Yes  14  29  Yes  2  91/323 (2.5/8.9)  No  15  22  Yes  12  <30/397 (<0.8/10.9)  No  16  34  Yes  33  <30/110 (<0.8/3.0)  No  Patient Number  Number of LNSC Samples  Three Peaks/Two Troughs, Yes/No  Number in Normal Range  Lowest/Highest ng/dL (nmol/L)  Cyclical, a Yes/No  1  12  No  0  134/653 (3.4/18.1)  No  2  36  Yes  0  110/857 (3.0/23.6)  No  3  11  No  2  66/155 (1.8/4.3)  No  4  42  Yes  3  83/493 (2.3/13.6)  No  5  31  Yes  2  86/784 (2.4/21.6)  Yes  6  43  No  25  <30/635 (<0.8/17.5)  No  7  42  Yes  1  88/1557 (2.4/43.0)  No  8  40  Yes  6  40/1403 (1.1/38.7)  No  9  40  Yes  35  <30/146 (<0.8/4.0)  No  10  39  Yes  2  81/331 (2.2/9.1)  No  11  8  No  4  46/160 (1.3/4.4)  No  12  24  Yes  7  <30/1276 (<0.8/338)  No  13  41  Yes  25  30/189 (0.8/5.2)  Yes  14  29  Yes  2  91/323 (2.5/8.9)  No  15  22  Yes  12  <30/397 (<0.8/10.9)  No  16  34  Yes  33  <30/110 (<0.8/3.0)  No  a Cyclical as defined by strict mathematical criteria [see Mullan et al. (9)]. View Large Six of the eight patients with recurrent/persistent CD had normal LNSC levels on two consecutive nights on at least one occasion; one had an instance of three normal on consecutive nights; one had four in a row, two had six in a row, and one patient had 31 consecutive normal LNSC levels. Only one patient with de novo CD had normal levels on more than four consecutive nights. Although the study was not designed to compare the study values of LNSC with those of UFC used in diagnosis, Supplemental Table 1 shows mean 24-hour UFC and range and median values of LNSC in the 13 patients in whom UFC was part of the diagnostic evaluation. There was a poor correlation between the mean 24-hour UFC and median LNSC (Pearson correlation coefficient R = 0.419; P = 0.15) Discussion The potential utility of salivary cortisol measurement in both the diagnosis and long-term monitoring of patients with “classical” cyclical CS has been ably demonstrated in studies of such patients for periods of ≥2 years (10, 11). Fluctuations have also been reported using 24-hour UFC, serum cortisol, and ACTH levels (3, 4, 9, 12, 13). Petersenn et al. (14) analyzed baseline (pretreatment) UFC data in 152 patients with CD from an international, multicenter trial of pasireotide treatment. The overall intrapatient coefficient of variation of four 24-hour UFC measurements was 52%, with one patient exhibiting a range of 217.5 to 5081.5 μg (600 to 14,020 nmol). Our prospective study of patients with CD, both de novo and recurrent/persistent, demonstrates the frequency with which wide fluctuations of cortisol levels, as measured by LNSC, occurs in patients with no suspicion of cyclical CS. It has been suggested that the demonstration of three peaks and two troughs in sequential biochemical work up is sufficient for diagnosis of cyclical CS. Graham et al. (15) investigated sequential LNSC and early morning urine cortisol/creatinine ratio for up to 28 days in 10 patients, 7 of whom had de novo CD. Similar to our findings, their patients showed fluctuation of measurements on a day-to-day basis. They also demonstrated a clear correlation (r = 0.79) between LNSC and urine cortisol/creatinine ratio. A previous report using urine cortisol/creatinine ratio found apparent cyclical hormonogenesis in five of nine patients investigated prospectively (13). To our knowledge, no other study has included as large a patient population of patients with CD studied prospectively with LNSC measurement. At first glance, the fluctuations in many of our cases fit with a previous definition of cyclical hormonogenesis: three peaks and two troughs. However, in only 2 of the 16 cases were the strict statistical criteria that we used for cyclical hormonogenesis fulfilled. In any event, because none of these patients had any apparent fluctuation in symptoms or clinical features, we would hypothesize that fluctuation in cortisol, although marked or extreme in some cases, generally occurs over a short enough time so as not to translate into symptomatic change. Therefore, the issue of whether the fluctuation is random or cyclical may be viewed as being of academic and pathophysiologic interest rather than affecting the immediate clinical management of these patients. Although we did not specifically set out to divide our patients into those with de novo and those with recurrent/persistent CD, it is clear that important differences, with potential clinical impact, emerged between these two groups. With respect to the de novo patients, even though fluctuations were just as extreme, more so in some cases, as in recurrent/persistent CD, there is no suggestion that such variable levels caused any diagnostic difficulty. This was because in seven of the eight de novo cases, the vast majority of results were above the upper limit of normal. In the eighth patient (patient number 6), the clinical features and prior biochemical tests left no doubt about the diagnosis of CS even though 25 of 43 LNSC values were in the normal range during the study period. Whether the diagnosis would be so clear-cut in de novo patients who report apparent short-term cyclicity of symptoms we cannot say, as there were no such patients in our cohort. Recurrence of CD after successful pituitary TSS is now recognized as being more common than previously thought, having been reported as ∼30% or higher up to 5 years after surgery (16, 17). LNSC measurement has been proposed as a convenient means to screen for recurrence of CD (18, 19). Amlashi et al. (20), in a retrospective study that included LNSC measurement in 68 patients with CD in remission after TSS, of whom 19 then had recurrence, concluded that LNSC would be a more convenient and more accurate test than 24-hour UFC for detecting recurrence. They also advocated that prospective studies are necessary. Therefore, the results in our patients with recurrent/persistent CD are potentially problematic. It should be no surprise that overall, the levels in these patients are lower than in de novo patients, because one might anticipate that the previous pituitary surgery would lower overall ACTH and cortisol production. What our results reveal, however, is the frequency with which patients with recurrent/persistent CD have LNSC within the normal range. This observation has particular relevance with regard to recent suggestions that LNSC measurement may represent the best test for early