Title:

The prevalence of structural pituitary abnormalities by MRI scanning in men presenting with isolated hypogonadotrophic hypogonadism

Short title:

MRI abnormalities in secondary hypogonadism

Authors:

Dr Mazhar Dalvi

Prof Brian R Walker

Prof Mark W J Strachan

Dr Nicola N Zammitt

Dr Fraser W Gibb

Department:

Edinburgh Centre for Endocrinology and Diabetes

Correspondence:

Dr Fraser W Gibb

Edinburgh Centre for Endocrinology and Diabetes

Royal Infirmary of Edinburgh

Edinburgh

EH16 4SA

United Kingdom Email: [email protected]

Keywords: Hypogonadism, pituitary, testosterone, Magnetic resonance imaging

Disclosure: nothing to declare

Word count: 1577 ABSTRACT

Objective:

Hypogonadotrophic hypogonadism (HH) is commonly associated with ageing, obesity and type 2 diabetes. The indications for pituitary imaging are controversial and current guidelines are based on small case series.

Design:

Retrospective case series from a secondary/tertiary Endocrinology referral centre.

Patients:

All men presenting to the Edinburgh Centre for Endocrinology and Diabetes with hypogonadotrophic hypogonadism (testosterone < 10 nmol/L and normal prolactin) from 2006 – 2013 in whom pituitary MRI was performed (n = 281). All

HH patients referred in 2011 (n=86) were reviewed to assess differences between those selected for pituitary MRI and those who were not scanned.

Results:

Pituitary MRI was normal in 235 men (83.6%), with 24 microadenomas (8.5%),

5 macroadenomas (1.8%) and 1 craniopharyngioma (0.4%) identified. The remaining 16 (5.7%) comprised a range of minor pituitary abnormalities including small cysts and empty sella. All men with abnormal imaging studies had otherwise normal pituitary function. Imaging abnormalities were associated with a significantly lower age at presentation (50 vs. 54 years, p = 0.02) but no differences in testosterone or gonadotrophin levels were observed. Current

Endocrine Society guidelines would have prompted imaging in only 3 of 6 patients with significant pituitary pathology.

Conclusions:

Structural pituitary disease is more common in isolated HH than in the general population and current guidelines do not accurately identify ‘at risk’ individuals.

Full anterior pituitary function testing has a low yield in patients presenting with hypogonadism. The optimal strategy for determining the need for pituitary imaging remains uncertain.

Introduction

Hypogonadotrophic hypogonadism (HH) is a common reason for referral to specialist endocrine services and is strongly associated with type 2 diabetes

(T2DM)1, obesity2 and ageing3. Significant controversy exists with respect to the appropriate investigation of this condition, particularly when deciding which patients require pituitary imaging and detailed assessment of anterior pituitary function. Endocrine Society guidelines4 suggest pituitary imaging and anterior pituitary function testing are reserved for men presenting with total testosterone concentrations less than 5.2 nmol/L (150 ng/dL) or with additional concerning clinical features such as headache or visual disturbance. This recommendation is based on a relatively small case series (n=164) from urological practice5, which is not necessarily representative of those patients referred to endocrine services.

We aimed to assess the diagnostic yield of pituitary imaging and anterior pituitary function testing in men presenting with HH. We also aimed to characterise the presenting clinical and biochemical features and how these related to imaging abnormalities in this population.

Methods

This was a retrospective case series comprising all men presenting to the

Edinburgh Centre for Endocrinology and Diabetes (a secondary and tertiary referral centre) for the assessment of HH between 2006 and 2013, in whom a pituitary MRI was performed (n = 281). HH was defined as a total testosterone concentration <10 nmol/L and a luteinizing hormone (LH) concentration <10 mU/L. Patients with an elevated serum prolactin at presentation were excluded.

Patients were identified from our comprehensive clinical database and a review of electronic pituitary MRI requests between 2006 and 2013. In addition, information was collected for all HH patients presenting in 2011, to assess differences between men receiving MRI scans (n = 44) and those who were not imaged (n = 42). Across the period studied, the decision to refer for pituitary MRI was largely at the discretion of the consultant endocrinologist, practice was variable and Endocrine Society guidelines were not typically adhered to. MRI scans were reported by consultant neuroradiologists affiliated with our centre.

