Neuromodulation for Cluster Headache—A Promising Area of Study

LAS VEGAS—In cluster headache, four neuromodulation techniques have been tested—deep brain stimulation (DBS), vagal nerve stimulation (VNS), sphenopalatine ganglion (SPG) stimulation, and occipital nerve stimulation (ONS). At the 17th Annual Meeting of the North American Neuromodulation Society, Stewart Tepper, MD, assessed the available evidence for each treatment modality.

Deep Brain Stimulation
“The idea of doing DBS for cluster headache was generated by the severity of what we saw in our patients,” said Dr. Tepper, who is Director of Research for the Neurological Center for Pain at the Cleveland Clinic. Early research was promising, but revealed drawbacks. DBS entails a risk of hemorrhage, transient ischemic attack (TIA), stroke, and death. Researchers also found that if the impulses were increased too quickly, patients would get oculomotor disturbances and vertigo.

“You couldn’t use DBS for acute treatment of cluster,” Dr. Tepper said. “And it also took a long time for the DBS to start to work in cluster headache prevention. The mean time to effectiveness was 42 days once it was on, and then it really seemed to work. If you turned it off, the clusters would come back.” One death, a hemorrhage, a stroke, and at least one TIA were reported.

Although 61% of the patients implanted seemed to benefit, the complications prompted researchers to examine other alternatives, including peripheral stimulation. “Small open-label case series do not justify widespread clinical use, but probably do justify a large prospective trial,” Dr. Tepper said. “Since the available evidence is from case series, the evidence is at level U—no randomized controlled data—but we think it works because if you turn it off, the clusters come back.”

Vagal Nerve Stimulation
The evidence for implantable VNS for cluster is even more scant. Two patients were implanted for chronic cluster. “They both responded, but the outcome measure was a disability measure that is primarily used for migraine, not validated in cluster, so the evidence level is U for implantable VNS for cluster,” said Dr. Tepper.

“The noninvasive VNS is called GammaCore, and there has been one open-label study presented in abstract form,” Dr. Tepper said. “We don’t have a complete peer-reviewed paper.” There were 21 patients in this study: 12 with chronic cluster and nine with episodic cluster. “The device may have preventive as well as acute effects, so in this study both were looked at,” Dr. Tepper noted.

For acute use, two to four cycles of 90 seconds each were given at the start of an attack ipsilateral to the pain. About two-thirds of the acute attacks seemed to respond; 47% of the attacks responded with pain freedom within 15 minutes. Tachyphylaxis did not appear to occur, and the effect persisted for six to 12 months. Four patients were able to stop their acute cluster medicine, two were able to stop oxygen, and four were able to stop taking triptans. “If this were to hold up in a randomized controlled trial, it would have a major impact on cluster headache treatment,” Dr. Tepper said.

For preventive use, two 90-second cycles were given twice daily—morning and afternoon—ipsilateral to the pain, and this treatment showed a significant decrease in 24-hour frequency of cluster attacks. Before the treatment, there were as many as four attacks per 24 hours; afterward, there were 2.5 per 24 hours, and that difference was statistically significant.

“So where do we stand? All patients who participated would recommend that other cluster patients try it,” Dr. Tepper said. Adverse events were local and short-lived. The mean estimated improvement was about 50%. Four randomized controlled trials are under way right now—three for cluster and one for chronic migraine. The noninvasive VNS device has a CE mark in Europe. “Right now the evidence level is still U for both migraine and cluster—we have no randomized controlled data,” Dr. Tepper said. “But stay tuned for next year because this is an area of great interest to clinicians and patients alike.”

Sphenopalatine Ganglion Stimulation
A randomized controlled trial of the Autonomic Technologies implanted SPG stimulator device was published in Cephalalgia in 2013. A preimplant baseline measurement was followed by the implant and stabilization. Next came the experimental period with a sham and subthreshold stimulation, followed by real stimulation that lasted from three to eight weeks. The primary end point was 15-minute relief. “Here are the most up-to-date numbers,” Dr. Tepper said. Data from 32 patients with experimental stimulation were published, an additional 11 subjects were subsequently recruited, and 24 completed 18 months of follow-up. At the end of the experimental period, 769 attacks were analyzed; 55% of the attacks achieved pain relief at 15 minutes versus 6% with sham.

