Non-Surgical Septal Reduction Therapy
in Hypertrophic Obstructive
Cardiomyopathy: Current practice and
future potential
1
Robert Michael Cooper
CID 744810
National Heart and Lung Institute
Imperial College London
PhD
2
This body of work is my own and represents original work. All else is appropriately referenced.
The copyright of this thesis rests with the author and is made available under a Creative Commons Attribution Non-Commercial No Derivatives licence.
Researchers are free to copy, distribute or transmit the thesis on the condition that they attribute it, that they do not use it for commercial purposes and that they do not alter, transform or build upon it. For any reuse or redistribution, researchers must make clear to others the licence terms of this work
3
Abstract
Obstruction in hypertrophic cardiomyopathy is associated with increased morbidity and mortality. Effective resolution of outflow tract obstruction can provide relief of symptoms and perhaps improve prognosis. Non-surgical septal reduction in the form of alcohol septal ablation (ASA) has been performed since 1994 with limited progress in the last decade.
ASA using traditional methods has an unacceptably high rate of failure to resolve LVOT gradient. By creating and populating a relational database I was able to show that failure to treat LVOT gradient satisfactorily was seen in 41% after one procedure and 18% after multiple procedures. This can be partly explained by inaccurate location of the iatrogenic infarction, seen on CMR.
Improving accuracy of infarction in ASA could be achieved by better peri- procedural imaging.
Intracardiac echocardiography (ICE) provides excellent image quality of the contact point of the mitral valve on the septum in HOCM, but is no better than standard transthoracic echocardiography in describing detail of the septum or other cardiac structures relevant to ASA. ICE cannot see myocardial contrast well and therefore cannot be used to guide ASA alone.
4
Computed tomography (CT) angiography can visualise small septal arteries.
The ability to merge angiographic images with structural detail allows description of the path of arteries to guide alcohol injection in ASA. The use of
CT planning improved the success rate of ASA after one procedure from 59% to
85%. We observed less RBBB (13% vs 62%) due to improved targeting of the LV septum, confirmed by CMR.
Patient selection in ASA is important. A standard operating procedure for assessment and treatment of HOCM patients is now part of routine clinical care.
Some patients cannot receive trans-coronary alcohol due to arterial anatomical restrictions. Direct endocardial radiofrequency ablation of the interventricular septum with merged ICE/CARTO electrophysiology mapping system guidance was explored with encouraging results.
5
Acknowledgments:
First and foremost I must acknowledge the input and guidance of my primary supervisor, Professor
Rod Stables. He was responsible for setting up an ASA service in Liverpool and it was with his reputation and funding that a research project was allowed at Liverpool Heart and Chest Hospital.
He was approachable and involved from day 1 to the end. He was enthusiastic and took his pastoral role very seriously. This project would not have been possible without him. I hope this working clinical and research relationship continues for many years.
With regard to the expert assessors for specific chapters in this thesis I must thank the following people:
Dr James Newton; Consultant Cardiologist Oxford John Radcliffe Hospital. Expert reviewer for ICE and TTE in chapter 4.
Dr Niels Vejlstrup: Consultant cardiologist at Rigshospitalet, Copenhagen. Expert reviewer for ICE and TTE in chapter 4.
Dr Jonathan Hasleton: Consultant Cardiologist at Royal Liverpool Hospital. Expert reviewer for assessment of myocardial infarction and CMR appearances post ASA Chapter 3.
Dr Sukumaran Binukrishnan: Consultant Cardiac Radiologist Liverpool Heart and Chest Hospital.
Expert reviewer for CT analysis in Chapters 5 and 6.
Lastly I must thank my wife Menna and children, Sam and Ben or their ongoing support. Without them this project would have been completed much sooner.
6
Author’s contributions:
As the author of this thesis I should clarify my role in the creation of the project, the collection of data and the writing of subsequent manuscripts.
I interviewed for a research post that was already approved at Liverpool Heart and Chest Hospital and was appointed ahead of other candidates. The basic structure of a project of appraising outcomes so far and introducing ICE was in place. The use of CT angiography was simply at a conceptual stage, as was the use of RF ablation. I was therefore involved in the ethical approvals for each section of study and the methods and data collection protocols.
I was present for and collected and collated the imaging data for each chapter. I did not personally score any data as I wanted to exclude this opportunity for bias. I also did not acquire the echocardiography images as part of chapter 4, this was to remove the possibility of acquisition bias.
An independent qualified sonographer therefore acquired all images, I was present for all echoes commented on in this thesis.
The CT scans were analysed by one expert radiologist – I am not formally trained in CT and therefore couldn’t perform this duty.
All ASA procedures were performed by Professor Rod Stables. The second operator varied somewhat according to lab rotas etc. I was present for all procedures for peri-ASA imaging guidance and to ensure research protocols were adhered to.
All RF procedures were performed by Dr Simon Modi with me as the second operator. I operated the
ICE catheter for these procedures. I then descrubbed to delineate the borders of the cardiac structures on the echo console.
7
Abbreviations list
HCM: Hypertrophic cardiomyopathy
HOCM: Hypertrophic obstructive cardiomyopathy
ASA: Alcohol septal ablation
LV: Left ventricle
RV: Right ventricle
MV: Mitral valve
SAM: Systolic anterior motion
CT: Computed tomography
ICE: Intracardiac echocardiography
NYHA: New York Heart Association
LAD: Left anterior descending
RCA: Right coronary artery
OTW: Over the wire (balloon)
LA: Left atrium
LHCH: Liverpool Heart and Chest Hospital
RF(A): Radiofrequency (ablation)
8
Contents
Acknowledgments: ...... 6
Author’s contributions: ...... 7
Abbreviations list ...... 8
Chapter 1: Current status of Non-surgical Septal Reduction Therapy (NSRT) in Hypertrophic
Obstructive Cardiomyopathy (HOCM) ...... 31
1.1 Introduction ...... 32
1.2 Pathophysiology of LVOT obstruction in HOCM ...... 34
1.3 Patient selection for NSRT ...... 36
1.4 Alcohol septal ablation: The procedure ...... 39
1.5 Results of ASA case series reports ...... 42
1.5.1 Survival ...... 42
1.5.2 Risk of ventricular arrhythmia...... 42
1.5.3 Symptom and gradient resolution ...... 43
1.5.4 Comparison to surgical myectomy ...... 46
1.5.5 Predictors of outcome in ASA ...... 49
1.5.6 Peri-procedural complications of ASA ...... 49
1.6 Alternative methods of NSRT ...... 51
1.7 Summary ...... 55
1.7.1 What questions must we seek to answer? ...... 55
1.7.1.1 What proportion of patients derives no benefit from ASA? ...... 55
9
1.7.1.2 Why do some ASA procedures fail to have the desired effect on LVOTO? ...... 56
1.7.1.3 If ASA procedures fail because of inaccurate infarct; how can we improve? ...... 56
1.7.1.4 What if patients cannot receive trans-coronary alcohol as part of ASA? ...... 57
Chapter 2: The Liverpool Experience: Describing outcomes in patients undergoing ASA during the period 2000-2011 ...... 59
2.1 Introduction ...... 60
2.2 Aims...... 63
2.3 Methods ...... 64
2.3.1 Database creation ...... 64
2.3.1.1 Flat versus relational database solutions ...... 64
2.3.1.2 Data requirements ...... 65
2.3.1.3 Relationships ...... 68
2.3.1.4 Dropdown menus ...... 72
2.3.1.5 Queries ...... 74
2.3.2 Patient selection: ...... 78
2.3.3 Statistical methods ...... 80
2.3.4 Ethical approval ...... 81
2.4 Results ...... 82
2.4.1 Patient details: ...... 82
2.4.2 Procedural details: ...... 83
2.4.2.1 Alcohol dose to CKMB correlation: ...... 86
2.4.2.2 Peri-procedural complications: ...... 87
10
2.4.3 Clinical outcomes ...... 88
2.4.3.1 Length of Follow up...... 88
2.4.3.2 Survival and risk of ventricular arrhythmia: ...... 90
2.4.3.3 Impact on symptoms:...... 92
2.4.3.4 Cardiopulmonary exercise testing: ...... 93
2.4.3.5 Echocardiographic parameters: ...... 94
2.4.4 Assessment of failure of treatment: ...... 96
2.4.4.1 LVOT gradient:...... 96
2.4.4.2 Dyspnoea: ...... 97
2.4.4.3 Combining clinical and echocardiographic outcome measures: ...... 97
2.5 Discussion ...... 99
2.5.1 Risk of ASA...... 99
2.5.2 Success and failure of traditional ASA methods ...... 99
2.5.3 Does aggregate outcome data mask a disappointing failure rate in individual patients?
100
2.5.4 Why does ASA fail to resolve LVOT gradients in some? ...... 101
2.5.5 What limitations do we face with ASA? ...... 101
2.5.5.1 Identification of septal artery targets: ...... 102
2.5.5.2 Technical instrumentation: ...... 102
2.5.5.3 Difficulty controlling infarct size: ...... 102
2.5.5.4 The need for precision ...... 103
2.5.6 Prospects for improvement: ...... 103
11
2.5.6.1 Procedural changes: ...... 103
2.5.7 Limitations of the analysis ...... 104
2.5.7.1 Control of outcome assessment ...... 104
2.7.1.2 Echocardiographic data ...... 105
2.7.1.3 Event reporting; ...... 106
2.7.2 Loss to follow up ...... 106
2.6 Summary ...... 107
Chapter 3: An exploratory study of the relationship between iatrogenic infarct location and the impact on LVOT haemodynamics post-ASA ...... 108
3.1 Introduction ...... 109
3.1.1 Mechanism of myocardial infarction in ASA ...... 109
3.2 Aims...... 111
3.3 Methods ...... 112
3.3.1 Choice of imaging modality ...... 112
3.3.1.1 Choice of LGE quantification methodology ...... 113
3.3.2 CMR acquisition methods ...... 113
3.3.3 Standard CMR image analysis ...... 116
3.3.4 Identification of target myocardium ...... 118
3.3.4.1 Basal target area definition ...... 120
3.3.4.2 Short axis area target definition ...... 123
3.3.4.3 LV versus RV definition...... 125
3.3.4.4 Summary of identification of target myocardium ...... 126
12
3.3.5 Image analysis for location of ASA induced infarct ...... 127
3.3.5.1 Accommodating pre-existing LGE not related to ASA-induced scar ...... 129
3.3.6 Patient selection and scan quality assessment ...... 129
3.3.7 Statistical analysis ...... 132
3.4 Results ...... 133
3.4.1 Assessment of LV size and function ...... 133
3.4.2 Correlation of infarct size markers ...... 134
3.4.3 Target area evaluation ...... 136
3.4.3.1 RV vs LV septum ...... 138
3.5 Discussion ...... 139
3.5.1 Limitations ...... 140
3.5.1.1 Hazards of retrospective data collection ...... 141
3.5.1.2 Restrictions of analysis methods ...... 142
3.6 Conclusion ...... 145
Chapter 4: Intra-cardiac Echocardiography to guide alcohol septal ablation in hypertrophic obstructive cardiomyopathy: A prospective validation study against trans-thoracic echocardiography
...... 146
4.1 Introduction ...... 147
4.2 Aims...... 149
4.3 Methods ...... 150
4.3.1 Defining the cardiac structures important to ASA ...... 150
4.3.1.1 The target myocardium ...... 150
13
4.3.1.2 Other structures ...... 150
4.3.1.3 A scoring system for assessing images...... 151
4.3.2 Intracardiac echocardiography ...... 153
4.3.2.1 Patient groups ...... 154
4.3.2.2 Power calculation ...... 154
4.3.3 Image acquisition ...... 156
4.3.3.1 Phase 1 ...... 156
4.3.3.2 Phase 2 ...... 156
4.3.4 Image analysis ...... 157
4.3.5 Statistical analysis ...... 159
4.3.6 Research approvals ...... 159
4.3.7 Research grants ...... 160
4.4 Results ...... 161
4.4.1 Phase 1 ...... 161
4.4.1.1 Work view ...... 161
4.4.1.2 All views ...... 163
4.4.1.3 Agreement between reviewers: ...... 164
4.4.2 Phase 2 ...... 171
4.4.2.1 Work view ...... 171
4.4.2.2 All views ...... 173
4.4.2.3 Agreement between reviewers...... 173
4.4.2.4 Contrast localisation...... 185
14
4.5 Discussion ...... 187
4.5.1 Phase 1 ...... 187
4.5.2 Phase 2 ...... 191
4.5.3 Myocardial contrast echocardiography ...... 192
4.5.3.1 Extensive acoustic shadowing ...... 192
4.5.3.2 Inadequate visualisation of injected contrast ...... 193
4.5.3.3 Contrast localising to distant structures ...... 194
4.5.3.4 Alteration of echocardiographic settings ...... 195
4.5.3.5 TTE vs ICE in myocardial contrast echocardiography ...... 196
4.5.4 Anatomical observations from ICE ...... 197
4.6 Conclusion ...... 200
4.7 Limitations...... 201
Chapter 5: An examination of the potential role of Computed Tomography (CT) angiography in the identification of target septal arteries prior to alcohol septal ablation for hypertrophic obstructive cardiomyopathy ...... 202
5.1 Introduction ...... 203
5.2 Aims...... 207
5.3 Methods ...... 208
5.3.1 Scoring system ...... 208
5.3.1.1 Required septal anatomy information: ...... 208
5.3.1.2 Assessment of image quality and ability to guide ASA ...... 212
5.3.2 Patient selection ...... 213
15
5.3.2.1 Sample size calculations ...... 214
5.3.2.2 Inclusion criteria ...... 215
5.3.3 CT image acquisition methods ...... 215
5.3.4 CT image analysis methods: ...... 216
5.4 Results ...... 223
5.4.1 Patient and criteria details ...... 223
5.4.2 Assessment of image quality ...... 223
5.4.3 Anatomy of septal arteries seen on CT ...... 226
5.4.3.1 Number of septal arteries per patient: ...... 226
5.4.3.2 Parent artery ...... 227
5.4.3.3 Bifurcation angle ...... 227
5.4.3.4 Septal artery diameter ...... 228
5.4.3.5 Length of septal vessel in fat and myocardium ...... 228
5.4.3.6 Length of septal vessel prior to bifurcation ...... 228
5.4.3.7 Landmarks ...... 228
5.4.3.8 Myocardial distribution ...... 229
5.4.3.9 Preferred projection for ostium of septal vessels ...... 237
5.4.3.10 Preferred projection for distal septal vessel ...... 237
5.5 Discussion ...... 239
5.5.1 Image quality ...... 239
5.5.2 Alternative epicardial artery source ...... 239
5.5.3 Septal vessels supplying multiple territories ...... 240
16
5.5.4 The ability to rule out alternative septal vessels ...... 241
5.6 Conclusion ...... 242
5.7 Limitations...... 243
Chapter 6: Clinical outcome of hypertrophic obstructive cardiomyopathy patients undergoing CT guided ASA ...... 244
6.1 Introduction ...... 245
6.2 Aims...... 246
6.3 Methods ...... 247
6.3.1 Patient selection ...... 247
6.3.1.1 Echocardiographic assessment ...... 248
6.3.1.2 Cardiopulmonary exercise (CPEX) testing ...... 248
6.3.1.3 Cardiac magnetic resonance imaging ...... 248
6.3.1.3 Research permissions ...... 249
6.3.1 CT methods ...... 249
6.3.1.1 CT image acquisition ...... 249
6.3.1.2 CT image analysis ...... 250
6.3.2 Alcohol septal ablation procedure ...... 255
6.3.4 Statistical analysis ...... 256
6.3.4.1 Comparison to traditional ASA ...... 256
6.4 Results ...... 258
6.4.1 CT image quality and anatomy ...... 258
6.4.1.1 CT image quality ...... 258
17
6.4.1.2 Anatomy of septal arteries seen on CT in HOCM patients ...... 260
6.4.2 Procedural details ...... 269
6.4.2.1 Procedural complications ...... 271
6.4.3 Clinical outcomes ...... 272
6.4.3.1 Survival and risk of ventricular arrhythmia: ...... 272
6.4.3.2 Symptomatic resolution ...... 272
6.4.3.3 Echocardiographic data ...... 274
6.4.3.4 Cardiopulmonary exercise testing ...... 274
6.4.3.5 Quality of life ...... 275
6.4.4 Comparison to traditional ASA ...... 275
6.5 Discussion ...... 278
6.5.1 Anatomical insights from CT alters the approach to ASA ...... 278
6.5.1.1 Dual vascular supply to RV and LV ...... 278
6.5.1.2 Alternative epicardial artery ...... 281
6.5.2 Effect of CT guidance on clinical outcomes ...... 281
6.5.2.1 CT guided approach improves control of infarct size and location ...... 281
6.5.2.1 Effect of better infarct localisation on clinical outcomes ...... 283
6.5.3 Exploring those with treatment failure ...... 283
6.5.3.1 Those who received alcohol ...... 283
6.5.3.2 Those who did not receive alcohol ...... 285
6.6 Conclusion ...... 289
6.7 Limitations...... 290
18
Chapter 7: Standard operating policy for assessment of patients referred for ASA at LHCH ...... 291
7.1 Introduction ...... 292
7.2 Aims...... 294
7.3 Section 1: Clinical review ...... 295
7.2.1 Background to HCM diagnosis ...... 295
7.3.2.1 Important clinical questions ...... 295
7.3.2 Symptom burden ...... 297
7.3.3 Medication review ...... 298
7.3.3.1 Current medications...... 298
7.3.3.2 Medication choice and titration...... 298
7.3.4 Exploration of pacing options ...... 299
7.3.5 Patient information ...... 301
7.3.6 Criteria for progression to ASA: Clinical history ...... 304
7.4 Section 2: Echocardiographic assessment ...... 305
7.4.1 Resting transthoracic echocardiogram (TTE) ...... 305
7.4.1.1 Septal size ...... 305
7.4.1.2 Systolic anterior motion (SAM) of the mitral valve ...... 306
7.4.1.3 LVOT gradient ...... 310
7.4.1.4 Diastolic function ...... 312
7.4.1.5 LVOT anatomical variants associated with HOCM ...... 312
7.4.2 Criteria for progression to ASA: Echocardiographic assessment ...... 313
7.4 Section 3: Functional testing ...... 314
19
7.4.1 Full pulmonary function testing ...... 314
7.4.2 Cardio-pulmonary exercise testing ...... 314
7.4.3 Criteria for progression to ASA: Functional testing ...... 317
7.5 Section 4: Cardiac Magnetic Resonance Imaging: ...... 318
7.5.1 Rule out phenocopies of HCM ...... 318
7.5.2 Alternative anatomical abnormalities responsible for LVOT gradient ...... 318
7.5.3 Visualise pre-existent fibrosis ...... 321
7.5.4 CMR protocol for pre-procedural investigation ...... 321
7.5.4.1 HCM pre-ASA ...... 322
7.5.4.2 HCM Immediate Post Ablation ...... 323
7.5.4.3 HCM 6 month scan ...... 324
7.6 Section 5: CT coronary angiography imaging ...... 326
7.6.1 Protocols for CT angiography ...... 326
7.6.1.1 Acquisition of images ...... 326
7.6.1.2 Protocol for analysis of images ...... 327
7.7 Section 6: Multi-disciplinary team meeting to discuss suitability for septal reduction therapy.
329
7.8 Conclusion ...... 330
Chapter 8: Radiofrequency ablation of the interventricular septum to treat outflow tract gradients in hypertrophic obstructive cardiomyopathy: A novel use of CARTOSound® technology to guide ablation ...... 331
8.1 Introduction ...... 332
20
8.2 Aims...... 334
8.3 Methods ...... 335
8.3.1 Patient selection ...... 335
8.3.1.1 Individual patient details...... 336
8.3.2 Research permissions ...... 339
8.3.3 Patient assessment: ...... 340
8.3.3.1 Imaging assessments ...... 340
8.3.3.2 Functional assessments ...... 340
8.3.4 Procedural details ...... 341
8.4 Results ...... 350
8.4.1 Procedural complications...... 350
8.4.2 Symptomatic resolution ...... 351
8.4.3 Echocardiographic parameters ...... 351
8.4.4 Exercise capacity ...... 354
8.4.5 Quality of life questionnaire ...... 354
8.4.6 CMR imaging ...... 354
8.5 Discussion ...... 356
8.5.1 Procedural risk ...... 358
8.6 Conclusion ...... 361
8.7 Limitations...... 362
Chapter 9: Conclusions ...... 363
References ...... 365
21
Appendices: ...... 375
Appendix 2.1: Database documenter ...... 375
Appendix 2.2: Service evaluation registry form NSRT at LHCH...... 386
Appendix 4.1: Intracardiac Echocardiography in LHCH permissions ...... 408
Appendix 4.2: Patient information leaflet for the use of intracardiac echocardiography ...... 411
Appendix 6.1: Permissions to use CT angiography prior to ASA at LHCH ...... 413
Appendix 6.2: Example case record form for CT details prior to ASA ...... 415
Appendix 7.1: Patient information leaflet for ASA at LHCH ...... 419
Appendix 8.1: Permissions to use RF ablation in HOCM...... 421
Appendix 8.2: Patient information leaflet for RF in HOCM ...... 423
Appendix 9: Published papers...... 424
Appendix 9.1: Current status of NSRT...... 424
Appendix 9.2: Historical control ASA results ...... 425
Appendix 9.3: ICE vs TTE ...... 426
Appendix 9.4: CT planning for ASA improves infarct location and patient outcome ...... 427
Appendix 9.5: Intervention in HCM; Patient selection and emerging techniques in NSRT ...... 428
Appendix 9.6: RF ablation to treat LVOT gradients in HOCM ...... 429
Appendix 10: Permissions for reproduction: ...... 431
Appendix 10.1: Permissions for review article ...... 431
Appendix 10.2: Permissions for historical patient group reporting ...... 432
Appendix 10.3: Permissions from Echo Research and Practice ...... 433
Appendix 10.4: Permissions for CT guidance of ASA paper ...... 433
22
Appendix 10.5: Permissions for RF ablation in HOCM ...... 434
Appendix 11: Required Imperial/NHLI courses attended ...... 435
Appendix 12: Awards and presentations based on work from this thesis ...... 436
Appendix 12.1: Awards: ...... 436
Appendix 12.2: National podium presentations: ...... 438
Appendix 12.3: International / National Poster Presentations: ...... 438
23
Index of Figures:
Figure 2.1: Total number of ASA procedures performed in the UK by year 2007-14...... 61
Figure 2.2: Volume of ASA procedures by centre in 2014 ...... 61
Figure 2.3: One-to-one relationships ...... 69
Figure 2.4: One-to-many relationships with a single data entry record...... 70
Figure 2.5: One-to-many relationship with multiple data entry records ...... 71
Figure 2.6: All relationships in the retrospective ASA patient database ...... 72
Figure 2.7: Process for creating dropdown menu I...... 73
Figure 2.8: Process for creating dropdown menu II...... 73
Figure 2.9: Simple query example – procedure related events ...... 75
Figure 2.10: Creation of a complex query step 1 ...... 76
Figure 2.11: Creation of a complex query step 2 ...... 76
Figure 2.12: Creation of a complex query step 3 ...... 77
Figure 2.13: Creation of a complex query step 4 ...... 78
Figure 2.14: Patient identification for entry into retrospective ASA outcome database ...... 80
Figure 2.15: Patient flow diagram detailing procedures and reasons for no alcohol delivery ...... 85
Figure 2.16: Correlation of alcohol dose to CKMB release ...... 86
Figure 2.18: Follow up data available...... 89
Figure 2.19: Kaplan-Meier survival curve for patients undergoing ASA 2000-2011 ...... 90
Figure 2.20: NYHA class pre- and post-ASA ...... 92
Figure 2.21: Change in NYHA class following ASA ...... 93
Figure 2.22: LVOT gradients pre- and post-ASA ...... 96
Figure 3.1: Schematic to demonstrate piloting and acquisition of a short axis stack ...... 115
Figure 3.2: Equation for calculation of ejection fraction ...... 117
Figure 3.3: Screenshot of typical volume and mass data output from CMRTools ...... 117
24
Figure 3.4: Definition of target myocardium in the long axis...... 121
Figure 3.5: Example of assessment in the long axis – infarct is outside the target area apical of the target basal septum...... 122
Figure 3.6: Example of assessment in the long axis - infarct is inside the target area ...... 123
Figure 3.7: Series of SAX slices from Aorta and atrial level (A) to Mid-ventricle (D)...... 124
Figure 3.8: Example of creation of target area in SAX slices ...... 124
Figure 3.9: Examples of infarct occupying most of the target (A) and a smaller area (B) ...... 125
Figure 3.10: Example of RV only infarct occupying target area in long and short axis ...... 126
Figure 3.11: Method for analysis of LGE in a SAX slice ...... 128
Figure 3.12: Consort diagram for patient/scan recruitment ...... 131
Figure 3.13: Alcohol dose to new LGE mass correlation in all patients ...... 134
Figure 3.14: Alcohol dose to new LGE mass correlation in Success patients ...... 135
Figure 3.15: Alcohol dose to new LGE mass correlation in Fail patients ...... 135
Figure 3.16: CKMB release to new mass LGE correlations ...... 136
Figure 3.17: Example of intended RV-LV delineation from concept phase ...... 143
Figure 4.1: Flow chart describing method of analysis for ICE and TTE in Phase 1 ...... 158
Figure 4.2: Flow chart describing method of analysis for ICE and TTE in Phase 2 ...... 159
Figure 4.3: Reviewers chosen work views TTE Phase 1 ...... 169
Figure 4.4: Reviewers chosen work views TTE Phase 2 ...... 178
Figure 4.5: Kappa agreement statistics for Phase 2...... 184
Figure 4.6: Demonstration of inferior septum vs target septum in ICE and CMR images ...... 188
Figure 4.7: Demonstration of papillary anatomy in standard echo and CMR planes ...... 190
Figure 4.8: Example of excessive contrast opacificaiton on ICE ...... 193
Figure 4.9: Example of inadequate contrast opacification on ICE ...... 194
Figure 4.10: Example ICE images using differing frequencies and tissue harmonics ...... 195
Figure 4.11: ICE images of SAM-septal contact ...... 198
25
Figure 4.12: Coronary wire visible with ICE ...... 199
Figure 5.1: Microsoft Access table, field structure and relationships for CT in non-HCM analysis .... 213
Figure 5.2: Example of septal vessel travelling to basal septum ...... 217
Figure 5.3: Example of vessel tracked from LAD travelling to distal RV septum ...... 218
Figure 5.4: Septal length in fat before travelling in to myocardium ...... 218
Figure 5.5: Septal length in myocardium prior to bifurcation (A) and after (B) ...... 219
Figure 5.6: Angle of bifurcation from parent vessel (A) and cross-sectional area (B) ...... 219
Figure 5.7: Example of CT angiogram created in non-HCM population ...... 220
Figure 5.8: Number of septals identified ...... 226
Figure 5.9: Parent artery to septal vessels ...... 227
Figure 5.10: Number of territories supplied by septal vessel ...... 229
Figure 5.11: Territories supplied in arteries with one distribution ...... 230
Figure 5.12: Territories supplied in arteries with 2 distributions ...... 231
Figure 5.13: Territories supplied in arteries with 3 distributions ...... 232
Figure 5.14: Distribution of all septal vessels identified ...... 233
Figure 5.15: LV basal septal vessels - number of territories supplied ...... 234
Figure 5.16: LV basal septal arteries - other territory in dual supply arteries ...... 235
Figure 5.17: LV basal septal arteries - other territories in arteries with 3 distributions ...... 236
Figure 5.18: Distribution of all arteries involving the LV basal septum ...... 236
Figure 5.19: Chosen view for ostium of vessels supplying LV basal septum ...... 237
Figure 5.20: Chosen view for distal septal vessels supplying LV basal septum ...... 238
Figure 6.1: Patient flow diagram ...... 248
Figure 6.2: CT image of target myocardium and identified septal vessel...... 251
Figure 6.3: Septal artery anatomical information taken from single plane reformatimages ...... 252
Figure 6.4: Examples of matched CT and invasive coronary angiography...... 254
Figure 6.5: Number of septal arteries identified per patient ...... 260
26
Figure 6.6: Parent artery to identified septals ...... 261
Figure 6.7: Method of identification of septal artery ...... 262
Figure 6.8: Number of territories supplied by all septal vessels ...... 264
Figure 6.9: Distribution of all septal vessels ...... 265
Figure 6.10: Number of territories supplied by vessels with target myocardium branch ...... 266
Figure 6.11: Distribution of septal vessels including branch to target myocardium ...... 267
Figure 6.12: Preferred projection for ostium of septal vessel ...... 268
Figure 6.13: Preferred projection for distal vessel ...... 269
Figure 6.14: Cranio-caudal tilt for ostium and distal vessel ...... 269
Figure 6.15: Patient consort diagram - referral to alcohol delivery ...... 271
Figure 6.16: Alcohol dose to CKMB release correlation ...... 271
Figure 6.17: NYHA following CT guided ASA ...... 273
Figure 6.18: Alcohol dose to CKMB correlation in traditional and CT guided ASA ...... 277
Figure 6.19: Effect of sub-selective contrast injection...... 279
Figure 6.20: CT angiogram, invasive angiogram and MCE ...... 280
Figure 6.21: CMR LGE SAX images post CT guided ASA ...... 282
Figure 6.22: Coronary angiogram showing guide wire entering target with acute angle ...... 284
Figure 6.23: Myocardial contrast venting directly into LV on injection ...... 285
Figure 7.1: Medication choice in treatment of LVOT gradients ...... 299
Figure 7.2: Patient journey from assessment to follow up ...... 304
Figure 7.3: Example of grade 0 SAM ...... 307
Figure 7.4: Example of grade 1 SAM ...... 308
Figure 7.5: Example of Grade 2 SAM ...... 309
Figure 7.6: Example of grade 3 SAM ...... 310
Figure 7.7: Example of 'scimitar' shaped CW envelope of LVOT obstruction ...... 311
Figure 7.8: Flow chart for CPEX investigation of cause of exercise restriction ...... 316
27
Figure 7.9: Example of altered MV geometry causing LVOT gradient ...... 320
Figure 8.1: ICE delineated LV endocardial contours transferred to CARTO ...... 342
Figure 8.2: SAM-septal contact area mapping ...... 343
Figure 8.3: CARTO mapping of conduction system...... 344
Figure 8.4: RF burns over SAM-septal contact area...... 346
Figure 8.5: LVOT gradient and exercise capacity change post RF ablation ...... 353
Figure 8.6: CMR imaging post RF ablation ...... 355
28
Index of Tables:
Table 1.1: Indications for NSRT in patients with HOCM ...... 38
Table 1.2: Long term results from observational studies post ASA ...... 46
Table 2.3: Basic demographic and chronic disease status data collected...... 66
Table 2.4: Procedural data collected ...... 66
Table 2.5: Symptomatic, structural/haemodynamic and functional outcome data collected ...... 67
Table 2.6: Mortality and arrhythmia data collected ...... 68
Table 2.7: Chronic disease status at time of referral for ASA ...... 82
Table 2.8: Peri-procedural complication of ASA in LHCH 2000-2011 ...... 88
Table 2.9: Cause of death during follow up after ASA ...... 91
Table 2.10: Two-way matrix of outcomes at completion of treatment according to echocardiographic and symptomatic data ...... 97
Table 3.11: Slice thickness and spacing of LGE stack for analysed scans...... 132
Table 3.12: Age and CMR measurements in those with success versus those with failure ...... 133
Table 3.13: LGE evaluation of post-ASA CMR scans ...... 137
Table 3.14: LGE evaluation post ASA with LGE from pre-existing scans removed...... 138
Table 4.15: Scoring system for section A: Mitral valve and SAM ...... 152
Table 4.16: scoring system for section B: Target septum ...... 153
Table 4.17: Scoring system for section C: Other key distant structures ...... 153
Table 4.18: Power calculations for echocardiographic assessment scoring system ...... 155
Table 4.22: Other structures visible as part of section C: Phase 1: All views ...... 163
Table 4.23: Agreement scores Phase 1: ICE work view ...... 166
Table 4.24: Agreement scores Phase 1: TTE work view ...... 167
Table 4.25: Agreement scores Phase 1: ICE all views ...... 170
Table 4.26: Key distant structures visible as part of section C: Phase 2: Work view ...... 172
29
Table 4.27: Mean scores ICE vs TTE Phase 2; work view analysis...... 172
Table 4.28: Mean scores ICE vs TTE Phase 2; work view analysis...... 173
Table 4.29 Agreement scores Phase 2: ICE work view ...... 175
Table 4.30: Agreement scores Phase 2: TTE work view ...... 176
Table 4.31: Agreement scores Phase 2: ICE all views ...... 179
Table 4.32: Agreement scores Phase 2: TTE all views ...... 181
Table 4.33 Agreement scores Phase 2: ICE work view ...... 183
Table 4.34: Observations on ICE following contrast injection ...... 186
Table 5.35: Scoring system for quality of CT images ...... 212
Table 5.36: Viewing angles for multiplanar reformat assessment: Pre-defined projections ...... 221
Table 5.37: Pre-defined territories for coronary artery supply ...... 222
Table 5.38: Summary of image quality scores in each of the 9 pre-defined domains...... 225
Table 6.39: Changes to standard ASA procedural details using information provided by CT ...... 256
Table 6.40 Summary of image quality scores in each of the 9 pre-defined domains...... 259
Table 6.41: Clinical outcomes pre- and post-CT guided ASA ...... 273
Table 6.42: CT guided ASA versus traditional ASA outcomes ...... 276
Table 7.43: SAM severity grading scale ...... 306
Table 7.44: Abnormal PFT triggers for referral to Respiratory team ...... 314
Table 8.45: Basic demographic and echo details of patients undergoing RF ablation ...... 336
Table 8.46: Procedural details ...... 348
Table 8.47: Measured parameters pre- and post-RF ablation ...... 351
30
Chapter 1: Current status of Non-surgical Septal
Reduction Therapy (NSRT) in Hypertrophic
Obstructive Cardiomyopathy (HOCM)
31
1.1 Introduction
Hypertrophic Cardiomyopathy (HCM) is an inherited disease characterised by otherwise unexplained hypertrophy of the myocardium. It is transmitted in an autosomal dominant pattern with variable penetrance, with an estimated phenotypic prevalence of 1 in 5001. Whilst the distribution of hypertrophy can be varied, involvement of the basal interventricular septum is common. The classical pattern of asymmetrical basal hypertrophy can narrow the left ventricular outflow tract
(LVOT) contributing to the pathology underlying LVOT obstruction (LVOTO). The prevalence of
LVOTO in HCM is 20-30% at rest2 and up to 70% with provocation3. LVOTO is associated with greater levels of dyspnoea, a greater incidence of stroke and higher mortality2.
