Abstract
Objective
To provide results from the largest study of new daily persistent headache patients to date and specifically evaluate if patients with primary new daily persistent headache develop white matter abnormalities or infarct-like lesions on neuroimaging.
Methods
Retrospective analysis of patient medical records utilizing an electronic medical record system. All patients were seen at a headache specialty clinic by a single headache neurologist and diagnosed with primary new daily persistent headache during the time period of January 2009 to January 2013.
Results
Altogether, 97 patients were diagnosed with primary new daily persistent headache (65 women and 32 men). The mean average age of onset was slightly younger in women than men: 32.4 years vs. 35.8 years. In total, 84 of the 97 new daily persistent headache patients had no white matter abnormalities or infarct-like lesions on magnetic resonance imaging with a gender distribution of 56 women and 28 men. The mean age of onset of this white matter negative subgroup was 31.1 years. Of these individuals, 36% had cardiovascular/cerebrovascular risk factors and 44% had a history of migraine. Only 13 new daily persistent headache patients (nine women, four men) demonstrated white matter abnormalities on magnetic resonance imaging. None had infarct-like lesions. The mean age of onset of this white matter positive subgroup was 54.2 years, significantly older than the white matter negative population (p < .05). All new daily persistent headache patients in the white matter positive subgroup had cardiovascular/cerebrovascular risk factors and dual risk factors were noted in seven of 13 patients. Only 23% had a migraine history. Almost 40% of the patients in the white matter negative group were imaged 3 years after headache onset and at least six patients were imaged at least 9 years or more after onset of new daily persistent headache. Triggering events in both white matter lesion positive and negative populations were typical of the new daily persistent headache population as a whole and not specific to the presence or absence of brain imaging lesions except for a post-surgery trigger, which was significantly more likely to occur in the white matter positive group. Migraine associated symptoms occurred in 77% of the white matter negative subgroup compared with 46% of the white matter positive subgroup, which was a significant difference.
Conclusion
White matter abnormalities and infarct-like lesions do not appear to occur in primary new daily persistent headache patients. Only new daily persistent headache patients with risk factors (cardiovascular/cerebrovascular or migraine) developed white matter abnormalities on brain magnetic resonance imaging. No patient with new daily persistent headache developed infarct-like lesions. New daily persistent headache triggering events (outside of possibly post-surgery) or the presence of migrainous symptoms did not appear to enhance the development of white matter abnormalities.
Keywords
Introduction
New daily persistent headache (NDPH) is a unique headache syndrome based on its temporal profile, which is daily and persistent from onset. At present, the underlying pathogenesis of this syndrome is unknown. Large clinic and population-based studies of NDPH are scant. To date, the largest clinic investigations have only involved upwards of 50–90 subjects (1–5). Migraine, on the other hand, has been very well studied in extensive populations and there has been a recognition of structural brain changes that occur in this headache disorder, including the development of white matter abnormalities (WMA) and infarct-like lesions (ILL) (6,7). There is some suggestion these lesions are dependent on migraine frequency (7). Little data actually exist on the imaging findings in NDPH. Most suggest normal studies, although the two largest investigations to date did note WMA in a very small percent of sufferers, but did not define if any of these patients had risk factors for these lesions outside of the syndrome itself (1,2). It would be important to note if WMA or ILL develop in NDPH as, if they did occur, then this condition may have some shared pathogenesis with migraine. The presented clinic-based study is the largest investigation of NDPH patients to date. The current manuscript will focus on the question: Do NDPH patients develop WMA and ILL and do specific triggering events for NDPH as well as the presence or absence of migraine associated symptoms alter the occurrence of imaging abnormalities on brain magnetic resonance imaging (MRI)?
