The Take Home: TCT
Many of interventional cardiology’s hottest topics came to the forefront at the Transcatheter Cardiovascular Therapeutics Scientific Symposium, held Oct. 29 to Nov. 2, 2017 in Denver. Cardiology Today’s Intervention obtained reaction from attendees, including Updesh S. Bedi, MD, from the University of Minnesota and Regions Hospital; Dmitriy N. Feldman, MD, from Weill Cornell Medical College; Cardiology Today’s Intervention Editorial Board Member William W. O’Neill, MD, from Henry Ford Hospital; Jordan Safirstein, MD, from Atlantic Health System/Morristown Medical Center; and Arnold H. Seto, MD, from the University of California, Irvine, and Long Beach VA Medical Center.
For a discussion of the controversial ORBITA trial, please see the At Issue feature.
O’Neill: The CULPRIT-SHOCK trial is the latest in large, randomized trials of immediate PCI in cardiogenic shock. The results clearly show that there is not significant benefit to immediate multivessel intervention, once the acute MI culprit lesion is addressed. In the 706-patient, multicenter, randomized trial, the composite primary endpoint of death or renal-replacement therapy at 30 days occurred in 45.9% of patients assigned PCI of the culprit-lesion-only and in 55.4% of those assigned multivessel PCI (RR = 0.83; 95% CI, 0.71-0.96).
This sends a message that if you are doing PCI alone, you should only do the culprit lesion. The CULPRIT-SHOCK results support treatment of other vessels later during hospitalization.
For 30 years, we have known that PCI in patients with cardiogenic shock results in a 50% survival. In CULPRIT-SHOCK, the mortality rate is still about 50%. So while the decision to do the culprit lesion only vs. multivessel intervention is helpful, it does not address the overall problem that even with optimal PCI strategy, we still have 50% mortality.
In the Detroit Cardiogenic Shock Initiative, we reported at TCT that a 76% survival rate can be achieved with use of immediate mechanical circulatory support prior to PCI. The incidence of acute MI with cardiogenic shock continues to increase in the U.S. And in the Medicare population, the incidence is climbing even faster. The CULPRIT-SHOCK trial helps in determining the optimal PCI strategy, but it does not move the ball forward in the big picture of this population. This is a national crisis and needs urgent, U.S.-wide attention. For this reason, we at Henry Ford and a coalition of the willing have started the National Cardiogenic Shock Initiative. Our aim is to decrease mortality from 50% to 20% on a nationwide basis.
Feldman: The CULPRIT-SHOCK trial randomly assigned 706 patients with acute MI, cardiogenic shock and multivessel CAD (at least 2 major vessels with > 70% stenosis of the diameter) to one of two revascularization strategies: PCI of the culprit lesion only (with the option of staged revascularization of nonculprit lesions) or immediate multivessel PCI. At 30 days, the composite primary endpoint of death or renal-replacement therapy had occurred in 45.9% in the culprit-lesion-only PCI group and in 55.4% in the multivessel PCI group (RR = 0.83; P = .01). This difference was primarily driven by a decrease in mortality (RR = 0.84; P = .03), with a trend towards a decrease in the rate of renal-replacement therapy within the culprit lesion only revascularization group (RR = 0.71; P = .07).
Given that approximately 80% of patients with cardiogenic shock have multivessel coronary disease and multivessel disease is associated with higher mortality, arguments have been made that immediate multivessel revascularization of all important coronary stenoses may provide best outcomes by improving myocardial perfusion and ventricular function. Recent appropriate-use criteria have suggested that it is appropriate to perform immediate revascularization of a nonculprit artery if cardiogenic shock persists after revascularization of the culprit artery. In the SHOCK trial, where a large number of patients had multivessel disease, the mortality rate was similar between PCI and CABG, despite a much higher rate of complete revascularization with CABG. Furthermore, a post-hoc analysis of the ISAR-SHOCK II trial did not observe a difference in mortality between immediate multivessel PCI vs. culprit lesion intervention strategies. A recent meta-analysis of cardiogenic shock patients observed an increase in 30-day mortality in those undergoing multivessel PCI compared to culprit vessel PCI.
The CULPRIT-SHOCK trial is the first randomized trial examining two revascularization strategies in those with MI and cardiogenic shock, expanding our understanding of shock in the setting of STEMI and non-STEMI patients and including chronic total occlusions in the nonculprit lesions. This trial suggests that routine immediate revascularization of culprit and nonculprit lesions in cardiogenic shock and MI is associated with higher composite rate of death or renal-replacement therapy as well as higher mortality. Importantly, staged revascularization was not counted as a disadvantage of the culprit-lesion-only PCI strategy and thus may still be of benefit in patients with persistent cardiogenic shock, but initially the focus should be on PCI of the infarct-related artery.
