Aortic stenosis is the most common valvular disease in the Western world. It is caused by progressive narrowing of the aortic valve leading to increased strain on the heart muscle which has to work increasingly hard to pump blood through the narrowed valve. Over time the heart muscle thickens to generate more force, but eventually the heart fails leading to death if the valve is not replaced with an operation. No medical treatments exist to stop or reverse the heart valve narrowing. Current clinical guidelines suggest that an operation should be performed only when symptoms develop or the heart muscle is visibly weak on cardiac ultrasound scanning. However, symptoms can be difficult to interpret and in many patients the heart muscle has become irreversibly damaged and the heart fails to recover following surgery.

Using MRI scans of the heart, the investigators have identified heart scarring which seems to develop as the heart muscle thickens. Several studies now show that people who have developed this scarring are more likely to suffer poor outcomes including death. The investigators have also identified clinical risks that predict the presence of scarring.

The investigators propose a study where patients with severe aortic stenosis but no indications for valve replacement as per current guidelines are assessed for those clinical risks. If a participant's risk of having scarring is higher they will undergo a cardiac MRI scan. If scarring is present participants will be randomised to routine clinical care, or referral for valve replacement surgery. Participants with no evidence of scarring will be randomised routine care with study follow or not. The investigators of this study hypothesize that early surgery will lead to fewer complications and reduced risk of death compared to standard care.

Inclusion Criteria:

1. Severe aortic stenosis (aortic valve jet velocity ≥4.0 m/s, or aortic valve area indexed to body surface area <0.6cm2/m2 with aortic jet velocity ≥3.5m/s)
2. Age over 18 years
3. No symptoms attributable to aortic stenosis that require aortic valve replacement

Exclusion Criteria:

1. Deemed lower risk for mid-wall fibrosis on screening
2. Planned cardiac surgery
3. Previous valve replacement
4. Severe hypertension (systolic >180 or diastolic >110 mmHg)
5. Acute pulmonary oedema or cardiogenic shock
6. Left ventricular ejection fraction <50% on cardiac MRI
7. Significant abnormalities on cardiac MRI that would prevent enrolment
8. Coexistent severe aortic regurgitation or mitral regurgitation
9. Coexistent mitral stenosis greater than mild in severity
10. Coexistent hypertrophic cardiomyopathy or cardiac amyloidosis
11. Any contraindication to MRI scanning (such as permanent pacemaker)
12. Advanced renal impairment (glomerular filtration rate <30 mL/min/1.73 m2)
13. Pregnancy or breast feeding
14. Patient judged to be unfit to be considered for aortic valve replacement or transcatheter aortic valve implantation
15. Patient declines to consider undergoing valve replacement surgery or transcatheter aortic valve implantation
16. Inability to give informed consent
17. Previous randomisation into this study

It is planned that an anonymised raw dataset will be shared under a controlled access model. It will be available following primary publication. Requests will be made in accordance with Edinburgh Clinical Trials Unit policy at the time of release.

The result of the cardiac MRI will be blinded to care provider, participant and investigator. Group A will be unblinded, as participants can only enter group A if mid-wall fibrosis is present. Groups B and C will appear identical and the groups combined so the presence of mid-wall fibrosis will be blinded. Group D will be unblinded, as participants can only enter group D if mid-wall fibrosis is not present.

Outcome assessors will be blinded to allocation in all groups where outcome is adjudicated (groups A, B and C)