Association of IGF-1 Level with Low Bone Mass in Young Patients with Cushing’s Disease

Posted on July 28, 2023 by MaryO

Abstract

Purpose. Few related factors of low bone mass in Cushing’s disease (CD) have been identified so far, and relevant sufficient powered studies in CD patients are rare. On account of the scarcity of data, we performed a well-powered study to identify related factors associated with low bone mass in young CD patients.

Methods. This retrospective study included 153 CD patients (33 males and 120 females, under the age of 50 for men and premenopausal women). Bone mineral density (BMD) of the left hip and lumbar spine was measured by dual energy X-ray absorptiometry (DEXA). In this study, low bone mass was defined when the Z score was −2.0 or lower. Results. Among those CD patients, low bone mass occurred in 74 patients (48.37%). Compared to patients with normal BMD, those patients with low bone mass had a higher level of serum cortisol at midnight (22.31 (17.95-29.62) vs. 17.80 (13.75-22.77), ), testosterone in women (2.10 (1.33–2.89) vs. 1.54 (0.97–2.05), ), higher portion of male (32.43% vs. 11.54%, ) as well as hypertension (76.12% vs. 51.67%, ), and lower IGF-1 index (0.59 (0.43–0.76) vs. 0.79 (0.60–1.02), ). The Z score was positively associated with the IGF-1 index in both the lumbar spine (r = 0.35153, ) and the femoral neck (r = 0.24418, ). The Z score in the femoral neck was negatively associated with osteocalcin (r = −0.22744, ). Compared to the lowest tertile of the IGF-1 index (<0.5563), the patients with the highest tertile of the IGF-1 index (≥0.7993) had a lower prevalence of low bone mass (95% CI 0.02 (0.001–0.50), ), even after adjusting for confounders such as age, gender, duration, BMI, hypertension, serum cortisol at midnight, PTH, and osteocalcin.

Conclusions. The higher IGF-1 index was independently associated with lower prevalence of low bone mass in young CD patients, and IGF-1 might play an important role in the pathogenesis of CD-caused low bone mass.

1. Introduction

Cushing’s disease (CD), caused by an adrenocorticotropic hormone (ACTH)-secreting pituitary tumor, is a rare disease with approximately 1.2 to 2.4 new cases per million people each year [1].

Osteoporosis has been recognized as a serious consequence of endogenous hypercortisolism since the first description in 1932 [2]. The prevalence of osteoporosis is around 38–50%, and the rate of atraumatic compression fractures is 15.8% in CD patients [3]. After cortisol normalization and appropriate treatment, recovery of the bone impairment occurs slowly (6–9 years) and partially [4, 5]. Hypercortisolemia impairs bone quality through multiple mechanisms [6]. Growth hormone (GH) and insulin-like growth factor 1 (IGF-1) play a crucial role in bone growth and development [7]. IGF-1 is considered essential for the longitudinal growth of bone, skeletal maturity, and bone mass acquisition not only during growth but also in the maintenance of bone in adults [8]. Previous research studies revealed that low serum IGF-1 levels were associated with a 40% increased risk of fractures [9, 10], and serum IGF-1 levels could be clinically useful for evaluating the risk of spinal fractures [11]. In Marl Hotta’s research, extremely low or no response of plasma GH to recombinant human growth hormone (hGRH) injection was noted in CD patients. This result suggested that the diminished hGRH-induced GH secretion in patients with Cushing’s syndrome might be caused by the prolonged period of hypercortisolemia [12]. Other surveys indicated that glucocorticoids, suppressing GH–IGF-1 and the hypothalamic-pituitary-gonadal axes, lead to decreased number and dysfunction of osteoblast [13].

However, the exact mechanism is still unclear, and few risk factors for osteoporosis in CD have been identified so far. Until now, relevant and sufficiently powered studies in CD patients have been rare [14, 15]. Early recognition of the changes in bone mass in CD patients contributes to early diagnosis of bone mass loss and prompt treatment, which could help minimize the incidence of adverse events such as fractures.

On account of the scarcity of data and pressing open questions concerning risk evaluation and management of osteoporosis, we performed a well-powered study to identify the related factors associated with low bone mass in young CD patients at the time of diagnosis.

2. Materials and Methods

2.1. Subjects

This retrospective study enrolled 153 CD patients (33 males and 120 females) from the Department of Endocrinology and Metabolism of Huashan Hospital between January 2010 and February 2021. All subjects were evaluated by the same group of endocrinologists for detailed clinical evaluation. This study, which was in complete adherence to the Declaration of Helsinki, was approved by the Human Investigation Ethics Committee at Huashan Hospital, Fudan University (No. 2017M011). We collected data on demographic characteristics, laboratory tests, and bone mineral density.

Inclusion criteria included the following: (1) willingness to participate in the study; (2) premenopausal women ≥18 years old, men ≥18 years old but younger than 50 years old, and young women (<50 years old) with menstrual abnormalities who were associated with CD after excluding menstrual abnormalities caused by other causes; (3) diagnosis of CD according to the updated diagnostic criteria [16]; and (4) pathological confirmation after transsphenoidal surgery (positive immunochemistry staining with ACTH). Exclusion criteria included Cushing’s syndrome other than pituitary origin.

2.2. Clinical and Biochemical Methods

IGF-1 was measured using the Immulite 2000 enzyme-labeled chemiluminescent assay (Siemens Healthcare Diagnostic, Surrey, UK). Other endocrine hormones, including cortisol (F), 24-hour urinary free cortisol (24hUFC), adrenocorticotropic hormone (ACTH), prolactin (PRL), luteinizing hormone (LH), follicle stimulating hormone (FSH), estrogen (E2), progesterone (P), testosterone (T), thyroid stimulating hormone (TSH), and free thyroxine (FT4), were carried out by the chemiluminescence assay (Advia Centaur CP). Intra-assay and interassay coefficients of variation were less than 8 and 10%, respectively, for the estimation of all hormones.

Bone metabolism markers included osteocalcin (OC), type I procollagen amino-terminal peptide (P1NP), parathyroid hormone (PTH), and 25-hydroxyvitamin D (25(OH)VD), measured in a Roche Cobas e411 analyzer using immunometric assays (Roche Diagnostics, Indianapolis, IN, USA).

The IGF-1 index was defined as the ratio of the measured value to the respective upper limit of the reference range for age and sex. Body mass index (BMI) was calculated using the following formula: weight (kg)/height2 (m2). The bone mineral density (BMD) measuring instrument was Discovery type W dual energy X-ray absorptiometry from the American HOLOGIC company. Quality control tests were conducted every working day. Before examination, the date of birth, height, weight, and menopause date of the examiner were accurately recorded, and then BMD (g/cm2) of the left hip and lumbar spine were measured by DEXA. Z value was used for premenopausal women and men younger than 50 years old, and Z-value = (measured value − mean bone mineral density of peers)/standard deviation of BMD of peers [17, 18]. In this study, low bone mass was defined as a Z-value of −2.0 or lower.

2.3. Statistical Analysis

The baseline characteristics were compared between CD patients with and without low bone mass by using the Student’s t-test for continuous variables and the χ2 test for category variables. Bone turnover markers, alanine aminotransferase (ALT), triglyceride (TG), IGF-1 index, thyroid stimulating hormone (TSH), free triiodothyronine (FT3), free thyroxine (FT4), testosterone (T), 24 hours of urine cortisol (24 h UFC), and serum cortisol at 8 a.m. (F8 am) and at midnight (F24 pm) were not in normal distribution, so variables mentioned above were Log10-transformed, which could be used as continuous variables during statistical analysis. Participants were categorized into three groups according to tertiles of the IGF-1 index: <0.5986, 0.5986–0.8380, and >0.8380. The linear trend across IGF-1 index tertiles was tested using linear regression analysis for continuous variables and the Cochran–Armitage test for categorical variables. We used a multivariate logistic regression model to identify related factors that are independently associated with the risk of low bone mass. Variables included in the multivariate logistic regression model were selected based on the Spearman rank correlation analysis and established traditional low bone mass risk factors as priors. The results were presented as odds ratios (OR) and the corresponding 95% confidence intervals (CI). Significance tests were two-tailed, with  value <0.05 considered statistically significant for all analyses. Statistical analysis was performed using SAS version 9.3 (SAS Institute Inc, Cary, NC, USA).

3. Results

3.1. The Prevalence of Low Bone Mass in Young Cushing’s Disease Patients

From the inpatient system of Huashan hospital, a total of 153 CD patients under the age of 50 for men and premenopausal women (some with menstrual abnormalities were associated with CD) were included, aged from 13 to 49 years, with an average age of 34.25 ± 8.39 years. There were 33 males (21.57%) and 120 females (78.43%). These CD patients included newly diagnosed CD, recurrences of CD, and CD without remission after treatment. There were no differences in the prevalence of different statuses of CD between the two groups (Table 1).

