Triptanophobia in migraine: A case-control study on the causes and consequences of the non-use of triptans in chronic migraine patients.

Abstract


Introduction:
Acute patient relief is a keystone of migraine management.Symptomatic treatment aims to abort migraine episodes and restore patients' normal function, minimizing migraine morbidity and impact.To date, different symptomatic treatments have been used 1 , representing triptan discovery a new era with more robust studies and better clinical results 2 .
Triptans are indicated in acute episodes of migraine with moderate to severe intensity, inadequate response or contraindication to other analgesics 3,4 .Pearls of triptans include the effective control in 9 to 48% of migraine attacks 5 , avoiding emergency department visits 6 .Triptans should be prescribed in every person with migraine, in particular those with chronic migraine (CM) 4 .
Pitfalls of triptans are their adverse effects, including the possibility of cardiovascular events 7 .Coronary artery diameter may decrease up to 26% in healthy subjects 8 , leading to the contraindication of triptans in patients with ischemic coronary artery disease or uncontrolled hypertension 9 .However, when used in healthy patients, they seem reasonably safe 7 .Frequency of side effects ranges from 17-72% 10 , being the most frequent nausea and dizziness 11 .
Even though triptans are effective, safe in selected patients, and generally well tolerated 12 , up to a fourth of CM patients have never used them 13,14 .The possible reasons for the non-use of triptans seem multifactorial, including formal contraindications, cost, and triptanophobia, among others.
Contraindications seem infrequent in most people with migraine 12 .Cost should not be the reason, as Spanish healthcare system subsidizes them 15 .Therefore, triptanophobia could be an important reason for not using triptans.Triptanophobia could be de ned as the excessive and inadequately justi ed concern of health professionals about the potential risks of triptans.
In the present study, we evaluated causes and consequences of the non-use of triptans in CM sufferers.
We analyzed frequency of triptan use in CM patients.We compared, between triptan users and triptan naïve patients, the presence of contraindications, frequency of vascular risk factors, and differences in management prior to the referral to a headache unit.

Material And Methods:
We conducted an observational analytic study with a case-control design, following the STrengthening the Reporting of OBservational Studies in Epidemiology (STROBE) statement 16 .The study took place in the Headache Unit of Valladolid University Hospital, a third-level academic hospital with a reference population of 280.000 people.It receives patients from both specialized care and primary care.We recruited patients between May 2013 and May 2019, using a non-probabilistic sampling method, screening all consecutive patients.We included patients that were: 1) Firstly referred to our headache unit, 2) De nitely diagnosed of CM according to the International Classi cation of Headache Disorders (ICHD) operating at the time, either 2nd version (ICHD-2) 17 , beta 3rd version (ICHD-3b) 18 or 3rd version (ICHD-3) 19 ; 3) Diagnosed of CM for the rst time; 4) Older than 14 years old; 5) Agreed to participate in the study and signed the informed consent form.Patients were considered cases if they were triptan naïve and controls if they had previously used triptans.
Patients were excluded if they had: 1) Hypersensitivity to triptans 2) a serious systemic disease with higher impact than migraine that could be life-threatening 3) severe psychiatric disorders, including bipolar and related disorders; dissociative disorders, schizophrenia spectrum and other psychotic disorders; drug abuse; addictive disorders, or intellectual disability 4) concurrent administration of a monoamine oxidase inhibitor 9 5) pregnancy or breastfeeding status.

Objectives:
The rst endpoint of the study was to analyze if non-use of triptans was justi ed by the presence of formal contraindications or previous adverse events.The second endpoint was to compare if triptan naïve patients had a higher frequency of vascular risk factors (hypertension, hyperlipidemia, diabetes, smoking or prior smoking or obesity), use of contraceptive drugs or older age.The third endpoint was to compare the clinical situation of patients and previous management in triptan naïve and current triptan users.In this analysis, we excluded patients that discontinued triptans due to inadequate tolerability.The compared variables included age, age of onset, sex, months of CM prior to the consultation, number of headache days per month, migraine days per month, number of acute medication days per month, frequency of medication overuse headache (MOH), and prior use of prophylactic drugs.Figure 1 depicts the study design.

