Abstract
Objective
The objective of this study was to analyse the prevalence and the clinical features of headache as a presenting symptom of glioma.
Methods
We interviewed 527 consecutive adult patients with histologically confirmed glioma, admitted to the Regina Elena National Cancer Institute between 2010 and 2015. We defined four headache phenotypes: Tension-type-like headache (TTH), migraine-like headache, worsening of a pre-existing headache (WPH) and classic brain tumour headache (BTH). Logistic regression analysis was carried out to investigate potential risk factors for headache at presentation of glioma.
Results
12.5% (n = 66) of patients with glioma indicated headache as a presenting symptom of their disease. Of these, 31 patients (47%) had TTH, while BTH and WPH were reported by 28 (42%) and seven (11%) patients, respectively. We did not find any case of migraine-like headache. Infratentorial (p = 0.038) and right-sided tumours (p = 0.013) were more frequently associated with the presence of headache at onset. Patients with TTH were older than patients with BTH and WPH (p = 0.035). BTH was less frequently associated with other neurological signs (p < 0.0001). The multivariate logistic regression analysis showed the localization of the brain tumour in the left hemisphere to be a protective factor for the development of headache.
Conclusions
Our study includes a very large series of patients with glioma, providing a description of headache phenotype at first presentation of disease and investigating possible factors that may influence the clinical features of headache.
Introduction
Gliomas are the most common primary tumours of the CNS, and are classified according to morphologic criteria established by the World Health Organization (WHO) into various histologic subtypes and grades, including astrocytomas, oligoastrocytomas, and oligodendrogliomas (1).
The prevalence of headache during the course of intracranial tumours varies among different epidemiological studies, ranging between 32% and 71% (2–6). Nevertheless, headache at presentation is less frequent, being observed in approximately 20% of patients with intracranial tumours (7,8); usually, at the time of diagnosis, other signs or symptoms have already manifested (2,5).
The Brain Tumour Headache (BTH) is classically described as progressive, worse in the morning and aggravated by Valsalva-like manoeuvres; however, its clinical features are not so specific and sometimes satisfy the criteria for primary headache such as migraine or tension-type headache (3,4,9,10).
Commonly, headache is more frequent in patients with infratentorial than with supratentorial tumours (4,11) and in glioblastomas than in slow-growing tumours (7). Headache side and localization may predict tumour location, at least in the early disease course (12). Furthermore, the elderly are less likely to present with headache associated with brain tumour (13).
It is not known whether the clinical phenotype of headache at presentation of glioma can be influenced by the brain tumour characteristics. In this study, we defined the prevalence and the clinical features of headache as a presenting symptom of glioma and the possible influence on headache phenotype of the neoplasm histology, grading and localization and of the demographic characteristics of patients.
Materials and methods
In this cross-sectional study, we considered all consecutive adult patients diagnosed with glioma who referred to the Neuro-Oncology Unit of Regina Elena National Cancer Institute in Rome between January 2010 and December 2015. The diagnosis of glioma was achieved through histological examination following biopsy or tumour resection. Gliomas (i.e. astrocytic, oligodendrocytic, or a mix of these two cell types) were categorised according to the 4th Edition of the WHO Classification of Tumours of the CNS (1).
Exclusion criteria were: (i) age < 18 years; (ii) disturbance of consciousness; (iii) poor general conditions, defined as Karnofsky Performance Status < 60. After a complete general and neurological examination, all patients underwent a face-to-face interview by one of the authors of the present study with expertise in headache diagnosis (VV).
Firstly, patients were asked if they had suffered from a new onset headache, or if their pre-existing headache had worsened, during the three months preceding the diagnosis of glioma. Headache could be an isolated presenting symptom of brain tumour, or could be associated with other neurological symptoms. In patients who responded affirmatively to this question, the interviewer investigated the clinical features of headache at onset and the personal history of headache.
We defined four headache phenotypes: Tension-type-like headache (TTH), migraine-like headache, worsening of a pre-existing headache (WPH) and classic brain tumour headache (BTH).
TTH and migraine-like headache were diagnosed when headache attacks fulfilled all the IHS criteria (14) for TTH and migraine except for the criterion E (“headache not attributed to another disorder”).
Worsening of a pre-existing headache was defined as an increase in frequency and/or intensity of a previous headache, which was present in the patient’s personal history for at least one year, and which could have migrainous or TTH features.
Finally, BTH was defined as a progressive headache, occurring daily or intermittently, and with at least one of the following three characteristics: Worse in the morning, triggered or aggravated by Valsalva-like manoeuvres, and associated with nausea and/or vomiting.
Other neurological signs and symptoms that had brought to the tumour diagnosis were also registered, including cognitive alterations, seizures, imbalance, sensory-motor disturbances or visual field deficits.
Written informed consent was obtained at the first visit from all subjects who participated in our study. The study was approved by the internal Regina Elena National Cancer Institute Ethics Committee in 2009.
