Stent length and -diameter and long-term clinical outcomes following percutaneous coronary intervention with drug-eluting stent implantation

Background

Long total stent length and small stent diameter have been associated with adverse events following percutaneous coronary intervention (PCI).

Purpose

To assess whether stent length and -diameter influence long-term target-lesion failure (TLF) following implantation of contemporary drug-eluting stents (DES) in an all-comers population undergoing PCI.

Methods

Patients included in the ReCre8 trial were stratified for troponin status and diabetes and randomized to implantation of a permanent polymer (PP-ZES) or polymer-free stent (PF-AES). Troponin negative patients were treated with dual antiplatelet therapy for one month, and troponin positive patients for twelve months. For the analysis on stent length, patients were divided in the quartiles of total stent length implanted per patient. Group 1a had a stent length of ≤18mm, Group 2a had a total stent length between 18 and 30mm, Group 3a had a total stent length of ≥30mm and lower than 49mm, and Group 4a had a total stent length of 49mm or more. For the analysis on stent diameter, patients were divided in the quartiles of the smallest stent diameter implanted per patient. Group 1b had a minimal stent diameter of ≤2.5mm, Group 2b had a minimal stent diameter between 2.5 and 3mm, Group 3b had a minimal stent diameter of ≥3mm and lower than 3.5mm, and Group 4b had a minimal stent diameter of 3.5mm or higher. The primary endpoint of TLF and its components – cardiac death, target-vessel myocardial infarction and target-lesion revascularization (TLR) – were assessed after three years.

Results

After division of patients in subgroups based on stent length, Group 1a included 409 patients (27.6%), Group 2a included 322 patients (20.7%), Group 3a included 376 patients (25.3%) and Group 4a included 377 patients (25.4%). After three years, TLF occurred more frequently in Group 4a with 6.6% in Group 1a, 8.4% in Group 2a, 7.7% in Group 3a and 18.0% in Group 4a (p<0.001) as shown in Figure 1. This was driven by a higher rate of TLR (p<0.001) and target-vessel myocardial infarction (p<0.001) in Group 4a. After division of patients in subgroups based on stent diameter, Group 1b included 408 patients (27.5%), Group 2b included 214 patients (14.4%), Group 3b included 477 patients (32.1%) and Group 4b included 386 patients (26.0%). After three years, TLF occurred more frequently in Group 1b with 14.0% vs. 7.9% in Group 2b, 8.6% in Group 3b and 9.3% in Group 4b (p=0.0241) as shown in Figure 2. The difference in TLF was driven by a higher rate of TLR in Group 1b (8.6% vs. 3.7% vs. 4.4% vs. 4.9%; p=0.016).

Conclusion

In an all-comers population undergoing PCI with implantation of contemporary DES, a stent length ≥49mm and a stent diameter ≤2.5mm were associated with a higher rate of TLF after three years.

Funding Acknowledgement

Type of funding sources: None.

Ischemic signs on the post-resuscitation ECG in absence of STEMI is associated with lower survival. A COACT trial's sub-study

Background

The recently published Coronary Angiography after Cardiac arrest (COACT) trial found that urgent coronary angiography did not improve 90-day survival in out-of-hospital cardiac arrest (OHCA) patients without STEMI. The prognostic value of signs of ischemia on the ECG in absence of STEMI, is yet to be determined.

Purpose

To assess whether ischemic ECG patterns such as ST-depression and T-wave inversion are predictors for survival after OHCA in patients without STEMI.

Methods

In the COACT trial, patients with return of spontaneous circulation after OHCA with initial shockable rhythm and absence of ST-segment elevation were included. In this sub-study, the first post-resuscitation ECG recorded at the hospital was analysed for signs of ischemia. Ischemia was defined as ST-depression or T-wave inversion >1mm in ≥2 contiguous leads, or both. Primary endpoint was 90-day survival. Secondary endpoints included angiographic outcomes and left ventricular function assessed by cardiac magnetic resonance imaging or echocardiography.

