Abstract
Objective
To evaluate clinical criteria for headache associated with pituitary adenoma (HaPA) in the International Classification of Headache Disorders (ICHD) 3rd edition version criteria and further determine whether elevations of plasma calcitonin gene-related peptide and pituitary adenylate cyclase-activating peptide 1-38 (PACAP1-38) concentration contribute to HaPA.
Methods
Demographic and clinical features of consecutive patients with pituitary adenoma were recorded. Plasma calcitonin gene-related peptide and PACAP1-38 concentrations in pituitary adenoma patients within 72 h pre- and post-operation were measured. Primary outcome for HaPA patients were 50% reduction of moderate-to-severe headache days at 3 months after discharge.
Results
Sixty-three patients with pituitary adenoma were recruited, 33 (52.4%) of whom had headache. The patients who had HaPA presented with migraine-like (32.9%), tension-type-like (12.1%), and stabbing headache (9.1%). Non-functional adenoma was present in the majority of cases (82.5%). Surgical resection improved headache in 83.3% of cases at 3 month follow-up. Pre- and post-operative calcitonin gene-related peptide and PACAP1-38 levels were significantly higher in patients with headache than in those without headache (p < 0.05). Plasma calcitonin gene-related peptide and PACAP1-38 levels at 72 h post-operation were lower at 72 h after operation in patients who had greater improvement in headache compared with those who had little improvement, while plasma calcitonin gene-related peptide and PACAP1-38 levels were similar between these two groups preoperatively.
Conclusions
Most pituitary adenoma patients have non-functional adenoma, and half of this group have HaPA, indicating that the ICHD-3 criteria for HaPA with the emphasis on secretion status need further modifications. Lower plasma calcitonin gene-related peptide and PACAP1-38 concentrations at 72 h after operation may predict a better outcome in patients with HaPA.
Introduction
Pituitary adenoma (PA) is one kind of common primary intracranial neoplasm, and the presence rate of headache in such cases is reported to be 33–72% (1–4). Clinicians are sometimes unable to determine the specific relationship between headache and PA, which makes it difficult to decide upon an appropriate treatment plan. Previous studies had shown that patients with a small PA, or a non-functional or secretory adenoma, often experienced refractory headache that showed significant improvement, or was completely relieved, after surgical resection of the PA; in such cases, postoperative complications were rare (2,5). This implies that the symptoms of headache in patients with PA are closely related to the presence of the adenoma itself. However, the pathological mechanisms underlying headache in patients with PA are still unclear and represent a significant hurdle in developing treatment plans. Conflicting data demonstrated the relationship between pituitary tumour–associated headache and potential mechanisms, including tumour size, cavernous sinus invasion, hormonal hypersecretion, increased intrasellar pressure, and traction or displacement of intracranial pain-sensitive structures (1,6–9). Of note, two headache phenotypes are specified as headache attributed to pituitary disorders in the International Classification of Headache Disorders (ICHD) 3rd edition criteria, 6.9 Headache attributed to pituitary apoplexy and 7.4.3 Headache attributed to hypothalamic or pituitary hyper- or hyposecretion (10). Previous study showed that non-functional adenoma comprised nearly 25% of PA, in which half of this group experienced HaPA (8,11). Criteria B and C, with an emphasis on endocrine status, were insufficient to deal with non-functioning adenoma, so Levy et al. (8) raised a new proposal for Headache attributed to pituitary diseases. However, HaPA criteria in the ICHD-3 remained similar to those in ICHD-3 beta and ICHD-2 due to the lack of evidence (8,12–14). Thus, whether headache is attributed to non-functioning adenoma, especially among Asian patients, remains uninvestigated.
A large number of preclinical and clinical trials have demonstrated that calcitonin gene-related peptide (CGRP) is involved in the transmission of pain signaling. CGRP is the strongest vasodilator of peripheral nerve endings and is closely related to headache. Thus far, CGRP has been the subject of a significant amount of research in the field of primary headache, especially in terms of migraine and cluster headache (13,15,16). Previous studies have demonstrated that the anterior pituitary contains a large number of CGRP immunoreactive fibres, that the pituitary gland cells contain a large number of CGRP receptors, and that CGRP could promote the secretion of growth hormone (GH) and prolactin (PRL) (17,18). Another study found that both pituitary gland cells and adenoma cells contain a large number of pituitary adenylate cyclase-activating peptide 1-38 (PACAP1-38) receptors, type I and type II binding sites, implying that PACAP1-38 could act upon pituitary gland cells (19). PACAP1-38 is also involved in pain conduction, and is known to induce headache (15,20). Therefore, we hypothesised that CGRP and PACAP1-38 might be involved in the pathogenesis of HaPA.