diagnosis of recurrent CD (20–22). Our study did not address that particular question, but the fact that half of our patients with recurrent/persistent CD had normal LNSC levels in >50% of their tests and that in 75% of the patients, normal levels were seen on at least two consecutive nights, indicates a need for caution. There is currently no consensus on what is the appropriate number of LNSC estimations to be used in screening for recurrence of CD. Danet-Lamasou et al. (23) propose that either the highest level from four samples or the mean of three may be needed, whereas our data suggest that in some cases, an even greater number may be required. We were surprised by the results in patient number 16, who had normal LNSC levels in 33 of 34 samples. This would naturally raise the question of whether this patient did indeed have recurrent/persistent CD. We are certain she did, based on symptoms and elevated UFC and ACTH, plus the fact that subsequent to this study, she has met criteria for participation in two clinical trials of investigational drugs for CD. Kidambi et al. (24) have pointed out that in cases in which there is a high index of clinical suspicion, large numbers of LNSC measurements may be necessary to confirm the diagnosis. Another implication from our study is the need to exercise caution in the interpretation of results when using medical therapy for CD. This would apply both in routine clinical practice and, perhaps particularly, in clinical trials, in which, because pituitary surgery remains the treatment of choice for de novo patients, it is likely that many or most patients will have recurrent/persistent CD as opposed to de novo. Doing clinical research prospectively in patients with CD for a period of time before they have surgery is challenging, given that one does not wish to delay treatment. Strengths of our study are that it was prospective and that all patients diagnosed at our clinic as having CD during the 6-month study period agreed to participate. There are also some potential weaknesses. Saliva samples were collected by the patients unsupervised (though after careful instruction). It has been noted that compliance with timing of salivary cortisol sampling can be substantially less than is reported by both patients (with fibromyalgia) and healthy subjects, though patients were more compliant than healthy volunteers (25). However, collection of saliva at home, unsupervised, is now accepted as being accurate for the diagnosis of CS in routine clinical practice (26). The number of patients we studied is relatively small, but to our knowledge, it is the largest prospective study looking at consecutive LNSC measurements. Clearly, only a minority of the patients completed 6 weeks of LNSC measurement, and five patients had gaps of 1 or 2 days in collection, a reflection of what is achievable in real-world clinical investigation; therefore, it is possible we might be underestimating the frequency of truly cyclical hormonogenesis. Our results relate to measuring LNSC by LC-tandem MS—studies using other methods of LNSC such as radioimmunoassay may be needed because the performance of other methodologies may be different. There has been recent interest in measurement of salivary cortisone, in addition to salivary cortisol, as a reflection of serum free cortisol (27, 28). The rationale for this is that salivary glands express the enzyme 11β-hydroxysteroid dehydrogenase type 2, which converts cortisol to cortisone. One can only conjecture on the role that conversion of cortisol to cortisone might have played in producing the variability in salivary cortisol we have described. Measurement of cortisol/cortisone ratio can be used to detect preanalytical contamination of saliva by, for example, hydrocortisone cream because collected saliva, as opposed to the salivary gland, would not contain the enzyme (29). In conclusion, our study highlights the frequency and extent of fluctuation of LNSC that occurs in CD. We have shown, also, that patients with recurrent/persistent CD may have LNSC levels in the normal range with considerable frequency. Both of these points are of potential importance in routine clinical practice. Finally, we suggest that a distinction should be made between cyclical (or variable) hormonogenesis and cyclical CS. Cyclical CS was not present in any of our patients, and, indeed, true cyclical CS, with remission of clinical features and symptoms in addition to cyclical hormonogenesis, is a relative rarity. In contrast, extremely variable hormonogenesis seems to be virtually the norm and may therefore be a greater clinical challenge in everyday practice than has hitherto been acknowledged. Abbreviations: ACTH adrenocorticotrophic hormone CD Cushing disease CRH corticotrophin-releasing hormone CS Cushing syndrome IPSS inferior petrosal sinus sampling LC liquid chromatography LNSC late-night salivary cortisol MRI magnetic resonance imaging MS mass spectrometry TSS transsphenoidal surgery UFC urinary free cortisol. Acknowledgments Disclosure Summary: The authors have nothing to disclose. References 1. Bailey RE. Periodic hormonogenesis--a new phenomenon. Periodicity in function of a hormone-producing tumor in man. J Clin Endocrinol Metab . 1971; 32( 3): 317– 327. Google Scholar CrossRef Search ADS PubMed  2. Brooks RV, Jeffcoate SL, London DR, Prunty FT, Smith PM. Intermittent Cushing’s syndrome with anomalous response to dexamethasone. J Endocrinol . 1966; 36( 1): 53– 61. Google Scholar CrossRef Search ADS PubMed  3. Oates TW, McCourt JP, Friedman WA, Agee OF, Rhoton AL, Thomas WC, Jr. Cushing’s disease with cyclic hormonogenesis and diabetes insipidus. Neurosurgery . 1979; 5( 5): 598– 603. Google Scholar CrossRef Search ADS PubMed  4. Atkinson AB, Kennedy A, Wiggam MI, McCance DR, Sheridan B. Long-term remission rates after pituitary surgery for Cushing’s disease: the need for long-term surveillance. Clin Endocrinol (Oxf) . 2005; 63( 5): 549– 559. Google Scholar CrossRef Search ADS PubMed  5. Alexandraki KI, Kaltsas GA, Isidori AM, Akker SA, Drake WM, Chew SL, Monson JP, Besser GM, Grossman AB. The prevalence and characteristic features of cyclicity and variability in Cushing’s disease. Eur J Endocrinol . 2009; 160( 6): 1011– 1018. Google Scholar CrossRef Search ADS PubMed  6. Meinardi JR, Wolffenbuttel BHR, Dullaart RPF. Cyclic Cushing’s syndrome: a clinical challenge. Eur J Endocrinol . 