Impact of Acute Total Occlusion of Culprit Artery on Outcome in NSTEMI – Results From a Large National Registry

The impact of acute total occlusion (TO) of culprit artery in non-ST-segment elevation myocardial infarction (NSTEMI) is not fully established. We aimed to evaluate clinical and angiographic phenotype and outcome of NSTEMI patients with TO (NSTEMI TO ) compared to NSTEMI patients without TO (NSTEMI NTO ) and those with ST-segment elevation and TO (STEMI TO ). infarction; artery; NSTEMI : non-ST-segment elevation myocardial infarction occluded infarct-related artery; ORPKI: Polish National Database of Invasive Coronary Procedures; PCI: percutaneous coronary intervention; RCA: right coronary artery; STE: ST-segment elevation; STEMI TO : ST-segment elevation myocardial infarction with occluded infarct-related artery; TIMI: Thrombolysis In Myocardial Infarction; TO: total occlusion of infarct-related artery.


Abstract Background
The impact of acute total occlusion (TO) of culprit artery in non-ST-segment elevation myocardial infarction (NSTEMI) is not fully established. We aimed to evaluate clinical and angiographic phenotype and outcome of NSTEMI patients with TO (NSTEMI TO ) compared to NSTEMI patients without TO (NSTEMI NTO ) and those with ST-segment elevation and TO (STEMI TO ).

Methods
Demographic, clinical and procedure-related data of patients with percutaneous coronary intervention (PCI) performed in acute myocardial infarction between 2014 and 2017 from the Polish National Registry were analysed.

Background
According to the European Society of Cardiology (ESC) guidelines patients with myocardial infarction (MI) and ST-segment elevation (STEMI) are eligible for emergency reperfusion therapy, whereas those with non-ST-segment elevation MI (NSTEMI) require further risk strati cation and thus the quali cation for invasive diagnosis and treatment is delayed. 1 The STEMI-NSTEMI paradigm is based on the observation that ST-segment elevation (STE) on the electrocardiogram (ECG) in the majority of patients with MI is associated with acute total occlusion (TO) of infarct-related artery (IRA), while subtotal IRA occlusion leads mostly to ST-segment depression and negative T-waves on the ECG. However, when quali cation for emergent reperfusion therapy is based on the ECG criteria, we lose around 25% of patients with acute TO of IRA who do not present STE. 2 NSTEMI patients are a very heterogeneous group in which ESC guidelines recommend urgent coronary angiography only in those with life-threatening ventricular arrhythmias, resistant angina pectoris and haemodynamic instability. These conditions may be accompanied by a total IRA occlusion, but this is not always the case. Thus, the percutaneous coronary intervention (PCI) may be deferred in signi cant subset of NSTEMI patients with TO of IRA which may result in delayed myocardial salvage and worse cardiovascular outcomes. 3 This study aimed to identify the key points of clinical characteristics, course of treatment and outcome of patients with NSTEMI with TO of IRA (NSTEMI TO ) by comparison with the two most outlying groups: patients with NSTEMI and non-occluded coronary artery (NSTEMI NTO ) and patients with STE and occluded IRA (STEMI TO ).

Methods
We analysed the data of patients with MI assembled within 48 months (2014-2017) into the ORPKI -Polish National Database of Invasive Coronary Procedures, coordinated by Jagiellonian University Medical College and endorsed by the Association of Cardiovascular Interventions of the Polish Cardiac Society. 4 All clinical data was collected by the operator and then uploaded into database after each procedure. The diagnosis of NSTEMI, STEMI, recognition of IRA, all clinical decisions during the coronary invasive procedure and de nition of periprocedural complications remained to the uploading ORPKI operators' experience and knowledge according to current ESC guidelines.
Acute TO of IRA was de ned in our study as Thrombolysis In Myocardial Infarction (TIMI) 0 ow during coronary angiography in patients with MI. 5 To achieve the aim of the study we compared 3 groups of patients: NSTEMI TO , NSTEMI NTO and STEMI TO and excluded from the analysis patients: with STEMI and non-occluded coronary artery (STEMI NTO ), without signi cant coronary artery stenosis, not treated with PCI and diagnosed as chronic total occlusion of IRA.
Our study was an observational, non-experimental, retrospective analysis and was performed in accordance with relevant guidelines and regulations. Only anonymized data was subjected to the research analysis and according to Regulation 2016/679 of the European Parliament and of the Council (EU) from 27 April 2016 on the protection of individuals with regard to the processing of personal data and on the free movement of such data, and with art. 9 Sect. 2 this study did not require any additional ethics board approval. All subjects of our study gave informed consent for personal data processing by the Association of Cardiovascular Interventions of Polish Cardiac Society before percutaneous coronary intervention.

