Clinical Variables Added to Syntax Score Increases Predictive Ability

Download this article's Factoid in PDF (& PPT for Gold Subscribers)

Adding clinical variables to the standard Syntax score boosts the ability to stratify risk for long-term adverse clinical outcomes, particularly 5-year mortality, in patients receiving drug-eluting stents. The findings are outlined in a study published online September 27, 2011, ahead of print in the European Heart Journal.

Researchers led by Patrick W. Serruys, MD, of Erasmus Medical Center (Rotterdam, the Netherlands), retrospectively calculated Syntax scores in 848 patients (1,792 lesions) enrolled in the all-comers’ SIRTAX (SIRolimus-eluting stent compared with pacliTAXel-Eluting Stent for coronary revascularization) trial. The investigators then calculated the clinical Syntax score for each patient using age, baseline LVEF, and creatinine clearance.

The median Syntax score was 10 (range, 1-42), while the median Clinical Syntax score was 11.6 (range, 0.7-272.2), with strong correlation between both measures (P < 0.001). Tertiles for the scores were defined as:

• Syntax score: low ≤ 7, mid ≤ 14, and high >14
• Clinical Syntax score: low ≤ 8.0, mid ≤ 17.0, and high 17.0

Using the standard score, MACE rates were significantly higher in the highest tertile compared with the lowest tertile at 5-year follow-up (24.2% vs. 12.5%; P < 0.01). The same also was true at 5 years for the endpoints of death, cardiac death, MI, and TLR. In addition, stratifying outcomes across Clinical Syntax tertiles led to similar results for the comparisons between high and low score tertiles (table 1).

Table 1. Clinical Outcomes at 5 Years Stratified by Clinical Syntax Score

There also were signs that Clinical Syntax score may help determine which stent is best for which patient. Among patients in the highest tertile, there was an increase in MACE with PES compared with SES at 1-year follow-up, which was mainly driven by higher TLR in the PES arm. Higher MACE risk in the PES vs. SES arm persisted at 5 years (34.7% vs. 21.3%; P = 0.008). However, the disparity in TLR was no longer significant (22.0% vs. 14.0%; P = 0.07).

Overall, the standard and the clinical scores were similarly equipped to predict 5-year MACE, while the clinical score also was an independent predictor for 5-year mortality.

In receiver-operator characteristic analysis, the area under the curve (AUC) for Clinical Syntax score was significantly larger compared with that for the conventional Syntax score for cardiac death (0.72 vs. 0.63; P = 0.002) and all-cause mortality (0.66 vs. 0.58; P < 0.001). The AUC for MACE was decreased for both scores, but not significantly larger for the clinical score (0.62 vs. 0.61; P = 0.24).

According to the study authors, in terms of calibration, Clinical Syntax score was more robust compared with normal Syntax score for all-cause mortality and slightly less robust for cardiac death. Calibration for MACE was worse for Syntax score and Clinical Syntax compared with that for mortality. They also point out that the clinical score showed diminished ability to discriminate between patients at low and intermediate risk.

Still Teasing Out The Best Combination

In an e-mail communication with TCTMD, Davide Capodanno, MD, of Ferrarotto Hospital (Catania, Italy), said the study is another piece of the puzzle in the move toward hybrid risk scores, which combine angiographic and clinical variables in order to stratify risk across the broad spectrum of CAD.

Dr. Capodanno said the use of a relatively simpler population compared with those seen in previous validation studies and the availability of long-term follow-up strengthen the current results. But he added that he was not particularly impressed by the incremental value of the new clinical score over the Syntax score.

“The [Clinical Syntax score] did not improve the discrimination for MACE and, consistent with previous series, suffered from poor calibration,” he said. “Also, the risk estimates for many endpoints were not in the anticipated order suggested by the tertiles. I think that one interesting take-away message, however, is that it's probably time to recognize that this kind of combined risk model works better when used to predict hard clinical outcomes, such as cardiac or all-cause mortality, rather than soft endpoints, such as MACE or revascularization.”

Dr. Capodanno also pointed to his own research on the Global Risk Classification, which was found to be associated with similar discrimination but better calibration than the Clinical Syntax score in patients with a higher angiographic complexity than that seen in the current study (Capodanno D, et al. J Am Coll Cardiol Intv. 2011;4:287-297). The Global score includes more clinical variables than the Clinical Syntax score, which he saw as a positive aspect despite the slightly longer time required for calculation. Although Dr. Capodanno believes the Global Risk Classification may be the best candidate to improve the performance of the Syntax score, he added that it requires validation in a prospective study.

Gauging Clinical Applicability

“The bottom line is we want to be able to predict who is going to do well from the standpoint of all endpoints,” said William F. Fearon, MD, of the Stanford Medical Center (Stanford, CA), in a telephone interview with TCTMD. “It’s not too surprising that adding clinical variables like age, LVEF, and creatinine would help better discriminate. The more accurate and specific [a score] is, the better.”

Dr. Fearon added that studies using the Syntax score in different ways and combinations will help gauge its importance as a tool and reveal how clinically applicable it may be in the long run. In a recent publication (Nam C-W, et al. J Am Coll Cardiol. 2011;58:1211-1218), Dr. Fearon’s group incorporated functional information from fractional flow reserve assessment into the standard Syntax score.

“These comparisons are all nice to have and they are showing us a variety of ways the Syntax score can be used. I think these studies will ultimately help us decide if we should take the time to do this in our practice,” he added.

 

---

Source:
Girasis C, Garg S, Räber L, et al. SYNTAX score and clinical SYNTAX score as predictors of very long-term clinical outcomes in patients undergoing percutaneous coronary interventions: A substudy of SIRolimus-eluting stent compared with pacliTAXel-eluting stent for coronary revascularization (SIRTAX) trial. Eur Heart J. 2011;Epub ahead of print.

 

 

Related Stories:

1. Syntax Score Incorporating Clinical Variables Improves PCI Risk Prediction
2. Syntax Score Validated in All-Comers LEADERS Trial