Results

In total, 552 patients were included in the COACT trial. For this sub-study, 510 OHCA-patients had an ECG available for assessment of whom 340 patients (66.7%) had signs of ischemia on the ECG and 170 patients (33.3%) were without signs of ischemia. Patients with signs of ischemia were significantly older (p=0.003) and more frequently had a history of CAD (p=0.009). Left ventricular ejection fraction was lower in those with signs of ischemia (p=0.007). The number of acute thrombotic occlusions did not differ between groups (p=0.34). Patients with signs of ischemia had a significantly worse 90-day survival compared to patients that showed no signs of ischemia (HR 1.51 (95% CI 1.08–2.12); log-rank p=0.02). Furthermore, larger ST-depression was found to be associated with worse survival (log-rank p=0.01). Neurologic injury was the most common cause of death and its incidence did not differ between the groups (p=0.77).

Conclusion

Signs of ischemia in absence of STEMI on the post-resuscitation ECG is a predictor for worse survival. Furthermore, a correlation was found between the sum of ST-depression and lower survival rate.

Funding Acknowledgement

Type of funding sources: Public Institution(s). Main funding source(s): Research grants of Netherlands Heart institue, Biotronic, AstraZeneca Survival plot signs of ischemia on ECGSurvival plot sum of ST-depression

Associations of observational and genetically determined caffeine intake with coronary artery disease and diabetes

Background

Caffeine is the most widely consumed psychostimulant and is associated with lower risk of coronary artery disease (CAD) and type 2 diabetes (T2D). However, whether these associations are causal remains unknown.

Objectives

This study aimed to identify genetic variants associated with caffeine intake, and to investigate possible causal links between genetically determined caffeine intake and CAD or T2D. Additionally, we aimed to replicate previous observational findings between caffeine intake and CAD or T2D.

Methods

Genome wide associated studies (GWAS) were performed on caffeine intake from coffee, tea or both in 407,072 UK Biobank participants. Identified variants were used in a two-sample Mendelian randomization (MR) approach to investigate evidence for causal links between caffeine intake and CAD in CARDIoGRAMplusC4D (60,801 cases; 123,504 controls) or T2D in DIAGRAM (26,676 cases; 132,532 controls). Observational associations were tested within UK Biobank using Cox regression analyses.

Results

Moderate observational caffeine intakes from coffee or tea were associated with lower risks of CAD or T2D compared to no or high intake, with the lowest risks at intakes of 120–180 mg/day from coffee for CAD (HR=0.77 [95% CI: 0.73–0.82; P<1e-16]), and 300–360 mg/day for T2D (HR=0.76 [95% CI: 0.67–0.86]; P=1.57e-5). GWAS identified 51 novel genetic loci associated with caffeine intake, enriched for central nervous system genes. In contrast to observational analyses, MR analyses in CARDIoGRAMplusC4D and DIAGRAM yielded no evidence for causal links between caffeine intake and the development of CAD or T2D.

Conclusions

MR analyses indicate caffeine intake might not protect against CAD or T2D, despite protective associations in observational analyses.

Manhattan_plot_CaffeineIntake

Funding Acknowledgement

Type of funding source: None

Risk results from screening for a high cardiovascular disease risk by means of traditional risk factor measurement or coronary artery calcium scoring in the ROBINSCA trial

Aims

Screening for a high cardiovascular disease (CVD) risk followed by preventive treatment can potentially reduce coronary heart disease (CHD)-related morbidity and mortality. ROBINSCA (Risk Or Benefit IN Screening for CArdiovascular disease) is a population-based randomized controlled screening trial that investigates the effectiveness of CVD screening in asymptomatic participants using the Systematic COronary Risk Evaluation (SCORE) model or Coronary Artery Calcium (CAC) scoring. This study describes the distributions in risk and treatment in the ROBINSCA trial.

