QAngio XA 3D QFR and CAAS vFFR imaging software for assessing coronary stenosis during invasive coronary angiography

Clinical experts explained that in general, physiological testing using fractional flow reserve (FFR) and instantaneous wave-free ratio (iFR) is available but not frequently used in the UK. People typically have an invasive coronary angiography after a previous functional test (see section 2.4). Sometimes decisions about revascularisation are based on the images from the invasive coronary angiography, results of the previous tests and patient history. If the revascularisation decision is still uncertain after invasive coronary angiography, people may be referred for FFR or iFR. Using QAngio XA 3D quantitative flow ratio (QAngio QFR) and CAAS vessel FFR (CAAS vFFR) during invasive coronary angiography may provide more information to help with decision making. It could also mean that in some cases clinical decisions could be made without needing FFR.

A patient expert explained the potential benefits of testing using QAngio QFR or CAAS vFFR. These included reduced anxiety, discomfort and distress than more invasive testing, which may be needed if a definitive treatment decision cannot be made during the initial invasive coronary angiography. Using an invasive test like FFR with a pressure wire means using an adenosine infusion. The committee noted that around 30% of people may experience chest pain and shortness of breath from this. These side effects usually pass quickly but can be distressing. Around 3% of people may experience discomfort from the pressure wire itself and there is a small risk of rupture of the blood vessel. By avoiding adenosine infusion and a pressure wire, QAngio QFR and CAAS vFFR could reduce unpleasant side effects and risk of complications.

The committee noted that there were only 3 studies using CAAS vFFR that matched the inclusion criteria for review in the diagnostics assessment report. These included 500 patients. The external assessment group (EAG) explained that there was notable heterogeneity across this small number of studies and that the meta-analyses of the CAAS vFFR studies should be interpreted with caution. Where reported, there was a high exclusion rate because of angiographic image processing issues. In 2 of the studies, the technology was not used in the way it was intended (ILUMIEN I and Jin et al. 2019). The committee concluded that the diagnostic accuracy of CAAS vFFR was highly uncertain and recommended further research (see research recommendation 5.4).

The committee noted that 39 studies using QAngio QFR matched the inclusion criteria for the review in the diagnostics assessment report. These included 5,440 patients. These studies showed that QAngio QFR had good diagnostic accuracy to predict the FFR result. The clinical experts explained that there was good agreement between QFR and FFR values particularly at the extremes of measurement. While there was some disagreement between QFR and FFR results within the grey zone, (in the range of 0.78 to 0.84), the clinical experts noted that there is likely to be a limited clinical effect of not identifying someone with an FFR of between 0.76 and 0.80, that is, a false negative result. A more significant effect could occur for people with an FFR result of less than 0.76 who have a negative result on QAngio QFR (0.80 or higher). The EAG noted that modelling suggested around 3% of people with an FFR result of less than 0.76 would be misdiagnosed if using QFR for functional imaging. Clinical experts also explained that with FFR values close to the 0.80 cut-off, it is unclear whether there is any added benefit of revascularisation compared to optimal medical therapy. The committee concluded that there was good agreement between QAngio QFR and FFR values. Although there was some uncertainty around the grey zone, this was not a particular concern.

In the UK, invasive coronary angiography is usually done in diagnostic-only catheter laboratories or in interventional catheter laboratories that can also do percutaneous coronary intervention in the same procedure. Clinical experts explained that the quality of angiography images from diagnostic-only centres was generally lower than those from interventional centres. This is because in the diagnostic centre, invasive coronary angiography is done so the information can be used to guide decisions about what further testing and treatment might be needed. In the interventional centre, invasive coronary angiography is often done to help plan percutaneous coronary intervention. Clinical experts noted that QAngio QFR and CAAS vFFR need high-quality angiography images so in diagnostic centres the tests may have a high technical failure rate. The committee concluded that because all the data considered were from interventional centres, it was not certain what the technical failure rate would be in diagnostic-only centres.

