Author + information
- George A. Beller, MD⁎ ( and )
- Michael Ragosta, MD
- ↵⁎Reprint requests and correspondence:
Dr. George A. Beller, Box 800158, Cardiovascular Division, University of Virginia Health System, Charlottesville, Virginia 22908
- coronary artery disease
- fractional flow reserve
- myocardial perfusion imaging
- positron emission tomography
Angiography forms the basis of most revascularization decisions. This approach is perfectly reasonable when the angiogram clearly demonstrates either a severely stenosed coronary artery or a normal one. However, angiography has well-known limitations and the significance of lesions of only moderate severity is often difficult to determine based on just the angiogram (1). This uncertainty may result in inappropriate care with stenting of nonflow limiting lesions or failure to revascularize significant ones.
When confronted with an ambiguous angiogram, additional testing is required to make a confident decision. Myocardial perfusion imaging (MPI) could be used to determine the presence of ischemia in the vascular territory supplied by the suspect artery. Although noninvasive imaging may help discern the significance of a coronary lesion when there is single-vessel disease, the report by Melikian et al. (2) in this issue of JACC: Cardiovascular Interventions is similar to our earlier study (3), suggesting that MPI cannot be used to make these decisions in the setting of multivessel disease. Because it relies on relative flow heterogeneity, MPI usually identifies ischemia caused by the most severe stenosis; it may misclassify as normal other vascular zones supplied by less diseased but still significantly narrowed arteries. In this study by Melikian et al. (2), MPI, compared with fractional flow reserve, underestimated the number of ischemic territories in 36% of patients with multivessel disease.
Invasive adjunctive techniques commonly used to assess lesion significance include intravascular ultrasound and fractional flow reserve (FFR). The anatomically based intravascular ultrasound identifies the most narrowed lumen from a specific cross-section of the artery; a minimal luminal area <4.0 mm2 is associated with ischemia but only applies to proximal and mid segments of major epicardial coronary arteries. Fractional flow reserve is physiologically based and describes the ratio of the maximum achievable flow in the presence of a stenosis to the theoretical maximum flow in the same vessel in the absence of a stenosis. It takes into consideration the multiple, complex variables influencing coronary flow including lesion severity, lesion length, and collateral flow. The mathematics, experimental basis, technique, limitations, and validation of FFR have been well described (4–6).
The unequivocal normal value of 1.0 is well accepted and has been firmly established in humans; the value below which a stenosis is deemed “significant” is of some debate. Although the initial validation studies determined that an FFR <0.75 most strongly correlated with ischemia, coronary stenoses with FFR between 0.75 and 0.80 have been considered “borderline” and may, in fact, be significant; currently, most clinicians and investigators consider an FFR <0.80 as “ischemic” (7). Importantly, revascularization of lesions with nonischemic FFR can safely be deferred, thereby establishing FFR as a valuable tool and important adjunct to angiography in clinical decision making (8).
The study by Melikian et al. (2) included patients with arteries with >50% stenosis determined by visual estimation. This is a potential limitation of the study not addressed by the investigators. Overestimation of stenosis severity by visual analysis might have led to the inclusion of an excessive number of lesions only mildly diseased. It is interesting to note that for the entire cohort of 67 patients, FFR was >0.80 (i.e., insignificant) in all 3 vessels in 20 patients and was <0.80 in just 1 vessel in 20 patients. Based on FFR criteria, 60% of the study cohort had either no or single-vessel ischemia. The quantitative coronary angiographic findings determined that the average coronary stenosis was 50% to 60%. Taken together, the patient population was one of more mild to moderate coronary artery disease (CAD). Thus, it is not surprising that single-photon emission computed tomography (SPECT) MPI, compared with FFR, missed some of these mild stenoses because the sensitivity of SPECT for detecting ischemia is highest for stenoses of ≥70% severity. Had the patient population in the study of Melikian et al. (2) exhibited more severe angiographic CAD, more reversible perfusion defects might have been observed in coronary territories with an FFR of <0.80. An additional limitation relates to the difficulty in measuring FFR in totally occluded arteries. The investigators assigned an FFR value of 0.5 to these occluded arteries. Although this is based on reasonable assumptions, some patients might have had well-developed collaterals and the true FFR, if measurable, might have been significantly higher. This issue would come into play for those patients with occluded arteries in whom the MPI failed to show ischemia in this zone and thus were classified as false negatives.
It is surprising that the investigators found a significant number of patients with reversible ischemia with a corresponding FFR ≥0.80. The conclusion that 20% (24 of 121) of coronary territories with reversible defects on MPI but a normal FFR represents an “overestimation” of ischemia by MPI may not be correct. Measurement of simultaneous FFR and absolute coronary flow reserve in these cases might have discovered microvascular dysfunction in the setting of a nonflow limiting epicardial coronary artery as an explanation. Severe endothelial dysfunction can also result in myocardial defects on SPECT in some patients (e.g., diabetic) who have no significant epicardial stenoses (9).
