Characterization of migraineurs presenting interictal widespread pressure hyperalgesia identified using a tender point count: a cross-sectional study

Migraineurs were recruited for the cross-sectional survey from March 2015 to September 2016 at Toriyama Clinic (local primary and secondary headache clinic in Komoro City, Nagano Prefecture, Japan, with a target population of around 100,000). All subjects had a standardized interview and careful clinical examinations were conducted by the first author to determine inclusion and exclusion criteria. This study was approved by the institutional review board of Shinshu University School of Medicine (approval number 3552-1). All patients provided informed consent prior to enrollment in the study.

Consecutive patients were considered eligible for inclusion in this study using the following inclusion criteria: episodic migraine as defined by the International Classification of Headache Disorders (ICHD)-III β criteria [27], aged 18-65 years, and subjects that had been migraine-free for at least 48 h to avoid the influence of migraine-related acute allodynia. Exclusion criteria were as follows: comorbid diagnosis of other primary or secondary headache, including trigeminal/occipital neuralgia, sinusitis, chronic migraine or medication overuse headache; history of cervical spondylosis; comorbid diagnosis of any local or systemic disease that may cause headache; and chronic widespread/regional pain or fibromyalgia. All patients were permitted to take migraine abortive and prophylactic drugs.

The comprehensive interview form included diagnostic questions based on the ICHD)-III-β criteria [27] for diagnosis of migraine, demographic characteristics, and migraine-specific features. Migraine-specific variables included migraine subtype, age at onset, duration of migraine illness, duration of migraine attacks, migraine attack frequency, headache intensity, headache disability, presence of cutaneous allodynia, presence of depression, migraine-associated symptoms, and use of headache medicine. The age at onset was divided into 5 categories. Duration of migraine history was divided into four categories. Duration of headache attack was classified into four classes. We measured migraine frequency by the number of migraine days per year and used four frequency classes. Medication intake was determined as clinical history. For women patients, no information of menstrual cycle was obtained in this study.

As no standardized measures are currently available to determine the presence of interictal WPH without using QST, the existence of WPH using TPC by MTPS for FM was investigated [28]. The author (TT) performed MTPS in a non-blinded manner. MTPS was performed systematically over the 18 FM tender points according to the published guidelines [28] in all migraineurs and controls. Digital pressure was incremented until the examiner’s nail bed became white (force of about 4 kg) [29]. Participant’s pain response at each site was classified as no pain, mild pain (complaint of pain without grimace, flinch, or withdrawal), moderate pain (pain plus grimace or flinch), and severe pain (pain plus marked flinch or withdrawal) [30]. Although TP with mild pain was rated as positive TP in the original guidelines [28], a mild pain response was considered potentially ambiguous because the objective of this study was to evaluate the relationships between WPH and migraine characteristics. Therefore, the cutoff response was set as greater than or equal to moderate pain according to Amris et al. [31] who suggested that a hyperalgesic pain response at palpation should be required for TP to be considered a primary identifier of pain hypersensitivity. Thus, the sum of the positive TPs, above cutoff thresholds set at ≥moderate, was defined as the TPC in this study. The maximum TPC was 18 (9 designated spots × right/left). With regard to the order of examinations, the first author checked the front of the body (in the order low cervical, second rib, lateral epicondyle and knee), and then examined the back of the body (in the order occiput, trapezius, supraspinatus, gluteal and greater trochanter) with the subject in a standing posture. In this study, only one examiner (TT) rated all TPs, indicating negligible inter-rater reliability. Prior to the study, a digital weight scale was used to train the rater.

As there were no standardized measures for evaluating WPH using MTPS, selection of threshold of TPC for WPH was needed. We recruited headache-free healthy controls to define TPC cutoff values for WPH in migraineurs. Control subjects that visited to Toriyama Clinic for health checkups were examined for TPs by the same rater blinded to their pain. After obtaining information on the current pain, subjects complaining of headache and other pain were excluded. To classify a subject’s response as WPH in a standardized manner, the cutoff value for WPH responder was defined as the TPC below which 95% of control subjects responded. Patients were classified into the migraine with interictal WPH subgroup, if they had a positive TPC above the threshold.

The comprehensive interview included a 12-item questionnaire similar to the items of the validated Allodynia Symptom Checklist [2]. Migraine patients that answered “yes” on two CA items were scored as allodynic [3].

