Comparison with other studies
Mechanical compression and damage of the adenohypophysis by the enlarging NFPMS can lead to endocrine dysfunction field [
18]. In addition, reduction of the pituitary portal circulation due to mass effect has been speculated to cause hypopituitarism [
19]. A substantial number of our patients had evidence of pituitary hormone deficiency on laboratory assessment at presentation despite being asymptomatic. Hypopituitarism often develops insidiously and remains undiagnosed until the patient undergoes full clinical and endocrine evaluation field [
18]. Interestingly in this study, male gender, older age, and NFPMs extending beyond the sella turcica were associated with a higher rate of anterior panhypopituitarism at presentation. Jahangiri et al. [
20] reported a correlation between any degree of preoperative hypopituitarism and old age, gender and large pituitary adenomas but not linked to complete hypopituitarism. Zhang et al. [
6] recorded significantly lower preoperative levels of thyroxine, GH, IGF-1, FSH, and LH in patients with giant NFPAs compared to those with macroadenoma, demonstrating the impact of larger NFPAs on pituitary function. Our cohort with reported partial and complete hypopituitarism at diagnosis is in line with other studies in the literature [
6,
8,
18]. The agreed sequential pattern of developing pituitary hormone deficiency in patients with pituitary adenomas at presentation is often debated. The classical description of losing GH axis followed by gonadotroph, thyrotroph and adrenocorticotroph impairment is generally agreed on [
21‐
24]; however, many studies demonstrate a different pattern of pituitary hormone insufficiencies [
6,
25,
26]. Carosi et al. [
25] showed that central hypogonadism was more frequently observed in NFPMs, followed by GH, ACTH and TSH deficiencies. A similar pattern of pituitary impairment was also demonstrated by Ferrante et al. [
26].
The definitive diagnosis of GH insufficiency at diagnosis of pituitary adenomas requires provocative testing, which is often performed in selected patients only when there is an intention to treat; this may lead to underdiagnosis of the condition. Yuen et al. [
27] demonstrated that 50% of their 38 patients with non-functioning pituitary microadenomas and normal IGF1 levels had GH insufficiency using growth hormone releasing hormone-arginine test. In contrast, adrenal insufficiency is investigated extensively, at presentation and after therapy, to establish the presence of hypocortisolism and to commence glucocorticoid therapy, when appropriate, to prevent adrenal crises. This study has a higher incidence of central adrenal insufficiency than reported in the literature. Our hospital is a large tertiary centre that receives a high volume of referrals from the local district hospitals. Patients who were referred from these hospitals while on glucocorticoid replacement were considered to have adrenal insufficiency. This could be one of the potential causes. The lower frequency of GH deficiency at presentation in this study may not reflect the overall incidence of the condition, as GH deficiency diagnosis requires dynamic tests, as discussed above, these may only be performed for selected cases before surgery.
Perioperative glucocorticoid replacement was traditionally used in all patients undergoing surgery for suprasellar and extrasellar tumours to prevent adrenal crisis, but this approach carries the risk of potential exposure to steroid side effects, including hyperglycaemia, peptic ulcers, sleep disturbance and osteoporosis. In addition, assessing adrenal function following surgery in patients treated empirically with exogenous steroids might be challenging and difficult to interpret. Many studies demonstrated the safety of performing surgical resection without glucocorticoid cover in those with normal hypothalamic–pituitary–adrenal axis before surgery [
28,
29]. In our centre, patients with confirmed ACTH deficiency before surgery were commenced on glucocorticoid therapy and received parenteral hydrocortisone 50–100 mg 6 h perioperatively. In addition, those patients received supraphysiological or stress doses of glucocorticoid therapy in the immediate postoperative phase. Two of the three neurosurgeons did not commence perioperative glucocorticoids in patients with normal adrenal function, while one surgeon’s approach was starting steroid cover in all patients perioperatively. 0800 a.m. cortisol level was assessed 48 h post-surgery in all patients after withdrawing steroid treatment for at least 18 h when applicable. Morning cortisol less than 350 nmol/L reflected central hypocortisolism in our centre and was managed with glucocorticoid replacement until definitive dynamic testing was performed 6–8 weeks following surgery.
The occurrence of new pituitary dysfunction post-therapy can vary according to treatment modality. Postoperatively, the likelihood of developing pituitary dysfunction has been linked to many factors: the operating neurosurgeon, pituitary tumour size, the degree of surgical manipulation, and the need for multiple surgeries to deal with recurrent disease [
30]. Similarly, the occurrence of new-onset pituitary insufficiency post-radiotherapy depends on many factors; radiation dose, technique and follow-up duration [
31,
32]. It has been estimated that around 30–60% of the patients may develop endocrine dysfunction after pituitary external beam irradiation [
32]. We demonstrated that the combination of surgery and radiotherapy is associated with a significantly higher risk of pituitary dysfunction than surgery alone, shorter hormone deficiency-free survival and less frequent hormone recovery over the long term. The pathophysiology of radiation-induced hypopituitarism is complex and not very well understood. Several mechanisms have been proposed in the literature to elucidate the relation between irradiation and hypothalamic–pituitary dysfunction; this includes thalamic vascular damage with subsequent pituitary atrophy, microstructural change and axonal loss of the hypothalamus, and alterations of hypothalamic neurotransmitters with subsequent endocrine and metabolic disturbance [
33‐
35]. In our centre, all patients are discussed in the pituitary multidisciplinary meeting to assess the need for surgery with careful consideration of the use of radiotherapy. We currently use proton beam therapy to treat children, teenagers and young adults up to their 25th birthday to reduce the late effects of radiation, according to the National Health Service (NHS) commissioning criteria. Recovery of pituitary dysfunction following treatment remains uncertain, with no apparent predictive clinical and radiological features. In this study, younger patients were more likely to regain normal endocrine function. Gender and postoperative evidence of residual tumour did not correlate with an overall recovery of pituitary hypofunction. This contrasts with the findings of Webb et al. [
36], who reported better improvement in pituitary function in those with no tumour remanent. Of note, Little et al. [
37] did not identify any predictors for regaining normal endocrine function. Berg et al. [
38] assessed the recovery of GH and ACTH axes in 36 patients with pituitary disease after surgery using ITT and demonstrated a significant increase in GH and cortisol levels at 12 months following resection, with the recovery of both axes in 11% of those with GH deficiency and secondary adrenal insufficiency. Therefore, it is of crucial importance to provide a long-term endocrine follow-up for patients with pituitary disease following therapy to assess for a new hormone deficiency or recovery and to avoid unnecessary hormone replacement therapy. In our centre, we routinely perform full baseline and dynamic endocrine testing for GH and ACTH axes 6–8 weeks after surgery and on annual intervals when clinically indicated during follow-up after surgery and radiotherapy.