Safirstein: The utility of bare-metal stents continues to wane, and the SENIOR trial may have served as the proverbial nail in the coffin. After assigning of 1,200 patients over the age of 75 with CAD and either stable angina, silent ischemia or ACS to 1 month or 6 months of DAPT, patients were randomly assigned to a drug-eluting stent (n = 596) or BMS (n = 604). The primary endpoint was a composite of all-cause mortality, target lesion revascularization, stroke, or MI. Perhaps equally important were the secondary endpoints of bleeding and stent thrombosis at 1 year.
The primary composite endpoint was significantly lower in the DES group (11.6% vs. 16.4%; P = .016), driven mostly by increased rate of TLR (1.7% vs 5.9%; P = .0002), while bleeding complications and stent thrombosis were low and similar in both groups. The results demonstrate that in elderly patients undergoing PCI with a short duration of DAPT, DES yield lower adverse events at 1 year without risk for stent thrombosis. For many of us who implant stents, it is nice to know that we do not have to sacrifice potential long-term benefits of DES compared to BMS in patients at high risk for bleeding and that we can safely stop DAPT if needed.
Seto: The SENIOR study demonstrates that there is no difference between a contemporary DES and a BMS in terms of stent thrombosis, but a significant difference in restenosis rates. Unique to this trial was the equal and early discontinuation of dual antiplatelet therapy regardless of the stent implanted, meaning planned discontinuation of DAPT at 1 month or 6 months for patients with stable angina or ACS, respectively.
The previously axiomatic concept that DES implantation mandates longer periods of DAPT than BMS to prevent late stent thrombosis is no longer applicable to contemporary DES. The modern second- or third-generation DES is much more biocompatible than the first-generation DES. We have multiple studies that have shown no excess in stent thrombosis with contemporary DES, with stent thrombosis rates below 0.5%. A large network meta-analysis for the cobalt chromium everolimus-eluting stent showed that it had a reduced rate of stent thrombosis compared with BMS.
The fear of late stent thrombosis that drove extended periods of DAPT is no longer justifiable. Many recent trials, including OPTIMIZE, have suggested that the shorter-duration DAPT is safe and effective in reducing stent thrombosis. Extended DAPT has a role in secondary prevention, as the DAPT trial has shown, but this should remain optional rather than mandatory.
All of the modern DES have the similar qualities of being more biocompatible with a low risk for stent thrombosis. There’s no evidence that using a BMS is any safer than a contemporary DES, while it is clear that BMS is inferior to DES in terms of restenosis. Now this is one study and the incidence of stent thrombosis is so low as to make these studies always underpowered, so you can take some caution in that. However, the stent thrombosis rates are low enough across multiple trials that an adequately powered trial would be neither feasible nor necessary.
This overall suggests that the role of BMS is very small and that the rate of BMS use, which is already less than 10% in many centers, is probably going to decline even further. I use BMS in less than 3% of cases at this point. Even for someone on anticoagulation or who is planning on having surgery after 1 month, DES with a short duration of DAPT would likely still be preferable over BMS.
This study answers the clinical question that continues to bother my referring physicians and fellows. They continue to instruct me: “Please don’t put in a DES” in patients who are on anticoagulants, who might need nonurgent knee surgery or who are at higher risk for bleeding. Data like those from SENIOR need to be disseminated widely to show that contemporary DES are not the same DES that we used 10 years ago, and that they now probably confer lower risk for stent thrombosis than BMS. For years, we did a good job training people that DAPT was necessary long term after DES to prevent a dire complication. Now we need to reeducate them that not only can DAPT durations be shortened, but that BMS is no longer any safer than DES.
Bedi: These findings are interesting because when bioresorbable vascular scaffolds (Absorb, Abbott Vascular) were introduced in the market, there were high hopes for them. The interventional community was looking at it very carefully, but consistently the ABSORB trials have shown poor performance of these BVS compared with the metallic everolimus-eluting stents.
The ABSORB III trial of 2,008 patients demonstrated that Absorb BVS is noninferior to cobalt-chromium DES (Xience, Abbott Vascular) at 1 year for CV outcomes. At 3-year follow-up, despite adequate vessel selection — meaning the investigators were very careful about not performing PCI in unstable lesions, bifurcations or a STEMI setting, instead performing PCI in patients with stable angina and nontortuous lesions — performance of the BVS was discouraging. At 3 years, for the BVS, there was a 13% rate of target lesion failure compared with 10% in the EES. Target vessel MI was 8.6% vs. 5.9%, a statistically significant difference. Even the TLR rates were higher, and this was predominantly driven by increased rates of device thrombosis. This was a statistically significant difference, 2.3% vs. 0.7%.
