Accurate characterisation of coronary artery disease (CAD) anatomy based on the diagnostic angiogram is essential to select the optimal strategy of revascularisation. Recently, the SYNTAX score generated a great amount of interest because of its ability to risk-stratify and discriminate outcomes of patients with complex CAD undergoing percutaneous coronary intervention (PCI) as compared to coronary artery bypass graft surgery.1–3 Additionally, it has been validated in different clinical settings of patients undergoing PCI, as well as in various subsets of lesions.4–11 Thus, SYNTAX score is a pioneer anatomical-based risk score that aids in the decision-making process. However, assessment of the SYNTAX score relies on pure visual interpretation of lesion severity and other semi-quantitative and subjective variables, which for even simple measures may be inaccurate. Although some studies report acceptable reproducibility when determined by angiographic core laboratory technicians,2 concerns remain regarding its reproducibility, especially among ‘SYNTAX-score naive’ or inappropriately trained cardiologists.12 The recent introduction of a ‘functional SYNTAX score’ that incorporates ischaemia-producing lesions as determined by fractional flow reserve (FFR), has brought some hope by better risk-stratifying of patients.13 Moreover, a limitation of the SYNTAX score relies on the fact that the score algorithm does not entail any clinical variable. Co-morbidities are known to impact early outcomes of patients undergoing revascularisation. Accordingly, to address the relative lack of discrimination and predictability of the pure SYNTAX score, some attempts to combine clinical variables into the SYNTAX score were reported recently. The ‘clinical SYNTAX score’, a combination of ACEF and SYNTAX score strata, parsimoniously combines three strong clinical predictors of clinical outcome (i.e., age, creatinine and ejection fraction). An alternative approach combines the SYNTAX score and the EuroSCORE, the so-called ‘Global risk classification’.14–16 Overall, their findings underline the potential importance of the interplay between clinical and angiographic data in predicting clinical outcomes after PCI.
Although the precise training requirements to optimise the performance of cardiologists using the SYNTAX score are unknown, some recent published data suggest that training beyond the standard on-line tutorial (www.syntaxscore.com) is warranted if the full clinical potential of the SYNTAX score is to be realised.12 The aim of this short article is to overview the basics and essentials of the SYNTAX score assessment, to ensure appropriate reading and, therefore, improve its reproducibility between readers.
SYNTAX Score – The Basics
It is of paramount importance to complete the entire recommended online tutorial (www.syntaxscore.com). This website takes the reader through all the variable definitions, followed by 13 schematic case examples. To conclude, a self-evaluation, including seven real cases with online angiograms, must to be performed. This mandatory training is vital and is the first step to adequate scoring. Importantly, when scoring an angiogram, the reader should focus strictly on the angiographic evaluation and not have therapeutic strategies in mind. Furthermore, the reader should be systematic, with an established ‘routine’. We strongly recommend starting from the right coronary artery (RCA) to qualify the dominance (right or left, no co-dominance exists with the scoring system of SYNTAX score). After a global review of the angiogram, the reader should restart, also from the RCA, and visually score each lesion. It is important that only lesions ≥50 % in a vessel ≥1.5 mm of diameter should be scored. If serial lesions occur, they should be considered as a single lesion if they are less than three vessel reference diameters apart (see Figure 1, lesion 2). Conversely, stenoses more than three vessel reference diameters apart are considered as separate lesions (see Figure 1, lesions 1, 2 and 3).
Total Occlusions and Collaterals
When a total occlusion (TO) is identified, the first segment involved in the TO, the age of the TO (most of the time unknown), the type of TO (blunt or non-blunt) and the presence or absence of bridging collaterals should be reported (see Figure 2, left panel). Afterwards, the reader should seek out collaterals; if present, the first SYNTAX segment beyond the TO that is visualised by antegrade or retrograde contrast should be identified (Figure 2, right panel).
Bifurcations and Trifurcations
Bifurcations and trifurcations are the SYNTAX score characteristic most associated with the highest degree of variability between readers.12 When scoring a trifurcation, the four segments involved should be assessed (the proximal segment and the three branches) and scored accordingly. Figure 3 illustrates a case in which four segments were diseased and have been scored. Importantly, only segments 3/4/16/16a, 5/6/11/12, 11/12a/12b/13, 6/7/9/9a and 7/8/10/10a should be considered for trifurcation scoring, and only segments 5/6/11, 6/7/9, 7/8/10, 11/13/12a, 13/14/14a, 3/4/16 and 13/14/15 should be considered for bifurcation scoring (see Figure 4).
Bifurcation scoring systems are based on the medina classification. After identifying the proximal segment, the reader must determine which distal branches have the smallest diameter and will, therefore, represent the side branch. The same rule applies for the distal left main lesion, where the left circumflex (LCx) and the left anterior descending (LAD) could be either a side branch or distal branch, according to their respective vessel diameter and their importance. Only vessels of diameter ≥1.5 mm should be included in bifurcation scoring. While no clear distance has been described by the original SYNTAX score algorithm system, the lesion must be in close contact with either the side branch (if the lesion is in the main vessel) or the main vessel (if the lesion is in the side branch) to be considered as a bifurcation.
The presence of an ostial lesion (only for the ostial right, left main or LCx or LAD if dual ostia is present), tortuosity (≥1 bends of ≥90°, or ≥3 or more bends of 45° to 90° proximal to the diseased segment) (see Figure 5), lesions >20 mm, heavy calcifications (visible in more than one projection that surrounds the complete lumen of the coronary artery at the site of the lesion) and thrombus should be assessed. When all the lesions are assessed, the reader must ascertain either the presence or absence of diffuse and narrowed disease. This characteristic is also a source of high variability.12 At this stage, the reader must be highly systematic, with a meticulous evaluation of all the vessels previously scored. Diffuse and narrowed disease should be scored when at least 75 % of the length of any segment distal to the lesion has a diameter of <2 mm because of atherosclerosis. Importantly, the vessel must be primarily a target for revascularisation (see Figure 2, left panel).
How do We Improve Accuracy and Reproducibility?
As with any other technique or test, ‘practice makes perfect’. Although the SYNTAX score tutorial exposes the ‘trainees’ to 20 case examples, recent data suggest that a learning curve can be overcome with at least 30 additional cases.12 Scoring by two or more colleagues, especially in the ‘initial experience’, may help to improve accuracy and, by consensus, will help to resolve conflict. Additionally, while scoring an angiogram, a reader should not hesitate to go back and consult the online definitions in cases of any doubt. The SYNTAX score has its own definitions, which need to be respected to ensure accuracy. Finally, the process of risk stratification is an integral part of the SYNTAX-pioneered heart team approach in selecting the most appropriate revascularisation modality in patients with complex CAD.
In conclusion, thus far the SYNTAX score is the most powerful angiographic tool in the prediction of events after PCI. Reproducibility is still an unsettled issue, and may be improved by adequate training, including the mandatory online tutorial. Additional training, supervised by knowledgeable mentors such as experienced core lab technicians or highly trained cardiologists, is highly desirable and will ensure the adequate use of the SYNTAX score. Adjunctive methodology, such as combining clinical data or the use of FFR, will enhance its power of discrimination and result in better patient stratification.