Methods
This was a retrospective analysis of patient medical records utilizing an electronic medical record system. All patients were seen at a headache specialty clinic by a single headache neurologist and diagnosed with NDPH during the time period of January 2009 to January 2013. All patients had to meet the ICHD 3 beta diagnostic criteria for primary NDPH (8). Each patient had to have a brain MRI with and without gadolinium and brain MR venography to rule out the major secondary causes of NDPH including intracranial hypotension from a cerebrospinal fluid leak and cerebral vein thrombosis. Some patients also had MR angiography of the intracranial and extracranial circulations if deemed by history (presence of a thunderclap headache) and examination (bruit) these studies were necessary. No patients were included in the investigation if they had secondary NDPH. The main imaging results to be verified were presence of WMA or ILL as defined by neuroradiology. All imaging reports were evaluated by the headache neurologist and most images were also viewed by the headache neurologist to verify the written results. Brain MRIs done through the primary hospital system in which the headache clinic was located were completed on a 1.5 Tesla unit and the brain images were acquired with 5-mm slices. This study was given approval with exempt status by the Geisinger IRB.
Statistical analysis
This was mostly a descriptive study but when statistical analysis was needed to look for any significant differences between NDPH subgroups the Chi-Square test was used for categorical data while the student T-test was used for continuous data. All p-values were two tailed and a p < .05 was considered statistically significant. SPSS for Windows (version 23) was used for statistical analysis.
Results
Altogether, 97 patients were diagnosed with primary NDPH during this time period. Gender distribution was 65 women and 32 men. Almost all patients were Caucasian except one patient was Hispanic while another patient was African-American. The mean average age of onset of NDPH was slightly younger in women than men, 32.4 years vs. 35.8 years, which was not a significant difference.
White matter negative subgroup (Table 1)
In total, 84 of the 97 NDPH patients had no WMA or ILL on neuroimaging (WMA negative subgroup) with a gender distribution of 56 women and 28 men. The mean age of onset of this WMA negative subgroup was 31.1 years. Of these individuals, 30 (36%) had cardiovascular/cerebrovascular risk factors for ischemic brain lesions, including hypertension, a history of current or past smoking or hyperlipidemia. Thirty-seven (44%) of these patients had a history of migraine; however, the percent of these patients who had migraine with and without aura could not be ascertained.
White matter positive subgroup (Table 1)
Only 13 NDPH patients (nine women, four men) demonstrated WMA on neuroimaging (WMA positive subgroup) and most were single to a few scattered lesions. None had ILL. The mean age of onset of NDPH in the WMA positive subgroup was 54.2 years; considerably older than the WMA negative subgroup and this was a significant difference (p = .00001). All NDPH patients in the WMA positive subgroup had risk factors for small vessel white matter ischemic lesions outside of migraine, including hypertension, hyperlipidemia, atrial fibrillation, inherited clotting disorders (factor V Leiden, prothrombin gene mutation), chronic smoking history and head trauma. The single patient with a head trauma history also had concomitant hypertension. Dual cardiovascular/cerebrovascular risk factors were noted in seven of 13 patients. There was a significant difference in the presence of cardiovascular/cerebrovascular risk factors between the WMA positive and negative NDPH subpopulations. No patient in the WMA positive subgroup had a stroke-like spell with headache. Only 3/13 (23%) of the WMA positive subgroup had a migraine history and thus migraine was unlikely the etiology of the WMA. The data set did not ascertain the number of patients who had migraine with and without aura. There was no significant difference between the percent of migraine sufferers in the WMA negative and positive NDPH populations.
MR venography studies in both WMA negative and positive populations were normal with no evidence of cerebral vein thrombosis or abnormal cerebral vein development issues. No significant cerebral or jugular vein stenosis was noted.
White matter lesions and NDPH triggering events (Table 2)
When looking at triggering events in the WMA negative population: none was noted in 44/84 patients (52%); flu-like illness/infection: 20/84 (24%); stressful life event: 9/84 (11%); surgical procedure: 5/84 (6%); other: 6/84 (7%). Triggering events in the WMA positive population: none: 7/13 (54%); flu-like illness/infection: 1/13 (8%); stressful life event: 0/13 (0%); surgical procedure: 4/13 (31%); other: 1/13 (8%). There were no significant differences between NDPH triggering events and the presence or absence of WMA except for the post-surgical triggering event, which was significantly more likely to occur in the WMA positive population (p = .004).
