Author + information
- aSection of Interventional Cardiology, MedStar Washington Hospital Center, Washington, DC
- bNational Heart, Lung, and Blood Institute, National Institutes of Health, Bethesda, Maryland
- ↵∗Address for correspondence:
Dr. Ron Waksman, MedStar Washington Hospital Center, 110 Irving Street, NW, Suite 4B-1, Washington, DC 20010.
- acute coronary syndrome
- implantable monitoring device
- ST-segment elevation myocardial infarction
- symptom-to-door time
Early intervention for patients presenting with acute coronary syndrome (ACS) can improve in-hospital and late outcomes. The concept of “time is muscle” is well-established for the treatment of patients presenting with acute myocardial infarction (AMI) (1–3). Therefore, early detection of and early intervention for ischemic events remains an unmet need and has been the subject of increased awareness among patients and physicians. The first clinical trigger for an ACS event is the patient’s symptoms, which are subject to patient interpretation and may or may not prompt a visit to the emergency room or the physician’s office. The heterogeneity in interpretation of symptoms and ambiguous electrocardiogram tracings impede the accuracy of diagnosis and often leads to significant delays in care for these patients. Further, the time from acute or partial closure of the vessel to onset of symptoms varies among patients and can be absent or prolonged in diabetic patients. Therefore, the concept of the AngelMed Guardian system for early detection of new onset of ACS events by continuous monitoring of ST-segment changes is brilliant, and the inventors should be congratulated on their vision and development of an algorithm that can facilitate early detection of acute coronary events (4,5).
The Guardian system is the first implantable device that continuously monitors the intracardiac electrogram for evidence of ST-segment shift indicating acute coronary artery occlusion. The system comprises 3 components: an implantable medical device akin to a single-chamber pacemaker, an external device that patients carry on their persons, and a programmer that the physician uses to program and retrieve data from the implantable device. The device algorithm generates an emergency alert if it identifies ST-segment change from the patient's personalized baseline. If an ST-segment shift is detected, the implantable device vibrates, and the external device flashes and sounds an auditory alarm, prompting the patient to call 911. Other abnormalities, such as high heart rate, arrhythmia, or artifact, trigger a less urgent alert on the external device. To evaluate the effectiveness and safety of the device, the investigators embarked on a large multicenter randomized controlled study in a high-risk population with prior history of ACS and with documentation of significant coronary artery disease (6).
The pivotal ALERTS (AngelMed for Early Recognition and Treatment of STEMI) trial (5), featured in this issue of the Journal, was a prospective multicenter randomized study with a Bayesian adaptive design used to adjust sample size based on interim treatment effect, with interim analysis planned every 300 patients after 600 patients randomized, up to a maximum of 3,000 patients. But the investigators did not believe that the predictive model accurately accounted for new Q waves at 6 months, and therefore, it could not predict the correct number of subjects to enroll. As a result, enrollment was stopped after 1,020 patients despite an interim analysis after 600 and 900 enrolled patients suggesting that enrollment should continue. Overall, only 910 patients had the device implanted. A total of 451 patients in the therapy arm had the device activated, and 456 patients had the device programmed to record but not to alert to serve as a control arm. The U.S. Food and Drug Administration (FDA) criticized the early termination and recommended caution interpreting the safety data.
The investigators set up safety endpoint criteria to be absent of device-related complications in >90% of the patients. Although the safety endpoint of the study was met, with an overall system-related complications rate of 3.3%, it actually was 3-fold higher than the absolute reduction of 1.1% of events resulting from the use of the device when compared with the control group’s adverse events rate. This questions the justification of the pre-specified safety threshold, especially with the presence of serious adverse events related to the device implantation, such as perforation, infection, and lead dislodgement or misplacement. After termination of the ALERTS study, patients had the option to leave the device in or to remove it (with or without the lead). Both options may be associated with long-term adverse events that should be considered when balancing risk versus benefit. Given these potential risks, only patients with a high risk for ACS events should be considered for the device.
With respect to efficacy, although within the first 90 days the alarms significantly decreased detection-to-arrival time at a medical facility (51 min vs. 30.6 h), which is impressive and emphasizes the potential of the device, only 1 of the 6 pre-specified primary endpoints was met. The pre-established primary efficacy endpoint was not significantly reduced (3.8% vs. 4.9%) in the treatment and control groups, respectively, posterior probability = 0.79. If the analysis were extended up to 90 days, no significant reductions were observed either. When the observation window was extended to 50, 70, and 90 days in a pre-specified analysis to include the majority of confirmed occlusive events in the control group and an exploratory dual-baseline electrocardiography analysis was used to reduce noise, a significant reduction in the primary endpoint was observed.
In an expanded analysis using data after the randomized period that compared patients with alarms disabled (control group during the first 6 months) with patients with alarms enabled (treatment group plus control patients after 6 months, mean follow-up 3.05 years), the positive predictive value for occlusive events was higher for device alerts (with or without symptoms) than for symptoms only (25.8% vs. 18.2%, respectively), while the false positive rate was lower (0.16 vs. 0.68 per patient/year). Importantly, the system did not cause excess use of resources or expose patients to unneeded risk.
