الفهرس 
IntroductionOnly 14 pages are availabe for public view
 |  | from | 151 |  |  |
| | | from | 151 | | |
Abstract
The Impact of Residual SYNTAX Score (rSS) and SYNTAX Revascularization Index (SRI) on In-Hospital Outcomes
Demographic DataWithin the demographic data, we only found a significant difference in the distribution of age within the three rSS groups (P = 0.002). We also found a similar significant difference between the SRI tertiles (P = 0.007). As we notice here, the mean age had the lowest value within the first rSS tertile and had the highest value within the third rSS tertile, while the relationship is inverse with the SRI tertiles, where the first tertile included the highest values, and the third tertile included the lowest values. This was in concordance with the study by Farooq and colleagues in 2013, which stated that the increase in rSS was associated with older age (P = 0.015). (10) Another study by Généreux and colleagues in 2015 also associated lower SRI values with older age (P < 0.001). (11) Other factors such as biological sex, BMI, dyslipidemia, smoking, hypertension, and diabetes showed no statistically significant difference in distribution within the three groups in both scores. This suggests that the rSS tends to increase with age since atherosclerosis is a progressive disease that is known to progress with age, among other factors. Factors that implicate aging as a factor in atherosclerosis include cellular senescence, in addition to the occurrence of several related biochemical reactions that are involved in the atherosclerotic process, including the release of reactive oxygen species (ROS), also known as the free radicals, and their role in the endothelial dysfunction caused by the oxidative stress on endothelial cells, in addition to the cumulative effect of other factors, such as smoking, diabetes, and inappropriate dietary habits. Genetic and racial factors are also involved in vascular aging, with a more accelerated aging process within African-Caribbean populations compared to European and Asian populations. (70)
In-Hospital Clinical FindingsWithin the three SRI groups, we found a significant difference between the distribution of systolic blood pressure on admission, with the highest values within the second SRI tertile (P = 0.038). Within the three rSS groups, we found a significant difference in the type of myocardial infarction (P = 0.034), with a positive relationship between the incidence of STEMI and rSS tertiles, in contrast to an inverse relationship with the incidence of NSTEMI. No previous studies were found to discuss this point. Therefore, it is worth noting that further studies are recommended to investigate this issue in detail.
In addition, we found a significant difference in the distribution of Killip classification of patients within the rSS tertiles (P = 0.024) and the SRI tertiles (P = 0.017). The Killip classification is a system used in individuals with an acute MI, considering physical examination and the development of heart failure to predict and stratify their risk of mortality. Individuals with a low Killip class are less likely to die within the first 30 days after their myocardial infarction than individuals with a high Killip class. (71) It is a powerful independent predictor of all-cause mortality in patients with non-ST-elevation acute coronary syndromes as well. (72) Although there was no significant difference in the distribution of class I, the highest number and percentage of class IV patients was found within the third tertile, indicating that patients with higher rSS values may present in a worse condition on admission. The relationship is inverse between the SRI tertiles, with more cases with Killip class IV within the first tertile than the second and the third tertile. Akgun and colleagues reported a similar finding but within the context of the bSS. (73) This could be attributed to the fact that most patients with higher rSS are patients with higher bSS and lower SRI, with a strong positive correlation with bSS, and a significant negative correlation with SRI.
