Removal of glucose testing as a safety assessment at each timepoint. Glucose assessment was unnecessary and was included in error in the previous version of the protocol. Inclusion of two additional timepoints for cGMP assessment (at 4 weeks and 3 months in addition to baseline, 9 and 12 month timepoints). On reviewing the results of the PARADIGM study (LCZ696 vs enalapril), the maximum change in cGMP was detected at approximately 4 weeks with the levels plateauing out at subsequent timepoints Replacement of the Mini-Mental State Examination (brief version) with the Standardised MMSE as the cognitive assessment tool. This is a more reliable cognitive assessment tool with less variability and variance than the brief version The ‘history of malignancy’ exclusion criteria was amended to exclude only those with active malignancy or other diseases that would compromise life expectancy. The larger PARADIGM study which randomised 8,442 patients to LCZ696 or enalapril did not exclude patients with a history of malignancy unless their life expectancy was compromised. Further clarification was given on how to measure waist and hip ratio to avoid any variability in this measurement

The inclusion criteria was amended to allow subjects with lower natriuretic peptide levels to be included in the study. People with BNP between 35 - 280pg/ml were now eligible for inclusion into the study (the previous cut-off was >50pg/ml) People with NT-proBNP between >125 - 1,000 pg/ml were now eligible for inclusion into the study (the previous cut-off was >250pg/ml)

Updates to the Reference Safety Information for LCZ696 (Investigators Brochure) were submitted to the HPRA. Changes to the RSI are always considered substantial.

Inclusion criteria amended to include subjects with lower natriuretic peptide levels. People with BNP 20-280pg/ml are now eligible (previous cut-off >35pg/ml). People with NT-proBNP 100-1,000 pg/ml are now eligible (previous cut-off >125pg/ml). Rationale: The aim of the study is to evaluate the effectiveness of LCZ696 in an asymptomatic population at risk of developing heart failure. Elevated levels of NP indicate a heightened risk of heart failure and other cardiovascular events. The results of a prospective study of 3,346 patients without heart failure, with a mean follow-up time of 5.2 years showed an excess risk at BNP levels of 20 pg/ml for men and 23.3 pg/ml for women (Wang, Larson et al. 2004) which is well below current thresholds used to diagnose heart failure. A sub-analysis of the STOP-HF data which included 429 patients with uncomplicated hypertension, showed the optimal threshold for BNP to predict major adverse cardiovascular events and death was 20pg/ml. (Heartbeat Trust, Data on File). Therefore, consistent with an increased risk of developing HF at BNP levels as low as 20pg/ml we would like to propose a lower NP threshold level. The risk thresholds for NT-proBNP are less well-defined. However, based on a correlation between 100 consecutive patients with simultaneous BNP and NT-proBNP measurements, a NT-proBNP level of 100pg/ml is comparable to a BNP level of 20pg/ml (Heartbeat Trust, Data on File). Exclusion criteria amendment. Previously, people with a “history of asymptomatic left ventricular systolic dysfunction defined as LVEF <50%, at any time” were excluded. This criteria was amended to only exclude patients with an LVEF <50% on the most recent measurement. Rationale: The target population is one at increased risk of heart failure which will include subjects with a history of coronary events. A transient drop in LVEF is not unexpected during and shortly after such cardiac events which would exclude patients under the previous criteria.

The study objectives were amended as follows: “To provide further information in the target population on the safety and tolerability of LCZ696 versus valsartan” was changed from an exploratory objective to a secondary objective. Rationale: The Investigators wished to emphasise the importance of safety by making it a secondary objective. The manner in which safety was assessed during the study was not changed in any way. The timelines for performing cardiac MRI were amended and the 9 month timepoint removed such that this procedure was performed at baseline and end-of-study only. An additional site was added for carrying out MRIs. Rationale: The Investigators had access to a limited number of MRI slots for PARABLE subjects. A second site was opened to increase availability of MRI slots. It is unlikely that any structural changes in the heart would be evident after 9 months of study drug. The need for gadolinium containing contrast during the MRI was removed from the protocol. Rationale: The use of gadolinium was included in the initial protocol in error. Gadolinium is not required for the measurement of left atrial volume index (LAVI). There are risks associated with the use of gadolinium including minor side-effects (nausea, headaches, paraesthesia, hypotension) and rarely anaphylaxis and nephrogenic systemic fibrosis.

