Introduction

The lifetime possibility of being diagnosed with an invasive cancer is around 40%¹. Approximately 10-30% of patients with cancer will develop brain metastases (BM) at some point during their disease. However, the true incidence of BM might be higher, with autopsy studies suggesting an incidence of up to 40%². The prognosis of patients who develop BM is usually poor, with a median overall survival of 12 months after being diagnosed with BM³.

Headache is the most prevalent neurological symptom experienced by almost everyone; the actual percentage of the adult population with an active headache is around 47%⁴. The principal general cause is a primary headache, accounting for at least 80% of cases⁵. Secondary causes of headache should be considered and excluded if the headache does not fulfill corresponding criteria established by the International classification of headache disorders-3 (ICHD-3)⁶; or if the headache presents with any simultaneous red flag, especially those abbreviated as systemic symptoms/signs and disease, neurologic symptoms or signs, onset sudden, onset after the age of 40 years, and change of headache pattern) which includes having co-occurrent or previous cancer⁵,⁷. Characteristics of headache often provide valuable information about the underlying etiology; therefore, the type of headache, onset, intensity, accompanying nausea/vomiting, location, temporality, exacerbation with Valsalva maneuver or cough, modifiers (such as position or Valsalva maneuver), or any other simultaneous neurologic symptom or sign should continuously be assessed⁸,⁹. Prior studies have described headache characteristics in patients with brain tumors¹⁰-¹⁴, also known as brain tumor-associated headache¹²; however, studies that report headache characteristics or accompanying symptoms or signs in patients with systemic cancer who are diagnosed with BM are scantly found.

The present study aimed to measure the association of headache characteristics and accompanying neurologic signs and symptoms with the diagnosis of BM in patients with systemic cancer.

Materials and methods

A retrospective observational study at a single center included patients with pathologically confirmed systemic cancer sent by their primary oncologist specialist for the evaluation by the neuro-oncology (NeOn) unit due to headache. Patients with a primary central nervous system (CNS) or hematologic malignancy were excluded from the study. To determine if patients presented with BM, most patients had a magnetic resonance imaging (MRI) of the brain done. The NeOn followed patients not evaluated by an MRI for at least 1 year. BM were excluded based on MRI results or if clinical/neurological deterioration was not seen during the 1^st year after the NeOn consultation.

We used the following definitions: synchronic tumor, more than one diagnosed active systemic malignant neoplasm (i.e., breast cancer and sarcoma). A visual complaint was determined when papilledema, diplopia, ptosis, or decreased visual acuity were present. Valsalva maneuver exacerbation if patients reported aggravation of the headache while performing a physical effort (i.e., during defecation, urination, or coughing). Headache intensity was measured with the visual analog scale (VAS) as rated by the patients.

Results

From 1,248 patients sent for NeOn consultation due to headache, 519 were diagnosed with a hematologic malignancy and 154 with a primary CNS tumor and thus excluded; therefore, we present the characteristics of 572 patients with cancer who complained of headache, of which 216 (38%) were diagnosed with BM.

The mean age at NeOn evaluation was 49.7 ± 13.1 years. Female gender was more prevalent (86%); the primary systemic tumor was breast in 261 patients (46%), gynecologic (cervix-uteri, ovarian, and endometrial) in 91 (16%), head and neck in 51 (9%), lung in 50 (9%), urologic in 34 (6%), skin (including melanoma) in 4%, other (i.e., bone, soft tissue, and primary undetermined) in 4%, synchronic (more than two primary sites) in 4%, and gastrointestinal in 3%.

Discussion

In a cohort of patients with systemic cancer sent for NeOn consultation due to headache, BM were found in 38%; clinical characteristics associated with a higher risk of BM were age < 65 years, male gender, a headache starting after a cancer diagnosis, oppressive type, and the presence of vomiting; having an exacerbation of headache with changes in position and generalized location correlated with a lower risk of BM. In addition, more than one neurologic complaint (other than a headache) was associated with a higher risk of BM, especially if patients had ataxia, visual complaint, altered mental status, seizures, focal motor weakness, and vertigo.

Headache is a well-recognized red flag in patients with cancer for a secondary cause, such as BM, and should permanently be excluded¹⁵. Headache is the most frequent neurological symptom associated with brain tumors (both primary or metastatic)¹⁶. Previous studies have reported a prevalence of headache in 8-71% of patients with brain tumors⁹,¹⁰,¹²,¹⁷-²¹. They have also reported the so-called "classic" brain tumor headache characteristics (described as severe, worse in the morning, with nausea and vomiting) to be uncommon, similar to our findings. On the other hand, following our results, other simultaneous neurological symptoms have been consistently associated with BM¹²,¹³,¹⁷,¹⁹. In our study, 73% of patients with a headache diagnosed with BM had another accompanying neurological symptom.

Headache in patients with cancer does not always mean having a BM; differential diagnoses include primary intracranial tumors (i.e., pituitary tumors, gliomas, and meningiomas)²¹, neuroinfections, vascular disease, radiotherapy, complications of systemic treatments, steroid use or withdrawal, intracranial hypertension, lumbar puncture, aseptic meningitis, systemic hypertension, and finally, primary headaches¹⁵,¹⁶,²¹. In addition, the pathogenesis of brain tumor headaches varies and has been reviewed elsewhere⁹,²².

One of the essential strengths of our study is that we looked directly for associations between headache characteristics and the presence of BM in numerous cohorts and compared them with those without BM. Previous studies¹⁰,¹⁷,¹⁸ have only reported frequencies or studied a smaller sample size²³; furthermore, it is not unusual to find many physicians and medical literature recognizing specific headache characteristics as associated with a brain tumor without checked bases. In future studies, this research can be used to verify the clinical significance of the headache features that are linked to BM.

An individualized validation for each cancer-specific site (i.e., melanoma, lung, and breast cancer) would be ideal, and multicenter prospective studies are encouraged to confirm our findings. The retrospective design, a referral bias, and a selection bias should be considered, for the study was done in patients treated at a single center, and the decision of referring patients for NeOn evaluation was determined by their primary treating physician (medical-, radio-, or surgical oncologist), according to their clinical judgment.

We did not classify the type of headache as others have done or attempted to do so because the criteria for primary headaches exclude the presence of secondary causes (i.e., not better accounted for another ICHD-3 diagnosis and other reasons have been excluded)⁶,¹⁴. The ICHD-3 has defined a headache attributed to intracranial neoplasia as one that occurs in a patient in whom an intracranial neoplasm has been diagnosed and in whom there is "evidence of causality was demonstrated by one or more of the following: (a) the headache symptom developed in temporal relation to the cranial neoplasm or led to its discovery; (b) the headache significantly emerged in parallel with the worsening of the intracranial neoplasm; (c) the headache significantly improved in temporal relation to the success of treatment of the intracranial neoplasm; and (d) another ICHD-3 diagnosis does not better explain it"⁶,²². If our current results are prospectively confirmed, other clinical characteristics might be considered an extra item for the following classification.

Finally, all of our patients were seen in the outpatient setting or during hospitalization; the emergency room approach to headaches ought to be managed according to best-practice evidence⁷,²⁴,²⁵.

Conclusion

Headache in patients with cancer is a warning sign, especially if it begins after a cancer diagnosis, male gender, and age < 65 years. The attending team should consider headache characteristics, accompanying symptoms, and signs to assess their risk of being diagnosed with BM as the etiology of their headache.