Other factors, such as ST-segment resolution, left ventricular ejection fraction (LVEF), and myocardial wall motion score index (WMSI), showed no statistically significant difference in distribution within the three tertiles, as shown in table 19 and table 26 for rSS and SRI tertiles. Other studies reported similar results regarding the LVEF, such as the study by Hayase and colleagues. (74)
In-Hospital Laboratory DataThere was no significant difference in the distribution of ischemic biomarkers, including admission and peak CK values, admission and peak CK-MB values, and serum troponin values. However, the mean and median values were higher within the third tertile. Upon inspecting boxplots of the previous items (not displayed hereby), the increased mean and median values are apparently affected by significant outliers, but the general distribution of the data showed insignificant differences in the CK and CK-MB values. This is not the case in the distribution of troponin, which shows higher overall values within the third rSS tertile and the first SRI tertile. While this may not be statistically significant in our study, possibly because of the relatively small sample size and significant outliers, however, a potential relationship might exist between the serum levels of troponin and rSS, which was previously confirmed by another study with a larger sample size (n = 538) by Altekin and colleagues in 2020, which reported significantly higher serum troponin T values in patients with rSS higher than 7. (75)
Indicators of renal function showed variable outcomes. While the distribution of serum urea and serum creatinine levels showed no statistically significant difference, there was a statistically significant difference in the distribution of creatinine clearance (CrCl) values (P = 0.018), with lower values in the third tertile. This goes in concordance with the previously mentioned study by Altekin and colleagues, which found a significant difference in the distribution of CrCl values (P < 0.001), with lower values in patients with rSS > 7.(75)
Coronary Angiographic DataAccording to our results, coronary dominance does not differ significantly between the three tertiles. However, the numbers of initial and residual lesions differ significantly between the three tertiles (P < 0.001 for both variables). This is normal and expected to exist because the lesion count is an essential factor in the estimation of the bSS and the rSS. In addition, there is a significant correlation between the bSS and the rSS (Pearson’s R = 0.782, P < 0.001, Spearman’s rho = 0.617, P < 0.001). This relationship was demonstrated before in the study by Yan and colleagues in 2019, with a Spearman’s rho of 0.695 and a P-value of < 0.001. (76)
The distribution of culprit vessels differed significantly between the tertiles (P = 0.009). Within the first and second tertile, more culprit lesions were found in the LAD (62.7% and 50.9%, respectively), unlike the third tertile, where most culprit lesions occured in the RCA (55.8%). These findings were similar to those in the study of Altekin and colleagues, which showed that RCA showed the highest frequency of being the culprit vessel in 39.7% of cases with rSS > 7. (75) Meanwhile, the baseline TIMI flow in culprit vessels did not significantly differ in distribution between the three tertiles, but the residual TIMI flow showed a significant difference (P = 0.047), with about 10.3% of the cases with TIMI flow score less than 2. Similar findings in residual TIMI flow were reported in the previous study by Altekin and colleagues in 2020. (75)
Left Main Coronary Artery (LMCA) LesionsThe impact of left main coronary artery (LMCA) lesions has been extensively studied. The LMCA supplies about 70 to 100% of the myocardium, depending on its dominance. (77) The LMCA lesions occur in 3% to 9% of patients undergoing coronary angiography (78), which is similar to the incidence within our study population (6%). Due to its extremely high risk, it was given particular attention in its management, and it was given the highest weighing factor in the SYNTAX score algorithm. (8) It was previously treated exclusively by CABG surgery until the advent of drug-eluting stents and improvements in coronary stenting techniques, which enabled the treatment of LMCA lesions by PCI. (79) The SYNTAX trial found no difference in five-year survival of patients with LMCA lesions treated by PCI or surgery, while patients with more complex lesions should benefit better from surgery rather than PCI. (80) However, a ten-year follow-up of patients with treated LMCA lesions showed a clear prognostic advantage of CABG over PCI in patients with higher complexity lesions as measured by the bSS. In addition, the bSS had a relevant discriminative capacity on long-term outcomes in patients treated with PCI but not in CABG-treated patients. (81) In our study, more cases with LMCA lesions were found in the third tertile than in the other groups (18.6% within the third tertile), with a highly significant difference. (P < 0.001). Moreover, it was a significant independent predictor of in-hospital composite MACE in univariate and multivariate logistic regression (Odds Ratio = 4.553, P < 0.001). This finding is similar to the study of Altekin and colleagues which stated that all of the LMCA lesions were within the higher rSS values. (75)
Proximal LesionsBoth the rSS tertiles and the SRI tertiles show similar results in terms of significance. No statistically significant difference in the distributions were found with the proximal LAD lesions in both the rSS and the SRI tertiles, while having the highest percentage within the third rSS tertile (43.1% vs. 38.2% vs. 60.5%, P = 0.077), but the numbers showed insignificant differences between the SRI tertiles (43.8% vs. 44.2% vs. 51%, P = 0.716). A study by Gao and colleagues in 2018 stated that the presence of residual proximal LAD lesions is associated with a higher incidence of death and MACE in addition to the higher risk associated with higher rSS values. (82) In addition, the proximal LAD lesions are heavily weighted within the SYNTAX score algorithm, second only to the LMCA lesions. (8) However, a study by Leborgne and colleagues in 2002 found that stenting of lesions in proximal LAD is as effective and safe as treatment of distal LAD lesions, and therefore the location of the LAD lesion in itself is not a predictor of worse outcomes when stenting is possible, and should not be considered upon selection of the revascularization strategy. (83) As a consequence, the proximal LAD lesions are recommended to be treated as soon as possible to avoid potential subsequent adverse events.