Until now, HaPA has received little research attention in China. Thus, the purpose of this study was to characterise clinical features of headache in Chinese patients with PA and evaluate whether ICHD-3 criteria or Goadsby 2005 proposals are applicable to identify HaPA patients (8). Then, plasma CGRP and PACAP1-38 concentrations were measured to investigate the protentional mechanism of HaPA genesis. Collectively, these data might provide us with an excellent opportunity to explore the pathogenesis of HaPA.
Methods
Study population
This was a prospective study conducted in the neurosurgery inpatients department of the First Affiliated Hospital of Chongqing Medical University in China between May 2015 and January 2016. All patients were seen in the same unit for treatment of newly diagnosed pituitary diseases, including either surgical or medical intervention.
The inclusion criteria were: (a) patients were aged ≥ 18 years; (b) initial suspicion of PA by magnetic resonance imaging (MRI) scans; (c) no pre-operative intracranial hemorrhage, aneurysm, infection, brain trauma or other secondary cause of headache. Exclusion criteria were: (a) patient was pregnant and lactating; (b) patients were unable to clearly describe the symptoms of their headache; and (c) histological diagnosis incompatible with a PA, such as Rathke cyst or craniopharyngioma.
Data collection
Tumour volume, cavernous sinus invasion, and chiasm compression were assessed by a neurosurgeon as described previously (9,11). Secretion status was determined by pre-operative laboratory tests for endocrine activities. Data were analysed prospectively by different investigators in a blinded fashion.
Clinical evaluation of headache in each eligible patient was conducted face to face by a trained headache fellow (YZ, QP and HJ) using a structured questionnaire before commencement of treatment. The questionnaire had been validated for the diagnosis of headache by previous studies (21,22), and included the following components: Demographic characteristics; clinical features of headache (onset age, course, location, duration, frequency, intensity, aggravation after physical activity and associated symptoms); history of head injury; past medical history and drug intake for headache; and family history. Moderate-to-severe headache was defined as Visual Analogue Scale ≥ 5. In each case, headache phenotypes were classified in accordance with ICHD-3 beta criteria, meanwhile ultimately considering the ICHD-3 criteria (10,14).
As described previously, all tumours were categorised as microadenoma (<10 mm) and macroadenoma (>10 mm) according to maximum tumour diameter, using 3 Tesla MRI with coronal and sagittal T1-weighted sequence before and after gadolinium-base contrast medium (8).
CGRP and PACAP1-38 plasma level estimation
Venous blood samples were collected from fasting patients 72 h ± 1 h pre- and post-operation. Patients were not permitted to take any acute analgesics for at least 24 h prior to the collection of blood samples. Blood samples were taken and processed within 1 h, followed by centrifugation at 4℃ for 15 min at 3000 rpm. Plasma was then stored at −80℃. Plasma sample CGRP and PACAP1-38 concentrations were estimated using a human CGRP enzyme-linked immunosorbent assay (ELISA) kit (XY-SJH-1692, Shanghai XY Biotechnology, Shanghai, China) and a human PACAP1-38 ELISA kit (XY-SJH-1202, Shanghai XY Biotechnology, Shanghai, China). The procedure was performed according to the instructions provided by the manufacturer.
Management
The decision to surgically or medically treat the tumour had been made based on the judgement of a surgeon and an endocrinologist. Patients were then followed up by telephone interview at a point 3 months after treatment. Treatment responses for headache were classified as significant effectiveness, partial effectiveness, and no improvement or aggravation. Significant effectiveness was defined as at least 50% reduction in the number of days with a moderate-to-severe headache.
Ethics, consent and permissions
Our study protocol was approved by the hospital ethics board. All patients gave their written informed consent prior to recruitment.
