Subjects and Methods TOSHIHIRO IMAKI*, **, MITSUHIDE NARUSE* AND KAZUE TAKANO*

Transcription

1 Endocrine Journal 2004, 51 (1), Adrenocortical Hyperplasia Associated with ACTH-dependent Cushing s Syndrome: Comparison of the Size of Adrenal Glands with Clinical and Endocrinological Data TOSHIHIRO IMAKI*, **, MITSUHIDE NARUSE* AND KAZUE TAKANO* *Department of Medicine, Tokyo Women s Medical University, Tokyo , Japan **Department of Bioregulation, Institute of Development and Aging Sciences, Nippon Medical School, Kawasaki , Japan Abstract. Diffuse or nodular hyperplasia of adrenal glands is associated frequently with ACTH-dependent Cushing s syndrome. We carried out a retrospective analysis of 28 patients with ACTH-dependent Cushing s syndrome admitted to our institution between 1984 and 1999 in order to clarify the incidence of adrenal hyperplasia in ACTH-dependent Cushing s syndrome and also to determine the correlation between adrenal gland images and clinical, biochemical and endocrinological data. Of the 28 patients, 16 (57%) showed diffuse adrenal hyperplasia while only 3 had focal adrenal nodules in the hypertrophied adrenals. There was a positive, significant correlation between the width of the adrenal glands measured on CT and circulating plasma ACTH, cortisol levels and urinary free cortisol (UFC) levels. Duration of the disease also correlated positively with adrenal width. No correlation was found between age and adrenal size and there was no difference in the prevalence of diffuse hyperplasia between normotensive and hypertensive patients. These results suggest that chronic ACTH hypersecretion may lead to diffuse adrenal hyperplasia in patients with ACTH-dependent Cushing s syndrome. Key words: Cushing s syndrome, Adrenocortical hyperplasia, CT scan, ACTH, Cortisol (Endocrine Journal 51: 89 95, 2004) Received: September 12, 2003 Accepted: October 30, 2003 Correspondence to: Dr. Toshihiro IMAKI, Department of Bioregulation, Institute of Development and Aging Sciences, Nippon Medical School, Kosugi-cho, Nakahara-ku, Kawasaki , Japan THE term Cushing s syndrome is most readily classified on the basis of whether excess glucocorticoid production is ACTH-dependent or -independent. Main causes of ACTH-dependent Cushing s syndrome are pituitary corticotroph adenoma (Cushing s disease) and ectopic ACTH production (ectopic ACTH syndrome) [1]. It is well established that chronic hypersecretion of ACTH causes diffuse bilateral adrenal hyperplasia [1]. The appearance of the adrenal glands in classic Cushing s disease as seen by computed tomography (CT) ranges from normal to diffuse symmetrical enlargement [2 4]. However, there have been only a few studies that have investigated the relationship between the size of adrenal glands in ACTHdependent Cushing s syndrome and changes in clinical, biochemical and endocrinological factors [4]. In this study, we measured the size of adrenal glands in patients with ACTH-dependent Cushing s syndrome using CT scanning and compared these measurements with clinical data, such as age, estimated duration of the disease and blood pressure and endocrinological data including the levels of plasma ACTH and cortisol and urinary free cortisol (UFC). Subjects and Methods We studied 28 patients (7 males, 21 females; mean age 43.5 ± 14.0) with proven ACTH-dependent Cushing s syndrome (Table 1). Patients nos had Cushing s disease while patients nos had ectopic ACTH syndrome. The majority of the patients were admitted to the Department of Medicine, Tokyo

2 90 IMAKI et al. Women s Medical University. The diagnosis of Cushing s syndrome was based upon standard criteria that included clinical features, elevated plasma and urinary cortisol levels with normal or high plasma ACTH levels, and a loss of circadian rhythm of plasma cortisol that was not suppressed by low dose dexamethasone. Other diagnostic tests included the high dose dexamethasone suppression test and corticotropin-releasing hormone (CRH) test. In the majority of patients the presence of pituitary adenoma was identified preoperatively by MRI. The diagnosis of the ectopic ACTH syndrome was established histologically following surgical resection of the tumors, 3 bronchial carcinoids (case nos ) and 1 malignant thymoma (case no. 25). The estimated duration of disease was calculated from patient records. Adrenal scanning was performed at 4 or 5 mm intervals from the level of the xiphisternum to the midportion of the kidneys in order to completely visualize both adrenal glands. The CT images of the adrenal glands were obtained prior to any surgical, radiological or medical treatment. Measurement of maximum adrenal body and limb widths were made