The NICE guideline on assessment and diagnosis of chest pain of recent onset recommends invasive coronary angiography as a third-line test. People who have invasive coronary angiography should have already had a previous assessment such as 64-slice coronary angiography and non-invasive functional imaging tests, but this may vary between centres. Some people may also have HeartFlow FFRCT which is recommended in the NICE medical technologies guidance on HeartFlow FFRCT for estimating FFR from coronary CT angiography. The clinical experts explained that these previous assessments can rule out the need for interventional treatment. Therefore, it is likely that people who do go on to have invasive coronary angiography have more severe disease than the people in the diagnostic accuracy studies. While the previous functional assessments may be used to guide further testing decisions such as whether to do an FFR, the QAngio QFR and CAAS vFFR results would be used to guide high-level treatment decisions with substantial consequences. Therefore, clinicians need to be confident that making a decision based on the results of these tests would lead to improved outcomes for patients. The committee commented that the diagnostic accuracy studies did not incorporate clinical history and the effect that symptoms had on decision making based on the QAngio QFR result. Therefore, it is unclear how this additional information combined with a QAngio QFR or CAAS vFFR result affects clinical decision making about revascularisation.

The EAG did a simulation study analysis to investigate the possible effect of using QAngio QFR compared with invasive coronary angiography and FFR on coronary outcomes such as revascularisation rates and major adverse cardiovascular events. QAngio QFR (with or without a grey zone) led to slightly more revascularisations compared with FFR (40.2% revascularisations using FFR compared with 42.0% for QAngio QFR and 43.2% using the grey-zone strategy). Both methods prevented broadly the same number of major adverse cardiovascular events (FFR may prevent more major adverse cardiovascular events but only for 1 in 1,000 people). However, the committee noted that the simulation study made numerous assumptions, so its results were uncertain.

The clinical experts noted the lack of prospective outcome data when a QAngio QFR or CAAS vFFR-based approach was used to guide revascularisation decisions after invasive coronary angiography. The clinical experts explained that there was a need for clinical outcome data from large endpoint studies comparing these imaging software with FFR or invasive coronary angiography‑guided treatment. There are already multiple tests in the care pathway, and it was unclear how QAngio QFR or CAAS vFFR could offer additional clinical benefit. There are currently 2 ongoing clinical trials of QAngio QFR. The FAVOR III Europe-Japan study will compare QFR with standard FFR-guided percutaneous coronary intervention, and the FAVOR III China study will compare QFR with angiography‑alone guided percutaneous coronary intervention. These trials will be completed in December 2023 and February 2023, respectively. There is 1 ongoing trial of CAAS vFFR. The LIPSIA STRATEGY trial will compare vFFR with FFR for the assessment of intermediate coronary stenosis and is due to be completed in November 2026. The committee concluded that data from trials like these are essential to be confident that revascularisation decisions based on QAngio QFR or CAAS vFFR results would improve patient outcomes (see research recommendation 5.3).

In the model the procedural disutility for FFR was assumed to be the same as for percutaneous coronary intervention. This was because no data were available on the disutility of FFR. The clinical experts explained that this assumption may not accurately reflect the actual side effects or people's experiences of the procedure. The EAG also looked at different scenario analyses where the disutility of FFR was increased. The committee commented that a disutility equivalent to percutaneous coronary intervention was likely too high. This affected the cost effectiveness of FFR more than might be expected in clinical practice. It noted further that there were important differences between FFR and iFR that were considered the same in the model. Because iFR does not need a hyperaemic agent such as adenosine, it avoids the associated unpleasant side effects. This disutility was a key driver of the cost-effectiveness results, but because of a lack of evidence it was uncertain what disutility should be used. The committee recommended further research into the disutility associated with FFR (see research recommendation 5.1).

Test failure rates were high in the studies, especially the retrospective ones, because the invasive coronary angiography images were not good enough to run QAngio QFR and CAAS vFFR. The committee noted that in clinical practice some images may not be of a sufficient quality for the software programs to produce a result. Therefore, the cost per test may have been underestimated because test failure rates were not factored into the model. Only people who had a QAngio QFR or CAAS vFFR result were included. The EAG did a scenario analysis in which patient throughput was varied, which affected the cost per test. However, the committee noted that this may not have explored a wide enough range to sufficiently capture this effect. It suggested that failure rates in routine clinical practice would be reduced over time as the operator gained experience in using the system. However, the clinical experts explained that this may be dependent on the setting, with sub-optimal angiography images more likely in diagnostic-only centres (see section 4.5). The committee concluded that further research on QAngio QFR and CAAS vFFR failure rates in clinical practice would be beneficial (see research recommendation 5.2).

