Abstract
Background
Since the first description of nummular headache (NH), more than 500 cases have been described, delineating its clinical phenotype and response to treatment. However, data on the natural history of NH and outcomes during long-term follow-up are not currently available. The present study aimed to describe the long-term outcomes and follow-up of a large series of patients with NH.
Methods
A descriptive observational ambisective study with a series of cases was conducted. The study population included adult patients with primary NH and a minimum of 12 months of follow-up. Demographic variables, previous medical history, clinical phenotype, diagnosis and treatment of NH, temporal pattern, and long-term evolution were analysed.
Results
In total, 168 patients were enrolled and followed for a median [interquartile range (IQR)] of 80.5 (55–118.5) months. The temporal pattern after NH onset was chronic in 67.9% and, at diagnosis, the median (IQR) number of pain days per month was 20 [10–30] days with 138 (82.1%) patients with ≥8 days of pain per month. Preventive treatment was needed by 112 (66.7%) patients. The most frequently used drugs were gabapentin (69/112; 61.6%), onabotulinumtoxinA (38/112; 33.9%), amitriptyline (31/112; 27.7%) and lamotrigine (21/112; 18.7%). Response to preventive treatment was at least partial in 91/112 (81.3%) patients. At the end of follow-up, 81 (48.2%) patients had inactive NH. Of patients with active NH, the median (IQR) number of headache days per month was 3 (1–12) days and patients had ≥8 days of pain in 35 (20.8%) cases.
Conclusions
Long-term outcomes of NH were positive in most patients. After a median of 6.7 years of follow-up, 48% of cases were inactive. Two-thirds of patients required preventive treatment, and 80% of them were treatment-responsive. In NH cases that remained symptomatic, the headache frequency was lower, and the proportion of patients with chronic NH decreased from 68% to 11%.
Introduction
Nummular headache (NH) is a headache disorder first described in 2002 by Pareja et al. (1), who identified 13 patients with a headache circumscribed within a small area of the scalp. The International Classification of Headache Disorders, 3rd edn (ICHD-3), defines nummular headache (NH) as a primary headache disorder, in which the pain is limited to a well-circumscribed area of the scalp: either (a) rounded, for which it is named (nummus, or “coin” in Latin) or (b) elliptical. The diameter of the painful area must be between 1 and 6 cm (2). NH incidence is estimated to be between six and nine cases per 100,000 people per year, although it may be underestimated (3). Since its first description, more than 500 cases have been reported in the literature, characterizing its clinical phenotype and response to acute and preventive treatments (4,5). However, data regarding the natural history of NH and outcomes during long-term follow-up are not currently available (6). The present study aimed to describe the long-term outcomes and follow-up period of a large series of patients with NH.
Methods
Study design and setting
A descriptive observational study with a series of cases was conducted. The study was done and reported according to the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) recommendations (7) and the Guidelines of the International Headache Society for Clinic-Based Headache Registries (8). The study population included patients with primary NH. The study setting was the Headache Unit of the Hospital Clínico Universitario, Valladolid, Spain, a tertiary public university hospital with a direct reference population of 270,000 individuals. This unit receives approximately 800 new headache patient visits per year, with patients referred from both primary care and general neurology offices.
Eligibility criteria
Patients were included if they fulfilled the following criteria: (a) age ≥18 years at the time of diagnosis; (b) diagnosis of NH according to the ICHD-3 criteria (2); (c) time since NH onset ≥1 month; (d) availability for follow-up period ≥12 months; and (e) consented to participate in the study.
Patients were excluded if they: (a) were unable to precisely describe the characteristics of the pain; (b) had another primary or secondary headache disorder from which the NH was not clearly distinguishable; (c) had a secondary disorder as the cause of the NH (9); (d) had a post-traumatic NH (10); (e) had any serious systemic or psychiatric condition that made the assessment imprecise; (f) had cognitive impairment; (g) passed away during follow-up; (h) declined to participate; (i) could not complete follow-up; or (j) had symptoms that were better explained by another disorder of the ICHD-3 (2).
Study period
All consecutive NH patients diagnosed between January 2008 and January 2021 were screened for eligibility. Follow-up evaluation was completed prospectively for all 168 patients between January and December of 2021, and patients were followed beyond that period when possible.