The intensity of headache was derived based on a numerical rating scale [32], with 0 representing “no pain” and 10 representing “the worst pain possible.” Mild headache was defined by scores 1–4; moderate by scores 5–7; and severe by scores 8–10 [33].

Headache Impact Test-6 [33, 34] was used to measure headache-related disability. Four groups have been derived to aid in the interpretation of Headache Impact Test −6 scores. In this study, patients were divided into two groups according to headache-related disability: “no severe impact” (score < 60) and “severe impact” (score ≥ 60).

Depressive symptoms were evaluated by the self-rating depression scale [35]. Patients were diagnosed as having depression, if they had a self-rating depression scale score ≥ 48 [36].

Descriptive data are expressed as means ± standard deviation or percentages. Differences in proportions were tested using the x² tests or Fisher’s exact test as appropriate. Differences in continuous data between the groups were analyzed by using Student’s t-test for normally distributed variables. The Mann-Whitney U test was used to compare differences between non-normally distributed variables. Logistic regression analysis was used to test which variables were independently associated with icterictal WPH. In logistic regression analysis, all variables for which the P value was less than 0.15 on the univariate analysis were included in the model, and variables with lower significance (higher P value) were sequentially removed from the model. To confirm whether the MTPS was administered consistently throughout the study period, the Mann-Whitney U-test was used to compare differences between TPC of the first half of the patients and those of the last half of the patients. Spearman’s correlation coefficient was used to investigate the test-retest reproducibility of MTPS in those examined on the initial and second interictal TPC at intervals of more than 1 month to within 6 months. Kappa values was used to investigate the intra-rater agreement of the initial and second diagnoses for positive WPH in the same patients with no change in medication at the same intervals described above. All P-values were two-sided, and P < 0.05 was taken to indicate statistical significance. All statistical analyses were performed with the free software, EZR [37].

All prominent symptoms of migraine are known to result from activation of the thalamus and/or cortex [62‐68] and they are associated with acute CA but not interictal WPH. Acute CA may be associated with dysfunction of regions involving pain modulatory systems affecting sensitization of the thalamus and/or cerebral cortex. In contrast, interictal WPH may be caused by dysfunction of different regions involving pain modulatory systems related to depression.

As interictal WPH is a feature of FM, putative candidates for the central nervous system regions involved in the mechanism underlying interictal WPH are presumed to be as follows.

The somatosensory cortex, with cerebral blood flow positively correlated with headache frequency [69], may be one of the regions related to interictal WPH. Hodkinson et al. [69] reported a significant increase in regional cerebral blood flow in the primary somatosensory cortex of migraineurs compared with healthy controls. They showed that the cerebral blood flow in the primary somatosensory cortex was positively correlated with headache attack frequency. However, it was unrelated to the duration of illness or age of the patients. In our study, the distribution of interictal positive TP corresponded to the distribution of homunculus of the primary somatosensory cortex. The prevalence of interictal WPH determined by positive TPC was related to the frequency of headache attacks, but not to current age or duration of disease. Based on these findings, the interictal increase in regional cerebral blood flow of the primary somatosensory cortex may cause interictal WPH. Patients with migraine and interictal WPH may include those with a high expectancy and hypervigilance for pain due to the nature of MTPS [61].

Blood oxygenation level-dependent signals of the posterior thalamus elicited by extracephalic skin stimuli were greater during migraine attacks than the interictal period. This suggests that sensitization of trigeminovascular thalamic neurons induces multimodal allodynia and hyperalgesia throughout the body [67]. However, it is unclear whether interictal blood oxygenation level-dependent signals in response to extracephalic skin stimuli in the trigeminovascular thalamic neurons of migraine patients are greater than those of headache-free controls. Any differences in blood oxygenation level-dependent signals of posterior thalamic neurons in response to extracephalic stimuli between migraineurs and controls may be related to interictal WPH. Further investigations are needed to determine whether blood oxygenation level-dependent signals are different between subjects with and without interictal WPH.

Diffuse noxious inhibitory control [70] failure may also be involved in extensive pain hypersensitivity, as chronic migraine patients frequently report diffuse somatic pain [73], suggesting that the same factors that facilitate the evolution from episodic to chronic migraine may also predispose patients to FM comorbidity.

