Low-dose intracoronary alteplase during primary percutaneous coronary intervention in patients with acute myocardial infarction: the T-TIME three-arm RCT

Article history

The research reported in this issue of the journal was funded by the EME programme as project number 12/170/45. The contractual start date was in November 2015. The final report began editorial review in July 2019 and was accepted for publication in January 2020. The authors have been wholly responsible for all data collection, analysis and interpretation, and for writing up their work. The EME editors and production house have tried to ensure the accuracy of the authors’ report and would like to thank the reviewers for their constructive comments on the final report document. However, they do not accept liability for damages or losses arising from material published in this report.

Copyright statement

© Queen’s Printer and Controller of HMSO 2020. This work was produced by McCartney et al. under the terms of a commissioning contract issued by the Secretary of State for Health and Social Care. This issue may be freely reproduced for the purposes of private research and study and extracts (or indeed, the full report) may be included in professional journals provided that suitable acknowledgement is made and the reproduction is not associated with any form of advertising. Applications for commercial reproduction should be addressed to: NIHR Journals Library, National Institute for Health Research, Evaluation, Trials and Studies Coordinating Centre, Alpha House, University of Southampton Science Park, Southampton SO16 7NS, UK.

2020 Queen’s Printer and Controller of HMSO

Research summary

During the Trial of low-dose adjunctive alTeplase during priMary PCI (T-TIME), the aims and objectives of the study were unchanged. A list of the amendments to the trial protocol is provided in Appendix 1.

To best ensure the safety of the participants, additional exclusion criteria were (1) a requirement for immunosuppressive drug therapy at any time during the past 3 months and (2) active or prophylactic treatment with oral or parenteral antibiotic, antifungal or antiviral therapy to prevent or treat infection. The change was implemented to ensure the safety of the participants.

Finally, on the recommendation of the Data and Safety Monitoring Committee, recruitment was discontinued on 21 December 2017 because a prespecified futility criterion for efficacy was met. Specifically, the conditional power for an analysis on the primary efficacy outcome based on 40% of the randomised population (n = 267) with follow-up to 3 months was < 30% in both treatment groups. The sponsor initially placed a temporary halt on enrolment pending a review by the Trial Steering Committee (TSC) and funder, and no further enrolment occurred following this determination.

Background

Ischaemic heart disease is a major cause of premature disability1 and death2. Acute coronary thrombosis causes ST-elevation myocardial infarction (STEMI). The evidence-based standard of care for STEMI is primary percutaneous coronary intervention (PCI) to emergently reopen the occluded coronary artery and secure vessel patency with a stent. 3 Primary PCI restores normal coronary artery blood flow in more than 9 in 10 patients. 4 However, failed microvascular reperfusion has been estimated to occur in almost half of all treated patients. 5,6 This complication, described as microvascular obstruction, is independently predictive of an unfavourable cardiac prognosis. 7 During primary PCI, displacement of thrombus within the lumen of the main coronary artery and microvascular thrombosis,8–12 notably of fibrin-rich microthrombi,9 are substrates for microvascular obstruction. Clinicians lack the therapeutic tools to treat microvascular obstruction. 3

Fibrinolytic therapy is also effective for acute coronary thrombosis. 13 Facilitated PCI involving half- or full-dose adjunctive fibrinolytic therapy that is given before PCI with stenting improves coronary flow acutely. 14,15 However, this therapeutic strategy16,17 causes paradoxical activation of thrombin, clot formation and bleeding. Sezer et al. 15 implemented a modified, pharmaco-invasive strategy by administering adjunctive low-dose thrombolytic therapy with 250 kU of streptokinase at the end of primary PCI in 41 patients with acute STEMI. This approach appeared to improve myocardial blood flow, as revealed by coronary angiography and invasive measures of coronary microvascular function during repeat coronary angiography 2 days after PCI.

However, this study had some limitations:

• The improvement in perfusion did not translate into improvements in long-term ventricular size or function.

• Streptokinase is not selective for fibrin and it has a higher risk of bleeding than fibrin-specific thrombolytics, such as alteplase and tenecteplase. 16

• Streptokinase was given through the guide catheter rather than through local delivery of the lytic drug directly into the culprit artery. Drug administration into the guide catheter means that the thrombolytic would be dispersed into coronary branches that were not connected to the culprit artery; that is, the drug was delivered to non-relevant coronary arteries, which potentially reduced drug exposure to the fibrin-rich thrombus in the culprit artery microcirculation.

• Given that streptokinase was given after the end of the PCI procedure, thrombolysis could only treat established microvascular thrombosis rather than prevent its occurrence, as would be the case if it were administered earlier, during PCI (as is proposed in our study).

These problems may explain why Sezer et al. ’s15 findings have not been adopted into clinical practice. Sezer et al. ’s15 findings indicate that intracoronary thrombolysis has the potential to reduce microvascular damage during primary PCI. Our proposal used alteplase, which is a more potent fibrin-specific second-generation lytic drug than streptokinase with comparable potency to third-generation drugs, such as reteplase; accordingly, our trial aimed to determine whether or not alteplase would reduce microembolic thrombi and improve microvascular function when delivered locally.

Rationale

Primary PCI is a multistep procedure. The first step involves coronary angiography to reveal the coronary anatomy and the culprit lesion implicated in the acute myocardial infarction (MI). Second, coronary reperfusion may be achieved using a thin (0.014-inch) coronary guidewire followed by balloon angioplasty and/or selective thrombus removal by transcatheter aspiration thrombectomy. Third, the standard of care procedure involves stenting followed by optimisation of the stent result by high-pressure balloon angioplasty in the stent.

Patients with acute STEMI who present with a blocked coronary artery and/or an artery with a heavy thrombus burden are at an increased risk of developing heart failure. This trial aimed to enrol patients with a heavy coronary thrombus burden at initial angiography to test the hypothesis that a therapeutic strategy involving reduced-dose alteplase, given early after coronary reperfusion as a single dose, would both prevent and treat distal microvascular thrombosis and microvascular obstruction. The trial aimed to determine the lowest dose of alteplase that would be effective in reducing microvascular obstruction.

The T-TIME trial aimed to investigate whether or not a therapeutic strategy involving low-dose intracoronary fibrinolytic therapy with alteplase infused early after coronary reperfusion would both prevent and reduce microvascular obstruction.

Hypothesis

We hypothesised that a therapeutic strategy involving low-dose intracoronary fibrinolytic therapy with alteplase infused early after coronary reperfusion will prevent and reduce microvascular obstruction. To test this hypothesis we enrolled patients with acute STEMI and a large thrombus burden evident at initial coronary angiography. A three-arm design was adopted in which 10 mg or 20 mg of alteplase (representing one-tenth and one-fifth of the standard dose, respectively) or placebo was administered following reperfusion and before stent implantation when residual thrombus burden is at its greatest. The trial aimed to determine the lowest effective dose of alteplase in reducing microvascular obstruction, and multiple surrogate outcomes for efficacy, safety and mechanisms were assessed.

Objectives

The T-TIME trial was a double-blind, randomised, parallel-group, placebo-controlled clinical trial that was designed to examine the efficacy and safety of reduced-dose intracoronary alteplase in STEMI patients receiving primary PCI.

Trial design

This was a randomised, double-blind, parallel-group, Phase II clinical trial of low-dose adjunctive alteplase during primary PCI (Figures 1 and 2).

FIGURE 1.

Graphical depiction of the trial protocol. Patients presenting with acute STEMI meeting the study eligibility criteria were enrolled by research staff in the cardiac catheterisation laboratory following reperfusion of the culprit artery. Blood tests were performed acutely (0, 2 and 24 hours after randomisation, which defines baseline) and again at the time of multiparametric magnetic resonance imaging (MRI) at 2–7 days and 3 months after enrolment. An electrocardiogram (ECG) was carried out prior to reperfusion (baseline), at 60 minutes and then again at 3 months. Parametric magnetic resonance imaging, starting from left: cine imaging allows calculation of left ventricular function and volumes; late gadolinium-enhanced MRI allows determination of infarct size (bright white area) and microvascular obstruction (hypointense, black core) within the infarct highlighted with the yellow arrow; T2* magnetic resonance imaging, far right image revealed myocardial haemorrhage (white arrow) within the infarct core. MVO, microvascular obstruction.