Table 1 
Clinical and biochemical preoperative characteristics of young Cushing’s disease patients according to status of bone mineral density at diagnosis.

Among these CD patients, low bone mass occurred in 74 patients (48.37%), including 24 men and 50 women. The prevalence of low bone mass was 41.67% and 72.73% in female and male CD patients, respectively, and 42 (56.76%) patients suffered from low bone mass in the lumbar spine only, while 10 (13.51%) patients had low bone mass in the femoral neck only, and 22 (29.73%) patients had low bone mass in both parts.

In female patients with low bone mass, 27 (54%) had low bone mass in the lumbar region only, 9 (18%) in the femoral neck only, and 14 (28%) had low bone mass in both parts. For male patients with low bone mass, 16 (66.67%) patients had low bone mass only in the lumbar region, and the rest (8, 33.33%) had low bone mass in both parts.

Ten patients had a history of fragility fractures (6 ribs, 3 vertebrae, 1 femoral neck, and ribs), and all of them achieved low bone mass in BMD.

3.2. Baseline Characteristics of Cushing’s Disease Patients with and without Low Bone Mass

These CD patients were divided into two groups with and without low bone mass (Table 1). Compared to patients without low bone mass, those low bone mass patients had a higher level of diastolic blood pressure (DBP) (97.07 ± 13.69 vs. 89.76 ± 13.43, ), serum creatinine (66.15 ± 24.33 vs. 55.90 ± 13.35, ), uric acid (0.36 ± 0.10 vs. 0.32 ± 0.10, ), cholesterol (5.57 ± 1.30 vs. 5.06 ± 1.47, ), testosterone in women (2.10 (1.33–2.89) vs. 1.54 (0.97–2.05), ), F24 pm (22.31 (17.95–29.62) vs. 17.80 (13.75–22.77), ), and higher portion of male (32.43% vs. 11.54%, ), as well as hypertension (76.12% vs. 51.67%, ). The low bone mass group had a lower IGF-1 index (0.59 (0.43–0.76) vs. 0.79 (0.60–1.02), ) and FT3 level (3.54 (3.16–4.04) vs. 3.98 (3.47–4.45), ) than those without low bone mass. CD patients without low bone mass were more likely to have serum IGF-1 above the upper limit of the normal reference range (ULN) with age-adjusted (18, 26.87% vs. 3, 4.84%, ). No differences of bone turnover makers were found between the two groups.

3.3. Association between Baseline Characteristics and BMD

Spearman’s rank correlation analysis was used to explore the related factors of low bone mass in young CD patients (Table 2). The results indicated that the Z score in the lumbar spine was positively associated with age at diagnosis (r = 0.18801, ), IGF-1 index (r = 0.35153, ), FT3 level (r = 0.24117, ), estradiol in women (r = 0.2361, ), and occurrence of normal menstruation in females (r = 0.2267, ). Meanwhile, SBP (r = −0.21575, ), DBP (r = −0.32538, ), ALT (r = −0.17477, ), serum creatinine (r = −0.36072, ), cholesterol (r = −0.20205, ), testosterone in women (r = −0.2700, ), F8 am (r = −0.18998, ), and serum cortisol at midnight (r = −0.27273, ) were negatively associated with the Z-score in the lumbar spine. The results also illustrated that the Z-score in the femoral neck was positively associated with BMI (r = 0.33926, ), IGF-1 index (r = 0.24418, ), FT3 level (r = 0.20487, ), and occurrence of normal menstruation in females (r = 0.2393, ). Serum creatinine (r = −0.1932, ), osteocalcin (r = −0.22744, ), and testosterone in women (r = −0.2363, ) were negatively associated with the Z-score in the femoral neck.

Table 2 
Spearman rank correlation of BMD and various variables in Cushing’s disease patients.
3.4. IGF-1 Index and Low Bone Mass

Participants were categorized into the following three groups according to tertiles of the preoperative IGF-1 index: <0.5986 (tertiles 1), 0.5986–0.8380 (tertiles 2), and >0.8380 (tertiles 3). With the IGF-1 index increasing, the level of PTH decreased (54.85 (38.35–66.2), 38.9 (26.6–66.9), 36 (25.5–47.05), and ), while other bone metabolism makers, including PINP, osteocalcin, and 25 (OH) VD, showed no differences among the three groups (Figures 1(a)–1(d)). With the increase in the IGF-1 index level, the Z-score of both vertebra lumbalis (tertiles 1: −2.4 (−3.3∼−1.5); tertiles 2: −1.9 (−2.3∼−1.0); tertiles 3: −1.15 (−1.9∼−0.4), ) and the neck of femur (tertiles 1: −1.7 (−2.3∼−0.95); tertiles 2: −1.2 (−1.9∼−0.5); tertiles 3: −1.0 (−1.5∼−0.5), ) increased gradually (Figures 2(a) and 2(b)). Meanwhile, prevalence of low bone mass decreased (68.29%, 53.33%, 23.81%, ) (Figure 3(a)) both in the vertebra lumbalis (63.41%, 48.89%, 16.67%, ) and the neck of femur (32.5%, 11.11%, 11.19%, ), with the increasing of the IGF-1 index level (Figures 3(b) and 3(c)).

Figure 1 
Bone turnover makers in three groups according to tertiles of the preoperative IGF-1 index. Tertiles 1: <0.5986, tertiles 2: 0.5986–0.8380, and tertiles 3 >0.8380. a for PINP; b for osteocalcin; c for PTH; d for VD-OH25. (a) p for trend = 0.2601. (b) p for trend = 0.1310. (c) p for trend = 0.008. (d) p for trend = 0.7956.
Figure 2 
Z-score of both the neck of femur and the vertebra lumbalis in three tertiles of the IGF-1 index. a for the neck of femur; b for the vertebra lumbalis. Tertiles 1: <0.5986, tertiles 2: 0.5986–0.8380, and tertiles 3 >0.8380. (a) p for trend = 0.0148. (b) p for trend < 0.0001.
Figure 3 
Prevalence of low bone mass according to tertiles of the preoperative IGF-1 index. With increment of the IGF-1 index level, prevalence of low bone mass decreased, both in the vertebra lumbalis and neck of femur. Tertiles 1: <0.5986, tertiles 2: 0.5986–0.8380, and tertiles 3 >0.8380. (a) p for trend = 0.0002. (b) p for trend = 0.0169. (c) p for trend < 0.0001.

In the logistic regression analysis of the related factors of low bone mass, most of the potentially relevant factors were put into this model; only the IGF-1 index was still significantly negatively associated with the prevalence of low bone mass after adjusting for covariables. The results indicated that compared to the patients in the lowest tertile of the IGF-1 index (<0.5563), those with the highest tertile of the IGF-1 index (≥0.7993) had a lower prevalence of low bone mass (95% CI 0.16 (0.06–0.41), ). After adjusting for age, gender, and BMI, the patients in the highest tertile of the IGF-1 index still conferred a lower prevalence of low bone mass (95% CI 0.15 (0.06–0.42), ). The association between the IGF-1 index and low bone mass still existed (95% CI 0.02 (0.001–0.5), ) even after adjusting for age, gender, CD duration, BMI, hypertension, dyslipidemia, diabetes, ALT, Scr, FT3, F24 pm, PTH, and osteocalcin (Table 3). In comparison to the reference population, the participants in the middle tertile of the IGF-1 index (0.5563–0.7993) had no different risk of low bone mass.

Table 3 
Association between the preoperative IGF-1 index and the risk of low bone mass.

4. Discussion

Our results revealed that low bone mass occurred in around half of young CD patients, affecting more males than females, and mostly in the lumbar spine. The CD patients in our study had a high prevalence (48.37%) of low bone mass at the baseline. This was in accordance with the findings of previous research, and the reported prevalence of osteoporosis due to excess endogenous cortisol ranges from 22% to 59% [19–25]. In this study, CD patients’ lumbar vertebrae were more severely affected than the neck of the femur. It is reported that lumbar vertebrae, containing more trabecular bone than femur neck, were more vulnerable to endogenous cortisol [26].

Our results also indicated that men were more prone to low bone mass than women in CD, which was in accordance with several other studies [23, 27, 28]; possibly, the deleterious effect of cortisol excess on BMD might overrule the protective effects of sex hormones, and men were more often hypogonadal compared with women in CD patients. In our study, patients with low bone mass had a significantly higher level of F24 pm. Both cortisol levels in the morning and at midnight, were negatively associated with the Z-score of BMD in the lumbar spine at diagnosis. But these results were not seen in the femoral neck at diagnosis. This further indicated that lumbar vertebrae were more vulnerable to endogenous cortisol. BMI was considered to be associated with bone mass [29]. In our study, higher BMI was associated with higher BMD at diagnosis in the femur neck but not in the lumbar vertebrae, consistent with other studies [30].