The brainstem involves nuclei closely related to dysfunction of the descending modulatory system [71]. Studies on periaqueductal gray matter [72], rostral ventromedial medulla [73] and nucleus cuneiformis [74] suggested that disorder of descending pain control plays a key role as a mechanism related to interictal hypersensitivity to pain in migraine patients. Moulton and colleagues reported interictal hypofunction of the nucleus cuneiformis in migraine patients using functional MRI [74]. They suggested that nucleus cuneiformis function could potentially serve as a diagnostic measure in migraine patients, even when not experiencing an attack. Based on the viewpoint that migraineurs with interictal WPH belong to a migraine subtype with dysfunction of pain modulatory systems sharing FM, we hypothesize that the brain region including the dysfunctional pain modulatory system underlying interictal WPH is the medulla. Recent morphological studies showed that migraineurs with more severe symptoms of allodynia and lower heat-pain thresholds show greater volume loss in specific brainstem regions (midbrain, medulla, and cerebellar peduncles), providing further evidence that the brainstem has a modulatory function in the development of central sensitization in migraine [75] and brainstem volume reduction is correlated with TPC in migraine subgroup with FM [76]. Notably, in another report indicating brainstem volume reduction in migraine patients, Bilgiç et al. reported that there was no correlation between brainstem volume and duration of disease [77]. As interictal WPH is associated with younger age at onset rather than duration of disease, structural changes in the brainstem associated with interictal WPH, if present, may be related to the process of migraine susceptibility.

Possible structural changes in the brainstem may be not simply a secondary phenomenon and may be also related to a genetically determined structural alterations in the migraine brain [78]. Schwedt et al. [17] suggested that lower pain thresholds beyond the trigeminal region would likely be mediated by brainstem regions of the descending pain modulatory system, such as the rostral ventromedial medulla, periaqueductal gray matter, and/or nucleus cuneiformis. Consistent with this hypothesis, functional imaging studies of migraineurs have shown that these brainstem regions have atypical pain-induced activation and atypical resting state functional connectivity, and implicated these brainstem regions in central sensitization [72, 74, 79].

Among the above mechanisms, we assume that the medulla is substantially involved as a hub structure related to interictal WPH, because similar brainstem volume reduction and structural alteration [75, 76] were observed in migraine and FM, and it includes the serotonergic system from the rostral ventromedial medulla, which has been implicated in migraine and FM pathophysiology [80]. There have been no previous reports regarding the structural changes in the medulla associated with interictal WPH. Further investigations to examine these association are required.

Schwedt et al. [17] hypothesized that the mechanisms underlying these lower pain thresholds could also predispose migraineurs to subsequent migraine attacks based on their observation of the positive correlations between pain thresholds and time to next migraine attack. They concluded that interictal hyperalgesia may be a part of the process of migraine development, and not simply a secondary phenomenon suggesting that the mechanisms underlying interictal hyperalgesia may predispose migraineurs to their next migraine attack and may be a marker of migraine activity and a target for treatment. Giamberardino et al. [58] showed that increased migraine frequency enhances both hyperalgesia and spontaneous FM pain reversed by effective migraine prophylaxis. Longitudinal investigations are necessary to determine whether effective migraine prophylaxis leads to reduce interictal WPH. In 1999, Okifuji et al. [81] concluded that extensive dysregulation of pain modulation is important for a substantial minority of recurrent headache patients, who seem to be quite similar to FM patients, and that differential treatment planning targeting generalized hyperalgesia may be useful for more effectively treating headache patients exhibiting generalized hyperalgesia. For migraines with interictal WPH, a possible subtype of migraineurs sharing FM feature, pharmacological and non-pharmacological [10, 82] effective interventions for FM may be considered. Candidate drugs to improve interictal WPH include current agents, such as pregabalin [83] and duloxetine [84], and/or future agents capable of improving the descending modulation system. Diagnosing interictal hypersensitivity to experimental pain using TPC at the primary clinic where headache patients first visit may prompt general practitioners to start tailored treatment considering central sensitization and underlying dysfunction of the pain modulatory system from an early stage of migraine.