In the base case QAngio QFR was within the range NICE considers cost effective compared with invasive coronary angiography alone in both the deterministic and probabilistic analyses. However, the clinical experts commented that the strategy of invasive coronary angiography alone was not representative of clinical practice, where the results of previous tests and people's preferences would also influence a treatment decision (see section 4.6). The committee noted that similar results were seen for CAAS vFFR but concluded that there was greater uncertainty in this result because of the lack of diagnostic accuracy evidence.

Compared with the reference standard of FFR or iFR, QAngio QFR (with and without a grey zone) and CAAS vFFR were less cost effective (generated less quality-adjusted life years [QALYs] but were slightly cheaper) in the base-case analysis. However, the committee noted that a review of the accuracy of FFR or iFR was not done by the EAG and they were assumed in the model to be 100% accurate. It also noted that the difference between the new technologies and the reference standard was small at 0.007 QALYs or £140 per person for QAngio QFR and 0.011 QALYs or £220 per person for CAAS vFFR. Following an update to the price structure of QAngio QFR by the company during consultation, QAngio QFR using an annual licence became slightly cheaper but remained less clinically effective than the reference standard of FFR or iFR. However, given the small difference in costs and outcomes, the committee reiterated the need for clinical outcome data from studies that directly compare QAngio QFR and CAAS vFFR with FFR or iFR (see research recommendation 5.3). These data would give clinicians confidence in their decision making based on the results of the tests. The committee concluded that given the uncertainty in clinical utility the cost-effectiveness results were also uncertain.

In a scenario analysis in which the tests were done in a diagnostic-only setting, QAngio QFR and CAAS vFFR became the most cost effective options in the fully incremental analyses. This was because of the additional cost of onward referral for the reference standard tests. However, clinical experts explained that fewer people are having invasive coronary angiography because it is recommended as a third-line test in the NICE guideline on assessment and diagnosis of chest pain of recent onset (see section 4.6). The appropriate use of CT coronary angiography and functional testing has resulted in a fall in the number of people having invasive coronary angiography in diagnostic-only centres. Having an angiographic procedure in an interventional centre means that invasive coronary angiography, FFR or iFR and percutaneous coronary intervention can be done in a single visit, if appropriate. This reduces the need for multiple hospital visits, which has the potential to reduce people's anxiety. A clinical expert explained that according to 2017 to 2018 data from the National Institute of Cardiovascular Outcomes Research, around 35,000 invasive coronary angiography procedures were done in diagnostic-only settings, compared with around 205,000 in interventional centres. The committee concluded that the future role of QAngio QFR and CAAS vFFR in a diagnostic-only setting is unclear because diagnostic-only catheter laboratories are likely to decline in number.

The committee noted that QAngio QFR and CAAS vFFR were more cost effective than invasive coronary angiography alone. QAngio QFR using the proposed annual licence was slightly cheaper but less clinically effective than FFR or iFR. However, the incremental difference in terms of costs and QALYs between the tests was small. For CAAS vFFR, the committee recalled that the diagnostic accuracy data was highly uncertain so it could not be recommended for routine use. The committee noted further that there were no clinical outcome studies for QAngio QFR or CAAS vFFR, which meant that the EAG had to make assumptions about treatment decisions and clinical outcomes, which led to uncertainty in the results. The clinical experts commented that these tests may be used to guide high-level clinical decisions about treatment, so clinicians need to be confident when making decisions based on the tests' results. The committee considered that clinical utility is uncertain and more data are needed. There were concerns around the results of the simulation study that showed that QAngio QFR could lead to an increase in revascularisations (see section 4.7). The committee also recalled the trend in clinical practice of moving away from diagnostic-only settings to interventional centres (see section 4.13) and considered that there was too much uncertainty to consider QAngio QFR and CAAS vFFR in this scenario. The committee concluded that because of the uncertainty in diagnostic accuracy and clinical evidence, CAAS vFFR was not recommended. Despite having good diagnostic accuracy evidence, QAngio QFR should not be recommended for use until further data showing that it improves patient outcomes are available.