Melikian et al. (2) defined an ischemic coronary supply region on SPECT MPI if only 1 segment showed reversibility. This may have lowered the specificity of MPI because just 1 segment exhibiting reversibility in an entire risk region could be a false positive finding. Requiring reversibility in ≥2 adjacent segments for an ischemic designation may have improved the specificity of reversibility for identifying ischemia. Some fixed defects on SPECT associated with an FFR of >0.80 may have represented attenuation artifacts that might have been eliminated with MPI attenuation correction.
A minor methodologic problem for studies that compare segments on a noninvasive myocardial perfusion scan with coronary supply regions on coronary angiography is a lack of perfect registration using the 17-segment model for both techniques. This may be due to overlap of segments between coronary territories. For example, in the event of stenoses in both the distal left anterior descending and posterior descending branch of the right coronary artery, if FFR is abnormal in both vessels, but the assignment of an apical defect was made solely to the right coronary artery, then this would yield a “false negative” SPECT study for left anterior descending stenosis detection.
Despite some of these limitations, the study by Melikian et al. (2) is clinically very valuable primarily because it again demonstrates that significant stenoses would be overlooked if a clinician relied solely upon perfusion scintigraphy to determine if an ambiguous stenosis causes ischemia. Myocardial perfusion imaging correctly identifies the most severe stenosis, and, as a diagnostic tool in patients with stable angina, likely gets patients “into the cath lab.” Yet, it cannot be subsequently used for revascularization decisions regarding vessels supplying other zones. This study also provides further evidence that FFR is indispensable to the current practice of interventional cardiology. Clearly, angiography alone is often inadequate for many revascularization decisions. This study (2) and others (e.g., Ragosta et al. ) confirm that angiography coupled with MPI is also inadequate for decision making in patients with multivessel disease.
Significant advances have been made in MPI that have enhanced the detection of ischemia and have improved the identification of physiologically important coronary stenoses that exhibit abnormal flow reserve with vasodilator stress. Positron emission tomography (PET) MPI has high temporal resolution permitting dynamic imaging for measuring absolute myocardial blood flow in ml/min/g at rest and stress at the time of tracer injection (10–12). As mentioned previously, some patients with 3-vessel disease (5% to 10%) have balanced ischemia yielding uniform uptake of an injected flow tracer at peak vasodilator stress (13). In this situation, coronary flow reserve is uniformly reduced in the supply regions of all 3 major coronary stenotic arteries. Quantifying absolute flow during rest and stress and measuring coronary flow reserve with dynamic PET MPI and tracer kinetic modeling can identify these patients with balanced ischemia thereby markedly increasing the detection rate of functionally significant multivessel coronary stenoses. One study (12) using N-13-ammonia PET MPI found that patients with uniform tracer uptake and abnormal flow reserve by quantitative PET perfusion analysis was independently associated with a higher annual event rate over 3 years than those with normal coronary flow reserve. Another study (11) reported that the area under the receiver-operator characteristic curve for peak adenosine absolute blood flow was 0.90 versus 0.69 for assessing relative perfusion by the conventional MPI imaging technique. It can be expected that when more quantitative methods for assessing coronary flow reserve are introduced into the clinical setting, the correlation between MPI ischemia and FFR will become greater with fewer functionally significant coronary stenoses missed.
In conclusion, it is common practice for physicians to make revascularization decisions in the catheterization laboratory after a cursory review of the angiogram. It is frightening to think how many patients undergo unnecessary revascularization procedures or whose symptoms are dismissed as noncardiac because their physician “guessed” the significance of an ambiguous lesion. It is easy to perform FFR at the time of the procedure, and, unlike conventional SPECT MPI, it is not influenced by disease in other vessels. With FFR we can be confident that a lesion requires revascularization and know that deferral of revascularization based on a nonischemic FFR is safe. Until MPI more reliably identifies all physiologically significant stenoses in patients with multivessel CAD, FFR remains the gold standard for this important evaluation.
↵⁎ Editorials published in JACC: Cardiovascular Interventions reflect the views of the authors and do not necessarily represent the views of JACC: Cardiovascular Interventions or the American College of Cardiology.
- American College of Cardiology Foundation
- Melikan N.,
- De Bondt P.,
- Tonino P.,
- et al.
- Pijls N.H.J.,
- van Son J.A.M.,
- Kirkeeide R.L.,
- De Bruyne B.,
- Gould K.L.
- Pijls N.H.J.,
- van Schaardenburgh P.,
- Manoharan G.,
- et al.
- Djaberi R.,
- Roodt J.,
- Schuijf J.D.,
- et al.
- Hajjiri M.M.,
- Leavitt M.B.,
- Zheng H.,
- et al.
- Herzog B.A.,
- Husmann L.,
- Valenta I.,
- et al.