2007; 157( 3): 245– 254. Google Scholar CrossRef Search ADS PubMed  7. Nieman LK, Biller BMK, Findling JW, Newell-Price J, Savage MO, Stewart PM, Montori VM. The diagnosis of Cushing’s syndrome: an Endocrine Society Clinical Practice Guideline. J Clin Endocrinol Metab . 2008; 93( 5): 1526– 1540. Google Scholar CrossRef Search ADS PubMed  8. Ahdesmäki M, Lähdesmäki H, Pearson R, Huttunen H, Yli-Harja O. Robust detection of periodic time series measured from biological systems. BMC Bioinformatics . 2005; 6( 1): 117. Google Scholar CrossRef Search ADS PubMed  9. Mullan KR, Atkinson AB, Sheridan B. Cyclical Cushing’s syndrome: an update. Curr Opin Endocrinol Diabetes Obes . 2007; 14( 4): 317– 322. Google Scholar CrossRef Search ADS PubMed  10. Hermus AR, Pieters GF, Borm GF, Verhofstad AA, Smals AG, Benraad TJ, Kloppenborg PW. Unpredictable hypersecretion of cortisol in Cushing’s disease: detection by daily salivary cortisol measurements. Acta Endocrinol (Copenh) . 1993; 128( 5): 428– 432. Google Scholar PubMed  11. Mosnier-Pudar H, Thomopoulos P, Bertagna X, Fournier C, Guiban D, Luton JP. Long-distance and long-term follow-up of a patient with intermittent Cushing’s disease by salivary cortisol measurements. Eur J Endocrinol . 1995; 133( 3): 313– 316. Google Scholar CrossRef Search ADS PubMed  12. Atkinson AB, Chestnutt A, Crothers E, Woods R, Weaver JA, Kennedy L, Sheridan B. Cyclical Cushing’s disease: two distinct rhythms in a patient with a basophil adenoma. J Clin Endocrinol Metab . 1985; 60( 2): 328– 332. Google Scholar CrossRef Search ADS PubMed  13. Atkinson AB, Kennedy AL, Carson DJ, Hadden DR, Weaver JA, Sheridan B. Five cases of cyclical Cushing’s syndrome. Br Med J (Clin Res Ed) . 1985; 291( 6507): 1453– 1457. Google Scholar CrossRef Search ADS PubMed  14. Petersenn S, Newell-Price J, Findling JW, Gu F, Maldonado M, Sen K, Salgado LR, Colao A, Biller BMK; Pasireotide B2305 Study Group. High variability in baseline urinary free cortisol values in patients with Cushing’s disease. Clin Endocrinol (Oxf) . 2014; 80( 2): 261– 269. Google Scholar CrossRef Search ADS PubMed  15. Graham UM, Hunter SJ, McDonnell M, Mullan KR, Atkinson AB. A comparison of the use of urinary cortisol to creatinine ratios and nocturnal salivary cortisol in the evaluation of cyclicity in patients with Cushing’s syndrome. J Clin Endocrinol Metab . 2013; 98( 1): E72– E76. Google Scholar CrossRef Search ADS PubMed  16. Patil CG, Prevedello DM, Lad SP, Vance ML, Thorner MO, Katznelson L, Laws ER, Jr. Late recurrences of Cushing’s disease after initial successful transsphenoidal surgery. J Clin Endocrinol Metab . 2008; 93( 2): 358– 362. Google Scholar CrossRef Search ADS PubMed  17. Dimopoulou C, Schopohl J, Rachinger W, Buchfelder M, Honegger J, Reincke M, Stalla GK. Long-term remission and recurrence rates after first and second transsphenoidal surgery for Cushing’s disease: care reality in the Munich Metropolitan Region. Eur J Endocrinol . 2013; 170( 2): 283– 292. Google Scholar CrossRef Search ADS PubMed  18. Carrasco CA, Coste J, Guignat L, Groussin L, Dugué MA, Gaillard S, Bertagna X, Bertherat J. Midnight salivary cortisol determination for assessing the outcome of transsphenoidal surgery in Cushing’s disease. J Clin Endocrinol Metab . 2008; 93( 12): 4728– 4734. Google Scholar CrossRef Search ADS PubMed  19. Raff H. Cushing’s syndrome: diagnosis and surveillance using salivary cortisol. Pituitary . 2012; 15( 1): 64– 70. Google Scholar CrossRef Search ADS PubMed  20. Amlashi FG, Swearingen B, Faje AT, Nachtigall LB, Miller KK, Klibanski A, Biller BMK, Tritos NA. Accuracy of late-night salivary cortisol in evaluating postoperative remission and recurrence in Cushing’s disease. J Clin Endocrinol Metab . 2015; 100( 10): 3770– 3777. Google Scholar CrossRef Search ADS PubMed  21. Carroll TB, Javorsky BR, Findling JW. Postsurgical recurrent Cushing disease: clinical benefit of early intervention in patients with normal urinary free cortisol. Endocr Pract . 2016; 22( 10): 1216– 1223. Google Scholar CrossRef Search ADS PubMed  22. Fleseriu M, Hamrahian AH, Hoffman AR, Kelly DF, Katznelson L; AACE Neuroendocrine and Pituitary Scientific Committee *. American Association of Clinical Endocrinologists and American College of Endocrinology disease state clinical review: Diagnosis of recurrence in Cushing disease. Endocr Pract . 2016; 22( 12): 1436– 1448. Google Scholar CrossRef Search ADS PubMed  23. Danet-Lamasou M, Asselineau J, Perez P, Vivot A, Nunes M-L, Loiseau H, San-Galli F, Cherifi-Gatta B, Corcuff J-B, Tabarin A. Accuracy of repeated measurements of late-night salivary cortisol to screen for early-stage recurrence of Cushing’s disease following pituitary surgery. Clin Endocrinol (Oxf) . 2015; 82( 2): 260– 266. Google Scholar CrossRef Search ADS PubMed  24. Kidambi S, Raff H, Findling JW. Limitations of nocturnal salivary cortisol and urine free cortisol in the diagnosis of mild Cushing’s syndrome. Eur J Endocrinol . 2007; 157( 6): 725– 731. Google Scholar CrossRef Search ADS PubMed  25. Broderick JE, Arnold D, Kudielka BM, Kirschbaum C. Salivary cortisol sampling compliance: comparison of patients and healthy volunteers. Psychoneuroendocrinology . 2004; 29( 5): 636– 650. Google Scholar CrossRef Search ADS PubMed  26. Raff H, Raff JL, Findling JW. Late-night salivary cortisol as a screening test for Cushing’s syndrome. J Clin Endocrinol Metab . 1998; 83( 8): 2681– 2686. Google Scholar PubMed  27. Perogamvros I, Keevil BG, Ray DW, Trainer PJ. Salivary cortisone is a potential biomarker for serum free cortisol. J Clin Endocrinol Metab . 2010; 95( 11): 4951– 4958. Google Scholar CrossRef Search ADS PubMed  28. Debono M, Harrison RF, Whitaker MJ, Eckland D, Arlt W, Keevil BG, Ross RJ. Salivary cortisone reflects cortisol exposure under physiological conditions and after hydrocortisone. J Clin Endocrinol Metab . 2016; 101( 4): 1469– 1477. Google Scholar CrossRef Search ADS PubMed  29. Raff H, Singh RJ. Measurement of late-night salivary cortisol and cortisone by LC-MS/MS to assess preanalytical sample contamination with topical hydrocortisone. Clin Chem . 2012; 58( 5): 947– 948. Google Scholar CrossRef Search ADS PubMed  Copyright © 2018 Endocrine Society http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Journal of Clinical Endocrinology and Metabolism Oxford University Press

