Impact of Post-Percutaneous Coronary Intervention Quantitative Flow Ratio of De Novo Coronary Artery Lesions Treated with Drug-coated Balloons: The QUADRIC Study

Background

Visual assessment of an angiogram to determine the significance of obstructive coronary artery disease has been the most widely used method for guiding decisions on percutaneous coronary intervention (PCI).¹ The 2018 European Society of Cardiology guidelines recommend using pressure wire-based physiology to assess the significance of intermediate-grade coronary artery stenosis before PCI.² However, this method remains underutilised due to potential side-effects from hyperaemic agents, cost and complications associated with the pressure wire.³ Studies have shown that pressure wire-based physiological assessment is significantly more accurate than visual estimation.⁴

Quantitative flow ratio (QFR) is a non-invasive technique that uses 3D coronary artery reconstruction and fluid dynamics computations from an angiogram to estimate fractional flow reserve without the need for a pressure wire or hyperaemic agent.^5,6 Whether QFR can predict major adverse cardiac events (MACE) in post-PCI patients with de novo lesions is unknown.

Therefore, we conducted the Quantitative Flow Ratio of de novo Coronary Artery Lesion Treated with Drug-Coated Balloon (QUADRIC) study to assess the use of post-PCI QFR in predicting MACE and the efficacy of drug-coated balloon (DCB) treatment for de novo coronary artery lesions.

Methods

Study Design

The QUADRIC study was an investigator-initiated, single-centre study conducted at the National Heart Institute, Kuala Lumpur, Malaysia. It was a real-world clinical assessment of post-PCI QFR use in predicting the outcomes of DCB treatment in patients. The study was approved by the institution’s ethics committee.

Patients

This study involved 414 patients aged 18 years and older who were admitted to the National Heart Institute, Kuala Lumpur, Malaysia between January 2020 and December 2020 for PCI with a DCB in a de novo lesion. Patients were excluded if they had flow-limiting dissection (thrombolysis in MI ≤2), residual stenosis greater than 30%, bail-out stenting following lesion preparation or treatment for in-stent restenosis.

Data Collection

A spreadsheet-based database (Microsoft Office 360 Excel) was created to collect relevant data. Most information was obtained from an in-house database including patient risk factors, vital signs, blood investigations and clinical outcomes from January 2020 to December 2020. Out-of-hospital data, including death or late complications, were obtained by tracing clinic notes.

QFR analysis was performed on all vessels pre- and post-PCI with a DCB by an independent, certified QFR provider. Intracoronary nitro-glycerine (100 µg) was administered before contrast injection to acquire cine images, which were analysed using QFR software version 2.2 (Medis Medical Imaging). The certified QFR providers were not involved in patient treatment and were blinded to clinical outcomes. The data was stored in spreadsheets and in the hospital computer system using a program called Synapse (FUJIFILM Medical Systems).

Outcome

The primary outcome was the 2-year MACE rate. MACE is defined as cardiac death, myocardial infarction distal to the culprit lesion and ischaemia-driven target lesion revascularisation during the respective follow-up period. We also aimed to assess the cumulative MACE rate at 24 months. The secondary outcome was the clinically-driven target lesion revascularisation rate (TLR) at 24 months.

Statistical Analysis

Multiple studies have shown that post-PCI QFR with a cutoff >0.90 may reduce MACE and improve long-term outcomes.^7,8 Data were extracted from an Excel database (Microsoft) with universal sampling and analysed using the Statistical Package for Social Sciences version 22.0 (IBM). Continuous variables were tested for normal distribution using the Kolmogorov–Smirnov test. All variables showed skewed distributions and were reported as the median and interquartile range. Comparisons between continuous variables were performed using the median test as appropriate. Categorical variables were reported as counts and percentages. Comparisons between categorical variables were carried out using Pearson’s χ² or Fisher’s exact tests, as appropriate. Cox regression analysis was performed to assess the relationships between predictor variables and MACE, providing hazard ratios and 95% confidence intervals. A p-value <0.05 was considered statistically significant.