FIGURE 2.

Study procedures. EDTA, ethylenediaminetetraacetic acid; NT-proBNP, N-terminal pro–B-type natriuretic peptide; NYHA, New York Heart Association.

Informed consent and study protocol

Screening, witnessed verbal informed consent, study drug administration and acute assessments of efficacy took place during the standard of care emergency primary PCI procedure in the cardiac catheterisation laboratory. A screening log was prospectively completed. Only patients who were sufficiently well to understand the information about the study, as described by the attending cardiologist, were eligible to participate. The decision of whether or not a patient was eligible to be included was made and documented by the cardiologist. The study information sheet that had been approved by the research ethics committee was subsequently provided to each participant on the ward, where written informed consent was obtained. The participants were followed up unless consent was withdrawn.

The protocol and statistical analysis plan are provided in Appendix 2. The trial was reviewed and approved by an ethics committee of the West of Scotland Research Ethics Service (reference 13-WS-0119), adhered to Guidelines for Good Clinical Practice in Clinical Trials24 and complied with the Declaration of Helsinki. 25

Participants and eligibility criteria

Patients with a clinical diagnosis of acute STEMI were eligible for randomisation according to the following eligibility criteria.

Setting

The participants were enrolled in 11 hospitals in the UK, and guideline-based medical and invasive management was recommended. 3 Enrolment started on 17 March 2016.

Randomisation, implementation and blinding

Participants were randomised by staff in the catheter laboratory using an interactive voice response-based randomisation system. The randomisation sequence was computer generated, using the method of randomised permuted blocks of length 6, with stratification by location of STEMI (anterior vs. non-anterior) and study site. The allocation sequence was on a 1 : 1 : 1 basis between the placebo, the 10 mg of alteplase and the 20 mg of alteplase groups, and the sequence was concealed electronically. The participants, staff and researchers were blinded to the treatment group allocation.

Standard care

Primary PCI followed contemporary practice guidelines. 3 Standard care for coronary reperfusion was recommended according to contemporary practice guidelines3 using either balloon angioplasty or aspiration thrombectomy for thrombus-containing lesions. A coronary balloon diameter (mm) versus lumen diameter (mm) relationship of < 1 : 1 and a low inflation pressure were recommended to minimise thrombus embolisation. The balloon angioplasty was intended to stabilise the thrombotic lesion and prevent vessel reocclusion prior to stent implantation. Antithrombotic therapy included oral antiplatelet drugs and intravenous heparin (5000 IU or as per standard practice) at the first medical contact. The target activated clotting time was 250 seconds.

Interventions

After successful reperfusion (TIMI flow grade of 2 or above) the participants received the allocated intervention immediately in the catheter laboratory. The study drug (placebo or 10 mg or 20 mg of alteplase) was manually infused before stent implantation. The drug was reconstituted by the clinical staff using 20 ml of sterile water for injection. The cardiologist then infused the solubilised drug over 5–10 minutes directly into the infarct-related artery proximal to the culprit lesion using either an intracoronary catheter or the guiding catheter if selectively engaged.

Outcomes

The methods for the assessments of the primary and secondary outcomes are described in Appendix 3.

Health status

Health-related quality of life (HRQoL) [EuroQol-5 Dimensions, three-level version (EQ-5D-3L)] was recorded at 2–7 days and 3 months post MI. The EQ-5D-3L is a standardised instrument that is used as a measure of health outcome, which is made up of the following two components. The health utility score is a descriptive system composed of five dimensions: mobility, self-care, usual activities, pain/discomfort and anxiety/depression. Scores for each are combined to give a maximum value of 1. Each dimension has three levels: no problems, some problems and extreme problems. The visual analogue scale reports the patient’s self-rated health from 0 (worst imaginable) to 100 (best health imaginable).

Bleeding and coagulation

Fibrinogen and other parameters of coagulation and haemostasis served as surrogate measures of bleeding risk. 20,21 These parameters were measured in blood samples when site logistics permitted blood sample collection. The sampling time points were at baseline before reperfusion (0 hours) and 2 and 24 hours post reperfusion. The parameters included fibrinogen and plasminogen (both measures of coagulation and systemic fibrinolysis), fibrin D-dimer (a measure of fibrin lysis), tissue plasminogen activator (tPA) (a measure of endogenous tPA and any circulating alteplase) and prothrombin fragment 1 + 2 (from here referred to as prothrombin F1 + 2) (a measure of thrombin activation).

Adverse events

Trial management

The trial was conducted in line with Good Clinical Practice in Clinical Trials24 guidelines and the study complies with the Declaration of Helsinki. 25 There was a trial management group for operational activity, an independent CEC to adjudicate on serious adverse events for safety and efficacy outcomes, an independent Data and Safety Monitoring Committee and a TSC to co-ordinate the trial and liaise with the sponsor and trials unit. Each committee had a charter that was established before enrolment started and an independent chairperson who had relevant expertise. The TSC also included two lay members representing patient and public involvement (PPI). The PPI representatives contributed actively to the meetings. The PPI representatives provided the patient voice when enrolment had been halted after December 2016.

The independent Data and Safety Monitoring Committee met before the enrolment began, and twice again during the active phase of the trial. This committee had responsibility for potentially recommending early discontinuation of the entire study or an individual group because of safety concerns or owing to futility (see Prespecified futility analysis).

The Robertson Centre for Biostatistics within the Glasgow Clinical Trials Unit provided the trial-specific electronic data collection system, acted as an independent co-ordinating centre for randomisation and data management, and conducted the statistical analyses. The trial was approved by the National Research Ethics Service (reference 13/WS/0119). The clinical trial registration number is NCT02257294, and the trial was co-sponsored by the University of Glasgow and Greater Glasgow and Clyde Health Board, NHS Scotland. The sponsors undertook feasibility assessments at each site. The sponsors prospectively monitored the study for safety and monitoring; visits were undertaken in all of the sites. All serious adverse events were prospectively reported to the Pharmacovigilance Unit.

Sample size and statistical methods

The target sample size was 618 patients, which was based on obtaining 186 patients per group with MRI at 2–7 days, allowing for approximately 10% missing data. This was designed to give 90% power at a 5% significance level to detect a difference between two groups of 1.72%, assuming a mean [standard deviation (SD)] of 3.2% (5.1%) for the extent of microvascular obstruction in the comparator group. This calculation was based on the amount of microvascular obstruction demonstrated in the subgroup of patients enrolled into the magnetic resonance in myocardial infarction (MR-MI) cohort study27 who fulfilled the enrolment criteria for the T-TIME trial.

Efficacy analyses were analysed according to randomisation group, that is in relation to randomised treatment allocation regardless of the treatment received. Safety data were analysed in relation to the treatment received. The primary outcome (extent of microvascular obstruction on MRI at 2–7 days post enrolment, as percentage of left ventricular mass) was compared between groups using a stratified Wilcoxon test (van Elteren test), stratified by the location of the MI. The 95% confidence intervals (CIs) for between-group differences in the mean extent of microvascular obstruction were derived by bootstrap resampling (10,000 replicates), stratified by location of MI; percentile CIs are reported. The primary analysis was to compare the 20 mg of alteplase group with the placebo group; if this was significant at a 5% level, the 10 mg of alteplase group would then be compared with the placebo group as a primary analysis. This hierarchical approach was used to preserve the overall type I error rate at 5%. However, if the 20 mg of alteplase group versus the placebo group comparison was not significant, the 10 mg of alteplase group versus the placebo group comparison would be considered as a secondary analysis. Primary and secondary outcomes were also analysed using linear regression (continuous outcomes), logistic regression (binary outcomes) or proportional odds logistic regression (ordinal outcomes). All models were adjusted for the location of the MI. In linear regression models for continuous outcome measures, data were transformed where necessary to improve model residual distributions and were further adjusted for the baseline value of the outcome (where appropriate). For the primary outcome, a post hoc analysis was performed with multiple imputation for the missing outcomes. Regression models were used to assess treatment effects within prespecified subgroups through the use of treatment-by-subgroup interactions (further details are provided in Appendix 3). All tests were two-tailed and assessed at the 5% significance level. Missing secondary outcome data were not imputed. Owing to the potential for type 1 error in the analyses of secondary end points, these end points should be interpreted as exploratory. All statistical analyses were carried out with R studio version 3.2.4 (R Development Core Team 2015, The R Foundation for Statistical Computing, Vienna, Austria) according to a prespecified statistical analysis plan (see Appendix 2).