As with the noninvasive VNS device, the SPG stimulator showed effects on cluster prevention. With as-needed stimulation, the average number of cluster attacks per week decreased by 31%. Forty-two percent of the patients had close to a 90% reduction in attack frequency versus baseline. “It was a nice problem to have—both an acute and a preventive outcome,” Dr. Tepper commented. Nearly 45% of the patients had a reduction in their acute medication use; 39% had stopped preventive medications or reduced the dose versus baseline. “The response was maintained, so two-thirds of patients kept their responder rate, and remember the responder rate now involves acute or preventive, during the long-term follow-up,” Dr. Tepper reported. Surgery effects were considered tolerable by 83% of participants, and more than 90% of study participants said they would recommend the surgery to other patients with cluster headache.

In another study published last summer in Cephalalgia, Danish researchers found that at low frequency, SPG stimulation triggered cluster attacks. At high frequency, it either terminated or prevented cluster attacks. “It seemed the low frequency stimulation was activating the outflow from the SPG, while the high frequency was either depleting the activation, causing an information block, or somehow inhibiting outflow,” Dr. Tepper said. “This [result] is a dramatic confirmation of the pathophysiology of the device and how the SPG neuromodulation affects the cluster.”

According to Dr. Tepper, the evidence for SPG stimulation for cluster is level C—possibly effective due to one Class II trial in which 63% of patients responded with acute relief, decreased attack frequency, or both. In addition, 42% of patients had close to a 90% reduction in attack frequency; 55% achieved pain relief in 15 minutes and maintained it through 90 minutes; 46% reduced acute treatments; and there was a decrease in preventive medications as well. Headache impact decreased, quality of life increased, and these changes were sustained through 18 months.

“Right now in Europe, not only does the device have a CE mark, it is also approved as a first-line therapy, and it is reimbursed in Denmark and in Germany,” Dr. Tepper said.

Occipital Nerve Stimulation
To date, 23 patients have been treated with ONS for chronic cluster, and 12 have responded, yielding a 52% response rate. But, as Dr. Tepper pointed out, these investigations were case studies. Because no randomized controlled trial data are available, the level of evidence is U.

The Current State of Evidence
“On cluster, we have level U for DBS, but we really think it works because you can turn it on and turn it off. For noninvasive VNS, we have four randomized controlled trials under way and open-label data that are encouraging. The noninvasive VNS would be a fantastic opportunity for patients, as would the implanted SPG stimulator, which would be level C based on one Class II study. And ONS is still at level U—no randomized controlled trials,” Dr. Tepper said.

—Glenn S. Williams

LAS VEGAS—In cluster headache, four neuromodulation techniques have been tested—deep brain stimulation (DBS), vagal nerve stimulation (VNS), sphenopalatine ganglion (SPG) stimulation, and occipital nerve stimulation (ONS). At the 17th Annual Meeting of the North American Neuromodulation Society, Stewart Tepper, MD, assessed the available evidence for each treatment modality.

Deep Brain Stimulation
“The idea of doing DBS for cluster headache was generated by the severity of what we saw in our patients,” said Dr. Tepper, who is Director of Research for the Neurological Center for Pain at the Cleveland Clinic. Early research was promising, but revealed drawbacks. DBS entails a risk of hemorrhage, transient ischemic attack (TIA), stroke, and death. Researchers also found that if the impulses were increased too quickly, patients would get oculomotor disturbances and vertigo.

“You couldn’t use DBS for acute treatment of cluster,” Dr. Tepper said. “And it also took a long time for the DBS to start to work in cluster headache prevention. The mean time to effectiveness was 42 days once it was on, and then it really seemed to work. If you turned it off, the clusters would come back.” One death, a hemorrhage, a stroke, and at least one TIA were reported.