Migraine-associated symptoms and white matter lesions with NDPH (Table 1)
NDPH and WMA.
CV: cardiovascular/cerebrovascular; NS: not significant.
Neuroimaging and NDPH triggering events.
Timing of brain imaging and onset of NDPH
In regard to when neuroimaging was completed and the onset of NDPH, for the WMA negative subgroup: <1 year: 15 patients; 1–3 years: 46 patients; 4–6 years: 15 patients; >6 years (with an average of 9.6 years): eight patients. At least six patients were imaged 9 years or more after the onset of NDPH. A further analysis of the imaging data demonstrates that for >4 years from onset of headache: 23/84 patients (27%) were imaged and for >3 years after onset of headache 31/84 (37%) were imaged. In the WMA positive subgroup: MRIs were completed <1 year after onset: 5 patients; 1–3 years: four patients; 4–9 years: two patients; >10 years: two patients. When looking specifically at those imaged within 1 year of NDPH onset, there was no significant difference between the WMA negative and positive subgroups (p > .05) even though, by percentage, fewer WMA negative (18%) vs. WMA positive (38%) patients where scanned during this time period.
Discussion
We still know very little about NDPH even though it is accepted as a distinct primary headache syndrome by the ICHD (8). What has become very evident from multiple clinical studies of NDPH patients is that migraine-associated symptoms are very common in this condition and almost as prevalent as in migraine itself (1–5, 9,10). If you remove the temporal profile of onset, then many patients with NDPH in both pediatric and adult populations would meet the ICHD criteria for chronic migraine (1,5,8). A stressful life event, which is a key triggering incident for NDPH, has also been linked to the onset of chronic daily headache, including chronic migraine in population-based studies (11). Thus, is there a common pathogenetic connection between migraine and NDPH? That would be an important finding as we know very little about the etiology of NDPH. A possible further link between NDPH and migraine would be the presence of WMA and ILL on brain imaging.
The current investigation is the largest study to date of NDPH, looking at 97 patients in total, diagnosed over a period of 4 years. The goal of the present manuscript was to define if WMA and/or ILL develop in this unique condition. From the study results of clinic-diagnosed NDPH patients, it can be stated with reasonable evidence that WMA do not occur in NDPH unless underlying cardiovascular/cerebrovascular/cerebrovascular risk factors are present. It appears that WMA and/or ILL do not occur in primary NDPH. No WMA were present in 87% of the studied patients. In this subgroup, 44% had migraine and 36% had cardiovascular risk factors and still none developed lesions on neuroimaging. In the subgroup of NDPH patients (13%) who had positive WMA, 100% had cardiovascular/cerebrovascular risk factors and over 50% had two risk factors, while only 23% had a history of migraine. This would appear to suggest that it was the stroke-related risk factors causing the white matter lesions in the NDPH patient population rather than underlying migraine. This could also be suggested by the ages of the two subgroups in that the WMA positive group was older (54 years) than the WMA negative group (31 years); thus, more time for the risk factors to take effect. There was a significant difference in mean ages between the WMA positive and negative subgroups. Migraine-associated symptoms were common in all NDPH patients but significantly more common in the WMA negative subgroup, 77% vs. 46%. Thus, the presence of migrainous symptoms did not influence the occurrence of WMA and, if anything, may have had a protective effect, but again more likely had no impact.
Of note, 11 of the NDPH patients experienced aura with their headaches but none developed imaging abnormalities. As several studies have suggested, it is the migraine with aura subgroup who have a higher prevalence of WMA on neuroimaging, the NDPH with aura subset would conceivably have been the NDPH population with the greatest risk of developing neuroimaging changes if any linked pathogenesis with migraine existed (6,12). What data could not be ascertained from this retrospective chart review was the percent of our NDPH patients with a history of migraine with aura. Thus, how a precedent migraine with aura history altered the prevalence of WMA in our NDPH population cannot be determined.