Early approaches to the treatment of symptomatic hypertrophic obstructive cardiomyopathy
(HOCM) were surgical, with good results 4-6. Whilst surgical techniques have advanced and mortality in expert centres has improved, the principle of septal myectomy through direct visualisation and incision remains unchanged. The desire for a less invasive option with reduced morbidity and faster recovery times were all drivers for the development of a non-surgical solution. Such an intervention could also provide an option for those with high surgical risk precluding them from myectomy. Case reports of myocardial infarction had been reported to cause resolution of the clinical signs associated with HOCM7; creating a localised infarction percutaneously could therefore alter LVOT haemodynamics. Early work in this field began in 1983, when inflation of an angioplasty balloon in a coronary artery was noted to cause reduction in regional wall motion abnormalities8. Some success had been reported with the creation of an infarct using trans-coronary alcohol delivery to remove a ventricular arrhythmogenic substrate9; this was therefore adapted to target the basal septum. The pioneering non-surgical septal reduction therapy (NSRT) was performed with trans-coronary alcohol septal injection by Sigwart in 1994. A 68-year-old female with HOCM who failed to respond to
32 medical therapy and pacing underwent alcohol septal ablation (ASA) with a good result. Her case and 2 others were reported in 199510. The index case remained well 10 years later11. An alternative to surgery was now available.
Current first line treatment in symptomatic HOCM patients is the introduction of negatively inotropic medications such as beta-blockers, verapamil or disopyramide12 13. If this fails to resolve symptoms and LVOT gradients then septal reduction may be required, current options include surgical myectomy and ASA. Both of these are relatively rare procedures in the UK, with approximately 70-80 ASA procedures performed each year14 15.
33
1.2 Pathophysiology of LVOT obstruction in HOCM
Basal septal hypertrophy and systolic anterior motion (SAM) of the mitral valve (MV) are the key components to LVOT obstruction in HCM. The septal hypertrophy causes abnormal posteriorly directed flow through the left ventricle (LV)16 17. This flow circulates around the MV and back towards the LVOT, dragging the MV apparatus towards the septum. The MV is proposed to be abnormal as part of the HCM phenotype, with a variety of changes such as longer leaflets, abnormal papillary muscle architecture and anterior displacement of the apparatus. This anatomic predisposition means that the valve is more easily pushed towards the septum by the aberrant flow in the LV. The abnormal septal and MV anatomy causes a smaller LVOT area, in order for blood to pass through here the velocity must increase in accordance with the principle of mass continuity.
However, its static pressure must decrease in accord with the principle of conservation of mechanical energy. By these theories the pressure drop and SAM of the MV could only happen in association with high velocities. SAM starts before velocities are notably elevated during systole so this cannot be the sole cause of the anterior movement of the leaflets. The MV moves towards, and in severe cases contacts, the hypertrophied septum. Once mitral-septal contact occurs, the LVOT orifice is narrowed further and greater obstruction to flow develops, resulting in a higher pressure difference. This pressure difference forces the leaflet further against the septum, further narrowing the orifice, exacerbating the haemodynamic abnormalities. This establishes an amplifying feedback loop until systole is complete. The greater the length of time the AMVL is in contact with the septum, the higher the pressure difference16.
Obstruction results in a pressure drop from the LV to the aorta and the magnitude of the problem will vary with myocardial contractility and loading conditions. In the vast majority of affected patients the degree of obstruction increases with exercise and hence cardiac output is compromised
34 at times of increasing demand. This obstruction has multiple deleterious effects including reduction of forward cardiac output, mitral regurgitation of varying degrees, load dependent diastolic dysfunction leading to an increase in LV end-diastolic pressure and coronary flow abnormalities.
These factors contribute to symptoms of dyspnoea, chest pain, pre-syncope and syncope18-20.
Alcohol septal ablation (ASA) reduces the size of the basal interventricular septum by inducing a localised infarction with trans-coronary alcohol injection. The myocardium is stunned and hypokinetic in the acute phase. Coagulation necrosis leads then to scar formation and subsequent thinning of the myocardium. The LVOT therefore widens and the shape of the basal septum changes.
This alters the flow dynamics and directs blood out of the aorta rather than posteriorly towards the
MV. The reduction in systolic excursion of the septum into the LVOT (due to akinetic myocardium) also contributes towards the improved flow dynamics in systole. The effects can continue to improve for months afterwards. This can partly be explained by the change in diastolic function; the drop in
LVOT gradient (and therefore afterload) has been shown to reduce secondary left ventricular hypertrophy away from the basal septum and therefore cardiac mass21 22. This shows that part of the hypertrophy in HOCM is afterload dependent and not entirely genetically pre-determined.
35
1.3 Patient selection for NSRT
The safety and efficacy of NSRT can be improved with fastidious patient selection and procedural planning. The 2011 ACCF / AHA Guidelines for the Diagnosis and Treatment of HCM stipulate that there must be a subaortic gradient of at least 50mmHg, at rest or with provocation 13. This was further supported by the 2014 ESC guidelines on HCM12. The 2003 ACC/ESC consensus document on
HCM first set this gradient requirement 23, whilst others have suggested a resting gradient of
>30mmHg and a rise with provocation to >50mmHg22 24 or >100mmHg25 will suffice. It is important that provocation should be physiological rather than pharmacological, as dobutamine can produce subaortic gradients in the normal heart26. Treating a symptomatic patient with no resting pressure difference, but with exercise induced gradient can have a good outcome 27.
Symptom burden is a significant part of selection criteria. Patients must have limiting symptoms that are refractory to medical therapy 23. Dyspnoea classed as NYHA grade III-IV, and to a lesser extent chest pain and syncope are indicators for treatment. Medical therapy uses negative inotropes and can include beta-blockers, calcium channel blockers and disopyramide.
Patients should have significant hypertrophy of the basal interventricular septum thought to be responsible for causing an LVOT gradient. The 2011 ACCF/AHA guidelines simply state that septal thickness is “sufficient to perform the procedure safely and effectively in the judgement of the individual operator”13. ESC guidelines state that those with modest septal thickness (≤16mm) should be considered for treatments other than septal reduction due to the risk of developing a ventricular septal defect12. . Some quote interventricular width of >18mm with clear protrusion into the LV cavity28, and others >15mm29.
36
The National Institute of Clinical Excellence states that NSRT can be performed in HOCM patients with symptoms refractory to medical therapy. They make a requirement this should only be performed in specialist units with clinicians who have adequate training 30. This was re-enforced in the ACCF/ACC guidelines stating an operator must have experience of at least 20 procedures or work in a centre with a cumulative procedural volume of 50 patients13. The learning curve in ASA can require the performance of at least 40 procedures 31.
It must be clear before progressing to NSRT that no other indication for cardiac surgery exists. MV disease amenable to surgical correction and significant coronary artery disease (CAD) requiring bypass grafting must be ruled out. If CAD exists an individual approach can be adopted, and percutaneous intervention considered. In a small number of patients abnormal mitral valve anatomy or a sub-aortic ring will be present and responsible for a large part of the LVOT gradient. Iatrogenic infarct of the basal septum in these patients will not affect the haemodynamics and should not be performed; they are better suited to surgery.
Patient demographics play a significant role in treatment selection. As the long-term sequelae of inducing a myocardial scar is still under debate, careful consideration should be exercised when proceeding with young patients, ASA is generally not performed in children23 32-34. Given the choice, most patients will favour ASA over cardiac surgery35. This seems driven by the fear of incision, general anaesthesia and prolonged recovery periods. Patients do respond to guidance from medical personnel and must be aware that NSRT is not always the correct choice.
37
Many patients referred for the procedure do not need invasive solutions and can be treated with more intensive medical therapy. Thirty-seven per cent of 249 patients referred for septal reduction therapy (including myectomy) were satisfactorily treated with increased oral medication in one reported series from a tertiary referral centre36.
The generally agreed indications for proceeding to NSRT are highlighted in Table 1.
Indications for NSRT
LVOT gradient >50mmHg at rest or with provocation
Severe dyspnoea or chest pain (usually NYHA III-IV) refractory to medical therapy
Sufficient septal size to perform procedure safely (usually >17mm)
Surgical option precluded by contraindication or patient choice
Table 1.1: Indications for NSRT in patients with HOCM
38
1.4 Alcohol septal ablation: The procedure
The approach to the classic procedure was standard to most centres at the start of this period of research. Most operators historically favoured right femoral arterial access; however the radial artery has been shown to be a safe and viable option in Liverpool Heart and Chest Hospital and other institutions37. Simultaneous monitoring of LV and central aortic pressures can be performed using catheters placed at the LV apex and ascending aorta. This will give continuous LVOT gradient readings throughout the procedure. This usually requires two arterial access points. The LV catheter should be end-hole. Pigtail catheters have multiple portals; the proximal holes may lie above the level of obstruction and provide falsely low gradient readings. LVOT gradient should be measured during Valsalva manoeuvre and following an ectopic beat for Brockenbrough-Braunwald phenomenon. An alternative method of using a single arterial puncture and measuring continuous gradients using a central aortic catheter and a pressure wire sited in the LV apex is also possible.
Sometimes a high ectopic burden or runs of ventricular tachycardia prevent continuous monitoring of LVOT gradients.
A venous sheath should be sited to allow insertion of a temporary pacing wire at the Right
Ventricular (RV) apex prior to any injection. Septal pacing must be avoided as early activation of the septum can increase LVOT obstruction. Heparin is used to prevent thromboembolic complication. A coronary guiding catheter should be positioned at the ostium of the left (or very rarely the right) coronary artery and a guidewire advanced to the septal artery of choice. A softer wire has been proven to reduce coronary complication but occasionally a stiffer wire will be needed38. An over- the-wire balloon is then advanced into the septal artery; the balloon must be compatible with alcohol injection. A slightly oversized balloon is preferred to overexpansion of an undersized balloon.
Historically ASA has been performed by inflating the coronary balloon as close as possible to the
39 ostium of the septal artery. This was based on the concept that a proximal balloon position would allow alcohol to travel to a greater proportion of the septal vessel bed and reach a greater volume of myocardium, more damage equals a greater chance of success. It is important to ensure the balloon does not impinge on the parent epicardial artery, so as not to risk dissection or occlusion of the parent vessel. Even partial occlusion of the LAD can promote collateral flow from septal arcade systems to the mid or distal LAD.
Myocardial contrast echocardiography is used to demonstrate the area of myocardium supplied by the chosen vessel. Contrast should be injected with continuous echocardiographic screening. An obvious opacification of the area of the septum involved in the contact point for SAM will be seen if the artery is the correct choice. If myocardial contrast is seen elsewhere an alternative artery will need to be explored.
The guide wire is then removed and a small amount of angiographic contrast (1-2mL) injected into the septal artery beyond balloon occlusion. The test injection should ensure no reflux of dye into the epicardial artery. Careful inspection for collateral connection to distal territories should be performed. Alcohol injection into such arteries will result in poorly localised, unwanted and ineffective infarction.
An alternative approach without the use of myocardial contrast echocardiography relies on the change in gradient measurements with occlusion of the chosen septal artery. Septal arteries originating from the LAD are balloon occluded in a step wise manner. If a ≤30% gradient drop was observed an alternative vessel was chosen. A drop in the LVOT gradient of >30% with balloon
40 occlusion was used to identify vessels supplying the target area39. This method has largely been superseded since the introduction of contrast echocardiography.
Once the target artery has been satisfactorily identified a small volume (1-3mL) of absolute alcohol is injected slowly in small increments. Analgesia can be given as injection of alcohol can cause some discomfort. The volume of alcohol injected will depend on the area of myocardium opacified with contrast injection, thickness of the septum and perceived target volume of myocardium, and ECG conduction changes in the lab with injection (complete heart block may halt injection). There is evidence to suggest small doses (1-2mL) of alcohol can have equally effective haemodynamic results as large (2-4mL) doses40 41. Balloon occlusion should be maintained for at least 5 minutes. A final angiographic image is taken to show the septal branch no-reflow and rule out any unwanted coronary artery damage. The patient is kept under coronary care supervision for 24-48 hours with the TPW in-situ. If no complete heart block is present at that time it can be removed. Temporary permanent pacing systems can be used for up to 4 weeks post procedure to allow intrinsic conduction to recover and may avoid the need for long term permanent pacemaker. Hospital stay is
4-5 days if no complication is observed.
Changes in the methodology of ASA have not been as pronounced as those observed in other procedures in cardiology. The introduction of myocardial contrast echocardiography was a significant milestone that reduced procedural complications42 43. The major other significant change was the recognition that small doses of alcohol can be effective, allowing operators to secure the same haemodynamic effect with less risk41.
41
1.5 Results of ASA case series reports
1.5.1 Survival
HCM patients with resting LVOTO have a higher mortality than matched HCM patients without a gradient2. The rate of mortality at 30 days in the Euro-ASA study was low (1.2%), and the 7057 patient years of follow-up identified an all-cause mortality rate of just 2.42% per 100 patient years44.
Another large series of ASA patients studied for survival (n=465) over a mean of 8.4 years showed a
1, 5 and 10-year survival of 99, 94, and 90%. This compares favourably to that of the age and sex matched general, non-HCM population figures of 99, 93, and 84% respectively45.
Some series have suggested that removing the gradient may have a beneficial effect on survival. Of
173 ASA patients, with mean age of 64 years, followed up for a median time of 5.7 years, survival was no different to that of the general, non-HCM population46. The survival was identical to those treated with surgical myectomy in the same time period. Residual LVOTO was a predictor of mortality in this series. A residual LVOT gradient has also been shown to be a predictor of cardiovascular events after ASA47. A further study comparing survival following invasive treatment with ASA versus conservative management found a benefit with ASA. This was explained by non- cardiac death48.
These recent series will go some way to reassuring operators that removing the LVOT gradient with
ASA has a beneficial effect on long term outcome.
1.5.2 Risk of ventricular arrhythmia
The medium term risk of ventricular arrhythmia has been studied with largely reassuring results.
Some observational studies provide the support that the risk of sudden cardiac death (SCD) is no
42 higher than HCM patients without an iatrogenic myocardial infarction34 49 . A registry of all ASA patients with ICDs for primary prevention highlighted the annual incidence of appropriate ICD discharge as 2.8%32. A comprehensive registry of 465 ASA patients in Denmark and Germany identified the SCD risk over 8.4 years as 0.5% per annum, with 16 sudden deaths and 3 appropriate
ICD discharges (representing 2.7% appropriate ICD discharge rate)45. The majority of ICD series identify an appropriate discharge rate of 2.5-3% per annum post-ASA.
The multi-centre HCM ICD registry in North America highlighted an increased risk of appropriate ICD activation post-ASA. A 4-fold increase in appropriate ICD treatment was observed with event rates of
10.3% per year compared to 2.6% per year in those treated with myectomy50. The ASA group, however, consisted of only 17 patients and 4 events, this may be a type 2 error due to the small numbers studied.
Meta-analyses have not indicated any difference between ASA and septal myectomy in the medium term incidence of SCD or all-cause mortality51 52. The overall trend in SCD and VT risk in reported series appears to favour no clear increase in arrhythmia post ASA.
1.5.3 Symptom and gradient resolution
The first series of 18 patients to undergo ASA was reported in 199753. This was promising and showed that the procedure had potential to reduce LVOT gradients, reduce symptom burden and increase exercise tolerance. A larger series of patients were reported in 1999; 50 in total were followed up for 7 months39, showing a clear improvement in LVOT gradients, interventricular septal
43 size, VO2MAX and pulmonary artery pressures. Perhaps more importantly the change in patient symptoms was reproduced, with NYHA status improving from a mean of 3 to 1.7.
Following these breakthrough studies, multiple early patient series were reported, and summarised in a systematic review collated in 200654. In total 2959 patients were summarised from 42 published studies, although the authors accept some were duplicated by involvement in more than one report.
A baseline assessment revealed a mean age of 53.5 years, an NYHA class of 2.9 despite medical therapy, and peak LVOT gradients of 65mmHg at rest and 125mmHg on provocation. Mean follow up was 12.7 months. Early mortality (defined as within 30 days) was reported to be 1.5%, with ‘late’
(30 days to a mean follow up of 12.7 months) all-cause mortality at 0.5% (0-9.3%). NYHA status improved significantly to 1.2 (p<0.001). A reduction in angina burden was also seen; Canadian
Cardiovascular Society score reduced from 1.9 to 0.4 (p<0.001). Echocardiographic gradients reduced to 15mmHg at rest and 31mmHg with provocation (both p<0.001). Repeat procedures were required in 7% of patients.
A 2011 report from the Multicentre North American registry detailed procedures in 874 patients. All had a resting gradient of ≥30mmHg or provocable gradient of ≥60mmHg with advanced symptoms of exertional dyspnoea and/or angina despite medical therapy55. Mean follow up duration was 26 months. Seventy-eight per cent of patients suffered NYHA class III or IV dyspnoea prior to the procedure, with a gradient at rest measuring 70mmHg, and 99mmHg on provocation. Following ASA
72.5% of patients were classified NYHA class I, 23% NYHA class II, 3.9% class III and 0.65% class IV. A repeat procedure was required in 12.8% of patients. The average peak resting gradients post procedure reduced to 35mmHg. The Scandinavian multicentre study included 279 patients and reported similar results, with NYHA III/IV breathlessness reduced from 94% to 21%, and outflow
44 tract gradients falling from 58mmHg to 12mmHg. Those persisting with NYHA III/IV breathlessness had a high prevalence of co-morbidities including chronic obstructive pulmonary disease and valve disease56. The Euro-ASA registry represents the largest multicentre series to date (n=1275) and reported an improvement of NYHA status from 2.9±0.5 to 1.6±0.7, 11% had persisting NYHA class 3 dyspnoea. LVOT gradient improved from 67±36 to 16±21mmHg44.
A number of series of medium term results are now available, with time periods ranging from 25-
141 months38 39 41 46 49 55-60(see Table 1.2). Improvement in dyspnoea is observed in most, with a clear trend towards lower LVOT gradients. The procedure does not provide uniform improvement for all, however, with large recent series displaying a mean post procedure resting gradient of 35mmHg55 (a gradient some would claim justifies repeat treatment), and persistent NYHA class III dyspnoea in
21%56. There is therefore scope for improvement in the treatment and subsequent outcomes for
HCM patients with LVOTO refractory to medical therapy.
45
Table 1.2: Long term results from observational studies post ASA
1.5.4 Comparison to surgical myectomy
According to the 2011 ACC/ACCF guidelines surgical myectomy remains the treatment of choice for
HOCM with symptoms of heart failure. Surgery has a class IIa recommendation, ASA is class IIb13.
This is based on excellent survival, symptom resolution, abolition of gradient and low operative mortality at centres of excellence61. The ESC guidelines do not make such a distinction and simply recommend septal reduction in patients fulfilling criteria, recognising that often patients will be appropriate for one modality and not the other.
The enthusiasm for ASA over the past decade has resulted in fewer centres performing septal myectomy, especially in Europe62. The performance of myectomy in lower volume centres is associated with higher risk; the US SCAI registry data up to 2011 highlighted an operative mortality of 5.3%. Two hundred and forty two myectomy procedures were performed in 97 centres at a rate of 2.5 operations per centre, per decade63. This compares to a mortality of 0.5% in high volume centres61. It is possible to set up a dedicated service de novo and produce good results64, but occasional myectomy in otherwise experienced surgeons should be discouraged.
The 2012 report of the Mayo Clinic experience from Sorrajja et al described outcome up to 5.7 years in 173 patients undergoing ASA and an age and sex matched group undergoing surgical myectomy46.
The proportion of patients alive and free of severe symptoms in the ablation group was 78%, versus
73% in the myectomy group (p=0.26). Procedural mortality occurred in 2 ASA patients and 1 myectomy patient. Both patients who died in the ASA group presented in cardiac failure. Pacemaker
46 requirement was higher in the ablation group, 20.3% versus 2.3% in the myectomy group. The median 30-day LVOT gradient was better in the myectomy group, 5mmHg versus 11mmHg in the
ASA group, a small but statistically significant difference. This non-randomised series from a centre with a high volume of both ASA and myectomy shows that in appropriately selected patients the outcome from ASA can be comparable to myectomy with regard to survival and symptom resolution.
Meta-analyses have compared symptom burden, complication rate and mortality between ASA and myectomy. A meta-analysis evaluating non-randomised studies comparing ASA and myectomy at individual institutions identified 5 such studies, with 183 patients undergoing ASA and 168 myectomy65. The age of the ASA cohort was markedly higher at 54.4 +/- 6.3 years versus 45.0 +/- 4.4 years in the myectomy group. Baseline NYHA status and echocardiographic characteristics were comparable between groups. Symptom improvement was similar, NYHA reducing to 1.5 +/- 0.3 in the ASA group and 1.3 +/- 0.2 in the myectomy cohort. Reduction in LVOT gradient was slightly superior in the myectomy group, resting LVOT gradient after intervention was 10.8mmHg versus
18.2mmHg. The requirement for post procedural permanent pacing was higher in the ASA group at
18.4% versus 3.3% in the myectomy cohort65. An updated meta-analysis of comparative observational studies in 2010 had higher numbers with 380 patients undergoing ASA and 326 patients undergoing myectomy51. The data supports the previous conclusion that the risk of mortality was similar at short and long term follow up, and that reduction of symptoms was comparable. The only significant differences to come out of detailed statistical analysis were a higher incidence of pacemaker implantation and RBBB in the ASA group, and a small, yet significantly higher, residual LVOT gradient in the ASA group51.
47
Leonardi et al selected 19 ASA case series and 8 independent myectomy case series for analysis52.
The follow up period for ASA was significantly shorter and again the ASA population was slightly older (55 versus 44 years). Baseline echocardiographic data was similar but with a statistically larger mean interventricular width in myectomy patients of 23mm versus 21mm. After adjustment for baseline characteristics the odds ratio for treatment effect on mortality of all-cause and SCD in ASA was 0.28 (95% confidence interval (CI) 0.16 – 0.46). This compares to a ratio of 0.32 (95% CI 0.11 –
0.97) in myectomy. Although not statistically significant the outcome favoured ASA52.
The numbers in these comparative meta-analysis are much lower that the total number of ASA procedures performed worldwide and reported on54, and perhaps reflect the small numbers of centres actively performing myectomy. This undoubtedly has an effect on choice of treatment. If one has no centre to offer myectomy due to dwindling numbers of operators, the expertise is then in
ASA and this can become the safer and more effective option.
Meta-analyses of non-randomised studies and case series are the highest level of evidence we currently have to work with when comparing these two strategies of septal reduction. The prospect of a randomised comparison has often been raised, but now seems unlikely66. The patient population requiring septal reduction therapy is small. There is a low incidence of HCM in the general population, and only a small section of these patients have LVOTO refractory to medical therapy. It is also difficult to get patients to accept random assignment to treatment strategies that seem so different in terms of patient experience and recovery time. The very low event rate observed over medium-term follow-up currently would mean a study to detect differences in mortality would be extremely difficult to power.
48
1.5.5 Predictors of outcome in ASA
Ability to predict symptomatic outcome has been investigated with limited success31 67 68. This may be due to relatively low numbers. Reduced ejection fraction before ASA predicts mortality [53].
Systolic dysfunction often reflects an advanced stage in disease progression with an increased risk of poor outcome69.
A multi-variate analysis highlighted that patients undergoing ASA with three or more of the following features; gradient <100mmHg, septal hypertrophy <18mm, age> 65 and left anterior descending artery diameter <4.0mm had a superior 4-year survival free of death and severe symptoms68.
Operator experience of >50 procedures is an independent predictor of survival free of severe symptoms68. Those with a higher gradient at baseline and post-procedure, a thicker septum and a younger age at ASA tend to have a higher risk of unsatisfactory outcome and need for further procedure67 68 70.
1.5.6 Peri-procedural complications of ASA
Early series described a high rate of pacemaker implantation, with reports of 40% of patients requiring permanent systems39. The introduction of myocardial contrast echocardiography (MCE) has reduced this rate to around 10-15%55 56. MCE has a class I recommendation in the ACC/ESC guidelines23. Predictors of heart block are pre-existing first degree AV block, LBBB, lack of use of
MCE, injection of alcohol by bolus rather than infusion, injection of more >1 septal artery and female sex71. However, patients with long term pacing seem to have a similar outcome clinically and haemodynamically as those without pacing following ASA72.
49
The incidence of RBBB following ASA has been reported as up to 54%71. The high incidence of RBBB at ASA probably explains why LBBB is a predictor of complete heart block[73]. There have been reports of numerous cases of delayed heart block post ASA, therefore normal conduction at discharge does not necessarily mean implantation of a pacemaker will never be required73 74.
Other peri-procedural complications have a significantly lower frequency. Early series reported ventricular fibrillation in 2.2%, LAD dissection in 1.8% and pericardial effusion in 1.8%54. The improvement in complication rates in recent years may be attributable to refinements in procedure and operator technique, more recently peri-procedural ventricular arrhythmia was seen in 0.02% and LAD dissection in 0.01%55.
50
1.6 Alternative methods of NSRT
Alternative approaches to NSRT have been sought to overcome the complications and limitations of
ASA. An alternative substance for arterial injection is one approach. This is an attempt to reduce the risk of alcohol escape causing unwanted distal infarction. The immediate polymerisation of glue lends itself to more accurate and proximal distribution75 76. In a series of 18 patients no heart block, or any other complication, was observed76. Glue has a different effect on myocardium; it simply restricts blood flow causing ischaemia and subsequent infarction. Alcohol however causes tissue injury by direct necrotising effect, with acute dehydration and fixation of surrounding tissues. The early results from the Cyanoacrylate studies are encouraging, with a reduction in LVOT gradient from
75.8mmHg to 18.0mmHg, and a reduction in symptom burden with NYHA status falling from a mean of 3.1 to 2.2 at 6 months. Three of the 18 patients showed a recurrence of LVOT gradients at 6 months with associated symptoms, having shown a good response initially. Cyanoacrylate injection for septal reduction therapy may be appropriate for those with septal collateralisation to distal arteries, a cohort previously inappropriate for NSRT in the form of ASA due to risk of alcohol escape into undesired tissues.
The injection of microspheres to restrict blood flow has also been investigated. Initial experience with polyvinyl alcohol foam particles was reported in 2004 with encouraging initial results. Gradients reduced from 83 +/-32mmHg to 31+/-35mmHg with improvement in mean NYHA from 3.3 to 1.3.
Also of note was the lack of peri-procedural complication in 18 patients77.
The success of coil embolisation of non-coronary arteries led to the concept being adapted to treatment of HOCM78. A pilot study investigating feasibility of coronary artery coil embolisation was
51 performed in 20 patients with drug-refractory symptoms attributable to HOCM79. Eighteen patients suffered with class III NYHA breathlessness at onset. Ten patients improved to class I and a further 6 to class II, two found no improvement. The reduction in LVOT gradients was relatively modest compared to most ASA studies, 80mmHg to 36mmHg.
One person died during admission from a confirmed ventricular septal defect post-procedure and subsequent operative complications. This patient was proposed to have had a large myocardial infarction diagnosed by high enzyme release, thought to be the cause of the septal defect. No patient required implantation of a pacemaker79. This is encouraging with respect to the lack of conduction system damage, but the relatively small drop in gradients and peri-procedural death may counteract this in terms of overall attraction to coil embolisation as a treatment modality.
The rich collateralisation of the septum may explain the difficulties in effectively reducing long term gradients by ischaemia alone. Intra-coronary glue and coil embolisation have a similar methodology to covered stents and simple balloon occlusion in this respect. Covered stents are placed in the parent epicardial artery (usually the LAD) at the origin of the target septal supply to restrict blood flow and induce ischaemia. These therapies have not been adopted due to the limited impact on gradients long term.
A significant limitation of ASA is reliance on coronary anatomy to provide access to a target for ablation. Five to eight per cent of patients have no septal vessel suitable for injection, due to inability to identify or instrument an artery supplying the basal hypertrophy at the SAM-septal contact point70 80. It is possible to bypass this reliance by accessing the septum from the ventricles
52 and provide a direct treatment. A feasibility study in canines assessed the role of a combined intra- cardiac echocardiography (ICE) and alcohol delivery catheter[80]. The catheter was placed in the right ventricle in 9 canines, and the left ventricle in 3. ICE was used to guide the operator to insert a needle directly into the basal septum of the beating heart, once in-situ a micro-bubble was injected to ensure safe position of the needle tip in myocardium. A varying dose of alcohol was then injected
(0.5, 1.0, 1.5mL) and the canines observed for 3 hours prior to sacrifice. No arrhythmia developed during this time. The operators were able to observe the depth of injection in real time during the procedure with ICE. Histological analysis confirmed the scar to be in the desired area, and size of scar correlated to volume of alcohol injected[80]. Whilst showing promise the procedure has not yet been trialled in humans.
An alternative, more advanced approach for treatment is direct, endocardial radiofrequency ablation of the septum 81 82. Lawrenz et al performed endocardial radiofrequency ablation (RF) of the septum in a cohort of 19 adults, 8 of whom had ineffective ASA due to inappropriate coronary anatomy or spill over of contrast into collateral branches 81. All patients were offered, and declined, surgery. Ablation was performed from the RV septum in 10 patients and LV septum in 9 patients.
Although follow up imaging showed a minor reduction in the magnitude of septal hypertrophy,
22.6mm to 21.4mm, deep scars were visible up to 28mm within the septum on cardiac MRI in some.
This is in contrast to ASA where a more marked change in septal thickness in diastole is seen. Despite a modest effect on septal width a reduction of resting gradients of 77.7mmHg to 26.5mmHg, and provoked gradients 157.5mmHg to 64.2mmHg was observed. NYHA status reduced from a mean of
3.0 to 1.7. Four patients required permanent pacing and remained reliant at 6 months, one pericardial tamponade due to perforation from RV pacing wire required surgical correction81.
53
RF ablation has also been used in the management of LVOTO in children with success82. Thirty-two children mean age 11.1 (2.9-17.5) years underwent RF ablation from the LV. Symptomatic improvement was remarkable, prior to treatment; 88% suffered excess lethargy, 50% suffered breathlessness, 25% non-arrhythmic syncope and 9% reported anginal chest pain. Over a median follow up period of 48 months (3-144 months) all symptoms were resolved bar one child (3%) with excess lethargy. Echocardiographic gradients reduced from 97mmHg (30-144mmHg) to 32mmHg (0-
140mmHg) at most recent follow-up. One procedural death occurred. At cessation of ablation LVOT gradients were noted to have increased and were presumed to be due to tissue oedema at the site of ablation, left-ventricular dysfunction ensued and the patient died 3 days later. Two patients required pacemaker implantation, one of whom was not reliant one week later.
Reduction of septal thickening in systole is proposed to be responsible for the reduced gradient. This akinesia of the basal septum prevents narrowing of the LVOT in systole, reducing SAM and therefore obstruction. The flexibility of RF ablation appeals to physicians dealing with HOCM, and early results are comparable to results of early ASA series39. The improvement in outcomes during the early evolution of ASA and the learning curve for operators performing the procedure83 would suggest that outcomes in RF can also be improved rapidly.
54
1.7 Summary
NSRT in the form of ASA can provide relief of symptoms attributable to LVOTO in a substantial proportion of patients in an appropriately selected group. Whilst initially viewed as an alternative to myectomy in patients unfit for surgery, the balance of preference has changed for some operators 24
84 85. The complication rate has much improved and NSRT may now be more accessible as the number of specialist surgical centres declines. When ASA is performed successfully, medium term survival is comparable to matched HCM controls45 46. The risk of ventricular arrhythmia associated with an ASA-induced myocardial scar has not been proven to be significantly elevated versus non- treated HCM and myectomy patients32 34 49 50 86.