Study intervention
All consecutive patients evaluated in the headache outpatient clinic on or after 1 January 2008, are systematically included in a clinical registry. Patients are routinely diagnosed and classified according to the ICHD criteria (2,11,12). A specific registry for NH patients is also in operation, and information is collected at the time of the patient diagnosis by using a structured questionnaire (see supplemental material). For all patients, an in-person clinical interview is conducted by a headache expert. The specific details of the registry and its results have been published elsewhere (3,10,13). For the present study, follow-up data were collected by a retrospective review of patients’ medical records, assessing all the study variables listed below. In addition, a subsequent evaluation of all the study patients was done between January 2021 and December 2021, collecting all the variables that were missing in the patients’ records or that were unclear, especially regarding follow-up and response to treatments.
Studied variables
The studied variables included (a) patients’ demographics and prior medical history; (b) clinical phenotype of the NH; (c) diagnosis, treatment type and treatment response; and (d) temporal pattern and long-term evolution of the NH after treatment, all described below.
Demographics and prior medical history
Demographic variables included age at onset, age at diagnosis, sex assigned at birth, and the duration of NH disease, expressed in months, prior to the NH diagnosis. Prior history of other primary or secondary headache disorders was assessed and classified according to the ICHD-3 criteria (2). In patients who were diagnosed with a NH prior to the publication of the ICHD-3 criteria, the headache diagnosis was revised and re-classified by using the ICHD-3 after a clinical in-person interview (2).
Clinical phenotype
Despite the ICHD-3 criteria do not differentiate between chronic or episodic NH, for the present study, we classified the temporal pattern as chronic, if patients had experienced ≥15 days per month for ≥3 months or episodic, if they had <15 days per month. The clinical phenotype of NH was systematically characterized, including the number of pain days per month and the pain quality (pressing, burning, stabbing, throbbing, electric or other). Intensity of pain was measured by using a verbal rating scale (VRS; 0, no pain to 10, the worst imaginable pain). Headache topography was assessed according to the precise location, based on the subjacent skull bone (frontal, parietal, temporal, occipital) and the laterality of the painful area (right, left or sagittal). NH shape (elliptical or circular) and size (cm) were collected.
The presence of exacerbations was assessed, and whenever these existed, their quality, duration (seconds) and intensity (0–10, VRS) were collected. The presence of associated symptoms, including photophobia, phonophobia, nausea, vomiting, cranial autonomic symptoms and/or worsening by physical activity, was systematically evaluated. The existence of superficial sensory disturbances, including allodynia, hyperalgesia or hypoesthesia within the painful area, was systematically assessed at diagnosis. The circadian temporal pattern of the headache and night sleep interruption were also assessed.
Diagnosis and treatment
According to the local standard of care and the national guidelines, cranial imaging is requested for every NH patient to rule out a secondary cause of the headache (14), either by magnetic resonance imaging (MRI) or computed tomography (CT), or both. The existence of possible triggers of NH, including craniofacial trauma, local space-occupying lesions, local surgery or viral infections, was assessed. For the present study, a secondary cause of the NH was considered an exclusion criterion. Details about secondary NH cases from this cohort have been published elsewhere (9).
Regarding treatment, prior use of acute and preventive medications was assessed, as well as the patient’s response to them after 3 months of treatment. Response was rated by adapting the International Headache Society guidelines for conducting clinical trials of acute medications for migraine (15) and preventive treatment of chronic migraine (16). Response to acute medication was defined as optimal if pain freedom occurred within 2 hours in >75% of uses, positive if it occurred in 50–74% of uses, partial if it occurred in 30–49% of uses, and no response if it occurred in <30% of uses. Response to preventive medication was defined as the proportion of patients who achieved a decrease in the number of headache days per month by ≥50% compared to the month prior to the start of treatment (i.e. response rate). Response was defined as optimal if the decrease in headache days per month was 75% or greater, standard if the decrease was between 50% and 74%; partial if between 30% and 49%; no response if below 30%; and no tolerability when adverse effects led to treatment discontinuation prior to receiving it for a sufficient length of time to assess effectiveness.
Temporal pattern and long-term evolution of NH
During follow-up, the NH was classified as inactive if the patients remained completely pain-free for ≥6 months, or active if they continued experiencing NH episodes. During follow-up, the NH was classified as inactive if the patients remained completely pain-free for ≥6 months, or active if they continued experiencing NH episodes. Patients were classified as inactive NH also if the headache had resolved between the general practitioner referral and the patient evaluation in the Headache Unit. These patients who are inactive at the beginning are patients who at the time of the first consultation did not have symptoms of NH, but had symptoms that fulfilled the ICHD compatible with the disease before this consultation.