Statistical analysis
Statistical analysis was performed using SPSS software, version 20.0 (Chicago, IL, USA) and GraphPad Prism 6 (GraphPad Software, La Jolla, CA, USA). The characteristics of the study population were analysed using descriptive statistics. Quantitative variables were expressed as mean ± SD. Qualitative variables were expressed as frequency (proportions). Student’s t test and the chi-squared test were used to compare clinical parameters between groups. Two-way ANOVA with Sidak’s post hoc test was used for multiple comparison. Pearson’s correlation was used to assess for correlations between continuous variables. p < 0.05 was considered to be statistically significant.
Results
During the study period (May 2015 to January 2016), 86 patients were initially suspected of having a PA (Figure 1). Twenty-three patients were excluded. Of these 23 patients, two (8.7%) had aneurysms, two (8.7%) had communication difficulties, four (17.4%) refused further examination and treatment, four (17.4%) were found during surgery to have an intrasellar cyst, and 10 (43.5%) had Rathke cysts or craniopharyngioma, as shown by pathological tests. Finally, we recruited 63 patients with PA. Of these 63 patients, 33 (52.4%) had headache and 30 (47.6%) had no headache.
Flow chart of the study protocol.
The clinical features of patients with PA
The clinical features of patients with pituitary adenoma.
ACTH: Adrenocorticotropic hormone; BMI: Body mass index.
Types of headache in patients with pituitary adenoma (according to the ICHD-3 beta and ICHD-3 classification).
ICHD: International Classification of Headache Disorders; NOS: not otherwise specified.
Clinical variables of patients with or without headache upon presentation with pituitary adenoma.
Treatment outcomes
Of the 63 patients, 58 received surgical treatment, while the others received bromocriptine treatment. One patient received somatostatin analogues after surgery. From the original 63 patients, 15 patients (23.8%) were lost to follow-up (surgical treatment, n = 13; bromocriptine, n = 2). Patients were lost because of incorrect telephone numbers (n = 14), or refusal to participate in follow-up (n = 1). Finally, 24 patients with headache, and 24 patients without headache, completed follow-up.
Follow-up results for headache patients.
CGRP and PACAP1-38 estimation
Plasma CGRP levels were significantly higher in patients with HaPA than in patients without HaPA at 72 h pre- (34.81 ± 15.95 pmol/L vs. 22.64 ±4.14 pmol/L) and post-operation (40.44 ± 19.76 pmol/L vs. 28.51 ± 9.76 pmol/L) (Figure 2(a)). No significant difference was found within headache or non-headache group at 72 h pre- (34.81 ± 15.95 pmol/L vs. 40.44 ± 19.76 pmol/L) and post-operation (22.64 ±4.14 pmol/L vs. 28.51 ± 9.76 pmol/L) (Figure 2(a)). Plasma PACAP1-38 levels followed a similar trend to that seen with plasma CGRP levels. Plasma PACAP1-38 levels were significantly higher in patients with HaPA compared with patients without HaPA at 72 h pre- (23.32 ± 6.51 pmol/L vs. 20.51 ± 4.73 pmol/L) and post-operation (34.86 ± 12.48 pmol/L vs. 26.83 ±4.85 pmol/L) (Figure 2(b)).
Plasma concentration of CGRP and PACAP1-38 in patients with pituitary adenoma pre- and post-operation. (a) Significantly increased plasma CGRP concentration in patients with HaPA compared to patients without HaPA pre- and post-operation. Significantly increased plasma CGRP concentration in patients without HaPA post-operation compared to pre-operation. (b) Significantly increased plasma PACAP1-38 concentration in patients with HaPA compared to patients without HaPA pre-operation. Significantly increased plasma PACAP1-38 concentration in patients without HaPA post-operation compared to pre-operation. No significant elevation was observed in patients with HaPA pre- and post-operation.
Eighteen patients reported significant improvement in headache at 3 months after operation, while four patients received little improvement in headache. Plasma CGRP concentrations were similar between these two groups at 72 h before operation (32.99 ±16.21 pmol/L vs. 51.44 ± 21.53 pmol/L, p = 0.07). However, patients who experienced little relief in headache had higher plasma CGRP concentration at 72 h after operation compared with those with significant improvement (68.71 ± 27.95 pmol/L vs. 36.40 ±10.72 pmol/L, p < 0.01). Similar results were also detected in plasma PACAP1-38 concentration. Seventy-two hours prior to operation, plasma PACAP1-38 concentration was similar for both groups (25.15 ± 6.55 pmol/L vs. 34.14 ± 1.73 pmol/L, p = 0.12). Plasma PACAP1-38 concentration remained high in patients who had little improvement in headache compared with those who had significant improvement at 72 h after operation (52.07 ± 14.64 pmol/L vs. 29.45 ± 6.51 pmol/L, p < 0.01). The above data indicated that most patients with HaPA could benefit from tumour removal, and this beneficial effect might, at least in part, rely on the significant reduction in plasma CGRP and PACAP1-38 concentrations after operation.