Methods and results

Individuals at expected elevated CVD risk were randomized (1:1:1) into the control arm (n=14,519; usual care); screening arm A (n=14,478; SCORE, 10-year fatal and non-fatal risk); or screening arm B (n=14,450; CAC scoring). Preventive treatment was largely advised according to current Dutch guidelines. Risk and treatment differences between the screening arms were analysed. 12,185 participants (84.2%) in arm A and 12,950 (89.6%) in arm B were screened. 48.7% were women, and median age was 62 (InterQuartile Range 10) years. SCORE screening identified 45.1% at low risk (SCORE<10%), 26.5% at intermediate risk (SCORE 10–20%), and 28.4% at high risk (SCORE≥20%). According to CAC screening, 76.0% were at low risk (Agatston<100), 15.1% at high risk (Agatston 100–399), and 8.9% at very high risk (Agatston≥400). CAC scoring significantly reduced the number of individuals indicated for preventive treatment compared to SCORE (relative reduction women: 37.2%; men: 28.8%).

Conclusion

We showed that compared to risk stratification based on SCORE, CAC scoring classified significantly fewer men and women at increased risk, and less preventive treatment was indicated.

ROBINSCA flowchart

Funding Acknowledgement

Type of funding source: Public grant(s) – EU funding. Main funding source(s): Advanced Research Grant

Relationship between cardiovascular risk factors and coronary calcification in a middle-aged Dutch population: the Imalife study

Background

Systematic COronary Risk Evaluation (SCORE) has been proposed to assess the 10-year risk of fatal cardiovascular diseases, with distinction between low-risk and high-risk countries. Risk modifiers are recommended to further improve risk reclassification, for example the coronary artery calcium (CAC) score. CAC scoring can significantly improve risk prediction for coronary events based on outcome studies. The impact of CAC scoring on risk classification in a middle-aged cohort from a low-risk country in comparison to SCORE is unknown.

Purpose

To assess presence of coronary calcification and association with cardiovascular risk factors and related SCORE risk in a middle-aged population from a low risk country.

Methods

Coronary calcification and classical cardiovascular risk factors were analyzed in 4,083 Dutch participants aged 45–60 years (57.9% women) without a known history of coronary artery disease in the population-based ImaLife (Imaging in Lifelines) study. Individuals underwent non-contrast cardiac CT using third generation dual-source CT. Coronary artery calcium (CAC) scores were quantified using Agatston's method. Age- and sex- specific distributions of CAC categories (0, 1–99, 100–299, ≥300) and percentiles were assessed. Distribution of CAC categories was compared to SCORE risk categories (<1%, ≥1% to 5%, and ≥5%) for low risk countries. Relationship between risk factors and CAC presence was evaluated by logistic regression models. Population attributable fractions (PAFs) of classical risk factors for CAC presence were estimated to investigate potential prevention strategy.

Results

CAC was present in 54.5% of men and in 26.5% of women. With increasing age, an increasing percentage had a positive CAC score, from 38.1% of men and 15.2% of women at age 45–49 years, to 66.9% of men and 36.6% of women at age 55–60. Mean SCORE was 1.3% (2.0% in men, 0.7% in women). In SCORE risk <1%, 32.7% of men and 17.1% of women had CAC. In men with SCORE risk ≥5% (0.1% of women), 26.9% had no CAC. Overall PAF for presence of CAC of the classical risk factors was 18.5% in men and 31.4% in women. PAF was highest for hypertension (in men 8.0%, 95% CI 4.2–11.8%; in women 13.1%, 95% CI 7.9–18.2%) followed by hypercholesterolemia and obesity.

Conclusion

In this middle-aged Dutch cohort, slightly over half of men and a quarter of women had any CAC. With age there was an increase in CAC presence for both sexes. Only a minor proportion of CAC presence was attributable to classical risk factors. This provides further support that CAC scoring can impact risk classification in a middle-aged population of a low-risk country.

Funding Acknowledgement

Type of funding source: Other. Main funding source(s): The ImaLife study is supported by an institutional research grant from Siemens Healthineers and by the Ministry of Economic Affairs and Climate Policy by means of the PPP Allowance made available by the Top Sector Life Sciences & Health to stimulate public-private partnerships.