Patients with inactive NH were subclassified as having spontaneous resolution if no preventive treatment had been administered, or as having resolution after treatment if the patient improved after receiving treatment with possible therapeutic effect, according to the local national guidelines (14).
Statistical analysis
Qualitative variables are reported as frequency and percentage. When a subgroup of patients was reported (e.g. patients who received acute medication), the fraction of patients was specified. Quantitative variables are presented as the mean ± SD if the distribution was normal, or the median and interquartile range (IQR) if the distribution was not normal. Normality was evaluated with the Kolmogorov–Smirnov test. For hypothesis testing, chi-squared tests were used to evaluate qualitative variables and an independent samples Student’s t-test or a Mann–Whitney U-test were used to evaluate quantitative variables when normally or not normally distributed, respectively. The statistical significance threshold was set to an alpha of 0.05. Statistical analyses were performed using SPSS, versus 26.0 (IBM Corp., Armonk, NY, USA).
Results
During the study period, 7115 new patients were evaluated in the Headache Unit and 282 patients received a NH diagnosis. Figure 1 shows the flow diagram of screened, included, excluded and enrolled patients. The final sample was composed of 168 patients who met the eligibility criteria. All eligible patients agreed to participate and completed the study.
Flow diagram of screened, included, excluded and enrolled patients. ICHD-3, International Classification of Headache Disorders, 3rd edn; NH, nummular headache.
Demographics and prior medical history
The median (IQR) duration of patient follow-up was 80.5 (55.0–118.5) months, accounting for a total of 1198.2 patient-years. Patients’ sex was female in 107 (63.7%) cases. Mean ± SD age of onset was 47.4 ± 18.0 years (range 7–84 years) and median (IQR) age at the time of NH diagnosis was 49 (35–64) years. The median (IQR) duration of NH disease at the time of diagnosis was 10 (5–24) months.
Prior history of other headache disorders was reported by 54 (32.1%) patients and, in five cases, multiple headache disorders had been diagnosed. Patients had received a diagnosis of migraine in 29/54 (49.2%), epicrania fugax in 16/54 (27.1%), tension-type headache in 4/54 (6.8%) and other headache disorders in 10/54 (16.9%) cases. The full list of prior history of headache disorders is available in the supplemental material.
Clinical phenotype
Table 1 summarizes the clinical variables. The temporal pattern after NH onset was chronic in 114 (67.9%) cases. At diagnosis, the median (IQR) number of pain days per month was 20 (10–30) days. Patients had ≥4 days of pain per month in 158 (94.0%) cases, ≥8 days of pain in 138 (82.1%) cases and ≥15 days in 114 (67.9%) cases. Pain was rated as at least moderate in 120 (71.4%) cases and severe in 36 (21.4%) cases. The predominant NH phenotype consisted of a pressing pain, moderate in intensity, predominantly located in the parietal region, circular in shape and about 4 cm in diameter. Associated symptoms were uncommon (Table 1).
Clinical phenotype of NH in the baseline situation (n = 168).
*Differing denominators indicate missing data from the total sample (n = 168). NH, nummular headache; IQR, interquartile range; VRS, verbal rating scale.
Diagnosis and treatment
All patients underwent cranial imaging at the time of diagnosis: CT in 80 (47.6%), MRI in 52 (30.9%) and both CT and MRI in 36 (21.4%) patients. Incidental imaging findings were observed in eight (4.8%) cases. The full list of incidental findings is available in the supplemental material.
Symptomatic treatment was needed by 112 (66.7%) cases. The symptomatic treatments used were NSAIDs in 45/112 (40.2%), paracetamol in 37/112 (33.0%), metamizole in 6/112 (5.4%), triptans in 2/112 (1.8%), opioids in 1/112 (0.9%) and other drugs in 21/112 (18.7%) cases. The response achieved corresponded to optimal response in 21/112 (18.8%), positive response in 21/112 (18.8%), partial response in 55/112 (49.1%) and no response in 15/112 (13.3%) cases.