Prespecified futility analysis

The funder, the EME programme of NIHR, required an interim analysis for futility and specified the criteria before the start of the trial. This analysis was scheduled for when approximately 40% of patients had been randomised and followed up to 3 months. Considering the primary outcome, each active treatment group was compared with the placebo group and if the conditional power for showing a benefit over placebo based on the current trend was less than 30%, then a recommendation would be made to halt that group.

Study intervention

The standard of care procedure and study intervention are illustrated in Appendix 3 and summarised in Table 3. Adjunctive study drug therapy was administered to 435 (98.9%) patients; five patients did not receive any drug (see Figure 3). Two patients (one randomised to the placebo group and one randomised to the 10 mg of alteplase group) received 20 mg of alteplase because an incorrect treatment pack had been selected.

Primary and secondary outcomes

Magnetic resonance imaging was performed in 400 (90.9%) patients at 2–7 days and in 367 (83.4%) patients at 3 months post enrolment. The primary end point was available in 396 patients, which meant that data were missing for the primary end point in 10% of patients. The median time to MRI was 4 days [interquartile range (IQR) 3–6 days]: placebo, median 4 days (IQR 3–5 days); 10 mg of alteplase, median 5 days (IQR 3–6 days); 20 mg of alteplase, median 4 days (IQR 3–6 days); and median 91 days (IQR 86–97 days) for the second MRI scan for the study population. Microvascular obstruction was observed in 176 (44.4%) patients and the amount of microvascular obstruction, expressed as the mean percentage of left ventricular mass, was 2.80%. Clinical case examples are illustrated in Figure 4.

FIGURE 4.

Clinical case examples. Two patients, both with acute anterior STEMI treated successfully with primary PCI within 4 hours of symptom onset. Each patient had thrombolysis in MI (TIMI) grade 3 flow at the end of PCI. MRI was performed at 5 days and 3 days post reperfusion, in a and b respectively. a, Patient with haemorrhagic infarction on MRI. Diagnostic angiogram (top far left image, yellow arrow) revealed an occluded left anterior descending artery with a TIMI 0 flow. T2*-MRI (top far right image) revealed myocardial haemorrhage (white arrow) within the infarct core. Late gadolinium-enhanced MRI revealed microvascular obstruction (top second from right image, red arrow) within the bright area of infarction. The microvascular obstruction within the infarct core spatially corresponded with the myocardial haemorrhage. There was evidence of persistent haemoglobin breakdown products (bottom far right image, white arrow), a reduction in ejection fraction, increased left ventricular end-diastolic volumes and thinned myocardium in the infarct territory on follow-up MRI at 3 months compared to baseline. This represents a case of failed myocardial reperfusion despite successful PCI. b, Patient with an anterior infarct but no MRI evidence of reperfusion injury. Diagnostic angiography (top, far left image, yellow arrow) revealed an occluded left anterior descending artery with a TIMI 0 flow. Late gadolinium-enhanced MRI revealed an anterior infarct (top second from right image, red arrow) with no evidence of microvascular obstruction and no evidence of haemorrhagic transformation on T2*-MRI. There was an improved ejection fraction and a reduction in infarct mass on follow-up MRI at 3 months. This represents a case of successful myocardial reperfusion.

Primary outcome

In the primary analysis, the amount of microvascular obstruction revealed by MRI did not differ between the 20 mg of alteplase group and the placebo group [mean 3.5% vs. 2.3%, estimated difference 1.16%, 95% CI –0.08% to 2.41%; Wilcoxon test (van Elteren test) p = 0.32]. The comparison of the 10 mg of alteplase group with the placebo group then became a secondary outcome [mean 2.6% vs. 2.3%, estimated difference 0.29%, 95% CI –0.76% to 1.35%; Wilcoxon test (van Elteren test) p = 0.74]. Similar results were obtained using a linear regression model, with no evidence of a difference in the primary outcome between all patients who were randomised to the alteplase group and those randomised to the placebo group (mean difference on square-root scale 0.15, 95% CI –0.12 to 0.42; p = 0.28).

Secondary outcomes

The seconday outcomes for efficacy and health outomes for alteplase are shown in Tables 6 and 7. The secondary outcomes for safety are shown in Tables 8 and 9.

Blood chemistry

The AUC for troponin T (ng/ml), which was measured at baseline and at 2 and 24 hours post reperfusion in 317 patients, was increased in both treatment groups compared with placebo (relative difference 1.53, 95% CI 1.16 to 2.01, p = 0.002, for both alteplase groups combined vs. the placebo group) (Figure 5). The AUC for troponin T was 35% higher in patients who were treated with 20 mg of alteplase than those in the placebo group (relative ratio 1.53, 95% CI 1.12 to 2.11; p = 0.008).

FIGURE 5.

Plot of troponin T AUC. The AUC for troponin T (ng/l) was measured from blood samples obtained immediately before reperfusion (0 hours) and then again at 2 and 24 hours. Analysed samples at all three time points were available in 317 patients.

Health-related quality of life

In unadjusted analyses, HRQoL scores were not significantly different between the groups at 3 months. The EQ-5D-3L health utility score was 0.88 in both the 20 mg alteplase group and the placebo group (mean difference –0.002, 95% CI –0.04 to 0.04; p = 0.93) (see Table 6).

Adverse events

Clinical events

The adverse events are described in Table 11. MACE occurred in 15 patients (10.1%) in the placebo, 18 (12.9%) in the 10 mg of alteplase and 12 (8.2%) in the 20 mg of alteplase groups. Two patients experienced a stroke. Aspiration thrombectomy was used in one of these patients, who developed a homonymous hemianopia after the procedure.

Safety

The potential for harm with facilitated PCI was highlighted in the ASSENT-416 and FINESSE17 trials. In ASSENT-4,16 compared with primary PCI (standard care), full-dose tenecteplase combined with PCI was associated with an increase in the primary end point of death, congestive heart failure or shock within 90 days. In those trials, ischaemic cardiac complications and ischaemic stroke were also increased in the intervention group. Despite more initial patency in the infarct-related artery, residual thrombus burden was higher in the facilitated PCI group, and tissue reperfusion and clinical outcomes were worse. 29 These results may be explained by comparatively inadequate anticoagulation and, potentially, the formation of fibrin and thrombus in the group that was treated with tenecteplase. 29,30 The importance of effective anticoagulation to mitigate the prothrombotic effects of fibrinolytic therapy with alteplase has been reported previously. 30

The targeted intracoronary infusion of the study drug was intended to minimise the systemic release of alteplase and minimise bleeding events. Alteplase was selected because it is a fibrin-specific fibrinolytic drug with a brief circulating half-life (≈5 minutes). To further reduce the possibility of harmful remote bleeds, patients with risk factors for bleeding were excluded and the PCI procedures were performed via the radial artery. The rates of bleeding events were within the expected range for primary PCI. 31

Safety assessments were a primary objective of our trial. Circulating fibrinogen concentrations were not different between the groups, indicating that there was no evidence of systemic fibrinolysis or fibrinogenolysis. There were no between-group differences in the haemoglobin concentration at 24 hours or bleeding events. We observed no short-term effect on NT-proBNP, making any compensatory natriuretic effect due to volume overload unlikely. The clinical significance of small increases in left ventricular end-diastolic volume and left ventricular end-systolic volume in the 10 mg of alteplase group at 2–7 days but not at 3 months is uncertain.