Validation of criteria in ICHD-3 and Goadsby 2005 proposals
Only nine of 33 HaPA patients fulfilled the ICHD-3 criteria. The remaining 24 patients had non-functioning adenoma. No significant changes of temperature regulation, emotional state and altered thirst and/or appetite were reported in these 24 patients. Meanwhile, all HaPA patients fulfilled criteria A–C in the Goadsby 2005 proposals, indicating that emphasis on secretion status excludes most HaPA patients from current ICHD-3 criteria (10). Most HaPA patients (83.3%) had significant improvements in headache after surgical resection and/or medical therapy.
Discussion
This prospective study analysed the presence rate and clinical features of headache in patients with PA. We also detected the plasma levels of CGRP and PACAP-38 to explore the potential pathogenesis of HaPA. In our study, of 63 patients with PA, the presence rate of HaPA was 52.4%. Headaches are common in patients with PA although the incidence can be variable, ranging from 33% to 72% (1–3). This variability may have arisen due to the different clinical settings.
Demographic characteristics in our cohort were consistent with previous studies, such as mean age of patients with PA and female-to-male ratio (3,8,23–25). However, discrepancy in headache characteristics was found between our study and previous studies. The most common type of HaPA in our patients was migraine-like headache followed by TTH-like headache, while none of our patients reported trigeminal autonomic cephalalgia-like (TAC-like) headaches. According to other descriptions, headache in prolactinoma patients could present as migraine-, TTH- and TAC-like headaches, such as cluster headache (CH), chronic paroxysmal headache (CPH), short-lasting unilateral neuralgiform headache with conjunctival injection and tearing (SUNCT) or hemicrania continua (HC) (26–29). In 2014, Bengt Edvardsson reported that a non-functional PA patient presented with a CH-like headache; following surgery to remove the adenoma, the patient’s headache attacks were completely resolved, thus suggesting that HC-like headaches are associated with adenoma (30). Another study reported similar findings, suggesting that only a small number (8%) of patients with headache could not be classified using the ICHD (12). In our setting, headaches in 39.4% of patients could not be identified by ICHD, partly because the referral system in China is quite different from the USA, UK and Canada, as described previously, or because headache might not be the only complaint when they were referred to the department.
It has been reported that hormonal hypersecretion is closely related to the development of HaPA. An earlier systematic review showed that prolactinoma accounted for 32–66% of cases, while somatotropinoma accounted for 8–16%, and ACTH accounted for 2–6% (31). Consistent with previous reports, we found that gender, cavernous sinus invasion and chiasm compression were not associated with HaPA, while hypersecretion of the PA, headache history and a family history of headache were significantly associated with the presence of headache (p < 0.05). Of note, a previous study that recruited patients presenting with both PA and troublesome headache showed that the majority had secretory tumours (37% prolactinoma, 33% somatotropinoma) (8), suggesting that secretory tumours might predispose patients to headache, especially moderate-to-severe headache. However, in our study, nonfunctional adenoma was the predominant type of PA, accounting for 82.5% of patients, while the proportion of secretory adenoma was low. This difference might be caused by the different referral system in China. Moreover, Ferrante et al. found that the presence rate of headache in patients with non-functional adenoma was as high as 41.4% (32). A recent study also confirmed this presence rate and demonstrated that over 70% of patients with non-functional adenoma experienced moderate-to-severe headache. These data implied that secretion status might not be the exclusive cause in the genesis of HaPA, and other associated causes require further clarification.