on the images that best showed each component of the gland, according to the method described by Vincent et al. [5]. The width of six components that included the body, medial and lateral limbs of the right and left sided glands were determined in each patient. Measurements that were derived from 5 mm-collimation scans were reported to be larger than those obtained from 10 mm-collimation scans (5.49 vs mm) [4]. Thus, the upper limit of the adrenals determined on 10 mm-collimation scans described in Sohaib s paper [4] were multiplied by 1.16 and the following upper limits were used: right body, 12.2 mm; right lateral limb, 4.6 mm; right medial limb, 5.2 mm; left body, 14.5 mm; left lateral limb, 5.75 mm; left medial limb, 5.75 mm. Mean plasma ACTH and cortisol levels were determined in at least 3 samples from each subject collected at 8 to 9 am in the morning. Plasma cortisol and urinary free cortisol (UFC) levels were measured by RIA (Amerlex RIA, Ortho Clinical Diagnostics Co., Tokyo) that had a detection limit of 1 g/dl. Plasma ACTH values were measured initially by RIA as described previously [6] with the limit of sensitivity of this assay being 6 pg/ml. After 1990, plasma ACTH was determined by a radioimmunometric assay (ACTH IRMA Mitsubishi, Tokyo Mitsubishi Chemical Co., Tokyo, or Allegro ACTH Kit, Nichols Institute Diagnostics, Los Angeles, CA) that had a limit of sensitivity of 5 pg/ml. Although correlation between RIA and IRMA was excellent (r = 0.953), plasma ACTH concentration was slightly higher when measured by RIA than by IRMA [7]. Therefore, plasma ACTH was converted to the value calculated by the formula: corrected ACTH concentration = (RIA ACTH value 14.0) ACTH concentration was virtually identical between the two IRMA kits (y = 1.00x , r = 0.979) [8]. All values are expressed as mean ± SE. Linear regression analysis was used to determine the correlation between the width of the adrenal glands and the clinical and endocrinological parameters. Differences between the groups were analysed by Student s t-test for normally distributed data and the Mann-Whitney U test for data with a skewed distribution. The level of significance was set at a p value <0.05. Results The clinical and laboratory data of the patients with ACTH-dependent Cushing s syndrome are summarized in Table 1. Measurements of adrenal glands could not be obtained in 32 (19.0%) of 168 components because of either the shape of the gland, distortion by nodules or other technical factors. Using the definition of enlargement reported by Vincent et al. [5] (i.e. at least 2 of 6 components larger than 95% of the normal width), 16 (57%) of the 28 patients with ACTH-dependent Cushing s syndrome had enlarged adrenal glands on CT. Three cases (nos ) had at least one small nodule with one of these cases (no. 24) having a macronodule >10 mm in diameter. None of these nodules were associated with atrophy of the remaining adrenal tissue. The correlation between the levels of plasma ACTH, cortisol and UFC and the six components of adrenal size is shown in Table 2. Plasma ACTH was significantly correlated with the mean width of the adrenal glands except for the left adrenal limb. Similarly, a positive, significant correlation was observed between the levels of plasma cortisol and UFC and the mean width of all six components of the adrenal gland (Table 2). Fig. 1 contains scattergram showing the relationship between right adrenal body width and plasma ACTH

3 ADRENAL HYPERPLASIA IN CUSHING SYNDROME 91 Table 1. Clinical data of patients with ACTH-dependent Cushing s syndrome No. Age/sex Plasma ACTH (pg/ml) Plasma cortisol ( g/dl) Urinary free cortisol ( g/day) Hypertension Duration of the disease (year) Adrenal hyperplasia 1 32F M F F F F M M F F F F M F F F M F F F F M * 23 44F * 24 60F * ± ± ± ± F F M F ± 51.6 # 45.4 ± ± ± : Cushing s disease, 25 28: ectopic ACTH syndrome *associated with nodular lesion in hypertrophied adrenal glands # P<0.05 vs. patients nos Data are presented as means ± SE Table 2. Correlation between Plasma ACTH, cortisol, or UFC and the width of six components of adrenal glands Plasma ACTH Plasma cortisol UFC Right body r = 0.662, P = r = 0.807, P< r = 0.739, P< Right lateral limb r = 0.514, P = r = 0.707, P = r = 0.588, P = Right medial limb r = 0.559, P = r = 0.771, P< r = 0.636, P = Left body r = 0.636, P = r = 0.638, P = r = 0.536, P = Left lateral limb *r = 0.217, P = r = 0.605, P = r = 0.725, P = Left medial limb r = 0.502, P = r = 0.685, P = r = 0.709, P< Endocrinological data were well correlated with the size of adrenal glands, and correlation was statistically significant except for that between plasma ACTH and left lateral limb*.