The increase in troponin T in the alteplase groups may represent a false-positive result within a series of multiple secondary outcome tests. Alternatively, it may be a true result and, therefore, mechanistically relevant. We think that this scenario is plausible. The AUC for troponin T is distinct from the other measures of infarct size, which were obtained at single time points. An alternative explanation, such as biomarker washout after fibrinolysis, may explain why this rise in troponin T was not associated with an increase in MRI measures of infarction. The dose-related increase in the systemic concentrations of fibrin D-dimer indicates that clot lysis had occurred. An increase in prothrombin F1 + 2 concentrations was observed in the alteplase groups, despite achieving therapeutic anticoagulation with unfractionated heparin. The undesired procoagulant effect of fibrinolytic therapy through thrombin activation30 may have led to microvascular thrombosis, limiting the efficacy of the intervention.

Statistical analysis of the primary outcome

As expected, around half of the participants did not have evidence of microvascular obstruction on the 2- to 7-day MRI scan. This creates zero values and a non-normal distribution in the data set. Different statistical approaches can be considered, each with strengths and limitations. We used a square-root transformation, which may not fully normalise the distribution of the data, back-transformation may be unclear and CIs on a square-root scale may not be easily interpreted. A log-transformation can be back-transformed to produce results as a ratio; however, zero values remain problematic with this approach. A boot-strapping approach is an alternative option.

Overview of current clinical trials

Contemporary practice guidelines call for more research to identify new treatments for microvascular obstruction. 3 There is a widespread, growing interest in the potential efficacy of adjunctive intracoronary fibrinolytic therapy during primary PCI. In our trial, the intervention was scheduled to be carried out at an early stage during primary PCI. The rationale was to infuse intracoronary fibrinolytic therapy upstream of the culprit lesion when thrombus burden was at its greatest and fibrinolytic therapy might be most effective. Intracoronary fibrinolytic therapy was infused before stent implantation to reduce mechanical distal embolisation of a thrombus. An alternative approach proposed by Sezer et al. 15 was to target the administration of adjunctive lytic therapy at the end of primary PCI. At this point, antegrade coronary flow is much improved compared with at the start of the procedure, implying the potential for improved delivery of fibrinolytic therapy to the microcirculation. Based on the new knowledge from the T-TIME trial, future trials should schedule adjunctive intracoronary antithrombotic therapy for the end of the PCI.

The T-TIME trial was designed as a Phase II clinical trial and it was powered to provide definitive information on the primary outcome. The trial was designed to inform the rationale for undertaking a larger Phase III trial that would ultimately be necessary to determine whether or not this strategy could be recommended in practice guidelines. Given that the results of this trial were not positive, there is no case for progressing to a Phase III trial of this intervention, as designed. The intervention should evolve in the light of the new knowledge arising from the T-TIME trial, and any future trial should reflect this.

Three trials are currently investigating this strategy (OPTIMAL, NCT02894138; ACTRN12618000778280; and STRIVE, NCT03335839). Two international, Phase III RCTs are investigating the efficacy of reduced doses of either alteplase or tenecteplase. The Adjunctive, Low-dose tPA in primary PCI for STEMI trial (STRIVE, NCT03335839) will enrol 4000 patients with acute STEMI who present < 6 hours after symptom onset with a large thrombus burden at angiography (TIMI thrombus grade of > 3). The intervention involves open-label intracoronary administration of alteplase (10 mg or 20 mg) or placebo during primary PCI, and the primary composite outcome is MACEs at 180 days. The Restoring Microcirculatory Perfusion in STEMI (RESTORE-MI; ACTRN12618000778280) trial will enrol 1666 patients undergoing primary PCI at < 12 hours after symptom onset. In this double-blind trial, the efficacy and safety of intracoronary administration of tenecteplase (one-third of the weight-adjusted systemic dose) will be compared with placebo in reperfused STEMI patients who have evidence of microvascular dysfunction that is reflected by an index of microcirculatory resistance (IMR) value of > 32 at the end of PCI. The primary outcome is cardiovascular mortality at 24 months. Finally, the Optimal Coronary Flow After PCI for Myocardial Infarction (OPTIMAL) study is a single-centre pilot. The intervention involves the selection of patients with an increased index of microvascular resistance (IMR of > 30) in the infarct-related artery at the end of PCI, who will receive an intracoronary infusion (10 ml) of either 20 mg of alteplase or sodium chloride. The primary outcome is the ratio of myocardial infarct size to area at risk, which is assessed by MRI early after enrolment (days 2–6) to assess the area at risk and again at 3 months to assess the final infarct size. The overall sample size is 90 patients (n = 40 per group), including 10 patients assigned to an observational group with no intervention.

The new knowledge from the T-TIME trial serves as a pause for thought and is relevant to the design of these trials. For example, we have identified a potential risk of intracoronary alteplase to promote myocardial haemorrhage in patients presenting with a prolonged ischaemic time.

Microvascular obstruction persists as an unmet therapeutic need, affecting around half of patients following an acute STEMI. We recommend further research towards therapy development for microvascular obstruction. Novel therapies could be designed to prevent the early occurrence of microvascular obstruction by limiting endothelial damage or by limiting the progression of microvascular obstruction by restoring reflow within the microcirculation. Based on the results of our trial, we support further research involving intracoronary antithrombotic therapy, including fibrinolytic drugs, when targeted only to appropriately selected patient groups who have a low likelihood of developing myocardial haemorrhage. In this regard, the designs of the RESTORE-AMI and STRIVE trials should be reconsidered in the light of the results from the T-TIME trial.

Discontinuation of the trial: the case to continue enrolment

The idea for the T-TIME trial was originally proposed in 2010 and, to date, the case to develop a treatment for failed microvascular reperfusion in patients with acute MI has grown even stronger. Ischaemic heart disease is the number one cause of death and disability in the UK and worldwide (as reported by The Global Burden of Disease Survey). 2 Myocardial infarction is the major cause of death from ischaemic heart disease and the dominant cause of heart failure in developed and developing countries. An increase in the rate of survival from acute MI has led to more patients surviving with persistent heart injury; it is these patients who are at risk of heart failure in the longer term. Despite advances in secondary prevention, the incidence of heart failure has remained persistently high in the past three decades. Failed myocardial reperfusion (or microvascular obstruction) affects half of all-comers with acute MI and has no known treatment, and microvascular obstruction is a driver of heart failure in the longer term. The lack of any available treatment to reduce infarct size in post-MI patients remains an unmet need.

The T-TIME trial was an exploratory Phase II trial that was designed and powered to gather evidence on the safety, efficacy and mechanisms of a potential new treatment approach for patients with acute MI. Given the knowledge gaps for therapy development, the NIHR EME board supported funding of a substantive clinical trial. The T-TIME trial intended to provide definitive results, and, in terms of the primary efficacy outcome, this objective was met. During meetings of the TSC (19 and 27 December 2017), members, including two lay representatives, affirmed that conclusive results, be they ‘positive’ or ‘negative’, would be in the public interest and inconclusive results would be undesirable.