Preventive treatment was needed by 112 (66.7%) patients. The total number of preventive treatments that each patient received was one in 57/112 (50.8%), two in 25/112 (22.3%), three in 17/112 (15.2%), four in 3/112 (2.7%), five in 7/112 (6.3%) and six in 3/112 (2.7%) cases. In 17/112 (15.2%) cases, a combination of drugs was used. The most frequently used drugs were gabapentin in 69/112 (61.6%), onabotulinumtoxinA in 38/112 (33.9%), amitriptyline in 31/112 (27.7%) and lamotrigine in 21/112 (18.7%) cases. Treatment response to these therapies is shown in Figure 2. The proportion of patients who experienced adverse events was 21/69 (30.4%) after gabapentin, 1/38 (2.6%) after onabotulinumtoxinA, 5/31 (16.1%) after amitriptyline and 6/21 (28.6%) after lamotrigine.
Response to preventive treatments. Response to preventive medication was defined as the proportion of patients who achieved a decrease in the number of headache days per month by ≥50% compared to the month prior to the start of treatment.
The best achieved response was optimal in 59/112 (52.7%), positive in 75/112 (66.9%), partial in 91/112 (81.3%) and no response in 10/112 (8.9%). Preventive treatment was not tolerated by 11/112 (9.8%) patients. A partial response was achieved in monotherapy in 82/91 (90.1%) patients and in polytherapy in 9/91 (9.9%) patients. The number of treatments that were needed to achieve at least a partial response was one in 57/91 (62.6%), two in 20/91 (22.0%), three in 9/91 (9.9%), four in 3/91 (3.3%) and six in 2/91 (2.2%).
Temporal pattern and long-term evolution of NH
Figure 3 summarizes the long-term follow-up. At the time of NH diagnosis, 164 (97.6%) patients presented with an active NH and four (2.4%) with an inactive NH. Among patients with active NH, 111/164 (67.7%) received preventive treatment. Patients with active NH who received preventive treatment had a median (IQR) of 25 (15–30) headache days per month and those who did not receive preventive treatment had a median (IQR) of 12 (6.2–20) headache days per month (p < 0.001). At the last recorded follow-up visit, 49/91 (53.8%) patients who had at least partial response to preventive treatment had an inactive NH, and 42/91 (46.3%) patients had an active NH. Among the four patients with inactive NH at onset, one (25%) patient experienced a relapse and required preventive treatment, with a non-sustained response, and had an active NH at the end of follow-up.
Sankey diagram of the follow-up and evolution of the patients with nummular headache (NH) (inactive NH if the patients remained completely pain-free for ≥6 months or active NH if they continued experiencing NH episodes).
Among the 21/111 patients with active NH at the beginning of the follow-up with no response to preventive treatment 10/21 (47.6%) or non-tolerance to it 11/21 (52.4%), 4/21 (19.0%) presented a non-sustained improvement with a median (IQR) of 91 (42.8–120.5) months. Among these patients, 3/4 (75%) continued with active NH at the end of follow-up and 1/4 (25%), despite presenting symptoms of the disease again, was inactive at the end of follow-up. Among the 17/21 (81.0%) with non-sustained improvement, 11/17 (64.7%) remained active at the end of follow-up and 6/17 (35.3%) were inactive.
Fifty-six patients did not require preventive treatment, due to the low intensity of the pain, because they presented a pain intensity ≤5 in 29/56 (51.8%) or due to the low frequency. At the end of follow-up, this group of patients had <3 days of pain per month in 45/56 (80.4%) and <15 days of pain per month in 54/56 (96.4%). At the end of follow-up, 33/56 (58.9%) had inactive NH.
At the end of follow-up, of the entire study sample, 81 (48.2%) patients had an inactive NH and 87 (51.8%) had an active NH. Of patients with an active NH, the median (IQR) number of headache days per month was 3 (1–12) days. Patients had ≥4 days of pain per month in 42 (25.0%) cases, ≥8 days of pain in 35 (20.8%) cases and ≥15 days in 19 (11.3%) cases. Twenty-six (15.5%) patients continued with preventive treatment after at the end of the follow-up.
Discussion
The present study provides data obtained during long-term follow-up of NH. After a median follow-up time of 6.7 years after diagnosis, NH was inactive in 48% of patients; in those who continued experiencing NH episodes, the median frequency of headache days per month was 3 days of pain, with only 21% of patients experiencing ≥8 headache days per month. The percentage of subjects with ≥15 days of headache was slightly higher than double of subjects with <15 days, similar to that reported in the literature (3). It has been observed that there is no association between the number of headache days and the probability of having an inactive NH at the end of follow-up (p = 0.6). Preventive treatment was needed by 67% of patients, who exhibited at least a partial response in 81% of cases and an optimal response in 53% of cases.
