2.3 Statistical Analysis
The statistical analysis was performed using the SPSS version 21 program.  Variables were presented as number and percentage or as mean ± standard deviation for discrete and continuous variables respectively. A two-sample t-test test was used for continuous data, and a χ2 test for categorical data. A two-sided p value less than o.o5 was considered significant.
This study was approved by the Institutional Committee on Human Research of our institution.
RESULTS
The HT and the No HT groups included 93 patients each. The mean age was 66 ±12 years and the majority of patients in both groups were men (65%). There were no differences in the percentage of patients with comorbidities and in other baseline clinical characteristics of both groups except for a higher rate of previous PCI in the No HT group (47% vs. 26%, p<0.05). The great majority of patients in both groups presented with acute coronary syndromes (90%) with a high prevalence of NSTEMI or evolved STEMI patients (64%). No differences were observed in the complexity of the coronary artery disease according either to the number of vessels with significant stenosis or the SYNTAX Score. Significant three vessel disease or left main disease was seen in 75% of the HT group and 80% of the No HT group (p=0.49). Baseline characteristics are presented in Table 1.
The mean SYNTAX score was similar between groups (19±6 in the No HT and 21±6 in the HT group, p=0.589) although there was a trend for a greater percentage of patients with low SYNTAX scores in the No HT group (75.3% in the No HT and 59.6% in the HT group p=0.06). The mean perioperative risk according to the STS score was low and similar in both groups (3±2 vs. 3±3). Again, there was a trend for a higher percentage of patients with intermediate and high STS scores in the No HT group, compared to the HT group (22% vs 8% and 12% vs 8% respectively, p=0.052). (Table 2)
A significant difference in the modality of treatment chosen in each period was observed. In cases treated without HT discussion, PCI was the most frequent method of treatment, being employed in 69% of cases. During the HT period we observed a radical change in treatment recommendations, with most patients being referred to CABG (63%) (Figure 1). This corresponds to a 174% increase in the use of surgical revascularization.
On the other hand, there was a significant increase in the time from diagnostic angiography to revascularization in the HT group (8.5 ±4 days) in comparison with the No HT group (1.8 ± 5 days). (Table 3). When examining this variable according to revascularization strategy, time to CABG was similar in both the HT and No HT groups (7±8 vs 8±5, p=0.75) but the time to PCI was substantially longer in the HT group compared to the No HT group (0.39±2 vs 8±6 p<0.01).
DISCUSSION
This study shows that the implementation of Heart Team discussion in the decision-making process of patients with complex coronary artery disease was associated with a profound change in the strategy for revascularization characterized by an increased rate of CABG with a reduction in the rate of PCI. Moreover, HT discussion was associated with a significant delay in the time to PCI in comparison with patients treated in earlier periods before the establishment of the HT.
The Heart Team has been recommended as standard care for patients with complex CAD since the publication of the European and American guidelines for myocardial revascularization in 2010 [6, 10]. This recommendation, while intended to provide interdisciplinary sharing of knowledge and experience in decision-making, is based solely on expert opinion. As such, little is known about the impact of this approach on rates of revascularization methods and on short and long-term outcomes as well as immediate peri-procedural risks and potential benefits in each individualized setting.
The main finding of our study was a significant change in the revascularization modalities before and after HT implementation. In fact, there was an increase in CABG recommendations from 23% without HT to 63% in the HT group with a parallel decrease in the rate of PCI. There are a few possible explanations for this substantial change: first, the baseline characteristics of the patients was slightly different. Despite individual matching, a trend for a lower STS score was observed in the patients in the No HT group. This difference however should have favored a surgical approach within this group, and these were not the results that were noted.
