Aortic Valve Replacement - Lund University



Aortic Valve Replacement

The Influence of Prosthesis-Patient Mismatch

for Left Ventricular Remodeling,

Cardiac Function and Survival

Shahab Nozohoor, M.D.

Department of Cardiothoracic Surgery

Faculty of Medicine

Lund University, 2009

Doctoral Dissertation

Department of cardiothoracic Surgery

Faculty of Medicine

Lund University

SE-221 85 Lund

SWEDEN

© Shahab Nozohoor, 2009 (pages 1-80)

Lund University

Printed by Media-Tryck, Lund, 2009

ISBN 978-91-86253-87-5

Dedicated to Ann and Saga,

the noor of my eyes

"A Parisian tailor, not yet old, having dined and left his house had walked hardly 40 paces when he suddenly fell to the ground and expired. His body was opened and no disease found except that the three semilunar cusps leading to the aorta were bony".

Théofile Bonet, 1679

List of Publications 7

Abstract 8

Abbreviations 9

Introduction 10

1.1 Prosthesis-patient mismatch – the concept 10

1.2 Classification of PPM 10

1.3 Determinants of prosthesis-patient mismatch 11

1.4 Hemodynamic impact of prosthesis-patient mismatch 11

1.5 The clinical influence of prosthesis-patient mismatch 14

1.5.1 Survival 14

1.5.2 Prosthesis-patient mismatch and morbidity 15

1.5.3 Prosthesis-patient mismatch and BMI 16

1.5.4 Prosthesis-patient mismatch and in vivo EOA 16

1.5.5 Severe prosthesis-patient mismatch 18

1.5.6 Homogeneity and propensity scoring 18

1.5.7 Prosthesis-patient mismatch and left ventricular mass regression 19

1.5.8 Prosthesis-patient mismatch quality of life 20

1.5.9 Prosthesis-patient mismatch and aortic valve insufficiency 20

1.6 Postoperative heart failure 20

1.7 Diastolic dysfunction in patients with aortic valve disease 21

1.8 Brain-type natriuretic peptide 22

1.8.1 Preoperative measurement of brain-type natriuretic peptide in aortic valve stenosis 22

1.8.2 Measurement of brain-type natriuretic peptide following aortic valve replacement 23

1.9 Aortic valve stenosis 23

1.9.1 Pathophysiology 23

1.10 Aortic valve insufficiency 24

1.10.1 Acute aortic valve insufficiency 24

1.10.2 Chronic aortic valve insufficiency 25

1.11 Aortic valve replacement 25

1.12 The small aortic root and alternative surgical strategies 27

Aims of this Research 29

Material and Methods 30

3.1 Patients 30

3.2 Study design 30

3.2.1 Paper I 30

3.2.2 Paper II 31

3.2.3 Paper III 31

3.2.4 Paper IV 31

3.3 Anesthetic management 32

3.4 Surgical management 32

3.5 Triage® BNP test 32

3.6 Echocardiography 33

3.7 Definitions 34

3.8 Statistical analyses 34

3.9 Ethical aspects 35

Results 36

4.1 Impact of patient-prosthesis mismatch on in-hospital complications 36

4.2 Risk factors for postoperative neurological events 36

4.3 Left ventricular mass regression and diastolic dysfunction 37

4.4 Left ventricular remodeling following AVR for severe aortic valve insufficiency 39

4.5 Predictors of postoperative heart failure 41

4.6 Impact of PPM on early mortality 41

4.7 Impact of PPM on late mortality 43

4.8 Postoperative appearance of BNP and the relation between BNP and PPM 44

Discussion 45

5.1 Postoperative morbidity 45

5.2 Impact of PPM on early mortality 46

5.3 Postoperative heart failure 47

5.4 Impact of PPM on diastolic heart failure 48

5.5 Stented bioprostheses for supra-annular implantation 49

5.6 Impact of PPM on LV remodeling in aortic valve insufficiency 50

5.7 Impact of PPM on mortality 51

5.8 General discussion 53

5.9 Limitations 55

Future perspectives 56

Conclusions 57

Populärvetenskaplig sammanfattning (Summary in Swedish) 58

Acknowledgements 61

References 63

Papers I-IV 81

List of Publications

This thesis is based on the following papers, which are referred to in the text by their Roman numerals:

I. Nozohoor S, Nilsson J, Lührs C, Roijer A, Sjögren J. The Influence of Patient-Prosthesis Mismatch on In-hospital Complications and Early Mortality after Aortic Valve Replacement. Journal of Heart Valve Disease 2007;16:475-482.

II. Nozohoor S, Nilsson J, Lührs C, Roijer A, Algotsson L, Sjögren J. B-type Natriuretic Peptide as a Predictor of Postoperative Heart Failure following Aortic Valve Replacement. Journal of Cardiothoracic and Vascular Anesthesia 2009;23:161-165.