NH is not as frequent as other prevalent headache disorders, such as tension-type headache or migraine, but its incidence could be estimated to be 8.6 cases per 100,000 patient-years, an estimation similar to those of central nervous system cancer (11 cases per 100,000 patient-years), myelopathies (13 cases per 100,000 patient-years) and amyotrophic lateral sclerosis (5 cases per 100,000 patient-years) (17). Previous studies have reported data on the clinical phenotype of NH and the patient response to treatment (1,3,6,10,18), but its long-term nature has not yet been characterized (5).
A topic of debate is whether NH is a localized form of migraine or a true separate disorder (19). Age of onset of NH appears to be a bit later than migraine (1,3,13–15,18,20–23). Some patients report NH as a consequence of a local secondary cause (9) and some patients describe NH onset after local trauma within the NH area (10). In the present study, the presence of typical migraine features was infrequent; however, these were evaluated in a cross-sectional manner and not prospectively or according to headache diaries. Interestingly, the proportion of patients with prior history of migraine was slightly higher than expected in our setting because the prevalence of migraine in Spain is 12.6% (24) and 17.2% of patients in our population had a previous history of migraine. On the other hand, NH is frequently associated with epicrania fugax (5,25).
NH punches above its weight. Despite its small size, 71% of patients rate the intensity of NH pain as moderate, 66–84% of patients need acute treatment (5,10) and 50–67% patients need preventive treatment (10). The most frequently employed preventive drugs are gabapentin, lamotrigine and onabotulinumtoxinA (10,26,27). Both tolerability and effectiveness favour onabotulinumtoxinA (4,18,27,28). In one study, most of the treated patients experienced at least a partial response, and more than half had an optimal response (4). In the present study, the proportion of patients who had ≥15 headache days per month changed from 68% to 11% by the end of follow-up, and, in 48% of patients, NH became inactive.
The present study has important limitations. First, it is an observational study with patients from a single centre. Treatment was open-label, and the selection of the employed drugs depended on the responsible physician’s judgement. To simplify the long-term outcome, treatment response was evaluated once after all therapies were administered and not after each individual intervention. The duration of follow-up was not equal across patients, but all patients had ≥1 year of follow-up.
Conclusions
Long-term outcomes of NH were positive in most patients. After a median of 6.7 years of follow-up, 48% cases of NH were inactive. Two-thirds of patients required preventive treatment and 80% of them were treatment responsive. In NH cases that remained symptomatic, the headache frequency was lower at the end of follow up than at diagnosis and the proportion of patients with a chronic NH decreased from 68% to 11%. The present study provides the first known observations of outcomes during long-term NH follow-up care. More studies are needed to confirm these results.
Clinical implications
Almost half of the patients who were diagnosed with NH experienced complete pain freedom by the end of the follow-up period. Two-thirds of NH patients required preventive treatment. Most of these patients responded to preventive treatment, and half of the patients who were treated experienced an optimal response. In patients who still suffered from NH episodes after treatment, the number of days of pain per month decreased over time, with only 21% of patients experiencing ≥8 days of pain at the end of the follow-up period.
Supplemental Material
sj-pdf-1-cep-10.1177_03331024231201576 - Supplemental material for Long-term outcomes of nummular headache: A series of 168 patients and 1198 patient-years of follow-up
Supplemental material, sj-pdf-1-cep-10.1177_03331024231201576 for Long-term outcomes of nummular headache: A series of 168 patients and 1198 patient-years of follow-up by Cristina García-Iglesias, Ana González-Celestino, Álvaro Sierra Mencía, Yésica González Osorio, Andrea Recio García, Cristina Martínez-Badillo, Ana Echavarría Íñiguez, Berta Varona-Galán, David García-Azorín Ángel Luis Guerrero-Peral in Cephalalgia
Footnotes
Declaration of conflicting interests
The authors declare that there are no potential conflicts of interest with respect to the research, authorship and/or publication of this article.
Ethical statement
The study was approved by the Ethics Committee for drug research of the Valladolid East Health Area (code PI 20-1980). The study was carried out in accordance with the principles of the Declaration of Helsinki. Patients participated voluntarily and verbal informed consent was obtained.
Funding
The study was funded by Gerencia Regional de Salud, Consejería de Sanidad, Salud Castilla y Leon, Salud Castilla y León (SACYL), Spain, code: GRS 2416/A/21.
Supplemental material
Supplemental material for this article is available online.
References
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