Results

The baseline characteristics of our study population are shown in Table 1. A total of 414 patients with 429 lesions were included. After excluding 135 lesions that were not optimised for QFR calculation, we analysed a total of 294 lesions (Figure 1). The mean patient age was 61 years (SD ± 10.1) and 234 (79.6%) were men. Malay patients comprised 65.6% of our study population. A majority of patients had hypertension and more than half had diabetes with dyslipidaemia (Table 1).

A total of 118 (40.1%) patients were admitted for unstable angina and 28 (9.5%) were admitted for non-ST-elevation MI. Seventeen (5.8%) patients were admitted for ST-elevation MI and subsequently underwent DCB during primary angioplasty. The mean left ventricular ejection fraction was 46% (SD ±11.3). Most patients were on aspirin (98%), with 79% receiving clopidogrel as part of dual antiplatelet therapy.

Procedures were mostly conducted via the radial route (228 patients [77.6%]). The majority of the lesions were either in the left anterior descending artery or left circumflex artery (101 [34.3%] and 88 [29.9%], respectively; Table 2). The median post-PCI QFR was 0.92 (interquartile range 0.86–0.96). The distribution of post-PCI QFR values is shown in Figure 2.

The procedures were performed with a mean fluoroscopy time of 43 minutes (SD 272.9) and contrast volume of 196 ml (SD 68.0; Supplementary Table 1). The mean estimated lesion length was 32 mm (SD 20.2) and the mean vessel size was 2.5 mm (range: 2.0 mm–4.0 mm; SD 0.4). The majority of cases were performed using a 6 Fr guiding catheter (SD 0.3).

Two-Year Outcome

At the vessel level, we observed three in-hospital mortalities; two patients had a QFR <0.90 while one patient’s QFR was ≥0.90 as shown in Table 3. Both the QFR <0.90 and QFR ≥0.90 groups had similar MI rates (two patients in each group, p=0.639). We performed a Pearson’s χ² analysis to assess MACE between the QFR <0.90 group and the QFR ≥0.90 group post-angioplasty. The 2-year primary endpoint occurred in 14 (12.6%) patients in the QFR <0.90 group and in 8 (4.5%) patients in the QFR ≥0.90 group (p=0.011). This finding was also consistent for the secondary endpoints. Post-PCI QFR values were significantly lower in patients who experienced TLR during follow-up compared to those without TLR (p=0.032). All patients who had a MI underwent concomitant TLR.

Patients with dyslipidaemia (p=0.01) or renal failure (p<0.001) showed a significant association with MACE (Supplementary Table 2). Procedural variables and lesion types did not have a significant effect on adverse events (p>0.05; Supplementary Table 2). A comparison of baseline characteristics between the QFR <0.90 and QFR ≥0.90 groups did not show any significant differences (Supplementary Table 3). Cox regression analysis demonstrated that patients with QFR ≥0.90 had a 73.4% lower hazard of experiencing MACE compared to patients with QFR <0.90 (HR 0.266, 95% CI [0.08–0.84]; p=0.024; Supplementary Table 4).

Discussion

The QUADRIC study was conducted to assess the potential role of QFR in predicting both MACE and the efficacy of DCB after successful PCI in de novo coronary artery lesions. To minimise errors, we excluded all patients with poor-quality images unsuitable for QFR calculations. Multiple studies have shown that despite successful angiography-based PCI, negative outcomes may still occur if the PCI is physiologically suboptimal.^6,9 QFR calculations were performed offline by an independent certified QFR provider. The main findings are as follows:

First, post-PCI QFR values varied significantly, although the majority of treated vessels had higher QFR values. Second, post-PCI QFR was an independent predictor of adverse events. Third, QFR can be used to predict MACE or TLR in complex coronary interventions, including bifurcations, ostial lesions, long diffuse lesions or chronic total occlusions (CTO). These findings require larger randomised trials to provide stronger clinical evidence.