Calcium scores distribution across coronary artery by age and sex: the ImaLife study

Background

The distribution of coronary artery calcium (CAC) across the coronary system increases the ability to predict coronary events compared to traditional CAC scoring alone. Reference values for regional distribution of CAC by age and sex are not yet available for a general European population.

Purpose

To investigate the distribution of CAC across the coronary arteries by age and sex in the population-based ImaLife study.

Methods

ImaLife is part of Lifelines, a multi-generational, prospective cohort study with over 167,000 participants from the northern Netherlands. From 2017–2019, 5,531 participants aged 45–84 years underwent non-contrast cardiac CT using third-generation dual-source CT. Total and vessel-specific CAC scores (Agatston's method) were acquired semi-automatically using dedicated software. Participants with a positive CAC score were classified into three groups: total CAC score 1–100, 101–300 and >300. The diffusivity index (equation: 1 – [highest one-vessel CAC/total CAC]) was calculated. The diffusivity index is an expression of the relative distribution of CAC across the coronary arteries. Data were analyzed for the whole population and by sex and age groups. Mann-Whitney U test was used to analyze the diffusity index in men and women. Kruskal-Wallis H tests were performed to test the diffusivity index in different age groups.

Results

In total 2,376 men (mean age 56.4±7.7 years) and 3,155 women (mean age 56.0±7.5 years) were analyzed. In participants with CAC, 1, 2, 3 or 4 vessels were affected in 523 (22.0%), 560 (17.7%), 371 (15.6%) and 257 (8.1%) of men, respectively, and in 385 (16.2%), 175 (5.5%), 185 (7.8%) and 81 (2.6%) of women, respectively (P<0.001). The number of 1, 2, 3 or 4 vessels affected were significantly different by age (p<0.001). In age category 45–49 years, CAC in 1, 2, 3, and 4 vessels was present in 60.1%, 21.6%, 15.5%, and 2.9%, respectively; for age 74+ years, these percentages were 19.3%, 19.3%, 31.1% and 30.3%, respectively. The number of affected vessels were significantly different in different CAC categories (p<0.001), see Figure. More vessels were affected in higher CAC categories. The median diffusivity index was higher in men than in women (0.10 (IQR: 0–0.36) vs 0 (IQR: 0–0.24), p<0.001) and increased by increasing age. For age categories of 45–49, 50–54, 55–59,60–64, 65–69, 70–74, and >74 years, diffusivity indexs were 0 (IQR: 0–0.12), 0 (IQR: 0–0.22), 0.02 (IQR: 0–0.28), 0.10 (IQR: 0–0.35), 0.16 (IQR: 0–0.42), 0.20 (IQR: 0–0.44), and 0.28 (IQR: 0.03–0.45) (p<0.001).

Conclusions

In this Dutch population-based study, male participants had higher prevalence of CAC with higher number of involved vessels, and a higher diffusivity index compared to women. For both sexes, involved vessels and diffusivity index increased with age. The reference values of this regional distribution of CAC in a European population can assist in risk categorization of cardiovascular events.

The CAC distribution in ImaLife

Funding Acknowledgement

Type of funding source: Other. Main funding source(s): Siemens Healthineers

Implementation study of CT calcium score in patients with atypical angina pectoris and non-specific thoracic complaints in primary care: rationale, objectives, and design of the CONCRETE study

Background

Identifying and excluding coronary artery diseases (CAD) in patients with atypical angina pectoris (AP) and non-specific thoracic complaints is a challenge for general practitioners (GPs). It is unclear what the best diagnostic and prognostic strategy is for these patients in primary care. Computed Tomography coronary calcium scoring (CT CCS) has a high sensitivity for early diagnosis and exclusion of CAD. However, CT CCS has not been tested in a primary care setting. In the CONCRETE study, the impact of direct access of GPs to CT CCS on management and diagnosis will be investigated. CONCRETE is the abbreviation for “COroNary Calcium scoring as fiRst-linE Test to dEtect and exclude coronary artery disease in GPs patients with stable chest pain.” Currently, we present the rationale, objectives and design of this study.