4 92 IMAKI et al. Fig. 1. A scattergram showing the relationship between the width of the right adrenal body and (a) plasma ACTH, (b) plasma cortisol, and (c) UFC. The data of 4 cases (nos. 3, 18, 24, 27) could not be obtained due mainly to the shape of the glands. The regression line of these associations is indicated with the shaded area representing the upper limit of normal right adrenal width reported by Vincent et al. (5). Open circles, Cushing s disease; closed circles, ectopic ACTH syndrome; arrow, a patient with marked hyperplasia and low normal plasma ACTH level (case 13); arrowheads, patients with ectopic ACTH syndrome with normal sized right adrenal body (case nos. 25 & 26). In case 25 the adrenal widths were enlarged with the exception of the right adrenal body. (Fig. 1a), cortisol (Fig. 1b) and UFC (Fig. 1c). Although there was a significant correlation between plasma ACTH and the width of the right adrenal body, one patient (no. 13) had marked adrenal hyperplasia associated with a low normal ACTH level of 21.0 pg/ ml. In contrast, right adrenal body size was within the normal range in two patients with the ectopic ACTH syndrome (nos. 25, 26), in spite of high plasma ACTH levels. We found a strong and significant degree of correlation between the estimated duration of disease and the width of the adrenal glands with the exception of the right lateral limb and the left adrenal body (Table 3). In contrast, no correlation was found between age and the width of the adrenal glands (Table 3). In addition, there was no significant difference in the prevalence of hyperplasia between normotensive and hypertensive patients (Fisher s exact probability = 0.57). Comparison of plasma ACTH in patients with either pituitary-dependent Cushing s syndrome (Cushing s disease) or ectopic ACTH syndrome showed these levels were significantly higher in the ectopic group (Table 1). However, the mean width of all of the adrenal components was similar in the two groups. For example, the mean width of the right adrenal body in patients with Cushing s disease was 11.0 ± 1.4 mm compared to 10.3 ± 2.5 mm in patients with the ectopic ACTH syndrome (P = 0.86). The duration of disease was marginally shorter in the patients with the ectopic ACTH syndrome, but this difference was not statistically significant (Table 1). Fig. 2 demonstrates a CT scan of a patient with Cushing s disease (case no. 22) showing marked bilateral hyperplasia. A micronodule was noted in the lateral limb of the right adrenal gland with the patient s plasma ACTH and cortisol levels being markedly elevated (Table 1). Ten years after transsphenoidal surgery and subsequent pituitary irradia- Table 3. Correlation between clinical data and the width of six components of adrenal glands Duration of disease Ages Right body r = 0.494, P = r = 0.067, P = Right lateral limb r = 0.368, P = r = 0.039, P = Right medial limb r = 0.475, P = r = 0.216, P = Left body r = 0.151, P = 0.50 r = 0.173, P = Left lateral limb r = 0.454, P<0.05 r = 0.015, P = Left medial limb r = 0.469, P = r = 0.003, P = 0.989

5 ADRENAL HYPERPLASIA IN CUSHING SYNDROME 93 Fig. 2. CT scans in a 33-year-old man with Cushing s disease (case no. 22). Both the right (a) and left adrenal glands (b) were markedly enlarged with a nodule being detected in the lateral limb of the right adrenal gland (asterisk). (c) Ten years after transsphenoidal surgery and pituitary irradiation, the adrenal glands had returned to normal size (white arrow). tion, endocrinological data showed remission with plasma ACTH 35.8 pg/ml, plasma cortisol 15.4 g/dl, UFC 55.0 g/day, and plasma cortisol after 1 mg dexamethasone 1.2 g/dl. As a result of this treatment the adrenal glands had returned to normal size (Fig. 2c). Fig. 3 shows a CT scan of a 60-year-old female patient (no. 24) who had a macronodule in the right adrenal gland associated with a plasma cortisol level of 23.9 g/dl, a baseline 24-h urinary free cortisol excretion of g/day and a normal plasma ACTH level of 41.0 pg/ml. The CT scan showed diffuse adrenal hyperplasia with a 2.0 cm nodule on the right side (Fig. 3a). The patient underwent transsphenoidal exploration with removal of a 4 mm pituitary adenoma located on the right side of the gland that stained positive in immunohistochemistry for ACTH. Six months after pituitary surgery, bilateral diffuse hyperplasia had regressed, although a repeat CT obtained 6 years after surgery showed the mass was stable but only slightly Fig. 3. CT scans in a 60-year-old woman with Cushing s disease associated with a normal plasma ACTH level of 41.0 pg/ml (case no. 24). (a) A macronodule (asterisk) is seen in the hyperplastic right adrenal gland and the hyperplastic left gland is shown by the white arrow. (b) Diffuse hyperplasia of the adrenal glands had regressed 6 years after transsphenoidal surgery, although a nodule still remained in the right adrenal gland (asterisk). reduced in size (Fig. 3b). This patient was considered to be in remission because of normalization of her plasma ACTH (44.5 pg/ml) and cortisol (9.0 g/dl) levels and suppressibility of plasma cortisol (1.1 g/dl) by 0.5 mg dexamethasone. Discussion CT scanning of the adrenal glands is generally reliable in the differential diagnosis of Cushing s syndrome. Patients with classic ACTH-dependent Cushing s syndrome have normal or symmetrically enlarged adrenal glands, while patients with ACTHindependent Cushing s syndrome have unilateral adrenal tumors [2 4]. In this study we found adrenal