Two secondary objectives added: 1) To assess response based on genetic variants of the NPPB, NPRA and NPRC genes; 2) Change in the incidence of progression of left ventricular dysfunction. Rationale: Genetic variant rs198389 of Nppb increases BNP and this is cardioprotective in at-risk patients. It is important to know if any benefits of LCZ696 are independent of genetic variants of key genes involved in the expression and metabolism of BNP. PARABLE is examining progression of left ventricular diastolic dysfunction. Endpoints include echo and cardiac MRI measures of diastolic dysfunction. An additional composite describing numbers of patients with progression of diastolic dysfunction is added. Two exploratory objectives added: 1) To assess the difference between groups in terms of atrial fibrillation (AF); 2) To assess response of immune cells to LCZ696 Rationale: As AF is influenced by atrial fibrosis, inflammation and hypertrophy, it is important to investigate relationships between AF, LAVI, natriuretic peptide and LCZ696. A non-invasive cardiac rhythm monitor will be placed on a sub-set of up to 60 patients per group for up to 6 days. Monocytosis is present in patients with left ventricular diastolic dysfunction, and monocyte derived macrophages are present in dysfunctional and fibrosed cardiac tissue. BNP can attenuate monocyte chemotaxis. We will examine the relative impact of LCZ696 on monocyte gene expression using next generation sequencing. Exclusion criterion amended. Hepatic dysfunction was defined as any LFT exceeding 3x the upper limit of normal (ULN). This was amended to exclude only with raised AST or ALT. Rationale: GGT is not an exclusion criterion; it is not specific in the diagnosis of liver disease. The statistical methods section was described in more detail. Additional information on sample size included following a blinded review of 9-month echo data/effect size of 125 subjects. However, the sample size remains at 250.

Changes to the site for QP release of the study drug. At study set-up, the study drug was released by Novartis in their London manufacturing facilities. In preparation for Brexit, an additional site for QP release was identified in the Republic of Ireland. Almac Clinical Services Limited continued to be the distributor but the IMP was transferred from the Craigavon site (Northern Ireland) to the Co Louth facility before 29 March 2019. Future batches of IMP from Novartis, UK, were shipped directly to this Almac site and the QP release was be performed there.

A nine month follow-up sub-study was added to the protocol. The objective was to evaluate if any of the treatment effects observed during the 18-month study persisted nine-months after study drug discontinuation or whether there was disease regression. This sub-study involved a single visit to the Investigator Site at month 27 (9 months after study drug discontinuation). The assessments and procedures carried out during this follow-up visit were identical to those carried out at baseline, 9 and 18 months (physical exam, measurement of BP, heart rate, height, weight, waist and hip, blood and urine for biomarkers, 24-hour ABPM, echocardiography, ECG and administration of questionnaires; 5-7 day holter monitoring was optional and cardiac MRI was not done).

Based on research conducted and published by the investigator team, and evidence that intervention can modulate the risk for progression of atrial cardiomyopathy and LAVI, further blood biomarkers of inflammation, fibrosis, inflammation, metabolism, platelet function, thrombosis and coagulation were added to the protocol. The primary endpoint of the study is LAVI, which is an important component of atrial cardiomyopathy. Furthermore, research by the Investigator team suggests that the intervention may favourably modulate the progression of other aspects of atrial cardiomyopathy such as electrophysiological measures and fibrosis, as well ALVDD. Therefore, additional secondary endpoints were added to the protocol to assess the impact of LCZ696 dependent on atrial cardiomyopathy at baseline as well as progression of atrial cardiomyopathy. Final changes, clarifications and additions were made to the secondary and exploratory endpoints. This acknowledges that the study is looking at a range of measures of atrial structure and function, as well as ventricular structure and function. In particular, this reflects new cMRI assessment techniques available. These prespecified endpoints maximise the value of the data in the context of the overall research aims and objectives.

Amendments relate to the secondary and exploratory objectives and endpoints of the study outlined in Section 10.1 and 10.2 of the study protocol.