21 Aug 2015

Since 2011, a pacemaker that controls diaphragm contractions has been available to help people with ALS breathe. A humanitarian exemption from the FDA allows clinicians to install the device before it has been extensively tested in clinical trials. Now, data from the first of such trials questions the value of the pacemaker. In the July 31 Lancet Neurology, researchers led by Christopher McDermott of the University of Sheffield report that not only did the surgically implanted device fail to extend life, but that safety monitors halted the U.K.-based, randomized clinical trial early because survival times were nearly one year shorter for people in the treatment arm. The findings have raised alarm bells in the ALS community and led to intense discussion among clinicians and researchers. However, the debate is far from over. Two other randomized trials are ongoing, as is a post-approval, non-randomized study required by the FDA. “There is definitely no consensus yet, and I would say no one is convinced yet whether the device is helping or hurting patients,” commented Robert Baloh of the Cedars-Sinai Medical Center, who is participating in two of the continuing studies.

Breathing Aid.

The external part of the diaphragm pacer includes a battery-powered pulse generator and cable. *[Courtesy of Raymond Onders.]*

As motor neurons wither, people with ALS have trouble breathing and their blood can be overloaded with toxic carbon dioxide. The diaphragm pacemaker is designed to assist with breathing, allowing recipients to delay having a mechanical ventilator attached through their throat, said Raymond Onders of the University Hospitals Case Medical Center in Cleveland. Onders invented the device and co-founded the company that makes it, Synapse Biomedical Inc. Called the NeuRx Diaphragm Pacing System, it consists of four electrodes that stimulate the diaphragm muscle and an external pulse generator. Surgeons insert it during laparoscopic surgery under general anesthesia.

Onders originally developed the pacer for people with spinal-cord injuries. The Food and Drug Administration approved it for ALS in 2011 based on a pilot trial (see Onders et al., 2014) and a prospective study (reported in part in an FDA document). The people who received the diaphragm pacer survived 16 months longer than predicted based on historical data. The FDA concluded that the probable benefits of the pacer—delaying use of a mechanical ventilator and prolonging survival—outweighed the minimal risk. Based on this, the FDA granted a humanitarian device exemption, meaning it approved the device for ALS without complete evidence of efficacy.

Clinicians hoped a randomized trial would provide that evidence. McDermott and colleagues enrolled patients who were not breathing well, randomizing half to receive the pacer surgery, and half to be controls. All received non-invasive ventilation as well, which provided positive-pressure oxygen through a mask. The authors used an “intent to treat” analysis, meaning they considered people members of the pacer group as soon as they were randomized to that treatment arm, even if they delayed the surgery or did not get it at all. They allocated 37 people to the pacer group before the study was halted. Five of those did not undergo the surgery, two because they chose not to, one because respiratory function declined too rapidly for an operation to be safe, and two because of the early termination. Of the 32 who received the pacemaker, most tolerated it quite well and were able to use it regularly.

Though no one died within 30 days of the implantation surgery, suggesting the operation itself was not fatal, people in the pacer group survived for an average of 11 months after they were randomized, compared to 22½ months for those in the control group. Based on these interim data, the study’s ethics committee terminated the trial. “Diaphragmatic pacing is harmful for patients with ALS and respiratory failure,” McDermott concluded.

“This is a very impressive study,” commented Richard Bedlack of the Duke ALS Clinic in Durham, North Carolina. “After years of offering this to all qualified patients and implanting approximately 60 diaphragm pacers at Duke, this study convinced me to stop.”

Michael Swash of Royal London Hospital, who was not involved in the research, did not find the negative results entirely surprising. “Driving a weak, partially denervated muscle by electrical stimulation … following a ‘minimally invasive’ surgical procedure was always unlikely to be helpful long-term,” he wrote in an email to Alzforum. “For the moment there is no reliable evidence to support the use of this invasive technique in ALS.”

Why did the new study find no survival benefit, when the earlier research did? Onders suspected some people in the study were not good candidates for the pacer. The device only works if the phrenic nerves, which control the diaphragm, are still working—and they often are not in people with ALS. “This works in the right patients,” he said. He tests phrenic nerve function by electromyography or nerve conduction testing before selecting someone for surgery, and again during the operation before implanting the pacer. Even in the midst of the operation, he stops before inserting the pacer for 10 to 20 percent of patients, he said, because it turns out their phrenic nerves do not respond to stimulation.