There remains however a significant proportion of HOCM patients for whom ASA does not provide any significant improvement, either symptomatically or haemodynamically. Despite the reassuring safety data and an increasing number of procedures performed internationally progress in ASA has stalled in recent years. Techniques have not advanced at the same rate as procedures in other fields of cardiology. The significant limitations in current practice could be improved by adapting approaches to ASA and incorporating alternative forms of NSRT into the care of HOCM patients.
1.7.1 What questions must we seek to answer?
1.7.1.1 What proportion of patients derives no benefit from ASA?
In order to understand the problem that we are facing a suitable method of describing longitudinal outcome of these patients must be designed. This database must then be used to accurately describe the individualised outcomes of all patients undergoing ASA. Liverpool Heart and Chest
55
Hospital has been performing ASA for 10 years prior to the initiation of this project, this will provide a suitable group of patients to populate this database.
1.7.1.2 Why do some ASA procedures fail to have the desired effect on LVOTO?
In those that have undergone ASA but still have a significant LVOT gradient we must consider:
Did the ASA procedure accurately damage the target myocardium?
Is there an alternative cause of LVOT gradient?
In order to answer these questions an imaging study must be designed to localise infarction in groups of patients with successful resolution of LVOT gradient versus those with a persisting gradient.
1.7.1.3 If ASA procedures fail because of inaccurate infarct; how can we improve?
Improved imaging and pre-procedural planning offers promise in the refinement of ASA.
Transthoracic echocardiogram images in the laboratory can be poor as the patient is lying prone. The operator and radiographer must also be careful not to irradiate the sonographer during image acquisition. Intracardiac echocardiography (ICE) produces images of high quality. Guiding ASA with
ICE has been suggested and warrants further exploration87. Better visualisation of the SAM-septal contact point, myocardial contrast and intra-procedural effects of balloon occlusion may help to refine localisation of infarct, improving long term outcome.
The only current method of identifying septal arteries for alcohol injection is invasive coronary angiography. Invasive angiography allows us only to detail the course of a coronary artery without
56 describing its interaction with myocardium. In ASA we rely on injection of echocardiographic contrast and ultrasound to highlight the area of myocardium perfused. Computed Tomography (CT) angiography can highlight the course of a septal coronary artery and show the area of myocardium supplied, guiding selection for alcohol delivery. This could improve accuracy of iatrogenic infarct and improve patient outcome.
Therefore we must consider:
Can intracardiac echocardiography provide superior images to transthoracic
echocardiography to guide ASA?
Can pre-ASA CT coronary angiography help us plan the procedure and guide alcohol delivery
to create a more accurate infarction of the target myocardium?
1.7.1.4 What if patients cannot receive trans-coronary alcohol as part of ASA?
A group of patients who cannot undergo ASA is likely to remain due to limitations of equipment design and anatomical variants. In various case series this represents 5-15% patients entering the lab with the intention of alcohol delivery70 80 88. RF ablation is used in electrophysiology labs to deliver myocardial damage and remove abnormal electrical circuits or areas of ectopy. This delivers shallow
‘burns’ in the myocardium. The technique has the flexibility of not being reliant on septal coronary anatomy, removing some of the restrictions of ASA.
Therefore:
What are the reasons for failure to deliver alcohol at ASA?
Can RF ablation provide a suitable alternative method of NSRT?
57
58
Chapter 2: The Liverpool Experience: Describing outcomes in patients undergoing ASA during the period 2000-2011
59
2.1 Introduction
In the search for improved outcome, a rigorous assessment of current performance will highlight areas for improvement and establish baseline values from which the impact of change can be assessed. Many series have sought to describe outcome following ASA38 46 49 55-59 80 86. In the main, results show an improvement in mean LVOT gradient and reduced symptoms of dyspnoea and chest pain. Most reports show that treatment is not perfect or uniform. There is very little reporting of individual patient outcome. Data are aggregated to provide summary statistics for the population, without mention of the outcome for individual patients (for example the proportion of patients with no or inadequate resolution of LVOT gradient). This is concerning as it may mask an unacceptably high failure rate.
Liverpool Heart and Chest Hospital (LHCH) first performed ASA in 2000. Many procedures had been performed up to 2011 with no method to report accurate ongoing audit or reporting of outcome data in the UK. During that time multiple international series had been reported, and despite ASA first being performed in London no large series had been reported from the UK. LHCH continues to be a relatively high volume operator with UK performance of ASA, although the national numbers amount to just 60-80 procedures per year (Figures 2.1, 2.2)14 15.
60
Figure 2.1: Total number of ASA procedures performed in the UK by year 2007-14.
A total of 66 procedures were performed in 2014.
Figure 2.2: Volume of ASA procedures by centre in 2014
(Red arrow is LHCH. BHL refers to LHCH; a previous moniker of ‘Broadgreen Hospital
Liverpool’).
61
According to British Cardiovascular Intervention Society audit returns LHCH performed 17 ASA procedures, accounting for 26% of procedures performed in the UK in the last reported year (2014).
Our case volume over the last 14 years would allow us to report a single-centre case series of reasonable size. Referrals came from a large area of the North of England and Scotland.
In the management of HCM, the collection of robust longitudinal data presents many challenges.
The spectrum of clinical manifestation (related to the heterogeneity and penetrance of the underlying genetic disorder) is wide and unpredictable. There is often a poor correlation between the anatomy of structural disease, measures of haemodynamic significance and patient symptoms.
Patient symptoms and limitations in HCM can be the result of cardiac and non-cardiac causes. A variety of therapies may be used over time and there is recognition that a proportion of patients managed with ASA will require repeat procedures for optimum clinical success. As a result patient characterisation is complex and will generally require the synthesis of information from a variety of observations and tests, often repeated over a prolonged natural history. Any database solution must be able to track clinical and adverse events in a longitudinal manner.
62
2.2 Aims
To develop, refine and populate a relational computer database to allow structured
collection, analysis and reporting of patient, investigation, procedural and outcome data
To examine and report the patient characteristics, procedural details and medium term
outcomes of a large, consecutive case series of ASA in HOCM in the UK
63
2.3 Methods
2.3.1 Database creation
2.3.1.1 Flat versus relational database solutions
It is largely recognised in clinical medicine that the ability to track outcomes is best served by a database with relational rather than flat-file design89.
Relational databases have several advantages in tracking longitudinal outcomes post-ASA. In a flat format a database consists of a single table of individual records, each comprising a number of data fields. In contrast to this a relational database allows data to be entered into a number of distinct tables. A single patient record can then be linked to multiple ‘episodes’. This is particularly crucial in the long term follow up of HCM patients as they can undergo more than one ASA procedure. This also allows us to measure outcome in multiple manners, of particular importance in this group is accurately reporting procedural complications along with symptomatic, haemodynamic, and functional outcome data.
Having understood that collecting data in a relational manner was required I chose Microsoft Access
2010 to create my database. This is standard market leading and easily available software. It is SQL compliant which is the NHS standard. It provided a good environment for new database development with online help and support groups. It has an excellent query by example interface to allow easier access to SQL language. It also includes tools for the creation of forms and publishing to the web. It also has excellent output formats for data sharing and for import to statistical packages.
64
Access has a weakness with very large databases (>20,000 records), data locking with simultaneous use by many individuals on a network and with some aspects of security when published to the web, this however was not perceived to be an issue for my intended use.
2.3.1.2 Data requirements
In order to describe outcome in a group of HOCM patients that have undergone ASA I had to first decide what data needed to be collected. A full list of all data fields collected in the database in available as an export from Microsoft Access in Appendix 2.1.
I first needed to describe my patient population. This required collection of basic demographics and baseline chronic disease status (see Table 2.3). Basic demographics and chronic disease status are important to understand the mortality data I collected. The presence of extra-cardiac morbidities also helped me to understand any persisting symptoms after ASA. I was not able to collect the presence of hypertension due to difficulties in getting information from patients notes, this is a limitation.
Basic Demographics Chronic disease status Name Diabetes Mellitus Date of Birth Lung disease Address/contact details Coronary artery disease Hospital ID Previous ASA at other institution Gender Previous cardiac surgery Ethnic origin Clinical heart failure
65
Table 2.3: Basic demographic and chronic disease status data collected.
I then needed to collect any relevant procedural data that may help explain why some procedures were successful and others failed.
Procedural data collection Invasive gradient measurement (+post ectopic) Number of procedures Use of myocardial contrast Alcohol delivery yes/no Reason for no alcohol injection Volume of alcohol injected CKMB enzyme release Complications
Table 2.4: Procedural data collected
Outcome is HOCM patients after ASA has traditionally been described by NYHA status and LVOT gradient derived by echocardiogram. Whilst these are valid tools for assessment some form of more objective measurement of functional status and symptom resolution are desirable. I therefore aimed to collect as much information about the status of each patient as possible from the retrospective nature of this study, whilst noting limitations for the prospective arms of the project. I divided the patient outcome measures into symptomatic, structural/haemodynamic and functional categories.
66
Symptomatic Structural/haemodynamic Functional NYHA status Echo LVOT gradient Peak VO2 CCS status Echo LVOT gradient Valsalva Exercise time Pre-syncope Echo LVOT gradient exercise Syncope Septal width SAM grade Left atrial size Mitral regurgitation grade
Table 2.5: Symptomatic, structural/haemodynamic and functional outcome data collected
One of the most robust methods of describing outcome from treatment is all cause mortality. This has its limitations but is a measure that is used throughout medicine for its reliability. I needed to be able to document this and describe cause of death if appropriate. Risk of arrhythmia is always relevant in a HCM population, and as some controversy exists as to whether ASA is proarrhythmic I decided to collect as much data as possible about presence of sudden cardiac death (SCD) or ventricular tachycardia. The inability to collect biomarkers of heart failure was a limitation of this study. No BNP data was routinely collected for any of these patients.
67
Mortality and arrhythmia markers Vital status Death – cardiac cause Death- non cardiac cause Sudden Cardiac Death Aborted SCD VT or NSVT – Holter monitor Presence of Implantable Cardioverter Defibrillator (ICD) or pacemaker VT or NSVT on device interrogation ICD treatment delivered
Table 2.6: Mortality and arrhythmia data collected
2.3.1.3 Relationships
In order to meet the challenge of describing the long term follow up of a series of patients undergoing a procedure that may be repeated the database needed to be able to record multiple, linked variables, episodes and events for each patient. Some variables can only be recorded once
(for example; disease state at time of referral) and require a one-to-one relationship ensuring no duplication. Others require multiple, time defined entries (for example; echocardiographic LVOT gradient) and are therefore linked with one-to-many relationships.
Microsoft Access automatically created a unique identification number for all patients assessed – named by us as the HOCMID. The first table I created is termed tblBasicDemographics, this acted as the central table in my relationships.
68
One-to-one relationships:
In a one-to-one relationship, a row in table A can have no more than one matching row in table B, and vice versa. I have created one-to-one relationships if the observation is a one-off (e.g. assessment of co-morbidity at presentation –tblChronicDisease) or to split basic demographics to promote confidentiality (identifiable details such as name are kept separate from the working demographics such as hospital ID).
Figure 2.3: One-to-one relationships
The HOCMID numbers are linked in each table in a 1-1 style with one-off observations of chronic disease status and further demographics to be kept separate.
One-to-many relationships:
This is a much more common occurrence in my database. In a one-to-many relationship table A can have many matching rows in table B, but a row in table B can have only one matching row in table A.
This is required for data entry points such as clinical or echocardiographic observations.
69
I created many tables linked to the tblBasicDemographics table via the HOCMID. In the first set of examples each of the linked tables provides an endpoint to the data entry. Although there may be many electrocardiograms per patient, I will collect all required data in one ECGID entry to this table.
The same is true for tblEvents, tblCPEx, tblHolterMonitor and tblObsStatus (see Figure 2.4).
Figure 2.4: One-to-many relationships with a single data entry record.
The one-to-many relationship is indicated on this diagram by the 1-∞ symbols.
Another variant of the use of one-to-many relationships is a further, secondary linked table. For example; each patient may have more than one ASA procedure, and each of those procedures may include multiple coronary injections. I therefore required a further level of flexibility. This is created
70 by using a one-to-many relationship using the unique patient identifier (HOCMID), and then using the second linked sheets unique ID (ProcID) to create a further one-to-many relationship (see Figure
2.5)
Figure 2.5: One-to-many relationship with multiple data entry records
In this example I have used the HOCMID from tblBasicDemographics to allow multiple procedures – each with a unique ID (ProcID) in tblProcedure. I then created a separate table for coronary injections (tblSeptalInjection) to allow me to add several injections per procedure. One patient
(HOCMID X) may receive 2 injections in procedure one and 3 injections in procedure two. This format is the best way to record when he/she had each of those 5 injections.
A full schematic of all table relationships can be seen in Figure 2.6.
71
Figure 2.6: All relationships in the retrospective ASA patient database
2.3.1.4 Dropdown menus
In order to remove the possibility of erroneous free text in cells I created a series of dropdown menus for all except one data collection point in each table (a free text ‘comments’ column). This approach creates uniform and structured data entry. This then ensures that searches and other manipulation of the database will produce consistent and reliable output. This was done by creating a separate table (termed ZRef) and creating a link in the lookup facility. Appropriate fields were allocated a ‘text’ data type. A ‘Combo Box’ option was chose and then the ‘Row Source’ directed to look at the lookup table ‘ZRefReasonNoAlcoholInjected’ (see Figure 2.7).
72
Figure 2.7: Process for creating dropdown menu I
I populated the ZRef lookup table ‘ZRefReasonNoAlcoholDelivered’ with the options seen in panel A of Figure 2.8. These options then became the only data that could populate that field in the septal injection table (Panel B Figure 2.8). Panel C shows an example of the data fields when viewing the septal injection table.
Figure 2.8: Process for creating dropdown menu II
73
Dropdown menus were created to facilitate data analysis. A ‘database documenter’ detailing the fields collected is presented in Appendix 1.
2.3.1.5 Queries
The query function within Microsoft Access allowed me to request data results from the database. I used simple select queries to retrieve data from a table. I then also used more complex select queries to answer questions that involved data in more than one table.
Simple queries
A select query allowed me to gather a count of events from a single table. In the example below I have used tblEvents to gather all procedural complications. In tblEvents I collected all adverse events with any relation to HCM, I then had a separate column denoting relationship to ASA procedure. This allows me to create a 3-column query with an event count and relationship to ASA
(‘ProcedureRelated’ - Yes/No/DataAwaited options). This allows me to accurately describe all complications (some procedures had >1 complication).
74
Figure 2.9: Simple query example – procedure related events
Complex queries
A query that required data from more than one table that was linked to an event and was time dependent became more complex. The relational nature of the database and a unique linked ID allowed me to do this. Symptom state was recorded in a longitudinal manner, with dates stipulated at each entry. I could therefore ‘query’ the database to tell me which entries were available either pre- or post-procedure, and could specify a date period from the procedure to assess outcome. An example is shown:
Aim: To identify long term (> 1 year) NYHA status after ASA.
Step 1:
Identify all observation status entries (‘ObsStatus’) >365 days since ASA procedure. ‘ObsStatus’ and
‘Procedure’ tables are linked according to HOCMID (unique identifier). A ‘datediff’ function is used to identify all symptom state entries >365 days since the NSRT procedure.
75
Figure 2.10: Creation of a complex query step 1
Composition of query in design view (left panel) and results following ‘run’ in datasheet view (right panel).
Step 2:
A list of entries is compiled by the ‘run’ function. I seek the most recent follow up data. I therefore run a query instructing Access to highlight the ‘max’ entry in days since the procedure. This provides me with a single entry per procedure.
Figure 2.11: Creation of a complex query step 2
Design view for step 2 of the query (left panel), and datasheet view following ‘run’.
Step 3:
The query results from step 1 and step 2 are then used to run a ‘query on query’ function. The
ProcID and HOCMID are linked. I can then search for the NYHA status of an individual patient at the
76 maximal follow-up after an ASA procedure. As some patients require more than one procedure I can review the symptom status at follow up after each procedure. Not all procedures have follow up for
>365 days, hence the need to add the caveat ‘is not null’.
Figure 2.12: Creation of a complex query step 3
Design view for ‘query on query’ in the left panel, results seen in worksheet view in the right panel.
Step 4:
The database structure allowed me to restrict results if required. In one analysis I only wanted to report on cases that received alcohol. In some cases a suitable septal artery could not be identified and no alcohol was given, I intend to highlight only those with alcohol delivered. This is done by linking the procedure table and adding ‘AlcoholDelivered’, and setting criteria to ‘yes’. I then run the query with the HOCMID.
77
Figure 2.13: Creation of a complex query step 4
Final design view and datasheet with last clinical review NYHA status
I then exported the data to IBM SPSS Version 20 for statistical analysis. Similar queries are written for short- and long-term symptomatic status (dyspnoea, chest pain and syncope), echocardiographic variables pre- and post-ASA, cardiopulmonary exercise test variables pre- and post-ASA and ECG details.
2.3.2 Patient selection:
All patients referred to LHCH for consideration of ASA from April 2000 to December 2011 were reviewed. A diagnosis of HCM was made according to typical clinical, electrocardiographic and echocardiographic features.
As all ASA procedures except one have been performed by the same operator at LHCH (and this procedure was done under the supervision of this operator). I was able to search this consultant’s
78 hospital records for suitable patients. All clinical letters and discharge summaries are kept in trust folders in Microsoft Word format. I was therefore able to search these folders for the following terms; ‘Alcohol septal ablation’, ‘Hypertrophic obstructive cardiomyopathy’, ‘Hypertrophic cardiomyopathy’, ‘Alcohol’, ‘Ablation’, ‘Left ventricular hypertrophy’, ‘Septal’. This consultant was acting as an interventionist and tertiary/quaternary referral point for ASA and did not run a general cardiomyopathy clinic. I sought the help of the clinical audit department in addition. They were able to search hospital databases for procedure with the same terms and identify anyone entering the lab for alcohol septal ablation using the catheter lab database tool TOMCAT.
One hundred and sixty four patients were identified in total. I personally reviewed clinical notes to decide if these patients were appropriate to enter in to the database. Fifty-eight patients did not have HCM, results were returned due to the presence of hypertension and left ventricular hypertrophy, comments about oral alcohol intake or previous alternative ablation procedures
(radiofrequency) for heart rhythm management. A further 18 HCM patients did not have significant
LVOT gradients to justify treatment, some of these patients had invasive pressure assessments, it was clearly stipulated that they entered the catheter lab for information gathering rather than with an option to proceed to ASA at the same sitting. Ninety potential subjects were identified. Two patients were excluded as they underwent alcohol ablation to the left anterior descending artery in an attempt to treat diastolic dysfunction and improve diastolic filling, treatment was not aimed at
LVOT gradients.
79
Total patients identified from key word search 164
Not HCM 58
Not taken to the lab 18
Additional identified from TOMCAT 2
Not LVOT gradient treatment 2
Final n 88
Figure 2.14: Patient identification for entry into retrospective ASA outcome database
Eighty-eight patients were therefore included.
2.3.3 Statistical methods
Continuous variables are presented as mean ± standard deviation for parametric data or Median ±
IQR for non-parametric data. Statistical analysis was performed using SPSS version 20 and Microsoft
Excel 2010. NYHA class changes and LVOT gradients were evaluated using Wilcoxon signed rank test as they had a non-parametric appearance, peak VO2 was normally distributed and hence a paired t- test was used. Assessment of volume of alcohol injected as a risk factor for complete heart block was assessed using Mann-Whitney U test. Pearson’s correlation was used to compare alcohol volume injected to CKMB release. All tests were two-sided and a p value of 0.05 was assumed to suggest statistical significance.*
80
2.3.4 Ethical approval
Appropriate ethical approvals to investigate outcome from ASA as an official audit were granted by the Research Committee at Liverpool Heart and Chest Hospital (see Appendix 2.2).
81
2.4 Results
2.4.1 Patient details:
Mean age of the group was 60.3(±14.3) years, 52% female. All had resting or exercise stress peak
LVOT gradient ≥50mmHg and basal interventricular septal diameter >15mm. All were trialled on negative inotropes prior to ASA.
Twenty-one patients had underlying lung disease, 3 had suffered previous CVA. Ten patients had undergone previous percutaneous coronary intervention and one patient had previous ASA at another centre with poor outcome.
Chronic disease n (%) COPD 13 (15%) Previous PCI (CAD) 10 (11%) Asthma 6 (7%) Diabetes Mellitus 3 (3%) Previous CVA 3 (3%) Peripheral arterial disease 2 (2%) Pulmonary fibrosis 2 (2%) Previous ASA elsewhere 1 (1%) Previous Cardiac surgery 1 (1%) (AVR)
Table 2.7: Chronic disease status at time of referral for ASA
82
2.4.2 Procedural details:
All procedures except one were performed by the same operator. A patient was taken to the catheterisation laboratory with the intention of delivery of alcohol on 125 occasions. Alcohol was delivered in 109 procedures (87%). The mean volume of alcohol delivered was 2.24 (±1.09)mL. Four patients could not receive alcohol at any procedure due to limitations identifying or instrumenting appropriate septal vessels. Reasons cited for withholding alcohol injection were; lack of angiographically identifiable septal, poor localisation of myocardial contrast on investigation or inability to access septal artery due to limitations of equipment (Figure 2.15).
Twenty-four patients (27%) required 2 or more procedures due to unsatisfactory outcome. Sixteen patients (18%) required two procedures. One patient did not receive alcohol at procedure 1 due to equipment limitations and subsequently received alcohol at procedure 2. The other 15 patients received alcohol at procedure 1 but progressed to a further procedure because of unsatisfactory outcome. Twelve of these 15 patients received a further dose of alcohol at procedure 2, 3 patients did not have an identifiable septal artery by angiography to explore. A further 7 patients had failure to resolve LVOT gradient satisfactorily and went for myectomy (3) or opted against a further procedure (4).
Seven (8%) patients returned for a third procedure, all had received 2 doses of alcohol previously but returned due to unsatisfactory outcome. Two of seven received a third dose of alcohol. Five patients could not receive alcohol due to lack of angiographically identifiable septal artery (3) or inaccurate localisation of myocardial contrast on injection (2).
83
One patient underwent four procedures. He received only one dose of alcohol. After failure to deliver alcohol at follow up procedures covered stents were delivered to the LAD.
84
1
1
5
0
1
2
2
0
3
1
0
0
2
4
1
3
0
88
No further proc Nofurther
1
1
1
2
2
0
1
1
0
2
3
3
0
Covered stent Covered
Covered stent Covered
No gradient with provocation
Unable to Unable septal access
Poor contrast localisation
No angiograhic septal identified
No gradient with provocation
Unable to Unable septal access
Poor contrast localisation
No angiograhic septal identified
Unable to Unable septal access
Poor contrast localisation
No angiograhic septal identified
1
0
1
5 4
Figure 2.15: Patient flow diagram detailing procedures and reasons for no alcohol delivery
1
2
6
Noalcohol
Noalcohol
Noalcohol
1
0
7
2
22
n = 88 n
Noalcohol
Proc1
0
2
18
82
n = 1 n
n = 8 n
n = 24 n
Proc4
Proc3
Proc2
Alcohol
Alcohol
Alcohol
Alcohol
2
13 60
No further proc Nofurther
85
2.4.2.1 Alcohol dose to CKMB correlation:
Creatine Kinase – MB was the cardiac enzyme used to measure myocardial damage after ASA in
LHCH (upper limit of local reference range =5). This was routinely measured at 12-16 hours post alcohol injection. In an ideal treatment the volume of alcohol injected determines the amount of damage delivered, this would therefore have a linear correlation.
The relationship between alcohol dose and CKMB release was examined by Pearsons correlation. In
92/109 (84%) procedures we had complete data for volume of alcohol injected and the subsequent
CK-MB release. The mean volume of alcohol delivered was 2.24 (±1.09)mL. Correlation was poor. R2 was low at 0.02 with a p value of 0.14 (see Figure 2.16).
y = 24.014x + 103.57 Alcohol dose to CKMB correlation R² = 0.0208 p=0.1434 700
600
500
400
300 CKMB release 200
100
0 0 1 2 3 4 5 6 Alcohol dose mL
Figure 2.16: Correlation of alcohol dose to CKMB release
86
2.4.2.2 Peri-procedural complications:
New complete heart block (CHB) was observed in 14/84 (17%) of patients. Permanent pacemaker
(PPM) implantation was required in 14 of 74 patients without pre-existing cardiac rhythm management devices. Those with pre-existing device had implantable cardioverter defibrillator (ICD) for primary prevention (6), or attempted right ventricular apical pacing with permanent pacemaker to treat LVOT gradients (4).
No new CHB was observed in those with a pre-existing device. The incidence of CHB reduced in later procedures, 9 of 14 occurred in the first half of the series, versus 5 of 14 in the latter half. There was a trend towards higher volumes of alcohol injected in those that developed CHB, 2.75 (±1.68) mL versus 2.16 (± 0.95) mL, but this did not reach statistical significance (p=0.24).
There were no procedural deaths. One inpatient death was reported following haemodynamic compromise as a complication of pacemaker implantation post ASA. VF was observed 3 hours after ablation in one patient, DC cardioversion was successful. Repeat angiogram revealed no reflow in the target septal artery only with no other changes to coronary flow. Distal infarction of the inferior wall was observed on echocardiogram 6 months post procedure in one patient. Pericardial effusion without tamponade was seen in one patient. This did not require further treatment. New bundle branch block was observed in 71% of patients without pre-existing CHB or paced rhythm or post- procedural CHB (n=68). There was a clear predominance to right bundle branch block (42/68) over left bundle branch block (6/68).
87
Complication Incidence (%) New Right bundle branch block 42/68 (62%) New CHB 14 (17%) New Left bundle branch block 6/68 (9%) Inpatient death 1 (1%) Inpatient ventricular arrhythmia 1 (1%) Distal unwanted myocardial infarction 1 (1%) Pericardial effusion without tamponade 1 (1%)
Table 2.8: Peri-procedural complication of ASA in LHCH 2000-2011
2.4.3 Clinical outcomes
2.4.3.1 Length of Follow up
Follow up data is presented for the 84 patients that received alcohol over a mean period of 4.20
(±3.33) years from the index procedure (range 0.13-12.29 years). Not all patients remained under the follow up of LHCH at the time of assessment. Some travelled from Scotland, the Isle of Man,
Northern Ireland and Yorkshire so ongoing follow up was taken on by the local cardiologists. For those that were not under local follow up I wrote to local cardiologists and GPs requesting an update. I also searched for mortality status for those that lived in England; this facility was not available for those from Scotland and Northern Ireland.
88
Patients entering the lab for ASA 88
Did not receive alcohol 4
Mortality status 84
No Satisfactory symptom data 2
Symptom status 82
No satisfactory echo LVOT data 8
LVOT gradient status 74 *
No satisfactory cardiac dimensions 11
Cardiac dimension status 63
No satisfactory CPEX data 39
CPEX data assessment 24
Figure 2.17: Follow up data available.
* This group of 74 patients represent the retrospective cohort with complete gradient and symptom data. This will be used as a comparator for the prospective group in chapter 6.
89
2.4.3.2 Survival and risk of ventricular arrhythmia:
Fifteen of the 84 (18%) patients that received alcohol died. One, 2, and 5 year survival rates were
96(n=68), 93(n=54) and 84%(n=33) respectively (see Figure 2.19).
Figure 2.18: Kaplan-Meier survival curve for patients undergoing ASA 2000-2011
No patient that received alcohol suffered sudden cardiac death (SCD). Nine suffered cardiovascular death. One inpatient death was reported due to complications of PPM insertion post ASA. Heart failure was the cause of death in 6; 4 of these had clinical evidence of decompensated heart failure with preserved ejection fraction before ASA. Myocardial infarction was reported as cause of death 9
90 years after ASA in one. Stroke was cause of death in one; this patient was in sinus rhythm. Non- cardiac death was reported in 6.
Age Time from at ASA until Source of HOCMID ASA death info Type of death Cause of death 1a Myocardial Infarction, 1b Coronary 55 55 8.74 Coroner Cardiac atheroma, 2 Alcohol abuse 1a Hypertrophic obstructive 57 69 0.01 LHCH notes Cardiac cardiomyopathy Local 58 73 1.14 Cardiologist Cardiac Congestive Cardiac Failure, renal failure 62 54 0.71 GP Cardiac Congestive Cardiac Failure 67 67 3.52 GP Cardiac Congestive Cardiac Failure 76 75 5.95 Coroner Cardiac 1a Congestive Cardiac Failure 79 67 4.54 GP Cardiac Acute left ventricular failure 1a Congestive Cardiac Failure 1b Hypertrophic obstructive 91 54 0.82 LHCH notes Cardiac cardiomyopathy Local 84 69 5.80 Cardiologist Cardiovascular CVA Local 13 70 2.74 Cardiologist Noncardiac Cancer 15 69 6.42 Family Noncardiac Metastatic bowel cancer 21 76 8.89 Family Noncardiac Lung cancer 60 72 1.54 GP Noncardiac Pulmonary fibrosis 68 64 4.15 LHCH notes Noncardiac Pulmonary fibrosis Peri-operative (AVR) complication - 90 75 3.48 LHCH notes Noncardiac sepsis
Table 2.9: Cause of death during follow up after ASA
One episode of sustained VT was seen 2 weeks after ASA. This was reviewed by independent electrophysiology specialists. Analysis of 12 lead ECG recordings suggested that the abnormal rhythm did not have a septal origin. An ICD was implanted. Fifteen patients had an ICD; 6 were in-
91 situ prior to ASA. No appropriate therapy (shocks or anti-tachycardia pacing) was delivered. Mean follow-up period between ASA procedure and last arrhythmia check was 3.08 years (±2.77).
2.4.3.3 Impact on symptoms:
Eighty-two of 84 patients that received alcohol had satisfactory follow-up data for clinical assessment. NYHA dyspnoea class pre-ASA was 2.80 (± 0.46), improving to 1.92 (± 0.84) post-ASA
(p<0.0001). Two (2%) of patients were in NYHA class 4 at presentation (in clinical heart failure),
62/82 (76%) were in class 3 and 18/82 (22%) were in class 2.
Fifty-nine of 82 (71%) patients had an improvement in dyspnoea of at least one grade; 43/82 (51%) improved by one NYHA class, a further 16/82 (20%) improved by 2 NYHA classes. Twenty-seven per cent reported no change in symptoms of dyspnoea. Two patients deteriorated from class 3 to class 4
(see Figures 2.20 and 2.21).
Pre-ASA NYHA class Post-ASA NYHA class 70 62 70 60 60 50 50
40 40 30 30 n 30 n 30 18 20 20 20 10 0 2 10 2 0 0 1 2 3 4 1 2 3 4 NYHA class NYHA class
Figure 2.19: NYHA class pre- and post-ASA
92
Figure 2.20: Change in NYHA class following ASA
2.4.3.4 Cardiopulmonary exercise testing:
There was no systematic use of CPEX either pre or post procedure. Often it was used to investigate persisting dyspnoea; there is therefore the potential for bias in interpreting these results. A standard bicycle ergometer protocol was used with increments of 10W added every minute. A respiratory exchange ratio (RER) of ≥1.1 was used to signify a satisfactory test. Only those with pre- and post-
ASA testing and satisfactory data were included.
93
Twenty-four patients were identified. Peak VO2 increased from 18.90 (±4.45) to 20.09 (±5.73) ml/min/kg (p=0.018), exercise time increased from 568 (±214) to 615 (±216)s (p=0.046).
2.4.3.5 Echocardiographic parameters:
Basal septal diameter
Basal septal diameter in diastole decreased from 22.35mm (± 5.08) to 17.27mm (±4.25) (p<0.0001) in 63 patients for whom data was available.
LVOT gradients
Seventy-four of 84 patients had satisfactory pre- and post-ASA echocardiographic assessment of
LVOT gradient. Those with a resting gradient ≥50mmHg and those with a resting gradient <50mmHg but exercise stress or Valsalva manoeuvre gradient ≥50mmHg will be discussed separately. Very few patients that had a resting gradient >50mmHg underwent stress testing as they already had an indication for treatment. I am therefore unable to accurately report the rest and exercise gradient in all patients. Assessment of efficacy in those that required exercise provocation to unmask a significant gradient after treatment was made under the same conditions. Amalgamating rest and stress gradients in reporting can be misleading so outcomes will be separated.
Resting gradient ≥50mmHg:
Sixty-one patients had a pre-procedural resting gradient of ≥50mmHg. Peak gradient was
99.80mmHg (±45.86), median gradient was 90mmHg (IQR 70-121). Following treatment peak gradient fell to 23.77mmHg (±41.87) (p<0.0001), median was 10mmHg (IQR 0-24) (see Figure 2.22)
94
Resting gradient <50mmHg with exercise stress or Valsalva gradient ≥50mmHg:
Thirteen patients were identified. Pre-procedure resting peak gradient was 27.50mmHg (±14.0), peak stress gradient was 102.00mmHg (±49.57), median of peak stress gradients was 89mmHg. Post-
ASA stress and Valsalva gradients were grouped, as Valsalva has been shown to be equivalent to stress echocardiography post-ASA(20). Post-ASA peak stress gradient was 16.92mmHg
(±30.65)(p=0.0005), median value was 0.
