Trial of Cardiac CT in Acute Chest Patients With Intermediate Level Initial High-sensitivity Cardiac Troponin

Overview

Patients who present to the emergency department (ED) with acute chest pain (ACP) possibly due to Coronary artery disease (CAD), with a normal heart tracing (ECG), need to have further troponin blood tests to confirm or exclude a heart attack. After initial troponin testing, a significant 50-85% of patients are said to be in an "observational zone" as one cannot confirm or exclude a diagnosis of a heart attack. Even after repeat blood testing, 22-33% remain in this "observational zone". These patients can be challenging to manage as they are not safe to be discharged home, but they also cannot be treated as a heart attack. This contributes to ED overcrowding and uncertainty in treatment plans.

1\. Background Coronary artery disease (CAD) remains the most common cause of mortality in the world according to the World Health Organisation (WHO). Chest pain accounts for a significant healthcare burden representing approximately 700,000 annual visits to the emergency department in England and Wales \[1\]. Patients with acute chest pain (ACP) of possible cardiac origin account for approximately 17% of all emergency department (ED) consultations, but less than 10% of these are eventually diagnosed with acute myocardial infarction (AMI). Hence a means of evaluating these patients in the ED in an efficient manner, whilst ensuring high sensitivity and specificity, is of paramount importance. Cardiac biomarkers e.g. cardiac Troponin (cTn) I or T along with electrocardiogram (ECG) remain the cornerstone in the evaluation of patients with suspected acute coronary syndrome (ACS). 1.1 Performance of high-sensitivity cardiac troponins High-sensitivity cardiac troponin (hs-cTn) assays enable the measurement of cTn at concentrations not detected with the former generation conventional cTn assays. In September 2015 hs-cTn assays were adopted in the European Society of Cardiology (ESC) guidelines for the management of patients with acute coronary syndrome (ACS) without persistent ST elevation. The proposed algorithms advocate either a single hs-cTn at ED presentation or repeat measurements after 1 or 2 hours thus enabling a more rapid "rule-in" and "rule out" of AMI compared with conventional cTn assays. The cut off values for the different hs-cTn assays are assay specific \[2\]. The performance of these algorithms (involving hs-cTn) has been evaluated in multiple studies. A prospective multicentre study by Gimenez et al looked at ruling out AMI using undetectable levels of hs-cTn (I and T) at presentation. With hs-cTnT, AMI was ruled out in 26.5% of cases with a negative predictive value (NPV) of 98.6%. Among three different hs-cTnI assays which were studied, the NPV ranged from 98.8% to 100%. No patient with undetectable levels of hs-cTnT died during the first 30 days and only 0.4% had died (2 patients not due to AMI) at 24 months' follow-up. Among the three hsc-TnI assays, mortality at 24 months ranged from 0 to 2.4% with only one death due to AMI (which occurred in the first 30 days) \[3\]. Although the more rapid risk-stratification with these algorithms (on the first sample of hs-cTn) helps in reduced time to rule-in or rule-out AMI there remains, however, (between the initial "rule-in" and "rule-out" categories) an intermediate "observational zone" category of patients, who do require a serial troponin test at 1 hour for further risk-stratification. Recent pilot data by Marjot et al have shown that, after initial hs-cTnT testing on presentation, there are a significant proportion of patients (54%) who would require further troponin testing after 1 hour as they were stratified in the observational zone on the initial troponin test. Despite the mandated repeated troponin at 1 hour, Marjot et al also showed that in real world practice, the mean time to repeat troponin was still 2.9 hours and that after training and implementation of the algorithm for 3 months, over 65% of patients still had their troponin taken at least 90 minutes after the first \[4\]. Similarly, in a sub-study of the ROMICAT II trial, Ferencik et al also found that a substantial 86.9% of patients had intermediate hs-cTn levels on initial testing and the addition of a second or third hs-cTn level did not improve risk stratification \[5\]. A study involving hs-cTnI (in a 2 hour algorithm) by Lindahl et al showed that, 47.1% remained in the observational zone on initial troponin. After a repeat troponin 2 hours later, 25.5% of patients, remained in the observational zone \[6\]. This presents an opportunity for a possible alternative means of further evaluating the initial observational zone cohort of patients in a more efficient manner. Furthermore, multiple studies also showed that, even after repeat serial 1-hour troponin, a significant proportion of patients remains in the intermediate "observational zone" category and this is associated with mortality and adverse cardiac event risks. Suggestions for further management of these patients is not standardised and is guided by possible further repeat troponin and/or invasive or non-invasive cardiac imaging \[2\]. In a prospective international multicentre trial, Mueller et al found that 22.5% of patients remained in the observational zone and cumulative mortality for this cohort was 0.7% at 30 days but