Interestingly, besides the above known related factors, we also found that a higher level of the IGF-1 index was strongly associated with a lower prevalence of low bone mass, both in the vertebra lumbalis and the neck of the femur, independently of age, gender, duration, BMI, hypertension, dyslipidemia, diabetes, level of ALT, creatinine, FT3, and F24 pm. The IGF-1 index was also positively associated with the BMD Z-score, both in the lumbar spine and the femoral neck. So far, there have been few studies concerning the association between IGF-1 and low bone mass in Cushing’s disease patients. As we know, GH [31, 32] and IGF-1 [33] have been demonstrated to increase both bone formation (e.g., collagen synthesis) and bone resorption. However, in CD patients, glucocorticoids resulted in decreased number and dysfunction of osteoblasts by inhibiting GH-IGF-1 axes [34, 35]. In vitro studies suggested that at high concentrations of glucocorticoids, a decreased release of GHRH had been reported [36–38]; therefore, GH-IGF-1 axes were inhibited. IGF-1 possessed anabolic mitogenic actions in osteoblasts while reducing the anabolic actions of TGF-β [39]. The decrease in IGF-1 might be a risk factor for low bone mass in CD patients. In vitro studies had also indicated that the suppressive effects of glucocorticoids on osteoblast function can be partially reversed by GH or IGF treatment [8]. In recent years, some studies have also shown that patients with untreated Cushing’s disease may have elevated IGF-1, and mildly elevated IGF-1 in Cushing’s disease does not imply pathological growth hormone excess. Higher IGF-1 levels could predict better outcomes in CD [40, 41]. Possible mechanisms were not clear, which might involve changes in IGF binding proteins (IGFBPs), interference in IGFBP fragments, IGF-1 synthesis or clearance, and/or the effects of hyperinsulinism induced by excess glucocorticoids. In our study, the results also showed that IGF-1 was an independent protective factor for low bone mass in CD patients.

Our study was one of the few well-powered research studies on the association of IGF-1 levels with low bone mass in young CD patients. These represented important strengths of our study, especially given the rarity of CD. The main limitation of this study was its retrospective nature. This could not prove causality. A prospective study should be conducted to explore the causality between IGF-1 and osteoporosis in CD patients. In addition, this study lacked morphometric data for spinal fractures in all patients, which may underestimate the incidence of fractures and osteoporosis. However, our study indicated that a lower IGF-1 index level was significantly associated with low bone mass in young CD patients, which might provide a new aspect to understand the possible risk factors and mechanism of osteoporosis in CD patients.

In conclusion, our study found that a higher IGF-1 index was independently and significantly associated with decreased prevalence of low bone mass in young CD patients, drawing attention to the role of IGF-1 in the pathogenesis of CD-caused low bone mass and may support the exploration of this pathway in therapeutic agent development in antiosteoporosis in CD.

Data Availability

The data used to support the findings of the study are available on request from the authors.

Additional Points

Through a retrospective study of a large sample of Cushing’s disease (CD) patients from a single center, we found that a higher IGF-1 index was independently associated with a lower prevalence of low bone mass in young CD patients and IGF-1 might play an important role in the pathogenesis of CD-caused low bone mass.

Disclosure

Wanwan Sun and Quanya Sun were the co-first authors.

Conflicts of Interest

The authors declare that they have no conflicts of interest.

Authors’ Contributions

Wanwan Sun analyzed the data and wrote the manuscript. Quanya Sun collected the data. Hongying Ye and Shuo Zhang conducted the study design and quality control. All authors read and approved the final manuscript. Wanwan Sun and Quanya Sun contributed equally to this work.

Acknowledgments

The present study was supported by grants from the initial funding of the Huashan Hospital (2021QD023). The study was also supported by grants from Multidisciplinary Diagnosis and Treatment (MDT) demonstration project in research hospitals (Shanghai Medical College, Fudan University, no: DGF501053-2/014).

References

  1. H. Nishioka and S. Yamada, “Cushing’s disease,” Journal of Clinical Medicine, vol. 8, no. 11, p. 1951, 2019.

    View at: Publisher Site | Google Scholar

  2. H. Cushing, “The basophil adenomas of the pituitary body and their clinical manifestations (pituitary basophilism),” Obesity Research, vol. 2, no. 5, pp. 486–508, 1994.

    View at: Publisher Site | Google Scholar

  3. R. A. Feelders, S. J. Pulgar, A. Kempel, and A. M. Pereira, “Management of endocrine disease: the burden of Cushing’s disease: clinical and health-related quality of life aspects,” European Journal of Endocrinology, vol. 167, no. 3, pp. 311–326, 2012.

    View at: Publisher Site | Google Scholar

  4. R. Pivonello, M. De Leo, A. Cozzolino, and A. Colao, “The treatment of Cushing’s disease,” Endocrine Reviews, vol. 36, no. 4, pp. 385–486, 2015.

    View at: Publisher Site | Google Scholar

  5. R. Pivonello, M. C. De Martino, M. De Leo, C. Simeoli, and A. Colao, “Cushing’s disease: the burden of illness,” Endocrine, vol. 56, no. 1, pp. 10–18, 2017.

    View at: Publisher Site | Google Scholar

  6. R. S. Hardy, H. Zhou, M. J. Seibel, and M. S. Cooper, “Glucocorticoids and bone: consequences of endogenous and exogenous excess and replacement therapy,” Endocrine Reviews, vol. 39, no. 5, pp. 519–548, 2018.

    View at: Publisher Site | Google Scholar

  7. R. Bouillon and A. Prodonova, “Growth hormone deficiency and peak bone mass: laboratory for experimental medicine and Endocrinology, catholic university of leuven, gasthuisberg, leuven, Belgium,” Journal of Pediatric Endocrinology and Metabolism, vol. 13, no. s2, pp. 1327–1342, 2000.

    View at: Publisher Site | Google Scholar

  8. A. Giustina, G. Mazziotti, and E. Canalis, “Growth hormone, insulin-like growth factors, and the skeleton,” Endocrine Reviews, vol. 29, no. 5, pp. 535–559, 2008.

    View at: Publisher Site | Google Scholar

  9. T. Sugimoto, K. Nishiyama, F. Kuribayashi, and K. Chihara, “Serum levels of insulin-like growth factor (IGF) I, IGF-binding protein (IGFBP)-2, and IGFBP-3 in osteoporotic patients with and without spinal fractures,” Journal of Bone and Mineral Research, vol. 12, no. 8, pp. 1272–1279, 1997.

    View at: Publisher Site | Google Scholar

  10. P. Garnero, E. Sornay-Rendu, and P. D. Delmas, “Low serum IGF-1 and occurrence of osteoporotic fractures in postmenopausal women,” The Lancet, vol. 355, no. 9207, pp. 898-899, 2000.

    View at: Publisher Site | Google Scholar

  11. C. Ohlsson, D. Mellström, D. Carlzon et al., “Older men with low serum IGF-1 have an increased risk of incident fractures: the MrOS Sweden study,” Journal of Bone and Mineral Research, vol. 26, no. 4, pp. 865–872, 2011.

    View at: Publisher Site | Google Scholar

  12. M. Hotta, T. Shibasaki, A. Masuda et al., “Effect of human growth hormone-releasing hormone on GH secretion in Cushing’s syndrome and non-endocrine disease patients treated with glucocorticoids,” Life Sciences, vol. 42, no. 9, pp. 979–984, 1988.

    View at: Publisher Site | Google Scholar

  13. L. T. Braun and M. Reincke, “The effect of biochemical remission on bone metabolism in Cushing’s syndrome: a 2-year follow-up study,” Journal of Bone and Mineral Research, vol. 36, no. 11, pp. 2281-2282, 2021.

    View at: Publisher Site | Google Scholar

  14. I. Kanazawa, T. Yamaguchi, M. Yamamoto, M. Yamauchi, S. Yano, and T. Sugimoto, “Serum insulin-like growth factor-I level is associated with the presence of vertebral fractures in postmenopausal women with type 2 diabetes mellitus,” Osteoporosis International, vol. 18, no. 12, pp. 1675–1681, 2007.

    View at: Publisher Site | Google Scholar

  15. A. Scillitani, G. Mazziotti, C. Di Somma et al., “Treatment of skeletal impairment in patients with endogenous hypercortisolism: when and how?” Osteoporosis International, vol. 25, no. 2, pp. 441–446, 2014.

    View at: Publisher Site | Google Scholar

  16. M. Fleseriu, R. Auchus, I. Bancos et al., “Consensus on diagnosis and management of Cushing’s disease: a guideline update,” Lancet Diabetes and Endocrinology, vol. 9, no. 12, pp. 847–875, 2021.

