Prognosis Effects of Interventional Therapy on Patients with Coronary Slow Flow: An Observational Cohort Study

Background: The phenomenon of slow coronary blood ow is common clinically and is related to the patient experiencing repeated chest distress. The mechanism of slow blood ow is still unclear, and the prognosis after combined interventional therapy is unknown. Therefore, we evaluated the role of interventional therapy in this part of the population through the results of a 2-year follow-up. Methods: A total of 4663 patients who underwent continuous coronary angiography (CAG) were identied. Those patients with primary slow coronary ow were included in the study (n=369). The population was then divided into the PCI group and the N-PCI group. CTFT is used to assess the severity of slow blood ow. The endpoint events were the occurrence of rehospitalization and out-of-hospital death within 2 years. The log-rank test, Kaplan-Meier method, and Cox regression were used to evaluate and analyze the nal results. Results: A total of 36 patients were readmitted to hospital, and 14 died suddenly outside the hospital during the follow-up period. Among these patients, 6 and 2 patients comprised the PCI group, while 30 and 12 patients comprised the N-PCI group. Comparison of the two groups showed no signicant superiority (15.1% vs 13%, P=0.73). In Cox regression analysis, high BMI (body mass index) was an independent predictor of adverse end events (P=0.024). Conclusions: Interventional therapy may not improve outcomes in patients with slow coronary blood ow. BMI plays an important role in the inuence of prognosis. Further research is needed to investigate this conclusion. Selective coronary angiography was performed by two experienced interventional cardiologists using a standard Judkins’ technique. When evaluating the results of coronary angiography, the left anterior descending (LAD) and the left circumex (LCX) should be projected in at least four angles; the right coronary artery (RCA), in at least two angles. The rst frame used to evaluate each TFC is when the dye enters the artery completely, and the last frame is when the dye enters the distal marker branch. In our study, the lming speed was 15 frames per second. The calculation of CTFC was performed by an observer who was blinded to the clinical details. The standard numbers of CTFC frames are dened as 36.2 ± 2.6 frames for LAD, 22.2 ± 4.1 frames for LCX, and 20.4 ± 3.0 frames or RCA. Researchers with TFC of any of the three vessels greater than twice the standard deviation of the normal published range were dened as patients with slow coronary blood ow.


Background
Coronary slow ow phenomenon (CSFP) is a common phenomenon in the process of coronary angiography, with an incidence of approximately 1-7% [1]. Patients usually present with typical angina symptoms and visit the hospital repeatedly. However, coronary angiography showed no signi cant intimal lesions but only a decrease in contrast agent lling.
The mechanism of CSFP is still unclear and mainly includes [2] the following aspects: microcirculatory dysfunction, endothelial dysfunction, subclinical atherosclerosis, in ammation, and anatomic factors. Pathophysiological inaccuracy leads to no targeted treatment. Previous studies have focused on the search for predictors [3][4][5][6][7] and potential therapeutic approaches [8][9][10], yielded signi cant results and received widespread recognition. Nevertheless, current guidelines do not provide clear treatment recommendations and evaluation procedures for this population, especially after percutaneous coronary intervention (PCI). It remains unclear whether interventional therapy can improve outcomes in patients with slow blood ow with coronary atherosclerosis.
Thus, we assessed the effect of interventional therapy on 2-year outcomes in patients with slow blood ow compared with conventional treatment.

Method Study Population
This is a retrospective cohort study conducted between January 2014 and September 2018. Out of 5663 consecutive patients who underwent coronary angiography in the Hangzhou First People's Hospital Heart Center, 409 patients with CSFP were assessed for eligibility. In total, 369 patients were enrolled in the analysis and divided into two groups according to the experimental design. A ow diagram of the progress through the phases is shown in Fig. 1.

Ethics and consent statements
Since this is a retrospective cohort study, ethical and informed consent is not required.
De nitions CSFP is de ned as primary CSFP [2], which essentially is a delay in contrast injection into the coronary arteries. It is important to identify primary CSFP from delayed progression of contrast agents in coronary reperfusion therapy (such as angioplasty or stent implantation for acute myocardial infarction) or from other secondary causes of coronary slow ow. Conditions associated with secondary coronary slow ow include the following categories: coronary ectasia, coronary spasm, embolism, heart failure, angioplasty and stenting of acute myocardial infarction, valvular heart disease and connective tissue disorders.
Corrected TIMI frame count (CTFC) [11] represents the number of frames required to reach the distal boundary marker of the coronary artery. Not only is CTFC a simple, repeatable, objective, quantitative coronary index, but it also standardizes TIMI grading and facilitates angiography time comparisons between ow tests.