Considering some of the secondary outcomes, premature discontinuation of the trial has the potential to lead to inconclusive results through lack of statistical power. This is not the case for all of the secondary outcomes. For example, the higher AUCs with troponin T in the alteplase groups, compared with placebo, present a clear result, albeit one that was not expected. Moreover, the explanation for this result is not fully understood, and more than one mechanism (or explanation) may be operative. If alteplase enhances microvascular perfusion then the higher AUC for troponin may reflect a ‘wash-out’ phenomenon. On the other hand, alteplase may increase the biochemical size of infarction through enhanced thrombosis. Indeed, both mechanisms may be operative. In the absence of any excess of adverse clinical events, allowing the trial to fully recruit would have conferred more power for the statistical analyses of the secondary, mechanistic outcomes, allowing a conclusive examination of the mechanisms and the effects of alteplase. Furthermore, the futility analysis on the primary outcome presented some issues that may not have been fully considered in the design of the analysis. Because microvascular obstruction is a binary outcome, it would occur in around half the participants, which would substantially reduce the power of an interim analysis. On reflection, we are not aware of any previous trial that has undertaken a futility analysis using microvascular obstruction. This gap increased uncertainty for the oversight committees at a crucial time in the determination of whether or not to discontinue enrolment in the trial. Representing a learning point from this experience, if futility analyses are required for clinical trials involving cardiovascular magnetic resonance imaging in patients following acute myocardial infarction, we advocate adoption of continuous surrogate outcome measures with a normal distribution, such as infarct size, rather than microvascular obstruction.

For the futility analysis, the median value of microvascular obstruction was based on 251 patients, and 124 had a value of zero. The median value was 0.05% of left ventricular mass (0–3.19% of left ventricular mass) and the range was 0.00–28.75% of left ventricular mass. The skewed, as expected, distribution with a wide spread may render a futility analysis based on conditional power highly likely to fail a 30% criterion. A calculation could be performed using a non-parametric test, but it is likely to give similar results to the binary conditional power calculation, given that half of the population have a zero value. Moreover, this scenario could be anticipated in other trials in which microvascular obstruction might be adopted, regardless of what the intervention may be. Therefore, this observation is relevant to investigators involved in the design of future trials that might consider adopting microvascular obstruction as a primary outcome measure. In this trial, the conditional power analysis based on 40% of the trial participants was consistent with the final analysis, including microvascular obstruction and other secondary outcome measurements, such as infarct size. Of note, the investigators had originally proposed infarct size (percentage of left ventricular mass) as the primary outcome in the trial and based on feedback during peer review the outcome measure had changed from infarct size to microvascular obstruction.

To our knowledge, this is the first report of a futility analysis based on microvascular obstruction. There is no other example that could serve as a reference to inform the situation that developed in our trial situation. Given that there were no prior studies to draw on, there was uncertainty relating to the validity of the futility analysis based on microvascular obstruction. In the end, there were no time-varying effects of the intervention and the futility analysis was subsequently proven to be valid.

The T-TIME trial was designed and powered to assess for between-group differences in predefined secondary outcome measures that had been identified as distinct measures of efficacy. These outcomes were measured by a range of methods, including (1) coronary angiography for TIMI flow grade, TIMI blush grade and TIMI thrombus grade, (2) ECG, for example the ST-elevation score at 60 minutes post reperfusion and the per cent ST-segment resolution on the 12-lead ECG at 60 minutes, and (3) biochemistry [troponin T (AUC) and NT-proBNP at 3 months]. These outcomes have comparable prognostic significance but reflect different aspects of cardiovascular function and pathobiology. These outcomes were purposefully selected to provide a broad range of information about efficacy and the evaluation of mechanisms, in line with the mission of the EME programme. Each of the secondary outcomes may be considered to have distinct importance.

Our intention a priori had been to analyse and interpret all of these outcomes to form an overall strategic view on the efficacy (and safety) of the intervention. Although there are inevitable correlations between these efficacy outcomes, they are nonetheless distinct. It is entirely possible that for one outcome there may be no obvious treatment effect on one or more of the secondary outcomes and a distinct effect on another. The statistically significant finding with the AUC for troponin T is a case in point. However, several of the other secondary outcomes exhibited directional trends that were not statistically significant. As an exploratory Phase II trial, the T-TIME trial was designed to have sufficient power to definitively answer questions on efficacy, safety and mechanisms across the full range of secondary outcomes. However, statistical power for the secondary outcome analyses was diminished when enrolment was discontinued with a sample size of 440 patients rather than 618. The Data and Safety Monitoring Committee noted ‘no safety concerns’. Nonetheless, given that the primary question had been answered, the funder and sponsor did not support continuing the study.

The T-TIME trial has provided clinically relevant results. Despite being discontinued prematurely, it is one of the largest Phase II therapeutic trials in acute STEMI. Experts, such as Herson et al. ,19 point out that a futility analysis may lead to premature discontinuation and equivocal results. Because the trial was discontinued with 72% of the population enrolled, there is an inevitable loss of statistical power. This means that prespecified secondary analyses, such as of the treatment interactions with key predefined presenting characteristics, for example the duration of ischaemia from symptom onset to first treatment to achieve coronary reperfusion, have reduced statistical power, limiting the strength of any conclusions. Given that other clinical trials are scheduled, there is community interest in learning as much as possible from the T-TIME trial to optimally inform the design of those other trials. Interestingly, Herson et al. 19 also suggest that different thresholds for conditional power on futility may be adopted. They seem to support a less stringent threshold (e.g. 15–25%) than the 30% level of conditional power that was designated during the peer review process in our trial, pointing out that a futility analysis may lead to premature discontinuation and equivocal results due to lack of statistical power.

Patient and public involvement in the decision to discontinue the trial

The TSC included two patient representatives. One of the representatives highlighted that > 400 patients had accepted a risk by agreeing to participate in the trial. These patients had given a personal commitment to the study; therefore, early termination of the study, leading to inconclusive answers, would not best represent their support. Given that the Data and Safety Monitoring Committee had reported ‘no safety concerns’, the patient representatives on the TSC described premature discontinuation of the trial as being highly undesirable. The patient representatives felt that either ‘negative’ or ‘positive’ results would be of value, whereas equivocal results would be unsatisfactory.

Overall, the discontinuation of enrolment in the T-TIME trial introduced an unexpected situation for the investigators, participants and stakeholders. In some respects, the situation was unprecedented. The experience has provided new knowledge and learning that will be relevant to the wider community involved in the design and conduct of clinical trials in patients with acute MI. There is evidence of good communication, shared timely decision-making, research governance and effective implementation of the decision to place a halt on enrolment without delay following receipt of the letter from the Data and Safety Monitoring Committee. There is also evidence that the trial management process between stakeholders was effective in forming a pragmatic decision, and the safety of the participants was assured.

Aim

The aim of the PPI in the T-TIME trial was to involve the public and service users to ensure that the trial was carried out ‘with’ and/or ‘by’ them.

Methods

The methods used for PPI in the study included involvement in the study design, implementation and conduct of the trial, and reporting the results. The study design involved informal and formal feedback from patients and the public. Informal feedback was gained during 2010–12 when the original idea for the intervention was conceived. The chief investigator (CB) sought the views of patients in his care during and after treatment for a heart attack (MI). Following favourable informal feedback, formal input was sought from members of the public who have designated roles as PPI representatives in the Golden Jubilee National Hospital, including Mr Gordon Baird and Mrs Sandra Pairman. PPI feedback was also provided through the ethics committee reviews of the research idea, trial design, patient information sheet and consent form. The idea for the current trial had been preceded by an earlier proposal involving tenecteplase, a third-generation lytic drug, rather than alteplase, a second-generation lytic drug. Our current trial was preceded by an earlier grant from the NIHR EME programme (10/90/12): ‘A randomised parallel group double blind placebo-controlled Trial of very low dose and low dose adjunctive Tenecteplase during priMary PCI’. The award was withdrawn for a number of reasons, for example delays in research and development (R&D) set-up. The design of that trial was developed with the involvement of Ms Brenda Rankin, who was a lay member of the R&D Steering Group in the Golden Jubilee National Hospital and a member of the NHS Board’s Quality Reference Group and Lay Research Group. The T-TIME trial was presented for review at PPI meetings outside the lead site. PPI review took place in Leeds General Infirmary and the Liverpool Heart and Chest Hospital. These meetings involved presentation of the trial to the local PPI group by the local principal investigator (Professor John Greenwood) or research lead (Professor Rod Stables). These PPI presentations formed an integral part of the site review process leading up to the approval of a new study in these hospitals. Therefore, the T-TIME trial has benefited from PPI feedback in multiple forums over an extended period of time.