Triggering events in both WMA positive and negative populations were overall typical of the NDPH population as a whole (as defined by previous investigations) and not specific to the presence or absence of WMA (3). The one interesting finding was the occurrence of post-surgical/procedure headache being significantly more predominant in the WMA positive subgroup, but this more likely had to do with the older age of this subpopulation and how that related to this particular triggering event rather than with the presence or absence of brain lesions. This will be explained further in a future manuscript on NDPH triggers. Overall, NDPH triggering events did not appear to alter the risk of the development of WMA.
In regard to the timing of imaging in the study population, there were more patients who were imaged 4 or more years after the onset of their NDPH (27 patients) vs. the number who were imaged less than 1 year into their headache syndrome (20 patients). When looking at those imaged within 1 year of onset of NDPH, there was a larger percent in the WMA positive subgroup (38%) vs. the WMA negative subgroup (18%), but this was not a significant difference. In addition, there were six patients who were imaged >9 years after headache onset in the WMA negative subgroup. The overall data would appear to be adequate time for the development of neuroimaging lesions, especially if WMA are headache frequency dependent as suggested by some of the migraine literature (7). The fact that lesions did not appear in the NDPH population without cardiovascular/cerebrovascular risk factors helps to suggest that they are not part of this headache syndrome complex.
The present study population had a fairly large percentage of comorbid migraine (41%), higher than that noted in previous studies (1–4). The reason for this is unknown and may just reflect a larger sample size. What is important to state is that a careful history was taken in each patient with a record of migraine to make sure that their current headache was a distinct change in pattern and frequency and thus met criteria for NDPH and did not reflect a rapid transformation to chronic migraine from episodic migraine.
Prior investigations of NDPH that have cited neuroimaging findings have not included any risk factor data for brain ischemic lesions. Peng et al. (2) in the previous largest study of NDPH from Taiwan looked at 92 patients, of which 73 had a brain MRI. Three patients had WMA while another three had ILL, but patient stroke risk factors for these lesions were not documented. This study did ascertain which NDPH patients had migrainous features and also who had a prior history of migraine but did not stratify these data to which individuals developed neuroimaging findings. Robbins and colleagues (1) documented 68 NDPH patients who had a brain MRI, of which five subjects demonstrated punctate WMA, while another two patients showed ILL (thus, 7% with abnormalities). However, there was no mention of patient cardiovascular/cerebrovascular risk factors to indicate if these lesions were probably linked to NDPH alone or from hyperlipidemia or chronic smoking, for example. Other prior studies of NDPH, which have documented neuroimaging findings, have not discussed WMA either because they were not present or because they were not felt to be clinically significant (3,4,10). Currently, the prevalence rates of WMA in migraine sufferers range from 4% to 59% in select studies and these appear independent of cardiovascular risk factors (6). ILL mostly involve the posterior circulation and occur in up to 8% of migraineurs vs. controls (6). Finally, WMA in migraine may depend on attack frequency of ≥1 per month (7). The etiology of such lesions is unknown at present. If NDPH shares any underlying pathogenesis with migraine and WMA are headache frequency dependent, then one could hypothesize that WMA should occur more commonly in NDPH as headaches are daily and persistent. On the converse, if WMA and ILL are absent in NDPH, which they appear to be, then that could or should indicate that this primary headache condition does not share major pathogenetic mechanisms with migraine. That is important to note, in that it would help guide further studies of pathogenesis and neuroimaging. It could also explain why treatment of NDPH with migraine protocols does not work in the majority of patients (1,3).