Variability of Late-Night Salivary Cortisol in Cushing Disease: A Prospective Study

Loading next page...
 
/lp/ou_press/variability-of-late-night-salivary-cortisol-in-cushing-disease-a-q7EGcxJmJ9
Publisher
Oxford University Press
Copyright
Copyright © 2018 Endocrine Society
ISSN
0021-972X
eISSN
1945-7197
D.O.I.
10.1210/jc.2017-02020
Publisher site
See Article on Publisher Site

Abstract

Abstract Background The frequency of variable hormonogenesis in patients with Cushing disease (CD) but without cyclical symptoms is unclear. Aim To assess the frequency of variable hormonogenesis in patients presenting with CD. Methods Over a 6-month period, patients with confirmed or suspected CD provided late-night salivary samples for up to 42 consecutive nights. Results Of 19 patients confirmed to have CD, 16 provided at least 7 consecutive salivary samples, and 13 provided at least 21; these 16 patients are the subjects of this report. Twelve patients had at least three peak and two trough levels of late-night salivary cortisol (LNSC) but in only two patients were strict criteria for cyclical hormonogenesis fulfilled; variation was assessed as random in the others. Eight patients had de novo CD, and eight had recurrent/persistent disease. All patients with recurrent/persistent CD had two or more normal results, and in four of these patients, >50% of LNSC were normal. In six patients with de novo disease with at least one normal LNSC level, the maximum levels ranged from 1.55 to 15.5 times the upper limit of normal. Conclusions Extreme fluctuations of cortisol production, measured by sequential LNSC, are common in CD. In newly diagnosed disease, this may only occasionally impair diagnostic ability, whereas in most patients with recurrent/persistent disease after pituitary surgery, LNSC is frequently within the reference range, with potential to cause diagnostic problems. Cyclical Cushing syndrome (CS) is said to be an uncommon disorder, characterized by intermittent hypercortisolism accompanied by clinical symptoms and signs that typically regress during periods of eucortisolism (1–3). Documentation of three hypercortisolemic peaks and two troughs with normal cortisol levels is an accepted way of confirming the diagnosis (4), though another retrospective study accepted the demonstration of two clinical and/or biochemical hypercortisolemic peaks separated by a period of clinical and biochemical remission as defining cyclicity (5). In a review of cases reported in the literature up to 2007, Meinardi et al. (6) pointed out that many patients might have cyclical or variable hormonogensis with nonvariable clinical expression. Although the term “cyclical CS” has been used somewhat indiscriminately in the literature to include cases both with and without variable symptoms, variable hypercortisolism with nonvarying symptoms may be considered a different phenomenon from cyclical CS and could be of potentially greater significance in routine practice. The Endocrine Society guidelines suggest measurement of late-night salivary cortisol (LNSC) as the appropriate method of investigating for cyclical CS or variable hormonogenesis, as it is simple to perform repeatedly and, when samples are collected appropriately, accurate (7). We set out to investigate prospectively the phenomenon of variable hormonogenesis in consecutive patients referred to our center with suspected CS over a 6-month period. Patients and Methods From January to June 2014, all patients investigated for a suspected diagnosis of CS in the Endocrine Department of the Cleveland Clinic were invited to participate in the study. This included patients with de novo CS and those with possible recurrent/persistent disease after previous transsphenoidal surgery (TSS). The study was approved by the Cleveland Clinic Foundation Institutional Review Board. Patients unable to cooperate with providing salivary samples and those doing shift work that interfered with normal sleep patterns were excluded. Informed written consent was obtained from each patient before study procedures began. Investigation for CS was by conventional methods, including urinary free cortisol (UFC), dexamethasone suppression tests, serum and salivary cortisol, plasma adrenocorticotrophic hormone (ACTH), corticotrophin-releasing hormone (CRH) stimulation, pituitary magnetic resonance imaging (MRI), and inferior petrosal sinus sampling (IPSS) (Table 1). Table 1. Patient Characteristics and Diagnostic and Confirmatory Tests/Features Patient Number  Sex/Age (y)  Diagnostic Tests That Establish Diagnosis  Confirmed by  De novo         1  F/57  UFC, LNSC, ACTH, Dex/CRH, MRI  Path, postop AI   2  F/46  UFC, ACTH, MRI  Path, postop AI   3  F/62  UFC, LNSC, ACTH, CRH, MRI  Path, postop AI   4  F/48  UFC, ACTH, Dex, IPSS  Path, postop AI   5  F/33  UFC, LNSC, ACTH, Dex/CRH, IPSS  Path, postop AI   6  F/49  UFC, LNSC, ACTH, IPSS  Path, postop AI   7  F/36  UFC, LNSC, ACTH, Dex/CRH, MRI  Path, postop AI   8  M/50  UFC, LNSC, ACTH, Dex/CRH, IPSS,  Postop AI        Prior Confirmation and Subsequent Treatment  Recurrent/persistent         9  F/27  UFC, LNSC, ACTH  Prior Path, no postop AI, radiation, bilateral adrenalectomy   10  F/46  UFC, LNSC, ACTH, MRI  Prior Path, no postop AI, repeat TSS, Path, no postop AI, radiation, bilateral adrenalectomy   11  F/38  UFC, LNSC, ACTH  Prior Path and postop AI, repeat TSS, Path, postop AI   12  F/50  LNSC, Dex, ACTH  Prior postop AI, radiation, cabergoline   13  F/59  LNSC, ACTH  Prior Path, no postop AI, radiation   14  F/56  LNSC, ACTH, Dex  Prior Path, no postop AI, radiation   15  F/56  UFC, ACTH, Dex  Prior Path, postop AI, radiation   16  F/44  UFC, ACTH  Prior TSS × 2, Path on each postop AI after first, not second TSS, investigational medications in clinical trial  Patient Number  Sex/Age (y)  Diagnostic Tests That Establish Diagnosis  Confirmed by  De novo         1  F/57  UFC, LNSC, ACTH, Dex/CRH, MRI  Path, postop AI   2  F/46  UFC, ACTH, MRI  Path, postop AI   3  F/62  UFC, LNSC, ACTH, CRH, MRI  Path, postop AI   4  F/48  UFC, ACTH, Dex, IPSS  Path, postop AI   5  F/33  UFC, LNSC, ACTH, Dex/CRH, IPSS  Path, postop AI   6  F/49  UFC, LNSC, ACTH, IPSS  Path, postop AI   7  F/36  UFC, LNSC, ACTH, Dex/CRH, MRI  Path, postop AI   8  M/50  UFC, LNSC, ACTH, Dex/CRH, IPSS,  Postop AI        Prior Confirmation and Subsequent Treatment  Recurrent/persistent         9  F/27  UFC, LNSC, ACTH  Prior Path, no postop AI, radiation, bilateral adrenalectomy   10  F/46  UFC, LNSC, ACTH, MRI  Prior Path, no postop AI, repeat TSS, Path, no postop AI, radiation, bilateral adrenalectomy   11  F/38  UFC, LNSC, ACTH  Prior Path and postop AI, repeat TSS, Path, postop AI   12  F/50  LNSC, Dex, ACTH  Prior postop AI, radiation, cabergoline   13  F/59  LNSC, ACTH  Prior Path, no postop AI, radiation   14  F/56  LNSC, ACTH, Dex  Prior Path, no postop AI, radiation   15  F/56  UFC, ACTH, Dex  Prior Path, postop AI, radiation   16  F/44  UFC, ACTH  Prior TSS × 2, Path on each postop AI after first, not second TSS, investigational medications in clinical trial  Abbreviations: Dex, dexamethasone suppression test; Dex/CRH, combined dexamethasone CRH test; F, female; M, male; MRI, pituitary MRI; Path, demonstration of pituitary adenoma tissue with positive immunohistochemical staining for ACTH; postop AI, development of postoperative adrenal insufficiency. View