Statistical analysis
Categorical variables are presented as numbers and percentages. Continuous variables were expressed as mean, standard deviation (SD) or median and interquartile range (IQR). Normality of continuous variables was assessed by the Kolmogorov-Smirnov-Lilliefors test. Equality of variances was assessed using the Levene's test. Differences between three groups were compared using the classical one-way analysis of variance (ANOVA) or the Welch's ANOVA depending on the equality of variances for normally distributed variables. The Kruskal-Wallis test was used for ordinal or non-normally distributed continuous variables. Categorical variables were compared by the Pearson's chi-square test. All post-hoc analyses were performed using the Benjamini-Hochberg procedure for controlling the False Discovery Rate (FDR).

Results
Results of 245,869 coronary angiographies performed in patients with MI were entered to the ORPKI registry. After exclusion of patients: a) without signi cant stenosis of coronary arteries, b) without occlusion of IRA in STEMI, c) without PCI treatment; d) with chronic total occlusion of IRA and e) with multivessel PCI treatment a total number of 131,729 patients with single-vessel PCI constituted the study group. Among them 65,206 (80.09%) patients with NSTEMI had no TO of IRA (TIMI > 0), while totally occluded IRA (TIMI = 0) were found in 16,209 (19.91%) patients with NSTEMI and in 50,314 (48.21%) with STEMI. The study owchart is shown in Fig. 1.

Clinical characteristic of the study groups
Patients with NSTEMI TO were younger than those with NSTEMI NTO but older than patients with STEMI TO .
The percentage of smokers was the highest in STEMI TO , lower in NSTEMI TO and the lowest in NSTEMI NTO . The prevalence of chronic diseases (arterial hypertension, diabetes, chronic kidney disease, chronic obstructive pulmonary disease) was the highest in NSTEMI NTO group, lower in NSTEMI TO group and the lowest in STEMI TO . All of the aforementioned differences were signi cant (p < 0.0001). Similar tendency was observed in the history of previous coronary revascularization (PCI or CABG), MI or stroke, which were the most frequent in patients with NSTEMI NTO , less frequent in NSTEMI TO and the least frequent in STEMI TO group (p < 0.0001).
Clinical status on admission in NSTEMI TO group was more severe than in NSTEMI NTO group but less serious than in STEMI TO group. More advanced Killip classes occurred with the highest frequency in patients with STEMI TO , lower frequency in NSTEMI TO and the lowest frequency in NSTEMI NTO . Cardiac arrest before admission was more frequent in patients with STEMI TO compared to NSTEMI TO , and more frequent in NSTEM TO than in NSTEMI NTO (See Table 1).   Fig. 3. NSTEMI TO was related predominantly to LCx artery occlusion, on the contrary LAD occlusion as a culprit lesion was observed the least often in this group. In patients with STEMI TO LCx occlusion was infrequent, while occlusion of RCA or LAD was prevalent.
NSTEMI NTO was related most often to LAD as the culprit lesion, less commonly to RCA and the least often to LCx (p for contingency analysis < 0.0001).