increased substantially to 9.6% at 365 days \[7\]. Similarly, Reichlin et al found that 24.1% of patients were found to be in the observational zone. In this cohort, the prevalence of acute MI was 18.6% and the cumulative mortality was 1.6% at 30 days, rising to to 16.5% at 2 years follow-up \[8\]. Mokhtari et al evaluated major adverse cardiac events (MACE) at 30 days in an observational study where the 1-hour hs-cTnT algorithm supplemented by patient history and ECG ("extended algorithm") was compared with an algorithm using hs-cTnT alone (troponin algorithm). Despite the addition of patient history and ECG, the proportion of patients remaining in the observation zone was not significantly different between the two algorithms and was found to be of the order of 25-27%. In the extended algorithm the 30 day MACE event rate including unstable angina was 10.1% in the observational zone cohort \[9\]. A study by Jaeger et al showed that 33% of patients still remained in the observational zone and that the cumulative mortality was found to be 0.6% at 30 days and 3.55 at 360 days \[10\]. These studies aptly demonstrate the presence of a significant cohort of patients who remain in the observational zone despite second troponin testing. Thus, again there is a pressing need to clarify risk stratification and further clinical management to reduce the proportion of patients in the observational zone. 1.2 Possible role for computed tomography coronary angiography The use of computed tomography coronary angiography (CTCA) in patients with acute chest pain has been shown to be safe \[11\], with high sensitivity and negative predictive value for coronary artery disease \[12-15\] and cost-effective with decreased time to diagnosis and earlier discharge from the ED \[11, 16\]. The finding of coronary artery disease on CTCA has been shown to predict prognosis, with significantly worse MACE for patients with \>50% stenoses, compared to those with \<50% \[17-19\]. Historically 10% of patients with clinical Non-ST-elevation myocardial infarction (NSTEMI) on conventional troponin analysis were found to have unobstructed (\<50% stenosis) coronary arteries on invasive coronary angiography. Subsequently it has been shown that approximately 10% of these have actual evidence of subendocardial infarction when investigated on late-gadolinium enhanced cardiac MRI (CMR) \[20\]. Therefore, in the era of high-sensitivity troponin assays, the percentage of patients with actual evidence of infarction would be even less in the ruled in group. Hence in our proposed research clinical pathway (detailed in section Experimental details and design of proposed investigation), the investigators have selected the conservative value of \<25% stenosis to rule out AMI. Very few studies have examined the possible role of CTCA in the era of hs-cTn's. A sub-study of the ROMICAT II trial by Ferencik et al, showed that CTCA, with advanced plaque assessment and hs-cTn, significantly decreased the proportion of patients who had been classified in the intermediate category on initial hs-cTn from 43.8% to 24.4% when compared with conventional slow release troponin and traditional CTCA assessment (based on luminal stenosis alone). However, significant drawbacks of the study included the observational design and the unlimited time for CT interpretation. \[5\]. The prospective randomised BEACON study compared the use of CTCA in addition to hs-cTn with a conventional management strategy involving hs-cTn alone. The authors concluded that the CTCA supplemented strategy did not meet the primary endpoint of identifying more patients with significant CAD requiring revascularisation. The use of CTCA also did not shorten hospital stay nor allow for more direct discharge from the ED, despite 42% of patients having no identifiable CAD and serial troponin testing being carried out at 3-6 hours. The main benefits of CTCA included significantly lower direct medical costs and less out-patient testing. However, duration of hospital stay was not the primary endpoint and the exclusion criteria did not include a specified lower limit for hs-cTn for ruling out AMI and only patients with no coronary disease were deemed to be suitable for A\&E discharge \[21\]. Given the drawbacks of these studies, there is a compelling need to compare the performance of a management strategy involving hs-cTn supplemented by CTCA in a direct prospective randomised fashion, with usual standard of care involving serial hs-cTn alone, in the cohort of acute chest pain patients deemed to be in the intermediate observational zone according to the initial hs-cTn result. 2\. Study Objectives and Design 2.1 Aim of the study In patients with ACP requiring serial hs-cTn testing, to perform a head-to-head comparison of a management strategy involving serial hs-cTn supplemented by CTCA versus the conventional standard of care management guided by serial second hs-cTn alone in a randomised prospective trial. To the best of the author's knowledge this study will provide the first prospective and randomised data pertaining to hospital length of stay as a primary outcome in the use of CTCA on this ACP cohort (with an intermediate observational zone category) on initial hs-cTn results presenting to the ED in a tertiary hospital (see Study 1 below). It will also provide further data on the influence of traditional and more advanced CTCA diagnostics (e.g. traditional CT stenosis assessment, advanced plaque characterisation) in clinical decision making, incremental to that provided by hs-cTn based care alone in patients with acute chest pain (see Study 2 below). 