In contrast, McDermott and colleagues tested phrenic nerve function by checking if the diaphragm moved independently of other abdominal muscles. Robert Miller of the Forbes Norris MDA/ALS Research Center in San Francisco, who was not involved in the U.K. study, speculated that they might have implanted some patients with advanced denervation, who would have been deemed unsuitable for a pacer in the U.S. McDermott discounted this. “When directly visualized at surgery, we could stimulate all diaphragms, otherwise we would not have implanted the device,” he told Alzforum. The authors did not measure whether the pacer recipients had better respiratory functions than controls, because they were interested in whether the procedure improved survival as a primary outcome, McDermott said.

Even if it might not have helped all recipients breathe, why did the pacer decrease survival? Swash and McDermott both speculated that the operation itself could have been detrimental. Supporting that idea, one review reported that ALS accelerates during the three months following surgery (Pinto et al., 2014). McDermott also theorized that stimulation of the phrenic nerve itself might be damaging.

The other explanation for the poor survival is simple chance, said Jonathan Katz, who also works at the Forbes Norris center. “In a 32-patient ALS study, chance is always an issue,” Katz told Alzforum. A handful of additional severe cases in the pacer arm compared to the control arm could make it seem as if the device was detrimental.

The authors noted that it was difficult to compare their data to the original study that led to FDA approval, because that research was not published in a peer-reviewed journal. Hiroshi Mitsumoto of Columbia University Medical Center in New York, who wrote a commentary in Lancet Neurology, expressed similar concerns. “That no data from the original study were published in scientific papers that had undergone vigorous peer review was an early warning sign,” Mitsumoto wrote. “A randomized controlled trial should have expeditiously followed the initial promising results.” Now that people can choose for themselves to get pacing or not, it makes it harder to recruit participants willing to undergo randomization.

Synapse Biomedical continues to monitor people who get the device in the follow-on study required by the FDA. Onders plans to present the latest data from 54 pacer recipients in that trial, as well as from 214 pacer recipients at the University Hospitals of Cleveland, at the International Symposium on ALS/MND in December 2015. Patients in those studies survived, on average, for more than 20 months after implantation, he told Alzforum. In addition, researchers in the United States and France are conducting randomized trials of the device. The French trial, which focuses on early implantation, has suspended recruitment, according to clinicaltrials.gov. Study leader Thomas Similowski at the Assistance Publique-Hôpitaux de Paris, which sponsors the trial, told Alzforum he would not discuss the reason, but that information on the suspension might be available soon.

Katz, co-leader on the U.S. randomized trial, discussed McDermott’s results and interim safety data from the U.S. study in a Webinar hosted by the ALS Association on August 17. “This has obviously created a lot of concern on this side of the Atlantic,” he said. The U.S. trial aims to enroll 120 people who would get pacers and 60 controls; so far, they have implanted 32 devices, the same number used in the British trial. Katz and colleagues have not seen the kind of reduced survival their U.K. counterparts observed. “We are seeing signs that the procedure is safe, but it is too early to draw conclusions about [survival],” he said, adding he could find no clear reason why the U.S. and British studies had such different results with respect to safety. Katz believes continuing with the randomized trial is important, though he noted the study’s safety board has the final call.

That leaves clinicians, and patients, still waiting for a firm conclusion on diaphragm pacing. Baloh, for one, told Alzforum that Cedars-Sinai would continue to perform the procedure as long as the trials continue.—Amber Dance

Comments

- Richard Bedlack
  Duke University
- Posted: 24 Aug 2015
DiPALS is a very impressive study. It is the first, and still only, randomized prospective trial of DPS in ALS. Although the sample size is small, this study showed a large negative effect from diaphragm pacing, with patients receiving diaphragm pacing and non-invasive ventilation living 11 months less than patients receiving non-invasive ventilation alone.

It is not yet clear why this data is so different from the results of the non-randomized studies that have been done. It is certainly not the first time that a randomized study showed something vastly different than a non-randomized one. One possible explanation is that there are biases in non-randomized studies that affect outcomes, independent of the intervention being studied. For example, in my own clinic I have found that patients who opted for diaphragm pacing were generally much more aggressive about fighting their illness than patients who opted against diaphragm pacing. That difference in attitude alone might have a significant effect on survival. Whatever the explanation, this is a great example of the importance of randomized studies.