    View at: Publisher Site | Google Scholar

  17. J. M. Liu, D. L. Zhu, Y. M. Mu, and W. B. Xia, “Chinese Society of Osteoporosis and Bone Mineral Research, the Chinese Society of Endocrinology, Chinese Diabetes Society, Chinese Medical Association; Chinese Endocrinologist Association, Chinese Medical Doctor Association,” Management of fracture risk in patients with diabetes-Chinese Expert Consensus Journal of Diabetes, vol. 11, pp. 906–919, 2019.

    View at: Google Scholar

  18. P. M. Camacho, S. M. Petak, N. Binkley et al., “American association of clinical endocrinologists and American college of endocrinology clinical practice guidelines for the diagnosis and treatment of postmenopausal OSTEOPOROSIS-2020 update,” Endocrine Practice, vol. 26, no. 1, pp. 1–46, 2020.

    View at: Publisher Site | Google Scholar

  19. C. V. dos Santos, L. Vieira Neto, M. Madeira et al., “Bone density and microarchitecture in endogenous hypercortisolism,” Clinical Endocrinology, vol. 83, no. 4, pp. 468–474, 2015.

    View at: Publisher Site | Google Scholar

  20. N. Ohmori, K. Nomura, K. Ohmori, Y. Kato, T. Itoh, and K. Takano, “Osteoporosis is more prevalent in adrenal than in pituitary Cushing’s syndrome,” Endocrine Journal, vol. 50, no. 1, pp. 1–7, 2003.

    View at: Publisher Site | Google Scholar

  21. M. E. Randazzo, E. Grossrubatscher, P. Dalino Ciaramella, A. Vanzulli, and P. Loli, “Spontaneous recovery of bone mass after cure of endogenous hypercortisolism,” Pituitary, vol. 15, no. 2, pp. 193–201, 2012.

    View at: Publisher Site | Google Scholar

  22. L. Tauchmanovà, R. Pivonello, C. Di Somma et al., “Bone demineralization and vertebral fractures in endogenous cortisol excess: role of disease etiology and gonadal status,” Journal of Clinical Endocrinology and Metabolism, vol. 91, no. 5, pp. 1779–1784, 2006.

    View at: Publisher Site | Google Scholar

  23. E. Valassi, A. Santos, M. Yaneva et al., “The European Registry on Cushing’s syndrome: 2-year experience. Baseline demographic and clinical characteristics,” European Journal of Endocrinology, vol. 165, no. 3, pp. 383–392, 2011.

    View at: Publisher Site | Google Scholar

  24. L. Trementino, G. Appolloni, L. Ceccoli et al., “Bone complications in patients with Cushing’s syndrome: looking for clinical, biochemical, and genetic determinants,” Osteoporosis International, vol. 25, no. 3, pp. 913–921, 2014.

    View at: Publisher Site | Google Scholar

  25. A. W. van der Eerden, M. den Heijer, W. J. Oyen, and A. R. Hermus, “Cushing’s syndrome and bone mineral density: lowest Z scores in young patients,” The Netherlands Journal of Medicine, vol. 65, no. 4, pp. 137–141, 2007.

    View at: Google Scholar

  26. P. G. Lacativa and M. L. F. Farias, “Office practice of osteoporosis evaluation,” Arquivos Brasileiros de Endocrinologia and Metabologia, vol. 50, no. 4, pp. 674–684, 2006.

    View at: Publisher Site | Google Scholar

  27. L. H. A. Broersen, F. M. van Haalen, T. Kienitz et al., “Sex differences in presentation but not in outcome for ACTH-dependent Cushing’s syndrome,” Frontiers in Endocrinology, vol. 10, p. 580, 2019.

    View at: Publisher Site | Google Scholar

  28. F. P. Giraldi, M. Moro, and F. Cavagnini, “Gender-related differences in the presentation and course of Cushing’s disease,” Journal of Clinical Endocrinology and Metabolism, vol. 88, no. 4, pp. 1554–1558, 2003.

    View at: Publisher Site | Google Scholar

  29. S. Morin, J. F. Tsang, and W. D. Leslie, “Weight and body mass index predict bone mineral density and fractures in women aged 40 to 59 years,” Osteoporosis International, vol. 20, no. 3, pp. 363–370, 2009.

    View at: Publisher Site | Google Scholar

  30. M. Zilio, M. Barbot, F. Ceccato et al., “Diagnosis and complications of Cushing’s disease: gender-related differences,” Clinical Endocrinology, vol. 80, no. 3, pp. 403–410, 2014.

    View at: Publisher Site | Google Scholar

  31. G. Amato, C. Carella, S. Fazio, G. La Montagna, A. Cittadini, and D. Sabatini, “Body composition, bone metabolism, and heart structure and function in growth hormone (GH)-deficient adults before and after GH replacement therapy at low doses,” Journal of Clinical Endocrinology and Metabolism, vol. 77, no. 6, pp. 1671–1676, 1993.

    View at: Publisher Site | Google Scholar

  32. S. A. Beshyah, E. Thomas, P. Kyd, P. Sharp, A. Fairney, and D. G. Johnston, “The effect of growth hormone replacement therapy in hypopituitary adults on calcium and bone metabolism,” Clinical Endocrinology, vol. 40, no. 3, pp. 383–391, 2010.

    View at: Publisher Site | Google Scholar

  33. P. R. Ebeling, J. D. Jones, W. M. O’Fallon, C. H. Janes, and B. L. Riggs, “Short-term effects of recombinant human insulin-like growth factor I on bone turnover in normal women,” Journal of Clinical Endocrinology and Metabolism, vol. 77, no. 5, pp. 1384–1387, 1993.

    View at: Publisher Site | Google Scholar

  34. G. Mazziotti and A. Giustina, “Glucocorticoids and the regulation of growth hormone secretion,” Nature Reviews Endocrinology, vol. 9, no. 5, pp. 265–276, 2013.

    View at: Publisher Site | Google Scholar

  35. N. A. Tritos, “Growth hormone deficiency in adults with Cushing’s disease,” Best Practice and Research Clinical Endocrinology and Metabolism, vol. 35, no. 2, Article ID 101474, 2021.

    View at: Publisher Site | Google Scholar

  36. K. Nakagawa, T. Ishizuka, T. Obara, M. Matsubara, and K. Akikawa, “Dichotomic action of glucocorticoids on growth hormone secretion,” Acta Endocrinologica, vol. 116, no. 2, pp. 165–171, 1987.

    View at: Publisher Site | Google Scholar

  37. G. Fernández-Vázquez, L. Cacicedo, M. J. Lorenzo, R. Tolón, J. López, and F. Sánchez-Franco, “Corticosterone modulates growth hormone-releasing factor and somatostatin in fetal rat hypothalamic cultures,” Neuroendocrinology, vol. 61, no. 1, pp. 31–35, 1995.

    View at: Publisher Site | Google Scholar

  38. S. K. Fife, R. S. Brogan, A. Giustina, and W. B. Wehrenberg, “Immunocytochemical and molecular analysis of the effects of glucocorticoid treatment on the hypothalamic-somatotropic axis in the rat,” Neuroendocrinology, vol. 64, no. 2, pp. 131–138, 1996.

    View at: Publisher Site | Google Scholar

  39. T. L. McCarthy, M. Centrella, and E. Canalis, “Cortisol inhibits the synthesis of insulin-like growth factor-I in skeletal cells,” Endocrinology, vol. 126, no. 3, pp. 1569–1575, 1990.

    View at: Publisher Site | Google Scholar

  40. K. English, V. Chikani, G. Dimeski, and W. J. Inder, “Elevated insulin‐like growth factor‐1 in Cushing’s disease,” Clinical Endocrinology, vol. 91, no. 1, pp. 141–147, 2019.

    View at: Publisher Site | Google Scholar

  41. E. Gezer, B. Çetinarslan, A. Selek et al., “The association between insulin-like growth factor 1 levels within reference range and early postoperative remission rate in patients with Cushing’s disease,” Endocrine Research, vol. 46, no. 3, pp. 92–98, 2021.

    View at: Publisher Site | Google Scholar

Copyright © 2023 Wanwan Sun et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Filed under: Cushing's, Diagnostic Testing, pituitary, symptoms | Tagged: , , , | Leave a comment »

High-resolution Contrast-enhanced MRI With Three-Dimensional Fast Spin Echo Improved the Diagnostic Performance for Identifying Pituitary Microadenomas In Cushing’s Syndrome

Posted on July 26, 2023 by MaryO

Abstract

Objectives

To assess the diagnostic performance of high-resolution contrast-enhanced MRI (hrMRI) with three-dimensional (3D) fast spin echo (FSE) sequence by comparison with conventional contrast-enhanced MRI (cMRI) and dynamic contrast-enhanced MRI (dMRI) with 2D FSE sequence for identifying pituitary microadenomas.