There was a signal that the BVS might be noninferior to the Xience EES at 1 year for CV interventions in noncomplex coronary lesions; however, at 3 years, there is a statistical significance showing the EES was superior to the BVS.
For the ABSORB IV trial of 2,604 patients more appropriately suited for the BVS, at 30 days in the intention-to-treat group, 5% of patients assigned BVS and 3.7% assigned EES had TLF with a noninferiority margin of 2.9%. Device thrombosis was seen in 0.6% of patients in the BVS group and 0.2% of those in the EES group (HR = 4.05; 95% CI, 0.86-19.07).
In the ABSORB II trial of 501 patients, at 4 years, the incidence of the patient-oriented composite endpoint of MI, death and revascularization was 24.9% in the EES group and 22.9% in the BVS group (HR = 0.9; 95% CI, 0.61-1.33). Definite or probable scaffold or stent thrombosis occurred in 2.8% of those in the BVS group, with none occurring in the EES group.
Ever since the ABSORB trials were first published, there’s been a concern among the interventional community. Abbott Vascular removed this stent from the market because of the poor outcomes, which had been consistently shown, and also poor sales. Despite using the current BVS technology in proper lesions, using proper sizing and performing proper lesion preparation, the outcomes of the BVS didn’t meet that of the current DES, predominantly because the DES have gotten so good.
What these findings add to our knowledge base is that the current technology is not there yet, but maybe we can improve the technology. I don’t think the results should kill the bioresorbable stents per se, but the company must determine how to make the BVS better so that it can perform comparably to or maybe even better than metallic DES.
Safirstein: The EXCEL Trial, presented initially at TCT 2016, showed that stenting was noninferior to surgery for the composite primary endpoint of death, MI or stroke at 3 years; in fact, stents were superior at 30 days for the same composite endpoint, but patients had to go back for repeat procedures more often over the course of 3 years. At TCT 2017, results of a quality of life substudy were presented. With regard to angina, the 1,788 patients in the trial scored similarly on the Seattle Angina Questionnaire at 1 year and 3 years. With regard to breathing, significantly more patients in the stent arm felt fine at 30 days compared to the surgical patients (60% vs. 44%; P < .01), but this difference was insignificant with further follow-up. Finally, clinically significant depression was more than double in the bypass group at 1 month (8% vs. 19%; P < .01) and still significant at 1 year, but similar at 36 months. This last finding is the most striking, as depression is no small issue at 12 months postoperatively.
Seto: The original EXCEL trial showed no difference in major adverse clinical events between CABG and PCI for left main disease. The EXCEL QOL substudy showed what many clinicians would expect, that CABG is associated with an upfront morbidity, which decreased quality of life at 1 month, but that also was relieved within 1 year. What was new was that the EXCEL QOL data also showed that late revascularizations after PCI didn’t seem to worsen quality of life at any time point.
Even with the 5% excess in repeat revascularization rate with PCI compared with CABG, there was no difference in quality of life at 3 years. These quality-of-life data support the separation of repeat revascularization from other clinical events such as MI or stroke in constructing composite endpoints for trials, as the consequences of revascularization were not equivalent.
Historically, left main with multivessel disease and left main disease in general was considered a class III indication for PCI, and only recently became class IIb (may be considered). The practice in the U.S., especially, is to send patients with left main and multivessel disease with a SYNTAX score to bypass surgery. Now, with no quality-of-life penalty at 3 years and equivalent clinical outcomes, you’re looking at potentially moving PCI up to IIa in any patient with concerns about upfront morbidity. For example, if someone had any qualms about taking on the upfront risk or if their life expectancy wasn’t going to be more than 3 years, you might favor PCI in that case.
This was a very high-quality study by a leading group of outcomes researchers. They measured multiple quality-of-life markers, both functional and symptom-driven — Seattle Angina Questionnaire, functional status — and they found no difference between the two revascularization techniques at 3 years. Whereas the current guidelines suggest CABG is preferred over PCI for multivessel disease, especially with left main, the results of the EXCEL-QOL study would justify patients having a choice between the two techniques.
After decades of trials comparing PCI vs. CABG, it is clear that revascularization rates with PCI are higher in the 1-to-5-year timeframe. The EXCEL study shows that this increase in revascularization rates was not associated with worse clinical outcomes or decreased quality of life at 3 years. Looking ahead, the ideal trial would also test the long-term (more than 10 years) outcomes of patients receiving CABG, as we know that 25% to 50% of saphenous vein grafts fail within 10 years. While these saphenous vein graft failures are largely clinically silent, it may very well be that a vein graft could be inferior to a DES, as bypass grafts are known to accelerate native vessel atherosclerosis, and potentially make a revascularization event more complicated. For a younger patient in particular, delaying bypass surgery with multivessel PCI may be worthwhile.