Intervention:
For the rst endpoint, every rst visit we assessed the presence of contraindications to triptans, including chronically uncontrolled hypertension, coronary artery disease, history of stroke and peripheral vascular disease.
Concerning the second objective, presence of vascular risk factors was systematically evaluated, including hypertension (systemic blood pressure higher than 140/90 mmHg in at least two prior determinations), hyperlipidemia (total cholesterol > 220 mg/dl, LDL > 150 mg/dl or HDL < 40 mg/dl, triglycerides > 200 md/dl), diabetes (fasting blood glucose level > 126 mg/dl on two separate tests, HbA1c > 6.5%, blood glucose level > 200 mg/dl with diabetes symptoms or blood glucose level > 200 mg/dl after oral glucose overload), current smoking habit, prior smoking habit (if the smoke cessation occurred > 3 months before) and obesity (Body Mass Index > 30).As in some studies, older age 20 and the use of contraceptive drugs 21 have been associated with a higher frequency of vascular events in migraine patients, we also compared mean age, frequency of patients older than 50 years old and frequency of patients older than 65 years.We compared frequency of patients under contraceptive therapies in both groups.
With regards to the third endpoint, we compared triptan naïve patients and patients using triptans at that moment.We gathered information about demographic data, including referral source considering primary care or secondary care, gender, age of onset of migraine, and age when CM started.We analyzed the clinical situation of the preceding month, in terms of number of headache days per month, migraine days per month and number of days using acute medication.We evaluated if patients ful lled MOH criteria according to ICHD-3 criteria 19 , by calculating it mathematically based on the number of acute medication days, as 10 or more days of triptan, opioids or combination analgesics use per month; or more than 15 days of analgesics or non-steroidal anti-in ammatory drugs (NSAIDs) days per month.We also studied the use of prior prophylactic migraine treatments, de ned according to the local guidelines 22 .We analyzed time in months since the patient experienced > 15 headache days per month and > 8 migraine days per month.
This study was approved by the local research ethics committee with the number PI 14-152 and was conducted in accordance with the Declaration of Helsinki.All patients were informed prior to the study and signed an informed consent form.

Statistics:
We present qualitative data as frequency and percentage and quantitative data as mean and standard deviation or median and interquartile range if the distribution was not normal, according to Kolmogorov-Smirnov test.In the rst endpoint, we describe frequency and percentage of contraindications in both triptan naïve patients and triptan users, differentiating current and former triptan users.
For the second and third endpoints, we employed the Chi-Squared test to compare qualitative variables and the Student-t test to compare qualitative and quantitative variables, providing degrees of freedom.We did a logistic and linear regression analysis.We considered an alpha error of 5%.We corrected for multiple comparisons by using Bonferroni method.We also did a survival analysis through a Kaplan-Meier analysis and Cox regression log-rank test, presenting the pertinent graphic representation.We anticipated as potential confounder in frequency of MOH the prior use of prophylactic drugs and adjusted for it.We also anticipated that patients being referred earlier to our headache unit might have essayed preventive medications less frequently.Missing data was managed by complete-case analysis.We did not estimate sample size in advance.Statistical analysis was conducted with SPSS v.26 (IBM Corp., Armonk, NY).

Results:
During the study period, 4138 patients were screened, ful lling inclusion/exclusion criteria 941 of them (22.7%),being 811 of them (86.2%)female.Mean age of migraine onset was 19.2 years (9.7), mean age at the moment of the rst visit was 39.9 years (13.9),duration of CM prior to diagnosis were 36.4 (50.4) months in mean and 12.0 months in median [6-36].Referral source was primary care in 574 patients (61.0%) and general neurology or other specialties o ces in 367 (39.0%).
Mean number of headache days per month was 24.1 (5.2), being 15.0 [9-20] of them migraine headaches in median.Patients were taking acute medications 16.6 (9.1) days per month in mean.In the total sample, 247 patients were under triptan use (26.2%) and 694 (73.8%) were not.Triptans had been previously used and stopped due to adverse events in 116 (12.3%) patients, leading to a total of 578 triptan naïve patients (61.4%).Triptans were used 9.6 (7.0) days per month in mean.

