The future of interventional cardiology

Ischemic postconditioning (iPost) was first described in 2003 as a strategy capable of reducing the size of infarction after prolonged coronary occlusion in dogs through the immediate application of reperfusion after 3 cycles of 30 seconds of coronary reocclusion followed by 30 seconds of reperfusion.1 These results were soon confirmed independently, and the potential mechanisms involved described including, among others, a delayed normalization of pH levels, less accumulation of intracellular calcium, inhibition of the mitochondrial permeability transition pore, and less oxidative stress.2 Compared to the robust protective effect of ischemic preconditioning, it was confirmed that iPost was only beneficial if the procedure started right after reperfusion. However, it was attenuated in elderly subjects or in the presence of comorbidities or certain drug therapies.2,3

Despite these limitations, iPost soon called the attention of interventional cardiologists because it was easy to apply during primary percutaneous coronary intervention. Back in 2005 the very first study ever conducted in humans was published. In this study, iPost reduced the size of creatine kinase release compared to the control group in patients with ST-segment elevation myocardial infarction (STEMI).4 However, successive trials that estimated the size of infarction using similar methods or was more reliably measured by contrast-enhanced cardiac magnetic resonance imaging showed contradictory results. Some of these confirmed iPost protective effect while others revealed the opposite or even less myocardial salvage in patients treated with iPost compared to those who were iPost-naive.5^-7 So far, no clinical trial has been able to demonstrate that iPost reduces clinical events. The largest trial ever conducted is the DANAMI-3–iPOST that included 1234 patients with STEMI treated with primary percutaneous coronary intervention within the first 12 hours of disease progression and with the culprit artery occluded at the beginning of the procedure. These patients were randomized to receive iPost or a conventional percutaneous coronary intervention.8 After a median of 38 months of follow-up, the rate of the primary endpoint (death or hospitalization due to heart failure) was similar in both the iPost and the control group (10.5% vs 11.2%, respectively; non-significant P value) with no differences being reported in their individual components, other events, ST-segment elevation resolution or in the size of infarction measured by cardiac magnetic resonance imaging in a subgroup. A follow-up meta-analysis confirmed the lack of tangible clinical benefits in iPost in an aggregate population of 3619 patients with STEMI.9

Given these results, clinicians have consequently lost interest in this strategy. Therefore, iPost has not joined the therapeutic arsenal for the management of patients with STEMI. However, the reason behind the contradictory results of the mentioned trials is worth analyzing to identify, if any, subgroups of patients who could benefit from the protective effect of iPost. A possible explanation could be that the benefit of iPost depends on the duration of previous ischemia.10

In an article recently published in REC: Interventional Cardiology, Nuche et al.11 put this hypothesis to the test by comparing the effect of iPost on the size of infarction in a series of pigs undergoing left anterior descending coronary artery occlusion through 30-min balloon inflation (N = 19) to a different series from a previous report12 where occlusion went on for 40 min (N = 10). Except for the duration of ischemia, the experimental protocol was identical. iPost consisted of 4 cycles of balloon reinflation and deflation (1 min each) started 1 min after reperfusion. The area at risk was measured on the contrast-enhanced multidetector computed tomography scan with contrast during ischemia while the size of infarction was measured on the contrast-enhanced cardiac magnetic resonance imaging at 7 days.

iPost did not reduce the size of infarction in animals with 30-min coronary occlusion (0.3% [0.0-3.9] vs 0.9 [0.0-2.6] of left ventricular mass in animals treated with iPost or in the control group, respectively) or 40-min coronary occlusion (31.1% [27.3-32.8] vs 27.3 [25.1-27.5], respectively; both with non-significant P values]). Overall, T1 relaxation times were longer in animals treated with iPost. Authors conclude that iPost did not reduce the size of infarction in any of the 2 series, which goes against the possible interaction between its effect and the duration of previous ischemia. Also, longer T1 relaxation times—a marker of interstitial fibrosis—in animals with iPost suggests potential damage associated with the procedure.

The trial11 comes from a group of researchers with solid experience in the area, it is technically demanding, and has been conducted following a highly sophisticated methodology, for which the authors should be credited. Results go against an interaction between the benefit of iPost and the duration of previous ischemia. However, before this becomes the definitive conclusion, some methodological considerations should be made. In the first place, to assess the effect of any protective procedures, the size of infarction in the control group should have certain variability and, on average, should not be too large or too small.13 However, in this trial, after 30 min of ischemia barely any infarction was reported (3.8% [0.0-8.5] of the area at risk) while after 40 min infarctions were massive (98.2% [70.7-98.8] of the area at risk). Although these ischemia times were selected because they had caused medium-sized infarctions in former trials,14 homogeneity of the infarction size seen in both series and the almost non-existent infarctions in the 30 min series complicate discarding a possible beneficial effect of iPost in the results reported. Secondly, and on this regard too, it was surprising to see that by increasing ischemia time in just 10 min we went from almost non-existent infarctions to infarctions that occupy the entire area at risk. Although the experimental protocol was the same, as both series were conducted in different moments in time, variations in the conditions of the experiment such as animal breed, room temperature, materials used, etc, may have impacted the results and, therefore, cannot be ruled out. In this sense, results stress out the possible setback associated with the use of historic series. Finally, for the lack of a targeted anatomopathological study, a possible explanation for the massive infarctions reported in the 40 min series is that maybe some animals had coronary reocclusions between the end of the experiment and when the size of infarction was estimated 7 days later. Reocclusion is a common occurrence in this experimental model, especially when ischemia has been prolonged, and although the risk of ischemia drops with antiplatelet therapy (3 doses of clopidogrel were used in this trial) it does not go away completely.15

Despite these considerations, the truth is that the results of this study11 do not offer any signs of a potential cardioprotective effect of iPost by changing the ischemia times in this experimental model. This, added to the lack of clinical benefits reported in the previously mentioned trials confirms that, currently, iPost should not be used in patients with STEMI. This anticipates that it will be difficult to find a population of target patients in whom this procedure might be beneficial.

FUNDING

J.A. Barrabés received research funds from Instituto de Salud Carlos III (PI20/01681 project) and Centro de Investigación Biomédica en Red de Enfermedades Cardiovasculares (CIBERCV) co-financed by the European Regional Development Fund.

CONFLICTS OF INTEREST

None reported.

REFERENCES

The use of drug-coated balloons (DCB) to treat stenotic coronary artery lesions is a treatment strategy whose main asset is to avoid leaving a permanent intracoronary stent device. Although highly effective in the percutaneous coronary intervention setting, it is associated with a risk of acute thrombosis, future events like restenosis, and late thrombosis following processes known as neointimal proliferation, neoatherosclerosis or fractures of material. This could be even more relevant in younger patients with a long trajectory of possible coronary events ahead of them.

The use of DCB is widely accepted to treat in-stent restenosis and de novo lesions in small vessels1, and it is considered an interesting option in patients with high risk of bleeding. Another possible indication currently under scrutiny due to its possible potential is the management of bifurcations—one of the most interesting indications of all. However, clearly defined recommendations have not been established yet.2

Over the last few years, small clinical trials have been published on the use of DCB in this indication; although they have not proven definitive for a strong guideline recommendation, they provide valuable data. In general, trials have grouped into those looking into the safety and efficacy profile of DCB—without comparison group—and trials that compared strategies with DCBs or conventional balloons (CB).

PROSPECTIVE NON-RANDOMIZED TRIALS WITHOUT COMPARISON CONTROL GROUP

Table 1 shows 5 small trials (between 28 and 50 patients) including this type of different strategies with acceptable results regarding late lumen loss and safety.3-9

 

Table 1. Non-randomized, prospective clinical trials without comparison control group

Trial	Name or author and DCB	No. of patients	LLL TLR events, and restenosis	Restenosis, and MACE
DCB into both branches and BMS into the main branch	PEPCAD-V4 (Sequent Please B. Braun, Germany)	28	0.21 ± 0.48 in the SB 0.38 ± 0.46 mm in the MB Only 1 TLR (3.57%) and 3 restenoses (10.7%)	2 patients (7.14%) had late thrombosis at 6 and 8 months
Paclitaxel DES into the MB, and DCB into the SB	DEBSIDE (NCT01485081) (Danubio, France)	50	LLL in the SB: -0.04 ± 0.34 mm and in the MB: 0.54 ± 0.60 mm TLR in 1 patient (2%) Restenosis, 7.5.	1 AMI (2%) without cardiac deaths
-limus DES into the MB, and DCB into the SB	BIOLUX-A (www.anzctr.org.au, ID 335843) (Pantera Lux, Biotronik AG, SwitzeSBand)	35	LLL in the SB: 0.1 ± 0.43 mm 1 TLR (2.85%) No restenosis	1 patient died, and 3 AMIs were reported in different vessels
	SARPEDON5 (Pantera Lux, BIOTRONIK AG, Bülach, Switzerland)	50	TLR, 5.2% at 1 year 4% of restenosis in the MB, and 6% in the SB	Stent thrombosis, 0%
	Estudio de Valencia et al.6 (Sequent Please)	54	TLR, 3.6%	Overall mortality, 3.7%
DCB alone into both branches	Schulz et al.7 (Sequent Please)	39	10% restenosis, and all in the left main coronary artery bifurcation
	Bruch et al.8 (Sequent Please)	127	TLR, 4.5	MACE, 6.1% Use of bailout stent in 45%
DCB alone into 1 branch	Her et al.9 (Sequent Please) (Only in the MB)	16	There was a significant increase in the SB luminal area at 9 months, 0.37 mm2 ± 0.64 mm2; (P = .013), with a similar increase in the MB luminal area	The use of DCB alone in the MB also had a favorable impact on an area gain of 52% in the SB ostium
	Vaquerizo et al. (NCT01375465) (Eurocor GmbH, Germany) (Only in the SB and 001 lesions)	31	LLL in the SB, 0.32 mm2 ± 0.73 mm2, and binary restenosis, and TLR of 22.5%	High need for bailout BMS (14%) 1 AMI (3.2%)
AMI, acute myocardial infarction; BMS, bare metal stent; CB, conventional balloon; DCB, drug-coated balloon; DES, drug-eluting stent; LLL, late lumen loss; MACE, major adverse cardiovascular events; MB, main branch; SB, side branch; TLR, target lesion revascularization.

 

TRIALS COMPARING THE RESULTS TO DIFFERENT STRATEGIES AND 2 COMPARISON GROUPS, MOST OF THEM RANDOMIZED

Table 2 shows the 6 landmark trials comparing different strategies, 5 of them randomized,10-14 and 1 non-randomized.15

 

Table 2. Trials that compared the results with different strategies in 2 randomized comparison groups (except for the one conducted by Li et al.15)

Trial	Name and no. of patients	LLL	Restenosis and MACE, TLR events	Takeaway
DCB alone vs CB as a first-line therapy in lesions without damage to the proximal segment	PEPCAD-BIF10 (Sequent Please) 64 patients	LLL in the DCB group, 0.08 mm ± 0.31 mm vs 0.47 ± 0.61 mm in the CB group (P = .006).	Rates of restenosis of 26% vs 6% Rates of TLR of 9% vs 3% Favorable to DCB	In this type of lesions, stents were required in < 10% of the cases only
DCB vs CB in the SB with the use of BMS in the MB	DEBIUT11 (Dior-I, Eurocor GmbH, Germany) 117 patients A) DCB in both branches and BMS in the MB B) BMS in the MB, and CB in the SB C) Paclitaxel DES in the MB, and CB in the SB	LLL in the SB was 0.19 mm ± 0.66 mm in group A, 0.21 mm ± 0.57 mm in group B, and 0.11 mm ± 0.43 mm in group C (P = .001) LLL in the MB, 0.31 mm ± 0.48 mm in group A vs 0.16 mm ± 0.38 mm in group B (P = .15)	The rates of binary restenosis were 24.2%, 28,6%, and 15%; (P = .45), and the rates of MACE were 20%, 29.7%, and 17.5%; (P = .40) in groups A, B, and C, respectively	With this strategy, pretreatment of both branches with DCB was not superior to conventional BMS with the provisional stenting technique. Also, the use of DES was superior to DCB plus BMS
	BABILON12 (Sequent Please) 108 patients A) DCB in both branches, and BMS in the MB B) Everolimus DES in the MB, and CB in the SB	LLL in the SB, –0.04 mm ± 0.76 mm in group A vs -0.03 mm ± 0.51 mm in group B (P = .983)	The rates of MACE and TLR were higher in group A in the MB (17.3% vs 7.1% [P = .10], and 15.4% vs 3.6%; [P = .045]) due to more restenosis in the MB (13.5% vs 1.8%; P = .027)	Bifurcation pretreatment with DCB with BMS in the MB had more LLL and higher rates of MACE vs DES in the MB and CB in the SB Also, both strategies gave similar and very good results in the SB
Paclitaxel DES in the MB with CB vs DCB in the SB	Herrador et al.13 (Sequent Please) 50 patients	LLL, 0.40 mm ± 0.50 mm vs 0.09 mm ± 0.40 mm, (P = .01) favorable to the DCB group	The rates of SB restenosis were 20% vs 7%, (P = .08), and the rates of TLR, 22% vs 12% (P = .16)	The rates of MACE at 12 months were 24% vs 11% (P = .11)
-limus DES in the MB with CB vs DCB in the SB	BEYOND14, (Bingo, Yinyi Biotech, China) 222 patients with coronary bifurcation lesions excluding the left main coronary artery	Significantly lower LLL in the DCB compared to the CB group (–0.06 mm ± 0.32 mm vs 0.18 mm ± 0.34 mm; P < .0001)	The rates of restenosis were 28.7% vs 40% (P < .0001)	No differences regarding MACE (0.9% vs 3.7%, P = .16) or non-fatal AMI were found (0% vs 0.9%, P = .49)
	Li et al.15 (Sequent Please) NON-randomized	LLL of SB in the DCB group was lower compared to the CB group (0.11 mm ± 0.18 mm vs 0.19 mm ± 0.25 mm; P = .024) at 12-month follow-up	Multivariate COX analysis indicated that the DCB group had less MACE (23.9% vs 12.8%; P = .03)	Better results in the SB with DCB and fewer composite endpoints, but basically at the expense of unstable angina
AMI, acute myocardial infarction; BMS, bare metal stent; CB, conventional balloon; DCB, drug-coated balloon; DES, drug-eluting stent; LLL, late lumen loss; MACE, major adverse cardiovascular events; MB, main branch; SB, side branch; TLR, target lesion revascularization.