In addition, the mean SYNTAX score and the proportion of patients with intermediate or high SYNTAX scores was numerically higher in patients discussed by the HT, although not statistically significant. This may reflect the fact that the patients in the HT group had more complex coronary anatomy hence being referred to surgery more often. In fact, the past two decades have seen a myriad of trials [2, 11-16] seeking to answer the question of which strategy provides better short and long term outcomes for complex CAD while balancing patient satisfaction and costs, mainly in the setting of stable disease. In summary, most trials on multivessel disease show a survival advantage of CABG over PCI according to anatomical complexity as assessed by the SYNTAX score, while there are no clear benefits to either treatment when it comes to isolated LM disease. Based on these findings, an intermediate or high SYNTAX score is recognized by recent revascularization guidelines as a class IA indication for CABG over PCI [4, 17]. Furthermore, the FREEDOM randomized trial [18] compared CABG to PCI solely in diabetic patients with multivessel disease and found reduced mortality associated with the surgical approach. The accumulation of data on the subject as well as updated guideline recommendations might have influenced the change in revascularization strategies seen in our study in the last few years. Supporting this observation is the fact that the operator in the HT period between 2016 and 2017 recommended CABG in 57% of the cases whereas only 23% of operators during the No HT period (2005-2015) referred the patients for surgery.
Other factors such as personal tendencies and beliefs of the cardiologists involved as well as financial or professional interests that could influence the revascularization decision are neutralized by the HT discussion leading to a change in treatment strategies. This point was emphasized in a study by Abdulrahman et al [19] that described how HT recommendations changed according to which professionals were present at the discussion.
Our findings are similar to recent reports in the literature. In 2016, Bonzel et al . [20] described that between 46-66% of patients discussed by the HT were referred to isolated CABG. They also demonstrated that the patients referred to PCI had a low rate of need for CABG during long term follow up with no increase in mortality after 2 years. Also, a possible intrinsic bias of the interventional cardiologist in recommending PCI as a primary revascularization approach was noticed in a study by Sanchez et al [21] that used the HT to review decisions on revascularization strategies based on SYNTAX and STS scores and on Appropriate Use Criteria for coronary revascularization. They found that 34.9% out of 301 patients who had undergone PCI in the past two years had an inappropriate or uncertain indication for angioplasty as recommended by the HT.
Another important finding of the present study is that the decision to bring the case to HT discussion was associated with a significant delay in the performance of revascularization. Indeed, we observed a significant increase in ”time to PCI” from an average of 0.36 days to 8 days, while the time to CABG did not differ substantially. This was most probably due to a limitation in institutional resources, but could also occur in other hospitals should the HT approach be systematically adopted. Taking into account that the population analyzed included 90% of patients with ACS, such a delay may be associated with potential risks for this population. According to the European Society of Cardiology/European Association for Cardio-Thoracic Surgery (ESC/EACTS) guidelines on myocardial revascularization [6], PCI or CABG should be performed within 6 weeks after angiography for patients presenting stable coronary disease and within 2 weeks for patients with a high-risk coronary anatomy. For patients presenting with ACS the recommended time frame for revascularization is shorter, based on adverse events that may occur while on the waiting list for revascularization. Guidelines for the management of patients with non-ST elevation acute coronary syndromes (NSTE-ACS) [22] recommend an invasive strategy in moderate to high-risk patients. The timing of the intervention in those patients is dependent on the patient’s baseline risk factors and extend up to 72 hours from initial presentation. The HT approach potentially promotes delays in time to revascularization which was evident in the population that was ultimately treated by PCI, in conflict with the aforementioned recommendations. Multiple trials investigated and confirmed the value of early revascularization in patients with NSTEMI (8), although the clinical significance and major events rate during the waiting period until revascularization after the HT have not been investigated. The present study was not designed to examine such clinical outcomes, which would be valuable data to be reported in follow up research.
Nevertheless, a 2018 single center [9] study of 1000 consecutive patients concluded that the HT approach was feasible, with decision making and treatment following within a short time after referral and largely in accordance with clinical guidelines. This study however, included mainly patients referred from a community setting and as such the mean time to revascularization of 6 weeks was considered safe and appropriate.
Our study has a number of limitations. First, this was an observational study with a retrospective attempt of matching patients on the basis of clinical and angiographic variables. Despite this attempt, numerical differences in the number of patients with previous PCI, the complexity of CAD and surgical risk were observed between groups. This potential selection bias could explain differences in the choice of type of revascularization between the two periods. Secondly, our sample size might have been small to detect differences in characteristics between the groups. Third, there was no long-term outcome follow up of the patients treated with and without HT discussion. Further research into the impact of HT discussions on clinical outcomes would be most certainly welcome.
Nonetheless, this is to the best of our knowledge the first study to analyze the change in revascularization policy associated with the introduction of systematic, institutional HT discussions for complex CAD.