2.2 Original hypothesis The use of CTCA will lead to improvements in hospital length of stay and risk stratification and clinical management of patients in the intermediate/observational zone category on initial hs-cTn when compared with standard of care involving serial hs-cTn alone. 2.3 Experimental Details and Design of the Proposed Investigation The proposed work is divided into two clinical studies: Study 1: Prospective, randomised single-centre trial to compare hospital length of stay, patient clinical management and outcomes between standard of care supplemented by CTCA versus standard of care alone, in ACP patients found to be in the intermediate observational zone category on initial hs-cTn in an acute hospital setting. The times for recruitment will be from 8am to 4pm, Mondays to Fridays (inclusive). If recruited, the patients will be randomised to either (Arm A): undergo early CTCA along with a serial second hs-cTnT; or (Arm B): undergo standard of care involving serial hs-cTnT alone. Patients in both arms will be consented to have CTCA. However, Arm B (standard of care arm) will be blinded from CTCA findings and will have standard of care based clinical management according to serial hs-cTn. The CTCA data in Arm B will be used for Study 2 (see below). Arm A: CTCA assessment will be carried out in Arm A while the patient would normally be waiting to have their repeat serial hs-cTn taken or waiting for the blood test result. Patients with \<25% stenosis on CTCA will have AMI ruled-out and may be considered for discharge or alternative reasons for their clinical presentation may be investigated. The results of these CTCA scans will be made available to the patients' clinical care team and further management decisions will be left to their discretion. Arm B: Patients in this arm will be managed according to standard of care, which includes serial hs-cTn testing. These patients will also have CTCA carried out, but the CTCA assessment will not form part of the patients' clinical management as clinicians will be blinded to CTCA findings. Furthermore, unlike Arm A, CTCA image interpretation will not take place in the acute hospital setting. It will be carried out in the following days. The whole CTCA procedure will be carried out after the patient has had their hs-cTn taken and while the patient is waiting for their hs-cTn result (as the investigators would not like the CTCA to delay the staff members from taking the blood test). Study 2: Sub-analysis of CTCA + biomarkers arm (Arm B): Analysis of the CTCA data-sets will be carried out including luminal stenosis, and plaque characterisation. Thereafter this information will be revealed to a select group of clinicians, who will be asked (in a virtual setting) to comment on possible changes to their original (hs-cTn based) clinical management plans in the light of the information gleaned from the existing CTCA datasets. From existing CTCA stenosis data, the cardiology clinicians will be asked to comment on any changes to their clinical management plan if they were given the following information: (i) hs-cTn + CTCA stenosis (ii) CTCA stenosis + CT plaque characterisation (iii) A combination of the above. The investigators will also make a comparison between these virtual plans of action and actual course of action among patients who undergo invasive coronary angiogram +/- invasive FFR assessment as part of their routine care. CT-Plaque Characterisation AMI results from sudden coronary luminal thrombosis, which can occur from any of three underlying pathological lesions: plaque rupture, plaque erosion and calcified nodules. Plaque rupture represents most of the underlying pathologies for AMI and the precursor coronary lesion is known as thin capped fibroatheroma (TCFA). These tend to be composed of a large lipid-rich necrotic core, thin and intact fibrous cap, spotty calcium, inflammation due to infiltration by macrophages and some smooth muscle cells \[24\]. High risk morphological features of TCFA that can be identified on CTCA plaque assessment include: (a) napkin ring sign (b) positive remodelling (c) spotty calcification and (d) low attenuation. 2.6 Study Statistics The investigators will first inspect the normality of the distribution of the outcome variable. If it appears as though the distribution is not normal as expected, then the investigators will use non-parametric tests (Mann-Whitney U test); differences between groups will be constructed using 10,000 bootstrap simulations on the difference in medians, and derive associated confidence intervals and p-values. If there are any significant imbalances in any covariate(s) between the two groups, then the investigators will also perform quantile (specifically median) regression analysis on the difference between median length of stay between the two groups, adjusted for these covariate(s). To estimate the standard errors for the difference in medians, 10,000 bootstrap simulations of this quantile regression will be performed. If there are any significant imbalances, then the adjusted analysis will be considered the main analysis, otherwise the univariate analysis will be taken to be