Although 61% of the patients implanted seemed to benefit, the complications prompted researchers to examine other alternatives, including peripheral stimulation. “Small open-label case series do not justify widespread clinical use, but probably do justify a large prospective trial,” Dr. Tepper said. “Since the available evidence is from case series, the evidence is at level U—no randomized controlled data—but we think it works because if you turn it off, the clusters come back.”

Vagal Nerve Stimulation
The evidence for implantable VNS for cluster is even more scant. Two patients were implanted for chronic cluster. “They both responded, but the outcome measure was a disability measure that is primarily used for migraine, not validated in cluster, so the evidence level is U for implantable VNS for cluster,” said Dr. Tepper.

“The noninvasive VNS is called GammaCore, and there has been one open-label study presented in abstract form,” Dr. Tepper said. “We don’t have a complete peer-reviewed paper.” There were 21 patients in this study: 12 with chronic cluster and nine with episodic cluster. “The device may have preventive as well as acute effects, so in this study both were looked at,” Dr. Tepper noted.

For acute use, two to four cycles of 90 seconds each were given at the start of an attack ipsilateral to the pain. About two-thirds of the acute attacks seemed to respond; 47% of the attacks responded with pain freedom within 15 minutes. Tachyphylaxis did not appear to occur, and the effect persisted for six to 12 months. Four patients were able to stop their acute cluster medicine, two were able to stop oxygen, and four were able to stop taking triptans. “If this were to hold up in a randomized controlled trial, it would have a major impact on cluster headache treatment,” Dr. Tepper said.

For preventive use, two 90-second cycles were given twice daily—morning and afternoon—ipsilateral to the pain, and this treatment showed a significant decrease in 24-hour frequency of cluster attacks. Before the treatment, there were as many as four attacks per 24 hours; afterward, there were 2.5 per 24 hours, and that difference was statistically significant.

“So where do we stand? All patients who participated would recommend that other cluster patients try it,” Dr. Tepper said. Adverse events were local and short-lived. The mean estimated improvement was about 50%. Four randomized controlled trials are under way right now—three for cluster and one for chronic migraine. The noninvasive VNS device has a CE mark in Europe. “Right now the evidence level is still U for both migraine and cluster—we have no randomized controlled data,” Dr. Tepper said. “But stay tuned for next year because this is an area of great interest to clinicians and patients alike.”

Sphenopalatine Ganglion Stimulation
A randomized controlled trial of the Autonomic Technologies implanted SPG stimulator device was published in Cephalalgia in 2013. A preimplant baseline measurement was followed by the implant and stabilization. Next came the experimental period with a sham and subthreshold stimulation, followed by real stimulation that lasted from three to eight weeks. The primary end point was 15-minute relief. “Here are the most up-to-date numbers,” Dr. Tepper said. Data from 32 patients with experimental stimulation were published, an additional 11 subjects were subsequently recruited, and 24 completed 18 months of follow-up. At the end of the experimental period, 769 attacks were analyzed; 55% of the attacks achieved pain relief at 15 minutes versus 6% with sham.

As with the noninvasive VNS device, the SPG stimulator showed effects on cluster prevention. With as-needed stimulation, the average number of cluster attacks per week decreased by 31%. Forty-two percent of the patients had close to a 90% reduction in attack frequency versus baseline. “It was a nice problem to have—both an acute and a preventive outcome,” Dr. Tepper commented. Nearly 45% of the patients had a reduction in their acute medication use; 39% had stopped preventive medications or reduced the dose versus baseline. “The response was maintained, so two-thirds of patients kept their responder rate, and remember the responder rate now involves acute or preventive, during the long-term follow-up,” Dr. Tepper reported. Surgery effects were considered tolerable by 83% of participants, and more than 90% of study participants said they would recommend the surgery to other patients with cluster headache.

In another study published last summer in Cephalalgia, Danish researchers found that at low frequency, SPG stimulation triggered cluster attacks. At high frequency, it either terminated or prevented cluster attacks. “It seemed the low frequency stimulation was activating the outflow from the SPG, while the high frequency was either depleting the activation, causing an information block, or somehow inhibiting outflow,” Dr. Tepper said. “This [result] is a dramatic confirmation of the pathophysiology of the device and how the SPG neuromodulation affects the cluster.”