PARTNER 2A/SAPIEN 3 INTERMEDIATE RISK
Bedi: This, a long-term follow up of transcatheter aortic valve replacement vs. surgical AVR, is an important trial. TAVR was performed with the second-generation balloon-expandable valve (Sapien XT, Edwards Lifesciences), which is the generation before the Sapien 3, the TAVR valve we’re using currently for the majority of our patients. There were 2,000 patients enrolled. The mean follow-up was 2 years, and the mean patient age was about 80 years.
Notably, these were intermediate-risk patients, with a 4% to 8% Society of Thoracic Surgeons Predicted Risk of Mortality. Patients with a score of less than 4% could be included if there were coexisting conditions where the surgeon felt that this patient is at intermediate risk.
For TAVR, the all-cause mortality plus disabling stroke was 19.3% vs. 21.1% in surgical AVR, and this reached statistical significance for noninferiority, but not for superiority.
We did see some important characteristics of TAVR vs. surgical AVR. We saw that all-cause mortality was statistically different. The length of stay was lower with TAVR, 6 days vs. 9 days. The ICU stay was also statistically lower, 2 days vs. 4 days. The vascular complications were higher in the TAVR arm, 7.9% vs. 5%. However, life-threatening or disabling bleeding was significantly higher in the surgical arm; at 30 days, it was 43% vs. 10% in TAVR. Also, atrial fibrillation was higher in the surgical arm, 26% vs. 9% in the TAVR arm. The need for rehospitalization was almost the same.
Paravalvular regurgitation at 2 years was higher in the TAVR arm, 8% vs. 0.6%, but that will keep on improving because the next generation, the Sapien 3, has a skirt around the bottom edge, which has the potential to reduce paravalvular regurgitation.
This is a good trial because it demonstrates that clinically, TAVR is not inferior to surgical AVR. The researchers also did a cost analysis, and it showed that although the upfront costs are higher, it was predominately driven by the cost of the valve. Most of the hospitals paid $32,000 to $33,000 for the TAVR valve vs. approximately $5,000 for the surgical valve.
Overall, at 1 year of follow up, the transcatheter strategy saved about $9,000 to $10,000. That demonstrates that the overall cost of care might be lower even if upfront costs are higher.
The newer-generation valves have reduced stroke rates, pacemaker implantation rates and paravalvular regurgitation rates further.
Boston Scientific and Abbott/St. Jude Medical are working on new valves. Those and other advances could be helpful moving forward.
Bedi: DAPT STEMI demonstrates that 6 months of dual antiplatelet therapy is noninferior to 12 months of DAPT for clinical outcomes among patients who underwent primary PCI for STEMI.
In this trial, the researchers aimed to assess the safety and efficacy of 6 months vs. 12 months of DAPT in patients who presented with STEMI and underwent PCI with a second-generation DES.
There were 433 patients randomly assigned DAPT for 6 months and 437 patients randomly assigned DAPT for 12 months. Follow-up was 1 year, and 23% of the patients were female.
The primary outcomes analyzed were all-cause mortality, MI, rate of revascularization, stroke and major bleeding at 18 months.
The rate of all of the primary outcomes combined was 4.8% for 6-month DAPT vs. 6.6% for 12-month DAPT; shorter DAPT reached significance for noninferiority, although it didn’t reach statistical significance for superiority.
All-cause mortality was also less in the 6-month group, but there was no statistical difference in all-cause mortality, MI and major bleeding. STEMI major bleeding was 0.2% in the 6-month group vs. 0.5% in the 12-month group, but it didn’t reach statistical significance. MI was exactly the same at 1.8% in both arms.
The decreased duration of DAPT did not lead to an increased rate of thrombosis. In the 6-month arm, it was 0.7% vs. 0.9% in the 12-month arm, with no significant difference. There was also no statistical significance in target lesion failure, 1.2% vs. 1.8%.
This trial demonstrated that 6-month DAPT is noninferior to 12-month DAPT. The trial was not powered enough to demonstrate superiority.
Currently, the guidelines are that for stable angina, it’s OK to stop DAPT after 6 months, but with ACS, to continue DAPT for 12 months. This trial invokes the thought that we need additional trials in this arena so we can have more robust data. Then maybe it might affect the guidelines.
Disclosures: Bedi, Feldman, Safirstein and Seto report no relevant financial disclosures. O’Neill reports he is a consultant for Abiomed and Boston Scientific.