the main analysis. If the data are Normally distributed then standard regression techniques will be used. 2.7 Cost and Economic Analysis The cost and cost-effectiveness analyses will assess whether the addition of CTCA within the ED setting to the conventional clinical pathway without acute imaging will produce any changes in terms of total costs and/or cost-effectiveness analyses. For the purposes of the secondary objectives of cost analyses and economic evaluations (consistent with secondary outcomes) quality of life and symptoms will be measured using the EQ-5D-5L questionnaire at baseline after the ED episode and then monthly for the first three months and three monthly thereafter. All relevant costs from an NHS and Personal Services perspective will be considered using a top-down costing strategy (consistent with GSTFT finance data). Cost-effectiveness will be estimated in terms of the incremental cost per quality-adjusted life year (QALY) of comparing both clinical pathways (with and without the use of CTCA in acute setting). This ratio will be calculated using the area under the curve for health utility using the EQ-5D-5L and health service costs up to one year. Sensitivity analyses will explore the potential impact of major adverse events upon lifetime costs and QALYs as well as the adoption of a societal perspective. Existing published models will constitute the base for long-term modelling of both clinical pathways. Lifetime QALYs and costs of surviving patients will be estimated from published sources of life expectancy, annual costs and corresponding annual utilities. It is hypothesised that patients in whom coronary artery disease is identified, will adhere better to strategies that include primary and secondary prevention. This means that the early use of CTCA might hold benefits in the short-term) as well in the medium and long-term. 2.8 Timeline In our internal audit, patients who presented with acute chest pain to the ED at GSTFT, and were found to be in an intermediate grey zone on initial troponin analysis and who required a second troponin amounted to 26 patients per week. As the investigators aim to recruit patients Monday to Friday from 8am to 4 pm, it is anticipated that it may be possible to recruit 5 patients per week in total. Given the target of 250 patients in the study, recruitment is likely to take approximately 52 weeks to achieve. Follow up Procedures Patients will be followed up at 1, 2, 3, 6, 9 and 12 months following the hospital visit, to capture all relevant costs and outcomes. It is estimated that 30% of participants enrolled in the study may be lost to follow-up. Data collection General and study specific data Data will be collected by the research team from routinely collected NHS records and will include several categories, such as: baseline demographics, co-morbidities, ECG results, admission and discharge diagnoses, cardiology and other relevant investigations or interventions, repeat hospitalisations and adverse events. Patient questionnaires Patients will receive a call (at 1, 2, 3, 6, 9 and 12 months) to collect key information around resource use, patient quality of life and patient satisfaction. During this contact, participants will be asked: i) about the presence and degree of pain/discomfort, their level of concern and patient satisfaction using a scale of 0-10; ii) EQ-5D-5L questionnaire, with five questions; iii) NHS resource use, due to the participant's ACP episode; and iv) the presence of any Major Adverse Cardiac Events (MACE). This information above will be required to achieve secondary objectives of the study. 3\. CTCA Procedures and findings 3.1 CTCA Procedure The CTCA exam will be performed on a new generation a multi-detector dual-source CT scanner. The investigators will try to ensure optimum CTCA images by attempting to minimise coronary and chest wall motion artefact through reducing heart rate to below 63 beats per minute (bpm) and by getting the patients to hold their breath for 10 - 12 seconds. If the heart rate is above 63 bpm and the systolic blood pressure (BP) is above 100 mmHg, intravenous beta-blockers) will be given to achieve the target heart rate. To further optimise coronary images, sublingual glyceryl trinitrate (GTN) will be given if the systolic BP is above 90 mmHg. Pre-CTCA renal function will be available from routine bloods samples that are taken as part of standard of care work-up of patients presenting to the ED with acute chest pain. Female patients of potential child bearing age will be screened for the possibility of pregnancy according to the local Guy's and St. Thomas' Radiology protocols. 