Purpose

The purpose of CONCRETE is to study the implementation of CT CCS in primary care, and determine the effects on GP office level. The primary objective is to determine the increase in detection/treatment rate of CAD in GP offices with CT CCS, compared to GP offices with standard of care.

Methods

CONCRETE is an implementation study with a cluster randomized design, in which direct access to CT CCS in a group of 40 GP offices is compared to the standard of care in a control group of 40 GP offices. In both arms, inclusion of 800 patients with atypical angina pectoris and non-specific thoracic complaints is intended.

Results

Recruitment of GP offices and participants started in January 2019. First results will be presented.

Conclusion

CONCRETE will determine whether access to CT CCS will lead to earlier and more effective detection or exclusion of CAD in GP patients with atypical angina pectoris and non-specific thoracic complaints, in comparison to the standard of care. Implementation of the study findings could initiate a change in the (Dutch) GP healthcare policy, for these patients in primary care.

Funding Acknowledgement

Type of funding source: Public grant(s) – National budget only. Main funding source(s): Dutch Heart Foundation

P3397Risk Or Benefit IN Screening for CArdiovascular disease (ROBINSCA): results from screening for a high cardiovascular disease risk by using a risk prediction model or coronary artery calcium scoring

Introduction

The ROBINSCA (Risk Or Benefit IN Screening for CArdiovascular disease) trial is a large-scale population-based randomized controlled screening trial with the aim to investigate whether screening for a high risk of cardiovascular disease (CVD) by means of either the Systematic COronary Risk Evaluation (SCORE) model or coronary artery calcium (CAC) scoring followed by preventive treatment is effective in reducing morbidity and mortality from coronary heart disease (CHD). This study shows the results of the CVD risks as assessed by the two screening tools.

Methods

Based on the Dutch population registry, 394,058 men aged 45–74 years and women aged 55–74 years received an information brochure, an invitation to participate in the trial, a baseline questionnaire with waist circumference tape and an informed consent form. Eligible individuals with an expected high CVD risk were randomized (1:1:1) into a control arm (n=14,519), intervention arm A (n=14,478) or intervention arm B (n=14,450). In the control arm, usual care was continued. In intervention arm A, participants were screened for a high risk of CVD using the SCORE model based on traditional risk factors. In intervention arm B, CAC scoring after computed tomography scanning was used for screening. After screening en risk communication, preventive treatment according to the Dutch guidelines is advised for high risk persons.

Results

Screening uptake was 84.2% in intervention arm A and 89.6% in intervention arm B. Of the screened participants, 48.7% was female, median age at screening was 62 (Interquartile Range 10), 35.2% was high educated, 19.6% was baseline smoker and 41.4% had a positive family history of myocardial infarction. The assessed CVD risk status according to SCORE screening was stratified into three risk categories; 45.1% was at low risk (SCORE<10%), 26.5% was at intermediate risk (SCORE 10–20%), and 28.4% was at high risk (SCORE ≥20%). According to CAC screening, 76.0% was at low risk (Agatston <100), 15.1% was at high risk (Agatston 100–399), and 8.9% was at very high risk (Agatston ≥400). Associations between baseline variables and increased CVD risk will be analyzed soon and will be available in summer 2019.

Conclusions

Using different screening tools resulted in reclassification of the CVD risk. CAC screening caused a substantial shift to more low risk individuals. This might, when screening is found to be effective, lead to less overtreatment in prevention of CVD events. Future 5-year follow-up data should provide evidence about whether population-based screening with subsequent preventive treatment is (cost-)effective in reducing CHD-related morbidity and mortality.

P5309Age and gender distributions of coronary artery calcium in the Dutch adult population: preliminary results of the ImaLife study

Background

Coronary artery calcium (CAC) scoring is a promising tool for cardiovascular risk classification. Population-based reference values are important for the clinical interpretation of CAC scores.

