Contrast Medium Reduction for CTA with Photon-counting CT: A New Opportunity or More of the Same?

See also the article by Higashigaito et al in this issue.

James Dundas, MBBS, MRCP, is an advanced cardiac imaging fellow at the University of British Columbia and St Paul’s Hospital, Vancouver, British Columbia, Canada. He graduated from the University of Newcastle-upon-Tyne, England in 2010 and completed training in cardiology in North East England in 2022. He is accredited by the British Society of Echocardiography in both transthoracic and transesophageal echocardiography and by the European Association of Cardiovascular Imaging at level 3 in cardiac MRI.

Jonathon A. Leipsic, MD, FRCPC, MSCCT, is professor and head of the department of radiology and professor of cardiology with the University of British Columbia. Dr Leipsic is also a Canada Research Chair in Advanced Cardiac Imaging. Dr Leipsic has more than 700 peer-reviewed manuscripts in press or in print and more than 350 scientific abstracts and is the editor of two textbooks. He speaks internationally on a number of cardiac imaging topics, with more than 150 invited lectures in the last 4 years. He is also past president of the Society of Cardiovascular CT (2015–2016) and was awarded its Gold Medal in 2019 and was recently given the prestigious top 1% most impactful scientists designation by the Web of Science for the 3rd year in a row (2019–2021).

An investment in knowledge always pays the best interest.

– Benjamin Franklin

Photon-counting detector CT (PCD CT) is an emerging technology with numerous potential applications for advances in cardiovascular CT. Increased spatial resolution and reduced blooming (1) compared with existing energy-integrating detector CT (EID CT) could bring improved specificity in the realm of coronary CT angiography (CTA), particularly in smaller arteries and those with stents or calcified atheroma. Exploiting the K edge of iodine and multiple energy-selective images to perform material decomposition (1), PCD CT can help differentiate contrast media from calcium. This can be used to generate virtual noncontrast imaging for coronary calcium scoring or acute aortic syndrome assessment, without a separate acquisition. The same principle allows PCD CT to acquire differently timed contrast phases in a single acquisition following administration of both gadolinium-based and iodinated contrast media, which could be particularly useful in the setting of complex congenital heart disease or could be exploited to rival cardiac MRI in the assessment of myocardial fibrosis and infarction (2).

As PCD CT helps discriminate between high- and low-energy photons, it can be used to reduce beam hardening and metallic artifacts (1) around implanted cardiac devices, which may improve the assessment of degenerating or infected bioprosthetic valves and pacing leads. Translating improved contrast-to-noise ratio (CNR) into reduced contrast media dosing with similar image quality may be useful in patients with impaired renal function or those who require serial assessments, such as following aortic surgery and endovascular procedures.

Prior randomized studies comparing the use of dual energy to that of standard single-energy EID CT in pulmonary (3) and coronary (4) CTA demonstrated that a reduction in contrast material dose could be achieved while maintaining both objective image quality and subjective image interpretability. However, the subjective image quality scores for coronary CTA were poorer with dual-energy reduced-contrast acquisitions, perhaps suggesting that the gain in CNR from dual-energy CT was insufficient for dose reduction when accurate assessment of small mobile vessels is required. The greater spectral separation of PCD CT might allow it to deliver in this arena.

In this issue of Radiology: Cardiothoracic Imaging, Higashigaito et al (5) have shown similar findings assessing photon counting for follow-up aortic imaging. By reconstructing at a variety of energy levels, they found that virtual monoenergetic images at 50 keV optimized for a balance between objective and subjective image quality. After translating this into a 25% reduction in contrast media volume, they report PCD CT to be noninferior for both objective CNR and subjective image quality when compared with EID CT with standard contrast media volume. Importantly, this progress was on top of an already conservative contrast media protocol and using optimized tube energy, showing that photon counting can be an incremental step even with the mature technology and protocolized environment of modern CT practice.

Nonetheless, although certainly welcome, a relatively modest reduction in contrast media volume alone is unlikely to drive the adoption of PCD CT. Like any new technology, it is not without financial outlay and technical hurdles to overcome. Delivering benefit in clinically relevant outcomes would make a stronger case for embracing photon counting in routine clinical practice. We would advocate for studies looking for decision-making equipoise in aortic imaging with reduced contrast media volume—for example, demonstrating noninferiority at helping discriminate patients on either side of binary cutoff points defined for surgical intervention.

These data serve as a framework to explore such opportunities in coronary CTA. Here, while clinicians and patients would likely prefer less contrast media, they would not welcome this without the preservation of anatomic accuracy and the ability to evaluate and characterize atherosclerosis. Studies in low-risk patients showing similar rates of major adverse cardiac events with reduced contrast media volume would certainly be welcome. We would also advocate attempting to demonstrate improved assessment of luminal stenosis in severely calcific vessels with comparison to the reference standard of invasive angiography, rather than technical or subjective measures of scan quality. Clearly, coronary CT provides more easily defined possible end points to ensure that this technique can be adopted while maintaining anatomic accuracy and clinical utility. For the aortic space, as next steps to build on this hypothesis-generating work, we call on the field to generate incremental evidence with clinical end points to allow for broader clinical adoption.