Large LNSC Once the diagnosis of CS had been established, patients were asked to collect late-night salivary samples, between 11 pm and midnight or as late as possible before going to sleep if they normally retired before 11 pm, for up to 42 consecutive nights. They were given 10-mL Sarstedt salivettes (Sarstedt, Newton, NC) for sample collection. Food was not to be eaten in the 2 hours before sample collection, and they were to avoid brushing teeth, smoking, and drinking alcohol or coffee during that time. Ten minutes before starting the sample collection, they were to rinse the mouth with water to remove any food remnants. Patients were also advised to avoid use of any skin creams and to wash their hands prior to collection. Patients stored these samples in their fridge or freezer and mailed them to our department at weekly intervals. Samples were subsequently stored at −70°C until analysis in our laboratory. If a patient occasionally missed a day, the data were still included. Analysis for salivary cortisol was by liquid chromatography (LC) and tandem mass spectrometry (MS) in our laboratory. Each subject’s samples were analyzed in a single assay run. Saliva (200 μL) was mixed with internal standard (50 μL; d4-cortisol 5 ng/mL in 1:1 methanol/water) and vortex mixed, followed by addition of 1 mL methyl-tert butyl ether. The mixture stood at room temperature for an hour, vortex mixed for 30 seconds, and centrifuged at 2000g for 3 minutes. The organic layer (900 μL) was transferred into a glass tube and dried under a flow of nitrogen at 35°C. The residue was reconstituted with 1:1 methanol/water (200 μL) and injected (25 μL). All reagents used were of LC-MS grade. This method was performed on a Prelude LC system (Thermo Fisher Scientific, Waltham, MA) coupled to a TSQ Quantiva MS with water and methanol as mobile phase using an Accucore C18 analytical column (Thermo Fisher Scientific). Analytical measurement range was determined to be 28 to 1562 ng/dL (0.8 to 43.1 nmol/L) with accuracy between 95% and 118%. No substantial carryover was observed up to 4748 ng/dL (131 nmol/L). The total coefficient of variation was <14.8% at levels of 68.9 to 2160 ng/dL (1.9 to 59.6 nmol/L). Comparison (n = 33) with an independent LC-tandem MS method offered by a separate clinical laboratory gave a Deming regression correlation coefficient of 0.925, slope of 1.04, an intercept of −0.006, a standard error of estimate of 0.86, and a mean bias of 3.6%. The reference interval was established by measuring LNSC in 41 healthy subjects aged 19 to 71 years. Normal LNSC in this assay was <101 ng/dL (2.8 nmol/L), with a lower limit of quantitation of 29 ng/dL (0.8 nmol/L). Cyclic patterns were evaluated using a robust, nonparametric method described by Ahdesmaki et al. (8). A periodogram was fit to each patient’s cortisol measures, and the magnitude of the maximum spectral estimator was tested to see if the observed pattern exhibited more periodicity than would be expected from a random pattern based on a permutation test. Analyses were performed using the GeneCycle package within R software (version 3.1; SAS Institute, Cary, NC), and a P value <0.05 was considered statistically significant. Table 1 summarizes the criteria on which the diagnosis of Cushing disease (CD), de novo or recurrent/persistent, was based in each case. In seven of the eight de novo cases, a pituitary adenoma that stained positively for ACTH but negatively for other anterior pituitary hormones was demonstrated, and all of these patients developed postoperative adrenal insufficiency requiring hydrocortisone replacement for variable lengths of time. In seven of the eight recurrent/persistent cases, confirmation of pituitary adenoma tissue, staining positively for ACTH, was available in previous records. The one patient in whom this was lacking had had postoperative adrenal insufficiency. Four of these patients did not have postoperative adrenal insufficiency, thereby blurring the distinction between persistent or recurrent CD. One patient had two transsphenoidal operations prior to participating in this study; she had adrenal insufficiency after the first and remained in remission for 9 years before having a second surgery, after which she did not develop adrenal insufficiency. Results Over the 6-month period, 27 patients signed consent to participate. Patients were not routinely asked whether they experienced fluctuation in their symptoms. In only one case was mention made of the possibility of cyclical CS, and this patient was eventually excluded from the study due to use of steroid cream during the collection period (discussed later). No patient with CD diagnosed during this period refused to participate. CD was confirmed in 19 patients (17 women and 2 men; age ranging from 27 to 62 years) by conventional investigations independent of the tests in this study. The diagnoses in the other eight patients were adrenal carcinoma (one), adrenal adenoma (two), initially undisclosed use of steroid cream (one), and obesity with fatigue (four). Of the 19 with confirmed CD, 16 (15 women and 1 man) submitted at least 7 consecutive LNSC samples, and 13 patients submitted >21 samples. Eight of these patients had de novo CD and subsequently underwent TSS at Cleveland Clinic, and eight had recurrent/persistent CD having had previous TSS. The results from these 16 patients are presented. Individual LNSC levels during the study are shown in Figs. 1 (de novo) and 2 (recurrent/persistent). On simple visual inspection, 12 of the 16 patients exhibited at least three peaks and two troughs. LNSC level was in the normal range [<101 ng/dL, (2.8 nmol/L)] on at least one occasion in 14 patients. This was the case in all eight patients with recurrent/persistent CD and in six of the eight with de novo CD. Applying the criteria of Ahdesmaki et al. (8), only 2 of the 16 patients exhibited fluctuations that were deemed cyclical: patients 5 (de novo) and 13 (recurrent/persistent), with the interval between peaks appearing to be ∼4 days with a little variation. Figure 1. View largeDownload slide Sequential LNSC levels in individual patients with de novo CD. The patient numbers at the top left-hand corner of each line graph correspond to the patient numbers in Tables 1 and 2. The fractions at the top right-hand corner of each line graph represent the fraction of the total number of tests that were elevated. The red lines indicate upper limit of normal for LNSC. Note that the scales for days on the abscissae and salivary cortisol on the ordinates vary from patient to patient for the purpose of formatting. To convert ng/dL to nmol/L, divide by 36.24. Figure 1. View largeDownload slide Sequential LNSC levels in individual patients with de novo CD. The patient numbers at the top left-hand corner of each line graph correspond to the patient numbers in Tables 1 and 2. The fractions at the top right-hand corner of each line graph represent the fraction of the total number of tests that were elevated. The red lines indicate upper limit of normal for LNSC. Note that the scales for days on the abscissae and salivary cortisol on the ordinates vary from patient to patient for the purpose of formatting. To convert ng/dL to nmol/L, divide by 36.24. Figure 2. View largeDownload slide Sequential LNSC levels in individual patients with recurrent/persistent CD. The patient numbers