3.2 CTCA Image Interpretation and Reporting The CTCA will be interpreted and reported by an experienced Radiologist or Cardiologist with a minimum of Level II certification in cardiac CT angiography. Angiograms will be reported using the standard 15 segment model \[22\]. A stenosis will be graded in severity according to the following classification \[23\]: 1. Minimal: 0-24% 2. Mild: 25-49% 3. Moderate: 50-70% 4. Severe: \>70% 5. Total Occlusion: 100% As discussed previously, patients with \<25% stenosis will have AMI ruled-out. 3.3 CTCA Results For Arm A, CTCA image interpretation and reporting will be carried out as early as possible in the acute hospital setting, while the patient is an in-patient. The results will be made available to the patients' clinical care team and further management decisions will be left to their discretion. Patients in Arm B will also have CTCA carried out but the CTCA assessment will not form part of the patients' clinical management as this arm will be blinded to CTCA findings. Furthermore, unlike Arm A, CTCA interpretation followed by reporting will not take place in the acute hospital setting. Should the CTCA be found to have significant high risk CAD e.g. \>50% stenosis in the left main (LM) coronary artery, and/or \>50% stenosis in the proximal left anterior descending (LAD) coronary artery, they will be un-blinded and kept in a separate registry. Their results will be discussed with the hospital care team and if required, an urgent cardiology out-patient referral will be made to enable further clinical management. 3.4 CTCA Incidental Findings Pooled studies show: (i) an incidental extra-cardiac finding in 44% of patients undergoing CTCA; and (ii) the diagnosis of a major finding in 16% of the CTCA exams \[25\]. Incidental findings on CTCA will be documented in the CTCA report. In the case of Arm A (Study 1), the clinical care team will be made aware of the finding through the report. In Arm B any clinically significant findings e.g. cancers and/or prognostically significant coronary artery disease will be notified as a Radiology alert to the clinical care team and the patient's general practitioner. 4\. Sample Size, Selection and Withdrawal of Subjects 4.1 Sample Size Waiting times often exhibit a skewed distribution, and so the sample size calculation was based on the difference in median waiting time. To estimate the sample size needed to observe a one-hour reduction in median hospital length of stay, normal techniques based on standard deviation estimates are not valid. The investigators therefore used a random sample of 49 patients undergoing the current pathway as the control 'population', and created an equivalent treatment 'population' by multiplying the waiting times of the 49 sampled patients by a constant such that the median was reduced by one hour; this constant was found to be 0.799. For a given sample size n, 10,000 Monte Carlo simulations were performed by sampling n patients with replacement from each of the two groups, and the p-value from a Mann-Whitney U test was calculated for each simulation. The proportion of these 10,000 simulations with a p-value below 0.05 was recorded as the power for that sample size n. The sample size was varied until a power of 0.8 was obtained, and was found to be 250 patients in total (125 patients in each arm of the study). Patient drop-out is anticipated to be minimal as all patients, by definition of the primary outcome, will be in hospital for the length of their hospital stay, and therefore their length of stay will be recorded. 5\. Study Procedures 5.1 Screening Procedures Patients with suspected ACS eligible for the study will enter GSTFT via the ED at St Thomas' Hospital. If the initial ECG shows no ischaemic changes and the initial hs-cTn result cannot rule in or rule out AMI, the patient will be identified as a potential recruit to the study by the clinical care team. A screening log will be maintained by the site and kept in the Investigator Site File. This will record all potentially eligible patients approached about the study and the reasons why they were not registered in the study if this is the case. 5.2 Consenting Participants Once a potential participant is identified by the clinical team, and if the patient meets the inclusion criteria (and none of the exclusion criteria), a trained member of research team will obtain signed informed consent from the patient. This procedure will be supported by a patient information sheet that appropriately explains the aims, methods, anticipated benefits and potential hazards of the study. It is anticipated that the consent process will take no longer than 15 minutes. It will be explained to the patient that it is his/her right to ask to be withdrawn from the study at any point in time. Written informed consent on the current approved version of the consent form for the study will be obtained before any study-specific procedures are conducted, and a copy will be given to the patient and kept in the patient's medical notes. The discussion and consent process will be documented in the patient notes. The patient's capacity will be assessed by trained and delegated clinical/research staff who have completed study specific training and have been delegated this responsibility by the Principal Investigator (PI). Research staff are responsible for: * Assessing the patient's capacity to provide informed consent. * Checking that the current approved version of the information sheet and consent form are used. * Checking that information on the consent form is complete and legible and the patient has completed/initialled all relevant sections and signed and dated the form. * Checking that an appropriate member of staff has countersigned and dated the consent form to confirm that they provided information to the patient. * Checking that an appropriate member of staff has made dated entries in the patient's medical notes relating to the informed consent process (i.e. information given, consent signed etc.). * Following registration: * Adding the patient study number to all copies of the consent form, which should be filed in the patient's medical notes and investigator site file. * Giving the patient a copy of their signed consent form and patient information sheet. * Respecting the right of the patient to refuse to participate in the study without giving reason as all patients are free to withdraw at any time. 5.3 Randomisation Procedures Once patients consent to participate in the study, they will be randomised into the intervention group (i.e. with CTCA) or the control group (i.e. hs-cTn based standard of care) on a 1:1 ratio. Randomisation will be carried out via the use of opaque sealed envelope block randomisation method. Both the block randomisation list and the sealed envelopes will be produced by the statistician. Each block will contain 5 envelopes, which would translate to 50 blocks. 