Given these results, one must question the accuracy of the algorithm used in the device and how the ALERTS study results corroborate other studies using the same or a similar algorithm. Prior to the ALERTS study, there were 2 phase I studies that supported the ability of the device to detect ST-segment shift. But a larger study, ANALYZE ST (ST Monitoring to Detect Acute Coronary Syndrome Events in Implantable Cardioverter Defibrillator Patients) (7), powered for sensitivity of the ST monitoring feature to detect clinical ACS events, was terminated prematurely in June 2017. The total sample size required for the study was 5,228, but the enrollment was suspended after only 2,258 subjects. The results of the study and the reason for the study termination have not been disclosed publicly. However, in ANALYZE ST, the algorithm was implemented within an implantable cardioverter-defibrillator, which is different from the standalone implantable detection device used in the ALERTS trial that cannot provide any treatment. Additionally, the patient populations differed between the ANALYZE ST and the ALERTS trials. This highlights the importance of accuracy of the ST-segment detection algorithm and the challenge of tailoring it to different patient populations. The ALERTS trial did not tease out which patient subgroups may best benefit from the device. Furthermore, it is not clear what the threshold is for early detection of ST changes on outcomes. For example, does early detection of acute occlusion of the proximal left anterior descending artery have more utility than nonocclusive events related to a diagonal branch, both of which may cause detectable ST changes?
FDA Advisory Panel and Approval Letter
Since the study did not meet its primary efficacy endpoints and was associated with safety concerns, it is imperative that the benefits of the device should outweigh the complications related to device implantation and use. To address these questions, the FDA assembled an advisory panel in March 2016 to discuss 3 fundamental questions.
1. On the question of whether there was reasonable assurance that the AngelMed Guardian System is safe for patients who meet the criteria specified in the proposed indication: the panel voted 4 to 8 against.
2. On the question of whether there was reasonable assurance that the device is effective for use in patients who meet the criteria specified in the proposed indication: the panel voted unanimously 0 to 12 against.
3. Finally, on the question of whether the benefits of the device outweigh the risks for use in patients who meet the criteria specified in the proposed indication: the panel voted unanimously 0 to 12 against approval of the premarket approval application.
The FDA reviewed additional data provided by the sponsor, and in May 2018 granted approval for marketing for the device with restricted labeling: “The Guardian System is indicated for use in patients who have had prior acute coronary syndrome (ACS) events and who remain at high risk for recurrent ACS events” (8).
In its notification letter, the FDA asked the sponsor to conduct a post-marketing prospective, nonrandomized, single-arm, event-based, multicenter trial. The purpose of the study would be to assess: 1) the diagnostic accuracy of the device; 2) the compliance of the prescribing physician; 3) the experience of the implanting physician; 4) the experience of the emergency department physician; and 5) the patient compliance for “emergency” and “see doctor” alerts. The approval letter summarized the latest efficacy and safety data.
In a press release issued by the sponsor on December 31, 2018, AngelMed announced that it had filed for Chapter 11 bankruptcy protection with a pre-packaged restructuring plan that had the support of the company’s senior creditors. In addition, it was announced that the battery life of the device will be extended from 4 to 6 years. The company announced that it was planning to bring the device to market in the fourth quarter of 2019. When the device is available, it will be imperative for physicians to carefully select the appropriate patients who can benefit from the device, focusing on those who are at high risk for subsequent coronary occlusion events. Perhaps more data from the post-marketing study and physicians’ and patients’ experiences will shed light on the utility of the device and will determine where this technology will land in the care of patients after AMI.
Finally, FDA approval of this first-of-its-kind device to detect and alert patients to early ACS events should stimulate industry to continue to develop sensors—preferably wearable, noninvasive devices with high levels of accuracy—and machine learning to alert patients and minimize delays in patient care when presented with ACS.
↵∗ Editorials published in the Journal of the American College of Cardiology reflect the views of the authors and do not necessarily represent the views of JACC or the American College of Cardiology.
Dr. Waksman has served on the advisory board of Abbott Vascular, Amgen, Boston Scientific, Cardioset, Cardiovascular Systems, Medtronic, Philips Volcano, and Pi-Cardia Ltd.; has served as a consultant for Abbott Vascular, Amgen, Biosensors, Biotronik, Boston Scientific, Cardioset, Cardiovascular Systems, Medtronic, Philips Volcano, and Pi-Cardia Ltd.; has received grant support from Abbott Vascular, AstraZeneca, Biosensors, Biotronik, Boston Scientific, and Chiesi; has served on the Speakers Bureau of AstraZeneca and Chiesi; and is an investor in MedAlliance. Dr. Rogers has served as a consultant and proctor for Medtronic; and has served as a proctor for Edwards Lifesciences.
- Faxon D.,
- Lenfant C.
- De Luca G.,
- Suryapranata H.,
- Zijlstra F.,
- et al.
- Gibson M.C.,
- Krucoff M.,
- Fischell D.,
- et al.
- Rogers T.,
- Steinvil A.,
- Torguson R.,
- Waksman R.
- Gibson C.M.,
- Krucoff M.,
- Kirtane A.,
- et al.
- U.S. Food and Drug Administration