III. Nozohoor S, Nilsson J, Lührs C, Roijer A, Sjögren J. Influence of Prosthesis-Patient Mismatch on Diastolic Heart Failure after Aortic Valve Replacement. The Annals of Thoracic Surgery 2008;85:1310-1317.

IV. Nozohoor S, Nilsson J, Lührs C, Roijer A, Sjögren J. Influence of Prosthesis-Patient Mismatch on Left Ventricular Remodeling in Severe Aortic Insufficiency. European Journal of Cardiothoracic Surgery, online publication: 19-AUG-2009; DOI: 10.1016/j.ejcts.2009.07.009

Abstract

Valve substitution due to aortic valve disease corrects anatomical defects, promotes regression of myocardial hypertrophy, recovery of left ventricular performance, and remission of symptoms. However, the best valve substitute in terms of hemodynamic performance, durability, incidence of complications, and survival remains the subject of much debate. It has been suggested that valve performance is influenced by the potentially modifiable variable prosthesis-patient mismatch (PPM). PPM has been reported to be detrimental for survival and symptom resolution mainly due to the promotion of unfavorable prosthesis hemodynamics with secondary impaired left ventricular remodeling. Nevertheless, an increasing number of studies with various study designs and outcomes present conflicting results. Thus, there is no convincing evidence for PPM’s detrimental effects.

The aims of this research were to evaluate the impact of PPM on in-hospital complications and survival, to analyze whether postoperative heart failure can be detected using brain-type natriuretic peptide (BNP) as a predictive biomarker, to investigate the influence of PPM in bioprostheses with respect to recovery of left ventricular diastolic function and left ventricular mass regression, and to evaluate the influence of prosthesis-patient mismatch on left ventricular remodeling following aortic valve replacement for severe valve insufficiency.

The present work demonstrated that PPM was not associated with low cardiac output syndrome, but rather an independent risk factor for a neurological event during the postoperative period after valve replacement. This finding probably reflects a more cumbersome surgical procedure in a small aortic root with extensive calcification, commonly observed in patients with native valvular stenosis. PPM had no impact on either early or late mortality. Postoperative heart failure following AVR was associated with a high early postoperative mortality and was predicted by elevated BNP levels on arrival in the ICU although the discriminatory ability of the biomarker was poor. PPM did not impair left ventricular mass regression or the recovery of diastolic function. PPM was surprisingly common in patients with severe aortic insufficiency undergoing AVR. In these patients, left ventricular remodeling was initiated regardless of preoperative left ventricular ejection fraction or PPM.

In conclusion, the clinical relevance and the prevention of PPM seem subordinate and to improve patient outcome, priority should be given to the design of a durable, non-thrombogenic prosthesis permitting easy handling and reducing surgical complexity.

Abbreviations

AS aortic stenosis

AVR aortic valve replacement

AVI aortic valve insufficiency

BNP brain-type natriuretic peptide

BSA body surface area

CABG coronary artery by-pass grafting

CFR coronary flow reserve

CPB cardiopulmonary bypass

COPD chronic obstructive pulmonary disease

CVI cerebrovascular insult

DHF diastolic heart failure

EOA effective orifice area

EOAi indexed effective orifice area (EOA/BSA)

GOA geometric orifice area

GOAi indexed geometric orifice area (GOA/BSA)