at the top left-hand corner of each line graph correspond to the patient numbers in Tables 1 and 2. The fractions at the top right-hand corner of each line graph represent the fraction of the total number of tests that were elevated. The red lines indicate upper limit of normal for LNSC. Note that the scales for days on the abscissae and salivary cortisol on the ordinates are vary from patient to patient for the purpose of formatting. To convert ng/dL to nmol/L, divide by 36.24. Figure 2. View largeDownload slide Sequential LNSC levels in individual patients with recurrent/persistent CD. The patient numbers at the top left-hand corner of each line graph correspond to the patient numbers in Tables 1 and 2. The fractions at the top right-hand corner of each line graph represent the fraction of the total number of tests that were elevated. The red lines indicate upper limit of normal for LNSC. Note that the scales for days on the abscissae and salivary cortisol on the ordinates are vary from patient to patient for the purpose of formatting. To convert ng/dL to nmol/L, divide by 36.24. In five of the eight patients with recurrent/persistent CD, the lowest LNSC level was at or below the limit of detection of the assay (29 ng/dL). This was the case in only one patient with de novo CD. Table 2 shows the lowest and highest LNSC levels in each patient and the number of LNSC values that were in the normal range out of each patient’s total number. Also, in five patients with recurrent/persistent CD, approximately one in three LNSC measurements were in the normal range, and four patients had normal levels >50% of the time. Table 2. Summary of LNSC Collections in Patients With De Novo (Patients 1 Through 8) and Recurrent/Persistent (Patients 9 Through 16) CD Patient Number  Number of LNSC Samples  Three Peaks/Two Troughs, Yes/No  Number in Normal Range  Lowest/Highest ng/dL (nmol/L)  Cyclical, a Yes/No  1  12  No  0  134/653 (3.4/18.1)  No  2  36  Yes  0  110/857 (3.0/23.6)  No  3  11  No  2  66/155 (1.8/4.3)  No  4  42  Yes  3  83/493 (2.3/13.6)  No  5  31  Yes  2  86/784 (2.4/21.6)  Yes  6  43  No  25  <30/635 (<0.8/17.5)  No  7  42  Yes  1  88/1557 (2.4/43.0)  No  8  40  Yes  6  40/1403 (1.1/38.7)  No  9  40  Yes  35  <30/146 (<0.8/4.0)  No  10  39  Yes  2  81/331 (2.2/9.1)  No  11  8  No  4  46/160 (1.3/4.4)  No  12  24  Yes  7  <30/1276 (<0.8/338)  No  13  41  Yes  25  30/189 (0.8/5.2)  Yes  14  29  Yes  2  91/323 (2.5/8.9)  No  15  22  Yes  12  <30/397 (<0.8/10.9)  No  16  34  Yes  33  <30/110 (<0.8/3.0)  No  Patient Number  Number of LNSC Samples  Three Peaks/Two Troughs, Yes/No  Number in Normal Range  Lowest/Highest ng/dL (nmol/L)  Cyclical, a Yes/No  1  12  No  0  134/653 (3.4/18.1)  No  2  36  Yes  0  110/857 (3.0/23.6)  No  3  11  No  2  66/155 (1.8/4.3)  No  4  42  Yes  3  83/493 (2.3/13.6)  No  5  31  Yes  2  86/784 (2.4/21.6)  Yes  6  43  No  25  <30/635 (<0.8/17.5)  No  7  42  Yes  1  88/1557 (2.4/43.0)  No  8  40  Yes  6  40/1403 (1.1/38.7)  No  9  40  Yes  35  <30/146 (<0.8/4.0)  No  10  39  Yes  2  81/331 (2.2/9.1)  No  11  8  No  4  46/160 (1.3/4.4)  No  12  24  Yes  7  <30/1276 (<0.8/338)  No  13  41  Yes  25  30/189 (0.8/5.2)  Yes  14  29  Yes  2  91/323 (2.5/8.9)  No  15  22  Yes  12  <30/397 (<0.8/10.9)  No  16  34  Yes  33  <30/110 (<0.8/3.0)  No  a Cyclical as defined by strict mathematical criteria [see Mullan et al. (9)]. View Large Six of the eight patients with recurrent/persistent CD had normal LNSC levels on two consecutive nights on at least one occasion; one had an instance of three normal on consecutive nights; one had four in a row, two had six in a row, and one patient had 31 consecutive normal LNSC levels. Only one patient with de novo CD had normal levels on more than four consecutive nights. Although the study was not designed to compare the study values of LNSC with those of UFC used in diagnosis, Supplemental Table 1 shows mean 24-hour UFC and range and median values of LNSC in the 13 patients in whom UFC was part of the diagnostic evaluation. There was a poor correlation between the mean 24-hour UFC and median LNSC (Pearson correlation coefficient R = 0.419; P = 0.15) Discussion The potential utility of salivary cortisol measurement in both the diagnosis and long-term monitoring of patients with “classical” cyclical CS has been ably demonstrated in studies of such patients for periods of ≥2 years (10, 11). Fluctuations have also been reported using 24-hour UFC, serum cortisol, and ACTH levels (3, 4, 9, 12, 13). Petersenn et al. (14) analyzed baseline (pretreatment) UFC data in 152 patients with CD from an international, multicenter trial of pasireotide treatment. The overall intrapatient coefficient of variation of four 24-hour UFC measurements was 52%, with one patient exhibiting a range of 217.5 to 5081.5 μg (600 to 14,020 nmol). Our prospective study of patients with CD, both de novo and recurrent/persistent, demonstrates the frequency with which wide fluctuations of cortisol levels, as measured by LNSC, occurs in patients with no suspicion of cyclical CS. It has been suggested that the demonstration of three peaks and two troughs in sequential biochemical work up is sufficient for diagnosis of cyclical CS. Graham et al. (15) investigated sequential LNSC and early morning urine cortisol/creatinine ratio for up to 28 days in 10 patients, 7 of whom had de novo CD. Similar to our findings, their patients showed fluctuation of measurements on a day-to-day basis. They also demonstrated a clear correlation (r = 0.79) between LNSC and urine cortisol/creatinine ratio. A previous report using urine cortisol/creatinine ratio found apparent cyclical hormonogenesis in five of nine patients investigated prospectively (13). To our knowledge, no other study has included as large a patient population of patients with CD studied prospectively with LNSC measurement. At first glance, the fluctuations in many of our cases fit with a previous definition of cyclical hormonogenesis: three peaks and two troughs. However, in only 2 of the 16 cases were the strict statistical criteria that we used for cyclical hormonogenesis fulfilled. In any event, because none of these patients had any apparent fluctuation in symptoms or clinical features, we would hypothesize that fluctuation in cortisol, although marked or extreme in some cases, generally occurs over a short enough time so as not to translate into symptomatic change. Therefore, the issue of whether the fluctuation is random or cyclical may be viewed as being of academic and pathophysiologic interest rather than affecting the immediate clinical management of these patients. Although we did not specifically set out to divide our patients into those with de novo and those with recurrent/persistent CD, it is clear that important differences, with potential clinical impact, emerged between these two groups. With respect to the de novo patients, even though fluctuations were just as extreme, more so in some cases, as in recurrent/persistent CD, there is no suggestion that such variable levels caused any diagnostic difficulty. This was because in seven of the eight de novo cases, the vast majority of results were above the upper limit of normal. In the eighth patient (patient number 6), the clinical features and prior biochemical tests left no doubt about the diagnosis of CS even though 25 of 43 LNSC values were in