Criteria for inclusion and exclusion
Inclusion criteria Patients with primary CSFP Patients who underwent interventional therapy showed CSFP on coronary angiography after 1 month

Exclusion criteria
Previous myocardial infarction or interventional therapy Heart pump failure Severe liver or kidney failure Life expectancy of less than 5 years Preoperative examination Venous blood samples are extracted within 24 hours of admission. White blood cells, red blood cells, platelets, total cholesterol, low-density lipoprotein, high-density lipoprotein, C-reactive protein, creatinine, uric acid, admission blood glucose, glycosylated hemoglobin, electrocardiogram and echocardiography must be completed and reviewed prior to coronary angiography.
Coronary angiography and evaluation of coronary ow velocity Selective coronary angiography was performed by two experienced interventional cardiologists using a standard Judkins' technique. When evaluating the results of coronary angiography, the left anterior descending (LAD) and the left circum ex (LCX) should be projected in at least four angles; the right coronary artery (RCA), in at least two angles. The rst frame used to evaluate each TFC is when the dye enters the artery completely, and the last frame is when the dye enters the distal marker branch. In our study, the lming speed was 15 frames per second. The calculation of CTFC was performed by an observer who was blinded to the clinical details. The standard numbers of CTFC frames are de ned as 36.2 ± 2.6 frames for LAD, 22.2 ± 4.1 frames for LCX, and 20.4 ± 3.0 frames or RCA. Researchers with TFC of any of the three vessels greater than twice the standard deviation of the normal published range were de ned as patients with slow coronary blood ow.

Endpoint
The mean follow-up time for all patients was 24 months. Data related to clinical outcomes were obtained mainly through telephone follow-up and partly by reviewing outpatient records. The endpoints of the study included cardiovascular events requiring hospitalization or death. The former includes the occurrence of ACS, heart failure requiring hospitalization, and angina requiring revascularization [12]. All deaths are considered cardiogenic unless an unequivocal noncardiac cause was established.

Statistics
Before statistical analysis, all continuous variables were plotted and tested for normality and homogeneity of variance. Depending on the results, continuous variables are expressed as the means ± standard deviations (or medians and interquartile ranges). Categorical variables are expressed in terms of frequencies and percentages. Differences were compared between the two groups. Continuous data were compared by the T test or the Mann-Whitney U test (according to data characteristics), and categorical variables were tested by the chi-square test and Fisher's exact test.
To compare the relationships between prognosis and factors, log-rank tests and Cox regression methods were performed. Log-rank tests are accurate based on hypergeometric probability. Time-to-event curves (i.e., age, treatment, diabetes, smoking and BMI) were evaluated by the Kaplan-Meier method. To identify independent risk factors for prognosis, a single-factor regression analysis was conducted on independent variables one by one, and corresponding p-values were recorded. Independent variables with p values less than 0.1 were incorporated into the nal regression equation to identify clinical predictors.
Receiver operating characteristic (ROC) curves were constructed for the prediction of endpoint events. Different prediction parameters were analyzed by the logistic regression model, and the prediction probability was calculated for ROC analysis. The area under the curve (AUC) and the sensitivity and speci city of predicting RP were calculated. AUC = 1.00 indicates the highest accuracy, while AUC = 0.50 indicates no accuracy. A p-value of < 0.05 (2-sided) was considered statistically signi cant. All statistical analyses were carried out using SPSS software (version 23.0).

Baseline Characteristics
During an average of 24 months of follow-up, a total of 369 people were included in the analysis, with 53 in the PCI group and 316 in the N-PCI group. The average age of the patients was 67 (56-78). The  Table 1.

Coronary angiography and medication
In this study, approximately half of the patients had slower blood ow in more than one vessel (50.4%), and most patients had only mild arteriosclerosis or even smooth intima (71.8%). Compared with the N-PCI group, the PCI group had higher use of antiplatelet agents, statins, and B-blockers, but there was no signi cant difference in the use of nicordil, and even lower use of nondihydropyridine calcium antagonists than the N-PCI group (1(1.9%) vs 41 (13.0%), p = 0.019) ( Table 2).

Endpoint
The overall prognosis for patients with slow blood ow was good, with the vast majority remaining in the outpatient follow-up; only 36 patients were rehospitalized. However, 14 patients died suddenly outside the hospital ( Table 3). The Kaplan-Meier curves of various in uencing factors for patients are shown in Fig. 2. Compared with the N-PCI group, the PCI group did not show an advantage on the prognosis, and there was no statistically signi cant difference (p = 0.73). Among the factors associated with the occurrence of slow blood ow, diabetes and smoking were not associated with the occurrence of events. Only age showed a signi cant difference, while BMI showed a trend of signi cant difference.