According to Dr. Tepper, the evidence for SPG stimulation for cluster is level C—possibly effective due to one Class II trial in which 63% of patients responded with acute relief, decreased attack frequency, or both. In addition, 42% of patients had close to a 90% reduction in attack frequency; 55% achieved pain relief in 15 minutes and maintained it through 90 minutes; 46% reduced acute treatments; and there was a decrease in preventive medications as well. Headache impact decreased, quality of life increased, and these changes were sustained through 18 months.

“Right now in Europe, not only does the device have a CE mark, it is also approved as a first-line therapy, and it is reimbursed in Denmark and in Germany,” Dr. Tepper said.

Occipital Nerve Stimulation
To date, 23 patients have been treated with ONS for chronic cluster, and 12 have responded, yielding a 52% response rate. But, as Dr. Tepper pointed out, these investigations were case studies. Because no randomized controlled trial data are available, the level of evidence is U.

The Current State of Evidence
“On cluster, we have level U for DBS, but we really think it works because you can turn it on and turn it off. For noninvasive VNS, we have four randomized controlled trials under way and open-label data that are encouraging. The noninvasive VNS would be a fantastic opportunity for patients, as would the implanted SPG stimulator, which would be level C based on one Class II study. And ONS is still at level U—no randomized controlled trials,” Dr. Tepper said.

—Glenn S. Williams

LAS VEGAS—In cluster headache, four neuromodulation techniques have been tested—deep brain stimulation (DBS), vagal nerve stimulation (VNS), sphenopalatine ganglion (SPG) stimulation, and occipital nerve stimulation (ONS). At the 17th Annual Meeting of the North American Neuromodulation Society, Stewart Tepper, MD, assessed the available evidence for each treatment modality.

Deep Brain Stimulation
“The idea of doing DBS for cluster headache was generated by the severity of what we saw in our patients,” said Dr. Tepper, who is Director of Research for the Neurological Center for Pain at the Cleveland Clinic. Early research was promising, but revealed drawbacks. DBS entails a risk of hemorrhage, transient ischemic attack (TIA), stroke, and death. Researchers also found that if the impulses were increased too quickly, patients would get oculomotor disturbances and vertigo.

“You couldn’t use DBS for acute treatment of cluster,” Dr. Tepper said. “And it also took a long time for the DBS to start to work in cluster headache prevention. The mean time to effectiveness was 42 days once it was on, and then it really seemed to work. If you turned it off, the clusters would come back.” One death, a hemorrhage, a stroke, and at least one TIA were reported.

Although 61% of the patients implanted seemed to benefit, the complications prompted researchers to examine other alternatives, including peripheral stimulation. “Small open-label case series do not justify widespread clinical use, but probably do justify a large prospective trial,” Dr. Tepper said. “Since the available evidence is from case series, the evidence is at level U—no randomized controlled data—but we think it works because if you turn it off, the clusters come back.”

Vagal Nerve Stimulation
The evidence for implantable VNS for cluster is even more scant. Two patients were implanted for chronic cluster. “They both responded, but the outcome measure was a disability measure that is primarily used for migraine, not validated in cluster, so the evidence level is U for implantable VNS for cluster,” said Dr. Tepper.

“The noninvasive VNS is called GammaCore, and there has been one open-label study presented in abstract form,” Dr. Tepper said. “We don’t have a complete peer-reviewed paper.” There were 21 patients in this study: 12 with chronic cluster and nine with episodic cluster. “The device may have preventive as well as acute effects, so in this study both were looked at,” Dr. Tepper noted.

For acute use, two to four cycles of 90 seconds each were given at the start of an attack ipsilateral to the pain. About two-thirds of the acute attacks seemed to respond; 47% of the attacks responded with pain freedom within 15 minutes. Tachyphylaxis did not appear to occur, and the effect persisted for six to 12 months. Four patients were able to stop their acute cluster medicine, two were able to stop oxygen, and four were able to stop taking triptans. “If this were to hold up in a randomized controlled trial, it would have a major impact on cluster headache treatment,” Dr. Tepper said.