25 blocks will contain 3 envelopes for Arm A and 2 envelopes for Arm B. The remaining 25 blocks will contain 3 envelopes for Arm B and 2 envelopes for Arm A. Each block will also be randomly arranged. The sealed opaque envelopes/blocks used to assign patients to either arm will be prepared by an individual external to the study. The recruiter will not be able to identify which arm a potential participant is going to be randomised to until after he/she has received informed signed consent from the potential participant. Once randomised onto the study, the patient will be given a study number. This will be documented in the enrolment log. 5.4 Radiation assessment The proposed study includes only one scan per patient covering approximately the cardiac anatomy from the aortic root/pulmonary arteries level down to below the inferior border of the heart. The CT scan protocol will use standard prospective ECG gating. If necessary intravenous beta-blockers will be given to achieve the target heart rate (below 63 beats per minute). If for clinical reasons (e.g. elevated heart rate despite the use of oral and/or intravenous beta-blockers) it is not possible to use the prospective CTCA scanning protocol, the patient will not undergo the CTCA examination and therefore will not take part in the study. The typical Dose-Length Product (DLP) for the prospective CTCA scan is expected to be in the region of 350 mGy.cm. This leads to an estimate cardiac CT dose of approximately 9mSv per scan. To provide an upper estimate of dose that is not expected to be exceeded, the 95th percentile plus 20% is taken. This gives a DLP of 640mGycm which corresponds to an effective dose of 16mSv. These effective doses estimates are based on a previous dose (DLP) audit of 30 CTCA scans performed on adult patients between 02/03/2017 and 20/03/2017 on the Siemens Force scanner at St Thomas' Hospital using the 'CaScore Turbo Flash' or the 'Coronaries Prospective' protocol. The total research protocol dose (TRPD) for this study is set at 16mSv all of which is due to the CTCA imaging research procedure, i.e. additional to standard care. It is noted that some patients involved in this study may go on to have other procedures involving radiation as part of their standard care. These are difficult to fully characterise and are not included in this dose calculation. The total risk expressed as the total detriment (cancer incidence weighted for lethality and life impairment and the probability over two succeeding generations of severe hereditary disease) is estimated at approximately 1 in 1100 for the full TRPD dose. The TRPD value corresponds to approximately 7 years of natural background radiation exposure, where the average natural background radiation per capita is 2.2 mSv in the UK. 6\. Assessment of Safety A serious adverse event is any untoward medical occurrence that: * Results in death; * Is life-threatening; * Requires in-patient hospitalisation or prolongation of existing hospitalisation; * Results in persistent or significant disability/incapacity; or * Other important medical events. 6.1 Ethics reporting A serious adverse event (SAE) occurring to participant would be reported to the REC that gave a favourable opinion of the study where in the opinion of the Chief Investigator the event was: 'related' - that is, it resulted from administration of any of the research procedures; and 'unexpected' - that is, the type of event is not listed in the protocol as an expected occurrence. Reports of related and unexpected SAEs would be submitted within 15 days of the Chief Investigator becoming aware of the event, using the NRES report of serious adverse event form. All related AEs that result in a patient's withdrawal from the study or are present at the end of the study, should be followed up until a satisfactory resolution occurs. A patient may also voluntarily withdraw from treatment due to what he, or she, perceives as an intolerable AE. If either of these occurs, the patient would undergo an end of study assessment and be given appropriate care under medical supervision until symptoms cease or the condition becomes stable. 7\. Study Steering Committee The Study Steering Group will meet at fixed points during the study and will include patient representatives. Data Monitoring and Ethics Committee (DMEC) functions will be embedded in the Study Steering Committee. The Study Steering Committee will have access to unblinded comparative data. The committee will monitor data collection methods and make recommendations regarding whether there are any ethical or safety reasons why the study should not continue.