Purpose

To establish standards of CAC distributions by age and gender in an unselected Dutch population, which can be used to determine reference values.

Methods

ImaLife (Imaging in Lifelines) is a computed tomography (CT) based substudy of the Lifelines cohort, with a primary aim to establish reference values of imaging biomarkers for early stages of coronary artery disease in adults (above 45 years old). In total, 12,000 participants will be enrolled from an unselected adult population in the northern Netherlands for CAC scoring with third generation dual-source CT. CAC is quantified with dedicated commercial software using the Agatston method.

Results

Included so far were 3,702 participants (57.5% females, mean age 54 years, range 45–82 years). CAC was present in 39.2% of participants, with a higher prevalence of CAC in men (55.3%) than in women (27.3%). CAC scores increased with increasing age in both genders. The percentiles of CAC scores by age and gender groups are summarized in the table.

Agatston CAC score percentiles by age and gender Percentiles Women – Age, years Men – Age, years 45–49 50–54 55–59 60–64 65∼ 45–49 50–54 55–59 60–64 65∼ N 505 634 719 260 10 355 473 543 185 18 25th 0 0 0 0 0 0 0 0 1 75 50th 0 0 0 0 4 0 1 6 22 556 75th 0 0 6 33 386 6 21 72 129 751 90th 4 26 77 120 1037 49 154 242 500 1803

Conclusion

This preliminary result presents CAC distribution by age and gender in a middle-aged unselected Dutch population. Compared with the Heinz Nixdorf Recall study, CAC scores in our cohort for both genders were lower in the 5-year age groups between 45 and 64 years. Based on the overall data, expected within 2 years, reference values of CAC for the Dutch population can be established.

2349Selatogrel, a novel P2Y12 inhibitor for emergency use, achieves rapid, consistent and sustained platelet inhibition following single-dose subcutaneous administration in stable CAD patients

Background

In the setting of AMI, rapid platelet inhibition is desirable but the onset of pharmacodynamic (PD) effect of oral platelet P2Y12 inhibitors is delayed, sometimes for hours. Subcutaneous (s.c) administration of a rapidly-acting P2Y12 inhibitor would overcome many of the limitations of available therapies. Patients with stable CAD were investigated initially.

Purpose

To characterise the inhibition of platelet aggregation and pharmacokinetics (PK) of a single dose of selatogrel, a novel s.c P2Y12 inhibitor, in patients with stable CAD.

Methods

Patients with stable CAD receiving oral antiplatelet therapy (aspirin and/or oral P2Y12 inhibitor) were randomized to 1 of 8 groups based on treatment (selatogrel or matching placebo), dose (8 mg or 16 mg) and s.c injection site (thigh or abdomen). Venous blood samples were collected into PPACK anticoagulant tubes. Platelet reactivity was assessed by VerifyNow PRU (P2Y12 reaction units) test before and 15 min, 30 min and 1, 2, 4, 8 and 24 h after injection. Light-transmittance aggregometry (LTA; ADP 20 uM) was also performed. PK samples were collected up to 24 h post-dose. Adverse events occurring within 30 days were recorded. Responders were defined as having PRU <100 at 30 min after injection and lasting ≥3 h.