95
Figure 2.21: LVOT gradients pre- and post-ASA
2.4.4 Assessment of failure of treatment:
2.4.4.1 LVOT gradient:
There are no established criteria for procedural success. For the purposes of this study, failure is defined as a persisting gradient of >50mmHg (the generally accepted level at which intervention is advocated) or failure to reduce the gradient by greater than 50%.
Pre-procedure resting gradient ≥50mmHg:
ASA was considered a success in 49/61 (80%) patients. Failure to reduce the gradient to <50mmHg was noted in 9 (15%), treatment did not reduce the gradient by greater than half in a further 3 (5%) patients.
Pre-procedure resting gradient <50mmHg with exercise stress or Valsalva gradient ≥50mmHg:
ASA was considered a success in 12/13 (92%), one failure (8%) to reduce the gradient to <50mmHg was observed.
Failure rate:
The overall failure rate in resolution of LVOT gradient at the end of ASA treatment options was therefore 13/74 (18%). Twenty four (32%) required a second procedure.
96
The failure rate after one procedure was 31/74 (42%), as 7 further patients progressed to surgical myectomy (3) or opted against a further procedure (4).
2.4.4.2 Dyspnoea:
Failure to improve symptoms of dyspnoea was defined as failure to improve NYHA status ≥1 class.
Fifty-nine patients (71%) improved by one or more NYHA class. Twenty-one (27%) found no difference, 2 (2%) deteriorated from class 3 to 4. The failure rate was therefore 29%.
2.4.4.3 Combining clinical and echocardiographic outcome measures:
Table 2.10 displays a 2-way matrix of combined outcomes in the 74 patients with complete echocardiographic and symptomatic follow up data. Sixty-six per cent of patients achieved a satisfactory outcome by both clinical and echocardiographic parameters. Eleven per cent fail by both clinical and echocardiographic criteria. Seven per cent achieve some improvement in symptom status (all improved by one NYHA category) with limited change in LVOT gradients. Sixteen per cent achieve resolution of LVOT gradient without change in symptom burden. Half of this category suffered significant lung disease, 3/12 had COPD, 3/12 had pulmonary fibrosis.
n=74 LVOT gradient
n (%) Success Fail
NYHA Success 48 (66) 5 (7) 53 (73) Fail 13 (16) 8 (11) 21 (17)
61 (82) 13 (18) 74 (100)
Table 2.10: Two-way matrix of outcomes at completion of treatment according to echocardiographic and symptomatic data
97
98
2.5 Discussion
2.5.1 Risk of ASA
ASA appears to have few serious complications. The risk of complete heart block necessitating pacemaker implantation remains around 15%. Lower doses of alcohol and increased operator experience appear to reduce this risk. Whilst the ramifications of a pacemaker should not be dismissed, we must bear in mind right ventricular apical pacing can be effective treatment to reduce
LVOT gradients in some. Others require ICD implantation due to risk of SCD. The use of cardiac rhythm management devices in this group may have an independent value and is not necessarily a classic ‘complication’.
The fears regarding the pro-arrhythmic risk of an iatrogenic infarct do not appear to have been substantiated in medium term follow up. No significant burden of VT was seen in those with internal monitoring systems in our group. Many other studies have come to similar conclusions33 46 49 86, whilst a smaller number claim and increased risk of ventricular arrhythmia90.
2.5.2 Success and failure of traditional ASA methods
ASA improves LVOT gradients in most; but 13 of 74 patients still had significant gradients after the completion of ASA, including multiple procedures. Three went on to have surgical myectomy; others were unwilling to submit to cardiac surgery or were unsuitable because of co-morbidity. Four patients did not receive alcohol and are not included in this number. They must also be deemed as failure as they did not receive intended treatment.
99
The failure rate after one procedure is much greater, with 31/74 (42%) failing to achieve adequate resolution of LVOT gradients.
A further 12 patients had technical success in resolution of gradient by the specified criteria but reported no improvement in dyspnoea. Half of this cohort has significant lung disease. Our population is representative of a real-world patient group, in which symptoms are often multi- factorial. Diastolic function can be improved by removal of the LVOT gradient and subsequent regression of afterload dependent hypertrophy91 92, but persisting diastolic dysfunction could account for a proportion of these on-going symptoms.
2.5.3 Does aggregate outcome data mask a disappointing failure rate in individual patients?
The failure of treatment for a significant proportion of patients could be easily overlooked in a classic aggregate analysis, reporting pre- and post-treatment mean values. The outcomes in our group when reporting in this manner are comparable to most medium term series published (see Table
1.2). When aiming to prove a treatment is efficacious it is attractive to report that mean LVOT gradients improved from 99.9mmHg to 23.3mmHg (p<0.0001) post-ASA. It is less attractive to hear that 5% could not receive the desired percutaneous treatment option, 42% failed to achieve satisfactory resolution of LVOT gradients after one procedure and 18% fail despite repeat intervention.
There is a continuing debate about the relative merits of surgical myectomy and ASA. Many centres cannot or do not offer myectomy and ASA is now the dominant option in terms of procedure
100 numbers62. Due to the relatively low international procedural volume of both interventions it would be extremely difficult to perform a prospective randomised controlled trial to establish superiority in reducing mortality66. We must therefore accept it is likely that cardiologists will continue to offer
ASA as a treatment option, often in preference to myectomy. Nineteen centres in the UK offered
ASA in 201114, and 17 centres performed ASA in 201415. As a result we must strive to improve ASA to provide satisfactory treatment to a greater proportion of patients.
The procedure went through an initial phase of rapid evolution. Over the last decade, a number of case series have reported similar results. It would appear that progress has stalled. There have been few substantial advances in procedural technique over recent years.
2.5.4 Why does ASA fail to resolve LVOT gradients in some?
The reasons for failure in those that receive alcohol as part of ASA have never been adequately investigated.
2.5.5 What limitations do we face with ASA?
It is difficult to perform predictable, high quality alcohol ablation due to inherent limitations associated with current techniques.
101
2.5.5.1 Identification of septal artery targets:
We have very little information about myocardium supplied from invasive angiography. We rely on myocardial contrast injection into a potential target vessel and subsequent visualisation of territory supplied from echocardiography; a process which depends on an initial selection from angiographic images alone. The optimum vessel may not be recognised, particularly if it is small or originates from an epicardial vessel other than the LAD. Five per cent of our study group and 5-15% of patients in reported series do not receive alcohol as no septal vessel can be identified or instrumented70 80 88.
Some areas highlighted by contrast may be close to the target area of myocardium but not ideal.
These may be accepted as the best available option. This is less likely to have a significant long-term effect on LVOT haemodynamics.
2.5.5.2 Technical instrumentation:
Some arteries cannot be injected as the operator is unable to safely balloon occlude the vessel.
Accessing the artery with coronary wires can be difficult if many turns need to be negotiated, and the operator cannot introduce contrast or alcohol if the balloon kinks on bends.
2.5.5.3 Difficulty controlling infarct size:
It is difficult to judge the size of myocardium supplied by the target vessel, and alcohol dose and injection rate are always an estimate. This is compounded by unpredictable run-off, venous drainage and hence tissue dwell time and absorption. This is exemplified by the poor correlation between alcohol dose injected and CK-MB release observed in this data (see Figure 2.16). The volume of alcohol to be delivered can also be restricted by conduction system disturbance.
102
2.5.5.4 The need for precision
In the ideal procedure we seek to impact a very small target zone and need very precise localisation.
An infarction missing the ideal location by as little as 3 mm in any direction may render the procedure of no long-term value.
2.5.6 Prospects for improvement:
2.5.6.1 Procedural changes:
A better method of visualising myocardial contrast after coronary injection might allow the operator to be more secure in the assertion that contrast localisation is correct and the choice of septal coronary artery is appropriate. On table transthoracic echocardiography poses many problems and can often lead to substandard quality images.
There is significant promise in the prospect of computed tomography guided ASA93. CT offers the dual benefit of viewing angiography and structural detail, describing the course of septal vessels supplying the target area. This has the potential to change the approach to an individual procedure and target vessels with increased accuracy.
Radiofrequency ablation in HOCM has been performed81. This has the advantage of independence from coronary anatomy. This method of septal reduction warrants further investigation.
103
2.5.7 Limitations of the analysis
As with any retrospective study there are significant limitations that need to be borne in mind when interpreting this data.
2.5.7.1 Control of outcome assessment
Symptoms
I was reliant on clinic letters and GP correspondence for a significant proportion of symptom assessment. Often I would collect an NYHA status based on a remote assessment by a doctor unknown to me; this would represent my most accurate method of data collection. This holds inherent risks which are difficult to overcome.
The NYHA class system of symptom reporting is rigid. This often makes it difficult to describe magnitude of improvement (or deterioration). The comparative statistical analysis also assumes that each class improvement (or deterioration) is equal. It is conceivable that moving from NYHA class 4 to 3 may bring more gain than 3 to 2, or even 2 to 1. No quality of life questionnaires were completed for this group of patients. Options would have included EQ5D-5L (used later) or the
Minnesota living with heart failure questionnaire.
104
2.7.1.2 Echocardiographic data
Labile gradients
Even in the most controlled situations LVOT gradients are labile94. Measurements vary with the hydration status of the patient, respiratory variation, room temperature and adrenergic drive95.
Acquisition bias
I could not control the sonographers’ awareness of the patient’s symptomatic outcome. If a patient tells the sonographer they feel better there is a possibility of the scan being acquired to fit the patients story; a cognitive bias.
There is an element of intra-observer variability when assessing Doppler signals in echocardiographic assessment96, a small difference in a measured velocity is magnified when the modified Bernoulli equation is applied.
LVOT gradient data:
Gradient values are a continuous variable with the potential for absolute change of reasonable magnitude. This facilitates the demonstration of change with statistical significance when perhaps the clinical impact and significance of change in symptoms may be less dramatic and certainly will not be uniform.
105
2.7.1.3 Event reporting;
All-cause mortality is a reliable adverse event outcome measure. Other outcomes beyond that carry some risk of reporting bias. The reported cause of death has, in some cases, been derived from third party reports and there may be may be some inaccuracy.
I was able to safely comment on the incidence of ventricular arrhythmia and ICD therapy in the patients that remained under follow up at LHCH as device interrogation reports were readily available. Cardiac arrest and ventricular arrhythmia reporting from patients no longer under follow up was more difficult. This relied on remote doctors providing me with information, and assumed they were up to date with the most recent events.
2.7.2 Loss to follow up
Patients that were no longer under the care of LHCH had less accurate clinical details, this is inevitable. The nature of the referral pattern meant that a significant proportion had to revert to local follow up due to distances travelled for treatment.
106
2.6 Summary
ASA has an unacceptably high failure rate.
The reasons for failure are not well understood.
The procedure has many limitations; some are unavoidable, others can be improved.
ASA appears safe.
107
Chapter 3: An exploratory study of the relationship between iatrogenic infarct location and the impact on
LVOT haemodynamics post-ASA
108
3.1 Introduction
The case series, presented in Chapter 2 suggests that an unacceptable proportion of patients undergoing ASA had significant, persisting LVOT gradients. A failure to resolve obstruction is likely to be associated with persistent symptoms and a less favourable prognosis 2 47 68. In order to better understand how we can improve LVOT gradient resolution and symptomatic outcome I must first suggest mechanisms that might explain the inconsistent efficacy of ASA.
3.1.1 Mechanism of myocardial infarction in ASA
In 1982 Come and Riley observed that clinical signs of obstruction in a patient with HOCM disappeared after myocardial infarction7. They proposed that the myocardium in the basal
‘subvalvular’ septum area was affected and became akinetic, thus affecting LVOT haemodynamics.
This lead to studies involving balloon occlusion of coronary arteries in an attempt to cause a localised infarction and regional wall motion abnormality, recreating this finding in a more controlled manner8. Whilst the investigators observed that regional wall motion abnormalities could be seen during prolonged balloon occlusion the function of this myocardium returned when the balloon was deflated. A similar reduction in invasively measured gradients was seen in a small series of patients with HOCM. Following 5 minutes of balloon occlusion in the first septal artery the mean gradient reduced from 67mmHg to 12mmHg10. This recovery may be due to septal arterial supply being linked via complex arcade systems97 98. The collateral septal vessels may provide some form of myocardial protection, and whilst temporary wall motion abnormalities may be seen it was not possible to create long term myocardial damage and dysfunction.
109
With the understanding that prolonged occlusion of the septal vessels did not result in predictable and persisting injury to the target myocardium alternatives were considered. Transcoronary alcohol injection had been used to create cell death or ‘myocardial infarction’ to abolish incessant ventricular arrhythmia9. This was then adopted to create infarction in the basal septum10, with the desired effect. The mechanism of chemically induced infarction must have differed from prolonged hypoperfusion.
Damage to the myocardium from Transcoronary alcohol injection is proposed to have two main mechanisms; direct toxicity and necrosis of myocardium and coronary artery necrosis99. After injection of alcohol, coagulative necrosis of the septal branches occurs, this can distinguish an alcohol induced infarction from that attributable to ischaemic damage. Myocardial samples from patients who have undergone failed ASA have been taken at septal myectomy operations. Histology performed >6 months after alcohol injection has confirmed focal areas of myocardial fibrosis similar in gross appearance to myocardial infarction. These areas, when examined by standard cardiac imaging techniques, would have the appearance of classic ischaemic-related infarction.
110
3.2 Aims
To characterise size and location of myocardial infarction post-ASA
Compare size and location of infarct in LVOT gradient responders and non-responder post-
ASA
111
3.3 Methods
3.3.1 Choice of imaging modality
Cardiac Magnetic Resonance (CMR) imaging is in frequent use in clinical medicine for the imaging of abnormal myocardium100-104. CMR offers excellent spatial and temporal resolution and can characterise myocardium with the use of Gadolinium contrast imaging. The accumulation of
Gadolinium based contrast with paramagnetic properties in abnormal myocardium leads to the appearance of late gadolinium enhancement (LGE) on CMR images. LGE correlates well to infarct and fibrosis in histological samples105-108. LGE assessment has been validated in many pathologies, including HCM100.
LGE in CMR has been used to image the myocardial infarction post-ASA with good results109-112. It has advantages when used in HCM as it can also provide information with regard to anatomical features and systolic function113. Alternative imaging modalities such as nuclear medicine imaging can localise infarction108 114. This technique is not available in Liverpool Heart and Chest Hospital.
Some patients that had undergone ASA with a persistent gradient had undergone CMR scanning in an attempt to understand reasons for failure before my project started.
For these reasons I chose CMR as the imaging modality to describe localisation of myocardial infarction post-ASA.
112
3.3.1.1 Choice of LGE quantification methodology
Quantification of extent of LGE has been reported with several methods. All use the principal that
LGE techniques make scar appear bright. The signal intensity (SI) difference between a chosen area of ‘abnormal’ myocardium and a chosen area of ‘normal’ myocardium is then used to differentiate the area of scar100 115 116.
Some methods rely on a standard deviation (SD) of signal intensity, ranging from 2SD to 6SD. Full width half maximum (FWHM) is an expression of the extent of a function given by the difference between the two extreme values of the variable. In the context of scar analysis it uses half the maximum signal within a manually highlighted scar as the threshold to include that signal/myocardium as abnormal.
In a canine model Amado et al. demonstrated that the FWHM technique was able to accurately and reproducibly quantify the amount of infarction that correlated with post mortem infarct size measured by tetrazolium chloride staining105. FWHM technique was found to be the most reproducible with results similar to manual quantification in acute myocardial infarction, chronic myocardial infarction and HCM100. I therefore chose FWHM as the method of analysis for this study.
3.3.2 CMR acquisition methods
Due to the retrospective nature of this section of study the time period covered is large (2006-2014).
During this time CMR protocols changed. This causes some inevitable limitations which are discussed later. This must be borne in mind when interpreting the results.
113
All CMR scans were performed at LHCH using a 1.5-Tesla scanner (Magneton AERA; Siemens,
Medical Imaging, Erlangen, Germany). To enable analysis of cardiac volumes, steady state free precession (SSFP) cine images were obtained with breath holding at end expiration. A perpendicular two chamber pilot view (horizontal long axis/HLA) was obtained through a perpendicular plane passing through the apex and centre of the mitral valve. Four perpendicular short axis pilot views were then acquired either side of the atrioventricular groove. The four chamber SSFP cine was obtained from the four short axis pilot views by cutting through the centre of the left ventricle below the aorta. Using the two chamber pilot view and the four chamber cine, a two chamber long axis
SSFP cine is obtained. The 4-Chamber cine and the short axis pilot were then used to obtain an LVOT long axis view. The LVOT long axis view was then used to obtain an LVOT cross cut cine. Left ventricular short axis SSFP cines were obtained using the two and four chamber cines. An initial perpendicular short-axis slice was placed on the atrioventricular groove through the back of the left and right ventricles. A short axis stack of slices from the atrioventricular ring to apex was then acquired. Slice thickness unfortunately varied through the scans. The slice thickness and inter-slice distance is detailed in Table 8.1 in section 3.3.5. The number of cardiac phases per acquisition was
80-90% of the RR interval divided by the temporal resolution (typically 48ms) with 8 to 12 slices to cover the whole LV. Typical fast imaging with steady-state free precessions (FISP) imaging parameters were TE 1.6ms, TR 3.2 ms, in plane pixel size was typically 2.3 x 1.4mm, flip angle 60°, acquired in 12 heart beats, with 15 to 20 phases. Parameters were optimised to take into consideration arrhythmia and inability of the patient to hold their breath.
114
Transverse test
VLA test
4 Chamber/HLA
SA test
Short axis stack
2 Chamber/VLA
Figure 3.22: Schematic to demonstrate piloting and acquisition of a short axis stack
Gadavist® Gadolinium-diethylenetriamine pentaacetic acid (gadobutrol) 0.1mmol/kg (Bayer®;
Germany) was administered via an antecubital vein followed by a 20ml bolus of saline. Late gadolinium images were taken 8-10 minutes following administration of contrast. Standard breath- hold, segmented inversion recovery gradient echo sequences were used with image parameters: slice thickness 8-10 mm, TR 9.8 ms, TE 4.6ms, Flip angle 21º, field of view 340 x 220 mm (transverse plane), sampled matrix size 256 x 115-135, 21 k–space lines acquired every other RR interval (21 segments with linear reordered phase encoding), spatial resolution 1.3 x 2.1 x 8 mm, no parallel
115 imaging. Images were obtained in the same planes as SSFP cine images. The spacing was set at 8-
18.4mm (see section 3.3.6). The inversion time was adjusted to null the myocardium. Where appropriate cerebrospinal fluid (CSF) ghosting was prevented with a presaturation band placed over the CSF space (normal CSF has long T1 and long T2 times that manifest as dark signal on T1-weighted images and bright signal on T2-weighted images)117. In the event of arrhythmias or the patient having difficulty with breath holding, sequences were optimised to obtain the best images including the use of prospective gating or single shot multi-slice non breath hold imaging respectively118.
3.3.3 Standard CMR image analysis
All CMR scans were analysed by a single independent operator with level 3 accreditation in CMR , he was blinded to outcome from ASA. Analysis of cardiac volumes was performed on a personal computer using the software package CMRtools (Imperial College©) and the plug-in LV tools (an automatic algorithm that allows for user-independent myocardial border delineation).
The ventricular volume was calculated as the sum of the endocardial areas multiplied by the interslice distance, and ventricular mass was the area occupied between the endocardial and epicardial border multiplied by the interslice distance. First the short axis stack and two long axis perpendicular slices that pass through the centre of the mitral valve (four chamber and two chamber) were selected along with the short axis stack. The apex was then manually selected and the axis of the LV manually adjusted so that slices were aligned correctly. The endocardial and epicardial borders were semi automatically segmented in both systole and diastole and a thresholding tool was used to differentiate blood pool from myocardium (to include papillary muscle) enabling the calculation of the LV mass. The long axis images were then used to identify the position and orientation of the mitral and aortic valve leaflets at end-diastole and end-systole to
116 exclude regions that belong to the left atrium or aorta. The stroke volume (SV) was calculated from the difference between the end diastolic volume (EDV) and the end systolic volume (ESV): Stroke
Volume (SV) = EDV – ESV.
The ejection fraction was then also calculated using the following equation:
Figure 3.23: Equation for calculation of ejection fraction
Figure 3.24: Screenshot of typical volume and mass data output from CMRTools
117
The calculation of LV mass can be performed manually, after planimetry of the epicardial and endocardial borders in end-diastole, by a summation of the discs technique and multiplying the myocardial muscle volume by the density of myocardial tissue (1.05 g/cm3)119.
The extent of late gadolinium enhancement was then calculated. Myocardial volume and mass analysis were carried out on cine images using standard techniques. Short-axis images were manually segmented for epicardial and endocardial borders (excluding papillary muscles) to obtain the myocardial volume. Where more than one image of the same slice position was present, the optimal image was selected for analysis. All further quantification occurred on the presegmented images. LGE was quantified manually by tracing around the borders of infarction or fibrosis. Semi- automated analysis was then carried out using purpose-written ImageJ (National Institutes of
Health, Bethesda, MD) software to determine full width half maximum (FWHM) derived volumes. A region of interest (ROI) was drawn around hyperintense myocardium and used to define maximal signal for the FWHM threshold. Manual corrections were then required and done twice for all automated regions of interest (ROIs) with any blood pool or pericardial partial voluming and artefact
(which occurred only rarely) manually removed from the ROI.
3.3.4 Identification of target myocardium
A precise description of the desired target myocardium in ASA has never yet been framed. Common terms in use are ‘basal septum’ and ‘SAM-septal contact point’. Clarification beyond this has not been agreed, partly due to the complex nature of LV haemodynamics in HCM and partly due to the low numbers of imaging studies available post-ASA.
118
Standard descriptive nomenclature for the left ventricle exist, such as the widely used 17-segment model120. This model splits the septum in the anterior and inferior septal segments. As the SAM- septal contact point is not one defined location in all patients and the position of the LVOT on short axis CMR scans differs subtly this can vary between these basal septal segments. This makes it challenging to accurately define a standard anatomical target for all patients.
Myectomy serves as a comparative model for ASA. The operation has been performed for a long time with excellent success in high volume centres64 121-123. It has been reported that the location of
ASA-induced infarct appears to be more inferior in the septum than the ‘groove’ of removed myocardium that results from myectomy, this is seen more anteriorly124. As myectomy serves as the gold standard to which ASA must be compared it is worth considering the reasons for this difference to try to understand the true target, as it is often commented that complete gradient resolution is seen more after myectomy than ASA51 52 63. The location of the ASA induced scar may relate to the septal arteries naturally draining towards the inferior septum and alcohol travelling inferiorly, there is therefore some lack of control when injecting fluid in to these arteries. The mechanism of alcohol induced damage to the intima of the arteries may restrict this to a certain extent, as we see the ‘no reflow’ phenomenon on contrast injection at the end of ASA99. An alternative view requires us to use the surgical perspective, quite literally. During a standard myectomy operation the surgeon uses direct visualisation of the septum to choose the piece of myocardium to remove; this is via aortotomy and aortic valve retraction. He or she creates a continuous channel for blood flow from the mid-LV level to the aorta by removal of the myocardium apparently obstructing flow. The concept of direct flow being the most efficient use of haemodynamic energy is not new and has been well studied in congenital heart disease125.
119
It therefore makes sense that the target infarct in ASA should also create this channel or groove, to allow laminar blood flow out of the heart and minimise interaction with the MV apparatus. The target then becomes the myocardium immediately below the AV and LVOT. This is in the base of the heart, and must be on the LV side of the septum. An infarct on the RV side would not create this channel or have any effect on the interaction with the MV apparatus.
The target therefore must be:
- In the base of the heart immediately below the AV and adjacent to the LVOT.
- Within the short axis area that the AV and LVOT occupy – i.e. not too anterior or inferior
within the septum.
- On the LV side of the septum.
3.3.4.1 Basal target area definition
I defined the basal septal target from diastolic still frames of the LGE LVOT/3 Chamber view. A
2x2cm box was placed at the LV side of the aortic valve (see Figure 3.4). Any slice of the LV that passed through this box was deemed to be in the target basal septum (represented by the green lines in Figure 3.4). Any slice that was apical of this box was deemed to be too far down the septum and not within the target box (red lines in Figure 3.4).
120
Figure 3.25: Definition of target myocardium in the long axis.
The relational nature of the CMR analysis software package allowed the operator to reference the short axis cut to the LVOT long axis view (see Figure 3.5 and Figure 3.6), this was crucial to enable the assessor to know what LGE was included in the target area. In Figure 3.5 panel A displays the
LVOT long axis view with a green line across, this green line represents the SAX LGE slice seen in panel B. This is within the target green box area - on the LGE slice no enhancement is seen. In panel
C and D the SAX view is more apical in the LV, beyond the target area. The LGE slice shows enhancement, this is included as infarct outside the target area.
121
Figure 3.26: Example of assessment in the long axis – infarct is outside the target area apical of the target basal septum.
In Figure 3.6 we see the SAX LGE slice is within the target green box in the long axis image. The infarct is therefore included as within the target area.
122
Figure 3.27: Example of assessment in the long axis - infarct is inside the target area
3.3.4.2 Short axis area target definition
In this section I approached the localisation of infarct by thinking about a short axis slice of the heart as a circle and splitting it into degrees of arc. The SAX slices from the aorta to the apex are therefore a series of circles (see Figure 3.7). The LVOT and AV will occupy a portion of the degrees of arc of the circle in the higher slices so that it becomes an incomplete circle, or reverse ‘C’ shape. When moving apically down the ventricle the area that was ‘empty’ in the C shape in aortic/LVOT views becomes the myocardium directly below the exit point from the LV. This is the area that we see the groove of myocardium excised after myectomy as the surgeon has moved directly from the aorta to the LV.
123
Figure 3.28: Series of SAX slices from Aorta and atrial level (A) to Mid-ventricle (D).
I used software callipers to create a degrees of arc ‘angle’ triangular shape in the aortic SAX sections to signify the outflow of the heart (see Figure 3.8). I then copied this angle to all SAX slices in the stack to indicate the target myocardium in the short axis plane. Panel A shows a SAX slice at
LVOT/AV level. In panel A the opening in to the aorta is marked with the green callipers from a central point in the LV to create an angle; this represents the path of blood flow. The callipers are placed to mark the borders of myocardium and LVOT (white arrows). These callipers are transferredon to all SAX slices moving apically down the ventricle (example shown in panel B).
Figure 3.29: Example of creation of target area in SAX slices
An example of LGE occupying a large proportion of the target myocardium in the short axis is shown in Figure 3.9A. An example of an infarct occupying a smaller area with a large portion of non-
124 infarcted and viable myocardium is seen in Figure 3.9B; the white arrow highlighting an area of non- infarcted viable myocardium within the target area.
Figure 3.30: Examples of infarct occupying most of the target (A) and a smaller area (B)
3.3.4.3 LV versus RV definition
I have defined the target area according to long axis and short axis directions. There is a third dimension in which the infarct can miss the target; this is related to the LV and RV sides of the septum. It is possible to have an infarct in the target areas in long and short axes, but if it occupies solely the RV septum then LV wall motion will not be affected and the desired ‘groove’ in the myocardium is not created (see Figure 3.10). I therefore stipulated that if the scar did not involve the
LV endocardium this was outside the target area.
125
In the example displayed in Figure 3.10 the scar is seen to be in the target area in the long axis
(within the green box) and the short axis (within the callipers), but in both images we can see that the infarct occupies only the RV side of the septum, this will have no significant effect on LV haemodynamics. This infarct was deemed to be outside the target myocardium.
Figure 3.31: Example of RV only infarct occupying target area in long and short axis
3.3.4.4 Summary of identification of target myocardium
This method of identifying target myocardium and its subsequent assessment is original to my thesis and has never been used before. In brief the target myocardium must be:
Within the 2x2cm box in the long axis LVOT image
Within the callipers representing the outflow of the heart in the short axis
Involving LV endocardium
126
3.3.5 Image analysis for location of ASA induced infarct
Standard CMR image analysis for LV function and mass was performed as described in section 3.3.3.
The total LV LGE burden in g of mass and % of LV was calculated. The operator then calculated the extent of LGE that was within the target myocardium. For each SAX LGE slice that was within the identified target area according to long axis criteria the operator first defined the presence of LV endocardium involvement in the infarct. Those with only RV infarct were highlighted and removed
(as stipulated in section 3.3.4.3). Then the process described in Figure 3.11 was completed. This was able to give me total g of infarct contained within the target myocardium. The total LV LGE burden had already been calculated. This data allowed me to calculate:
Total LGE burden/infarct associated with ASA
% of LV occupied by LGE
Total mass of the target myocardium
Total mass of infarct within the target area
% of target area occupied by infarct.
% of total infarct within target area
127
Figure 3.32: Method for analysis of LGE in a SAX slice
128
3.3.5.1 Accommodating pre-existing LGE not related to ASA-induced scar
LGE burden in HCM patients may be in part due to replacement fibrosis associated with HCM, with an observational burden in an unselected population seen to progress from 4% LV mass to 8% LV mass over a year in some populations126. Ideally I would like to remove this progressive LGE burden from my measurements to be able to comment on just ASA related infarct LGE. Pre-ASA CMR was available for interpretation in 13/22, these pre-ASA scans underwent LGE assessment according to the methods described. In these I subtracted any LGE measured on the pre-ASA scan from the post-
ASA scan to give a ‘new’ mass of LGE observed and attributed this to alcohol related infarct. The remaining 9/22 had a one-off measurement of LGE mass.
3.3.6 Patient selection and scan quality assessment
Prior to the start of my project there was no standard operating policy for performing CMR after
ASA. Some patients who had persisting gradients underwent CMR in an attempt to understand the reasons for failure but no patient who underwent successful ASA was scanned. These scans were performed between 2006 and September 2011. Not all failed procedures underwent CMR. I did not have ethical permission or funding to call all patients from the retrospective group back to be scanned.
I wanted to create 2 groups of scans to analyse; ‘success’ and ‘fail’. I classified success and fail according to the echocardiographic criteria discussed in chapter 2. All patients that underwent
129 scanning prior to my arrival were classified as ‘fail’. As I was missing a success group for comparison I decided to include all scans taken prospectively as part of the standard operating policy that was introduced at the start of my project. These scans were performed between January 2012 and
January 2014. This included some patients referred from other centres, some had ASA in another institution and were referred to LHCH with a persisting gradient.
I assessed all scans available for image completeness. Twelve retrospective patients had been scanned. The image quality was poor and short axis SSFP stacks were incomplete in 3 of these. I collected 15 patients prospectively (inclusive of patients who underwent CMR at LHCH to investigate failure of ASA elsewhere). I had a total of 24 patient scans available for analysis. Twelve patients fell into each category, this was not intentional and they are not matched. A patient consort diagram can be seen in Figure 3.12.
In each group there was one scan that had an incomplete LV LGE stack (the SSFP cine stack was complete). The LGE images focussed only on the base and mid-LV. I could ask the operator to complete some of the analysis on these scans but not all (e.g. % total LV infarct in target area) as we were unable to comment on the apical potions of the LV.
130
Retrospective review 12
Unsuitable 3
Scans suitable for analysis 9
Prospective collection 15 Total number 24
Success 12 Failure 12
Incomplete LGE stack 1
Incomplete LGE stack 1 RV infarct only 2
11 Final analysis 9
Figure 3.33: Consort diagram for patient/scan recruitment
As stipulated in section 3.3.2 the slice thickness and spacing of the LGE stack was variable due to the lack of set protocol prior to initiation of the project. The observed thickness and spacing is detailed in Table 3.11. There was considerable variation.
131
Thickness Spacing Thickness Spacing (mm) (mm) (mm) (mm)
Success 8 8 Fail 8 14 8 12 8 12 8 12 8 18.4 10 10 8 8 8 10.4 8 10 8 10.4 10 10 10 18 10 18 8 16 8 16 10 18 8 18 10 18 8 18.4 8 18 8 10 Mean 8.67 13.4 Mean 8.4 14.28 Min 8 8 Min 8 8 Max 10 18 Max 10 18.4
Table 3.11: Slice thickness and spacing of LGE stack for analysed scans.
There are several limitations associated with this methodology that are discussed in section ‘3.7
Limitations’.
3.3.7 Statistical analysis
Statistical analysis was performed with StatsDirect version 2.8.0 and Microsoft Excel. Normally distributed data for CMR analysis was assessed using unpaired t test, non-parametric data was assessed using Mann-Whitney U test. Assessment of normality of data was performed using the
Shapiro-Wilk test, a p value of <0.05 was used to indicate a non-parametric data set. Correlation assessments were performed using Pearson’s test. Fisher’s exact test was used to compare incidence of isolated RV infarcts in each group.
132
3.4 Results
3.4.1 Assessment of LV size and function
There was no significant difference between the success and fail groups in terms of age or time from
ASA to CMR scanning (see Table 3.12). When assessing CMR measurements of myocardial volume and systolic function there was no significant difference. Parameters such as EF, diastolic and systolic volumes were similar. There was however at trend towards a higher LV mass (295.20 ±95.48g vs
242.25 ±66.21g) and higher stroke volume (102.60 ±21.98g vs 88.08 ±21.54g) in the fail group, this did not reach statistical significance. The higher stroke volume may be due to a higher amount of mitral regurgitation, as I did not have phase contrast velocity mapping aortic flow sequences I was unable to calculate the mitral regurgitant fraction.
