COMPLETE Beats Culprit-Only PCI in Multivessel STEMI and Timing Doesn’t Matter

The results justify waiting to completely revascularize if a complex patient presents in the middle of the night, an expert says.

COMPLETE Beats Culprit-Only PCI in Multivessel STEMI and Timing Doesn’t Matter

PARIS, France—For STEMI patients presenting with multivessel disease, complete revascularization is superior to culprit-only PCI with regard to the composite of cardiovascular death or MI as well as cardiovascular death, MI, or ischemia-driven revascularization over a median follow-up of 3 years, according to results of the randomized COMPLETE trial. Importantly, the timing of complete revascularization—whether during or after the initial hospitalization—does not seem to affect outcomes.

The debate over whether this patient population should be treated with a complete revascularization strategy has leaned toward the affirmative in recent analyses, with the caveat that most studies in this field have been small or retrospective. Yet another question has been whether nonculprit lesions need to be treated immediately or whether operators can wait and address them in a staged procedure.

“The results are definitive . . . for hard endpoints,” Shamir Mehta, MD (McMaster University and Health Sciences, Hamilton, Canada), who presented the results today during the first Hot Line session here at the European Society of Cardiology (ESC) Congress 2019, told TCTMD. “As far as the timing goes, we certainly have a clue that performing [complete PCI] later is probably not harmful in a major way. On the other hand, you can turn it around and say there's no downside to doing it earlier, meaning 1 day after the procedure.”

COMPLETE was simultaneously published in the New England Journal of Medicine.

Complete Wins at 3 Years

For the study, the researchers randomized 4,041 STEMI patients with multivessel CAD who previously had successful culprit-lesion PCI at one of 140 centers in 31 countries to either undergo further complete revascularization or be managed on guideline-directed medical therapy alone. All patients had nonculprit lesions with at least 70% diameter stenosis or an fractional flow reserve (FFR) measurement of 0.80 or less, and the timing of complete PCI was left to operator discretion. No patients with cardiogenic shock were enrolled in this study.

Notably, patients were younger (mean age 62) and most were “exceptionally well treated,” according to Mehta, who highlighted that almost all patients received aspirin and about 71% received either ticagrelor or prasugrel at discharge. The largest proportion of nonculprit lesions, about 40%, were located in the LAD, followed by the circumflex (about 36%) and the RCA (about 24%). Also, all but approximately 10% of patients assigned to nonculprit-lesion PCI received complete revascularization (SYNTAX score = 0).

Over a median follow-up of 3 years, rates of the co-primary outcomes of cardiovascular death or new MI as well as cardiovascular death, new MI, or ischemia-driven revascularization were statistically lower for patients who received complete versus culprit-only PCI. Additionally, there was an advantage for complete PCI with regard to the key secondary outcome of cardiovascular death, MI, ischemia-driven revascularization, unstable angina, or NYHA class IV heart failure.

COMPLETE: Median Follow-up of 3 Years


Complete PCI

Culprit-Only PCI



95% CI




CV Death or New MI






CV Death, New MI,

or Ischemia-Driven Revascularization






CV Death, New MI, Ischemia-Driven

Revascularization, Unstable Angina,

or NYHA Class IV HF





Neither co-primary endpoint was affected by the timing of nonculprit PCI—during the initial hospitalization (median 1 day after the culprit-lesion PCI) or after discharge (median 23 days)—with interaction P values of 0.62 and 0.27, respectively.The largest effect seen on an individual outcome was the benefit of complete revascularization on MI, which was reduced by 32% (HR 0.68; 95% CI 0.53-0.86). There were no individual differences observed for either CV or all-cause death, but Mehta noted that COMPLETE was not powered for these endpoints.

“The main message here is that the action is late,” Mehta said. “These patients have multivessel disease. This isn't an issue of rushing to do the nonculprit lesion to prevent these events that are happening within the first day or the first month. These events occur over the long term, and that's really the value of a complete revascularization strategy.”

There were also no differences noted between the groups with regard to major bleeding, stroke, acute kidney injury, or stent thrombosis.

‘Solidify the Guidelines’

Speaking with TCTMD, B. Hadley Wilson, MD (Sanger Heart & Vascular Institute, Charlotte, NC), called COMPLETE “one of the most impactful, certainly for interventional cardiologists, trials” to come out of the ESC meeting. “It's welcome because it solidly answers this complete revascularization or ‘early-on’ strategy for patients that have nonculprit lesions that come in with an acute myocardial infarction,” he said.