Does the lack of development of WMA or ILL tell us anything about NDPH pathogenesis? Several factors thought to be involved with lesion pathogenesis in migraine, including neuropeptide and cytokine release, which have also been linked to NDPH pathogenesis (glial activation and tumor necrosis factor alpha production), may still be occurring in both disorders, but based on the present study results are not leading to WMA production in NDPH (13,14). Also from the present study findings, clotting factor deficiencies, platelet aggregation issues and cortical spreading depression, which have been linked to WMA pathogenesis in migraine, are not likely to be involved with NDPH pathogenesis (14). One has to consider the possibility that NDPH may have WMA or ILL that develop early in the course of the illness but then disappear over time. Only 18% of our WMA negative study population was imaged within 1 year of headache onset. The disappearance of WMA and ILL has been documented to occur in migraine (15,16). The goal should be to image the NDPH patient early in the course of their illness but, in many cases, these patients do not reach the headache specialist or general neurologist until many years into their pain syndrome.
The main issue about the results of the present study is that every patient did not have their neuroimaging completed at the headache clinic institution (>90% did) and thus one cannot assume that, for the studies done elsewhere, those images were read by a neuroradiologist. Thus, some subtle WMA or ILL could have been missed. The author did try to review most of the studies personally to verify the negative and positive findings and of those perused the findings were correct, but not every study was visualized (some of the images could not be collected). It is possible some WMA or ILL did occur in NDPH patients without cardiovascular/cerebrovascular risk factors and were missed, so it cannot be stated with maximum certainty this never occurs. In addition, our institution did not utilize the thin section (3 mm) MRI imaging that has been utilized in prior studies looking for WMA and ILL in migraine; thus, some lesions could have been missed based on imaging technique (14). Other limitations of this study include the retrospective design, thus limiting some of the data on comorbid history of migraine with aura and also not being able to completely rule out that some of the patients diagnosed as NDPH actually had rapidly transformed chronic migraine. This, however, is not likely, with one headache specialist being the sole diagnostician for the entire study. Migraine- associated symptoms were significantly more common in our WMA negative subgroup; however, the WMA positive population was older. With age, migraine-associated symptoms may diminish in migraine patients. An age-related reduction in migrainous symptoms has yet to be studied in NDPH patients (1,2). Without multivariate analysis to account for this confounder, this specificity may not be valid and thus could be considered a possible limitation of our study findings.
There was also a lack of a control population (such as chronic migraine), which would have been imaged at the same time as the NDPH patients to assess prevalence of WMA and ILL from MRI brain images being performed at the same institution.
Conclusion
NDPH is a syndrome in its relative infancy compared to migraine, cluster headache and tension-type headache and thus reporting issues about its basic presentation including its neuroimaging patterns are still essential and will hopefully help lead to a better understanding of its underlying pathogenesis. From the current clinic-based study of NDPH, it appears that WMA and ILL do not occur in this condition even when patients have been imaged for more than a decade after headache onset. Only those NDPH patients with cardiovascular/cerebrovascular risk factors or a history of migraine developed WMA on MRI. This information suggests a disparate pathogenesis between NDPH and chronic migraine, in which these lesions can be quite prevalent. However, based on the present study design limitations, these results will need to be confirmed by future investigations before it can be absolutely stated that WMA and ILL do not occur with primary NDPH. What is suggested is a systematic, prospective study of NDPH, which incorporates neuroimaging structural biomarkers to verify our surprising results from this retrospective study, which did not account for pre-morbid headache frequency and migraine diagnosis, which can be major confounders.
Clinical implications
Patients with primary NDPH do not appear to develop white matter abnormalities or
infarct-like lesions on brain MRI. Triggering events for NDPH do not influence the presence or absence of white matter
abnormalities on brain MRI except possibly a post-surgical trigger; although this is
more likely an age related phenomenon. The presence of migrainous symptoms does not appear to enhance the development of
white matter abnormalities in patients with NDPH.
Footnotes
Declaration of conflicting interests
The author declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
This work received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.