the normal range during the study period. Whether the diagnosis would be so clear-cut in de novo patients who report apparent short-term cyclicity of symptoms we cannot say, as there were no such patients in our cohort. Recurrence of CD after successful pituitary TSS is now recognized as being more common than previously thought, having been reported as ∼30% or higher up to 5 years after surgery (16, 17). LNSC measurement has been proposed as a convenient means to screen for recurrence of CD (18, 19). Amlashi et al. (20), in a retrospective study that included LNSC measurement in 68 patients with CD in remission after TSS, of whom 19 then had recurrence, concluded that LNSC would be a more convenient and more accurate test than 24-hour UFC for detecting recurrence. They also advocated that prospective studies are necessary. Therefore, the results in our patients with recurrent/persistent CD are potentially problematic. It should be no surprise that overall, the levels in these patients are lower than in de novo patients, because one might anticipate that the previous pituitary surgery would lower overall ACTH and cortisol production. What our results reveal, however, is the frequency with which patients with recurrent/persistent CD have LNSC within the normal range. This observation has particular relevance with regard to recent suggestions that LNSC measurement may represent the best test for early diagnosis of recurrent CD (20–22). Our study did not address that particular question, but the fact that half of our patients with recurrent/persistent CD had normal LNSC levels in >50% of their tests and that in 75% of the patients, normal levels were seen on at least two consecutive nights, indicates a need for caution. There is currently no consensus on what is the appropriate number of LNSC estimations to be used in screening for recurrence of CD. Danet-Lamasou et al. (23) propose that either the highest level from four samples or the mean of three may be needed, whereas our data suggest that in some cases, an even greater number may be required. We were surprised by the results in patient number 16, who had normal LNSC levels in 33 of 34 samples. This would naturally raise the question of whether this patient did indeed have recurrent/persistent CD. We are certain she did, based on symptoms and elevated UFC and ACTH, plus the fact that subsequent to this study, she has met criteria for participation in two clinical trials of investigational drugs for CD. Kidambi et al. (24) have pointed out that in cases in which there is a high index of clinical suspicion, large numbers of LNSC measurements may be necessary to confirm the diagnosis. Another implication from our study is the need to exercise caution in the interpretation of results when using medical therapy for CD. This would apply both in routine clinical practice and, perhaps particularly, in clinical trials, in which, because pituitary surgery remains the treatment of choice for de novo patients, it is likely that many or most patients will have recurrent/persistent CD as opposed to de novo. Doing clinical research prospectively in patients with CD for a period of time before they have surgery is challenging, given that one does not wish to delay treatment. Strengths of our study are that it was prospective and that all patients diagnosed at our clinic as having CD during the 6-month study period agreed to participate. There are also some potential weaknesses. Saliva samples were collected by the patients unsupervised (though after careful instruction). It has been noted that compliance with timing of salivary cortisol sampling can be substantially less than is reported by both patients (with fibromyalgia) and healthy subjects, though patients were more compliant than healthy volunteers (25). However, collection of saliva at home, unsupervised, is now accepted as being accurate for the diagnosis of CS in routine clinical practice (26). The number of patients we studied is relatively small, but to our knowledge, it is the largest prospective study looking at consecutive LNSC measurements. Clearly, only a minority of the patients completed 6 weeks of LNSC measurement, and five patients had gaps of 1 or 2 days in collection, a reflection of what is achievable in real-world clinical investigation; therefore, it is possible we might be underestimating the frequency of truly cyclical hormonogenesis. Our results relate to measuring LNSC by LC-tandem MS—studies using other methods of LNSC such as radioimmunoassay may be needed because the performance of other methodologies may be different. There has been recent interest in measurement of salivary cortisone, in addition to salivary cortisol, as a reflection of serum free cortisol (27, 28). The rationale for this is that salivary glands express the enzyme 11β-hydroxysteroid dehydrogenase type 2, which converts cortisol to cortisone. One can only conjecture on the role that conversion of cortisol to cortisone might have played in producing the variability in salivary cortisol we have described. Measurement of cortisol/cortisone ratio can be used to detect preanalytical contamination of saliva by, for example, hydrocortisone cream because collected saliva, as opposed to the salivary gland, would not contain the enzyme (29). In conclusion, our study highlights the frequency and extent of fluctuation of LNSC that occurs in CD. We have shown, also, that patients with recurrent/persistent CD may have LNSC levels in the normal range with considerable frequency. Both of these points are of potential importance in routine clinical practice. Finally, we suggest that a distinction should be made between cyclical (or variable) hormonogenesis and cyclical CS. Cyclical CS was not present in any of our patients, and, indeed, true cyclical CS, with remission of clinical features and symptoms in addition to cyclical hormonogenesis, is a relative rarity. In contrast, extremely variable hormonogenesis seems to be virtually the norm and may therefore be a greater clinical challenge in everyday practice than has hitherto been acknowledged. Abbreviations: ACTH adrenocorticotrophic hormone CD Cushing disease CRH corticotrophin-releasing hormone CS Cushing syndrome IPSS inferior petrosal sinus sampling LC liquid chromatography LNSC late-night salivary cortisol MRI magnetic resonance imaging MS mass spectrometry TSS transsphenoidal surgery UFC urinary free cortisol. Acknowledgments Disclosure Summary: The authors have nothing to disclose. References 1. Bailey RE. Periodic hormonogenesis--a new phenomenon. Periodicity in function of a hormone-producing tumor in man. J Clin Endocrinol Metab . 1971; 32( 3): 317– 327. Google Scholar CrossRef Search ADS PubMed  2. Brooks RV, Jeffcoate SL, London DR, Prunty FT, Smith PM. Intermittent Cushing’s syndrome with anomalous response to dexamethasone. J Endocrinol . 1966; 36( 1): 53– 61. Google Scholar CrossRef Search ADS PubMed  3. Oates TW, McCourt JP, Friedman WA, Agee OF, Rhoton AL, Thomas WC, Jr. Cushing’s disease with cyclic hormonogenesis and diabetes insipidus. Neurosurgery . 1979; 5( 5): 598– 603. Google Scholar CrossRef Search ADS PubMed  4. Atkinson AB, Kennedy A, Wiggam MI, McCance DR, Sheridan B. Long-term remission rates after pituitary surgery for Cushing’s disease: the need for long-term surveillance. Clin Endocrinol (Oxf) . 