Methods

This single-institutional retrospective study included 69 consecutive patients with Cushing’s syndrome who underwent preoperative pituitary MRI, including cMRI, dMRI, and hrMRI, between January 2016 to December 2020. Reference standards were established by using all available imaging, clinical, surgical, and pathological resources. The diagnostic performance of cMRI, dMRI, and hrMRI for identifying pituitary microadenomas was independently evaluated by two experienced neuroradiologists. The area under the receiver operating characteristics curves (AUCs) were compared between protocols for each reader by using the DeLong test to assess the diagnostic performance for identifying pituitary microadenomas. The inter-observer agreement was assessed by using the κ analysis.

Results

The diagnostic performance of hrMRI (AUC, 0.95–0.97) was higher than cMRI (AUC, 0.74–0.75; p ≤ .002) and dMRI (AUC, 0.59–0.68; p ≤ .001) for identifying pituitary microadenomas. The sensitivity and specificity of hrMRI were 90–93% and 100%, respectively. There were 78% (18/23) to 82% (14/17) of the patients, who were misdiagnosed on cMRI and dMRI and correctly diagnosed on hrMRI. The inter-observer agreement for identifying pituitary microadenomas was moderate on cMRI (κ = 0.50), moderate on dMRI (κ = 0.57), and almost perfect on hrMRI (κ = 0.91), respectively.

Conclusions

The hrMRI showed higher diagnostic performance than cMRI and dMRI for identifying pituitary microadenomas in patients with Cushing’s syndrome.

Key Points

• The diagnostic performance of hrMRI was higher than cMRI and dMRI for identifying pituitary microadenomas in Cushing’s syndrome.

• About 80% of patients, who were misdiagnosed on cMRI and dMRI, were correctly diagnosed on hrMRI.

• The inter-observer agreement for identifying pituitary microadenomas was almost perfect on hrMRI.

Introduction

Cushing’s syndrome, caused by excessive exposure to glucocorticoids, is associated with considerable morbidity and increased mortality [1]. Cushing’s syndrome has diverse manifestations, including central obesity, moon facies, purple striae, and hypertension [2]. Cushing’s disease, due to adrenocorticotropic hormone (ACTH) hypersecretion from pituitary adenomas, is the most common etiology of ACTH-dependent Cushing’s syndrome [1, 2]. According to the Endocrine Society Clinical Practice Guideline, transsphenoidal surgery is the first-line treatment for Cushing’s disease [3]. The identification of pituitary adenomas on preoperative MRI can significantly increase the postoperative remission rate from 50 to 98% [4]. Therefore, it is critical to identify pituitary adenomas on MRI before surgery.

However, there are considerable challenges in identifying ACTH-secreting pituitary adenomas. This is because about 90% of the tumors are microadenomas (less than 10 mm in size) and the median diameter at surgery is about 5 mm [5, 6]. Conventional contrast-enhanced MRI (cMRI) using a two-dimensional (2D) fast spin echo (FSE) sequence has been routinely used to acquire images with 2- to 3-mm slice thickness, but some microadenomas are difficult to be identified on cMRI, resulting in false negatives reported in up to 50% of patients with Cushing’s disease [7]. Dynamic contrast-enhanced MRI (dMRI) increases the sensitivity of identifying pituitary adenomas to 66% [8], but it also increases false positives at the same time [9, 10]. The 3D spoiled gradient recalled (SPGR) sequence has been introduced in high-resolution contrast-enhanced MRI (hrMRI) to acquire images with 1- to 1.2-mm slice thickness. It is reported that the 3D SPGR sequence is superior to the 2D FSE sequence in the identification of pituitary adenomas with a sensitivity of up to 80% [11,12,13], but it cannot satisfy the clinical needs that about 20% of the lesions are still missed. Therefore, techniques are needed that can help better identify pituitary adenomas, particularly microadenomas. Previously, the 3D FSE sequence was recommended in patients with hyperprolactinemia [14]. Recently, the 3D FSE sequence has developed rapidly and can provide superior image quality with diminished artifacts [15]. Sartoretti et al demonstrated in a very effective fashion that the 3D FSE sequence is a reliable alternative for pituitary imaging in terms of image quality [16]. However, to our knowledge, few studies have investigated the diagnostic performance of 3D FSE sequences for identifying ACTH-secreting pituitary adenomas, particularly microadenomas.

The aim of our study was to assess the diagnostic performance of hrMRI with 3D FSE sequence by comparison with cMRI and dMRI with 2D FSE sequence for identifying ACTH-secreting pituitary microadenomas in patients with Cushing’s syndrome.

Materials and methods

This single-institutional retrospective study was approved by the Institutional Review Board of our hospital. The study was conducted in accordance with the Helsinki Declaration. The informed consent was waived due to the retrospective nature of the study.

Study participants

We retrospectively reviewed the medical records and imaging studies of 186 consecutive patients with ACTH-dependent Cushing’s syndrome, who underwent a combined protocol of cMRI, dMRI, and hrMRI from January 2016 to December 2020. Postoperative patients with Cushing’s disease (n = 97), patients with ectopic ACTH syndrome who underwent pituitary exploration (n = 2), and patients with macroadenomas (n = 5) or lack of pathology (n = 13) were excluded from the study. Finally, 69 patients with ACTH-dependent Cushing’s syndrome were included in the current study (Fig. 1) and the patients included were all surgically confirmed.

Fig. 1
figure 1

Flowchart of patient inclusion/exclusion process and image analysis. ACTH adrenocorticotropic hormone, CD Cushing’s disease, EAS ectopic ACTH syndrome, T1WI T1-weighted imaging, T2WI T2-weighted imaging

MRI protocol

All the patients were imaged on a 3.0 Tesla MR scanner (Discovery MR750w, GE Healthcare) using an 8-channel head coil. The MRI protocol included coronal T2-weighted imaging, coronal T1-weighted imaging, and sagittal T1-weighted imaging before contrast injection. After contrast injection of gadopentetate dimeglumine (Gd-DTPA) at 0.05 mmol/kg (0.1 mL/kg) with a flow rate of 2 mL/s followed by a 10-mL saline solution flush, dMRI and cMRI with 2D FSE sequence were obtained first, and hrMRI with 3D FSE sequence using variable flip angle technique was performed immediately afterward. Detailed acquisition parameters are presented in Table S1.

Image analysis: diagnostic performance

Image interpretation was independently conducted by two experienced neuroradiologists (F.F. and H.Y. with 25 and 16 years of experience in neuroradiology, respectively), who were blinded to patient information. The evaluation order of cMRI, dMRI, and hrMRI sequences was randomized. The identification of pituitary microadenomas on images was scored based on a three-point scale (0 = poor; 1 = fair; 2 = excellent). Scores of 1 or 2 represented the identification of the lesion. Reference standards were established by using all available imaging, clinical, surgical, and pathological resources, with a multidisciplinary team approach.

Image analysis: image quality

Two readers (Z.L. and B.H. with 4 years of experience in radiology, respectively) were asked to assess the image quality of cMRI, dMRI, and hrMRI. Before exposure to images used in the current study, these readers underwent a training session to make sure that they were comparable to the experienced neuroradiologists in terms of image quality assessment. Images were presented in a random order. Image quality was assessed by using a 5-point Likert scale [17], including overall image quality (1 = non-diagnostic; 2 = poor; 3 = fair; 4 = good; 5 = excellent), sharpness (1 = non-diagnostic; 2 = not sharp; 3 = a little sharp; 4 = moderately sharp; 5 = satisfyingly sharp), and structural conspicuity (1 = non-diagnostic; 2 = poor; 3 = fair; 4 = good; 5 = excellent). An example of image quality assessment is shown in Table S2. Final decision was made through a consensus agreement.

The mean signal intensity of pituitary microadenomas, pituitary gland, and noise on cMRI, dMRI, and hrMRI was measured using an operator-defined region of interest. For noise, a 10-mm2 region of interest was placed in the background, and noise was defined as the standard deviation of the signal intensity of the background [17]. For pituitary microadenomas and pituitary gland, the region of interest should include a representative portion of the structure. The mean signal intensity of the pituitary microadenoma was replaced with that of the pituitary gland when no microadenoma was identified. A signal-to-noise ratio (SNR) was defined as the mean signal intensity of the pituitary microadenoma divided by noise. A contrast-to-noise ratio (CNR) was defined as the absolute difference of the mean signal intensity between the normal pituitary gland and pituitary microadenomas divided by noise [17]. Supplementary Fig. 1 shows how to measure the SNR and CNR with the region of interest in a contrast-enhanced pituitary MRI. Supplementary Fig. 2 shows the selection of images for the SNR and CNR calculation.