Results

345 patients (mean age 65 y; 20% female; 31% diabetes) received selatogrel 8 mg (n=114), selatogrel 16 mg (n=115) or placebo (n=116). 97% were on background therapy with aspirin (or its derivative carbasalate) and 35% with oral P2Y12 inhibitor (clopidogrel 23%, prasugrel 4%, ticagrelor 8%). 89% of subjects were responders to selatogrel 8 mg, 90% to selatogrel 16 mg and 16% to placebo (P<0.0001). At 15 min post-dose, PRU values (mean±SD) were 10±25 with selatogrel 8 mg, 5±10 with selatogrel 16 mg and 163±73 with placebo (Figure). PRU levels were maintained at 2 and 4 h for both doses and gradually returned to pre-dose levels by 24 h post-dose (Figure). LTA results were consistent with the VerifyNow results. PD responses were similar for thigh and abdomen injection sites. Selatogrel was well tolerated: mild dyspnoea (or moderate dyspnoea, n=1, with 16 mg) occurred in 5% and 9% with selatogrel 8 mg and 16 mg, respectively, vs 0% with placebo; dizziness occurred in 4% and 4% vs 1%, respectively, without significant haemodynamic or ECG changes. Bleeding events occurred in 9.6% and 4.3% with selatogrel 8 mg and 16 mg, respectively, vs 6.9% with placebo. Pharmacokinetic data will be presented.

Conclusions

Selatogrel has a rapid PD effect following s.c injection in patients with stable CAD, within 15 min in most patients. The consistent and high levels of P2Y12 inhibition with a single 8 mg or 16 mg dose are sustained for over 4 hours, following which platelet reactivity progressively recovers over 24 h. Selatogrel was well tolerated, with mostly mild, transient dyspnoea observed in <10% patients. These data support further studies of selatogrel for emergency treatment of AMI patients.

Acknowledgement/Funding

Fully funded by Idorsia Pharmaceuticals Ltd

P1218Deep learning survival analysis enhances the value of hybrid PET/CT for long-term cardiovascular event prediction

Background

Deep Learning (DL) is revolutionizing cardiovascular medicine through complex data-pattern recognition. In spite of its success in the diagnosis of coronary artery disease (CAD), DL implementation for prognostic evaluation of cardiovascular events is still limited. Traditional survival models (e.g.Cox) notably incorporate the effect of time-to-event but are unable to exploit complex non-liner dependencies between large numbers of predictors. On the other hand, DL hasn't systematically incorporated time-to-event for prognostic evaluations. Long-term registries of hybrid PET/CT imaging represent a suitable substrate for DL-based survival analysis due the large amount of time-dependent structured variables that they convey. Therefore, we sought to evaluate the feasibility and performance of DL Survival Analysis in predicting the occurrence of myocardial infarction (MI) and death in a long-term registry of cardiac hybrid PET/CT.

Methods

Data from our PET/CT registry of symptomatic patients with intermediate CAD risk who underwent sequential CT angiography and 15O-water PET for suspected ischemia, was analyzed. The sample has been followed for a 6-year average for MI or death. Ten clinical variables were extracted from electronic records including cardiovascular risk factors, dyspnea and early revascularization. CT angiography images were evaluated segmentally for: presence of plaque, % of luminal stenosis and calcification (58 variables). Absolute stress PET myocardial perfusion data was evaluated globally and regionally across vascular territories (4 variables). Cox-Nnet (a deep survival neural network) was implemented in a 5-fold cross-validated 80:20 split for training and testing. Resulting DL-hazard ratios were operationalized and compared to the observed events developed during follow-up. The performance of Cox-Nnet evaluating structured CT, PET/CT, and PET/CT+clinical variables was compared to expert interpretation (operationalized as: normal coronaries, non-obstructive CAD, obstructive CAD) and to Calcium Score (CaSc), through the concordance (c)-index.

Results

There were 426 men and 525 women with a mean age of 61±9 years-old. Twenty-four MI and 49 deaths occurred during follow-up (1 month–9.6 years), while 11.5% patients underwent early revascularization. Cox-Nnet evaluation of PET/CT data (c-index=0.75) outperformed categorical expert interpretation (c-index=0.54) and CaSc (c-index=0.65), while hybrid PET/CT and PET/CT+clinical (c-index=0.75) variables demonstrated incremental performance overall independent from early revascularization.

Conclusion

Deep Learning Survival Analysis is feasible in the evaluation of cardiovascular prognostic data. It might enhance the value of cardiac hybrid PET/CT imaging data for predicting the long-term development of myocardial infarction and death. Further research into the implementation of Deep Learning for prognostic analyses in CAD is warranted.