Previous studies have shown that there is a large number of CGRP nerve fibers in the anterior pituitary, which could regulate pituitary cells via the hypothalamus. In addition, it has been confirmed that PAs contain a large number of PACAP1-38 receptors and that PACAP1-38 can act upon pituitary glandular cells (19). CGRP and PACAP1-38 have been proved to be involved in headache initiation (33,34). It has been described that CGRP and PACAP1-38 are involved in central sensitisation, increasing the neuronal excitation and facilitating induction of chronic pain (33). Exogenous administration of CGRP or PACAP1-38 induces headache in healthy controls or patients with migraine (35,36). In our study, HaPA was primarily migraine-like headache, which is consistent with the prior findings (8,11). Therefore, we hypothesised that CGRP and PACAP1-38 might be involved in the development of HaPA. This study was the first to measure plasma CGRP and PACAP1-38 levels in patients with PA and showed that both plasma CGRP and PACAP-38 levels were significantly higher in patients with headache than in patients without headache (p < 0.05) at 72 h pre- and post-operation. Both CGRP and PACAP1-38 plasma concentrations were significantly increased in patients without HaPA after surgery, which might be due to the surgical stress or inflammatory status. Taken together, besides secretion status, upregulation of plasma CGRP and PACAP1-38 concentrations might be an essential factor in the development of HaPA in patients with PA.
The ICHD-3 classification defines headache caused by PA as attributed to pituitary hyper- or hyposecretion (including PRL, GH and/or ACTH) (10). In our present study, none of the patients had pituitary hyposecretion; however, data showed that pituitary hypersecretion was associated with HaPA, which was in accordance with the ICHD-3 criteria for headache caused by a PA (10). As stated previously, 82.5% of patients in this cohort were diagnosed with nonfunctional adenoma, half of whom manifested HaPA. Current ICHD-3 (10) only includes HaPA related to secretory disturbances. Consistent with the Goadsby 2005 proposals (8), criterion C in ICHD-3 beta was insufficient to deal with non-functional adenoma, which constituted the majority type of PA patients. Thus, we suggested that criterion C might be modified as “headache develops in close temporal proximity to endocrine abnormality or with symptoms attributable to pituitary disease, such as visual loss.”
This study was a cross-sectional hospital study. Consequently, data arising from patients at the First Affiliated Hospital of Chongqing Medical University cannot represent the general population. Due to limited conditions, this study was not able to specify if these cases had increased saddle pressure. Furthermore, we were not able to collect tissue specimens from PAs. We were only able to measure plasma CGRP and PACAP1-38 levels. If conditions permitted, we could have collected PA tissue samples to detect CGRP, PACAP1-38 and their receptors in tissue, and allow the immunolocalisation of innervations and receptor distribution.
Conclusions
We have described the clinical characteristics and treatment outcomes observed in 63 patients with PA, half of whom had HaPA. The majority of HaPA patients presented with migraine-like headache, although a range of headache phenotypes was observed. Most patients in this study had non-functional adenoma, half of whom had HaPA, indicating that the ICHD-3 criteria for HaPA, with the emphasis on secretion status, exclude most HaPA patients from current ICHD-3 criteria. Thus, the ICHD-3 criteria might consider modifications to criterion C to include headache attributed to non-functional adenoma. Furthermore, patients with HaPA had higher plasma CGRP and PACAP levels than patients without HaPA, and lower plasma CGRP and PACAP1-38 concentrations after operation suggested a better outcome in patients with HaPA. Therefore, CGRP or PACAP1-38 antibodies might be future targets for the acute treatment of HaPA.
Clinical implications
Both surgical and medical treatment for tumour resection significantly reduce headache severity in the majority of patients with HaPA at 3-month follow-up, and additional medication on HaPA might not be a necessity. The ICHD-3 criteria for HaPA with the emphasis on secretion status might lead to the misdiagnosis of headache attributed to non-functional adenoma, which comprised the majority patients with HaPA in our cohort. Lower plasma CGRP and PACAP1-38 concentrations at 72 h after operation may predict a better outcome in patients with HaPA, although further studies are needed to confirm this hypothesis.
Footnotes
Acknowledgments
Appreciation is expressed to all the members in the headache clinic and department of neurosurgery of The First Affiliated Hospital of Chongqing Medical University for their assistance in screening for eligible participants. We also would like to thank all the subjects in this study.
Author contributions
YZ and HJ designed the study, extracted the data and performed the analyses. YZ and HJ wrote the manuscript. QP, HJ, GY, LC, GQ, and JZ critically revised the manuscript. JZ has full access to all the data in this study and takes responsibility for the integrity of the data as well as the accuracy of analysis. All authors read and approved the final manuscript.
Declaration of conflicting interests
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This study was supported by National Natural Science Foundation of China (No: 81671092).