6 94 IMAKI et al. hyperplasia in 57% of patients with ACTH-dependent Cushing s syndrome assessed by CT scan, with this prevalence being slightly lower than that reported in an earlier study [4]. It is apparent that pituitary-derived factors play a crucial role in adrenal growth as hypophysectomy leads to adrenal atrophy, whereas chronic adrenal stimulation in ACTH-dependent Cushing s syndrome causes adrenal hyperplasia [9]. It has been proposed that of those factors, ACTH is the major tropic and mitogenic factor that stimulates the adrenal cortex. Evidence for such actions is that an inactivation mutation of the ACTH receptor is associated with adrenocortical hypoplasia [10]. Our finding of a positive correlation between plasma ACTH and adrenal gland size in patients with ACTH-dependent Cushing s syndrome provides further support for this notion. In addition, the size of adrenal glands returned to normal in one patient in our study following a decrease in plasma ACTH, an observation that implied the hormone had an important role in adrenal growth. POMC-derived peptides are also known to be involved in pituitary-dependent adrenal growth. Short N terminal-pomc peptides are a mitogenic stimulus for the adrenal cortex [11] and recently the mitogen action of two peptides, 1-28 POMC and 1-48 POMC has been shown to be mediated by activation of MAPKs [12]. Indeed, POMC-peptides have been detected by two-site IRMA for POMC in patients with Cushing s disease [13, 14]. Adrenal hyperplasia may therefore occur in patients with low-normal plasma ACTH levels when a corticotrophic adenoma secreting large amounts of N-POMC peptides is present. This may have been the case for patient no. 13 in the present study who had moderate diffuse hyperplasia in spite of a low-normal plasma ACTH level. Our data showed that adrenal size was related to duration of disease. Although plasma ACTH level was significantly higher in patients with the ectopic ACTH syndrome than those with Cushing s disease, there was no tendency for massive adrenal hyperplasia in the ectopic ACTH syndrome group. On the contrary, the size of adrenal glands was within the normal range in spite of high plasma ACTH levels as in case no. 26. Usually the ectopic ACTH syndrome is associated with a more rapid progression. In our patients the mean duration of this condition was marginally shorter than in the patients with pituitary Cushing s syndrome, suggesting that long-standing stimulation by ACTH may be crucial for the development of diffuse hyperplasia in Cushing s syndrome. The reported incidence of macronodular hyperplasia in all patients with Cushing s disease varies between 8% and 40% [15]. In our study the incidence was considerably lower as only 1 (3.6%) of the 28 cases had macronodular hyperplasia. The tumor mass in this female patient remained stable and had decreased only slightly 6 years after surgery when it appeared she was in clinical and endocrinological remission. The presence of unilateral adrenal macronodules associated with Cushing s disease is coincidental, as they represent approximately 1% of the non-functional adrenal mass found on routine CT [16]. Therefore, the coexistence of adrenocortical incidentaloma may be most considerable diagnosis at present. Nevertheless, adrenal scintigraphy is required to ascertain whether a nodule is functional or not. In our patients adrenal gland size did not correlate with age, nor did the prevalence of adrenal hyperplasia differ between normotensive and hypertensive patients. Since aging and hypertension are associated with an increased prevalence of adrenal nodules [15], it is possible that both factors are involved in the induction of nodular lesion, but not in the development of diffuse hyperplasia. We also observed a positive correlation between plasma and urinary free cortisol and the size of all 6 components of the adrenal glands. This finding suggests that adrenal hyperplasia may well reflect the degree of cortisol secretion from adrenal glands. This is consistent with a previous study by Sohaib et al., demonstrating a positive correlation between mean width of the limbs and mean cortisol levels [4]. In summary, our study demonstrated that while long-standing hyperstimulation of the adrenal cortex by hypersecretion of ACTH may lead to the development of diffuse adrenal hyperplasia in patients with ACTH-dependent Cushing s syndrome, other factors that predispose to adrenal hyperplasia cannot be excluded.

