Abstract
Background:There is an imperative to improve risk stratification and timing of intervention among aortic stenosis patients. Previous research demonstrated transient left ventricular dysfunction (TLVD) after aortic valve intervention (surgical or transcatheter aortic valve implantation (TAVI)). The definition of TLVD and its clinical relevance are not established.
Objectives and Methods:
I aimed to test the hypothesis TLVD peri-TAVI is a true phenomenon and that it influences patients’ outcome. To minimise confounders, I opted for a load-independent invasive technique (the pressure-volume loop (PVL) studies) to assess the left ventricular (LV) function before and after aortic valve intervention; I obtained baseline measurements and then studied the force-frequency relationships (FFR) with incremental pacing (contractility reserve assessment). I also studied these severe aortic stenosis (AS) patients with serial echocardiography and serial blood biomarkers.
To inform the above discussion, I first delineated the LV function in severe AS with preserved ejection fraction (EF) and compared the AS group to a group of controls (Ctrl) and a group of heart failure (HF) patients using the invasive PVL studies. After that, to examine the primary hypothesis, I performed and compared the PVL studies before and after TAVI among the AS patients.
Considering the invasive nature of these PVL studies, I validated the methods of non-invasive measurement of end-systolic pressure-volume relationship (ESPVR; also known as a single-beat estimate of Ees(SB)), correlating this to the invasive PVL ESPVR in severe aortic stenosis.
I finally sought to answer the question of whether PVL indices could predict the short-term clinical course post-intervention. I correlated the invasive and non-invasive indices of LV contractility to blood biomarkers and echocardiographic measures up to one year. I then studied whether the PVL indices and the occurrence of TLVD can predict our pre-defined clinical outcomes (NYHA class improvement, 6-minute walk test improvement, acute heart failure syndrome (AHFS) occurrence and survival).
Best fit linear mixed effect modelling was used to compare the serial measures among groups (SPSS statistical software (IBM Corp. Version 22.0)).
Results:
To compare the LV contractility among groups, twenty-six (26) AS, 15 Ctrl and 12 HF patients were recruited. The HF group had the lowest EF, Starling contractility index (SCI), and pre-load recruitable stroke work (PRSW). SCI dropped more in AS and HF groups compared to the Ctrl with FFR (β coefficient = 1.12 ±0.47 for Ctrl Vs AS, p =0.007; β coefficient = -1.6 ±0.53 for HF Vs Ctrl, p =0.003). The difference between AS and HF groups was insignificant.
Twenty-four (24) AS patients completed the pre-and post-intervention study protocol. Fifty-eight per cent of patients (58%; 14/24) met the criterion of TLVD (the null hypothesis was rejected). Post-TAVI, EF and PRSW dropped significantly, and the left ventricular (LV) diastolic profile worsened. With FFR, cardiac output had a more blunted response post-TAVI (β coefficient = 0.237 ±0.08 for the pre-Vs post, p =0.005). The LV of patients with TLVD showed a more blunted SCI response to FFR pre-TAVI (β coefficient = 0.633 ±0.3 for no-TLVD Vs TLVD, p =0.049). TLVD patients were more symptomatic (NYHA class ≥3, 65% Vs 10%, p =0.013; 6 min walk test 135 Vs 280, p =0.037).
The non-invasive ESPVR Kelly's method (Ees(sb) = 0.9 × systolic blood pressure/end systolic volume) had the best agreement with invasive ESPVR. Systolic blood pressure, as measured by the brachial blood pressure cuff, had the best agreement with end-systolic pressure in severe aortic stenosis.
Baseline NT-pro-BNP correlated negatively with SCI (Spearman’s R, -0.41, p =0.017) and PRSW (Spearman’s R, -0.459, p =0.007). The invasive indices of contractility correlated well with the volumetric measures of the LV, such as end-diastolic volume and LV wall stress.
Finally, AHFS admissions (15%) and one-year mortality (9%) were not different between the TLVD groups (TLVD versus no-TLVD). Using univariate analysis, SCI and ESPVR at peak pacing rate were different between one-year survivors and non-survivors. Stress studies (incremental pacing) identified a unique group of patients with higher contractility indices but smaller LV volumes (e.g. EDV at peak pacing rate) that faired a worse prognosis.
Conclusion:
Patients with severe aortic stenosis and preserved left ventricular ejection fraction have impaired contractile reserve. Transient left ventricular dysfunction (TLVD) after aortic valve intervention is a true phenomenon and occurs at an intrinsic myocardial level. TLVD correlates well with the severity of symptoms pre valve intervention. Stress studies in aortic stenosis, incorporating pressure-volume loop indices, identified a specific subgroup of patients who faired a worse prognosis. Kelly’s method (the non-invasive measure of ventricular elastance) has the best agreement with the invasive estimate of ventricular elastance in severe aortic stenosis.
| Date of Award | 1 May 2021 |
|---|---|
| Original language | English |
| Awarding Institution |
|
| Supervisor | Philip MacCarthy (Supervisor) & Rafal Dworakowski (Supervisor) |