As for the impact of DiPALS on my own practice, after years of offering this to all qualified patients and implanting approximately 60 diaphragm pacers at Duke, this study convinced me to stop. I will not be offering new diaphragm pacing implants in my clinic again unless or until an even better-designed study shows a benefit.

View all comments by Richard Bedlack

- Robert Miller
  California Pacific Medical Center
- Posted: 24 Aug 2015
We are happy to see the first controlled clinical trial examining the benefit and safety of diaphragm pacing in patients with ALS.

The DiPALS study was a well-designed and -executed study, but it actually showed shorter survival in the patients receiving diaphragm pacing plus non-invasive ventilation compared with those receiving non-invasive ventilation alone. At first blush, this would appear to cast doubt on the value of pacing in ALS. However, there are some issues that mitigate the impact of the finding.

First, the study is very small for a survival study, raising the possibility that the finding was due to chance, as we have seen before with lithium and with the first riluzole study. Second, many patients in the study did not adhere to protocol and failed to utilize non-invasive ventilation. Third, all patients taken to surgery were implanted with pacing electrodes, suggesting that some patients with advanced denervation, who would not have been implanted in the United States, were implanted and unlikely to benefit.

Two registry studies of diaphragm pacing in ALS patients have demonstrated a survival advantage with pacing that led the FDA to conclude 1) the device is of probable benefit and 2) there were no safety concerns.

Two other randomized controlled trials are underway and hopefully with all of this additional data, the role of diaphragm pacing will become clearer. In particular, if there is a subset of responders to pacing, these patients might be more adequately characterized by further studies, which I believe should be done.

View all comments by Robert Miller

- Katalin Scherer
- Posted: 25 Aug 2015
In this well-designed clinical trial, McDermott and colleagues found that ALS patients with respiratory insufficiency who received the diaphragm pacing simulation (DPS) system in addition to non-invasive ventilation (NIV) had a significantly shortened survival (11 months) versus the group who received NIV alone (22 months). The median survival from symptom onset was similarly shorter in the DPS+NIV group (28 months versus 45 months in the NIV-alone group). Minor differences amongst the two treatment groups (age, use of cough-assist machines, breath stacking and suction devices) are unlikely to have caused an almost doubling of the length of survival in the NIV-alone group, and the patient demographics data makes this study generalizable to ALS patients as a whole (as opposed to the previous selected DPS patient cohorts). Another clinical trial, (RepStimALS, early phrenic nerve stimulation) has suspended enrollment as of this writing.

The authors stipulate several possible mechanisms that may explain why the DPS worsened survival: Diaphragm stimulation may cause excessive muscle fatigue, may contribute to increased demise of already ill motor neurons, may cause asynchrony with the patient’s own breathing, or may accelerate neuromuscular dysfunction that has been shown in laboratory research animals.

Based on this study, DPS is not recommended for ALS patients suffering from respiratory insufficiency. NIV is a proven and beneficial treatment that can prolong survival by almost two years in ALS patients who tolerate it and by seven months in all ALS patients as a group. Together with other, safe adjunct respiratory therapies (breath stacking, mechanical and manual cough assistance, and airway clearance), NIV remains the primary treatment for respiratory insufficiency in ALS.

View all comments by Katalin Scherer

- Robert Baloh
  Director of Neuromuscular Medicine, Cedars-Sinai Medical Center
- Posted: 01 Sep 2015
There is definitely no consensus on the benefit of diaphragm pacing in ALS. There is still a study ongoing in the United States run by the North East ALS consortium (NEALS). To my knowledge, they are planning to look at the data and decide if any changes need to be made or if the trial should continue.

The bottom line is the survival for the group receiving diaphragm pacing in the DiPALS study was shorter than what has been reported previously for DPS in the retrospective studies. The reason for this remains unclear.

That said, I would say no one is convinced yet whether DPS helps or hurts patients. We continue to do the procedure at our site as part of the NEALS U.S. trial, and will continue to provide that option until the trial is completed or stopped, because that is the only way to get definitive information that we can use to guide patients.

View all comments by Robert Baloh