 

CONCLUSIONS FROM TRIAL RESULTS

1. 1. The use of bare metal stents (now in disuse) neutralizes all positive effects from the DCB in the main or side branch (DEBIUT11 and BABILON trials.12)
2. 2. In lesions without proximal damage to the bifurcation, an early strategy of DCB can only be considered in 1 or in both branches (PEPCAD-BIF.10) Also, non-flow-limiting dissections have good prognosis at follow-up.
3. 3. The use of DCB alone into the main branch can also have positive effects on the side branch ostium. Even using a limus-eluting stent in the main branch can only have a positive remodeling effect on the side branch ostium (aside from the study conducted by Her et al.,9 the BABILON trial already suggested it.12). In any case, the use of a DCB as a single stent-less strategy (unless results are poor or in the presence of flow-limiting dissections) seems like a reasonable option with a favorable long-term remodeling both in the main and side branches.
4. 4. The use of a limus-eluting stent in the main branch with a DCB implanted in the side branch (currently the most widely used strategy) can improve angiographic intraluminal parameters like late lumen loss or minimum lumen diameter without any significant clinical repercussions on the long-term events (the BEYOND trial.14). This is probably so because, in the other group, late lumen loss in the side branch is also small since events are more conditioned by the main compared to the side branch (BABILON12), and also because there are barely any myocardial infarctions or target lesion revascularizations associated with the side branch in any of the 2 groups.
5. 5. The results obtained with different balloons could also be different.

However, we should mention other aspects like vessel length, and not only vessel diameter since some studies demonstrate that length—and not diameter—can be a more important predictor of the impact side branch occlusion. Moreover, almost all these trials included side branch lesions < 10 mm in length, which is a well-known favorable predictor for the provisional stenting technique. Side branch lesions > 10 mm plus other signs of complexity like calcium, etc. can require the double stenting strategy, especially in left main coronary artery bifurcation lesions.16

Its role in more complex settings like left main coronary artery bifurcations or in-stent restenosis in bifurcations has also been studied, with reasonably good results.17,18

The article by Valencia et al.6 recently published in REC: Interventional Cardiology falls within the category of observational studies without control group that do not include angiographic measurements to allow, at least, a rough result comparison with other studies. This article combines treatment strategies like drug-eluting stent implantation into the main vessel in 71% of the cases or DCB alone into the main branch in 29% of the cases followed by DCB implantation into the side branch or DCB alone into the side branch, since 18% of the lesions were Medina 0,0,1 while, overall, 37.5% had no proximal damage.

According to the authors, this article contribution is the presentation of the clinical results of a small series of 54 patients with 55 lesions and the authors’ management of this type of lesions without excluding patients with higher risk of restenosis, as 32.1% of the patients with in-stent restenosis in the bifurcation and 8.9% with left main coronary artery lesions showed. Nevertheless the clinical outcomes are good with a median follow-up of 12 months. The rates of all-cause mortality, lesion thrombosis or infarction, and target lesion revascularization were 3.7%, 0%, and 3.6%, respectively, precisely in the most unfavorable cases of all, patients with in-stent restenosis.

The study limitations are obvious and well-established by the authors in the corresponding section. In brief, a small number of patients, no control group or angiographic follow-up, and the assumption that asymptomatic patients had no side branch restenosis. Also, since follow-up was not conducted on-site, possible developments of new Q waves associated with the side branch segment could not be detected. However, the study shows what many interventional cardiologists currently do in their cath labs and maintains interest for this strategy that should undoubtedly be taken into consideration when treating bifurcations. The most recent trials on drug-eluting stent and DCB implantation into the main and side branch, respectively, show good results in both branches, though with small differences in the repercussion of clinical events. Randomized clinical trials with a large cohort of patients are needed so that all possible trends favorable to the side branch become significant. Despite the presence of complex patients, the results from the trial conducted by Valencia et al.6 are good, promising, and their data welcome.

FUNDING

None whatsoever.

CONFLICTS OF INTEREST

None reported.

REFERENCES

Endorsed by the current clinical practice guidelines, the indication to perform percutaneous coronary intervention (PCI) of intermediate coronary stenosis should be guided by either fractional flow reserve (FFR) or instantaneous wave-free ratio (iFR) if evidence of ischemia is lacking.1 Despite these clear recommendations, the uptake of physiology in clinical practice remains low supporting the development of new non-invasive tools that no longer mandate the need for dedicated coronary guidewires or microcatheters along with the need to administer hyperemic agents in case of FFR.1

Advances in computational power and three-dimensional quantitative coronary angiography has facilitated the development of angiography-based-FFR indices, thus allowing easy, online physiological lesion assessments. Besides anatomical and angiographic exclusion criteria like severe tortuosity, aorto-ostial lesions or overlapping vessels, pivotal studies demonstrated that with angiography-based-FFR indices the need for invasive coronary artery instrumentation and hyperemic agents can, in most cases, be avoided.2

Currently, 4 angiography-based-FFR indices have emerged and are currently commercially available.1 Despite workflow differences and embedded simplified computational fluid dynamics models, these indices demonstrated to have a good diagnostic performance with pressure guidewire based FFR as a reference.1

Among these, vessel fractional flow reserve (vFFR, CAAS Workstation 8.5 Pie Medical Imaging, Netherlands) uses a computational fluid dynamic approach based on simplified Navier–Stokes equations and 2 angiographic views separated, at least, 30 degrees to generate a 3D reconstruction of the coronary artery. Using aortic pressure as inlet boundary condition, the algorithm applies automated and harmonized optimal end-diastolic frame selection in the 2 views by electrocardiogram triggering, thus allowing physiological lesion assessment without the need for full cardiac tree assessment or manual frame counting.3

This review provides an overview of the currently available clinical evidence on the use of vFFR (table 1 and figure 1).

Vessel fractional flow reserve was first validated in 2 retrospective, single-center studies where the technology demonstrated an excellent diagnostic performance in intermediate coronary artery lesions compared to FFR, which was consistent among different anatomical and patient subsets including tandem lesions, and patients presenting with non-ST-segment elevation acute coronary syndrome.3,4 These findings were later confirmed in the multicenter, prospective FAST II study, in which vFFR computed offline by local site personnel and a blinded core lab showed excellent diagnostic accuracy in identifying lesions with invasive guidewire-based FFR ≤ 0.80 (area under the curve [AUC], 0.93; P < .001). Positive and negative predictive values, sensitivity and specificity of vFFR were 90%, 90%, 81% and 95%, respectively.5 The system allows accurate automated vessel contour detection with manual correction required in merely 9.3% of vessel contours.5 Regarding reproducibility, vFFR showed a low inter-observer variability when computed offline by blinded academic operators (r = 0.95; P < .001) or local personnel vs a blinded core lab (r = 0.87; P < .001). Additionally, a low coefficient of variation (3.92%) was observed when vFFR was analyzed at 2 different timeframes by an independent core lab.6

Following these promising data, we explored the potential value of vFFR in a variety of clinical and procedural settings (table 1 and figure 1).

 

Table 1. Major studies trials investigating the diagnostic performance of vessel fractional flow reserve (vFFR)

Study/Author	Year	Study design	Number of vessel (patient)	Primary endpoint
Pre-PCI setting
FAST study	2019	Retrospective	100 (100)	AUC = 0.93 (95%CI, 0.88-0.97)
FAST EXTEND	2020	Retrospective	294 (294)	AUC = 0.94 (95%CI, 0.92-0.97)
FAST II	2021	Prospective	334 (334)	AUC = 0.93 (95%CI, 0.90-0.96)
FAST Heart Team	2022	Retrospective	1248 (416)	Mismatch between vFFR and revascularization = 29.8%
FAST III	Ongoing	Prospective
Imaging
Tomaniak et al.  (Left main coronary artery disease)	2022	Retrospective	63 (63)	AUC = 0.95 (95%CI, 0.89-1.0)
FAST OCT	Ongoing	Prospective
Post-PCI setting
FAST POST	2021	Retrospective	100 (100)	AUC = 0.98 (95%CI, 0.96-1.0)
FAST OUTCOME	2022	Retrospective	832 (748)	vFFR tertiles = TVF 24.6%, 21.5% vs 17.1%
STEMI and multivessel disease
FAST STEMI II	Ongoing	Prospective
95%CI, 95% confidence interval; AUC, area under the curve; PCI, percutaneous coronary intervention; STEMI, ST-segment elevation myocardial infarction; TVF, target vessel failure.

Figure 1. Clinical application of vessel fractional flow reserve (vFFR). LM, left main; PCI, percutaneous coronary intervention; STEMI, ST-segment elevation myocardial infarction.

 

First, the evaluation of left main coronary artery (LMCA) lesions remains challenging and often warrants a multimodality approach, including physiological assessment and intravascular imaging. Since patients with LMCA disease are often under-represented in the studies, a dedicated analysis comparing vFFR to intravascular ultrasound in patients with non-ostial LMCA disease was performed. vFFR was shown to correlate well to the LMCA minimum lumen area (MLA) as assessed by intravascular ultrasound (r = 0.79; P = .001) and to have excellent diagnostic accuracy identifying LMCA lesions with MLA < 6.0 mm² (AUC = 0.95; P = .001).7

Second, the use of physiology in the ACS setting has been topic of disussion as the benefit of physiology-guided-PCI has been mainly demonstrated in patients with stable disease.1 The latter is an important limitation since most patients present with ACS, which in up to 31% of cases occurs in the context of plaque rupture/erosion or calcium nodules located in intermediate coronary artery lesions. Conversely, a thrombotic component was identified in 602/695 of the culprit lesions (87%), which may affect the validity of both the pressure guidewire and the angiography-based-FFR assessments (TACTIS Registry, TCT 2022). In that perspective, the FAST OCT study (NCT04683133) will assess the agreement between vFFR and optical coherence tomography detected causes of luminal obstruction in intermediate lesions of patients presenting with ACS.

Whether the use of vFFR can be extended to patients with ST-segment elevation acute coronary syndrome and multivessel disease will be explored in the ongoing FAST STEMI program.

Next to the potential for online use, the concept of angiography-based-FFR caries significant potential in an offline setting where this technology could be used for clinical decision-making in patients with multivessel disease or those referred for heart team discussion. In a recent retrospective analysis, 3-vessel vFFR screening demonstrated a discordance between lesion significance and revascularization in 30% of the cases.8

Third, post-PCI physiological assessment has gained attention since several studies demonstrated that low post-PCI FFR values are detectable in up to 58% of vessels.9 Although the relevance of low post-PCI FFR was demonstrated by a significantly increased risk for future adverse cardiovascular events, the uptake of post-PCI FFR in the routine clinical practice is still limited.9 Hypothetically, the concept of having a wire-free method to detect suboptimal stent deployment, residual disease, and additional procedural optimization is promising. In the retrospective, single-center FAST POST study, vFFR demonstrated a good correlation with conventional invasive post-PCI FFR (r = 0.88), and a higher accuracy in the identification of patients with FFR values < 0.90 (AUC = 0.98) compared to three-dimensional quantitative coronary angiography (AUC = 0.62).10 In the light of these results, the hypothesis that post-PCI vFFR may predict future adverse cardiac events was proven in the FAST OUTCOME study.11

Fourth, the ability to predict functional outcomes of PCI may entail another step forward in the identification of patients who could benefit the most from PCI and thereby avoid the risk of a futile invasive procedure. Recent developments in vFFR software have allowed to simulate the effects of a ‘virtual’ PCI and estimate post-PCI FFR (residual vFFR). Using pre-PCI virtual pullbacks, residual vFFR showed a good correlation with invasive post-PCI FFR and post-PCI vFFR values (r = 0.84, and r = 0.77, respectively), and good discriminative ability to identify post-PCI FFR < 0.90 (AUC = 0.93).12 Of note, the current algorithm assumes an almost perfect PCI result, and thus, cannot account for heavy calcifications or stent underexpansion suggesting a potential need for future hybrid technologies combining multimodality invasive and non-invasive imaging modalities and physiology tools.

Finally, following the positive data from the FAVOR III outcome trial that proved the superiority of quantitative flow ratio (QFR, Pulse Medical Imaging Technology, China) vs angiography-guided-PCI in a Chinese population, the results of, at least, 5 currently ongoing angiography based FFR outcome trials (FAVOR III Europe Japan trial [NCT03729739], PIONEER IV [NCT04923191], FAST III [NCT04931771], LIPSIA STRATEGY [NCT03497637], FLASH FFR II [NCT04575207]) are eagerly awaited and may enhance guideline adoption.2 Specific to vFFR, the ongoing multicenter, randomized FAST III trial will assess whether a vFFR-based diagnostic strategy yields non-inferior clinical outcomes compared to an FFR-based strategy.

Up until the results of these studies will be released, angiography-based-FFR indices, including vFFR, remains an appealing alternative to conventional physiological indices in a broad selection of anatomical and clinical scenarios with the potential to increase the use of physiology and improve patient outcome.

FUNDING

None whatsoever.

AUTHOR’S CONTRIBUTIONS

A. Scoccia contributed to the drafting of this manuscript, and made a critical review of its intellectual content. J. Daemen also contributed to the drafting of this manuscript, made a critical review of its intellectual content, and gave his final approval to the version that would eventually be published.

CONFLICTS OF INTEREST

J. Daemen received institutional grant/research support from Astra Zeneca, Abbott Vascular, Boston Scientific, ACIST Medical, Medtronic, Microport, Pie Medical, and ReCor medical; and consultancy and speaker fees from Abiomed, ACIST medical, Boston Scientific, ReCor Medical, PulseCath, Pie Medical, Siemens Health Care and Medtronic. A. Scoccia declared no conflicts of interest.

REFERENCES

Diabetes mellitus is a comorbidity that is present in 20% to 30% of the patients with coronary artery disease and an indication for revascularization. Also, it poses a scenario of greater complexity for several reasons. The presence of diabetes is associated with more extensive, diffuse, calcified coronary artery disease, and graft and stent failure. All of it is associated with a higher risk of repeat revascularizations and worse prognosis for the patients, which is why diabetes is a differential element here since it establishes the revascularization method in patients with multivessel disease based on the clinical practice guidelines.1 Currently, the recommendation of coronary artery bypass graft (CABG) is superior to percutaneous coronary intervention (PCI) in diabetic patients. This indication comes from numerous studies being the FREEDOM trial2 one of the most important of all. However, are patients from the routine clinical practice or real-world patients similar to those included in these clinical trials?