Value Normal Success Fail p distribution mean ±SD/(Median mean ±SD/(Median IQR) IQR) Age Yes 59.37 ±12.83 56.05 ±14.28 0.57 Time from ASA to scan (yrs) No 1.25 (0.43-0.64) 1.22 (0.46-1.18) 0.43 Body surface area (BSA) Yes 1.99 ±0.15 2.00 ±0.22 0.95 EF (%) Yes 76.00 ±6.31 81.10 ±6.97 0.91 LV mass (g) Yes 242.25 ±66.21 295.20 ±95.48 0.14 LV mass normalised (g/m2) Yes 121.79 ±34.07 148.39 ±46.91 0.14 End diastolic volume (EDV) Yes 115.50 ±24.16 126.80 ±27.26 0.31 EDV normalised (EDV/m2) Yes 58.07 ±11.76 63.96 ±14.96 0.32 End systolic volume (ESV) Yes 27.50 ±8.61 24.10 ±11.96 0.45 ESV normalised (ESV/m2) Yes 13.90 ±4.62 12.07 ±5.96 0.44 Stroke volume (SV) Yes 88.08 ±21.54 102.60 ±21.98 0.13 SV normalised (SV/m2) Yes 44.20 ±10.13 51.68 ±12.49 0.13 LA area (cm2) Yes 34.29 ±5.39 31.86 ±7.18 0.49
Table 3.12: Age and CMR measurements in those with success versus those with failure
133
3.4.2 Correlation of infarct size markers
In Chapter 2 the correlation analysis of volume of alcohol injected and CKMB rise suggested no reliable relationship in the retrospective ASA population, suggesting a lack of control of the effects of alcohol. A similar message is seen when correlating alcohol dose to the measured mass of LGE
(interpreted as infarct) on CMR. Full data on CMR LGE, alcohol dose and CKMB was available in 20 patients. There was no significant correlation of alcohol dose to LGE mass when including all patients; R2 0.03, p=0.44 (see Figure 3.13). This was also true when splitting the patients into the success (R2 0.02, p=0.46) and fail groups (R2 0.03, p=0.43) (See Figures 3.14 and 3.15)
Alcohol Dose to new mass LGE correlation: All patients y = 1.6744x + 7.5398 R² = 0.0329 p=0.44 40
35
30
25
20
15 New mass LGE (g) 10
5
0 0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5 Dose alcohol mL
Figure 3.34: Alcohol dose to new LGE mass correlation in all patients
134
y = 1.2116x + 9.9948 Dose alcohol to new LGE correlation: Success R² = 0.0193 p=0.46 40 35
30 25 20 15
New mass LGE (g) 10 5 0 0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5 Alcohol dose mL
Figure 3.35: Alcohol dose to new LGE mass correlation in Success patients
Dose alcohol to new LGE correlation: Fail y = 2.2896x + 4.6096 R² = 0.0397 p=0.43 25
20
15
10 New mass LGE (g)
5
0 0 0.5 1 1.5 2 2.5 3 Alcohol dose mL
Figure 3.36: Alcohol dose to new LGE mass correlation in Fail patients
135
In Chapter 2 I used CKMB as a marker of control of infarct size. CKMB to new mass of LGE measured shows a much better correlation (see Figure 3.16), with an R2 value of 0.20 and p=0.06. This suggests there is some mild correlation between the two markers of infarct size in this population.
y = 0.0762x + 3.5771 CKMB to new mass LGE correlation R² = 0.2035 p=0.06 40
35
30
25
20
15 New mass LGE (g) 10
5
0 0 50 100 150 200 250 CKMB Release ng/dL
Figure 3.37: CKMB release to new mass LGE correlations
3.4.3 Target area evaluation
The target area was identified according to the methods specified in section 3.3.4. The size of the target area showed a trend towards being slightly bigger in the fail group; 16.64 ±5.09g vs 13.35
±6.66g (p=0.2), this was perhaps related to the fact that the total mass of the LV was greater in this group. The amount of LGE in the target angle trended towards being larger in the success group
(p=0.15) (see Table 3.13). Given the smaller target mass and the increased size of LGE occupancy the
136
% of the target myocardium occupied by LGE was also higher in the success group, with p value reaching 0.07. Although these measurements showed trends towards significance, none reached conventional levels of statistical significance (p<0.05).
The total size of infarct was similar between groups, but the success group had a trend towards higher % LV occupancy due to the overall smaller mass. The % of the total volume of LGE inside the target area also showed a trend towards being higher in the success group, but this also did not meet statistical significance.
Value Success Failure P n Mean ±SD n Mean ±SD Mass of target 11 13.35 ±6.66 9 16.64 ±5.09 0.20 Mass Gd in target (g) 11 5.42 ±4.47 9 3.92 ±3.94 0.15 % of target occupied by Gd (%) 12 40.60 ±18.63 10 23.61 ±19.01 0.07 Total LV Gd (g) 11 12.65 ±9.67 9 11.74 ±8.46 0.55 Total LV Gd (%) 11 5.41 ±3.17 9 3.03 ±2.97 0.15 Mass Gd outside target (g) 11 7.23 ±4.27 9 7.82 ±4.67 0.72 % of total Gd in target area 11 42.85 ±24.24 9 28.53 ±27.42 0.18
Table 3.13: LGE evaluation of post-ASA CMR scans
N.B. One patient in each had an incomplete LGE stack. We were therefore able to complete the % occupancy of the target area as this was covered, but other measurements such as mass and total LV observations were not possible.
Some patients had CMR prior to ASA (n=13, 8 in success group, 5 in fail group). These scans were also analysed by the same operator in the same manner. Any LGE was subtracted from the post-ASA numbers to give a ‘new LGE since ASA’. This may better reflect the change associated with the intervention of ASA.
137
When the pre-ASA LGE is removed both groups show a reduction in observed LGE. The mass of LGE in the target area reduces in both groups, but more so in the fail group – indicating a slightly higher proportion of the LGE observed there was pre-existing and not due to alcohol induced infarct. The trend for the success group to have a greater size of LGE in the target area became more apparent, and the % of target area occupied by LGE became statistically significant, with a p value of 0.03 (see
Table 3.14).
Value Success Failure p n Mean ±SD n Mean ±SD Mass of target 11 13.35 ±6.66 9 16.64 ±5.09 0.20 Mass Gd in target (g) 11 4.83 ±4.15 9 3.00 ±3.94 0.12 % of target occupied by Gd (%) 12 36.18 ±16.21 10 18.02 ±17.11 0.03 Total LV Gd (g) 11 8.83 ±4.75 9 8.96 ±8.46 0.46 Total LV Gd (%) 11 3.97 ±1.81 9 3.12 ±2.10 0.24 Mass Gd outside target (g) 11 4.00 ±3.65 9 5.96 ±4.67 0.26 % of total Gd in target area 11 54.70 ±24.24 9 33.48 ±27.42 0.19
Table 3.14: LGE evaluation post ASA with LGE from pre-existing scans removed.
3.4.3.1 RV vs LV septum
There were 2 CMR scans with isolated RV infarction only. Both patients were in the fail group. The infarcts were in the basal septum but no LV endocardial involvement was seen. The incidence of isolated RV infarct in the success group was 0/12, and 2/12 in the fail group (p=0.23).
138
3.5 Discussion
There are numerous limitations to this study; these are discussed in section 3.5.1. Accepting these limitations we observe that a greater proportion of the target volume of myocardium is infarcted in those that have successful resolution of LVOT gradients versus those that do not. Infarction of myocardium in the basal septum reduces the septal width, opening the LVOT. It also reduces the systolic excursion of the septum in to the LVOT in systole, reducing its interaction with the mitral valve and reducing SAM, this further reduces obstruction to flow out of the LV and the gradient is therefore lower.
Other series have attempted to describe the appearance of LGE on CMR post ASA109-111 124 127. These series report a lower LV mass (183-216g) and larger Infarct size (20-29.5g) than observed in our population111 127. The total LV LGE burden does not seem to be critical to affect LVOT haemodynamics, as the size of infarct did not differ between success and fail groups (8.83 ±4.75 in the success group vs 8.96 ±8.46 in the fail group). The location is however very important. In the success group we see twice the % of target volume occupied by infarct (36.18% vs 18.02%). This target volume also involves some RV septal myocardium that is irrelevant to the LV haemodynamics; this suggests that the 36.18% occupancy is likely an underestimate of the infarct in the true target zone.
It is claimed that ASA creates an infarct more inferiorly in the septum when compared to the trench created by septal myectomy124. This group also suggested that the majority of the infarcts extended to the right ventricular myocardium, suggesting a lack of the desired control when attempting to mimic the structural effects of myectomy. The inferior infarcts seen in these groups may extend
139 beyond the target area I had set if they were to be analysed in the same manner. Although this would have some effect on LV haemodynamics it would be unlikely to be as pronounced as those involving true target myocardium and hence reducing SAM-septal contact. This may explain part of the slightly higher persisting LVOT gradients after ASA when compared to myectomy51 52.
It has also been observed that infarcts that solely involve the RV septum have a poor outcome and do not affect LVOT haemodynamics111. This was true in my population but did not reach statistical significance due to the low numbers involved.
The observation made in section 2.4.2.1 that alcohol dose does not correlate well with size of infarction as assessed by CKMB release was re-iterated by measurements of infarct size by LGE assessment. No significant correlation between dose of alcohol injected and extent of LGE was seen, regardless of population chosen (all, success only, fail only). There was however a better correlation between CKMB and LGE extent as markers of size of infarct (R2 0.20 p=0.06).
The observations of RV only infarct and lack of correlation of alcohol dose to LGE/infarct size suggests there is an uncontrolled element to the myocardial damage delivered using traditional ASA methods.
3.5.1 Limitations
There are numerous limitations associated with the methodology used in this chapter.
140
3.5.1.1 Hazards of retrospective data collection
Patient related
There was no routine method or standard operating policy (SOP) in place for the CMR scanning of
ASA patients prior to the initiation of my project. I was reliant on some historical patients undergoing CMR scanning to investigate the reasons for failure of ASA, not all with persisting gradients were scanned however. This may represent some selection bias as the recruitment was not consecutive.
The low numbers scanned retrospectively and the lack of ethical approval or funding to bring historical patients (both success and failure) back to LHCH for CMR meant I had to recruit from the prospective patient group undergoing ASA. A new SOP was introduced with these patients and all underwent CMR prior to ASA. They therefore had pre-alcohol scans available for LGE assessment whereas the historical group did not. This meant that some of the scans assessed in section 3.4.3 had removal of pre-existing LGE whilst others did not; this created a ‘mixed’ population.
I did not perform any power calculations to detect a significant difference in target infarct size as I was unsure how to recruit patients at the initiation of this section of the project. I was also unsure what size of infarct would be observed, and given the low numbers of previous studies I was not comfortable basing calculations on the observations of others with slightly different methodologies.
The observed infarct size in my population was smaller than that available in the literature; this would have caused significant problems if I had used these to model power calculations for this study.
141
CMR acquisition related
The mixed nature of the recruitment for this section resulted in differing acquisition protocols. A clear example of this was the differing slice thicknesses chosen for SSFP and LGE stacks. This means a different volume of myocardial signal was used to create a single short axis slice – this was then assessed for infarct. It also means there is differing partial volume effect between scans.
The spacing between slices also differed, varying from 8-18mm. In those with 18mm gap and 10mm slice thickness there is also the possibility of missing areas of infarct due to incomplete LV coverage.
These were all historical scans.
3.5.1.2 Restrictions of analysis methods
Software
In the conception phase of this chapter I wanted to create a target area in the long and short axis as described, but intended to create an RV-LV delineation for further assessment. We only had access to ImageJ software for LGE analysis and this package did not allow us to do this. There was no funding to seek alternative packages and this therefore had to be left out. Those with clear RV only infarcts were therefore removed from the analysis.
In Figure 3.17 I have displayed an example of some of the images from the conception phase. The green callipers representing the LVOT in degrees of arc are seen as discussed in section 3.3.4.2. I had intended to add a further dividing line in the septum to split this in to LV and RV – this is represented
142 by the light blue line. In panel A we see an example of a success; most of the area within the green lines and on the LV side of the septum is occupied by LGE. In panel B we see an example of an isolated RV infarct. It is wholly within the green lines and would represent part of the ‘%target occupied by LGE’ statistic, but it is almost exclusively outside the LV target area.
Figure 3.38: Example of intended RV-LV delineation from concept phase
Reproducibility
I have no studies of reproducibility. I was unable to recruit another assessor to repeat the assessments for inter-observer reproducibility. We were also unable to perform intra-observer variability assessment. This was a restriction of time and finance.
143
All current LGE quantification methods are limited by the fact that essential steps like delineation of myocardial contours, drawing a region of interest in remote myocardium and the exclusion of artefacts are all performed manually.
144
3.6 Conclusion
A greater proportion of the target myocardium is infarcted in those with successful
resolution of gradients versus those with persisting gradients.
There is no correlation between alcohol dose and extent of infarction suggesting a lack of
control.
145
Chapter 4: Intra-cardiac Echocardiography to guide alcohol septal ablation in hypertrophic obstructive cardiomyopathy: A prospective validation study against trans-thoracic echocardiography
146
4.1 Introduction
I have shown that a proportion of patients treated with ASA will be left with LVOT obstruction of clinical significance and that imperfect localisation (rather than size) of the iatrogenic myocardial damage may be a factor in the failure of this therapy.
In this chapter I explore the potential of intracardiac echo to better guide the procedural performance of ASA, seeking better localisation of alcohol delivery and consequent heart muscle damage.
I have described the procedure for performance of ASA in Chapter 1. International guidelines recommend the use of myocardial contrast echocardiography (MCE) during ASA12 13. MCE is used to confirm the myocardial distribution and blood supply of individual septal vessels prior to injection of alcohol. This allows operators to evaluate the location and size of the territory supplied by the chosen vessel. This gives an estimate of the distribution and extent of damage alcohol injection will cause.
Transthoracic echocardiography (TTE) is the current imaging modality of choice for guiding ASA.
There are significant difficulties in using TTE for this procedure. The images must be taken on-table, with the patient in a supine position. The optimal position for transthoracic image acquisition is usually in the left lateral decubitus position with the patient at 45 degrees. There are further logistical difficulties resulting from narrow catheterisation tables, lack of arm support and cramped space in most laboratories. These difficulties can result in substandard image quality.
147
In most cases, the images obtained are deemed to be of sufficient quality to comment on the localisation of contrast and likely alcohol distribution. We must consider the prospect that substandard procedural imaging may contribute to the observed inaccurate location of iatrogenic infarction. If the operator is unable to see the highlighted myocardium with the level of accuracy required then he or she is unable to confidently say the chosen septal vessel supplies the target area. We must therefore investigate alternative approaches to procedural imaging.
Other echocardiographic modalities available are trans-oesophageal echocardiography (TOE) and intra-cardiac echocardiography (ICE). Most centres have access to TOE imaging and established skills in its use, but it also has limitations. The need for sedation and intubation carries a risk. This increases the risk in HOCM population due to their complex cardiac physiology; the group provide a greater challenge in terms of anaesthetic safety. Anaesthetic agents can also alter loading conditions and therefore affect LVOT gradients, leading to inaccurate assessment. ICE has been shown to provide high quality images in other interventional procedures such as patent foramen ovale and left atrial appendage closure128 129. No sedation or anaesthetic agents are required for the use of ICE; it is therefore an attractive modality to guide ASA.
148
4.2 Aims
To perform a prospective validation study of ICE vs TTE for guiding ASA.
149
4.3 Methods
4.3.1 Defining the cardiac structures important to ASA
In order to perform accurate and safe ASA, procedural imaging must provide clear information on several aspects of cardiac anatomy.
4.3.1.1 The target myocardium
Systolic anterior motion of the mitral valve (MV) is critical to the pathophysiology of LVOT gradient generated in HCM. Visualisation of the motion of the MV, the SAM-septal contact point and the surrounding basal septum is critical as this forms the target for infarction.
4.3.1.2 Other structures
It is also important to clearly see areas of the surrounding myocardium that could receive blood supply from septal arteries. If an incorrect artery is chosen the operator must have sight of the highlighted myocardium, remote from the intended target site to be able to dismiss this artery. It is therefore important to be able to see the mid-septum, as an infarct too apical in the septum will have a minimal effect on LVOT gradients. Infarcting the mid-septum will not help resolve obstruction, but it is unlikely to cause complication. It is also important to consider structures that can cause significant problems if infarcted. The anterior papillary muscle receives blood from the left coronary arteries in most people, whereas the posterior muscle receives supply from the right coronary system in 70% cases130. As septal arteries are usually from the left system the anterior papillary muscle is at greater risk of inadvertent contrast or alcohol injection.
150
The most common site for undesirable contrast localisation has been reported to be the RV septum and moderator band131. I have observed RV infarction on CMR scans in patients with persisting gradients. Most septal arteries will have sub-branches serving the right and left ventricular myocardium. When blood flows into the right sided branches it often empties into the right ventricle via Thebesian veins. It is therefore essential to see the right ventricular cavity on contrast injection to ensure injected fluid will not simply empty into the RV without dwelling in the myocardium.
4.3.1.3 A scoring system for assessing images
After considering the structures relevant to intra-procedural imaging in ASA I devised a scoring system to help describe image quality in TTE and ICE. This was split in to three sections:
A: Mitral valve and SAM
B: Target septum
C: Other key, distant structures: Mid-septum, anterior papillary muscle, RV cavity
The scoring system had 3 levels:
2: Optimal
1: Moderate quality
0: Poor
151
The definitions devised to achieve a score in each section are detailed in the tables below.
Section A: Mitral valve and SAM
Score Observation 2 Able to determine SAM – septal contact point with precision such that images could allow determination of the length of anterior mitral valve leaflet (AMVL) in contact with septum at maximum excursion (or if no contact - ability to measure with precision the distance from AMVL tip to septum (Analogous to parasternal long axis m-mode measurement of E-point septal separation [EPSS] distance)) 1 Able to localise approximate SAM - septal contact point (or, if there is no contact with the septum, able to localise approximate anterior leaflet tip position at point of maximum excursion) 0 Unable to determine tip position of the anterior leaflet in relation to the interventricular septum
Table 4.15: Scoring system for section A: Mitral valve and SAM
152
Section B: Target septum
Score Observation 2 Able to define endocardial border of basal ventricular septum in both RV and LV with visualisation of central intra-septal ‘fusion line’ between RV and LV. On-axis images 1 Able to define endocardial border of proximal ventricular septum in both RV and LV but no visualisation of central intra-septal ‘fusion line’ between RV and LV 0 Unable to clearly define endocardial border of one or both sides of the ventricular septum
Table 4.16: scoring system for section B: Target septum
Section C: Other key distant structures: Mid-septum, anterior papillary muscle, RV cavity
Score Observation 2 3 key structures 1 2 key structures 0 0 or 1 key structure
Table 4.17: Scoring system for section C: Other key distant structures
4.3.2 Intracardiac echocardiography
ICE images are taken using a side facing echocardiographic array probe mounted on an 8Fr or 10Fr catheter. Images are routinely taken from the right atrium or right ventricle, the catheter is passed
153 to the right heart via the right femoral vein in most cases. Connection to a standard echocardiogram console with additional ICE software allows real time visualisation of images.
4.3.2.1 Patient groups
Two groups of patients were selected for study. In phase 1 I chose a population of patients in whom
ICE was already being used for another reason. I wanted to assess the feasibility of ICE and the use of the scoring system in a series of patients that are already undergoing ICE guidance for another procedure before introducing it as a new technology to a group of patients undergoing ASA. I therefore used a population of 25 consecutive patients undergoing patent foramen ovale or atrial septal defect closure using ICE guidance at The John Radcliffe Hospital, Oxford.
Following the use of ICE in Phase 1 I was able to transfer the technology to guide ASA in a consecutive series of 20 patients entering the catheterisation laboratory with the intention of alcohol delivery at Liverpool Heart and Chest Hospital – this formed Phase 2.
4.3.2.2 Power calculation
As in an adaptive study design, following the first 10 Phase 1 patients I undertook some calculations to estimate total sample size necessary to power the study appropriately. I am scoring in 3 domains;
I consider it clinically significant if one domain changes by one grade score.
After 10 patients the mean total working view score for TTE was 4 (SD = ±1.1), e.g. the images scored
1, 1, and 2 in the three categories A, B, and C. I have stipulated that I would like to see a change in
154 score by one full point. I therefore require a 25% effect on the mean in ICE for it to be a significant change. Using Excel I constructed a power calculation table (Table 4.18).
No of No of No of No of Enter Absolute SD of Patients in Patients in Patients in Patients in known Expected difference in known Each Group Each Group Each Group Each Group mean % effect the means mean (Two Sided) (Two Sided) (Two Sided) (Two Sided) value on mean (Limb 1 +/- (Limb 1) Alpha 0.05, Alpha 0.05, Alpha 0.05, Alpha 0.05, (Limb 1) Limb 2) Power 50% Power 80% Power 90% Power 95% 4 1.1 10 0.40 58 119 159 197 4 1.1 15 0.60 26 53 71 87 4 1.1 20 0.80 15 30 40 49
4 1.1 25 1.00 9 19 25 31 4 1.1 30 1.20 6 13 18 22 4 1.1 35 1.40 5 10 13 16 4 1.1 40 1.60 4 7 10 12 4 1.1 45 1.80 3 6 8 10 4 1.1 50 2.00 2 5 6 8 4 1.1 55 2.20 2 4 5 6 4 1.1 60 2.40 2 3 4 5
Table 4.18: Power calculations for echocardiographic assessment scoring system
I have highlighted the relevant section in red. I have entered the mean and standard deviation values of 4 and 1.1 respectively. I require a change from the mean of 1; 1/mean(4) x 100 = 25 % expected change on mean. The 4 levels of power from 50% to 95% are represented in the table, with number of patients needed in each group. In my study each patient will be represented in both groups as unpaired data, as I am comparing TTE and ICE, not pre- and post-intervention change. The conventional statistical norm is to use 80% power; this gives me a sample size of 19 patients.
155
4.3.3 Image acquisition
4.3.3.1 Phase 1
During clinically indicated PFO/ASD closure procedures ICE images were taken of the left ventricle in the long axis view. The ICE catheter (8Fr AcuNav™ (Siemens Acuson, Mountain View, CA,USA)) was inserted via the right femoral vein. This was manipulated into the right ventricular inlet area for image acquisition. The phased array probe was then directed to visualise the inter-ventricular septum. This view displays the inter-ventricular septum and potential target myocardium for ASA.
No SAM-septal contact was anticipated in these patients. To comment on the ability to identify the
SAM-septal contact point in this group we used the ability to comment on the E-point septal separation distance as a surrogate. Other aspects of the scoring system were applicable to Phase 1 patients. On table paired TTE (Siemens Acuson Cypress) images were also taken. The patients were lying supine with femoral vascular access in situ. The position mimics that of ASA. Operators recorded standard TTE views as per British Society of Echocardiography standards132. The views obtained focussed on the LVOT and were predominantly the parasternal long axis (PLAX), apical 5 chamber and apical 3 chamber. All echocardiographic images were acquired at the time of cardiac procedure as described above and analysed at a later sitting be a separate pair of expert cardiologists (see 4.3.4).
4.3.3.2 Phase 2
ICE images were obtained as for Phase 1 patients using 8Fr AcuNav™ (Siemens Acuson, Mountain
View, CA,USA) and VIVID Q console, GE. TTE images (VIVID I, 3Sc RS probe, GE) were acquired as in phase 1. The ability of the operators to use myocardial contrast echocardiography to localise the intended area of infarction using ICE and TTE was also assessed in this group. Myocardial contrast
156
(0.5mls Sonovue®, Bracco) was injected into previously identified septal vessels. Separate coronary injections were performed for ICE and TTE.
4.3.4 Image analysis
Two expert reviewers experienced in the use of ICE independently performed off line analysis of all images. ICE and TTE were assessed according to the pre-defined scoring system. Due to the difference in appearance of TTE and ICE images blinding to the echocardiographic modality was not possible. The comparison of each modality to identify relevant anatomy using the scoring system was performed in 2 pre-determined stages.
In the primary analysis, the reviewers applied the scoring system using only the ‘work view’. This was a single view chosen from a standard range of TTE and ICE images prior to contrast injection. This was designed to mimic the approach used in ASA as a single contrast injection can highlight perfused myocardium for <10 seconds. This only allows acquisition of a single view in most cases. The chosen loops were recorded and declared (the chosen cines for Phase 2 are available in supplemental file 1).
If required multiple contrast injections can be performed in ASA allowing multiple views to be obtained. Therefore a secondary analysis was performed allowing reviewers to choose separate views in order to maximise the scores for each of the three domains (A: Mitral valve , B: Target septum and C: Other structures). This was termed ‘all views’ analysis.
157
Figure 4.39: Flow chart describing method of analysis for ICE and TTE in Phase 1
In Phase 2, I assessed the potential of ICE to be used in ASA instead of TTE. Therefore in Phase 2 ICE images were acquired prior to TTE. Analysis of sections A, B and C was performed as in Phase 1. The interventional cardiologist performing ASA was asked to comment on the ability to perform the procedure using ICE as the sole imaging modality. This relied on visualising relevant anatomy and highlighted myocardium with myocardial contrast injection. If this was not possible the reasons were documented as ‘myocardial contrast not visualised satisfactorily’, ‘inability to see detail beyond contrast due to acoustic shadowing’, or ‘n/a’ if no alcohol was delivered for technical reasons. TTE was the gold standard for this aspect of analysis.
158
Figure 4.40: Flow chart describing method of analysis for ICE and TTE in Phase 2
4.3.5 Statistical analysis
Statistical analysis was performed using Stats Direct version 2.8.0.0. Wilcoxon’s signed rank test was used to compare the non-parametric ICE and TTE scores. Agreement between reviewers was assessed using descriptive statistics and Kappa statistics.
4.3.6 Research approvals
Appropriate permissions to perform this study were approved by the Research and Development board of Liverpool Heart and Chest Hospital (see Appendix 4.1). A patient information leaflet was also created to assist in patient education prior to informed consent to use this new technology (see
Appendix 4.2).
159
4.3.7 Research grants
This section of study was supported by an Investigator Initiated Study Grant award from Biosense
Webster.
160
4.4 Results
4.4.1 Phase 1
4.4.1.1 Work view
TTE scored better when analysing the ability of echo to comment on the relationship of the AMVL with the target septum in section A; scoring 1.4 (±0.63) vs ICE score of 0.4 (±0.65) (p<0.0001). ICE images were acquired in a long axis view as per the standard set of views, this was in addition to the views used for the ASD closure procedure. In most of the acquired loops the view was similar to a
TTE 4-chamber. This displays the inferior septum. This is not the correct target zone for myocardial infarction during ASA – a more focussed view of the septum with the LVOT in view is required to see the true target. This was not the intention of the ICE operator but the relevant structures for ASA were not in view. With the less favourable ICE images a score of 0 was awarded as the reviewers felt unable to comment on the motion of the MV in this view rather than one that incorporated the
LVOT. A further exploration of this is detailed in the discussion section (4.5.1).
n=25 ICE TTE p A – MV 0.4 1.4 <0.0001 B – Target septum 1.66 1.08 <0.0001 C – Other structures 1.12 1.04 0.38
Table 4.19: Mean scores ICE vs TTE Phase 1: Work view analysis
161
ICE scored better when describing the detail of septal architecture in section B; scoring 1.66 (±0.59) vs 1.08 (±0.37) (p<0.0001). ICE was able to delineate the RV/LV fusion line in most cases, enabling higher scores than TTE which generally lacked this definition. TTE had an advantage in a small number of cases when ICE could not visualise the RV endocardial border as the probe position was too close to the septum.
There was no significant difference in viewing distal structures relevant to ASA in section C, ICE scoring 1.12 (±0.63), TTE scoring 1.04 (±0.29) (p=0.38). Whilst the total scores were similar, there was some difference in which of the three structures was visible with each imaging modality. A total of 50 assessments were possible; 2 expert reviewers opinions on 25 sets of images. ICE was able to visualise the mid-distal septum in 38/50 scores (76%), whilst TTE saw this in nearly all; 48/50 (96%).
The RV cavity was seen in 40/50 (80%) ICE cines and 47 (94%) of TTE films. Visualising the anterior papillary muscle was deemed possible in 28/50 (56%) of ICE films, whereas TTE was only able to see this in 10/50 (20%). This was mostly due to a misinterpretation of papillary muscle anatomy during collection of the images (see discussion).
n=50 ICE n(%) TTE n(%) Mid-distal septum 38 (76) 48 (96) Anterior papillary muscle 28 (56) 10 (20) RV cavity 40 (80) 47 (94)
Table 4.20: Other structures visible as part of section C: Phase 1: Work view
162
4.4.1.2 All views
The ability to use multiple views made little difference to scores. TTE again scored better in section
A; 1.4 (±0.63) vs 0.64 (±0.91) in the ICE section (p<0.0001). The score for ICE increased slightly with the target septum becoming visible in some views, reviewers were then able to score >0 for section
A. ICE again scored better in section B; 1.66 (±0.59) vs 1.08 (±0.37) (p<0.0001), scores were identical to work view analysis. There was no significant difference between ICE and TTE in section C; ICE scored 1.48 (±0.77) vs 1.24 (±0.43) in TTE (p=01893). The scores for both modalities increased in section C, the flexibility of multiple views allowing visualisation of other structures. This was solely due to the ability to see both papillary muscles.
n=25 ICE TTE p A – MV 0.64 1.40 <0.001 B – Target septum 1.66 1.08 <0.001 C – Other structures 1.48 1.24 0.19
Table 4.21: Mean score ICE vs TTE Phase 1: All views
n=50 ICE n(%) TTE n(%) Mid-distal septum 39 (78) 48 (96) Anterior papillary muscle 34 (68) 15 (30) RV cavity 40 (80) 47 (94)
Table 4.192: Other structures visible as part of section C: Phase 1: All views
163
4.4.1.3 Agreement between reviewers:
Agreement of reviewers’ scores was assessed using descriptive statistics. The number and percentage of scores agreed upon between reviewers is displayed (i.e. if the reviewers agreed that the score was the same in 20 sets of images the display would be 20/25 (80%)).
Two basic calculations were then performed:
Summative difference; describing one reviewer consistently scoring higher:
- A total score was calculated by subtracting the score of reviewer B from reviewer A. The
scoring system ranges from 0-2. There are 25 patients. If there was complete disagreement
the maximum score if one reviewer scored all images as 2 and the other rated all images as 0
is therefore 25x2=50. When the trend towards agreement was high this score was close to 0.
A negative value represents reviewer B scoring higher than reviewer A. The mean difference
between scores is also calculated, this is the total score/number of observations.
- This allows me to describe any pattern towards higher scoring. If the absolute agreement
level is low, a high total and mean difference will point towards one reviewer scoring
significantly higher in most cases.
Absolute difference; describing total difference between reviewers:
- It is possible that the reviewers will disagree and vary in how highly they score each set of
images. For example Reviewer A may score 2, 0, 2, 0, 2, 0. If Reviewer B scores 0, 2, 0, 2, 0, 2
the differences will be calculated as 2, -2, 2, -2, 2, -2; when added up the score will be 0 and
appear that the agreement was excellent, when they actually disagreed on every score. The
164
absolute difference allowed me to observe total disagreement over the 25 sets of scores,
but did not allow me to comment on any trend to one reviewer scoring more highly. The
absolute difference in this example is 12.
Work view
ICE:
The reviewers agreed on 19/25 (76%) of scores when describing the mitral valve in relation to the septum in section A. The summative total score difference was -4, the mean difference was therefore -0.16 (see Table 4.9), suggesting a mild pattern of Reviewer B scoring higher. The absolute difference suggested the reviewers disagreed on the score by an average of 0.48, meaning each time they disagreed it was by 2 points.
There was 18/25 (72%) agreement when describing the basal septal structure in section B. The summative total score was -3, giving a mean score difference of -0.12. The absolute difference was higher at 9 with an average score of 0.36. There was lower agreement in 13/25 (52%) of scores when describing ability to see relevant distal structures in section C. The summative total score difference was 10, and mean score difference was 0.4. The absolute difference was 12 (mean 0.48) suggesting reviewers differed in nearly half of scores allocated. The agreement in sections A and B is good, with reviewer B tending to score slightly higher than reviewer A. There is less agreement in section C, reviewer A scored higher than reviewer B in 11 cases.
Phase 1:
ICE work view A
165
Agreement 19/25 (76%) Summative total score -4 Summative mean difference -0.16 Absolute total score 12 Absolute mean difference 0.48 B Agreement 18/25 (72%) Summative total score -3 Summative mean difference -0.12 Absolute total score 9 Absolute mean difference 0.36 C Agreement 13/25 (52%) Summative total score 10 Summative mean difference 0.4 Absolute total score 12 Absolute mean difference 0.48
Table 4.20: Agreement scores Phase 1: ICE work view
TTE:
The reviewers agreed on the score in 15/25 (60%) cases for section A, the summative total score was
-8 giving a mean score of -0.32. The absolute total score was 10 (mean 0.4) confirming that the difference in agreement was predominantly due to one reviewer (B) scoring higher in most images.