On the other hand, the distribution of proximal LCx lesions differ significantly in both the rSS and SRI tertiles (21.6% vs. 34.5% vs. 62.8%, P < 0.001 for rSS tertiles, 50% vs. 40.4% vs. 24.5%, P = 0.033 for SRI tertiles). Cases with LCx stenosis may present with non-significant ECG changes, and that may result in delayed intervention and thus worse prognosis, especially when compared with RCA lesions. (84) Therefore, it could be possible that a higher number of proximal LCx lesions in the higher rSS and lower SRI tertiles may carry a higher risk of adverse events to the patients.
In addition, the distribution of proximal RCA lesions showed a statistically significant difference between both the rSS and the SRI tertiles. (25.5% vs. 34.5% vs. 53.5%, P = 0.018 for rSS tertiles, 52.1% vs. 34.6% vs. 24.5%, P = 0.017 for SRI tertiles).
In-Hospital Adverse Events and Composite MACEThere was no significant difference in the distribution of any of the in-hospital events between the SRI tertiles. Therefore, it is unlikely that the SRI could have a significant impact on in-hospital events. This was later confirmed by univariate logistic regression, which yielded insignificant outcomes, and denoted that it was unlikely that the SRI should be considered as a predictor of in-hospital MACE and all-cause mortality. This was confirmed further by ROC curve analysis, which yielded statistically insignificant outcomes. It is worth mentioning here that the bSS had the best discrimination power in our study to predict in-hospital MACE in univariate logistic regression (Odds Ratio = 1.048 (95% CI = 1.01 – 1.088), P = 0.014, C-index = 0.65), and it had the best classification power by ROC curve analysis (AUC = 0.65, P = 0.018). In contrast, logistic regression and ROC curve analysis for both rSS and SRI yielded statistically insignificant results. The univariate logistic regression model for rSS showed statistically significant outcomes but failed to provide adequate discrimination (C-index = 0.587, P-value for C-index = 0.17). In addition,based on multivariate logistic regression results, it was found that hypertension, LV Ejection Fraction, ST segment resolution, and the presence of significant LMCA lesions were the independent early predictors for In-Hospital Composite MACE (Table 30). This is discordant with other studies in this context, including the study by Khan and colleagues (63) and the study by Altekin and colleagues (75). Both studies stated that the rSS was a good predictor for in-hospital death and MACE, with a good discrimination power based on ROC curve analysis. The main difference points here are that the two previous studies included all patients with STEMI who underwent primary PCI but were not specific to the multi-vessel disease group. In addition, the sample sizes included in both studies were larger than ours because our study did not include patients with single-vessel disease. On the other hand, our results were concordant with a similar study by Loutfi and colleagues that stated similar in-hospital outcomes. (85) The study by Loutfi and colleagues worked with similar sample numbers and worked on the same population of patients with coronary MVD. Therefore, the similarity in results may indicate that the MVD patient population might have different characteristics that are worth investigating, especially when it comes to in-hospital outcomes and their relationship with risk scores. Further studies within the population of patients with MVD with larger sample sizes are recommended to evaluate this issue adequately.