Statistical analysis

The κ analysis was conducted to assess the inter-observer agreement for identifying pituitary microadenomas. The κ value was interpreted as follows: below 0.20, slight agreement; 0.21–0.40, fair agreement; 0.41–0.60, moderate agreement; 0.61–0.80, substantial agreement; greater than 0.80, almost perfect agreement.

To assess the diagnostic performance of cMRI, dMRI, and hrMRI for identifying pituitary microadenomas, the receiver operating characteristic curves were plotted and the area under curves (AUCs) were compared between MR protocols for each reader by using the DeLong test. Sensitivity, specificity, positive predictive value, and negative predictive value were calculated. The Mann–Whitney U test was used to evaluate the difference in image quality scores and the Wilcoxon signed-rank test was used to evaluate SNR and CNR measurements between MR protocols. A p value of less than 0.05 was considered statistically significant. Statistical analysis was performed using MedCalc Statistical Software (version 20.0.15; MedCalc Software) and SPSS Statistics (version 22.0; IBM).

Results

Clinical characteristics

A total of 69 patients (median age, 39 years; interquartile range [IQR], 29–54 years; 38 women [55%]) with ACTH-dependent Cushing’s syndrome were included in the study and their clinical characteristics are shown in Table 1. Among the 69 patients, 60 (87%) patients were diagnosed with Cushing’s disease and 9 (13%) were ectopic ACTH syndrome. The median disease course was 36 months (IQR, 12–78 months). The median serum cortisol, ACTH, and 24-h urine free cortisol level before surgery were 33.0 μg/dL (IQR, 25.1–40.1 μg/dL; normal range 4.0–22.3 μg/dL), 77.2 ng/L (IQR, 55.0–124.0 ng/L; normal range 0–46 ng/L), and 422.0 μg (IQR, 325.8–984.6 μg; normal range 12.3–103.5 μg), respectively. The median serum cortisol and 24-h urine free cortisol level after surgery were 3.0 μg/dL (IQR, 1.8–18.4 μg/dL) and 195.6 μg (IQR, 63.5–1240.3 μg), respectively. The median diameter of pituitary microadenomas was 5 mm (IQR, 4–5 mm), ranging from 3 to 9 mm.

Table 1 Clinical characteristics of the patients

Diagnostic performance of cMRI, dMRI, and hrMRI for identifying pituitary microadenomas

The inter-observer agreement for identifying pituitary microadenomas by κ statistic between two readers was moderate on cMRI (κ = 0.50), moderate on dMRI (κ = 0.57), and almost perfect on hrMRI (κ = 0.91), respectively.

The diagnostic performance for identifying pituitary microadenomas on cMRI, dMRI, hrMRI, and combined cMRI and dMRI is summarized in Table 2. For reader 1, the diagnostic performance of hrMRI (AUC, 0.95; 95%CI: 0.87, 0.99) was higher than that of cMRI (AUC, 0.75; 95%CI: 0.63, 0.85; p = 0.002), dMRI (AUC, 0.59; 95%CI: 0.47, 0.71; p < 0.001), and combined cMRI and dMRI (AUC, 0.65; 95%CI: 0.53, 0.76; p = 0.001). For reader 2, the diagnostic performance of hrMRI (AUC, 0.97; 95%CI: 0.89, 1.00) was higher than that of cMRI (AUC, 0.74; 95%CI: 0.63, 0.84; p = 0.001), dMRI (AUC, 0.68; 95%CI: 0.56, 0.79; p = 0.001), and combined cMRI and dMRI (AUC, 0.70; 95%CI: 0.58, 0.80; p = 0.003).

Table 2 Diagnostic performance of cMRI, dMRI, and hrMRI for identifying pituitary microadenomas

For reader 1, 23 of the 69 patients (33%) were misdiagnosed on both cMRI and dMRI, but 18 of the 23 misdiagnosed patients (78%) were correctly diagnosed on hrMRI. For reader 2, 17 of the 69 patients (25%) were misdiagnosed on both cMRI and dMRI, but 14 of the 17 misdiagnosed patients (82%) were correctly diagnosed on hrMRI.

Figure 2 shows that a 5-mm pituitary microadenoma was identified on preoperative pituitary MRI. The margin of the lesion was fully delineated on hrMRI, but not on cMRI and dMRI. Figure 3 shows that a 3-mm pituitary microadenoma was missed on cMRI, but identified on dMRI and hrMRI. Figure 4 shows that a 5-mm pituitary microadenoma was correctly diagnosed on hrMRI, but missed on cMRI or dMRI. Figure 5 shows that a 4-mm pituitary microadenoma was evident on coronal images as well as reconstructed axial and reconstructed sagittal images on hrMRI.

Fig. 2

figure 2

Images in a 56-year-old man with Cushing’s disease. The 5-mm pituitary microadenoma (arrow) can be identified on (a) coronal contrast-enhanced T1-weighted image and (b) coronal dynamic contrast-enhanced T1-weighted image obtained with two-dimensional (2D) fast spin echo (FSE) sequence, but the margin is not fully delineated. The lesion (arrow) is well delineated on (c) coronal contrast-enhanced T1-weighted image on high-resolution MRI obtained with 3D FSE sequence. d Intraoperative endoscopic photograph during transsphenoidal surgery after exposure of the sellar floor shows a round pituitary microadenoma (arrow)

Fig. 3

figure 3

Images in a 34-year-old woman with Cushing’s disease. No tumor is identified on (a) coronal contrast-enhanced T1-weighted image obtained with two-dimensional (2D) fast spin echo (FSE) sequence. The 3-mm pituitary microadenoma (arrow) with delayed enhancement is identified on the left side of the pituitary gland on (b) coronal dynamic contrast-enhanced T1-weighted image obtained with 2D FSE sequence and (c) coronal contrast-enhanced T1-weighted image on high-resolution MRI obtained with 3D FSE sequence. d Intraoperative endoscopic photograph during transsphenoidal surgery shows a 3-mm pituitary microadenoma (arrow)

Fig. 4

figure 4

Images in a 43-year-old man with Cushing’s disease. The lesion is missed on (a) coronal contrast-enhanced T1-weighted image and (b) coronal dynamic contrast-enhanced T1-weighted image obtained with two-dimensional (2D) fast spin echo (FSE) sequence. c Coronal contrast-enhanced T1-weighted image on high-resolution MRI obtained with 3D FSE sequence shows a round pituitary microadenoma (arrow) measuring approximately 5 mm with delayed enhancement on the left side of the pituitary gland. d Intraoperative endoscopic photograph for microsurgical resection of the 5-mm pituitary microadenoma (arrow)

Fig. 5

figure 5

Images in a 48-year-old woman with Cushing’s disease. Preoperative high-resolution contrast-enhanced MRI using three-dimensional fast spin echo sequence shows a 4-mm pituitary microadenoma (arrow) with delayed enhancement is well delineated on the left side of the pituitary gland on (a) coronal, (b) reconstructed axial, and (c) reconstructed sagittal contrast-enhanced T1-weighted images. d Intraoperative endoscopic photograph during transsphenoidal surgery after exposure of the sellar floor shows a round pituitary microadenoma (arrow)

Image quality of cMRI, dMRI, and hrMRI

Image quality scores of cMRI, dMRI, and hrMRI are presented in Table 3. Scores for overall image quality, sharpness, and structural conspicuity on hrMRI (overall image quality, 5.0 [IQR, 5.0–5.0]; sharpness, 5.0 [IQR, 4.5–5.0]; structural conspicuity, 5.0 [IQR, 5.0–5.0]) were higher than those on cMRI (overall image quality, 4.0 [IQR, 3.5–4.0]; sharpness, 4.0 [IQR, 3.0–4.0]; structural conspicuity, 4.0 [IQR, 4.0–4.0]; p < 0.001 for all) and dMRI (overall image quality, 4.0 [IQR, 4.0–4.0]; sharpness, 4.0 [IQR, 4.0–4.0]; structural conspicuity, 4.0 [IQR, 4.0–4.5]; p < 0.001 for all).

Table 3 Image quality scores on cMRI, dMRI, and hrMRI

The SNR and CNR measurements are shown in Table 4. The SNR of the pituitary microadenomas on hrMRI (67.5 [IQR, 51.2–92.1]) was lower than that on cMRI (82.3 [IQR, 61.8–127.2], p < 0.001), but higher than that on dMRI (53.9 [IQR, 35.2–72.6], p = 0.001). The CNR on hrMRI (26.2 [IQR, 15.1–41.0]) was higher than that on cMRI (10.6 [IQR, 0–42.6], p = 0.023) and dMRI (11.2 [IQR, 0–29.8], p < 0.001).