In section B there was again agreement in 15/25 (60%) cases, The summative and absolute scores were identical to section A. There was better agreement in section C, scores were identical in 21/25
166
(84%) cases, a summative total score of -2 and mean score -0.08. The absolute difference was 4 with a low mean difference of 0.16.
Phase 1:
TTE work view A Agreement 15/25 (60%) Summative total score -8 Summative mean difference -0.32 Absolute total score 10 Absolute mean difference 0.4 B Agreement 15/25 (60%) Summative total score -8 Summative mean difference -0.32 Absolute total score 10 Absolute mean difference 0.4 C Agreement 21/25 (84%) Summative total score -2 Summative mean difference -0.08 Absolute total score 4 Absolute mean difference 0.16
Table 4.21: Agreement scores Phase 1: TTE work view
167
Agreement on chosen cine loops:
Each reviewer was able to choose the best cine loop from a selection of those acquired. This was the projection that allowed the reviewer to maximise the score according to the predefined scoring system.
ICE: Most ICE images were of a very similar projection. Multiple loops were taken to give reviewers as much choice as possible. There was therefore minimal agreement between reviewers with regard to the optimal loop to maximise scores. The reviewers agreed on the best loop in just 5/25 (20%) cases.
TTE: The reviewers agreed on chosen views in 18/25 (72%) cases. The predominant view chosen was the parasternal long axis (PLAX). Figure 4.3 shows the chosen views of each reviewer.
168
Figure 4.41: Reviewers chosen work views TTE Phase 1
All views
ICE:
The reviewers agreed on the scores for all 25 sets of images when allowed to use multiple ICE cines to score section A. The total and mean summative and absolute scores are therefore both 0. There was agreement in 18/25 (72%) cases when scoring section B, a total summative score of -3 and mean of -0.12. The absolute difference was 9 with a mean of 0.36: Reviewer B scored higher by 1 point in 6 images and reviewer A scored higher in 3 images. There was total agreement again in section C.
Phase 1:
ICE All views A Agreement 25/25 (100%) Summative total score 0 Summative mean difference 0 Absolute total score 0 Absolute mean difference 0 B Agreement 18/25 (72%) Summative total score -3 Summative mean difference -0.12 Absolute total score 9 Absolute mean difference 0.36
169
C Agreement 25/25 (100%) Summative total score 0 Summative mean difference 0 Absolute total score 0 Absolute mean difference 0
Table 4.22: Agreement scores Phase 1: ICE all views
TTE:
There was a similar pattern across sections A-C when assessing scores allocated for all available views for TTE. In section A 15/25 (60%) scores were identical, a total summative score of -8 was observed and a mean score of -0.32. An absolute difference of 10 and mean of 0.4 was seen:
Reviewer B scored higher on 9 occasions and reviewer A scored higher once. Section B scores were exactly the same as section A. In section C the summative scores and absolute scores were similar; total score of -8 with mean of -0.32 suggested Reviewer B scored higher in most circumstances.
When reviewing the absolute score we see a total of 8 and mean of 0.32, this means the reviewers disagreed on 8 occasions with Reviewer B scoring higher each time.
The trend in this section was for reviewer B to give higher scores than reviewer A. This was seen in all sections with equal spread.
170
Phase 1:
TTE All views A Agreement 15/25 (60%) Summative total score -8 Summative mean difference -0.32 Absolute total score 10 Absolute mean difference 0.4 B Agreement 15/25 (60%) Summative total score -8 Summative mean difference -0.32 Absolute total score 10 Absolute mean difference 0.4 C Agreement 15/25 (60%) Summative total score -8 Summative mean difference -0.32 Absolute total score 8 Absolute mean difference 0.32
4.4.2 Phase 2
4.4.2.1 Work view
ICE was superior when assessing scores in section A; a mean of 1.88 (±0.28) was higher than the TTE value of 1.60 (±0.45) (p=0.02) (see Table 4.13). The lessons learned from Phase 1 were critical in optimising images to ensure the true target septum was in view. There was no significant difference in assessing ability to view the detail of the target septum in section B; the ICE score was slightly
171 higher at 1.28 (±0.64) vs the TTE score of 1.08 (±0.18), but this did not reach statistical significance
(p=0.30). The ability of ICE to see the intra-septal fusion line was greater in this section, but it often lost points on the inability to clearly see the RV endocardial border. The probe was usually positioned very close to the septum; the echo window therefore did not include any RV cavity. TTE scored higher when describing key, distant structures in section C; 1.03 (±0.20) vs 0.63 (±0.39) in ICE
(p=0.002). The low scores were due to the inability of both modalities to see the anterior papillary muscle. The mid septum and RV cavity were visible in almost all TTE images, and in a smaller number of ICE images (see Table 4.12).
n=40 ICE n(%) TTE n(%) Mid-distal septum 27 (67.5) 40 (100) Anterior papillary muscle 1 (2.5) 1 (2.5) RV cavity 31 (77.5) 40 (100)
Table 4.23: Key distant structures visible as part of section C: Phase 2: Work view
ICE TTE p A – MV 1.88 1.60 0.02 B – Target septum 1.28 1.08 0.30 C – Other structures 0.63 1.03 0.002
Table 4.24: Mean scores ICE vs TTE Phase 2; work view analysis
172
4.4.2.2 All views
The ability to adapt the probe/catheter position and use different views to focus on each section of scoring had a greater effect in ICE than in TTE. ICE remained superior in section A, scoring 1.93
(±0.18) vs 1.63 (±0.39) in TTE (p=0.007). The difference in section B was more marked, with the p value of 0.06 nearing significance at 0.05 level; ICE now scored 1.35 (±0.49) vs 1.1 (±0.21) in TTE
(p=0.06). There was no difference between scores in section C when allowed multiple views; ICE scoring 0.85 (±0.43) vs 1.075 (±0.24) in TTE (p=0.13). This is in contrast to the work view analysis when TTE scored significantly better.
ICE TTE p A – MV 1.93 1.63 0.007 B – Target septum 1.35 1.10 0.06 C – Other structures 0.85 1.08 0.13
Table 4.25: Mean scores ICE vs TTE Phase 2; work view analysis
4.4.2.3 Agreement between reviewers
Agreement of reviewers’ scores was assessed in the same manner as Phase 1.
Work view
ICE:
173
The reviewers agreed on 16/20 (80%) of scores in section A. The summative total score was -1, the mean score is therefore -0.05. The absolute total score was 5 with a mean of 0.25. There was less agreement in section B; 9/20 (45%) scores were identical. There wasn’t a clear pattern towards one reviewer scoring higher than the other however; as the summative total score was still just 1 and mean score was 0.05. The absolute difference corroborates this with a total score of 13 and mean of
0.65. There was agreement in 13/20 (65%) of scores in section C, with a low summative total and mean score of -1 and -0.05 respectively. The absolute score was higher with a total of 7 and mean of
0.35.
Phase 2: n=20 ICE Work view A Agreement 16/20 (80%) Summative total score -1 Summative mean difference -0.05 Absolute total score 5 Absolute mean difference 0.25 B Agreement 9/20 (45%) Summative total score 1 Summative mean difference 0.05 Absolute total score 13 Absolute mean difference 0.65 C Agreement 13/20 (65%) Summative total score -1 Summative mean difference -0.05 Absolute total score 7 Absolute mean difference 0.35
174
Table 4.26 Agreement scores Phase 2: ICE work view
The pattern of agreement observed indicates that whilst the absolute agreement between individual scores ranged between 25-80%, there was no clear trend towards one reviewer allocating higher scores. This is displayed by the low summative scores and higher absolute scores in this section. In general, the reviewers agreed upon the quality of the batch of images acquired for all 20 patients.
TTE:
There was agreement between scores in 11/20 (55%) of patients in section A. The total score difference was relatively high at -8, and mean score was -0.4. Reviewer B scored higher in this category, the absolute difference scores were similar to the summative. There was greater agreement in section B; 17/20 (85%) scores were identical. The total difference was -1 with a mean score of -0.05. The absolute difference was also low at 3 and mean of 0.15. The reviewers agreed on
13/20 (65%) scores in section C; the total difference was 3 with a mean score of 0.15. This was mirrored in the absolute difference scores of 3 and 0.15.
175
Phase 2: n=20 TTE Work view A Agreement 11/20 (55%) Summative total score -8 Summative mean difference -0.4 Absolute total score 10 Absolute mean difference 0.5 B Agreement 17/20 (85%) Summative total score -1 Summative mean difference -0.05 Absolute total score 3 Absolute mean difference 0.15 C Agreement 13/20 (65%) Summative total score -3 Summative mean difference -0.15 Absolute total score 3 Absolute mean difference 0.15
Table 4.27: Agreement scores Phase 2: TTE work view
The disagreement in scores achieved in section A stands out most in work view analysis of TTE for
Phase 2. Reviewer A felt less able to comment on the intricate detail of the SAM-septal contact point; there was the ability to localise the contact approximately (score = 1), but not to measure length of the AMVL in contact (score = 2). Reviewer B felt more confident allocating a score of 2. This
176 occurred 7 times. There was better agreement in the other sections, with low summative and absolute score differences.
Chosen cine loops:
Each reviewer was again able to choose the best cine loop from a selection.
ICE: A similar pattern to Phase 1 was seen, as many cines were available the chosen loop correlated in only 8/20 (40%) cases. A full avi film of each reviewer’s chosen ICE and TTE images for this assessment is available as supplemental file 1.
TTE: The reviewers agreed on chosen views in 18/25 (72%) cases. The predominant view chosen was the parasternal long axis (PLAX). Figure 4.4 shows the chosen views of each reviewer.
177
Figure 4.42: Reviewers chosen work views TTE Phase 2
All views
ICE:
Absolute agreement on scores allocated was seen in 17/20 (85%) cases in section A. In each disagreement reviewer A scored higher, the summative and absolute differences were therefore both 3 and 0.15 respectively. There was agreement in scores allocated in 6/20 (30%) cases in section
B. There was variation around a central scoring threshold though as the summative total and mean differences were 0. This is exemplified by the absolute score of 16 and mean of 0.8. The reviewers therefore disagreed by an average of 0.8 marks out of 2, but with no clear pattern towards one scoring higher than the other as they were balanced out to a summative score of 0. 12/20 (60%) scores in section C were identical, with a summative total and mean score of 2 and 0.1. The absolute differences were slightly higher at 8 and 0.4 respectively.
178
Phase 2: n=20 ICE all views A Agreement 17/20 (85%) Summative total score -3 Summative mean difference -0.15 Absolute total score 3 Absolute mean difference 0.15 B Agreement 6/20 (30%) Summative total score 0 Summative mean difference 0 Absolute total score 16 Absolute mean difference 0.8 C Agreement 12/20 (60%) Summative total score 2 Summative mean difference 0.1 Absolute total score 8 Absolute mean difference 0.4
Table 4.28: Agreement scores Phase 2: ICE all views
TTE:
Agreement was seen in 12/20 (60%) cases in section A; a summative total and mean difference score of -9 and -0.45 were observed. The absolute score was similar at 10 and 0.5 indicating the difference in the summative score was due to reviewer B again scoring higher.
179
Section B was scored identically in 16/20 (80%) cases, the summative total difference was 0. The absolute difference was 4 with mean score of 0.2, this indicates the reviewers disagreed 4 times with each reviewer scoring higher twice. Section C scores were agreed upon in 15/20 (75%) cases, a summative total and mean score of 3 and -0.15 were seen. The absolute scores were 5 and 0.25 respectively.
The pattern of TTE scoring in this section was very similar to the work view analysis. Reviewer B scored higher than reviewer A, this was most marked in section A.
180
n=20 All TTE A Agreement 12/20 (60%) Summative total score -9 Summative mean difference -0.45 Absolute total score 10 Absolute mean difference 0.5 B Agreement 16/20 (80%) Summative total score 0 Summative mean difference 0 Absolute total score 4 Absolute mean difference 00.2 C Agreement 15/20 (75%) Summative total score -3 Summative mean difference -0.15 Absolute total score 5 Absolute mean difference 0.25
Table 4.29: Agreement scores Phase 2: TTE all views
Kappa agreement
Due to the relatively low number of observations in each category and the narrow spread of values
Kappa agreement statistics for each section (modality ICE/TTE, work view/all views and Phase
1/Phase 2) were not thought to be as representative as the chosen descriptive statistics detailed in sections 4.4.1.3 and 4.4.2.3.
181
I have analysed the Kappa data as a statistical exercise for Phase 2. Although the values observed are statistically significant in the areas where the reviewers agreed on all scores (mainly ICE all views), it takes very little fluctuation in the reviewers scores to reduce the agreement values considerably.
If I use the Phase 2 ICE work view as an example:
In the descriptive statistics we see that the mean difference between these reviewers over 20 observations is just 0.05 in each category, including the overall score. This would suggest an overall reasonable agreement. However when using the Kappa statistics we see that there is at best a moderate agreement with predominantly ‘fair’ values of 0.26-0.50. Many of the p values are also insignificant. The chosen descriptive statistics gives more information about the overall agreement of the image quality.
182
Phase 2: n=20 ICE Work view A Agreement 16/20 (80%) Summative total score -1 Summative mean difference -0.05 Absolute total score 5 Absolute mean difference 0.25 B Agreement 9/20 (45%) Summative total score 1 Summative mean difference 0.05 Absolute total score 13 Absolute mean difference 0.65 C Agreement 13/20 (65%) Summative total score -1 Summative mean difference -0.05 Absolute total score 7 Absolute mean difference 0.35
Table 4.30 Agreement scores Phase 2: ICE work view
183
Kappa ICE TTE
Work All Work All Score View views View Views
Kappa p Kappa p Kappa p Kappa p
Non- A weighted 0.26 0.0920 1.00 <0.0001 0.33 0.0120 0.33 0.0120
Weighted 0.26 0.0920 1.00 <0.0001 0.47 0.0007 0.46 0.0007
Non- B weighted 0.27 0.0071 0.27 0.0071 0.09 0.1155 0.09 0.1155
Weighted 0.37 0.0071 0.37 0.0071 0.14 0.0112 0.14 0.0112
Non- C weighted 0.28 0.1080 1.00 <0.0001 0.28 0.0281 0.37 0.0021
Weighted 0.41 0.0006 1.00 <0.0001 0.30 0.0120 0.40 0.0008
Weak Fair Moderate Strong Very strong
Figure 4.43: Kappa agreement statistics for Phase 2.
Overall agreement
When we have a greater number of values to compare the Kappa statistic becomes more useful. The k statistic value over all observed scores throughout Phase 1 and Phase 2 was 0.55, (p<0.0001). For each pair of observations, the observed mean difference was 0.36 (95% CI 0.29–0.43). This suggests that, on average, for each three sets of paired score groups, there is an isolated difference of a single
184 point (e.g. reviewer A scores 2, 2, 1 in the respective three sections and reviewer B scores 2, 1, 1).
This demonstrates a good clinical agreement.
4.4.2.4 Contrast localisation
ICE
Good contrast localisation with sufficient image quality to inject alcohol with ICE guidance alone was seen in 2 cases. One patient has significant left main stem disease identified at the time of ASA – no injections were performed as he was subsequently sent for surgical myectomy and coronary artery bypass grafting. In 8 patients the injection of contrast resulted in the inability to see sufficient detail of contrast localisation due to significant acoustic shadowing. In some ICE images the contrast did not opacify the myocardium adequately (n=6). The operator could not confidently comment on localisation in these cases (n=14). Contrast could be clearly seen in paired, simultaneous TTE images.
In 3 cases the ICE failed to see the contrast in the field of view, the catheter was focussed on the target myocardium. In these 3 cases nothing was seen in the ICE images and TTE confirmed that contrast was visible in cardiac structures away from the target (RV cavity in 2, mid septum in 1).
The ability to comment on contrast localisation is critical to safe and accurate ASA and was significantly better with TTE.
185
Observation Incidence (n=20) Significant acoustic shadowing 8 Inadequate opacification 6 Contrast localised to distal structures 3 Possible using just ICE 2 No injection 1
Table 4.31: Observations on ICE following contrast injection
TTE
TTE performed very well in this section with good visualisation of MCE in all. Ultimately all alcohol injections were guided by TTE as contrast could be seen in all 19 patients without the phenomena described in ICE.
186
4.5 Discussion
4.5.1 Phase 1
TTE scored higher in section A – observing motion of the MV. ICE was superior in viewing detail of the septum and there was no difference in ability to see other structures relevant to ASA or the overall score. The agreement between reviewers was good in most sections. The predominant TTE view to comment on relevant anatomy was the parasternal long axis view.
Phase 1 was designed partly as a feasibility study. Some important anatomical lessons were noted. In the majority of ICE images acquired the focus of the echocardiographic field was the inferior septum.
When the catheter is positioned in the RV inlet and rotated to face the septum the natural position focusses here. This is a view akin to a 4-chamber in TTE. To see appropriate target myocardium in continuity with the LVOT further right rotation is required on the ICE catheter. This brings the true
LVOT in to view. An example of this is shown in figure 4.6. The reviewers felt unable to comment on the relationship of the MV with the target septum as the LVOT was not in view.
187
Figure 4.44: Demonstration of inferior septum vs target septum in ICE and CMR images
Panel A shows a typical ICE image taken from phase 1. It is similar projection to a TTE 4-chamber view taken from inside the RV. Panel C shows a typical 4 chamber view In CMR. Panel E shows a short axis cut. The blue line represents the cut of the LV that is shown in panel C. The shaded area is the target for ASA. The blue line clearly passes through the inferior septum, missing the target area. Panel B shows an ICE image with the true LVOT in view. This is represented in the 3-chamber CMR image in panel D. The shaded area is the target for ASA. Panel F shows the corresponding short axis view, the blue line this time passes through the target myocardium in the true septum.
188
A papillary muscle was often easily visible, and interpreted initially as the anterior muscle. A further investigation of papillary muscle architecture using relational images from CMR showed this to be the posterior papillary muscle (see Figure 4.7). Historically the posterior muscle has been described in anatomical studies as receiving blood supply from the right coronary system130. It is unlikely to be of vascular relevance when injecting contrast into septal vessels originating from the left coronary system. This was therefore not part of the scoring system. It was rare to visualise the anterior papillary muscle in the same plane as the target septum and LVOT. It is usually possible to seek out this structure with TTE using a 2-chamber view, but less straightforward with ICE. The misinterpretation by those acquiring the images meant that many of the echo cines showed only the posterior muscle.
189
Figure 4.45: Demonstration of papillary anatomy in standard echo and CMR planes
Panel A shows a typical TTE PLAX echo image, the papillary muscle in view is the posterior muscle. Panel B shows a typical ICE long axis view of the left ventricle, with both heads of the posterior muscle visible. Panel C shows a 3-chamber CMR still with a papillary muscle labelled. Relational short axis CMR images show this to be the posterior muscle (PP), the blue line represents the plane displayed in the 3 chamber view in panel C. The anterior muscle (AP) is out of shot and would be very difficult to see in the same plane as the LVOT. The same perspectives are displayed in the non- HCM heart in panels E and F.
190
4.5.2 Phase 2
During phase 2 images were optimised to ensure the target septum in connection with the LVOT was in view. This optimisation of image allowed ICE to score much better. ICE was now superior to TTE when visualising the SAM-septal contact area.
The ability of ICE to describe septal architecture was reduced in phase two. There was no significant difference between ICE and TTE in section B. This was because ICE was often not able to comment on the RV endocardial border. The catheter seemed to lie on the RV endocardial septal surface on more occasions. This is perhaps a reflection of the altered anatomy seen in HCM. The RV inlet area was smaller and less able to accommodate the ICE catheter. The geometry of the heart in the thoracic cage and hence angle of approach may also have been altered by the grossly hypertrophied left heart. When it was possible to create space between the catheter and the septum the ICE images were superior in detailing LV endocardium and intra-septal fusion line.
TTE was superior to ICE in detailing distant structures in section C. In the HCM population the anterior papillary muscle was seen only once in all of the working views that focussed on the LVOT.
The anterior displacement of papillary muscle architecture previously described in HCM may explain this133. As any structure moves anteriorly in the LV it moves further away from a plane of image that incorporates the LVOT (see Figure 4.7). The proximity of the ICE catheter to the target area meant that the echocardiographic field was more limited. We again noted the inability to see RV cavity in
22.5% cases, and were unable to see the mid-septum in 32.5% of cases. TTE saw the RV and mid- septum in all cases. The predominant TTE work view for assessing sections A-C was again the parasternal long axis view (80-90% cases).
191
Agreement between the reviewers was good in the ICE section. TTE displayed a trend towards reviewer B scoring higher, this was most notable when commenting on section A.
4.5.3 Myocardial contrast echocardiography
Visualisation of contrast distribution following coronary injection is imperative to ASA. It provides an accurate assessment of the perfusion bed of the chosen septal artery. The localisation of contrast predicts dispersion of alcohol and therefore subsequent infarct. Its use has improved accuracy of iatrogenic infarct and reduced complications such as pacemaker implantation42. Several problems were encountered when using ICE to visualise contrast.
4.5.3.1 Extensive acoustic shadowing
In 8 patients the contrast opacification was vivid but the acoustic shadowing was so pronounced it was impossible to comment on cardiac structures beyond the contrast. Figure 4.8 gives an example of this phenomenon with comparative TTE injections.
192
Figure 4.46: Example of excessive contrast opacification on ICE
Panels A and B show ICE images. Panels C and D show paired TTE images. Panels B and D are simultaneous images taken with myocardial contrast injection. There is significant acoustic shadowing seen following contrast injection in the ICE image (panel B). There is minimal shadowing seen on TTE (panel D).
4.5.3.2 Inadequate visualisation of injected contrast
This was observed in 6 cases. When contrast was injected we did not see any significant change to the composition of target myocardium on ICE images. When the same injection was repeated on TTE the contrast was seen quite clearly (see Figure 4.9).
193
Figure 4.47: Example of inadequate contrast opacification on ICE
Panels A and B show ICE images. Panels C and D show paired TTE images. Panels B and D are simultaneous images taken with myocardial contrast injection. The contrast is much more vivid in TTE (panel D) than ICE (panel B).
4.5.3.3 Contrast localising to distant structures
There were 3 examples of contrast localising to distant structures only rather than the target area.
These patients did not receive alcohol injection. This was not visualised on the ICE working view but was seen on TTE. The contrast had passed by the time the ICE catheter was manipulated into the
194 correct position. It is impossible to know in these cases whether the contrast is in the correct area and cannot be seen due to issues with the use of ICE, or if it localised outside the target area and echocardiographic window. There were only two cases in which the procedure could potentially have been performed using just ICE contrast images.
4.5.3.4 Alteration of echocardiographic settings
We adapted frequency settings on the ICE probe to attempt to see contrast in those cases with inadequate opacification of myocardium. The baseline of 11MHz was reduced sequentially down to
5MHz. This allowed us to see contrast in some, but the image quality was so poor we were unable to relate it to relevant anatomy (see Figure 4.10). Using tissue harmonics settings (5.0/9.5MHz) had a similar effect.
Figure 4.48: Example ICE images using differing frequencies and tissue harmonics
195
Panel A shows a long axis ICE view of the left ventricle using a frequency of 11MHz. Panel B shows the same transducer positon with a frequency of 5MHz. Panel C shows a myocardial contrast injection at a frequency of 5MHz. Myocardial contrast is visible in the basal septum in panel C, but to allow the operator to see this a reduction in frequencies is required. This degrades the image to the extent that it cannot be relied upon to accurately comment on cardiac anatomy.
4.5.3.5 TTE vs ICE in myocardial contrast echocardiography
TTE is the gold standard in MCE. Contrast agents are predominantly used in TTE to opacify the LV cavity when assessment of function is difficult due to low image quality. This is often seen in obese patients and in stress echocardiography when positioning may not be ideal. The development of these agents has been targeted at providing appropriate backscatter when ultrasound is projected from several centimetres away, as is the case in TTE. When image quality is good using TTE for ASA, contrast localisation can be seen well.
The technology of the Acunav™ ICE transducer is similar to that of TTE, a miniaturised phased-array probe. It is recommended as a single use catheter for venous insertion. The major differences are the proximity to the visualised structures and higher frequencies used. The signal processing techniques used in modern echocardiographic modules have previously allowed assessment of MCE in high- and low-mechanical index settings134. We therefore did not anticipate a problem visualising contrast with ICE. The paradoxical events of inability to visualise contrast and acoustic shadowing were also difficult to explain. The explanation of these phenomena is an area for future study.
The images acquired were deemed to have been of sufficient quality to progress to alcohol injection in only 2/19 (11%) cases. One patient had significant left main stem disease and required bypass
196 surgery and concomitant myectomy, he therefore did not have any MCE studies. It could be argued that no procedure could be carried out with ICE alone. If the contrast in these 2 patients had localised elsewhere we would be unable to confidently say if it was inaccurate location, or in the target area but invisible by ICE studies.
This is a major restrictor of the isolated use of ICE for echocardiographic guidance in ASA.
4.5.4 Anatomical observations from ICE
ICE images of the interaction of the anterior MV leaflet and the septum are of such high quality that they provide new insights. The target for alcohol septal ablation has previously been identified as the SAM-septal contact point, as the phenomenon of contact leads to the majority of the intra-cavity gradient formation. This can be seen in several planes of ICE imaging and could be marked and projected onto available software to accurately detail the contact area in three dimensions. The damage we aim to deliver could then be more accurate and potentially more likely to succeed. This would depend upon developing a method of non-surgical septal reduction that does not rely on septal vascular supply or the visualisation of myocardial contrast.
The ability to visualise the abnormal MV anatomy was extremely clear on ICE. It has been previously reported that the AMVL is elongated in HCM, and was theorised to be part of the genetically determined phenotype rather than the effects of haemodynamic disturbance in the LVOT135. A long
AMVL would lend itself to the pathophysiology seen in obstruction in HCM. The extra length would result in coaptation with the posterior leaflet towards its mid- point, leaving a redundant tip. This
197 redundant tip is then caught in the abnormal flow in the LV. The flexibility of the leaflet to bend back on itself is remarkable (Figure 4.11) and seen with clarity on ICE.
Figure 4.49: ICE images of SAM-septal contact
Both panels show ICE images in systole with SAM-septal contact. In each example the AMVL is folded back and has a bend of nearly 90. In panel B an area of iatrogenic infarct in the mid septum from previous ASA is seen. The contact area for the AMVL is further basal; there is viable, contracting myocardium in this area.
ICE can provide such accurate detail of myocardial composition that a coronary wire measuring
0.014” can be seen in target myocardium (see Figure 4.12).
198
Figure 4.50: Coronary wire visible with ICE
The arrow points towards the coronary wire. Acoustic shadowing is also seen. This is in the target myocardium.
199
4.6 Conclusion
There was no significant difference between ICE and TTE in the overall ability to determine
cardiac anatomy relevant to the performance of ASA. ICE visualises the SAM-septal contact
area better than TTE.
ICE does not visualise myocardial contrast well and therefore cannot be used in the
assessment of myocardial supply by septal coronary arteries during ASA
200
4.7 Limitations
In the planning stage of this study I had intended to use the same ICE operator for all patients in
Phase 1, and the same ICE operator for all patients in Phase 2. I achieved this. I also intended to use the same sonographer for each set of acquired TTE images in each phase. Unfortunately this was not possible as this relied on the availability of sonographers for clinical cases. Due to the need for patient recruitment and the desire to continue consecutive enrolment I accepted that TTE would be performed by a variety of British Society of Echocardiography accredited sonographers.
I was not able to perform intra-observer reproducibility. This was due to restrictions of the expert reviewers’ time.
201
Chapter 5: An examination of the potential role of
Computed Tomography (CT) angiography in the identification of target septal arteries prior to alcohol septal ablation for hypertrophic obstructive cardiomyopathy
202
5.1 Introduction
Accurate localisation of the iatrogenic infarct is critical in impacting LVOT flow dynamics from ASA.
We have seen that patients with persisting gradients have infarcts that occupy less of the target myocardium than those with successful treatment. A further 5 – 15% of patients entering the cardiac catheterisation laboratory with the intention of alcohol delivery receive no treatment as no suitable septal vessel can be identified or used70 80 88. In our retrospective group 18% had failure to resolve the LVOT gradient at end of treatment with ASA (with multiple procedures), a further 5% were unable to receive treatment. Twenty-three per cent therefore had unsatisfactory treatment of LVOT gradient after multiple attempts at treatment.
The use of intra-procedural intra-cardiac echocardiography (ICE) did not provide any tangible benefits in ASA beyond the use of transthoracic echocardiography. The ability to see the relevant facets of anatomy was similar overall and ICE did not provide suitable quality images of myocardial contrast. Attempts to visualise the contrast better using ICE as part of ASA did not work. We must therefore consider alternative approaches to improve accuracy of the iatrogenic infarct.
The only current method of identifying septal arteries for alcohol injection is invasive coronary angiography. Invasive angiography allows us only to detail the course and lumen of a coronary artery. We then rely on myocardial contrast echocardiography to highlight the area of myocardium perfused. CT angiography can highlight the course of a coronary artery and show the area towards which and through which it runs. This has the potential to guide selection for alcohol delivery.
Developments in CT have enabled the non-invasive imaging of coronary arteries in the investigation of chest pain136. The technology has advanced to allow imaging of smaller structures, with a collimation distance of 0.6mm and temporal resolution of 75msec. This detail might allow us to track
203 septal vessels in to target myocardium, and offers the dual benefit of angiographic and anatomic information. If it is possible to accurately describe the course of septal vessels we can plan the chosen artery for alcohol delivery before entering the cardiac catheterisation laboratory.
It is increasingly recognised that septal vascular supply can take its origin from vessels other than the left anterior descending (LAD) artery. Although the anterior two-thirds of the septum are usually supplied by vessels from the LAD, septal arteries have been reported to originate from diagonal arteries, the circumflex artery, left main stem or even right coronary artery97 137-139. The optimum target vessel may, in some cases, never even be considered for exploration. This might explain the inability to identify a target vessel in 5-15% cases.
We have seen poor outcome following ASA with right ventricular septal infarct localisation on CMR imaging in our group and the work of others111. Right bundle branch block was seen in 62% of our patients and is reported in other series in approximately half of patients who receive alcohol, this is likely to be indicative of RV septal damage54. It has been reported that ASA causes a transmural infarct pattern on CMR, often with the majority of damage localising to the RV endocardial surface124. Although this has been observed in successfully treated patients, theoretically the RV insult should have minimal effect on LV haemodynamics. The damage localised to the basal portion of the LV septum interacting with the MV in SAM remains the target for ASA. The RV damage seems to be a bystander effect of the inability to control alcohol dispersion with current methods. This could explain the poor correlation of alcohol volume delivered and myocardial damage as measured by enzyme release. Some of the alcohol injected enters into LV myocardium with densely packed, hypertrophied muscle - a high pressure environment. Some alcohol enters in to the RV myocardium with less hypertrophy and direct drainage to the RV cavity - a low pressure environment. The inability to predict what proportion of alcohol goes to which environment leads to unpredictable tissue dwell times and hence unpredictable infarction size and localisation. The level of precision of
204 high quality CT images with current technology might provide us information about the dual vascular territory supply of septal arteries. If we can localise alcohol delivery to the LV arteries only it may translate to improved accuracy and clinical outcome.
Perhaps just as importantly, it is possible to characterise other potential septal vessels with incorrect myocardial distribution and hence eliminate the need for them to be subjected to invasive exploration. Instrumentation of any coronary artery carries a risk. Exploration with x-ray and echocardiographic contrast takes time and increases the risk of contrast related complications. This could be avoided by correctly identifying an inappropriate distribution from pre-procedural CT planning, potentially reducing procedural time and contrast load.
Most analysis of coronary arteries with CT is based on large, epicardial vessels responsible for transporting large volumes of blood to myocardium. This is relevant when investigating the possibility of coronary insufficiency in chest pain. The assessment of calcium load and stenosis is performed on 2-dimensional and then 3-dimensional imaging. Once CT images are acquired they can be manipulated and reconstructed to allow a rotational 3D map.
If septal arteries can be correctly identified in target myocardium in single plane reformat images, a multiplanar reconstruction will allow the accurate prediction of the angiographic appearance for any given projection in the catheterisation lab. The CT coronary tree can then be rotated to allow the operator to visualise the path of target arteries before ASA. Optimal projections can then be chosen to remove foreshortening and overlap and provide accurate information when attempting to access these arteries with coronary wires.
This combination of potential benefits from pre-procedural planning based on CT angiography could allow us to improve accuracy of infarct, and minimise risk associated with the instrumentation of
205 inappropriate coronary arteries. This may help to achieve a more predictable and uniform outcome when treating LVOT gradients in HOCM.
To assess the feasibility of using CT to guide ASA procedures I chose to first examine a series of existing CT coronary scans performed for other reasons.
206
5.2 Aims
To examine pre-existing CT coronary angiography scans performed at LHCH to:
Assess the ability of CT angiography to comment on septal coronary anatomy pertinent to
ASA.
