Angina: the impact of baseline symptoms on the effect of PCI, relationships between ischaemia tests, design of a trial of PCI for stable angina, and development of a symptom app

Abstract

Current assessment and treatment of angina is based on substantial, but primarily observational, data for the mechanisms of coronary artery obstruction, ischaemia and chest pain, coupled with the sincere desire to reduce myocardial infarctions and deaths. The belief that percutaneous coronary intervention (PCI) provides event reduction and angina relief has been questioned following surprising results from randomised controlled trials. This thesis addresses why unblinded trials find that PCI improves symptoms in stable angina but a blinded study, the Objective Randomised Blinded Investigation with optimal medical Therapy of Angioplasty in stable angina (ORBITA), did not find an exercise time benefit. I examine the link between symptoms, anatomy, and ischaemia which appears to be more complex than previously assumed. Historical symptom descriptions, e.g. by epidemiologist Geoffrey Rose, gave rise to the concept of typical and atypical angina. This classification is intended to reflect the likelihood of the symptoms being due to coronary artery obstruction and, although not the original intention, to infer how likely revascularisation is to relieve the symptoms. In Chapter 3, I test whether the nature of symptoms predicts the placebo-controlled benefit of PCI. Invasive coronary pressure indices such as Fractional Flow Reserve (FFR) were developed to determine the clinical significance of a coronary artery stenosis and are now considered gold standard for assessing coronary obstruction. FFR was originally mapped against multiple ischaemia tests treated dichotomously, in 45 patients. In Chapter 4, I examine the associations between the different ischaemic tests in ORBITA, treated dichotomously or continuously. I also assess the ability of anatomical severity, measured by quantitative coronary angiography (QCA), to predict the placebo-controlled benefit of PCI, using the ORBITA dataset. It has been suggested that features of the ORBITA trial design contributed to the lack of treatment effect observed with PCI. ORBITA participants also felt that the trial design could be improved. They recommended using symptoms rather than exercise time as the primary endpoint. In Chapter 5 I describe my work on the design of the ongoing ORBITA-2 trial which addresses these design features. I incorporate daily documentation of symptoms on a smartphone application and a novel ordinal clinical outcome scale for angina as the primary endpoint, both developed in partnership with patients and experienced statisticians to provide a more accurate and patient-centred measure of health status in angina. In Chapter 6 I present the development and validation of the symptom smartphone app using data from ORBITA-2 participants. This is the first study of a smartphone app for monitoring angina in a clinical trial. I assess the ability of ORBITA-2 participants in the Completion Assessment Group to complete the app, demonstrating feasibility, and the ability of ORBITA-2 participants in the Recall Assessment Group to recall numbers of episodes, showing the advantage of a daily documentation approach using the app. This exploration of symptoms, anatomy and ischaemia using the ORBITA dataset, and experience with patients during the design and conduct of ORBITA-2, provides insight into ways management and research of angina could be improved.