Contraindications for triptan use:
Formal contraindications for triptan use were found in 23 patients in the whole sample (2.4%), being one of those patients under triptan use.Frequency of contraindications was 22/694 subjects, leading to 3.2% of all triptan naïve patients and former triptan users.The contraindications included uncontrolled hypertension in fourteen cases, severe peripheral artery disease in four cases, antiphospholipid syndrome with associated peripheral vascular disease in three cases, and prior stroke in two patients.One of the patients with severe peripheral artery disease was under triptan treatment.Vascular risk factors, contraceptive use and old age: We found no statistically signi cant differences in the frequency of vascular risk factors between triptan users and triptan naïve patients (all corrected p > 0.1).Table 1 shows the relative frequency of each vascular risk factor in triptan users vs. triptan naïve patients.Discussion: In this study, we evaluated if the non-use of triptans was justi ed by the presence of contraindications or prior adverse events.We analyzed if triptan naïve patients had a higher frequency of vascular risk factors associated with stroke or cardiovascular conditions.Finally, we analyzed differences in clinical situation and previous management between the groups of triptan users and triptan naïve patients.We speci cally studied CM patients because in that population, there is no doubt on the need of a triptan prescription 3,4 .
The main results of our study were that the percentage of triptan naïve patients was surprisingly high (74%), that the frequency of formal contraindications was sparse (3%), that the percentage of patients that discontinued triptans because of adverse events low (12%), and that the non-use of triptans was associated with a worse clinical situation and worse previous management, indicated by a higher frequency of acute medication use, a higher frequency of MOH and a lesser prior use of prophylactic drugs.
Headache disorders are the second cause of years lived with disability worldwide 23 .CM patients are frequently underdiagnosed and undertreated, not only in primary care, but also at the specialized level 13 .
Only 40.8% of CM patients have a consultation with a healthcare professional for headaches, and 4.5% receive adequate diagnosis and treatment 24 .
The main options in the acute treatment of migraine attacks are NSAIDs, ergots and triptans 25 , although recently ditans and gepants have been used 26,27 .Although NSAIDs are the most widely used treatment in migraine attacks 28 , triptans are more migraine speci c and its e cacy rate is overall higher than NSAIDs or ergots 29,30 .In our sample, only a fourth of all CM patients were under triptan use.National studies conducted in Spain showed a frequency of use of 17%-48% 14,15 , compared to 14 to 25% in other countries 13,31 .It is also surprising that opioids, which are not speci c migraine treatments and have more frequent and serious adverse effects, are frequently used in migraine 32 , showing that opioidophobia seems uncommon.
Adverse events of triptans are related with their effect on 5-HT 1B receptors 35 , being contraindicated in subjects with coronary artery disease, coronary vasospasm, prior cerebrovascular events, peripheral arteriopathy or uncontrolled hypertension 7 .Despite its potential risks, serious cardiovascular effects occur in < 1/1.000.000exposures, the most severe events being stroke and myocardial infarction 36,12 .In our sample, only a minority of patients without triptan use presented contraindications (3.2%), so contraindications seem insu cient to explain the non-use of triptans.
The main vascular risk factors associated with both stroke and ischemic coronary disease are hypertension, hyperlipidemia, diabetes, smoking or obesity 37 .Therefore, we analyzed if the non-use of triptans could be associated with an excessive concern for vascular events.Nonetheless, frequency of vascular risk factors was similar between the two groups, without signi cant differences.Evidence suggests that migraine patients have increased vascular risk 38 , and especially those with aura, those who smoke, those treated with contraceptives, and those older than 40 years old 21 .Therefore, some authors do not recommend triptans in old patients 39 .Nevertheless, we could not nd differences in age, use of oral contraceptives, or seniority.