Describe the anatomy of septal arteries
207
5.3 Methods
5.3.1 Scoring system
5.3.1.1 Required septal anatomy information:
In order to assess the ability of CT to provide useful information on septal arterial anatomy to guide
ASA we must first decide what information is desirable. An initial exploration phase involved an expert cardiac Radiologist examining a series of 5 CT scans in order to assess feasibility of visualising septal arteries. During this phase I was able to define what details of the cardiac anatomy we require from the images.
Parent artery
Anatomical observations describe septal arterial supply to originate predominantly from the LAD artery97. This has been observed to be the parent artery for the majority of ASA procedures in ours and many other groups. Case reports have highlighted target arteries originating from alternative epicardial vessels137-139. If it is known that the septal vessel takes it origins from the circumflex, right coronary, diagonal or intermediate arteries before ASA it would reduce procedural time and potentially open the procedure up to those who previously could not receive alcohol70 80.
Bifurcation angle
The angle of the septal artery from its parent artery has implications for intervention. If the angle is obtuse the septal vessel can usually be accessed with standard coronary wires and appropriate wire shaping. If the angle is acute entry in to the vessel is more difficult. Techniques developed in coronary intervention such as the use of microcatheters and wire deflectors have allowed wire
208 manipulation into target vessels. Advanced warning of this anatomical variant is useful in planning appropriate equipment and approach to accessing these vessels.
Septal artery area / diameter
Knowledge of the cross-sectional dimensions of the target vessel will help with planning appropriate over-the-wire balloon size for the procedure. This knowledge will improve planning and identification of the artery once invasive angiography pictures are acquired.
Length of the septal vessel in epicardial fat and myocardium
The traditional teaching for procedural technique in ASA is to site the balloon at the ostium of the septal vessel near its origins from the parent vessel28. When following these recommendations it has been observed that the balloon often migrates back into the parent vessel, a phenomenon known as
‘melon-pipping’. When this occurs it partially occludes flow down the parent vessel. This theoretically can cause ischaemia in the distal territories. Ischaemic myocardium responds by recruiting collateral circulation, often via septal arcades. Septal arcades are well described and extensive97. If a septal vessel has been recruited to assist with blood supply in the distal portion of the parent vessel any alcohol injected will travel to the epicardial coronary circulation and cause unwanted, distal infarction.
Most septal arteries travel initially in epicardial fat before entering myocardium. These are very different environments. If a balloon is inflated with the distal portion in muscle, and the proximal portion in fat it will move towards the area of lower resistance, sliding back in to the epicardial fat towards the parent vessel. Armed with the knowledge of the length of the septal vessel in fat prior to entering muscle we can actively prevent this happening by siting the balloon wholly in muscle.
209
Knowledge of the septal vessel length in the myocardium will help to identify the correct vessel when reviewing invasive angiograms in the cardiac catherisation lab. It will also give us information to guide positioning of the coronary balloon to occlude the vessel prior to contrast and alcohol injection.
Length of vessel prior to bifurcation
It is important to know if a septal vessel splits and provides blood to different territories. Injection of alcohol into the ostium of a vessel that supplies territories such as the LV basal septum and RV cavity will have varying effects. The run off into such territories is unpredictable and the lack of correlation of enzyme release to alcohol dose may reflect the inability to control infarct size when injecting into the common stem of such arteries. If we are armed with the knowledge of an important early bifurcation into different destinations we can manipulate the coronary wire to direct the balloon into the appropriate branch.
Landmarks
Being able to reference target septal arteries to local angiographic landmarks will allow the operator to be more secure in choosing the predicted artery. The vessel dimensions are small and there are often millimetres between vessels leaving the LAD. Knowing the relationship of our target to nearby diagonals vessels, or being able to measure from an important bifurcation will reduce risk of misinterpretation.
210
Myocardial distribution
This is critical. Knowing where a vessel travels to will allow us to choose or dismiss an artery on first angiographic image acquisition. Being able to track CT contrast into target myocardium will predict blood supply and provide a target for alcohol.
Preferred projection ostium
Invasive angiography converts a three dimensional structure in to a two dimensional image. In two dimensions it is difficult to be certain of the true orientation of the vessels. This has implications for identifying the origins and angle of entry to subsidiary coronary arteries. It is easy to be misled by foreshortening and overlap. This can lead to futile attempts from the operator to wire the artery before changing projection and gaining more information. With a CT based three dimensional rotational coronary artery map we can be certain of the origins of vessels and remove foreshortening or overlap. The projections can be noted from CT and taken in to the lab to predict optimal angles.
Preferred projection distal
The same issues apply to the distal target vessels. As the operator may be required to navigate several turns within the septal vessel we need information throughout the full length, which can be up to 60mm. As such the optimal angle to comment on the ostium may not be the same as that for the distal vessel.
211
5.3.1.2 Assessment of image quality and ability to guide ASA
Having identified the relevant structures and anatomy I needed to be able to comment on I then set a method of scoring the image quality. This was in a similar manner to the assessment of ICE in
Chapter 4.
All data elements in the scoring system describe a specific aspect of anatomy. The scoring system will examine the ability of each CT image to provide a numerical measurement for each individual element (except landmarks). The scoring system will again range from 0-2, definitions for each score are provided in Table 5.35.
Score Definition 2 Septal artery appearance commented on with high level of confidence of accuracy, clearly able to guide invasive procedure. 1 Septal artery appearance commented on with moderate level of confidence, able to provide some useful information to invasive procedure operator. 0 Unable to comment on an aspect of septal artery appearance, or commented on with very little or no confidence of accuracy. Minimal use when providing guidance to invasive procedure operator
Table 5.32: Scoring system for quality of CT images
A single experienced cardiac radiologist specialising in coronary CT scans was responsible for providing all scores. I was present for all CT analysis. CT scans were reported by a consensus. We agreed on the absolute numerical value for each constant and the confidence with which we reported each detail. This data were recorded in individual case record forms (CRF). Data was then transferred from the CRF to a specifically designed Microsoft Access Database to allow query tests and analysis. The data collected and the database relationships are displayed in Figure 5.1.
212
Figure 5.51: Microsoft Access table, field structure and relationships for CT in non-HCM analysis
5.3.2 Patient selection
The assessment of CT scans in non-HCM patients was intended as a proof of concept study. I needed to be certain images were of sufficient quality to allow us to describe septal arterial anatomy before
213 commencing study in the HCM population. We analysed scans of patients without HCM who had undergone CT angiography for any indication.
5.3.2.1 Sample size calculations
To help analyse how many non-HCM scans would need to be assessed to predict scores achieved in the HCM population we performed basic sample size calculations.
A total image quality score for each scan will be a composite of the 9 sections. The total score possible will be 18 for target arteries. We know from preliminary work on CT scans in HOCM patients that the standard deviation is likely to be low, 1 or 2. I can therefore calculate a standard error of the mean for a proposed sample size of 30. I will use a standard deviation of 2 for the following equation:
Standard error of the mean (SEM) = sample standard deviation / √n
= 2 / √30 = 0.37
Assuming the population is normally distributed I can then calculate approximate confidence intervals:
Upper 95% limit:
= Sample mean + (0.37 x 1.96) = 0.72 (2DP)
Lower 95% limit:
= Sample mean – (0.37 x 1.96) = 0.72 (2DP)
214
I can therefore predict that with n=30 I will be able to use the mean sample population score to estimate the mean true population score to +/- 0.72 with 95% confidence.
A prospective series of 30 consecutive CT angiograms undertaken at Liverpool Heart and Chest
Hospital from 1/1/12 were assessed according to the protocol.
5.3.2.2 Inclusion criteria
To ensure we were analysing suitable CT scans that were representative of the HCM population we are likely to encounter in the next phase of CT study I set inclusion criteria.
Patients must be in sinus rhythm
Heart rate at time of CT must be < 70bpm
Native vessels only will be scored, grafts studies will be excluded
Calcium score <400, a higher score may indicate significant coronary artery disease; this
could affect ability to comment on the septal arterial anatomy.
We therefore analysed 30 consecutive CT scans from 1/1/12 that met the inclusion criteria. There were no exclusions based on image quality or artefact.
5.3.3 CT image acquisition methods
Acquisition was according to standard CT methods, these CT scans had already been acquired for clinical reasons. Coronary CT angiography (CTA) was performed using a dual- source CT system
(‘Definition Flash’, Siemens Healthcare, Forchheim, Germany) with 2 x 128 x 0.6mm collimation, 0.28
215 second (s) rotation time, a pitch of 3.2-3.4, 100 KV tube voltage and current of 320 mAs. Data acquisition was prospectively triggered at 60 % of the RR interval. This was performed either as high
– pitch spiral acquisition (‘Flash spiral’) or sequentially with a table feed of 38.4mm.
All the data was acquired in deep inspiration. Calcium scoring was performed prior to progression to coronary angiography. Coronary CTA was performed using intravenous contrast enhancement. To synchronise acquisition of the coronary CTA data set to arterial enhancement, a test bolus protocol was used; 15mls contrast agent (Optiray 350, Covedian, Mallinckrodt Medical Imaging Ireland) was followed by 40mls saline solution at 6mls /s. The time to peak enhancement in the aorta was measured using a series of trans-axial scans acquired in 2s increments, with first image acquired after 12s. For CTA 60mls of contrast was injected followed by a saline flush of 40mls, both at flow rates of 6mls/s. Image acquisition was started with a delay of measured transit time plus 2s. For image reconstruction a half scan reconstruction algorithm was used which provides a temporal resolution of 75ms. Reconstructed slice thickness was 0.6mm, slice increment was 0.3mm. The
Reconstructed images were analysed with SYNGO Via a multimodality workstation.
5.3.4 CT image analysis methods:
In this non-HCM population the target myocardium was defined as the basal third of the LV septum.
This was felt to be representative of the target myocardium in a HCM population. Scans were analysed off-line using the SYNGO Via workstation (VB10B). Two-dimensional images were used to study the optimum target area of myocardium, identify its vascular supply and characterise the course and origin of these vessels. We started by scouring the myocardium for vessels and tracking back to parent vessels (see Figure 5.2).
216
Figure 5.52: Example of septal vessel travelling to basal septum
If no vessel was visible by this method we followed parent vessels until a branch heading towards the septum was identified (see Figure 5.3). This was tracked and analysed. Vessels travelling to areas other than the basal septum were identified in the same manner.
217
Figure 5.53: Example of vessel tracked from LAD travelling to distal RV septum
A measurement in each score category was taken in the axial cuts (see Figures 5.4, 5.5, 5.6). All vessels tracking towards the basal and mid septal sections were analysed in this manner.
Figure 5.54: Septal length in fat before travelling in to myocardium
218
Figure 5.55: Septal length in myocardium prior to bifurcation (A) and after (B)
Figure 5.56: Angle of bifurcation from parent vessel (A) and cross-sectional area (B)
Septals travelling to target myocardium were tracked and a set of multiplanar reformatted images akin to a 3D CT angiogram was constructed (an example is shown in Figure 5.7). This allowed definition of the optimum angiographic projections to identify and wire the vessel(s), removing foreshortening and overlap. I analysed these images and rotated the coronary tree through 3 dimensions to first define optimal angles of the septal-parent artery origin. This is useful information when negotiating a coronary wire in from the parent vessel. I then rotated the coronary tree to define the best angles to view the distal vessel. These are often different from the views that display
219 the ostium. The vessel can travel several centimetres and important intra-septal bifurcations could be missed through overlap if the region of interest was not changed to the distal vessel.
Figure 5.57: Example of CT angiogram created in non-HCM population
Familiar projections from the catheterisation laboratory were pre-defined before data collection
(see Table 5.36).
220
Viewing angles for multiplanar reformat assessment: Pre- defined projections Antero-posterior (AP) Antero-posterior Cranial Antero-posterior Caudal Left Anterior Oblique (LAO) Left Anterior Oblique Cranial Left Anterior Oblique Caudal Right Anterior Oblique (RAO) Right Anterior Oblique Cranial Right Anterior Oblique Caudal
Table 5.33: Viewing angles for multiplanar reformat assessment: Pre-defined projections
All major epicardial arteries were then surveyed to identify any other vessels tracking towards the septum, noting the vessel course and myocardial supply. This identified additional potential target vessels but, in the main, served to exclude septal vessels with an inappropriate distribution, avoiding their instrumentation at the time of the ASA procedure. Any vessels not tracking to target myocardium were not routinely analysed with three dimensional coronary images as we would not aim to instrument these as part of ASA.
Areas of myocardial distribution pre-defined before collecting data are summarised in Table 5.37.
221
Myocardial distribution: Pre-defined territories
Left ventricle; basal septum Left ventricle; mid-septum Left ventricle; distal septum Right ventricle; basal septum Right ventricle; mid-septum Right ventricular cavity Right ventricular outflow tract Anterior wall Inferior wall Aortic valve annulus
Table 5.34: Pre-defined territories for coronary artery supply
We recognised from feasibility sessions viewing scans prior to anaylsis that the same septal vessel could supply more than one territory. I allowed up to 4 territories when collecting data.
222
5.4 Results
5.4.1 Patient and criteria details
The mean age of patients at the time of scanning was 53.71 (±15.13) years. Fifty per cent were male.
Twenty-nine scans were performed for investigation of chest pain, 1 scan was performed to assess coronary arteries prior to cardiac surgery for aortic valve endocarditis.
All patients were in sinus rhythm at the time of scanning. Mean heart rate was 58 (±6.19)bpm, range
44-69. Mean calcium score was 26.43 (±51.15). No patients with coronary artery grafts were included. The mean basal septal thickness in these patients was 9.47 (±1.68)mm, range 6.00-
13.60mm.
5.4.2 Assessment of image quality
A total of 86 septal arteries were identified. Each artery was scored in available sections of single plane reformatand multiplanar reformat analysis. In 2 small septal arteries there was no visible distal branching, we could not assess the ability to comment on the distance to the bifurcation in these. We only assessed the multiplanar reformat images in target arteries traveling to the basal septum (n=47). The other 39 septal arteries were assessed in single plane reformatimages and travelled to other territories.
The confidence with which we could comment on the pre-defined domains was high; the image score was therefore also high. Each septal vessel could be tracked easily. A numerical value and description of anatomy was achieved in nearly all cases. The following 7 domains achieved a score of
2 in 100% of cases (defined as ‘Septal artery appearance commented on with high level of
223 confidence of accuracy, clearly able to guide invasive procedure and viewing of septal anatomy’ in methods):
Parent artery
Bifurcation angle
Pre-intraseptal split length
Myocardial distribution
Landmarks
Preferred projection ostium
Preferred projection distal
A further 2 domains (artery diameter and length of septal vessel) achieved near perfect scores with
98% and 99% respectively achieving 2.
224
Category n available for n (%) score Mean score assessment = 2 (±SD)
single plane reformatimage assessment
Parent artery 86 86 (100%) 2 (±0)
Bifurcation angle 86 86 (100%) 2 (±0)
Artery diameter 86 84 (98%) 1.96 (±0.24)
Length of the septal vessel in epicardial fat 86 85 (99%) 1.98 (±0.15) and myocardium Pre-intraseptal split length 84 84 (100%) 2 (±0)
Myocardial distribution 86 86 (100%) 2 (±0)
Landmarks 86 86 (100%) 2 (±0)
multiplanar reformat CT coronary angiogram image assessment
Preferred projection ostium 47 47 (100%) 2 (±0)
Preferred projection distal 47 47 (100%) 2 (±0)
Table 5.35: Summary of image quality scores in each of the 9 pre-defined domains.
The mean score achieved over the 9 categories was therefore 17.94. Using the sample size calculations detailed in section 5.3.2.1 I can predict that if these methods were to be applied to a series of HCM hearts the 95% CI for the score would be 17.22 – 18.66 (maximal score being 18). This gives a high level of confidence that the images could be used to describe arterial anatomy.
225
5.4.3 Anatomy of septal arteries seen on CT
5.4.3.1 Number of septal arteries per patient:
Eighty-six septal arteries were assessed in total. This was an average of 2.87 per patient, the median value was 3. The pattern of number of septals per patient is displayed in Figure 5.8. Only one patient had just one vessel visible, and one patient had 6 vessels visible. Twenty-nine (34%) were found in images in the basal septal myocardium and tracked back to the parent artery. Fifty-seven (66%) were identified by assessing epicardial arteries for vessels travelling towards the septum.
Number of septal arteries identified 18 16 16 14
12 10 9 8 6 Numberof patients 4 2 2 1 1 1 0 1 2 3 4 5 6 Number of septals identified
Figure 5.58: Number of septals identified
226
5.4.3.2 Parent artery
The parent artery was predominantly the left anterior descending artery; this was observed in 76
(88%) arteries. There was a second vessel following the same course as the primary LAD that acted as a parent vessel in 6 (7%) further cases, this was termed a ‘LAD 2’. The right coronary artery was the parent vessel in 2 (2%) and the circumflex and first diagonal in 1 (1%) case each.
Parent artery to septal vessels assessed by CT in non-HCM 80 76 70 60 50 40 30 20 6 10 2 1 1 0 LAD LAD 2 RCA D1 Circ
Figure 5.59: Parent artery to septal vessels
5.4.3.3 Bifurcation angle
The mean observed bifurcation angle was 112.30° (±24.64). The minimum observed angle was 51° and maximum was 170°. Seventeen (20%) were acute angles of <90°. This has implications for entering the vessel with a coronary wire when performing ASA. Seventy-nine (80%) were >90°.
227
5.4.3.4 Septal artery diameter
The mean septal artery area observed was 2.02mm2 (±1.43). The mean septal artery diameter was
1.51mm (±0.52). The minimum area and diameter were 0.43mm2 and 0.76mm respectively. The maximum area and diameter were 6.95mm2 and 2.97mm.
5.4.3.5 Length of septal vessel in fat and myocardium
The mean length of septal vessel in epicardial fat prior to entering myocardium was 4.78mm (±3.40).
Thirteen (15%) were completely intra-myocardial vessels. The range of length prior to entering myocardium was 0-13.80mm.
The mean length of vessel after entering myocardium was 32.09mm (±18.08). The range of lengths visible after entering myocardium was 0-83.70mm. The mean measured total length of visible septal vessel was 36.90mm (±20.00), with a range of 5.80-92.20mm.
5.4.3.6 Length of septal vessel prior to bifurcation
The mean length of septal vessel prior to a branching point was 7.20mm (±4.96). There were 2 vessels that branched immediately but had the same ostium. The range of lengths prior to a bifurcation point was 0-22.20mm.
5.4.3.7 Landmarks
No numerical data was collected for this section. Common landmarks used to describe position of the chosen septal were:
Distance from LAD / circumflex bifurcation
228
Distance from first diagonal branch
Distance from previous septal vessel from the same parent artery
A landmark was identifiable in all cases.
5.4.3.8 Myocardial distribution
A septal artery supplying one territory was seen in 17 (20%) of vessels. Dual supply with branches of the vessel travelling to more than one territory was seen in 34 (39%). A vessel travelling to three territories was seen in 35 (41%). Figure 5.10 displays number of territories supplied by identified septal vessels.
Number of territories supplied by septal vessel
17, 20%
35, 41% 1 2 3
34, 39%
Figure 5.60: Number of territories supplied by septal vessel
229
In those with a single territory supplied the predominant distribution was the LV basal septum.
Other vessels that appeared to be tracking towards target myocardium when followed from parent arteries supplied areas such as the aortic valve annulus, RV basal septum, RV cavity, RV outflow tract, anterior and inferior walls. The breakdown is displayed in Figure 5.11.
Territories supplied in arteries with 1 distribution
1, 6% 1, 6%
1, 6% LVBasalSeptum 6, 35% LVMidSeptum 1, 6% AnteriorWall InferiorWall AVAnnulus
2, 12% RVBasalSeptum RVCavity RVOT
2, 12% 3, 17%
Figure 5.61: Territories supplied in arteries with one distribution
In those arteries with dual supply the vessel branched to travel to left and right ventricular myocardium in 29 of 34 (85%) cases. In the other 5 cases we saw branches traveling to left ventricular septum and anterior wall. The dual supply to right and left ventricular myocardium has implications for ASA; a further exploration of this is in the discussion section.
230
Territories supplied in arteries with 2 distributions
1, 3% 1, 3% 1, 3% LVBasalSeptum / RVBasalSeptum 1, 3% 1, 3% LVMidSeptum / RVMidSeptum 10, 29% 2, 6% LVMidSeptum / RVBasalSeptum LVBasalSeptum / AnteriorWall
LVMidSeptum / RVCavity 3, 9% AnteriorWall / RVOT
LVBasalSeptum / RVCavity
LVBasalSeptum / RVOT 4, 12% LVMidSeptum / DistalSeptum
10, 29% LVMidSeptum / RVOT
Figure 5.62: Territories supplied in arteries with 2 distributions
In those arteries with a vessel supplying three territories we saw a recurring pattern of LV septal myocardium, RV septal myocardium and anterior wall. This occurred in 33 of 35 (94%) of vessels; 25
(71%) at the basal septal level and 8 (23%) at the mid septal level. In the other two cases the vessel supplied LV septal myocardium, RV septal myocardium and the inferior wall.
231
Territories supplied in arteries with 3 distributions 1, 3% 1, 3%
8, 23% LV septum/RV septum/Anterior wall at basal level LV septum/RV septum/Anterior wall at mid chamber level LV septum/RV septum/Inferior wall at basal level LV septum/RV septum/Inferior wall at mid chamber level
25, 71%
Figure 5.63: Territories supplied in arteries with 3 distributions
A display of the distribution of all identified septal arteries in seen in Figure 5.14.
232
Figure 5.64: Distribution of all septal vessels identified
Myocardial distribution assessment limited to vessels supplying LV basal septum
In this section I limited the assessment of myocardial distribution to vessels that had a branch to LV basal septum; this is acting as a surrogate for the target myocardium in ASA. This could help predict
233 what other territories will be involved if we were to explore these vessels with myocardial contrast as part of ASA.
Forty-seven vessels were assessed. Six of 47 (13%) were single vessels direct to target myocardium with no visible bifurcations. Fifteen (32%) of vessels supplying the basal septum had a branch travelling to another territory. Twenty-six (55%) of vessels supplying the basal septum had branches supplying two other territories.
Number of territories supplied by vessels with target myocardial involvement
6, 13%
1 2
26, 55% 15, 32% 3
Figure 5.65: LV basal septal vessels - number of territories supplied
Of the 15 vessels with LV basal septum supply and one other territory supplied the most common other territory was the RV basal septum (67%). In a smaller number of vessels the other branch travelled to the anterior wall (20%), RV cavity (7%) and RV outflow tract (7%).
234
LV basal septal arteries - other territory in dual supply arteries
1, 7% 1, 6% LVBasalSeptum / RVBasalSeptum LVBasalSeptum / 3, 20% AnteriorWall LVBasalSeptum / RVCavity
10, 67% LVBasalSeptum / RVOT
Figure 5.66: LV basal septal arteries - other territory in dual supply arteries
Twenty-six vessels supplying the LV basal septum had branches to three territories. In these the most common pattern was to see one branch to the RV basal septum and another to the anterior wall (77%). Other combinations included the LV basal septum, RV outflow tract and the anterior wall
(15%) and LV basal septum, inferior wall and one of the RV outflow tract or RV basal septum (4% respectively).
LV basal septal arteries - other territories in arteries with 3 distributions
1, 4% 1, 4% LVBasalSeptum / RVBasalSeptum / AnteriorWall 4, 15% LVBasalSeptum / RVOT / AnteriorWall
LVBasalSeptum / RVBasalSeptum / InferiorWall
LVBasalSeptum / RVOT / InferiorWall 20, 77%
235
Figure 5.67: LV basal septal arteries - other territories in arteries with 3 distributions
A full display of the distribution of all arteries supplying the LV basal septum (and therefore surrogate marker for the target myocardium in ASA) is seen in Figure 5.18.
Figure 5.68: Distribution of all arteries involving the LV basal septum
236
5.4.3.9 Preferred projection for ostium of septal vessels
The most common chosen view for viewing the ostium of septal vessels was RAO Cranial; this was the optimal angle in 23/47 septals (49%). The next most common view was LAO cranial with 17/47
(36%). Other views were deemed to be the optimal angle in smaller number of cases; RAO caudal in
4/47 (9%), AP Cranial in 2/47 (4%) and LAO caudal in 1/47 (2%).
Chosen view for ostium of vessels supplying LV basal septum 25 23
20 17
15
10
5 4 2 1 0 RAO Cranial LAO Cranial RAO Caudal AP Cranial LAO Caudal
Figure 5.69: Chosen view for ostium of vessels supplying LV basal septum
5.4.3.10 Preferred projection for distal septal vessel
The most common preferred projection for viewing the distal course of the vessel was LAO caudal, used in 21/47 (47%). This was followed by the left cranial view in 13/47 (28%) of vessels. The caudal
237
RAO view was used more than the cranial RAO, 6/47 (13%) vs 5/47 (11%). Straight LAO was used
3/47 (6%). AP was used once (2%).
Chosen view for distal septal vessels supplying LV basal septum 25 21 20
15 11 10 6 5 5 3 1 0 LAO Caudal LAO Cranial RAO Caudal RAO Cranial LAO AP Caudal
Figure 5.70: Chosen view for distal septal vessels supplying LV basal septum
238
5.5 Discussion
5.5.1 Image quality
The image quality of this series of CT scans acquired for non-HCM related reasons was sufficient to allow analysis of the septal arterial anatomy. It became clear when viewing the first 5 scans during the early feasibility stage that contrast can be seen in septal arteries with standard protocol CT scanning. The cardiac radiologist was able to comment on each facet of the septal anatomy with a high level of confidence in almost every case. This extremely high level of confidence then allowed me to detail the septal arterial anatomy in order to try to understand what patterns we might see in
HCM. This may then allow me to address some of the problems associated with unsatisfactory outcome in ASA.
5.5.2 Alternative epicardial artery source
Up to 15% of patients entering the catheter lab for ASA do not receive an alcohol injection. In our retrospective series that was true in 5% cases. Often the reason cited for not delivering alcohol is the inability to identify a suitable vessel after exploration of visible septal arteries with myocardial contrast echocardiography. Standard techniques involve following an epicardial artery from the left main stem, to the LAD and then to a vessel that the operator proposes may supply the target septal myocardium. In the CT technique described this process is reversed. The septal artery is identified in the target area and tracked back to the parent vessel. Using this approach the source of the septal artery will not be missed, regardless of parent artery.
This may help to reduce the amount of patients that do not receive treatment.
239
5.5.3 Septal vessels supplying multiple territories
In this exploratory series of non-HCM patients a septal vessel supplying LV basal septum also had a branch to another distinct territory in 84% of arteries. This has ramifications when considering the traditional techniques employed in ASA. Coronary balloons are sited at the ostium of a chosen septal vessel. Contrast and subsequently alcohol are then injected and can travel to any territory the artery travels to. If the chosen artery supplies a larger area than required then alcohol will be wasted infracting myocardium that will not affect LVOT haemodynamics. This unwanted infarction could also potentially cause complication.
When an artery supplies multiple territories within the left ventricle the end environments are similar. In HCM this is hypertrophied, densely packed myocardium with high resistance to flow and microvascular abnormalities140. Seventy-six per cent of septal arteries supplying the LV basal septum also had branches travelling to the right ventricle. This is a very different environment. The RV myocardium is less hypertrophied in HCM; the incidence of RVH in HCM is reported to be 28%141 in
CMR based studies. The RV blood supply drains directly into the RV cavity via Thebesian veins more often than the left142. Less hypertrophy and direct drainage into a low pressure cavity both lead to lower pressure end environment for the branch of the septal vessel travelling to the RV. A significant differential in the microvascular resistance of end environments will lead to preferential flow of any injected fluid in to the RV branch. This can have two consequences in ASA:
1. Myocardial contrast is injected – this flows to the RV cavity and the artery is dismissed as
inappropriate for alcohol injection. Engaging the sub branch that travels to the LV septum
may allow the operator to deliver contrast and then alcohol to the perfect target
myocardium.
240
2. A weak contrast signal is seen in the LV myocardium on echocardiography (compounded by
poor on table echo images). This is accepted as no alternative vessel is identified. Alcohol
predominantly localises to the RV branches and the subsequent infarct is on the RV side of
the septum (as seen in Chapter 3) – this will probably have limited effect on LVOT
haemodynamics and ASA fails.
Knowledge of this branching pattern prior to ASA could allow the operator to engage a sub branch of the septal vessel and deliver the iatrogenic infarction with greater accuracy.
5.5.4 The ability to rule out alternative septal vessels
Pre ASA CT assessment would detail the course and supplied territories of all vessels leaving the LAD.
The operator at ASA would then be able to dismiss inappropriate vessels travelling from the LAD without the need to engage the coronary artery and investigate with myocardial contrast. This could reduce complications associated with instrumentation and also reduce procedure time.
241
5.6 Conclusion
Standard CT coronary angiography images are of sufficient quality to be able to comment of
septal arterial anatomy
Septal vessels travelling to the LV basal septum supplied additional alternative territories in
87% of non-HCM hearts
242
5.7 Limitations
The assessment of septal arteries in the initial phases was very time consuming. The scans for a single patient could take 3-4 hours to analyse. The time and financial constraints meant that progressing to intra- and inter-observer variability was not possible.
The accuracy of the septal arterial anatomy has not been verified against a better validated model.
Options for this would include histological angiographic studies or standard invasive coronary angiography with myocardial contrast echocardiography. Neither are possible with these patients as this was an observational study. I will apply these methods to a population of HOCM patients undergoing ASA in the next chapter – they will have invasive angiography and myocardial contrast echocardiography.
This data has only been produced at one centre, and analysed with one radiologist and one cardiologist. This needs to be reproduced at another centre with different scanners and software packages.
243
Chapter 6: Clinical outcome of hypertrophic obstructive cardiomyopathy patients undergoing CT guided ASA
244
6.1 Introduction
The case series, presented in Chapter 2 suggests that an unacceptable proportion of patients undergoing ASA had significant, persisting LVOT gradients with standard techniques. Chapter 3 suggested that this is in part due to inaccurate location of the iatrogenic infarction. There is a clear need to improve treatment methods to achieve a more predictable and uniform outcome following
ASA.
The data in Chapter 5 suggests that standard CT scanning technologies can be used to describe septal arterial anatomy with a high level of confidence. It also suggests that the origins and branching pattern of septal arteries are more complex than we had previously acknowledged.
Knowing the details of origins and branching patterns may guide ASA and allow the operator to deliver accurate iatrogenic infarction in a greater proportion of patients. The greater accuracy could result in successful ASA in a higher proportion of patients.
In Chapter 6 I will describe the use of these CT techniques in a population of HOCM patients undergoing ASA and detail clinical outcomes with CT guided procedures.
245
6.2 Aims
To use CT methods described in Chapter 5 to describe septal anatomy in a series of HOCM
patients.
To use information derived from CT to guide ASA
To report clinical outcomes for patients undergoing CT guided ASA and compare to historic
controls
246
6.3 Methods
6.3.1 Patient selection
Patients were recruited from the Cardiology clinics of Liverpool Heart and Chest Hospital between
January 2012 and January 2014. Forty-nine patients were referred to a specific quaternary service for consideration of ASA. A diagnosis of HCM was made according to typical clinical, electrocardiographic and echocardiographic features. Each patient was assessed according to new local protocols (detailed in Chapter 7). Mean age was 57.41(±14.84) years, 62% male. One patient had significant lung disease (pulmonary fibrosis), 2 had previous PCI. Three patients had attempted
ASA previously with unsatisfactory outcome (2 received alcohol).
Twenty patients (40%) were treated successfully with tailored medical therapy. Three (6%) required surgical management for concomitant cardiac disease (bypass grafting x1, Subaortic ring x1, patient choice myectomy x1). Twenty six patients entered the catheter lab with the intention of delivering alcohol. All patients taken to the laboratory had resting, Valsalva manoeuvre or exercise stress peak
LVOT gradient ≥50mmHg and basal interventricular septal diameter ≥15mm.
247
Figure 6.71: Patient flow diagram
6.3.1.1 Echocardiographic assessment
Resting echocardiogram was performed >24 hours prior to ASA to evaluate LVOT gradients and SAM of the MV (Phillips ie33 scanner, Phillips S5-1 probe). SAM severity grading was adapted to provide a binary option of ‘contact with septum’ or ’no contact with septum’ to allow statistical analysis143.
This is based on the observation that the majority of the increased flow velocity in the LVOT is related to SAM-septal contact16.
6.3.1.2 Cardiopulmonary exercise (CPEX) testing
CPEX was performed using bicycle ergometer with 10W increments in workload every minute.
Patients were exercised until exhaustion with RER >1.1, a mean of the last 7 readings over the final
30 seconds of exercise was used. If patients did not have an LVOT gradient >50mmHg at rest or with
Valsalva manoeuvre upright echocardiography was performed at peak exercise to assess gradients.
6.3.1.3 Cardiac magnetic resonance imaging
Cardiac magnetic resonance (CMR) was performed prior to and 6 months after ASA. Examinations were performed using a 1.5-Tesla scanner (Magneton AERA; Siemens, Medical Imaging, Erlangen,
Germany). Left ventricular volumes, ejection fraction and LV mass were determined using SYNGO
Via (Siemens Healthcare) version VA30A_HF05 multimodality workstation. For late gadolinium enhancement (LGE) imaging, 0.1 mmol/kg gadolinium-diethylenetriamine pentaacetic acid
248
(Gadovist, Bayer Schering, Berlin, Germany) was administered intravenously and standard breath- hold inversion recovery imaging was performed.
