What’s your cath lab’s normal/non-actionable coronaries rate? I don’t know what ours is, so we’ve sent a registrar to gather some of our data to discuss in the department.
Are you like me, disappointed when you see an out-patient whose invasive angio finds minimal CAD, where even a pressure wire doesn’t seem justified? It really isn’t the proceduralist that simply sees that the patient doesn’t need a PCI, but the physician that thinks “this must be a failing in our diagnostic testing, and did we just subject this patient to an invasive test, with a very small, but appreciable risk of potentially important complications, when they really didn’t need it?”
There’s contemporary data that shows a percentage of patients with actionable coronaries at catheter is only around 45%, surely that’s not good enough? But what is the optimal rate?
Thinking mostly about out-patients investigated for chest pain (but also those having a routine angiogram prior to valve surgery) do we find too many normals, showing that we don’t use our resources well enough? With the cath lab being the most labour-intense, costly and often most pressurised resource in NHS departments, how to keep it for the most appropriate patients, in whom further action is taken, is an important consideration.
With increasingly sensitive and specific non-invasive diagnostic tests why would elective chest pain patients come to the cath lab and be found to have non-obstructive coronaries?
Well, some patients with a high CV risk will be indicated to have invasive coronary angiography (ICA) initially by NICE guidance, and a small number of these will have non-obstructive disease. And in the lower risk groups having anatomical or ischaemia testing, there may be an equivocal result, or a negative result but ongoing symptoms, and with no alternative diagnosis, they will go to have an ICA as final adjudication. However, these probably don’t account for the whole group of non-actionable coronaries.
In the last year, evidence has moved forward our knowledge of not only the increasing accuracy of non-invasive testing, but specifically the better gate-keeping of access to the cath lab, and improving the selection of patients who were found to have actionable CAD.
For example: the CE-MARC 2 study used functional imaging (with MPI and Stress MR) compared to NICE guideline-directed care in patients with suspected CAD, finding there was significantly more unnecessary ICAs in the guidelines group compared to those having MPI or MRI, with no difference in short-term MACE.
Also, the SCOT-HEART study randomly assigned chest pain clinic patients to CTCA compared to standard care. In the CT group, as well as showing an increase in the certainty of attributing symptoms to angina, it described a more frequent change in diagnostic and therapeutic approaches, with an increase in the number of ICAs, but most of those having obstructive CAD. The CTCA group was probably associated with an increase in coronary revascularisations, suggesting more appropriate selection of patients with actionable CAD.
And further, with CTCA, the additional use of FFRCT can make the selection of those who will have actionable coronary disease even more precise. In the PLATFORM trial incorporating FFRCT was associated with a significantly lower rate of ICAs with non-obstructive CAD.
At the moment of course, access to FFRCT has clear limitations both in cost and being available from only one vendor, but particularly with recent NICE approval, it is likely the cost will fall, and there will be wider access once competing systems roll out.
But FFRCT aside, access to the expertise of standard imaging modalities is variable and limited in some parts of the country: the BSCI recently highlighted that there would need to be a 700% increase in current CTCA provision to respond to the expected change in practice suggested by the updated NICE guidelines on chest pain assessment.
Shouldn’t we also be able to make a difference in reducing the normals in those who have a routine ICA pre-cardiac valve surgery? Guidelines recommend all patients with symptoms of angina or coronary risk factors, including men aged > 40 years and postmenopausal women, have evaluation for CAD prior to valve surgery. But there are data showing that a significant number of patients could have this by CTCA and not need to have it done invasively when using appropriate selection criteria.
The limiting step may also be convincing the cardiac surgeons that the data is robust and accurate, as of course, they feel familiar and comfortable with an ICA. Getting our surgeons down to the CT reporting room to see the reconstructed images and convincing them it is as sensitive as the catheter study has been difficult. But, it does seem that more surgeons are accepting of CTCA data now. This is likely to be quite centre-specific, I wonder what the proportions are across the UK, and compared to other similar countries?
So, although we recognise that ICAs are increasingly safe, and an expert can take less than 10-15 minutes to do a simple diagnostic test, I think it is definitely time to reduce the number of non-actionable coronaries being found in the cath lab. We need to be more systematic in our gate-keeping, help all practitioners to trust the increasingly accurate non-invasive imaging, which need to be adequately resourced, for the benefit of our patients. I’ll be interested to hear your thoughts on here, or on Twitter.