For preventive use, two 90-second cycles were given twice daily—morning and afternoon—ipsilateral to the pain, and this treatment showed a significant decrease in 24-hour frequency of cluster attacks. Before the treatment, there were as many as four attacks per 24 hours; afterward, there were 2.5 per 24 hours, and that difference was statistically significant.

“So where do we stand? All patients who participated would recommend that other cluster patients try it,” Dr. Tepper said. Adverse events were local and short-lived. The mean estimated improvement was about 50%. Four randomized controlled trials are under way right now—three for cluster and one for chronic migraine. The noninvasive VNS device has a CE mark in Europe. “Right now the evidence level is still U for both migraine and cluster—we have no randomized controlled data,” Dr. Tepper said. “But stay tuned for next year because this is an area of great interest to clinicians and patients alike.”

Sphenopalatine Ganglion Stimulation
A randomized controlled trial of the Autonomic Technologies implanted SPG stimulator device was published in Cephalalgia in 2013. A preimplant baseline measurement was followed by the implant and stabilization. Next came the experimental period with a sham and subthreshold stimulation, followed by real stimulation that lasted from three to eight weeks. The primary end point was 15-minute relief. “Here are the most up-to-date numbers,” Dr. Tepper said. Data from 32 patients with experimental stimulation were published, an additional 11 subjects were subsequently recruited, and 24 completed 18 months of follow-up. At the end of the experimental period, 769 attacks were analyzed; 55% of the attacks achieved pain relief at 15 minutes versus 6% with sham.

As with the noninvasive VNS device, the SPG stimulator showed effects on cluster prevention. With as-needed stimulation, the average number of cluster attacks per week decreased by 31%. Forty-two percent of the patients had close to a 90% reduction in attack frequency versus baseline. “It was a nice problem to have—both an acute and a preventive outcome,” Dr. Tepper commented. Nearly 45% of the patients had a reduction in their acute medication use; 39% had stopped preventive medications or reduced the dose versus baseline. “The response was maintained, so two-thirds of patients kept their responder rate, and remember the responder rate now involves acute or preventive, during the long-term follow-up,” Dr. Tepper reported. Surgery effects were considered tolerable by 83% of participants, and more than 90% of study participants said they would recommend the surgery to other patients with cluster headache.

In another study published last summer in Cephalalgia, Danish researchers found that at low frequency, SPG stimulation triggered cluster attacks. At high frequency, it either terminated or prevented cluster attacks. “It seemed the low frequency stimulation was activating the outflow from the SPG, while the high frequency was either depleting the activation, causing an information block, or somehow inhibiting outflow,” Dr. Tepper said. “This [result] is a dramatic confirmation of the pathophysiology of the device and how the SPG neuromodulation affects the cluster.”

According to Dr. Tepper, the evidence for SPG stimulation for cluster is level C—possibly effective due to one Class II trial in which 63% of patients responded with acute relief, decreased attack frequency, or both. In addition, 42% of patients had close to a 90% reduction in attack frequency; 55% achieved pain relief in 15 minutes and maintained it through 90 minutes; 46% reduced acute treatments; and there was a decrease in preventive medications as well. Headache impact decreased, quality of life increased, and these changes were sustained through 18 months.

“Right now in Europe, not only does the device have a CE mark, it is also approved as a first-line therapy, and it is reimbursed in Denmark and in Germany,” Dr. Tepper said.

Occipital Nerve Stimulation
To date, 23 patients have been treated with ONS for chronic cluster, and 12 have responded, yielding a 52% response rate. But, as Dr. Tepper pointed out, these investigations were case studies. Because no randomized controlled trial data are available, the level of evidence is U.

The Current State of Evidence
“On cluster, we have level U for DBS, but we really think it works because you can turn it on and turn it off. For noninvasive VNS, we have four randomized controlled trials under way and open-label data that are encouraging. The noninvasive VNS would be a fantastic opportunity for patients, as would the implanted SPG stimulator, which would be level C based on one Class II study. And ONS is still at level U—no randomized controlled trials,” Dr. Tepper said.

—Glenn S. Williams