Table 4 SNR and CNR on cMRI, dMRI, and hrMRI

Discussion

The identification of pituitary microadenomas is considerably challenging but critical in patients with ACTH-dependent Cushing’s syndrome. Our study demonstrated that hrMRI with 3D FSE sequence had higher diagnostic performance (AUC, 0.95–0.97) than cMRI (AUC, 0.74–0.75; p ≤ 0.002) and dMRI (AUC, 0.59–0.68; p ≤ 0.001) for identifying pituitary microadenomas. To our knowledge, there are no previous studies specifically evaluating the identification of pituitary microadenomas on hrMRI with 3D FSE sequence by comparison with cMRI and dMRI in patients with ACTH-dependent Cushing’s syndrome, and this is the largest study conducted in ACTH-secreting microadenomas with a sensitivity of more than 90%.

Recently, techniques for pituitary evaluation have developed rapidly. Because of false negatives and false positives on cMRI and dMRI using 2D FSE sequence [7, 9, 10], a 3D SPGR sequence was introduced for identifying pituitary adenomas. Previous studies demonstrated that the 3D SPGR sequence performed better than the 2D FSE sequence in the identification of pituitary adenomas with a sensitivity of up to 80% [11,12,13]. In patients with hyperprolactinemia, the 3D FSE sequence was recommended [14] and the 3D FSE sequence has rapidly developed recently with superior image quality [15, 16], suggesting that the 3D FSE sequence may be a reliable alternative for identifying pituitary adenomas. However, to our knowledge, few studies have investigated the diagnostic performance of the 3D FSE sequence for identifying ACTH-secreting pituitary adenomas. To fill the gaps, we conducted the current study and revealed that images obtained with the 3D FSE sequence had higher sensitivity (90–93%) in identifying pituitary microadenomas, than that in previous studies using the 3D SPGR sequence [8, 11,12,13].

There is a trade-off between spatial resolution and image noise. The reduced slice thickness can overcome the partial volume averaging effect, but it is associated with increased image noise [17]. Strikingly, our study showed that hrMRI had higher image quality scores than cMRI and dMRI, in terms of overall image quality, sharpness, and structural conspicuity. The SNR of the pituitary microadenomas on cMRI was slightly higher than that on hrMRI in our study. This is because the SNR was calculated as the mean signal intensity of the pituitary gland (instead of the pituitary microadenoma) divided by noise when no microadenoma was identified, and the mean signal intensity of the pituitary gland is higher than that of the pituitary microadenoma. About 40% of pituitary microadenomas were missed on cMRI, whereas less than 10% of pituitary microadenomas were missed on hrMRI. Given the situation mentioned above, the SNR on hrMRI was lower than that on cMRI. However, the CNR on hrMRI was significantly higher than that on cMRI and dMRI. Therefore, hrMRI in our study can dramatically improve the spatial resolution with high CNR, enabling the better identification of pituitary microadenomas.

The identification of pituitary adenomas on preoperative MRI in patients with ACTH-dependent Cushing’s syndrome could help the differential diagnosis of Cushing’s syndrome and aids surgical resection of lesions. It should be noted that most of the pituitary adenomas in patients with Cushing’s disease are microadenomas [5, 6]. In our study, all the tumors are microadenomas with a median diameter of 5 mm (IQR, 4–5 mm), making the diagnosis more challenging. The sensitivity of identifying pituitary adenomas decreased from 80 to 72% after excluding macroadenomas in a previous study [12], whereas the sensitivity of identifying pituitary microadenomas in our study was 90–93% on hrMRI. In the current study, hrMRI performed better than cMRI, dMRI, and combined cMRI and dMRI, with high AUC (0.95–0.97), high sensitivity (90–93%), and high specificity (100%), superior to previous studies [8, 11,12,13]. The high sensitivity of hrMRI for identifying pituitary adenomas will help surgeons improve the postoperative remission rate [4]. The high specificity of hrMRI will assist clinicians to consider ectopic ACTH syndrome, and then perform imaging to identify ectopic tumors. Besides, the inter-observer agreement for identifying pituitary microadenomas was almost perfect on hrMRI (κ = 0.91), which was moderate on cMRI (κ = 0.50) and dMRI (κ = 0.57). Therefore, hrMRI using the 3D FSE sequence is a potential alternative that can significantly improve the identification of pituitary microadenomas.

Limitations of the study included its retrospective nature and the relatively small sample size in patients with ectopic ACTH syndrome as negative controls. The bias may be introduced in the patient inclusion process. Only those patients who underwent all the cMRI, dMRI, and hrMRI scans were included. In fact, some patients will bypass hrMRI when obvious pituitary adenomas were detected on cMRI and dMRI. These patients were not included in the current study because of lack of hrMRI findings. Given the situation, the sensitivity of identifying pituitary adenomas will be higher with the enrollment of these patients. Besides, the timing of the sequence acquisition after contrast injection is essential [16] and bias may be introduced due to the postcontrast enhancement curve of both the pituitary gland and the microadenoma [14]. In the future, a prospective study with different sequence acquisition orders is needed to minimize possible interference caused by the postcontrast enhancement curve. Moreover, a larger sample size is also needed to verify the diagnostic performance of hrMRI using 3D FSE sequence for identifying pituitary microadenomas and to determine whether it can replace 2D FSE or 3D SPGR sequences for routinely evaluating the pituitary gland.

In conclusion, hrMRI with 3D FSE sequence showed higher diagnostic performance than cMRI and dMRI for identifying pituitary microadenomas in patients with Cushing’s syndrome.

Abbreviations

ACTH:
Adrenocorticotropic hormone
AUC:
Area under the receiver operating characteristics curve
cMRI:
Conventional contrast-enhanced MRI
CNR:
Contrast-to-noise ratio
dMRI:
Dynamic contrast-enhanced MRI
FSE:
Fast spin echo
hrMRI:
High-resolution contrast-enhanced MRI
IQR:
Interquartile range
SNR:
Signal-to-noise ratio
SPGR:
Spoiled gradient re

called

References

  1. Lacroix A, Feelders RA, Stratakis CA, Nieman LK (2015) Cushing’s syndrome. Lancet 386:913–927

    Article CAS PubMed Google Scholar

  2. Loriaux DL (2017) Diagnosis and differential diagnosis of Cushing’s syndrome. N Engl J Med 376:1451–1459

    Article CAS PubMed Google Scholar

  3. Nieman LK, Biller BM, Findling JW et al (2015) Treatment of Cushing’s syndrome: an Endocrine Society clinical practice guideline. J Clin Endocrinol Metab 100:2807–2831

    Article CAS PubMed PubMed Central Google Scholar

  4. Yamada S, Fukuhara N, Nishioka H et al (2012) Surgical management and outcomes in patients with Cushing disease with negative pituitary magnetic resonance imaging. World Neurosurg 77:525–532

    Article PubMed Google Scholar

  5. Vitale G, Tortora F, Baldelli R et al (2017) Pituitary magnetic resonance imaging in Cushing’s disease. Endocrine 55:691–696

    Article CAS PubMed Google Scholar

  6. Jagannathan J, Smith R, DeVroom HL et al (2009) Outcome of using the histological pseudocapsule as a surgical capsule in Cushing disease. J Neurosurg 111:531–539

    Article PubMed PubMed Central Google Scholar

  7. Boscaro M, Arnaldi G (2009) Approach to the patient with possible Cushing’s syndrome. J Clin Endocrinol Metab 94:3121–3131

    Article CAS PubMed Google Scholar

  8. Kasaliwal R, Sankhe SS, Lila AR et al (2013) Volume interpolated 3D-spoiled gradient echo sequence is better than dynamic contrast spin echo sequence for MRI detection of corticotropin secreting pituitary microadenomas. Clin Endocrinol (Oxf) 78:825–830

    Article CAS PubMed Google Scholar

  9. Lonser RR, Nieman L, Oldfield EH (2017) Cushing’s disease: pathobiology, diagnosis, and management. J Neurosurg 126:404–417

    Article PubMed Google Scholar

  10. Potts MB, Shah JK, Molinaro AM et al (2014) Cavernous and inferior petrosal sinus sampling and dynamic magnetic resonance imaging in the preoperative evaluation of Cushing’s disease. J Neurooncol 116:593–600

    Article PubMed Google Scholar

  11. Grober Y, Grober H, Wintermark M, Jane JA, Oldfield EH (2018) Comparison of MRI techniques for detecting microadenomas in Cushing’s disease. J Neurosurg 128:1051–1057

    Article PubMed Google Scholar

  12. Fukuhara N, Inoshita N, Yamaguchi-Okada M et al (2019) Outcomes of three-Tesla magnetic resonance imaging for the identification of pituitary adenoma in patients with Cushing’s disease. Endocr J 66:259–264