“This will really solidify the guidelines in terms of culprit versus nonculprit at the time of initial presentation, or in the immediate period thereafter, that complete revascularization is superior for both the primary and secondary endpoints of CV death and also new MIs and ischemic revascularization,” Wilson continued. “That's really what we're all interested in: trying to prevent second events and reduce mortality and improve patient experience and outcomes.”

Also, the COMPLETE results can justify an operator’s decision to wait to address nonculprit lesions if, for example, a patient presents with a complex set of comorbidities and/or in the middle of the night when the STEMI team might not be optimally staffed, he said.

“We have to take the whole situation into consideration, which is not just the lesions or the blockage or the stenoses, but we have to [think about] the patient—how they're doing with other factors outside of just their immediate heart attack. How is their kidney function? How [are] their other comorbidities?” Wilson said. “In the middle of the night you have to consider your team and the performance of the cardiologists. Are they going to be as efficient and do as well with a second blockage as they did with the first considering all those other fatigue factors and contrast and radiation?”

While Wilson said he would have liked to have seen a drop in the individual endpoints of all-cause and cardiovascular death with complete PCI, he acknowledged that the benefit seen with regards to the other endpoints is “super strong.” Add in the fact that the researchers primarily used the latest antiplatelet agents as well as radial access most of the time, these results make up “certainly the largest and the most powerful study in this area that answers a lot of questions solidly and supports the guidelines,” Wilson said. “It's going to be very difficult to show any other major additions or improvements in this area, because this was a much larger study than any others that have been done and it'll be difficult to reproduce this.”

Questions Remain About FFR, Complex Lesions

In an accompanying editorial, Lars Køber, MD, and Thomas Engstrøm, MD (University of Copenhagen, Denmark), write that the results of this trial support the guidelines recommending “a strategy of full revascularization in patients with STEMI and multivessel disease, at least in those who have suitable nonculprit lesions.”

In the past, many have believed that the decision to do nonculprit PCI should be based solely on FFR, they note. Because “nearly 60% of the lesions had at least 80% stenosis of the vessel diameter on visual estimation and 38% were in the left anterior descending artery [in COMPLETE,] . . . most lesions were angiographically significant, and FFR may still have an important role in diagnosing lesions of intermediate severity,” Køber and Engstrøm say.

Discussing the study during the main session, Stefan James, MD (Uppsala University, Sweden), agreed, saying “it's unclear today if stratification by intracoronary physiology such as FFR would have changed the results in a more favorable direction.”

Additionally, patients included in trials are sicker than the general population, “and extrapolation of the results to patients with a greater risk of complications may not be safe,” the editorialists caution. Given the “relatively low” mean SYNTAX score of around 16 in COMPLETE, “more complex nonculprit lesions (associated with higher SYNTAX scores) may be different physiologically and may be less suitable for routine treatment. Also some patients may benefit more from firm adherence to high-potency dual antiplatelet therapy with either prasugrel or ticagrelor,” they add.

With the apparent safety and event risk reduction associated with complete PCI in this study, “it appears to be appropriate to recommend complete revascularization for patients similar to those included in the COMPLETE trial,” they conclude. “We hope that the investigators will be able to obtain data from longer follow-up in order to evaluate whether the tendency toward a small reduction in all-cause mortality becomes significant over time.”

Ultimately, COMPLETE has several strengths including “definitive, very clear results [and] hard clinical endpoints with very large reductions,” James said. “It had a very low crossover rate, and it was consistent across all major subgroups and with no safety concerns.” Because of this, he argued that the next update of international guidelines should include complete revascularization in a second procedure as a “class 1a” recommendation in for STEMI patients with multivessel disease.

  • Mehta SR, Wood DA, Storey RF, et al. Complete revascularization with multivessel PCI for myocardial infarction. N Engl J Med. 2019;Epub ahead of print.

  • Køber L, Engstrøm T. A more COMPLETE picture of revascularization in STEMI. N Engl J Med. 2019;Epub ahead of print.

  • COMPLETE was funded by the Canadian Institutes of Health Research.
  • Mehta reports receiving grants from AstraZeneca and Boston Scientific.
  • Wilson reports no relevant conflicts of interest.
  • Køber reports serving as a speaker for Novartis and Bristol-Myers Squibb and receiving grants from AstraZeneca.
  • Engstrøm reports receiving speakers fees from Abbott, AstraZeneca, and BMS and serving on the advisory board for Bayer AS.