The conduct of the T-TIME trial was overseen by a TSC. The responsibility of the TSC was to supervise the trial and ensure that it was conducted according to the Medical Research Council Guidelines for Good Clinical Practice. The TSC had two PPI representatives, Mr Gordon Baird and Mrs Sandra Pairman. Mr Baird is a former patient, based in Stranraer, and Mrs Pairman has a designated role as a lay adviser in the Golden Jubilee National Hospital. They participated in the TSC meetings, including one before the start of the trial and twice-yearly during the course of the trial. The meetings had an agenda and minutes. Finally, these PPI representatives received the plain English summary of the report and provided feedback.

During the past 3 years, the T-TIME trial has been presented at a number of PPI events hosted by the sponsor in Glasgow, the lead site in Clydebank and in Leeds.

Study results

The overall results of the study are reported in the Plain English summary. The T-TIME trial benefited from PPI input before, during and after the end of the trial. In general, service users recognised that blocked microvessels in the distribution of the ‘culprit’ heart artery represented a problem of unmet need. They supported the rationale to dissolve the microthrombi using low doses of the ‘clot buster’, alteplase, while placing patient safety as a paramount consideration. They positively contributed to optimise the plain English summary content of the patient information sheet and consent forms. Their feedback on these documents was included in the final, approved versions that were used in the trial. PPI representatives positively contributed to the TSC meetings. In December 2016, following a recommendation from the chairperson of the independent Data Monitoring Committee (IDMC), it was advised that the trial be discontinued based on preliminary evidence of futility. In an urgently convened meeting of the TSC held on 28 December 2017, the PPI members strongly advocated that the trial should continue. The IDMC had noted that there were no safety concerns, and on this basis the PPI representatives voiced concern that, by discontinuing the trial prematurely, all of the trial’s objectives may not be met, reducing the value of the trial’s results. In this regard, they were particularly concerned that by discontinuing the trial the contribution of those patients who had already been enrolled (n = 440) would be undermined. The sponsor took account of this feedback while enrolment was temporarily suspended. Nonetheless, EME programme management required that the recommendation of the IDMC be implemented; therefore, the trial was discontinued.

The PPI representatives and service users contributed to PPI meetings that focused on the trial. Dates of PPI meetings include 24 March 2017 and 25 April 2019 (both in the Queen Elizabeth University Hospital, Glasgow) and 20 May 2019, for the SofTMech Patient Participation Day in the Golden Jubilee National Hospital (URL: www.softmech.org/events/headline_596271_en.html; accessed 29 June 2020).

Discussion and conclusions

In the SoftMech PPI event that took place in the Golden Jubilee National Hospital (20 May 2019), in the light of the T-TIME trial’s neutral results that were presented by Professor Colin Berry, the service user representatives very much supported the case for more research into therapy development for microvascular obstruction. They also supported the role of mathematics and biostatistics for computational heart modelling to better understand the nature of heart injury, to create computational models to predict the effects of novel therapies and, finally, to enhance the efficiency of randomised trials, for example through efficient trial design to reduce the sample size of clinical trials, wherever possible.

On reflection, PPI very much influenced the study design across a broad range of areas. PPI brought added value to the trial. The PPI representatives reviewed the Plain English summary of the trial and provided feedback on the text. There were no negative effects of PPI input.

Reflections/critical perspective

The PPI input enhanced the case for support and helped make the case for funding stronger. Prior PPI input may have facilitated the ethics review, which also benefited from lay/service user appraisal. PPI input enhanced the discussions during the TSC meetings and positively supported the trial when enrolment was suspended. PPI user feedback continues to advocate for more research into microvascular obstruction, which affects half of all patients affected by STEMI (heart attack) and, in the absence of any effective therapy, represents an unmet therapeutic need.

Role of sponsor statement

The University of Glasgow and Greater Glasgow and Clyde Health Board were independent co-sponsors of the trial. The co-sponsors shared oversight and responsibility for the design and conduct of the study, data collection and management, research governance, analysis, and final approval of the manuscript. The sponsor did not have the right to preclude submission of the manuscript.

Access to data and data analysis

Colin Berry, who is the principal investigator for the trial, had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Peter J McCartney MRCP; British Heart Foundation Glasgow Cardiovascular Research Centre, University of Glasgow, Glasgow, UK; salaried, no personal compensation.

Hany Eteiba FRCP; West of Scotland Heart and Lung Centre, Golden Jubilee National Hospital, Clydebank, UK; salaried, no personal compensation.

Annette M Maznyczka MRCP; British Heart Foundation Glasgow Cardiovascular Research Centre, University of Glasgow, Glasgow, UK; salaried, no personal compensation.

Margaret McEntegart FRCP; West of Scotland Heart and Lung Centre, Golden Jubilee National Hospital, Clydebank, UK; salaried, no personal compensation.

John P Greenwood FRCP; Leeds University and Leeds Teaching Hospitals NHS Trust, Leeds, UK; no personal compensation; salaried, no personal compensation.

Douglas F Muir FRCP; The James Cook University Hospital, South Tees Hospitals NHS Foundation Trust, Middlesbrough, UK; salaried, no personal compensation.

Saqib Chowdhary FRCP; Wythenshawe Hospital, Manchester University NHS Foundation Trust, Manchester, UK; salaried, no personal compensation.

Anthony H Gershlick FRCP; University of Leicester, Leicester Biomedical Research Centre, University Hospitals of Leicester NHS Trust, Leicester, UK; salaried, no personal compensation.

Clare Appleby FRCP; Liverpool Heart and Chest Hospital NHS Foundation Trust, Liverpool, UK; salaried, no personal compensation.

James M Cotton FRCP; Wolverhampton University Hospital NHS Trust, Wolverhampton, UK; salaried, no personal compensation.

Andrew Wragg FRCP; Barts Health NHS Trust, London, UK; salaried, no personal compensation.

Nick Curzen FRCP; University Hospital Southampton NHS Foundation Trust, Southampton, UK; salaried, no personal compensation.

Keith G Oldroyd MD; West of Scotland Heart and Lung Centre, Golden Jubilee National Hospital, Clydebank, UK; salaried, no personal compensation.

Mitchell Lindsay FRCP; West of Scotland Heart and Lung Centre, Golden Jubilee National Hospital, Clydebank, UK; salaried, no personal compensation.

J Paul Rocchiccioli FRCP; West of Scotland Heart and Lung Centre, Golden Jubilee National Hospital, Clydebank, UK; salaried, no personal compensation.

Aadil Shaukat FRCP; West of Scotland Heart and Lung Centre, Golden Jubilee National Hospital, Clydebank, UK; salaried, no personal compensation.

Richard Good FRCP; West of Scotland Heart and Lung Centre, Golden Jubilee National Hospital, Clydebank, UK; salaried, no personal compensation.

Stuart Watkins FRCP; West of Scotland Heart and Lung Centre, Golden Jubilee National Hospital, Clydebank, UK; salaried, no personal compensation.

Keith Robertson FRCP; West of Scotland Heart and Lung Centre, Golden Jubilee National Hospital, Clydebank, UK; salaried, no personal compensation.

Christopher Malkin FRCP; Leeds University and Leeds Teaching Hospitals NHS Trust, Leeds, UK; salaried, no personal compensation.

Lynn Martin BN; West of Scotland Heart and Lung Centre, Golden Jubilee National Hospital, Clydebank, UK; salaried, no personal compensation.

Lynsey Gillespie PhD; Greater Glasgow and Clyde Health Board, Glasgow, UK; salaried, no personal compensation.

Thomas J Ford FRACP; British Heart Foundation Glasgow Cardiovascular Research Centre, University of Glasgow, Glasgow, UK; salaried, no personal compensation.

Mark C Petrie FRCP; British Heart Foundation Glasgow Cardiovascular Research Centre, University of Glasgow, Glasgow, UK; salaried, no personal compensation.