Overall, none of the angiographic scores showed better predictive performance than the in-hospital GRACE score, which showed a statistically significant outcome (OR = 1.058 (95% CI = 1.032 – 1.085), P < 0.001, C-index = 0.883, P-value for C-index < 0.001). Based on these findings, it was apparent that the combination of clinical variables may add more power to the prediction model. A study by Aktürk and colleagues in 2018 found that the SSII, which included clinical factors as well as angiographic factors, was an independent predictor of in-hospital MACE, mortality, non-fatal MI and stent thrombosis. (86) The study by Altekin and colleagues found a higher incidence of death and MACE in patients with rSS > 7. (75) The difference here is that he divided his groups in a different manner than our study, and his target population is more inclusive than ours. They had no preference in their patient selection criteria, unlike our study that focused mainly on patients with MVD. Further studies are recommended on larger scales to investigate the impact of angiographic and derived clinical scores on in-hospital outcomes on larger samples. Follow-Up OutcomesThe difference in the distribution of follow-up MACEs is more statistically significant between the rSS tertiles rather than the SRI tertiles. However, numbers are still larger in the first SRI tertile with more CAD residual burden than other tertiles. Patients with adverse events had higher mean and median rSS values and lower mean and median SRI values. In addition, survival analysis using Kaplan Meier curves at six-month, one-year and two-year follow-up for both rSS and SRI and log rank tests showed that the impact was more pronounced between the rSS tertiles, with a statistically significant difference between the first two tertiles and the third tertile in the incidence of MACE and all-cause mortality. On the other hand, the SRI tertiles did not exhibit a significant difference between the three tertiles, but with an overall statistically significant impact on follow-up MACE and all-cause mortality. This indicated that the rSS had more pronounced impact on the incidence of adverse events than the SRI. Moreover, the comparison of test results showed that both the rSS and the SRI have the highest impact on the incidence of death and MACE at one-year follow-up, but that impact declines afterward, probably due to the influence of other confounding factors such as aging, smoking, the decline in LVEF, and natural disease progression over time. This was concordant with several studies that affirmed the predictive ability of rSS and SRI to predict short- and long-term morbidity and mortality. (69,75,87–89) It should be noted that the SRI in itself is a percentage of the treated burden of CAD, but it does not give a solid idea on the extent of the existing disease itself (89). For instance, treating a patient with two lesions with a low total bSS will result in a small or even nonexistent rSS, but a very high SRI, which may sound tricky if used as a measure to evaluate the patient’s overall condition. On the other hand, treating a critical lesion in a high bSS patient may not result in a major SRI reduction, but the SS reduction could be quite significant with improved overall outcomes. In other words, the rSS is a quantifying tool for the residual disease post-intervention that may be used effectively for risk stratification and treatment strategy planning, while the SRI is a quantifying tool for the treated burden of the disease but should not be used as an indicator of the severity of present disease by itself, rather than being a prognostic indicator for what was treated already. Comparison of Scores’ Performance
Follow-up MACEFew studies were present that compared the impact of various scores on patient outcomes. (90,91) The difference in post-discharge GRACE score between rSS and SRI groups was not statistically significant. Further analysis by Cox regression for all scores showed that both the rSS and the SRI are independent predictors for MACE and death at all time points of follow-up, with increasing impact over time. While at earlier time points (six-months and one-year points) the rSS and SRI have good AUC values, the bSS showed better performance in this regard. At two-year follow-up, the rSS had the best performance among the angiographic scores. Multivariate Cox regression models at different time points found that LVEF was the only common significant predictor of MACE throughout the entire follow-up period of two years. Therefore, clinical scores that include the LVEF as a factor are likely to predict the outcome in a much better way. There were very limited data available about comparing clinical and angiographic scores. Therefore, we recommend further studies with a large, multicenter design to evaluate this issue with more adequacy.
Follow-up MortalityThe results with all-cause mortality were like those of MACE, with some difference. The predictive performance indices of rSS and SRI were significant at all time points of follow-up. Upon comparison of the performance of angiographic scores, it was found that the bSS performed better than rSS and SRI in terms of predicting mortality, with better overall performance at one-year follow-up. Apart from the bSS, the rSS had the best performance over the rest of the scores at six-month and two-year follow-up. However, it was superseded by the bSS at one-year follow-up. This goes in concordance with the study by Martins and colleagues in 2021 which stated that the performance of rSS performed better than bSS and SSII in prediction of adverse events on the long term. (91) Still, comparative data in this regard within the ACS patients with MVD were limited and deemed insufficient. Therefore, further studies in this topic are recommended for better evaluation.