IABP intra-aortic balloon pump

ICU intensive care unit

IVSd interventricular septum at end-diastole

LA left atrium dimension at end-systole in parasternal long axis view

LCOS low cardiac output syndrome

LFLG AS low-flow, low-gradient aortic stenosis

LPWDd left ventricular posterior wall dimension at end-diastole

LV left ventricle

LVEDD left ventricular end-diastolic diameter

LVESD left ventricular end-systolic diameter

LVEF left ventricular ejection fraction

LVH left ventricular hypertrophy

LVIDd left ventricular internal dimension in diastole

LVMI left ventricular mass index

LVMR left ventricular mass regression

LVOT left ventricular outflow tract

MI myocardial infarction

PHF postoperative heart failure

PPM prosthesis-patient mismatch

QoL quality of life

ROC receiver operating characteristic

SVD structural valve deterioration

TPG transprosthetic gradient

TVI time-velocity integral

Introduction

1.1 Prosthesis-patient mismatch – the concept

Prosthesis-patient mismatch (PPM) was first described by Rahimtoola in 1978 (1) who stated that “mismatch can be considered present when the effective prosthetic valve area, after insertion into the patient, is less than that of a normal human valve”. Rahimtoola suggested that the degree of PPM could be quantified, which would aid in identifying patients at risk of clinical sequelae caused by this condition. The pathophysiology of mismatch was subsequently proposed to be related to persistent valve gradients based on in vitro studies conducted by Dumesnil and Yoganathan (2). They demonstrated an exponential relationship between the mean transprosthetic pressure gradient and the indexed effective orifice area (EOAi, i.e. PPM) for aortic bioprostheses in an in vitro physiologic pulse-duplicator system. Their findings led to the recommendation that the EOAi should ideally not be less than 0.9 to 1 cm2/m2 for aortic bioprostheses to minimize residual postoperative transprosthetic pressure gradients. This subsequently led to the premise that there may be a correlation between the decrease in transvalvular gradient and the clinical improvement seen after surgery (3;4). With the development of Doppler echocardiography, in vivo observations demonstrated that normally functioning valve prostheses could have relatively high postoperative transvalvular gradients corresponding to the phenomenon previously referred to as prosthesis–patient mismatch (5-7). PPM was suggested to occur more often in patients with large body surface area (BSA), in whom a high cardiac output across a small orifice area may produce high transprosthetic gradients (1;8). Hence, the calculated effective orifice area (EOA) of a specific prosthesis has frequently been adjusted for BSA to ensure its hemodynamic performance for an individual patient. The most widely accepted and validated parameter for identifying PPM is the indexed EOA, which is the EOA of the prosthesis divided by the patient’s BSA (9;10). Prosthesis–patient mismatch has been recognized as a functional hemodynamic abnormality rather than being due to an intrinsic defect of the prosthesis and is identified as a nonstructural dysfunction by the Society of Thoracic Surgeons (11). Previous studies have demonstrated that mismatch is a common phenomenon when using a relatively conservative definition (i.e., EOAi≤0.85 cm2/m2), observed in 20 to 70% whereas the prevalence of severe PPM ranges from 2 to 10% (9;12).

1.2 Classification of PPM

It has previously been demonstrated by Pibarot et al. (9) that the relation between the transprosthetic gradients and the EOAi is nonlinear and that the gradient increases exponentially when the EOAi falls below 0.8 to 0.9 cm2/m2 as shown in Figure 1.1. The value of EOAi≤0.85 cm2/m2 is thus generally regarded as the threshold for PPM with values between 0.65 and 0.85 cm2/m2 being classified as moderate PPM and 21 mm (3;7). Mismatch occurs more frequently in patients with stenotic native valves as they generally have smaller valvular annuli than those with regurgitant valves (16). Furthermore, calcific aortic stenosis is by far the most prevalent lesion in older patients undergoing aortic valve replacement (AVR).

1.4 Hemodynamic impact of prosthesis-patient mismatch

The main consequence of prosthesis–patient mismatch is the generation of a high transvalvular gradient through a normally functioning prosthetic valve. Assuming a normal cardiac index of 3 l/min/m2, implantation of a prosthesis with an EOA of 1.3 cm2 in a patient with a BSA of 1.5 m2 will theoretically result in a mean transprosthetic gradient (TPG) of about 13 mmHg. The mean TPG would theoretically be 35 mm Hg if the same prosthesis were to be implanted in a patient with a BSA of 2.5 m2 (10).

The increased transvalvular gradient associated with PPM has been shown to result in an increased left ventricular (LV) work, which in turn influenced the regression of LV hypertrophy (LVH) (17-20). LVH is in turn a strong independent risk factor for mortality as well as a major determinant of systolic and diastolic function and exercise capacity in patients undergoing valve replacement (21;22). Normalization of LV mass is therefore a crucial goal of AVR. The persistence of LVH associated with PPM has been proposed to be one of the factors contributing to adverse outcomes related to PPM. However, Sharma et al. (23) reviewed the published literature on LV mass regression (LVMR) after valve replacement for aortic stenosis over the past 23 years. They found that surgical correction of stenosis by valve replacement led to unequivocal regression of LV mass regardless of the type of valve inserted with the bulk of the hypertrophy regressing within the first 6 months of operation. These findings are supported by more recent publications also demonstrating that the extent of LVMR is maximal during the first 6 postoperative months and influenced only by the preoperative degree of hypertrophy and the presence of hypertension (24). In other studies, neither prosthesis size nor type was correlated with LVMR (25;26). One explanation of the conflicting resluts related to LVMR and PPM may be that many studies showing long-term detrimental effects of LVH have been conducted in patients with hypertensive and ischemic heart disease (27;28). It remains to be seen whether similar consequences are observed with respect to the hypertrophy due to valvular disease.

The difference in TPG between patients with PPM and those without may be even more important during exercise, given that gradients are a square function of flow. Recently, Bleiziffer et al. (29) were the first to report that the presence of PPM significantly influenced the peak physical exercise capacity, according to stress test echocardiography, following AVR. The authors suggested that their findings could be explained by an increase in hemodynamic burden resulting from higher gradients in patients with PPM. Another possible explanation is that PPM may limit the increase in cardiac output during exercise, similar to that observed in native aortic stenosis. This may in turn limit the capacity of the cardiac function to match the increasing metabolic demand during intense exercise. Mannacio et al. (30) evaluated the impact of PPM defined as EOAi ................
................

In order to avoid copyright disputes, this page is only a partial summary.

Google Online Preview   Download