Triptan non-use was not justi ed by a worse e cacy compared with NSAIDs or ergots, presence of contraindications, or an increased frequency of vascular risk factors.Two other explanations seem possible: a worse tolerability pro le or treatment expense.Concerning tolerability, frequency of adverse is estimated around 17-41% in ergots 40 , 5-28% in NSAIDs and 40% in triptans 7 .In our sample, adverse events led to triptan discontinuation in 12% of patients.Adverse event pro les differ: NSAIDs typically cause gastrointestinal AE 11 , and ergot derivatives cause nausea or vomiting 41 .The most frequent triptan AE are nausea (10%), dizziness (4%), fatigue (3%) and feeling of heaviness (3%) 11 .
Triptans can cost between 4 and 62 times higher than NSAIDs.Still, the Spanish healthcare system covers between 0%-60% of the total price 42 .Retired workers with less than 18.000 € of annual income only contribute a 10% of the total cost, with a maximum amount of 8.23 € per month 42 .The total annual cost of CM is estimated to be 12,922 € per patient.Only 43% corresponds to the direct cost 15 .Despite the higher cost of triptans, they have proven to be cost-effective, due to the decrease of direct costs related with emergency department 6 consultations and indirect costs 13,43 .Therefore, neither tolerability nor cost seem to fully explain the underutilization of triptans.
All CM patients should be under prophylactic medication 44 .In our sample, only 51% of the patients were under preventive treatment, in line with comparable series (49%) 15 and more than primary care-based series (1.6-14%) 13 .One of the most remarkable ndings of our study was the signi cant difference of prior prophylactic treatment in patients treated with triptans compared with triptan naïve subjects, being almost twice as frequent (79% vs. 35%).
We examined whether differences in triptan use were related with a delayed referral to Headache specialized care.In our setting, general practitioners can refer patients directly to our Headache Unit without prior consultation with a general neurologist, reducing the wait time.In our sample, we could not explain differences in time prior to the referral by the use or non-use of triptans, being 36 and 38 months in both groups.
We used the term triptanophobia to de ne the excessive and inadequately justi ed concern of health professionals about the potential risks of triptans.It should not be interpreted in a pejorative or dismissive way.The reasons for the non-use of triptans seem multifactorial; however, in the present study we observed that the relative weight of many of them seem insu cient, so there must be some degree of triptanophobia.
Causes of triptanophobia can be related to education on headache medicine.Headache education is not an obligatory rotation during neurology residence curriculum.The lower percentage of preventive use in triptan naïve patients may also suggest this worse management of CM patients.It could also be motivated by the atypical commercial presentation, in blisters of two to nine pills, unlikely for a medication that should be used frequently.
Our study had several limitations: it is a single-center study with an unusual referral pathway.In some patients, there could exist some degree of memory bias, but we reviewed referral charts and doublechecked prescription history in every case.The study could fail to detect differences in the frequency of vascular risk factors, given that the sample was not sized primarily for that, but all the relative frequencies were similar with the sole exception of hypertension (4.9 vs 7.6%).Prior use of preventive medications could also in uence the results, but all included patients met CM criteria when included, suggesting that almost all preventive drugs failed in them.Triptanophobia is a new term that may not fully correspond to the classic de nition of phobia.We use this concept in a practical way to try to explain the underutilization of triptans, without assuming the negative connotations of phobia.Among the strengths of our study, we analyzed an important subject not previously studied in detail.Our study design allowed us to study multiple effects and multiple exposures, given the high prevalence of migraines.The large sample size also gave robustness to the analysis.Patients were diagnosed in an experienced setting and quality of information was high.More research on this subject is needed to de nitely demonstrate some of this hypothesis.Future studies should analyze the presence of triptanophobia in larger samples and different settings, speci cally analyzing if triptan use might change the clinical situation.

Figures Figure 1
Figures

Table 1
Frequency of vascular risk factors in triptan users compared with triptan naïve patients.Frequency and percentage within each group.

Table 2
Comparison of demographic and clinical variables between triptan users and triptan naïve patients.