2005; 63( 5): 549– 559. Google Scholar CrossRef Search ADS PubMed  5. Alexandraki KI, Kaltsas GA, Isidori AM, Akker SA, Drake WM, Chew SL, Monson JP, Besser GM, Grossman AB. The prevalence and characteristic features of cyclicity and variability in Cushing’s disease. Eur J Endocrinol . 2009; 160( 6): 1011– 1018. Google Scholar CrossRef Search ADS PubMed  6. Meinardi JR, Wolffenbuttel BHR, Dullaart RPF. Cyclic Cushing’s syndrome: a clinical challenge. Eur J Endocrinol . 2007; 157( 3): 245– 254. Google Scholar CrossRef Search ADS PubMed  7. Nieman LK, Biller BMK, Findling JW, Newell-Price J, Savage MO, Stewart PM, Montori VM. The diagnosis of Cushing’s syndrome: an Endocrine Society Clinical Practice Guideline. J Clin Endocrinol Metab . 2008; 93( 5): 1526– 1540. Google Scholar CrossRef Search ADS PubMed  8. Ahdesmäki M, Lähdesmäki H, Pearson R, Huttunen H, Yli-Harja O. Robust detection of periodic time series measured from biological systems. BMC Bioinformatics . 2005; 6( 1): 117. Google Scholar CrossRef Search ADS PubMed  9. Mullan KR, Atkinson AB, Sheridan B. Cyclical Cushing’s syndrome: an update. Curr Opin Endocrinol Diabetes Obes . 2007; 14( 4): 317– 322. Google Scholar CrossRef Search ADS PubMed  10. Hermus AR, Pieters GF, Borm GF, Verhofstad AA, Smals AG, Benraad TJ, Kloppenborg PW. Unpredictable hypersecretion of cortisol in Cushing’s disease: detection by daily salivary cortisol measurements. Acta Endocrinol (Copenh) . 1993; 128( 5): 428– 432. Google Scholar PubMed  11. Mosnier-Pudar H, Thomopoulos P, Bertagna X, Fournier C, Guiban D, Luton JP. Long-distance and long-term follow-up of a patient with intermittent Cushing’s disease by salivary cortisol measurements. Eur J Endocrinol . 1995; 133( 3): 313– 316. Google Scholar CrossRef Search ADS PubMed  12. Atkinson AB, Chestnutt A, Crothers E, Woods R, Weaver JA, Kennedy L, Sheridan B. Cyclical Cushing’s disease: two distinct rhythms in a patient with a basophil adenoma. J Clin Endocrinol Metab . 1985; 60( 2): 328– 332. Google Scholar CrossRef Search ADS PubMed  13. Atkinson AB, Kennedy AL, Carson DJ, Hadden DR, Weaver JA, Sheridan B. Five cases of cyclical Cushing’s syndrome. Br Med J (Clin Res Ed) . 1985; 291( 6507): 1453– 1457. Google Scholar CrossRef Search ADS PubMed  14. Petersenn S, Newell-Price J, Findling JW, Gu F, Maldonado M, Sen K, Salgado LR, Colao A, Biller BMK; Pasireotide B2305 Study Group. High variability in baseline urinary free cortisol values in patients with Cushing’s disease. Clin Endocrinol (Oxf) . 2014; 80( 2): 261– 269. Google Scholar CrossRef Search ADS PubMed  15. Graham UM, Hunter SJ, McDonnell M, Mullan KR, Atkinson AB. A comparison of the use of urinary cortisol to creatinine ratios and nocturnal salivary cortisol in the evaluation of cyclicity in patients with Cushing’s syndrome. J Clin Endocrinol Metab . 2013; 98( 1): E72– E76. Google Scholar CrossRef Search ADS PubMed  16. Patil CG, Prevedello DM, Lad SP, Vance ML, Thorner MO, Katznelson L, Laws ER, Jr. Late recurrences of Cushing’s disease after initial successful transsphenoidal surgery. J Clin Endocrinol Metab . 2008; 93( 2): 358– 362. Google Scholar CrossRef Search ADS PubMed  17. Dimopoulou C, Schopohl J, Rachinger W, Buchfelder M, Honegger J, Reincke M, Stalla GK. Long-term remission and recurrence rates after first and second transsphenoidal surgery for Cushing’s disease: care reality in the Munich Metropolitan Region. Eur J Endocrinol . 2013; 170( 2): 283– 292. Google Scholar CrossRef Search ADS PubMed  18. Carrasco CA, Coste J, Guignat L, Groussin L, Dugué MA, Gaillard S, Bertagna X, Bertherat J. Midnight salivary cortisol determination for assessing the outcome of transsphenoidal surgery in Cushing’s disease. J Clin Endocrinol Metab . 2008; 93( 12): 4728– 4734. Google Scholar CrossRef Search ADS PubMed  19. Raff H. Cushing’s syndrome: diagnosis and surveillance using salivary cortisol. Pituitary . 2012; 15( 1): 64– 70. Google Scholar CrossRef Search ADS PubMed  20. Amlashi FG, Swearingen B, Faje AT, Nachtigall LB, Miller KK, Klibanski A, Biller BMK, Tritos NA. Accuracy of late-night salivary cortisol in evaluating postoperative remission and recurrence in Cushing’s disease. J Clin Endocrinol Metab . 2015; 100( 10): 3770– 3777. Google Scholar CrossRef Search ADS PubMed  21. Carroll TB, Javorsky BR, Findling JW. Postsurgical recurrent Cushing disease: clinical benefit of early intervention in patients with normal urinary free cortisol. Endocr Pract . 2016; 22( 10): 1216– 1223. Google Scholar CrossRef Search ADS PubMed  22. Fleseriu M, Hamrahian AH, Hoffman AR, Kelly DF, Katznelson L; AACE Neuroendocrine and Pituitary Scientific Committee *. American Association of Clinical Endocrinologists and American College of Endocrinology disease state clinical review: Diagnosis of recurrence in Cushing disease. Endocr Pract . 2016; 22( 12): 1436– 1448. Google Scholar CrossRef Search ADS PubMed  23. Danet-Lamasou M, Asselineau J, Perez P, Vivot A, Nunes M-L, Loiseau H, San-Galli F, Cherifi-Gatta B, Corcuff J-B, Tabarin A. Accuracy of repeated measurements of late-night salivary cortisol to screen for early-stage recurrence of Cushing’s disease following pituitary surgery. Clin Endocrinol (Oxf) . 2015; 82( 2): 260– 266. Google Scholar CrossRef Search ADS PubMed  24. Kidambi S, Raff H, Findling JW. Limitations of nocturnal salivary cortisol and urine free cortisol in the diagnosis of mild Cushing’s syndrome. Eur J Endocrinol . 2007; 157( 6): 725– 731. Google Scholar CrossRef Search ADS PubMed  25. Broderick JE, Arnold D, Kudielka BM, Kirschbaum C. Salivary cortisol sampling compliance: comparison of patients and healthy volunteers. Psychoneuroendocrinology . 2004; 29( 5): 636– 650. Google Scholar CrossRef Search ADS PubMed  26. Raff H, Raff JL, Findling JW. Late-night salivary cortisol as a screening test for Cushing’s syndrome. J Clin Endocrinol Metab . 1998; 83( 8): 2681– 2686. Google Scholar PubMed  27. Perogamvros I, Keevil BG, Ray DW, Trainer PJ. Salivary cortisone is a potential biomarker for serum free cortisol. J Clin Endocrinol Metab . 2010; 95( 11): 4951– 4958. Google Scholar CrossRef Search ADS PubMed  28. Debono M, Harrison RF, Whitaker MJ, Eckland D, Arlt W, Keevil BG, Ross RJ. Salivary cortisone reflects cortisol exposure under physiological conditions and after hydrocortisone. J Clin Endocrinol Metab . 2016; 101( 4): 1469– 1477. Google Scholar CrossRef Search ADS PubMed  29. Raff H, Singh RJ. Measurement of late-night salivary cortisol and cortisone by LC-MS/MS to assess preanalytical sample contamination with topical hydrocortisone. Clin Chem . 2012; 58( 5): 947– 948. Google Scholar CrossRef Search ADS PubMed  Copyright © 2018 Endocrine Society

Journal

Journal of Clinical Endocrinology and MetabolismOxford University Press

Published: Mar 1, 2018

There are no references for this article.

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Search

Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly

Organize

Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.

Access

Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve

Freelancer

DeepDyve

Pro

Price

FREE

$49/month
$360/year

Save searches from
Google Scholar,
PubMed

Create lists to
organize your research

Export lists, citations

Read DeepDyve articles

Abstract access only

Unlimited access to over
18 million full-text articles

Print

20 pages / month

PDF Discount

20% off