Although the diagnostic value of MTPS is decreasing in the field of FM [85], TPC is worth revaluating as a practical alternative for QST to diagnose interictal WPH in migraine. We expect general practitioners to be able to identify migraine patients with interictal WPH using TPC without QST, like a McBurney sign for the diagnosis of appendicitis. If the primary care physician can diagnose the presence of interictal WPH, many migraineurs will be able to receive tailored treatment considering dysfunction of the pain moderation system underlying interictal hyperalgesia, from the early stage of migraine. Finally, we endorse the conclusion of de Tommaso et al. [86] that “Improved insight into the interictal dysfunction of temporal information processing in individuals with migraine will pave the way for novel therapeutic targets, and could herald improved migraine management.” The first step toward “improved insight” is for general practitioners examining headache patients to routinely diagnose the presence of WPH in migraineurs using practical methods.

The strengths of this study include the large cohort size, the well-defined migraine status, detailed information on TP characteristics in comparison with healthy controls, and that the presence of WPH was defined in a standardized manner, i.e., the cutoff value of WPH responders was defined as TPC to which 95% of pain-free healthy controls showed a response. Most importantly, this is the first study to investigate the prevalence of interictal WPH in episodic migraineurs using TPC and demonstrating that the independent factors associated with an increased interictal WPH prevalence were female gender, higher frequency of migraine attacks, and younger age at onset. However, this study had some potential limitations. First, only one rater was involved, resulting in lack of evaluation of inter-rater reliability and limited generalizability. Second, the standardized TP assessment was conducted as part of the clinical examination, and therefore the examiner could not be blinded to headache diagnoses of subjects other than controls. It is possible that the examiner would expect hyperalgesia at TP in migraine patients. Therefore, the results must be confirmed in a blinded manner. Third, we recruited our patients from a secondary headache clinic seeking further medical therapies. As refractory patients visiting a tertiary headache center and mild headache patients improving with non-prescription drugs were excluded, the headache intensity, disability, and depression of our cohort may have been biased toward homogenized moderate state compared to those treated in the general population. This may explain why neither interictal WPH nor acute CA was associated with headache intensity. Therefore, further multicenter studies including individuals from the general population would help to reconfirm the results. Fourth, the cross-sectional nature of this study prevented us from drawing conclusions about causality in the relationship between interictal WPH and high attack frequency. Fifth, we did not consider the time interval to next migraine attack. A previous study suggested that episodic migraineurs have reduced pain thresholds in the 24 h prior to migraine onset [60]. Although we cannot entirely exclude this as a confounder, it is unlikely that a substantial proportion of our subjects (median frequency and mean duration were 2.0 attacks month and 24 h/attack, respectively) developed a migraine within 24 h of examination.

Although this was a non-blinded study by one rater in a small city in Japan, the demographic and clinical features of migraine, including prevalence of acute allodynia, were consistent with those of large scale population studies performed in several countries. In addition, the good test-retest reproducibility of TPC using TP cutoff threshold set at ≥ moderate pain response and substantial intra-rater agreement in WPH diagnosis were confirmed. Therefore, it seems that the results of this study were generalizable. As this should be considered a preliminary study, future research is warranted to evaluate whether the results could be replicated by other raters in cohorts of other nationalities and ethnicities. More refined randomized controlled trials, focusing on methodological issues relating to the determination of WPH, are warranted to elucidate the true relationship between interictal WPH and susceptibility to higher frequency of migraine attacks and to migraine onset at a younger age.

Future research should focus on identifying the pathophysiological mechanisms linking migraine and interictal WPH. Given the results of this study, patients were divided into four subtypes: 1) migraine with both acute CA and interictal WPH; 2) migraine with acute CA only; 3) migraine with interictal WPH only; and 4) migraine with neither acute CA nor interictal WPH. Direct comparison of high resolution functional MRI results between the four groups in a large-sized cohort, may delineate functional and structural alterations and characteristics of the brain related to interictal WPH and acute CA, which would lead to elucidation of the pathophysiology of migraine. As WPH is not related to the duration of disease, it is not due to barrage of headaches but is presumed to be related to the structural changes in the “migraine brain” [78]. Direct comparison using MRI between the subgroups with and without WHP may also reveal genetically determined functional and structural alterations in the “migraine brain.” New interventions can then be tested for their capacity to normalize the anatomical and functional alterations associated with migraine subtypes. However, these concepts await experimental evidence.