According to the third report of the DCB consensus group, DCB is only considered when an acceptable angiographic result has been attained, as evidenced by the absence of flow-limiting dissections, residual stenosis of less than 30% and fractional flow reserve (FFR) of more than 0.80.¹⁰ Multiple studies have been conducted to optimise post-PCI outcomes.^11,12 The rationale for post-lesion preparation FFR measurements is to evaluate residual disease burden which cannot be fully assessed through angiography alone. Despite extensive evidence supporting the use of FFR in post-PCI assessment, its adoption in daily practice remains negligible.¹³ Compared to FFR, QFR does not require adenosine administration, is non-invasive and can be performed within minutes. These features make QFR appealing for post-PCI assessment.¹⁴ Multiple studies have shown a strong correlation between QFR and pressure wire-based FFR measurements.^5,15

QFR is a novel approach based on dedicated software to estimate coronary physiology. Image acquisition and computation are relatively quick after adequate training and are reproducible.¹⁵ We found that QFR measurement following optimal PCI was feasible and easy. The accuracy and reproducibility of QFR depend on the quality of angiographic image acquisition, which requires two projections for each vessel.⁶ Lower QFR values predict a higher risk of adverse events.⁵ QFR is also useful for selecting lesions for PCI, as a previous study showed that a QFR-guided strategy can improve 1-year clinical outcomes compared to standard angiographic guidance.⁶ This study evaluates QFR with DCB in de novo lesions whereas previous studies have focused on stented vessels, small vessels or ISR lesions.^6,14,15

The QUADRIC study adds evidence that QFR can be used to predict future adverse events, not only in de novo small vessels but also in larger vessels, bifurcation lesions, long diffuse lesions and even in CTO. However, this evidence is preliminary, based on a small number of lesions (n=294) and should be confirmed in larger, randomised studies. Multiple studies have confirmed that imaging-guided PCI is associated with better outcomes in stented vessels.^16–18 Nevertheless, the clinical evidence of usage of imaging-guided PCI with DCB remains limited.¹⁶ Hence, non-invasive functional assessment of coronary arteries such as QFR following DCB may become the preferred treatment approach.

In our study, we performed QFR for all DCB treatments in de novo lesions, including hybrid strategy, bifurcation lesions, ostial lesions and CTO. In these cases, QFR helped us identify previously unnoticed lesions and quantify diffuse disease burden. The next generation of QFR systems will require only one projection and will be more automated, further reducing analysis variability and time.¹⁹ Currently, multiple ongoing studies are evaluating DCB in de novo large vessels (≥2.5 mm diameter), such as the REVERSE trial, and investigating hybrid strategies, such as in the Hyper II study. These trials will provide additional insights into our study results, as our study included DCB in large vessels up to 4.0 mm in diameter and in long diffuse lesions as part of a hybrid strategy.

In our QUADRIC study, we noticed that the 2-year cardiac death rate was relatively high in the QFR ≥0.90 group compared to the QFR <0.90 group. Upon further analysis, these patients were very ill, presenting with cardiogenic shock, recurrent ventricular arrhythmia and a history of multi-vessel PCI, findings consistent with current literature.^19,20 In the AQVA trial, the authors compared post-PCI outcomes between QFR-based and angiography-based PCI, demonstrating that QFR-based PCI was superior in achieving optimal post-PCI physiological results without significantly increasing procedural time, contrast dye use or radiation exposure.⁷

A study from South Korea also demonstrated the importance of statin therapy in reducing MACE, all-cause mortality and cardiac death compared with non-users after PCI in acute myocardial infarction patients.²¹ Zhu et al. reported that acute myocardial infarction is the leading cause of death in haemodialysis patients, attributed to high underlying levels of inflammatory cytokines, immune dysfunction and changes in mineral ion homeostasis.²² The Hawkeye group demonstrated that post-PCI QFR is a viable alternative to FFR in daily practice for predicting adverse events and poor outcomes.⁸

Study Limitations

Due to the single-centre nature of this study, a larger randomised trial is recommended to further evaluate the use of QFR in DCB-treated de novo coronary artery lesions. Additionally, patients with images not suitable for QFR calculations were excluded.

Conclusion

In this study, a post-PCI QFR <0.90 could predict MACE events in patients with de novo coronary artery lesions treated with DCBs and thereby improve patient outcomes within the first 2 years.

Click here to view Supplementary Material.