Anterior pituitary function testing was requested in the majority of patients, including measurement of LH, follicle stimulating hormone (FSH), total testosterone (TT), sex hormone-binding globulin (SHBG), prolactin, ACTH- stimulated cortisol (measured at any time of day, 30 minutes after the intramuscular administration of 250 micrograms Synacthen®), thyroid stimulating hormone (TSH), free thyroxine (fT4) and, in a minority, insulin-like growth factor (IGF-1). All assays were immunometric with the exception of total testosterone, which was measured before 2010 by immunometric assay and since 2010 by liquid chromatography – mass spectrometry (LC-MS); the same reference range was applied to interpretation of both assays.

Data are presented as median (inter-quartile range). Between group comparisons were analysed by Independent-Samples Mann-Whitney U test. A p value of < 0.05 was considered statistically significant.

Results

The median age of men in whom a pituitary MRI was performed was 53 years

(44 – 60). Median weight and BMI were 84.2 kg (69.0 – 81.5) and 30.0 kg/m2

(27.0 – 33.0), respectively. The median total testosterone was 6.2 nmol/L (5.0 –

7.5) and LH was 2.6 U/L (1.6 – 4.0). There was no significant difference in age (53 years [47 – 60]), BMI (29.5 kg/m2 [25.0 – 37.7]) or LH (2.8 U/L [1.8 – 3.8]) in the cohort of men who were not referred for pituitary imaging, however, total testosterone was a median of 2.1 nmol/L higher in this group (p <0.0001). The full biochemical characteristics of men in whom pituitary imaging was performed is summarized in table 1. Only 81 (28.8%) men had a morning total testosterone lower than the Endocrine Society threshold of 5.2 nmol/L. Seven patients had a cortisol, after exogenous ACTH, below the reference range, all of whom had normal pituitary imaging. All but one were being treated with opioid analgesia and, in 2 cases, repeat testing was normal.

84% of men had normal pituitary imaging with the remainder reported as having a spectrum of abnormalities (figure 1). 6 men were found to have structurally significant pituitary disease (including non-functioning macroadenomas and craniopharyngioma); their clinical and biochemical features are summarized below:

Patient 1: A 44 year-old man (BMI 25 kg/m2) with a morning TT of 7.7 nmol/L

(CFT 164 pmol/L) and LH 5 U/L. All other anterior pituitary function was normal (details of anterior pituitary results for all patients are provided in supplementary table 1). Imaging revealed a 12 x 8 x 7mm intrasellar adenoma.

To date, this has been managed conservatively with interval imaging.

Patient 2: A 53 year-old man (BMI 25 kg/m2) with a nadir morning TT of 6.4 nmol/L (CFT 137 pmol/L) and LH 1.5 U/L. All other anterior pituitary function was normal, although TSH and free thyroxine were borderline low at 0.26 mU/L and 10 pmol/L respectively. Imaging demonstrated a 25 x 22 x 20mm sellar mass extending into the right cavernous sinus and superiorly displacing the optic chiasm. One week following initial assessment, he was admitted to hospital with pituitary apoplexy. He was managed conservatively and remains under regular imaging follow-up.

Patient 3: A 31 year old man (BMI 36 kg/m2) with a morning TT of 0.5 nmol/L

(CFT 14 pmol/L) and LH 0.7 U/L but otherwise normal anterior pituitary function. Imaging revealed a 37 x 25 x 30mm pituitary lesion with right cavernous sinus invasion and optic chiasm displacement. Trans-sphenoidal resection (null cell adenoma) was performed, with subsequent pituitary radiotherapy.

Patient 4: A 54 year-old man (BMI 32 kg/m2) with a nadir TT of 7.9 nmol/L at presentation (CFT 120 pmol/L) and LH of 1.8 U/L. Anterior pituitary function was otherwise normal. Imaging revealed a 13 x 12 x 12mm pituitary lesion distorting the stalk and immediately inferior to the optic chiasm. Trans- sphenoidal resection (null cell adenoma) was performed without complication.

Patient 5: A 47 year-old man (BMI 32 kg/m2) with a short history of hypogonadal symptoms and an undetectably low total testosterone (<0.3 nmol/L), with undetectable gonadotrophins. Prolactin, short synacthen test and

IGF-1 were all normal although TSH was low (0.02 mU/L) in the context of normal fT4 (13 pmol/L). Imaging revealed a 35 x 25 x 26 suprasellar mass, which required resection by craniotomy (craniopharyngioma), resulting in post- operative panhypopituitarism and diabetes insipidus.