7 ADRENAL HYPERPLASIA IN CUSHING SYNDROME 95 References 1. Stewart P (2003) Adrenocortical disease. In: Reed Larsen P, Kronenberg HM, Melmed S, Polonsky KS (eds) Williams Textbook of Endocrinology. Saunders, Philadelphia, Doppmann JL, Miller DL, Dwyer AJ, Loughlin T, Nieman L, Cutler GB, Chrousos GP, Oldfield E, Loriaux DL (1988) Macronodular adrenal hyperplasia in Cushing s disease. Radiology 166: Komiya I, Takasu N, Ohara N, Nagasawa Y, Asawa T, Yamada T, Hashizume K (1994) Analysis of adrenocortical hyperplasia by computed tomography in patients with Cushing s disease, idiopathic hyperaldosteronism and adrenogenital syndrome. Endocr J 41: Sohaib SA, Hanson JA, Newell-Price JDC, Trainer PJ, Monson JP, Grossmann AB, Besser GM, Reznek RH (1999) CT appearance of the adrenal glands in adrenocorticotrophic hormone-dependent Cushing s syndrome. Am J Roentogenol 172: Vincent JM, Morrison ID, Armstrong P, Reznek RH (1994) The size of adrenal glands on computed tomography. Clin Radiol 49: Imaki T, Tsushima T, Hizuka N, Odagiri E, Murata Y, Suda T, Takano K (2001) Post-operative plasma cortisol levels predict long-term outcome in patients with Cushing s disease and determine which patients should be treated with pituitary irradiation after surgery. Endocr J 48: Ushiyama T, Suda T, Sumitomo T, Nakano Y, Tozawa F, Demura H (1990) Comparison of plasma ACTH determination between IRMA and extraction method. ACTH Related Peptides 1: (In Japanese). 8. Sato Y, Suda T, Tozawa F, Nakano F, Demura H (1991) Examination of high sensitive ACTH assay ACTH IRMA Kit Mitsubishi Yuka and experience of using ACTH II IRMA kit Mitsubishi Yuka. ACTH Related Peptides 2: 6 10 (In Japanese). 9. Estivariz FE, Lowry PJ, Jackson S (1992) Control of adrenal growth. In: James VHT (ed) The adrenal gland. 2nd ed. Raven Press, New York, Clark AJ, Weber A (1988) Adrenocorticotropin insensitivity syndromes. Endocr Rev 19: Lowry PJ, Silas L, McLean C, Lintin EA, Estivariz FE (1983) Pro- -melanocyte-stimulating hormone cleavage in adrenal gland undergoing compensatory growth. Nature 306: Fassnacht M, Hahner S, Hansen IA, Kreutzberger T, Zink M, Adermann K, Jakob F, Troppmair J, Allolio B (2003) N-terminal proopiomelanocortin acts as a mitogen in adrenocortical tumor cells and decreases adrenal steroidogenesis. J Clin Endocrinol Metab 88: Gibson S, Crosby SR, Stewart MF, Jennings AM, McCall E, White A (1994) Differential release of proopiomelanocortin-derived peptides from the human pituitary: evidence from a panel of two-site immunoradiometric assays. J Clin Endocrinol Metab 78: White A, Gibson S (1998) ACTH precursors: biological significance and clinical relevance. Clin Endocrinol 48: Smals AGH, Pieters GFFM, Van Haelst UJG, Kloppenborg PWC (1984) Macronodular adrenocortical hyperplasia in long-standing Cushing s disease. J Clin Endocrinol Metab 58: Borretta G, Terzolo M, Cesario F, Meineri I, Pia A, Angeli A (1996) Coexistence of adrenal macronodule and Cushing s disease. Report of two cases. J Endocrinol Invest 19: