Dr. Buse: I am Dawn Buse, PhD, an Associate Professor of Neurology at Albert Einstein College of Medicine and Director of Behavioral Medicine at the Montefiore Headache Center in New York City. Today we are going to talk about nonpharmacologic treatments for migraine and other forms of headache, specifically the empirically supported behavioral interventions, and to do so we are privileged to have three expert clinician-scientists with us.
Dr. Frank Andrasik is a Distinguished Professor and Chair of the Department of Psychology at the University of Memphis. He also serves as Director of the Center for Behavioral Medicine at UM. Dr. Alvin Lake is a founding member of the Michigan Head-Pain and Neurological Institute and is both Director of the Behavioral Medicine Division and Associate Program Director of the inpatient Head Pain Treatment Unit at Chelsea Community Hospital. Dr. Donald Penzien is a Professor of Psychiatry and Human Behavior at the University of Mississippi Medical Center and founder and Director of the Head Pain Center. All three of our panelists have extensive clinical experience working with patients with headache and pain disorders. They have also conducted research in these areas and collectively published hundreds of peer-reviewed scientific articles. Welcome, gentlemen.
Although the range of acute and preventive pharmacologic treatments for migraine continues to grow, nonpharmacologic treatments play an important role in the comprehensive effective management of primary headache disorders including migraine. Nonpharmacologic treatments include behavioral approaches such as cognitive behavioral therapy (CBT), biofeedback, relaxation training, stress management, lifestyle modification and patient education. They also include physical and occupational therapies, acupuncture, and other non-drug modalities. Today we will focus on the behavioral and cognitive interventions with empirical evidence for their use. For ease of parsimony, we will refer to this group of treatments collectively as “behavioral treatments.”
In today’s conversation we will review the data on efficacy of empirically supported behavioral therapies for migraine, discuss how to identify appropriate patients by matching patient characteristics and needs with available treatments, and review important areas for clinical assessment. We will also review suggestions for healthcare providers (HCPs) for enhancing adherence and improving medical communication among other topics.
There is a large body of published evidence examining the use of behavioral and cognitive therapies for migraine (and other forms of primary headache) including meta-analytic studies and evidence-based reviews. The US Headache Consortium developed evidence-based guidelines for the treatment and management of migraine headache based on an extensive review of the medical literature and compilation of expert consensus.1 In the published guidelines, they suggest that nonpharmacologic treatments2 might be particularly well suited for patients who have a preference for nonpharmacologic interventions; display a poor tolerance for specific pharmacologic treatments; exhibit medical contraindications for specific pharmacologic treatments; have insufficient or no response to pharmacologic treatment; are pregnant, are planning to become pregnant, or are nursing; have a history of long-term, frequent, or excessive use of analgesic or acute medications that can aggravate headache problems (or lead to decreased responsiveness to other pharmacotherapies); or exhibit significant stress or deficient stress-coping skills.
Dr. Andrasik, what are the behavioral approaches with empirical evidence supporting their use for headache and migraine and what does the evidence show?
Dr. Andrasik: There are several behavioral treatments that have been examined starting with biofeedback beginning in the late 1970s. There are a number of different biofeedback approaches. Researchers have also looked at relaxation training and cognitive behavior therapy.
When I think of the evidence base I like to examine it from two standpoints. One includes reviews by special panels of experts convened to conduct evidence-based reviews according to specific criteria. Also, another way is to look at large-scale statistical analyses or meta-analyses of what the data show. So this way we have both qualitative and quantitative reviews.
The expert review of most interest to our audience perhaps would be the evidence-based guidelines developed by the US Headache Consortium2 based on evidence reports produced under the auspices of the Agency for Healthcare Research and Quality.3 The Consortium consisted of seven member organizations where they strongly endorsed various cognitive and behavioral procedures for treatment of migraine. They also listed some situations in which behavioral treatments were particularly indicated, such as those mentioned in the opening comments.
Then if we look at the meta-analytic side, we also have a large number of analyses that have been conducted.4,5 When I talk about this to audiences I jokingly refer to the fact that we have so many meta-analyses that the next one I expect to see is a meta-analysis of the meta-analyses. I wanted to mention one in particular, just briefly, that shows not only the support for various biofeedback procedures in improving pain parameters, but also demonstrates that biofeedback can impact variables not directly targeted but are nonetheless important to consider.6 In this most recent meta-analysis, we were able to examine some of the co-occurring symptoms associated with migraine headache. We found that upon receiving biofeedback treatments, anxiety scores improved, depression improved, and self-efficacy also improved. This meta-analysis was probably one of the first to look at correlated or concomitant changes in addition to headache parameters.
Dr. Penzien: I agree, there are many different ways to look at these data. But many do not recognize that there is a wealth of published evidence examining data on behavioral interventions for primary headache disorders. Empirically validated behavioral treatments include biofeedback training, relaxation training, combinations of the two, stress management training, and cognitive behavioral therapy.
Those behavioral interventions have yielded improvements on the order of 35% to 55% reductions in headache from pretreatment to post-treatment indicating that they're viable interventions. Those observed outcomes are approximately equivalent to the best prophylactic medications for primary headache. By comparison, wait list control treatments show almost no evidence of improvement. In fact, Dr. Andrasik was involved in a couple of trials years ago where he and his colleagues looked at the wait list control groups for periods ranging from a few months up to 3 years post treatment and found no improvements.7,8
Then medication placebo for migraine prophylactics yields, on average, approximately 12% to 14% improvement.9-11 So all of the active behavioral interventions show much higher outcomes with respect to medication placebo. In fact, those outcomes tend to be sustained over time, at least over the short run of 2 years. And there is some evidence that the improvements with behavior therapy are well maintained without necessity of additional therapy over time.
Dr. Andrasik: Dr. Penzien made a good point that in these meta-analyses, he and others have been able to directly compare literatures on the behavioral treatments with the most common prophylactic treatments, and almost to a decimal point they come out at the same efficacy level.
Dr. Lake: One question that treatment providers may raise relates to the efficacy of combining pharmacologic and behavioral treatments. Some of the research with which we are all familiar shows synergistically better outcomes from the combination of the two approaches, as in a now classic study by Holroyd and colleagues published in JAMA over a decade ago.12 Combining tricyclic antidepressant therapy with behavioral stress management led to a 50% or greater reduction in headache index scores for 64% of chronic tension headache patients randomly assigned to that condition, compared to 38% of those who only received tricyclic medication, 35% of those who only received stress management therapy, and 29% of those who were given placebo medication. The gain from the combination of behavioral therapy with a tricyclic antidepressant was both statistically significant and clinically meaningful.
I think we all agree that empirically sound evidence-based behavioral treatments should be included as a standard part of the basic headache-care toolkit, and need to be made available to more patients before the provider moves on to more esoteric, less proven, and sometimes far more expensive medical therapies. However, the value of behavioral therapy is not limited to the basics of headache management, but may need to be a critical part of the intensive treatment of intractable or refractory headache on both outpatient and inpatient levels as well.13
Dr. Penzien: You raise a good point, Dr. Lake, regarding the value of combining behavioral and pharmacological treatments for headache. The US Headache Consortium made the following recommendation2: “Behavioral therapy (i.e., relaxation, biofeedback) may be combined with preventive drug therapy (i.e., propranolol, amitriptyline) for patients to achieve additional clinical improvement for migraine relief (Grade B).” The only reason the Consortium’s recommendation regarding the use of combined drug and non-drug therapies was not stronger is because so few studies have directly evaluated the combined therapies. But all of the published evidence seems quite consistent in showing that benefits accrue when drug and non-drug therapies are combined for primary headache disorders.
So, as you're pointing out, these are viable interventions that have stood the test of time and are supported by ample empirical evidence but frankly remain underutilized.
Dr. Buse: Do you have any hypotheses about why these treatments are underutilized given their strong evidence base?
Dr. Andrasik: Well, I think one simple explanation is that the word is just not getting out. Perhaps we've not done a good job in this respect. The direction that some of my and the research of others has gone recently is to address the issue of costs of these treatments. A lot of patients are looking for a very quick and less time demanding and less expensive cure.
Our treatments work but they take a motivated patient. They take time. They take effort. They take multiple trips to the clinic. So one of the things that I and others have been doing is looking at ways that we can trim back the amount of treatment time, without reducing effectiveness.
I now use the acronym “PLOT” to describe this approach; prudent, limited, office treatment (a term originated by a close friend and biofeedback pioneer, Dr. Mark S. Schwartz, formerly of the Mayo Clinic). We and others have found that we can teach these behavioral procedures with fewer sessions if we prepare training materials that patients can study at home or work. The impetus for this approach came after examining the self-help literature, which showed that a number of conditions could be successfully treated in this manner, but that attrition was quite high. We reasoned that a few office visits, judiciously scheduled, had potential to engage and hold patients in treatment, ensure understanding, and allow the therapist to trouble-shoot with them as needed.14 Two reviews showed that PLOT was essentially equivalent to more intensive in-office treatments and resulted in considerable cost savings.15,16 This conserves monetary costs for the patient and can also conserve time on the therapist's part. But it doesn't conserve time on the patient's part. They still have to work hard at learning the techniques.
But, as we explore those kinds of treatments—and people are also looking at Internet-based treatment now—and investigate those new and different ways to deliver treatment, I think we may actually extend the reach of our treatments.14 The few published internet-based studies have shown good outcomes, but nearly all have revealed high rates of attrition (similar to pure self-help approaches). Hedborg and Muhr is the notable exception. They, too, added some appropriately timed face-to-face contacts and reduced their dropout rate to under 10%, which is much lower than any prior investigators obtained.17
Dr. Lake: Dr. Andrasik mentioned cost, which includes expenses in terms of effort and time. But we also have to think about the cost of medical treatment, which goes way beyond the charge for physician consultations, and includes the enduring costs of the various pharmacologic options that we use, some of which can be quite expensive and may not be covered well by some insurance providers. From both pharmacoeconomic and behavioral-economic perspectives, it makes sense to make these behavioral treatments available to as many headache sufferers as possible.
Dr. Penzien: We've recently published a study focused upon empirically modeling relative costs of behavioral treatments compared with drug therapies over time.18 We found that even the more expensive of the clinic-based behavioral therapies within a couple of years post treatment are no more expensive and often are less expensive to administer than continuing drug prophylaxis. In fact, the minimal-contact behavioral therapies for migraine are cost-competitive relative to pharmacological therapies by one year post treatment, and they become more cost efficient as the years of treatment accrue. Again, this is statistical modeling, not evidence from real-world experience, but there nevertheless is ample evidence that behavioral interventions can prove considerably more cost-effective than medication therapies.