Analyses of PCI results
The successful revascularization outcome de ned as TIMI ow grade after PCI in the NSTEMI TO group was worse than in STEMI TO and NSTEMI NTO (Table 3). TIMI ow grade 3 in IRA after PCI was reached with the lowest frequency and TIMI ow grade 0 after PCI was noticed with the highest occurrence rate in NSTEMI TO group compared with both STEMI TO and NSTEMI NTO groups. No-re ow phenomenon, cardiac arrest during PCI and death during invasive procedure in NSTEMI TO were less frequent than in STEMI TO but more frequent than in NSTEMI NTO . Higher total radiation dose and total amount of contrast used during procedure were observed in NSTEMI TO compared with NSTEMI NTO and STEMI TO . Time from pain to FMC was the longest in NSTEMI TO group, even longer than in NSTEMI NTO group. In NSTEMI TO group patients postponed decision to seek medical help probably because of younger age (than in NSTEMI NTO group) and lack of previous experience with stenocardial pain. Longer time delay from pain to FMC in NSTEMI TO than in STEMI TO may be explained by lower severity of symptoms due to lower extent of ischemia in case of LCx occlusion (typical for NSTEMI TO in our study) in contrary to LAD or dominant RCA occlusion typical for STEMI TO .
Time delay to achieve the opening of the occluded artery in NSTEMI TO group in comparison to STEMI TO was ampli ed during in-hospital management what is noticeable as the pronounced difference (almost three times longer median time from FMC to balloon in ation in NSTEMI TO group).
In contrast, patients with NSTEMI TO in comparison to NSTEMI NTO were earlier considered as candidates for invasive management. The potential explanation is more severe clinical presentation caused by total artery occlusion. Higher frequency of cardiac arrest before admission and more advanced Killip class in NSTEMI TO group than in NSTEMI NTO group in our study con rms this hypothesis. Similar results were obtained by Shin et al. in the COREA-AMI Registry. 16 Other commonly used parameter of time delay in MI is the percentage of patients who receive PCI within 120 min. since the onset of symptoms. In the study of Terkelsen et al. approximately 50% of STEMI patients had balloon in ation within 120 min. 17 In our study almost 70% of STEMI TO , but only 25% of NSTEMI TO patients had PCI within 120 minutes. Terkelsen and other investigators con rmed that time delay to PCI worsened prognosis causing increased risk of mortality especially in patients with totally occluded artery. 3,15,17 In our NSTEMI patients approximately 20% had acute coronary artery occlusion which is less than previously reported by Khan 19 We must acknowledge that ECG has unsatisfactory sensitivity to diagnose coronary artery total occlusion, especially in posterolateral distribution. 20 It has been shown that the presence of STE on ECG enables to detect acute coronary TO in 70%-92% of cases for the LAD and RCA, but the ability of 12 lead ECG to diagnose LCx-related MI with coronary occlusion of IRA is below 50%. 2,21 Explanation is that LCx supplies the region of the heart placed more distally to the chest wall with no corresponding leads in standard ECG.
In our study patients with NSTEMI TO demonstrated more severe clinical condition on admission than those with NSTEMI NTO (more advanced Killip class, higher prevalence of death and cardiac arrest prior admission or during invasive procedure, no-re ow phenomenon), which is in concordance with prior studies showing that prognosis of patients with total occlusion without ST segment elevation is worse than in NSTEMI NTO patients. 6-7 We con rmed that the outcome after PCI (lower frequency of achieving TIMI 3 and higher frequency of TIMI 0) in NSTEMI TO is even inferior to STEMI TO . Possible explanation is that unrecognized acute coronary artery occlusion is associated with high morbidity and mortality 15 and the outcome in this group is worse than in those who received timely revascularization. 11,22 Two additional results of our study in NSTEMI TO group are noteworthy, i.e.: increased total radiation dose and higher amount of contrast media during PCI compared with both STEMI TO and NSTEMI NTO . It may be due to predominance of LCx as IRA in NSTEMI TO . Fetterly et al. showed that PCI of LCx correlates with increased total radiation dose due to anatomy and need for speci c oblique projections consuming higher radiation doses. 23 Furthermore, it has been proven that patients with longer time to reperfusion (NSTEMI TO patients in our study) are prone to receive signi cantly more contrast media during PCI. 24

Study limitations
Our study has several limitations. First, we should deduce very cautiously about detailed in-hospital prognosis because our analysis is based on data from the structured registry of prespeci ed clinical and periprocedural data spectrum only, without longitudinal follow-up, but with the largest number of evaluated patients. Second, the registry was created and ful lled by several operators, also quality of data depends on their individual knowledge; however only the most experienced operators collected the data.

Conclusions
Approximately one-fth of NSTEMI patients had acute total coronary artery occlusion (NSTEMI TO ).
According to the clinical characteristics NSTEMI TO seems to be an intermediate condition between   Comparison of median time form pain to rst medical contact (FMC) and form FMC to balloon in ation or angiogram, data presented as median, p < 0.0001 for all post-hoc analyses;