Abstract
Many patients present to their physician with headache. One of the challenges for the physician is to identify those patients whose headaches are secondary to a structural lesion such as a brain tumour. Headache is a common symptom in patients with brain tumours. Most case series of patients with unselected tumours report 50–60% of patients have headache (1–3). Yet primary brain tumours are rare, with an annual incidence of only 10–12/100 000 population per year (4). Metastatic brain tumours are more common, but most present sometime after the diagnosis of cancer, when other metastatic disease has also developed. The incidence of patients presenting with a brain metastasis as a first manifestation of cancer or developing one within 2 months of diagnosis is only 4/100 000 (5).
Previous studies of headache in patients with brain tumours have failed to identify distinctive characteristics of the headaches that might help physicians decide which new headache patients require further investigation. With the exception of the study by Pfund et al., none has used the International Classification of Headache Disorders (ICHD) criteria to diagnose headaches attributable to intracranial neoplasms and have not limited their studies to patients who meet the criteria (2). Information on previous headache history and headache response to tumour surgery is often minimal or absent.
In this issue of Cephalalgia, Valentinis et al. have presented the results of a prospective cohort study of patients with headache attributed to intracranial neoplasms. All patients presenting with a solitary brain tumour had a preoperative interview with a neurologist. All those with new or changed headache had a second interview 3 months after surgery. If their headache resolved or improved postoperatively, they were diagnosed with ‘headache attributable to intracranial neoplasm’ according to the 2004 ICHD-II 7.4 definition (6). Of 206 patients, almost half had new headache meeting the ICHD criteria. The authors collected data on previous headache history, medical history, clinical features and tumour characteristics and then analysed the data to characterize brain tumour headache (BTH) and its risk factors.
They did not find a distinctive headache pattern in their patients. The typical BTH is a non-specific intermittent headache of moderate intensity lasting a few hours that responds to common analgesics and does not meet ICHD-II criteria for a primary headache disorder. If it does meet criteria for migraine, usually there are atypical features. The classic description of severe headache, worse upon awakening, associated with nausea and vomiting, is uncommon. Of those patients with BTH, the headache was the initial symptom of the tumour in about half but, by diagnosis, almost all had other symptoms or signs. Unfortunately, they did not report the duration of the headache before diagnosis.
Their results are interesting, as they suggest there may be several different mechanisms for production of BTH. While increasing size of the tumour within the same histology is associated with an increased risk of headache, size is not significantly associated with headache when all tumour types are combined. Rate of growth of the tumour may also increase the likelihood of headache, as headache was more common in patients with glioblastoma and metastases than in slow-growing, low-grade gliomas or meningiomas. The effect of tumour size appears to be moderated by the age of the patient, as younger patients are more likely to have BTH than older patients. In younger but not older patients, headaches are more frequent with larger tumours as measured by tumour volume. As Valentis et al. have suggested, this may be related to the presence of age-related brain atrophy in older patients. Other studies have not found an association between tumour size and headache, but they did not analyse by individual tumour type or by age. Also, they used the largest diameter as a measurement of size rather than calculating tumour volume, as Valentis et al. did (2).
Patients with a history of longstanding primary headache are more likely to have BTH, as also reported by others (1,3). Schankin et al. have reported that a family history of headache is associated with an increased risk of BTH also, but Valentis et al. did not report family history (3).
Most BTH are intermittent, lasting a few hours and occurring a few times a week. Only about one-third are progressive in nature. This suggests that despite the association with tumour size and speed of growth, the presence of the tumour alone may not be enough to cause headache. Perhaps, in susceptible patients, the tumour lowers the headache threshold and other factors trigger the headache. Migraineurs and patients with other primary headache disorders already have a genetic predisposition to headache and so might be even more likely to have BTH.
This study, like previously published studies, provides no information on headache after the initial tumour resection and 3-month follow-up. Although tumour recurrence following resection would not be expected with some pathologies, it would be interesting to see if the headaches relapsed with recurrence and if the headache characteristics changed at relapse. Headache has been reported in 36% of patients dying of a brain tumour (7).
Despite the non-specific nature of the typical BTH, which causes difficulty in distinguishing it from benign headaches, the risk of BTH is low. A recent case–cohort study examining new-onset undifferentiated headaches presenting to primary care physician in the UK found that the 1-year risk of a malignant brain tumour was only 0.15%, increasing to 0.28% for patients > 50 years old. For new onset of headache meeting ICHD-II criteria for a primary headache disorder, the risk was even lower (0.045%) (8).
This study has provided further information on the characteristics of headache attributable to intracranial neoplasms and has confirmed many of the findings of previous studies. Unfortunately, it appears that BTH does not have distinctive characteristics, and physicians will continue to have to rely on their clinical acumen to decide which patients require further investigation. Brain tumours should always be considered in patients with new headache, especially if associated with other neurological symptoms or signs.