Peter W Macfarlane DSc; Electrocardiology Core Laboratory, University of Glasgow, Glasgow, UK; unsalaried (honorary staff), no personal compensation.

R Campbell Tait FRCPath; Department of Haematology, Royal Infirmary, Glasgow, UK; salaried, no personal compensation.

Paul Welsh PhD; British Heart Foundation Glasgow Cardiovascular Research Centre, University of Glasgow, Glasgow, UK; salaried, no personal compensation.

Naveed Sattar FMedSci; British Heart Foundation Glasgow Cardiovascular Research Centre, University of Glasgow, Glasgow, UK; salaried, no personal compensation.

Robin A Weir MD; University Hospital Hairmyres, East Kilbride, UK; salaried, no personal compensation.

Keith A Fox FRCP; University of Edinburgh, Edinburgh, UK; salaried, no personal compensation.

Ian Ford PhD; Robertson Centre for Biostatistics, Institute of Health and Wellbeing, University of Glasgow, Glasgow, UK; salaried, no personal compensation.

Alex McConnachie PhD; Robertson Centre for Biostatistics, Glasgow Clinical Trials Unit, University of Glasgow, Glasgow, UK; salaried, no personal compensation.

Colin Berry FRCP; British Heart Foundation Glasgow Cardiovascular Research Centre, University of Glasgow, Glasgow, UK; salaried, no personal compensation.

Contributions of authors

Peter J McCartney (https://orcid.org/0000-0001-7482-1811) (Cardiology Specialty Trainee) contributed to screening and enrolment, collected the data, undertook data analyses, interpreted the data and contributed to the manuscript.

Hany Eteiba (Consultant Cardiologist) was a local principal investigator and obtained informed consent, randomised patients, collected data, interpreted the results and contributed to the manuscript.

Annette M Maznyczka (https://orcid.org/0000-0002-0118-5153) (Cardiology Specialty Trainee) contributed to screening and enrolment, collected the data, undertook data analyses, interpreted the data and contributed to the manuscript.

Margaret McEntegart (https://orcid.org/0000-0001-5135-3322) (Consultant Cardiologist) obtained informed consent, randomised patients, collected data, interpreted the results and contributed to the manuscript.

John P Greenwood (https://orcid.org/0000-0002-2861-0914) (Consultant Cardiologist) was a local principal investigator and obtained informed consent, randomised patients, collected data, interpreted the results and contributed to the manuscript.

Douglas F Muir (https://orcid.org/0000-0003-1871-2869) (Consultant Cardiologist) was a local principal investigator and obtained informed consent, randomised patients, collected data, interpreted the results and contributed to the manuscript.

Saqib Chowdhary (Consultant Cardiologist) was a local principal investigator and obtained informed consent, randomised patients, collected data, interpreted the results and contributed to the manuscript.

Anthony H Gershlick (https://orcid.org/0000-0002-8044-9048) (Consultant Cardiologist) was a local principal investigator and obtained informed consent, randomised patients, collected data, interpreted the results and contributed to the manuscript.

Clare Appleby (https://orcid.org/0000-0002-3838-1174) (Consultant Cardiologist) was a local principal investigator and obtained informed consent, randomised patients, collected data, interpreted the results and contributed to the manuscript.

James M Cotton (https://orcid.org/0000-0001-6548-2723) (Consultant Cardiologist) was a local principal investigator and obtained informed consent, randomised patients, collected data, interpreted the results and contributed to the manuscript.

Andrew Wragg (https://orcid.org/0000-0003-0067-1771) (Consultant Cardiologist) was a local principal investigator and obtained informed consent, randomised patients, collected data, interpreted the results and contributed to the manuscript.

Nick Curzen (https://orcid.org/0000-0001-9651-7829) (Consultant Cardiologist) was a local principal investigator and obtained informed consent, randomised patients, collected data, interpreted the results and contributed to the manuscript.

Keith G Oldroyd (https://orcid.org/0000-0002-7842-3463) (Consultant Cardiologist) obtained informed consent, randomised patients, collected data, interpreted the results and contributed to the manuscript.

Mitchell Lindsay (https://orcid.org/0000-0002-4654-712X) (Consultant Cardiologist) obtained informed consent, randomised patients, collected data, interpreted the results and contributed to the manuscript.

J Paul Rocchiccioli (https://orcid.org/0000-0003-1350-1544) (Consultant Cardiologist) obtained informed consent, randomised patients, collected data, interpreted the results and contributed to the manuscript.

Aadil Shaukat (Consultant Cardiologist) obtained informed consent, randomised patients, collected data, interpreted the results and contributed to the manuscript.

Richard Good (https://orcid.org/0000-0002-3906-1025) (Consultant Cardiologist) obtained informed consent, randomised patients, collected data, interpreted the results and contributed to the manuscript.

Stuart Watkins (https://orcid.org/0000-0003-1579-6294) (Consultant Cardiologist) obtained informed consent, randomised patients, collected data, interpreted the results and contributed to the manuscript.

Keith Robertson (https://orcid.org/0000-0002-6557-8298) (Consultant Cardiologist) obtained informed consent, randomised patients, collected data, interpreted the results and contributed to the manuscript.

Christopher Malkin (https://orcid.org/0000-0002-6643-6830) (Consultant Cardiologist) obtained informed consent, randomised patients, collected data, interpreted the results and contributed to the manuscript.

Lynn Martin (Research Nurse) supported screening and enrolment, data collection and event reporting.

Lynsey Gillespie was project manager for the trial.

Thomas J Ford (https://orcid.org/0000-0003-4009-6652) (Research Fellow) was a co-ordinator for the Clinical Event Committee.

Mark C Petrie (https://orcid.org/0000-0002-6333-9496) (Consultant Cardiologist) was a co-ordinator for the Clinical Event Committee.

Peter W Macfarlane (https://orcid.org/0000-0002-5390-1596) directed and participated in the ECG core laboratory analyses.

R Campbell Tait (https://orcid.org/0000-0003-4842-6783) directed the haematology core laboratory analyses.

Paul Welsh (https://orcid.org/0000-0002-7970-3643) directed the biochemical analyses and contributed to interpretation and to the manuscript.

Naveed Sattar (https://orcid.org/0000-0002-1604-2593) directed the biochemical analyses and contributed to interpretation and to the manuscript.

Robin A Weir (https://orcid.org/0000-0001-9895-2795) (Consultant Cardiologist) chaired the Clinical Event Committee.

Keith A Fox (https://orcid.org/0000-0002-0140-2752) (Emeritus Professor of Cardiology) chaired the Trial Steering Committee and contributed to the interpretation of the data and the manuscript.

Ian Ford (https://orcid.org/0000-0001-5927-1823) (Emeritus Professor of Statistics) contributed to the study design, data analyses and interpretation, and contributed to the manuscript.

Alex McConnachie (https://orcid.org/0000-0002-7262-7000) (Biostatistician) analysed and interpreted the data, and contributed to the manuscript.

Colin Berry (https://orcid.org/0000-0002-4547-8636) (Consultant Cardiologist) conceived the study, obtained the funding, recruited and randomised patients, collected and assessed data blind to treatment group assignment, interpreted the results and wrote the manuscript.

Colin Berry is chief investigator and Saqib Chowdhary, Raj Das, Ian Ford, John P Greenwood, Peter W Macfarlane, Douglas Muir, Keith G Oldroyd, Naveed Sattar and Campbell Tait were co-applicants on the grant from the NIHR EME programme (reference 12/170/45).

Publications

McCartney PJ, Eteiba H, Maznyczka AM, McEntegart M, Greenwood JP, Muir DF, et al. Effect of low-dose intracoronary alteplase during primary percutaneous coronary intervention on microvascular obstruction in patients with acute myocardial infarction: a randomised clinical trial. JAMA 2019;321:56–68.

Berry C, McCartney PJ, Maznyczka AM, T-TIME Group. Fibrinolytic therapy to reduce microvascular obstruction after myocardial infarction-reply. JAMA 2019;321:2033–34.

Berry C, Maznyczka AM, McCartney P. Failed myocardial reperfusion during primary PCI: an unmet therapeutic need. EuroIntervention 2019;14:1628–30.

Data-sharing statement

All data requests should be submitted to the corresponding author for consideration. Access to anonymised data may be granted following review.

Patient data

This work uses data provided by patients and collected by the NHS as part of their care and support. Using patient data is vital to improve health and care for everyone. There is huge potential to make better use of information from people’s patient records, to understand more about disease, develop new treatments, monitor safety, and plan NHS services. Patient data should be kept safe and secure, to protect everyone’s privacy, and it’s important that there are safeguards to make sure that it is stored and used responsibly. Everyone should be able to find out about how patient data are used. #datasaveslives You can find out more about the background to this citation here: https://understandingpatientdata.org.uk/data-citation.

Disclaimers

This report presents independent research. The views and opinions expressed by authors in this publication are those of the authors and do not necessarily reflect those of the NHS, the NIHR, the MRC, NETSCC, the EME programme or the Department of Health and Social Care. If there are verbatim quotations included in this publication the views and opinions expressed by the interviewees are those of the interviewees and do not necessarily reflect those of the authors, those of the NHS, the NIHR, NETSCC, the EME programme or the Department of Health and Social Care.

Amendment 1 (prior to regulatory green light)

Date: 18 July 2014.

Protocol v2.0 and associated documents (short patient information sheet v2.0, patient information sheet v2.0, patient alert card).

Amendment 2 (prior to regulatory green light)

Date: 6 October 2015.

Protocol v3.2 and associated documents (short patient information sheet v3.0, full patient information sheet v3.0, EQ5D questionnaire, patient alert card v2.0).

Amendment 3 (prior to regulatory green light)

Date: 21 January 2016.

Changes to participating sites.

Amendment 4 (prior to regulatory green light)

Date: 16 February 2016.

Revised general practitioner letter (v2.0) and clinical information sheet (v2.0).

Amendment 5

Date: 5 July 2016.

Protocol v4.0 and associated documents (short patient information sheet v3.1, full patient information sheet v3.1). Changes to participating sites.

Amendment 6

Date: 31 August 2016.

Addition of participating site.

Amendment 7

Date: 8 November 2016.

Addition of participating sites.

Amendment 8

Date: 13 December 2016.

Protocol v5.0 and associated documents (short patient information sheet v4.0, full patient information sheet v4.0). Following medical advice from the MHRA (Medicines and Healthcare products Regulatory Agency), the amendment was submitted as an urgent safety measure.

Amendment 9

Date: 22 February 2017.

Addition of participating site.

Amendment 10

Date: 21 March 2017.

Addition of participating site.

Amendment 11

Date: 27 July 2017.

Protocol v6.0 and associated documents (short patient information sheet v5.0, full patient information sheet v5.0, general practitioner letter v3.0, clinical information sheet v3.0).

Amendment 12

Date: 27 July 2017.

Addition of participating site.

Amendment 13

Withdrawn on advice of Research Ethics Committee: removal of participating site.

Amendment 14

Date: 21 December 2017.

Temporary halt to recruitment.

Amendment 15

Date: 30 March 2018.

Protocol v7.0: clarification of the designation of outcome variables as secondary or exploratory.

Secondary outcomes

Cardiac magnetic resonance acquisition and analysis

We used cardiac MRI to assess left ventricular dimensions, function and pathology 2–7 days and 3 months post MI. MRI was performed using 1.5-T platforms [MAGNETOM^® Avanto (Siemens Healthineers AG, Erlangen Germany) and Philips Intera, Best, the Netherlands]. The imaging protocol followed a standard operating procedure that included planning and localisers, T1-mapping, T2*-mapping, cine MRI with steady-state free precession (SSFP) and late-gadolinium enhancement imaging 10–15 minutes after administration of contrast medium. The scan acquisitions were spatially co-registered and included different slice orientations to enhance diagnostic confidence. The cardiac magnetic resonance imaging protocol is shown in Figure 6.

FIGURE 6.

Flow diagram of the cardiac MRI protocol. FLASH, fast low-angle shot; GRA, gradient recalled echo; i.v. intravenous; LAx, long axis; LV, left ventricular; MOLLI, modified Look–Locker imaging; SAx, short axis.

Cardiac magnetic resonance imaging at 3 months

Cardiac MRI at 3-month follow-up involved the same scan protocol as was used in the baseline MRI scan. Renal function was known or checked. A full blood count was checked to provide haematocrit for the extracellular volume analysis from the post-contrast T1 map.

Magnetic resonance imaging analysis

The MRI analyses were undertaken using Medis^® Suite MR (Medis, Leiden, the Netherlands), which is vendor-independent post-processing software. Peter J McCartney undertook the primary analyses of the scans and related analyses and they were reviewed by Colin Berry (second observer). The research staff were blinded to treatment allocation.

Coronary angiogram acquisition and analyses

The acquisition and analyses of the coronary angiograms were performed in accordance with the standard operating procedures. Coronary angiograms were acquired during emergency care with cardiac catheter laboratory X-ray and information technology equipment. Feedback was provided to sites on the quality and completeness of the angiograms.

The angiograms were analysed by trained observers (AM, CB and MMcE) using imaging post-processing software (QAngio^® XA Medis, Leiden, the Netherlands). The observers were blinded to treatment assignment. Catheter calibration was performed using the catheter calibration function on QAngio. For each lesion, a view perpendicular to the long axis of the vessel was used to avoid foreshortening and overlap of branches. Therefore, the single-plane projection showing the best opacified and most severe lesion with minimal foreshortening and minimal branch overlap was selected.

The TIMI coronary flow grade and thrombus grade were assessed following initial angiography and at the end of the procedure. TIMI myocardial perfusion grade and TIMI frame count were assessed at the end of the procedure. The TIMI frame count and perfusion grade are angiographic measures of microvascular function.

TIMI coronary flow grade

The TIMI coronary flow grade of the infarct-related artery is independently predictive of prognosis.

TIMI myocardial perfusion grade

The TIMI myocardial perfusion grade provides a score for ground-glass appearance ('blush') of contrast entering the microvasculature and contrast washout, it is predictive of prognosis. 8

Corrected TIMI frame count

The TIMI frame count is an objective continuous variable index of coronary blood flow, representing the amount of time (in frames) for contrast dye to reach a standardised distal landmark, corrected for vessel length (corrected TIMI frame count, normal < 27 frames).

Electrocardiogram acquisition and analysis

Twelve-lead ECG was carried out before coronary reperfusion and 60 minutes afterwards, and again at 3 months. The ECGs were analysed in the University of Glasgow ECG core laboratory, Glasgow, UK.

Safety

Parameters of haemostasis and coagulation, including fibrinogen concentration, plasminogen activity, fibrin D-dimer and prothrombin F1 + 2, served as a surrogate measure of bleeding risk and safety. tPA antigen levels were assessed as a potential measure of the systemic overflow of the alteplase administered into the culprit coronary artery. Haemostasis and coagulation parameters were measured in blood samples, when site logistics permitted. The sampling time points were 0, 2 and 24 hours post reperfusion.

A depletion of fibrinogen and plasminogen following thrombolysis correlates with systemic fibrinolysis and may correlate with bleeding risk. Fibrin D-dimer concentration represents a specific measure of fibrinolysis. Fibrin D-dimer concentrations may correlate with the amount of clot lysis and may therefore represent a measure of residual clot burden. Fibrin D-dimer concentrations have the potential to correlate with efficacy and outcome. Prothrombin F1 + 2 is a measure of thrombin activation and correlates with the (undesired) procoagulant effect of thrombolysis. Prothrombin F1 + 2 is depressed by anticoagulants administered before and during PCI. Prothrombin F1 + 2 concentrations may be associated with the dose of alteplase and placebo, and potentially could correlate with adverse thrombotic events.