Patient 6: A 54 year-old type 2 diabetic man (BMI 26 kg/m2) with a TT of 3.4 nmol/L and LH of 1.2 U/L. Anterior pituitary function was otherwise normal. Imaging demonstrated a 14 x 9 x 9 mm pituitary adenoma, which has been managed conservatively with imaging follow-up.

Patients 1, 2 and 4 would not have been recommended for pituitary imaging, at presentation, based on current Endocrine Society guidelines.

Age at presentation was significantly lower in men with a reported abnormality on pituitary MRI, although no other clinical parameters were associated with imaging abnormalities (table 2).

Discussion

In this series, pituitary pathology was present in a relatively low proportion of men presenting with hypogonadotrophic hypogonadism, however this included a small number of cases requiring surgical intervention, sometimes in the absence of ‘red flag’ clinical or biochemical features. This cohort of men presenting with isolated hypogonadotrophic hypogonadism are likely to be typical of those referred to endocrine services across the United Kingdom. It is likely that such referrals will increase in the context of an ageing population, with rising rates of obesity and T2DM. Hypogonadism is present in between

17% 6 and 33%1 of men with type 2 diabetes mellitus, which has prompted the

Endocrine Society recommendation to screen for testosterone deficiency in this group4. Testosterone deficiency is also present in 40% of non-diabetic obese men over the age of 452. The vast majority of such cases will not be a consequence of structural pituitary disease but rather result from hypothalamic- pituitary dysfunction, perhaps related to a pro-inflammatory state and adiposity- associated total body aromatase excess7.

Identifying the small minority of patients with significant pituitary pathology presenting with isolated hypogonadism is a significant clinical challenge. A policy of performing pituitary imaging in all patients is neither clinically justifiable nor likely to be cost effective. A recent international survey of endocrinologists confirmed marked variation in practice in relation to a testosterone threshold for imaging8. Current Endocrine Society guidelines recommend pituitary function tests and imaging when high-risk symptoms are present (headache and visual disturbance) or when the morning total testosterone is less than 5.2 nmol/L4. The imaging recommendation is based on a series of 164 men presenting with erectile dysfunction (age 27 – 79 years) with persistently low total testosterone (< 8 nmol/L). 6 men (3.7%) had significant pituitary pathology and all presented with total testosterone levels less than 3.6 nmol/L5. Our series had a lower prevalence of significant pituitary pathology

(2.2%) than the previously published case series on which Endocrine Society guidelines are based, perhaps reflecting the higher testosterone threshold at which imaging was performed. However, three of the six cases were associated with total testosterone significantly greater than the 5.2 nmol/L threshold (6.4,

7.7 and 7.9 nmol/L) and would not have been imaged at presentation if

Endocrine Society guidelines had been adhered to. The prevalence of pituitary macroadenoma does appear to be greater in hypogonadal men than the reported

‘incidentaloma’ rate of between 0.16 and 0.2%9. Across a range of clinical and biochemical features, only younger age at presentation was actually predictive of a pituitary imaging abnormality. Whilst our data may ostensibly suggest more liberal recourse to pituitary imaging in men with HH, it is impossible to quantify whether the small number of extra cases detected would justify the expense and inconvenience of additional imaging. Establishing a large multi-centre data collection network for HH would be of value in developing an evidence base, to permit a more judicious approach to the investigation and management of this condition.

References

1. Dhindsa S, Prabhakar S, & Sethi M et al. (2004) Frequent Occurrence of

Hypogonadotropic Hypogonadism in Type 2 Diabetes. Journal of Clinical

Endocrinology & Metabolism, 89(11), 5462-5468.

2. Dhindsa S, Miller MG & McWhirter et al. (2010) Testosterone

concentrations in diabetic and nondiabetic obese men. Diabetes care,

33(6), 1186-1192.

3. Wu FC, Tajar A & Pye SR et al. (2008) Hypothalamic-pituitary-testicular

axis disruptions in older men are differentially linked to age and

modifiable risk factors: the European Male Aging Study. The Journal of

clinical endocrinology and metabolism. 93(7), 2737-2745 4. Bhasin S, Cunningham GR & Hayes FJ et al. (2010) Testosterone Therapy

in Men with Androgen Deficiency Syndromes: An Endocrine Society

Clinical Practice Guideline. The Journal of Clinical Endocrinology &

Metabolism. 95(6), 2536-2559.