Statistical analysis
Descriptive statistics were produced for all variables of interest. The Pearson’s Chi-squared test of independence or the Mann-Whitney non-parametric test, when appropriate, were employed to investigate the potential differences between patients with and without headache and to compare patients with different headache phenotypes.
Univariate logistic regression models were performed to identify potential risk factors for headache at presentation of brain tumour. Multivariate models were built by including variables testing significant at the univariate analysis.
The level of significance was set at p < 0.05. The collected data were analysed using SPSS version 21.0 (SPSS Inc., Chicago, Illinois, USA).
Results
Of the 598 eligible patients, 527 (88.1%) responded to our face-to-face interview. Seventy-one patients (11.9%) were unwilling to participate; these cases were not significantly different from the participants by demographic characteristics (i.e. gender and age) and by tumour histology, grading and location.
Demographic and clinical features of the study sample (n = 527).
SD: standard deviation.
Clinical and demographic variables in patients with or without headache at presentation of glioma.
Pearson’s Chi-square test.
Mann-Whitney non-parametric test.
SD: standard deviation. Figures in bold: p > 0.05.
Logistic regression analysis: Risk factors for headache at presentation in glioma.
Figures in bold: p > 0.05.
Distribution of demographic and clinical features of patients with headache at presentation according to the headache phenotype.
p-value was calculated using Mann-Whitney non-parametric test.
p-value was calculated using Pearson’s Chi-square test for a contingency table with ≥ 2 rows (each demographic or clinical variable) and 3 columns (each headache phenotype).
SD: standard deviation; BTH: brain tumour headache; TTH: tension-type headache. Figures in bold: p > 0.05.
Discussion
In our study, 66 out of 527 (12.5%) cases with glioma indicated headache as a presenting symptom of their disease; this finding is in accordance with the literature data that report headache at presentation of brain tumours in around 20% of patients (7,8). The prevalence of headache as the sole presenting symptom of brain tumours is even lower (2–8%) (2,5).
The patients with headache weren’t significantly different from patients without headache for age, gender, histology and grading of the tumour, as shown in Table 2. In accordance with other studies (4,11), infratentorial gliomas were frequently associated with headache, probably because of cerebrospinal fluid flow dysfunction and consequent increased intracranial pressure, even if this finding was not confirmed on multivariate analysis.
The multivariate logistic regression analysis showed the localization of the brain tumour in the left hemisphere to be a protective factor for development of secondary headache (Table 3). Patients with gliomas in the dominant hemisphere have more cognitive deficits than those with non-dominant hemisphere lesions (15). Such deficits could conceal or outweigh the symptom of “headache” at onset. Besides, patients with language deficits can be unable to report headache.
In our sample, headache resembled TTH in 31 patients (6% of all glioma cases), while the classic BTH was found in 28 cases (5% of all glioma cases). In their case series, Forsyth and Posner (3) found similar proportions of BTH, but a higher frequency of TTH; however, in their study, it is not clear whether headache was considered as the first presentation of brain tumour or as a symptom appearing during the course of the disease.
We did not find any patient with migraine-like headache at onset, probably because brain tumours are not common migraine mimics (16). Headache at presentation represented the worsening of a previous headache in only seven cases (1% of the overall sample). We did not find any significant difference in the headache type between males and females.
Patients with TTH were older than patients with classic BTH and WPH (p = 0.035). Probably our result reflects the more insidious nature of TTH, which may delay the diagnosis of the underlying brain tumour. The clinical phenotype of headache at presentation was not related to histology, grading, location and side of the brain tumour.
The classic BTH was less frequently associated with other neurological signs (p < 0.0001); this finding could be explained by the fact that these patients are probably diagnosed earlier due to the worrisome features of their headache (progressive course, morning onset, aggravation by Valsalva-like manoeuvres, nausea or vomitus).
Our study has methodological strengths and limitations that need to be addressed. Among the former are: (1) our study includes a very large series of patients with glioma, focusing on headache; (2) the diagnosis of headache type was made by a face-to-face interview; (3) all cases were personally seen by the same author (VV).
Among the latter are: (1) our study population is a selected subgroup of patients with glioma who referred to the Neuro-Oncology Unit for clinical evaluation following surgery and for therapeutic purposes (radiotherapy and/or chemotherapy); (2) we couldn’t define an etiologic association between headache and brain tumour because we didn’t collect information on the course of headache after the tumour removal; (3) the classic BTH was a clinical diagnosis, and we had no information on whether these cases had an increased intracranial pressure.
Our study provides a detailed description of headache phenotypes at first presentation of a large series of gliomas. Headache as the first and sole presenting symptom of gliomas will be dealt with in an upcoming study.
Clinical implications
Headache is observed in approximately 20% of patients with glioma at disease onset. In patients with glioma, headache at onset may be similar to tension-type headache, especially in older people. Infratentorial and right-sided gliomas are more frequently associated with the presence of headache at onset.
Footnotes
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 received no financial support for the research, authorship, and/or publication of this article.