In this sense, the study conducted by Puyol-Ruiz et al.3 recently published in REC: Interventional Cardiology provides valuable observational information on the results of coronary revascularization in diabetic patients in the routine clinical practice. This study shows the results from a historical cohort (2012-2014) of 733 patients with diabetes and multivessel coronary artery disease with a clinical indication for coronary angiography. Authors divide the study population based on the degree of revascularization (complete or incomplete) and the clinical profile consistent, or not, with the inclusion criteria of the FREEDOM clinical trial.2 In this cohort, 80.8% and 14.5% of the patients were revascularized percutaneously and surgically, respectively compared to 4.8% who received medical therapy only. Authors found a tendency towards a lower rate of clinical events at 35-month follow-up in patients with complete revascularization. Also, both the risk profile and the rate of events of the FREEDOM study population (41%) was significantly lower compared to the non-FREEDOM study population (59%): lower rate of death (5.5% vs 38.4%; P = .006), cardiac death (3.2% vs 31.2%; P = .002), and major adverse cardiovascular events (6.5% vs 40.0%; P = .012). Therefore, we can deduce that patients from the FREEDOM trial are a selected subpopulation of lower risk representative of less than half of the real-world patients with diabetes and multivessel disease. Other studies that have tried to identify, in a population from the real-world clinical practice, the group of patients potentially eligible for clinical trial show similar prevalences (around 50%) of selection criteria for clinical trials on coronary revascularization, a population that also shows a significantly lower rate of cardiovascular adverse events.4

On the other hand, regarding the interpretation of these data, we should remember that over 10 years have passed since the FREEDOM trial, and the recruitment phase into the cohort of the study conducted by Puyol-Ruiz et al.3. Let’s see what elements have changed in the revascularization of diabetic patients through all this time.

Modern PCI is not similar to the one described in the FREEDOM trial that used first-generation drug-eluting stents (sirolimus in 51%, and paclitaxel in 43%). Current platforms have exceeded paclitaxel-eluting stents in multiple clinical settings including diabetic patients.5 Sirolimus-eluting stents had higher rates of thrombosis and stent failure compared to current stents in relation to the mechanisms of hypersensitivity to polymer.6 Also, ultrathin strut drug-eluting stents have proven to be associated with a lower rate of adverse events compared to first-generation stents (> 120 μm).7 As a matter of fact, more recent studies with all-comers design have demonstrated that state-of-the-art stents like the polymer-free amphilimus-eluting stent improves results even more (target lesion failure) compared to second-generation reference stents.8 This huge improvement in stent technology seen over the last few years, the calcified plaque modification techniques used, and the intracoronary imaging modality-guided PCI performed or with pressure guidewires lead us to think that the current PCI results improve significantly those reported both in the FREEDOM trial and in this cohort of patients. Such statement can be confirmed in the comparison between the SYNTAX II cohort and the SYNTAX PCI group that used paclitaxel-eluting stents.9,10

However, CABG results have also improved, at least, in the clinical trials. For example, the 1-year rate of adverse events (death, myocardial infarction, stroke or repeat revascularization) has dropped from 12.4% in the CABG group of the SYNTAX trial10 down to 6.9% in the CABG group of the FAME 3 trial.11 This reduction is probably due to better perioperative care and the optimal medical therapy since no major changes in the surgical technique have been reported.

Regarding the impact diabetes has on the results of complete revascularization, the results from the study conducted by Puyol-Ruiz et al.3 are consistent with a meta-analysis of 28 studies and 83 695 patients published by Zimarino et al. This analysis revealed that complete revascularization produced similar benefits in diabetic and non-diabetic patients in terms of mortality and adverse events reporting, in the former, significantly lower rates of new myocardial infarctions.12 Despite this benefit, the numbers of residual coronary artery disease are still high both in the present study (CABG 49/106 [46.2%], PCI 396/592 [66.9%]), and in other PCI cohort studies (28.6%)13 or CABG (33.1% with residual SYNTAX score > 18.514). Therefore, there is this doubt on whether incomplete revascularization is just a technical problem or else a risk marker associated with a more advanced stage of the disease.

In conclusion, the therapeutic management of multivessel coronary artery disease in diabetic patients is still challenging for cardiology today. While clinical trials keep being conducted in selected low-risk populations it’ll be of paramount importance to complement them with information from the results of the actual clinical practice as this article did. In future studies we should make all the necessary efforts to use pragmatic designs where exclusion criteria are minimized to encourage their immediate applicability to the routine clinical practice.

FUNDING

None whatsoever.

AUTHORS’ CONTRIBUTIONS

P. Salinas designed, supervised, reviewed, and drafted the manuscript. A. Travieso drafted and reviewed the original manuscript version.

CONFLICTS OF INTEREST

None reported.

REFERENCES