5. Citron JT, Ettinger B & Rubinoff H et al. (1996) Prevalence of

hypothalamic-pituitary imaging abnormalities in impotent men with

secondary hypogonadism. The Journal of urology. 155(2), 529-533.

6. Kapoor D, Aldred H & Clark S et al. (2007) Clinical and Biochemical

Assessment of Hypogonadism in Men With Type 2 Diabetes. Diabetes

Care. 30(4), 911-917.

7. Gibb FW & Strachan MWJ. (2014) Androgen deficiency and type 2

diabetes mellitus. Clinical Biochemistry. 47, 940 – 949.

8. Grossmann M, Anawalt BD & Wu FCW. (2015) Clinical practice patterns in

the assessment and management of low testosterone in men: an

international survey of endocrinologists. Clinical Endocrinology.

82(2),234-241.

9. Ezzat S, Asa SL & Couldwell WT et al. (2004) The prevalence of pituitary

adenomas: a systematic review. Cancer. 101(3), 613-619. Figures:

Table 1: Clinical characteristics of men who underwent pituitary MRI following diagnosis of hypogonadotrophic hypogonadism. *IGF-1 range is age-specific for men aged 40 – 54.

Median (IQR) Reference range N Age (years) 53 (44 – 60) 281 Duration of symptoms 18 (10 – 36) 279 (months) Weight (kg) 84.2 (69.0 – 81.5) 281 Body mass index (kg/m2) 30.0 (27.0 – 33.0) 281 Total testosterone (nmol/L) 6.2 (5.0 – 7.5) 10.0 – 30.0 281 Calculated free testosterone 162 (127 – 196) 245 – 785 252 (pmol/L) Estradiol (pmol/L) 61 (40 – 77) <160 70 LH (U/L) 2.6 (1.6 – 4.0) 1.0 – 9.0 281 FSH (U/L) 3.7 (2.5 – 5.9) 1.0 – 10.0 281 SHBG (nmol/L) 21 (15 – 27) 6 – 45 242 Prolactin (mU/L) 170 (131 – 233) 60 – 500 276 ACTH-stimulated cortisol 623 (540 – 702) >430 248 (nmol/L) Free thyroxine (pmol/L) 13 (12 – 14) 9 – 21 274 TSH (mU/L) 1.5 (1 – 2) 0.2 – 4.5 274 IGF-1 (μg/L) 108 (80 – 145) 63 – 201* 86 Figure 1: Outcome of pituitary MRI in 281 men presenting with hypogonadotrophic hypogonadism. Table 2: Comparison of clinical and biochemical features in men with normal and abnormal (including macroadenoma, craniopharyngioma, microadenoma, small cysts and empty sella) pituitary MRI. Data are median (IQR) compared by Independent-samples Mann-Whitney U Test.

Normal MRI (n=237) Abnormal MRI P (n=44) Weight (kg) 78.0 (69.0 – 91.0) 81.8 (68.0 – 93.0) 0.883 BMI (kg/m2) 30.0 (27.0 – 33.0) 29.5 (26.0 – 35.9) 0.693 Age (years) 54.0 (45.0 – 61.0) 50.0 (40.0 – 54.0) 0.021 Duration of 18 (10 – 36) 12 (11 – 30) 0.454 symptoms (months) Total testosterone 6.3 (5.0 – 7.5) 5.7 (3.9 – 7.6) 0.167 (nmol/L) LH (U/L) 2.7 (1.6 – 4.0) 2.2 (1.2 – 3.5) 0.149 FSH (U/L) 3.8 (2.6 – 6.3) 3.1 (2.3 – 4.7) 0.074 Prolactin (mU/L) 171 (131 – 229) 166 (136 – 270) 0.578

Supplementary materials

Supplementary table 1: Full anterior pituitary function test results for men with large pituitary tumours (Reference ranges are reported in table 1).

Patient FSH Prolactin Basal Stimulated Free T4 TSH IGF-1 (U/L) (mU/L) cortisol cortisol (pmol/L (mU/L) (μg/L) (nmol/L) (nmol/L) ) 1 2.4 342 342 657 15 0.97 75 2 2.4 256 251 685 10 0.29 NA 3 2.6 448 269 596 10 2.6 74 4 3.8 280 146 724 13 0.64 NA 5 <0.5 163 213 551 13 0.20 63 6 2 302 226 654 13 0.66 NA