Amendment 1: Left ventricular ejection fraction (LVEF) <50% was removed from exclusion criterion (8). The purpose of exclusion criterion 8 was to eliminate known/confirmed heart failure. The inclusion of an EF lower limit to specify which subjects should be included did not address this aim. Ejection fraction defines different heart failure phenotypes (e.g. heart failure with preserved, reduced or mid-range ejection fraction) and as such was not relevant to the aim of excluding those with a heart failure diagnosis as a whole. Even if the purpose was to exclude only subjects with reduced EF, the EF of 50% corresponds to ‘normal’ EF when echocardiography is performed but does not correspond to the appropriate value to demark normal and abnormal ejection fractions for other modalities such as SPECT. • Clarification of follow-up period for the study. • Clarification of medicinal products in the study and update of storage and handling conditions for Flurpiridaz (18F) Injection. • Corrections to text to ensure recording of concurrent medications to study completion. • Removal of requirement for drug and alcohol screening. Drug and alcohol abuse screening is useful to address: (1) unique safety concerns associated with potential interactions of IMP with illicit drugs, (2) concerns regarding confounding of an efficacy signal, and (3) concerns that follow-up would be compromised given occult substance abuse. GE Healthcare determined that these concerns were minimal, given: (1) there were no unique concerns regarding drug/illicit drug interactions, particularly given that Flurpiridaz is administered to subjects in one sitting (as opposed to repeatedly) at a tracer dose;

Amendment 1 (Continued): (2) concerns of confounding the correlation between the anatomical gold standard of the QCA and the PET MPI blinded reads were minimal. It was possible that toxic effects of illicit drugs (such as cocaine) could affect the microvascular function that in turn could lead to perfusion abnormalities as seen on MPI in the absence of significant epicardial coronary stenoses evidenced on QCA. GE Healthcare believed that these discrepancies were likely to be minimal and encountered infrequently, and (3) subjects were already being screened for psychiatric conditions which could impair participation in all study visits (exclusion #12). Substance abuse is an axis II disorder and investigators were counselled to exclude subjects with ongoing drug abuse that may have led to poor compliance in the manual of procedures. Given the short-term follow up of this study, it was unlikely that occult substance abuse (missed as part of the medical history) would frequently impair subject follow up. Since the Sponsor believed that active substance abuse was likely to be rare and to have minimal if any effect on efficacy, no effect on safety, and minimal if any effect on follow-up compliance, the collection of this sensitive health information was not justified. • Guidance regarding use of beta blocker therapy was added. To ensure that the PET and SPECT MPI were conducted according to guidelines and in accordance with standard clinical care, study sites were advised to withhold the beta-blocker when possible for at least 24 hours prior to the stress test. Evidence demonstrates that beta-blockade at the time of stress testing may reduce the sensitivity of MPI. GE Healthcare acknowledged that withholding the beta-blocker might not always be possible due to clinical concerns such as difficult to control hypertension or arrhythmia (as per the guidelines).

Amendment 1 (Continued): • Clarification that an additional blood sample at screening could be analysed by the local lab to determine if the subject met exclusion criteria. Dependence on central lab results for screening purposes would result in a delay from screening to the earliest performance of in study visits (including the PET or on-study SPECT) of at least 48 hours from the time of the lab draw (in most cases). This delay added a significant hurdle to subject recruitment and retainment in a study where all study visits must occur prior to a prescheduled invasive coronary angiography. In most cases the window between screening and the clinical intracoronary angiography was expected to be less than 7 days. Permitting screening through the use of local labs permitted subject to be screened, enrolled and have a SPECT scan in a minimal amount of time (or even in the same day) with a PET scan following as closely after as doses are available. Central labs results could still be used for screening if local labs were not drawn. The determination of whether to draw local labs for this purpose rested with the investigator and depended on the rapidity with which the study visits occurred (e.g., if the window between screening and ICA was brief, local labs were advised, if this window was more lengthy than local labs were not drawn). Significant discordance between local and central lab results was unlikely since the central and local labs would be drawn in the same sitting at screening. If they did occur the local lab would take precedence for the screening purposes. The safety data set was to use exclusively the central labs to analyse changes in biochemistry, since all patients would systematically have labs analysed centrally at screening and pre- and post-scan timepoints. • Safety reporting for AEs and SAEs was updated and clarified.

Amendment 02: • Clarification of time points for recording vital signs. • Clarification in text that urine would be collected at pre-treatment time points only. • Medical Director details were updated following a change in personnel.

Amendment 03: • Editorial correction to clarify that rest and stress SPECT MPI procedures can take place on 2 separate days that do not have to be consecutive. • Correction of typographical error in the description of the semi-quantitative read (exploratory analysis).