ROC analysis and prediction of endpoint
We constructed the ROC curve to explore the possibility of predicting the occurrence of terminal events through the area under the curve. No strong predictors were found (AUC of 0.661, AGE + BMI, P = 0.042) (AUC of 0.683, TC + BMI, P = 0.039), even though we used joint diagnoses to increase the e cacy of the diagnosis (Fig. 3). Through a single-factor regression analysis, we extracted the following potential variables: age, BMI, Lvids, FS, EF, Hb, Plt, LDL, HDL, CRP, TC, and Scr. Combined with clinical factors, smoking and sex were added. BMI was found to be an independent risk factor for endpoint events.

Discussion
The main ndings of our study are as follows: 1. Interventional therapy did not show an advantage over drug therapy during the two-year observation period. 2. BMI was an independent risk factor for endpoint events in patients with slow blood ow. 3. Patients with slow coronary blood ow may be at potential risk of sudden death.

Prevalence of CSFP
Compared with previous epidemiological ndings [1], the CSFP incidence in our heart center was quite common, at approximately 8%. At the same time, the slow ow in our center usually involved more than one vessel. It is not unusual for different populations and different studies to have slight uctuations in the incidence of CSFP, and there are some signi cant differences in our study that may be due to the following two factors. First, the epidemiological data included the normal population, and the inclusion of coronary angiography in our heart center was more rigorous. Second, because of the differences in inclusion criteria of research design, our research subjects were often at higher risk. According to the pathophysiology of small vessels in CSFP, slow coronary blood ow may be an early manifestation of atherosclerosis [13,14], which is associated with more risk factors due to the presence of xed stenosis in our enrolled population. Our reasonable conjecture is that it is due to the mapping of the atherosclerotic conditions to the microcirculation, which is manifested as diffuse small vessel lesions, so that the angiography is more likely to show multiple branches of slow blood ow. Of course, these bold conjectures need to be con rmed by further basic research.

Outcome
Based on 2 years of follow-up, we did not nd a strong association between interventional therapy and a reduction in adverse events, suggesting that we should not be too aggressive when patients with slow blood ow are combined with critical lesions and should use intravascular ultrasound (IVUS) and fractional ow reserve (FFR) comprehensive judgment to maximize the bene t to patients. This strategy can also avoid the adverse reactions and poor compliance problems caused by patients taking many drugs. From a pathophysiological point of view [15,16], stents can solve the problem of large vessels, while the fundamental problem of microcirculation is not improved.
Previous studies [14,17,18] have shown that obesity is an important component in the development of slow blood ow. Obesity will lead to insulin resistance, produce a large number of adipocytokines, and then induce the production of reactive oxygen and lead to the breakdown of endogenous vasodilator nitric oxide, resulting in the imbalance between endothelial systolic and diastolic substances and ultimately leading to endothelial dysfunction. Studies have shown that adiponectin is secreted by adipose tissue and that plasma adiponectin levels are negatively correlated with obesity. Adiponectin [18] protects the body through anti-in ammatory and anti-atherosclerosis mechanisms, and low levels of adiponectin [19] are associated with cardiovascular events. The conclusion of this paper is a good proof of this view. Lower BMI has been shown to improve markers of endothelial dysfunction [20], and further randomized controlled trials are needed to demonstrate clinical e cacy.
Current studies suggest that slow blood ow may lead to sudden death due to malignant arrhythmias by affecting the length of QT [21] in the electrocardiogram or by inducing Brugada syndrome [22]. Slow blood ow should not be de ned only as a functional diagnosis because the majority of patients have a relatively good prognosis[23]. Long-term and chronic management of these patients should be taken seriously, especially as the use of B blockers, which can improve symptoms and reduce the risk of sudden death, may be a good option.

Study Limitations
This study has the following de ciencies. First, the nature of this study is a retrospective cohort study, and the e cacy of the conclusions is insu cient; therefore, a multicenter prospective randomized controlled trial is needed to further con rm the conclusions. Second, the outcome events in the follow-up events of this study were relatively insu cient, and further extension of follow-up events and an increased number of samples are needed to obtain higher positive results. Third, after two years of follow-up, we did not reassess cardiac function or perform coronary angiography in all patients; thus, it cannot be ruled out that in some patients, slow coronary blood ow has been alleviated. Finally, the low proportion of patients using nondihydropyridine calcium antagonists and nicordil in the population may in uence the results to a certain extent, resulting in a certain deviation.

Conclusion
Interventional therapy may not improve the prognosis of patients with slow coronary blood ow, and BMI is an independent predictor of readmission and sudden death in patients with CSFP. Further randomized controlled studies are needed to determine the mechanisms of these relationships and to determine whether treatment for slow coronary ow improves outcomes.  Tables   Table 1   BMI Table 2