    Article PubMed Google Scholar

  13. Patronas N, Bulakbasi N, Stratakis CA et al (2003) Spoiled gradient recalled acquisition in the steady state technique is superior to conventional postcontrast spin echo technique for magnetic resonance imaging detection of adrenocorticotropin-secreting pituitary tumors. J Clin Endocrinol Metab 88:1565–1569

    Article CAS PubMed Google Scholar

  14. Magnaldi S, Frezza F, Longo R, Ukmar M, Razavi IS, Pozzi-Mucelli RS (1997) Assessment of pituitary microadenomas: comparison between 2D and 3D MR sequences. Magn Reson Imaging 15:21–27

    Article CAS PubMed Google Scholar

  15. Lien RJ, Corcuera-Solano I, Pawha PS, Naidich TP, Tanenbaum LN (2015) Three-Tesla imaging of the pituitary and parasellar region: T1-weighted 3-dimensional fast spin echo cube outperforms conventional 2-dimensional magnetic resonance imaging. J Comput Assist Tomogr 39:329–333

    PubMed Google Scholar

  16. Sartoretti T, Sartoretti E, Wyss M et al (2019) Compressed SENSE accelerated 3D T1w black blood turbo spin echo versus 2D T1w turbo spin echo sequence in pituitary magnetic resonance imaging. Eur J Radiol 120:108667

    Article PubMed Google Scholar

  17. Kim M, Kim HS, Kim HJ et al (2021) Thin-slice pituitary MRI with deep learning-based reconstruction: diagnostic performance in a postoperative setting. Radiology 298:114–122

    Article PubMed Google Scholar

Download references

Acknowledgements

We thank Dr. Kai Sun, Medical Research Center, Peking Union Medical College Hospital, for his guidance on the statistical analysis in this study.

Funding

This study has received funding from the National Natural Science Foundation of China (grant 82071899), the National Key Research and Development Program of China (grants 2016YFC1305901, 2020YFA0804500), the Chinese Academy of Medical Sciences Innovation Fund for Medical Sciences (grants 2017-I2M-3–008, 2021-I2M-1–025), the Beijing Natural Science Foundation (grant L182067) and National High Level Hospital Clinical Research Funding (2022-PUMCH-B-067, 2022-PUMCH-B-114).

Author information

Author notes

  1. Zeyu Liu and Bo Hou contributed equally to this work and share first authorship
  2. Hui You and Feng Feng contributed equally to this work and share corresponding authorship

Authors and Affiliations

  1. Department of Radiology, Peking Union Medical College Hospital, Chinese Academy of Medical Sciences and Peking Union Medical College, No. 1 Shuaifuyuan Wangfujing Dongcheng Distinct, Beijing, 100730, China

    Zeyu Liu, Bo Hou, Hui You, Mingli Li & Feng Feng

  2. Department of Endocrinology, Peking Union Medical College Hospital, Chinese Academy of Medical Sciences and Peking Union Medical College, No. 1 Shuaifuyuan Wangfujing Dongcheng Distinct, Beijing, 100730, China

    Lin Lu, Lian Duan & Huijuan Zhu

  3. Department of Neurosurgery, Peking Union Medical College Hospital, Chinese Academy of Medical Sciences and Peking Union Medical College, No. 1 Shuaifuyuan Wangfujing Dongcheng Distinct, Beijing, 100730, China

    Kan Deng & Yong Yao

  4. State Key Laboratory of Complex Severe and Rare Disease, Peking Union Medical College Hospital, Chinese Academy of Medical Sciences and Peking Union Medical College, No. 1 Shuaifuyuan Wangfujing Dongcheng Distinct, Beijing, 100730, China

    Yong Yao, Huijuan Zhu & Feng Feng

Corresponding authors

Correspondence to Hui You or Feng Feng.

Ethics declarations

Guarantor

The scientific guarantor of this publication is Feng Feng.

Conflict of interest

The authors of this manuscript declare no conflict of interest.

Statistics and biometry

No complex statistical methods were necessary for this paper.

Informed consent

Written informed consent was waived by the Institutional Review Board.

Ethical approval

Institutional Review Board approval was obtained.

Methodology

• retrospective

• diagnostic or prognostic study

• performed at one institution

Additional information

Publisher’s note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary Information

Below is the link to the electronic supplementary material.

Filed under: Cushing's, Diagnostic Testing, pituitary | Tagged: , , , | Leave a comment »

Happy 23rd Birthday!

Posted on July 21, 2023 by MaryO

happybirthday-2015

It’s unbelievable but the idea for Cushing’s Help and Support arrived 23 years ago last night.  That’s a long time for anything online.

I was talking with my dear friend Alice, who ran a wonderful menopause site called Power Surge, wondering why there weren’t many support groups online (OR off!) for Cushing’s and I wondered if I could start one myself and we decided that I could.

The first website (http://www.cushings-help.com) first went “live” July 21, 2000 and the message boards September 30, 2000. Hopefully, with these sites, I’m making some helpful differences in someone else’s life!

The message boards are very active and we have weekly online text chats, occasional live interviews, local meetings, email newsletters, a clothing exchange, a Cushing’s Awareness Day Forum, podcasts, phone support and much more.

Whenever one of the members of the boards gets into NIH, I try to go to visit them there. Other board members participate in the “Cushie Helper” program where they support others with one-on-one support, doctor/hospital visits, transportation issues and more.

Of course, we now have a Facebook page and 2 groups.  Both are secret, so if you want to join, please email  or PM me for an invitation.

Other sites in the Cushing’s Help “Family”

maryo colorful zebra

Filed under: Cushing's | Tagged: , , , , , , , , | 5 Comments »

Cushing Syndrome – A Review

Posted on July 18, 2023 by MaryO
Abstract

Importance  Cushing syndrome is defined as a prolonged increase in plasma cortisol levels that is not due to a physiological etiology. Although the most frequent cause of Cushing syndrome is exogenous steroid use, the estimated incidence of Cushing syndrome due to endogenous overproduction of cortisol ranges from 2 to 8 per million people annually. Cushing syndrome is associated with hyperglycemia, protein catabolism, immunosuppression, hypertension, weight gain, neurocognitive changes, and mood disorders.

Observations  Cushing syndrome characteristically presents with skin changes such as facial plethora, easy bruising, and purple striae and with metabolic manifestations such as hyperglycemia, hypertension, and excess fat deposition in the face, back of the neck, and visceral organs. Cushing disease, in which corticotropin excess is produced by a benign pituitary tumor, occurs in approximately 60% to 70% of patients with Cushing syndrome due to endogenous cortisol production. Evaluation of patients with possible Cushing syndrome begins with ruling out exogenous steroid use. Screening for elevated cortisol is performed with a 24-hour urinary free cortisol test or late-night salivary cortisol test or by evaluating whether cortisol is suppressed the morning after an evening dexamethasone dose. Plasma corticotropin levels can help distinguish between adrenal causes of hypercortisolism (suppressed corticotropin) and corticotropin-dependent forms of hypercortisolism (midnormal to elevated corticotropin levels). Pituitary magnetic resonance imaging, bilateral inferior petrosal sinus sampling, and adrenal or whole-body imaging can help identify tumor sources of hypercortisolism. Management of Cushing syndrome begins with surgery to remove the source of excess endogenous cortisol production followed by medication that includes adrenal steroidogenesis inhibitors, pituitary-targeted drugs, or glucocorticoid receptor blockers. For patients not responsive to surgery and medication, radiation therapy and bilateral adrenalectomy may be appropriate.

Conclusions and Relevance  The incidence of Cushing syndrome due to endogenous overproduction of cortisol is 2 to 8 people per million annually. First-line therapy for Cushing syndrome due to endogenous overproduction of cortisol is surgery to remove the causative tumor. Many patients will require additional treatment with medications, radiation, or bilateral adrenalectomy.

Filed under: Cushing's | Tagged: , , , | Leave a comment »

Minimizing the Number of False Positives in Dexamethasone Suppression Testing for the Diagnosis of Cushing’s Syndrome

Posted on July 12, 2023 by MaryO

In this application note, Tecan presents a method for diagnosing Cushing’s syndrome efficiently and accurately. The approach involves simultaneous the measurement of cortisol and dexamethasone levels using LC-MS/MS, which reduces false positives in dexamethasone suppression test (DSTs). The described LC-MS/MS method enables the tracking of multiple analytes, including cortisol, cortisone, and dexamethasone, in serum or plasma. Implementing this analytical approach offers clinical laboratories a straightforward means of performing DSTs, and the availability of a commercially available kit ensures reliable and reproducible results.

 

Filed under: Diagnostic Testing | Tagged: | Leave a comment »

Next Page »