6.3.1.3 Research permissions
Appropriate permissions to perform CT prior to clinically indicated ASA were provided by the research and development board of Liverpool Heart and Chest Hospital (see Appendix 6.1).
6.3.1 CT methods
6.3.1.1 CT image acquisition
This was performed in a very similar manner to that described in Chapter 5. Coronary CT angiography was performed using a dual- source CT system (‘Definition Flash’, Siemens Healthcare,
Forchheim, Germany). All patients with a heart rate greater than 60 beats per minute were given either oral or intravenous metoprolol. Axial data acquisition was prospectively triggered at 70 % of the RR interval. A lower 25% dose was given during 30 to 80% of the cardiac cycle to get systolic frames and SAM-septal contact area.
To synchronise acquisition of data set to arterial enhancement, a test bolus protocol was used;
15mls contrast agent (Optiray 350, Covedian, Mallinckrodt Medical Imaging Ireland) was followed by
40mls saline solution at 6mls /s. The time to peak enhancement in the aorta was measured using a series of trans-axial scans acquired in 2s increments, with first image acquired after 12s. For CTA
60mls of contrast was injected followed by a mixed flush (10mls contrast and 30mls saline), both at flow rates of 6mls/s. For image reconstruction a half scan reconstruction algorithm was used, this
249 provides a temporal resolution of 75ms. Reconstructed slice thickness was 0.6mm, slice increment was 0.3mm.
6.3.1.2 CT image analysis
The reconstructed images were analysed with SYNGO Via (Siemens Healthcare) version VA30A_HF05 multimodality workstation. This was a software upgrade from the package used for the CT in non-
HCM population in Chapter 5. The target area of myocardium was identified using a short section of systolic imaging (Figure 6.2A). This target myocardium at the SAM-septal contact area was examined in diastole to identify a segment of its arterial branch supply (Figure 6.2B, C and D). This vessel was tracked back to its parent artery.
250
Figure 6.72: CT image of target myocardium and identified septal vessel
Panel A: 3-chamber systolic CT image displaying SAM of the MV, the contact area is seen in the basal septum. Panel B; the target area of myocardium in the basal septum is located at the centre of the pink and purple lines – a short axis view is displayed in the pink box (image C).The centre point of these lines is the same area of myocardium in each image. This is the target myocardium. This myocardium is then surveyed for evidence of a coronary artery; the vessel is traced back to its parent epicardial vessel and examined as per protocol.
Other characteristics of the septal vessel including the angle of bifurcation from its parent artery, the length of its course in epicardial fat (before septal penetration) and myocardium, its branch pattern, and ultimate myocardial territories of all branches were noted (see Figure 6.3).
251
Figure 6.73: Septal artery anatomical information taken from single plane reformatimages
Panel A: Short axis view at the level of target myocardium. A septal artery was seen and highlighted in panel B. Panel C: Septal vessel with a measurement of the distance travelled from parent vessel origin to entry into myocardium. Panel D: Total identifiable length of septal vessel from ostium. Panel E: Angle of entry from LAD in to septal. Panel F: Area of septal vessel at ostium.
The target artery was traced and marked in the CT software package and a coronary angiogram
‘map’ created. The coronary map was then rotated through horizontal and vertical planes to allow
252 optimal visualisation and remove any overlap or foreshortening. The optimum angiographic projections were noted, these were then used as the ‘working views’ in the catheterisation laboratory (Figure 6.4). The major epicardial arteries were then surveyed for other vessels tracking towards the septum. All potential target vessels were followed and ultimate distribution noted, this was to allow identification or rejection as an additional target vessel.
253
Figure 6.74: Examples of matched CT and invasive coronary angiography
Panel A: CT angiogram. The traced septal vessels from single plane reformat images were projected on to the coronary angiogram ’map’. This CT angiogram was rotated to open the angles, minimise foreshortening and remove overlap (in this example to RAO cranial). The equivalent invasive angiogram projection is shown in panel B. The target artery is identified and only this sub-branch is occluded for alcohol delivery. Further examples are shown in panels C/D and E/F.
254
6.3.2 Alcohol septal ablation procedure
The working projections identified from CT to visualise the origin of the target septal from its parent vessel were used. A coronary wire was passed into the target septal. A traditional approach to contrast echo studies was employed first. An over the wire balloon was inflated in the common ostium of the target septal with subsequent contrast injection and echocardiographic studies. If contrast was seen to travel to the RV septum or RV cavity (as predicted by CT) alcohol was not injected. This was often in addition to some opacification of the LV septum. The wire was then passed in to the target sub branch of the septal vessel. A second view was identified from CT to visualise the distal branch pattern. Once the wire was in the proximal portion of the vessel the angiographic projections were changed to allow navigation into the chosen sub branch as advised by
CT. When the wire was secure in the desired sub branch the over the wire balloon was inserted and inflated to safely occlude the artery and allow delivery of contrast and alcohol. If the target myocardium wasn’t completely covered during myocardial contrast echo studies additional target vessels were explored (n=1). The delivery of alcohol was guided by CT and confirmed by myocardial contrast studies in all patients.
For 17/21 patients CT angiography highlighted the target for alcohol to be a specific sub-branch of a septal vessel (see Table 6.39). The parent septal vessel also had other divisions travelling to the right ventricular septum and sometimes to the anterior wall. A sub-selective alcohol injection into the specific target branch was performed in these patients. In 2 procedures the parent epicardial vessel for the target septal branch was the circumflex artery. For a further 2 procedures a small septal artery was identified from the proximal LAD. This did not differ from the technique of a standard
ASA (i.e. CT suggested that a proximal/ostial balloon position in a small septal artery was required).
Other aspects of the procedure were as described in Chapter 1.
255
Procedural change n (%) Multiple territory distribution with sub-selective injection 17 (80%) Alternative epicardial parent vessel (Cx) 2 (10%) Standard approach 2 (10%)
Table 6.36: Changes to standard ASA procedural details using information provided by CT
6.3.4 Statistical analysis
Statistical analysis was performed using StatsDirect version 2.8.0. Continuous data are presented as means and SD, or median and IQR for data that are non-parametric. Paired T tests were used for parametric data and Wilcoxon signed ranks for non-parametric data. Normality was assessed using
Shapiro-Wilk test, a value of <0.05 was used to decide data was non-parametric. Change in the categorical extent of SAM16 was assessed using sign test. Correlation of alcohol dose to CKMB release was assessed using Pearson’s test.
6.3.4.1 Comparison to traditional ASA
Our experience (prior to the introduction of CT) mirrors international norms55-57 80 . We compared the results of the CT-guided cohort against our historic controls. Fisher’s exact test was used to compare success in treating LVOT gradients after one procedure, failure was defined as per Chapter
2; a persisting gradient of >50mmHg or failure to reduce the gradient by greater than 50%. Fisher’s exact test was also used to compare incidence of procedural related complication. We compared the relationship between alcohol dose and myocardial damage as a reflection of the control of the
256 procedure using Pearson’s correlation between alcohol dose injected and CKMB release in each group.
257
6.4 Results
6.4.1 CT image quality and anatomy
6.4.1.1 CT image quality
Image quality was high in the analysis of CT in the HOCM population. A total of 49 septals were identified and tracked. Twenty-four of these arteries were identified as target arteries and went on to invasive assessment as part of ASA – multiplanar reformat image assessment was performed in only these arteries.
The following domains achieved a score of 2 in 100% of cases (see Table 6.40):
Parent artery
Bifurcation angle
Myocardial distribution
Landmarks
Preferred projection ostium
Preferred projection distal
The other categories also scored highly, with few images unable to provide the information required to guide an operator for ASA. In the 24 arteries with a full assessment of all 9 sections the mean score was 17.71 ±0.63 (maximum score 18).
258
Category n available for n (%) score Mean score assessment = 2 (±SD)
single plane reformatimage assessment
Parent artery 49 49 (100%) 2 (±0)
Bifurcation angle 49 49 (100%) 2 (±0)
Artery diameter 49 46 (94%) 1.88 (±0.48)
Length of the septal vessel in epicardial 49 45 (92%) 1.92 (±0.28) fat and myocardium
Pre-intraseptal split length 39 38 (97%) 1.97 (±0.16)
Myocardial distribution 49 49 (100%) 2 (±0)
Landmarks 49 49 (100%) 2 (±0)
multiplanar reformat CT coronary angiogram image assessment
Preferred projection ostium 24 24 (100%) 2 (±0)
Preferred projection distal 24 24 (100%) 2 (±0)
Table 6.37 Summary of image quality scores in each of the 9 pre-defined domains.
259
6.4.1.2 Anatomy of septal arteries seen on CT in HOCM patients
Number of septal arteries per patient
The most common observation was to identify 2 septal vessels, with a mean observation of 2.33
±1.55 septals per patient (see Figure 6.5).
Number of septal arteries per patient 12
10
8
6 Incidence 4
2
0 1 2 3 4 5 6 Number of septals identified
Figure 6.75: Number of septal arteries identified per patient
Parent artery
The most common parent artery was again the LAD (see Figure 6.6), this accounted for 90% of described septal vessels. The circumflex artery was the parent artery in 4 (8%) and the first diagonal artery in 1 (2%).
260
Parent artery to identifed septals 50 44 45 40 35 30 25 20 15 10 4 5 1 0 0 LAD Cx D1 RCA
Figure 6.76: Parent artery to identified septals
The described method of identifying the septal artery in the myocardium and tracking it back to its parent artery was performed in 65% of vessels. The other 35% were identified by following parent arteries and looking for a vessel travelling towards the septum (Figure 6.7).
261
Method of identification of septal artery
17, 35%
Tracked back Traced from parent artery
32, 65%
Figure 6.77: Method of identification of septal artery
Bifurcation angle
The mean angle of bifurcation from the parent artery was 115.80 ±21.76°. Five of 49 (10%) arteries had an acute angle of entry (i.e. <90°).
Septal artery diameter
The mean septal artery diameter measured by CT was 1.63 ±0.54mm (range 0.54 – 2.69mm).
Length of septal vessel in fat and myocardium
The mean length of vessel in epicardial fat was 6.30 ±4.00mm (range 0-16.80mm). The measured length in myocardium was 26.85 ±18.48mm (range 0-67.30mm). The total length of septal artery was 33.15 ±18.69 (Range 3.80-74.10mm).
262
Length of septal vessel prior to bifurcation
Forty vessels had an identifiable intra-septal split. The distance to the bifurcation was 11.06 ±6.31
(range 0-26.80mm).
Landmarks
No numerical data was recorded for this section. Common landmarks used to identify correct septal arteries were proximity to diagonal arteries, distance from bifurcation of parent vessel (i.e. LAD- circumflex bifurcation), and proximity to other septal vessels.
Myocardial distribution
Eighteen (37%) of identified septals supplied a single territory (LV basal septum in 4). Twenty-four
(49%) had a dual territory supply and 7 (14%) had supply to three territories (See Figure 6.8).
263
Number of territories supplied by septal vessels - all
7, 14%
18, 37% 1 Territory 2 Territories 3 Territories
24, 49%
Figure 6.78: Number of territories supplied by all septal vessels
Of those vessels with just one territory supplied 4/18 travelled to the basal septum. Of those with 2 territories supplied 19/24 (79%) had dual RV and LV supply. In the 7 arteries with 3 territories supplied all had a branch supplying RV and LV myocardium (see Figure 6.9). Therefore of all septal arteries identified 26/49 (53%) had dual RV and LV septal myocardial supply.
264
Figure 6.79: Distribution of all septal vessels
Myocardial distribution limited to vessels supplying target myocardium.
Twenty-four vessels were identified to travel to target myocardium. Four (14%) of these had a single branch pattern travelling to target myocardium only, the other 20/24 (86%) had branches travelling to other territories (see Figure 6.10).
265
Number of territories supplied by vessels with target myocardium branch
4, 17% 7, 29%
1 Territory 2 Territories 3 Territories
13, 54%
Figure 6.80: Number of territories supplied by vessels with target myocardium branch
Of the 20 septals with branches to other territories 18 (18/24 – 75%) were identified to have RV septal supply – offering differing end environments and run off pressures to consider when injecting contrast or alcohol (see Figure 6.11).
266
Figure 6.81: Distribution of septal vessels including branch to target myocardium
Preferred projection for ostium of septal vessels
The preferred projection to visualise detail of the ostial portion of the septal vessel was the RAO cranial projection (see Figure 6.12). The cranial views tended to remove overlap with diagonal and circumflex arteries to allow the operator to see branching points, this helped to manipulate the wire in to the chosen vessel.
267
Preferred projection for ostium 12
10
8
6
4
2
0 RAO LAO RAO AP LAO AP Cranial RAO Cranial Cranial Caudal Caudal
Figure 6.82: Preferred projection for ostium of septal vessel
Preferred projection for distal septal vessel
RAO cranial was also the most commonly chosen view for seeing the detail of the distal vessel (see
Figure 6.13).
Preferred projection for distal vessel 9 8 7 6 5 4 3 2 1 0 RAO Cranial LAO Caudal LAO Cranial RAO Caudal AP Caudal AP
268
Figure 6.83: Preferred projection for distal vessel
There was a slight shift towards using more caudal views in viewing the distal vessel; 5/24 (21%) for ostium and 10/24 (42%) for distal vessel (see Figure 6.14).
Figure 6.84: Cranio-caudal tilt for ostium and distal vessel
6.4.2 Procedural details
All procedures were performed by a single operator. All procedures were guided by CT with a prepared report and images reviewed pre-procedure (an example is shown in Appendix 6.2).
In order to maximise research opportunities I used each patient entering the lab as a subject for
Chapter 4 (ICE vs TTE) and CT guidance for alcohol ablation. We therefore would get the echo probes in to position and site the OTW balloon at the ostium of the septal vessel. Myocardial contrast was
269 injected for the ICE images, a separate 0.5mL injection was then used for the TTE images. We then re-sited the OTW balloon in the chosen subselected septal vessel and repeat the process, with separate injections for ICE and TTE.
Twenty-one patients received alcohol to the target septal artery. One patient underwent a second
ASA procedure in the study period. Alcohol was injected into 2 septal arteries at the same procedure in one further patient. In 2 patients the target septal artery was at an acute angle and wire and balloon access were not possible. In a further two patients the myocardial contrast vented directly in to the LV cavity with no dwell time in the myocardium. One patient was found to have significant coronary artery disease following intravascular ultrasound assessment and required bypass grafting).
Twenty-one patients received alcohol to the target artery (see Figure 6.15).
270
Figure 6.85: Patient consort diagram - referral to alcohol delivery
Mean volume of alcohol delivered was 1.95mL (±0.62). Mean CKMB release was 116.65 (±63.30) ng/dL (reference range <5ng/dL). A correlation between alcohol dose and CKMB release was seen, R2 value was 0.46 (p=0.03) (see Figure 6.16).
Figure 6.86: Alcohol dose to CKMB release correlation
6.4.2.1 Procedural complications
Complete heart block requiring permanent pacemaker implantation was seen in 2 of 21 patients.
Two patients developed minor pericardial effusion without tamponade, no treatment was required.
This was presumed to be due to perforation of the right ventricular wall by temporary pacing wires.
271
New RBBB was seen in 2/16 (3 paced prior to ASA to treat LVOT gradients, 2 with CHB and PPM as a result of ASA), new LBBB was seen in 1/16. There was therefore some conduction complication in
5/21 patients.
6.4.3 Clinical outcomes
6.4.3.1 Survival and risk of ventricular arrhythmia:
Follow up data is presented for a mean period of 375 (±137) days, all patients had assessment >180 days from ASA. One patient was lost to follow up (but has been confirmed to be alive). No patient that received alcohol to the target artery died or suffered ventricular arrhythmia in the study period.
6.4.3.2 Symptomatic resolution
Symptoms of dyspnoea improved in 18 (90%), mean NYHA improved from 2.85 (±0.11) to 1.45
(±0.39) (p<0.0001) (Table 6.41). Ten patients improved from class 3 to class 1 (see Figure 6.17). Five patients improved from class 3 to class 2. Three patients improved from class 2 to class 1 and experienced resolution of recurrent pre-syncope. Two patients found no benefit and remained in class 3 (one developed pulmonary fibrosis and had persistent class 3 dyspnoea despite successful treatment of gradient). Chest pain was reported in 11/20 prior to ASA, all resolved with ASA
(p=0.005).
272
Figure 6.87: NYHA following CT guided ASA
Parameter Pre-ASA SD/IQR Post-ASA SD/IQR p
NYHA class (mean) 2.85 0.11 1.45 0.39 <0.0001 LVOT gradient (mmHg) (median) 98 89.50-111.50 14 8.50-22.0 0.003 Provoked LVOT gradient in those with 82 73.25-108.75 22.50 16.25-31.5 0.007 resting gradient <50mmHg (median) Presence of severe SAM 18/20 - 2/20 - 0.0008 Septal thickness in diastole (mm) 21 20.0-23.25 17 15.0-18.50 <0.0001 (median) Left atrial diameter (mm) (mean) 48 7.07 42.10 7.14 0.0002
Peak VO2 (mL/min/Kg) (mean) 19.09 6.34 21.45 6.59 <0.0001 % Predicted VO2 (mean) 79.19 14.01 91.62 12.02 <0.0001 Exercise time (secs) (mean) 715.59 252.35 837.35 265.10 <0.0001 EQ5D-5L index value (0-1) (mean) 0.51 0.24 0.78 0.16 <0.0001 EQ5D-5L health score (0-100) (mean) 49.00 16.32 71.67 16.0 0.0012
Table 6.38: Clinical outcomes pre- and post-CT guided ASA
273
6.4.3.3 Echocardiographic data
Twelve of 20 patients had a resting gradient ≥50mmHg, a further 8 had a resting gradient of
<50mmHg and a Valsalva or exercise stress induced gradient ≥50mmHg. Those with a resting gradient ≥50mmHg improved from 98 (IQR 89.50-111.50) mmHg to 14 (IQR 8.50-22) (p=0.003). Two of 12 had a persisting gradient ≥50mmHg at the end of the study period. In those with a provoked gradient we saw an improvement from 82 (IQR73.25-108.75) to 22 (IQR16.25-31.50) (p=0.007) mmHg. None had a provoked gradient ≥50mmHg after ASA.
SAM improved in 18/20 patients (p=0.0008). Eighteen of 20 patients had SAM-septal contact at rest prior to ASA, 2/20 had contact after treatment. Two patients had SAM-septal contact associated with high LVOT gradients on exercise, neither had contact on exercise after ASA. Those with persisting SAM-septal contact had significant LVOT gradients.
Interventricular septal thickness in diastole decreased from 21 (IQR 20-23.25)mm to 17 (IQR 15-
18.50)mm (p<0.0001). Left atrial diameter decreased from 48 (±7.07)mm to 42.10 (±7.14)mm
(p=0.0002).
6.4.3.4 Cardiopulmonary exercise testing
Satisfactory CPEX data was available in 17 patients. Three patients failed to reach an RER ≥1.1 on either pre- or post-ASA testing. Peak VO2 increased from 19.09 (±6.34) to 21.45 (±6.59) mL/min/Kg
(p<0.0001). This translated into an increase from 79.19% of predicted value (±14.01) to 91.62%
274 predicted (±12.02) (p<0.0001). Exercise time improved from 715.59 (±252.35) to 837.35 (±265.10) seconds (p<0.0001).
6.4.3.5 Quality of life
EQ5D-5L questionnaires were completed before and >6 months after ASA in 15 patients. The index value improved in all patients, with an increase from 0.51 (±0.24) to 0.78 (±0.16) (p<0.0001). The overall health score increased in 14/15 patients, values improving from 49.00 (±16.32) to 71.67
(±16.00) (p=0.001).
6.4.4 Comparison to traditional ASA
Successful LVOT gradient resolution after a single ablation procedure according to pre-specified criteria was observed in 17/20 (85%) cases in CT-guided ASA and 44/74 (59%) (see Figure 2.18 for breakdown) in the traditional methods groups (p=0.02).
There is a learning curve associated with ASA, as with most procedures. To ensure the improvement in outcomes post ASA was related to the introduction of CT methods and not the well-recognised learning curve associated with ASA I analysed the last 20 patients treated with traditional methods.
This group represented procedure numbers 105-124 performed at our centre at a rate of 10/year.
This is in keeping with international guidance for adequate training. These state an operator must have a minimum experience of 20 procedures and work in a centre with a total procedural number of 5013, and that a centre should perform at least 10 procedures/year12. This group had the same success at first procedure rate; 12/20 (60%) compared to 44/75 (59%) (p=0.92). When comparing the CT-guided group to the most recent traditional method group we see a trend towards improvement with CT; 17/20 (85%) vs 12/20 (60%) (p=0.09).
275
Symptom improvement defined by change in NYHA class improved in 18/20 (90%) patients with CT guided ASA and 53/74 (71%) patients with compete follow up treated with traditional methods
(p=0.09). Direct comparisons are difficult as patients in the traditional group received a greater number of doses of alcohol; 1.23 (±0.48) vs 1.05 (±0.22) in the CT guided group (p=0.17).
New RBBB (or complete heart block in those with pre-existing LBBB) was observed in 2/16 (13%) patients treated with CT guided ASA and 42/68 (62%) patients treated with traditional methods
(p=0.0004) (see Table 6.4). Complete heart block requiring pacemaker implantation was seen in 2/21
(10%) in the CT guided group and 14/74 (17%) of the traditional ASA group (p=0.17).
Traditional ASA CT guided ASA p LVOT gradient success after one procedure 44/75 (59%) 17/20 (85%) 0.02 NYHA class improvement ≥1 53/74 (73%) 18/20 (90%) 0.09 ASA induced RBBB 42/68 (62%) 2/16 (13%) 0.0004 PPM requirement 14/74 (17%) 2/21 (10%) 0.17 Alcohol dose 2.24 (±1.08) 1.95 (±0.62) 0.20 Alcohol to CKMB correlation R value 0.14 (p=0.15) 0.46 (p=0.03) -
Table 6.39: CT guided ASA versus traditional ASA outcomes
Total alcohol dose delivered was not significantly different between groups (2.24±1.08 in traditional methods vs 1.95±0.62 in CT guided ASA (p=0.20)). Alcohol dose to CKMB release had a significant
276 correlation in the CT-guided group (R value 0.46, p=0.03), whereas no significant correlation was seen in the traditional group (R value 0.15, p=0.14) (see Figure 6.18).
Figure 6.88: Alcohol dose to CKMB correlation in traditional and CT guided ASA
277
6.5 Discussion
CT angiography can provide high quality images of septal coronary arteries in HCM. The images provide anatomic insights that have adapted our approach to alcohol delivery in ASA. This may be associated with better outcomes and less complications.
6.5.1 Anatomical insights from CT alters the approach to ASA
6.5.1.1 Dual vascular supply to RV and LV
A septal artery was seen on CT to supply both the RV and LV septum in 75% of septals with a branch to the target myocardium. These are very different environments. This has implications for variable
‘run-off’, as discussed in Chapters 3 and 5. Traditional teaching for alcohol ablation is to site a coronary balloon at the ostium of the septal artery, and inject beyond28. This results in echocardiographic contrast highlighting the RV; a recurring pattern when injecting contrast in to the common ostium of a septal vessel with this branching pattern (see Figures 6.19 and 6.20). The vessel is then dismissed as serving an incorrect area of myocardium. Engaging the sub branch of the septal vessel that supplies the LV myocardium can result in target myocardium being highlighted.
278
Figure 6.89: Effect of sub-selective contrast injection
279
Figure 6.90: CT angiogram, invasive angiogram and MCE
Panel A shows the pre-ASA CT angiogram, Panel b shows the corresponding RAO30, Cranial 30 projection. When the over-the-wire balloon in left in the proximal section of the septal vessel the contrast runs to the RV cavity – as seen in Panel C. When the over-the-wire balloon is advanced to the sub-branch of the chosen septal artery contrast then localises to the target myocardium.
The variable branching pattern involving right and left ventricular myocardium also has an effect on the ability of injected alcohol to damage myocardium. The differing and unpredictable coronary pressures affect vascular resistance, the route and rate of venous drainage and hence the ‘dwell time’ of alcohol in the myocardium. This was reflected in the poor correlation between alcohol dose and CKMB release in our traditional ASA group. This is much improved when injecting into selective
LV branches as identified by CT, with a more predictable amount of damage observed (see Figure
6.18).
280
6.5.1.2 Alternative epicardial artery
The parent vessel for the appropriate septal artery is not always the LAD. Case reports have highlighted septal supply from the circumflex and intermediate arteries28 and right coronary artery139. In this CT guided series 5/49 (10%) of all identified septals originate from vessels other than the LAD (predominantly circumflex - 8%), and 2/21 (10%) of target vessels originate from the circumflex artery. This is higher than previously thought.
These patients may be dismissed as not having an appropriate artery by operators that may not consider parent arteries other than the LAD. The approach to identifying the correct artery is reversed in CT. In traditional ASA a vessel is followed from its origins looking for a course that may travel towards the target myocardium in the septum. Using CT we identify the appropriate vessel for alcohol delivery in the target myocardium and track it back to its source, wherever that parent artery may be. This will avoid missing vessels due to unexpected origins.
6.5.2 Effect of CT guidance on clinical outcomes
6.5.2.1 CT guided approach improves control of infarct size and location
CT guided ASA appears to give greater control of the size and location of infarct. The increased control is suggested by the better correlation of alcohol dose to CKMB release. Previously it was difficult to predict the extent of myocardial damage, in part due to the inability to control variable run off in to different myocardial territories. We can now predict extent of myocardial damage as indicated by CKMB release.
281
RBBB was a very common procedural observation using traditional methods in many case series, including ours110 144 145. Infarction of the right bundle and right ventricular myocardium is collateral damage and should not be the primary target. The true target is the left ventricular myocardium, aiming to reduce size and systolic excursion into the LVOT. The significant reduction in RBBB in CT guided ASA is indicative of more targeted infarct location in the LV myocardium. This is confirmed by
CMR studies. Seven of 9 patients who were able to undergo CMR post ASA showed RV endocardial sparing (see Figure 6.19). We were unable to perform CMR in 11; ICD in 5, PPM in 4, claustrophobia in 2. CT allows us to target the appropriate LV myocardium with greater accuracy.
Figure 6.91: CMR LGE SAX images post CT guided ASA
282
All SAX LGE slices are in the plane immediately below the LVOT. The target septum is the same as that detailed in Chapter 3. All images show LGE on the LV endocardial surface, panels A,B,C,F,G,H,I show RV septal sparing. All of the patients displayed had successful treatment of LVOT gradients.
6.5.2.1 Effect of better infarct localisation on clinical outcomes
A more accurate infarct will create a more favourable effect on LVOT haemodynamics and result in lower gradients. LVOT gradients and symptom burden were much improved after CT guided ASA.
This has been true for most case series reported historically. The average gradient in the CT group is composed of a greater proportion of successful procedures than our traditional group however. This is despite there being a reduced need for additional procedures required to reach this outcome. This greater procedural success rate results in an improved exercise capacity and a trend towards a greater proportion of patients reporting a symptomatic improvement. This does not reach statistical significance compared to the effect of a traditional approach in this relatively small group of patients.
6.5.3 Exploring those with treatment failure
6.5.3.1 Those who received alcohol
Two patients had LVOT gradient treatment failure according to pre-defined criteria. Both of these patients reported an improvement in symptoms from NYHA class 3 to 2. One patient subsequently underwent a second ASA procedure. He reported a complete resolution of dyspnoea up to 4 months post procedure 1, and the late return of NYHA class 2 symptoms. Echocardiogram at one month had shown resolution of LVOT gradient, with return of significant pressures at 6 months. CMR imaging showed a small infarct at the original SAM-septal contact point with a new, more basal contact
283 point. A repeat CT was performed and an alternative septal artery at the new SAM-septal contact point in the basal septum was injected with alcohol.
The other patient also had initial success with resolution of gradient at one month. A significant gradient returned at 6 months post ASA. A second ASA procedure was attempted. CT identified 2 small, basal septal arteries. Each had an acute angle of entry from the LAD (74° and 57° respectively).
Neither could be safely engaged, a coronary guide wire prolapsed out of the vessel when a balloon was advanced (see Figure 6.21). Other vessels were explored with myocardial contrast – no contrast localised to the target myocardium as predicted by CT.
Figure 6.92: Coronary angiogram showing guide wire entering target with acute angle
284
6.5.3.2 Those who did not receive alcohol
Five patients taken to the lab with the intention of performing ASA did not receive alcohol. One patient had left main stem stenosis confirmed by IVUS and did not have any contrast studies; he was referred for bypass grafting and surgical myectomy. CT and invasive angiography did not highlight the stenosis well. He was referred for surgical myectomy and bypass grafting.
The other 4 patients should be considered a failure for treatment with ASA. Three of these had undergone attempted ASA with unsatisfactory outcome previously. Two had received alcohol at prior ASA; in one no septal target could be identified with myocardial contrast echocardiography at previous attempts.
Two patients displayed direct venting of myocardial contrast in to the left ventricular cavity on arterial injection. The vessel injected was highlighted by CT to travel to the target area but had no dwell time to allow alcohol to cause damage (see Figure 6.23).
Figure 6.93: Myocardial contrast venting directly into LV on injection
285
Panel A shows the parasternal long axis on table image. Panel B shows the same image during contrast injection into the target septal artery. This vents directly into the LV cavity (arrow).
Two patients had vessels shown by CT to travel to target myocardium that could not be safely accessed with over the wire coronary balloons. The vessels highlighted originated from a diagonal and obtuse marginal branch respectively. The coronary wire had to pass too many bends when sited in these vessels, advancement of the balloon caused prolapse of the wire back in to the parent artery. These patients were unable to receive percutaneous treatment with ASA.
Options for treatment in those that did not receive alcohol
Five patients (4 did not receive alcohol and one did not receive alcohol at second CT guided procedure) have failed to respond to maximal medical therapy and cannot receive further transcoronary alcohol. They require an alternative method of LVOT gradient treatment. The most frequently available alternative method of septal reduction is surgical myectomy. This is available as an option at LHCH. Three of these patients were deemed too high risk for cardiac surgery – one died
3 months after failed ASA following a large cerebrovascular accident. The other 2 refused surgery after a full and frank discussion about risks and benefits.
Radiofrequency ablation offers a method of delivering myocardial damage that is independent of coronary anatomy. I will explore this option in Chapter 8.
6.5.4 Resolution of CT versus invasive angiography
286
The spatial resolution quoted by CT manufacturers and invasive angiography texts are very similar. A number of 0.3mm2 is provided by the manufacturer of our CT scanner (Siemens), and a value of
0.3mm is often quoted in invasive angiography. There are however some critical points in the interpretation of this basic data when considering invasive angiography:
Resolution is far better in real time, in the lab than on subsequent DICOM images. There is
substantial image degradation in the action of saving and potential further degradation of
images if they are viewed on suboptimal machines.
Even in the catheter lab there is a difference between screening and acquisition. Resolution
is better on acquisition.
The exact details of the resolution will also be affected by issues such as zoom. This is less of
an issue on the more modern flat plate systems.
The heart is moving. There are therefore issues both of spatial resolution and temporal
resolution. Resolution will be improved by a faster frame rate. We traditionally shoot at 15
frames per second. Increasing this to 30 frames per second will give better resolution, at the
expense of more radiation exposure.
There is also one further major point to consider when comparing CT to invasive angiography. This relates to the fact that we view flat, two-dimensional representations of the cardiac three- dimensional object. Even if the heart was not moving this would create major problems, as segments of the artery will simply be hidden by other portions of the same artery that lie either in front or behind in the current projected image. The apparent edge of an artery will also be rendered inaccurate when a portion of the same (or a different) vessel lies either in front or behind in the
287 missing third. The true contour of the vessel, and its path, cannot be determined. Furthermore it would be possible to make incorrect assumptions about the location on nature of branch points.
A major advantage of CT is that it allows an iterative process of continuous three-dimensional reconstruction. The process is more akin to old-fashioned tracking. You may cast around for the apparent path of the vessel only to be disappointed. This causes the observer to go back and reassess the situation, looking in a different dimension. Eventually the truth will emerge. This process is not possible with standard angiography. I believe this to be the basis of CT superiority. The magnitude of the advantage cannot simply be expressed on the basis of apparent pixel size
288
6.6 Conclusion
CT angiography provides anatomical insights that alter the approach to ASA. This
change improves control and location of infarct and translates into better clinical
outcomes.
A small proportion of HOCM patients require a method of percutaneous treatment that
is independent of coronary anatomy.
289
6.7 Limitations
The relatively low numbers of ASA performed in the study period make it difficult to come to conclusions with a high level of security. The numbers of ASA procedures performed per year in the
UK are low14 15. Despite LHCH being a quaternary referral centre for ASA the patient volume is still low. I was dependent upon referral patterns to recruit patients. The majority of patients referred to the clinics did not require septal reduction therapy. This method of guiding a low-volume procedure needs to be used in a greater volume of patients to assess if the perceived benefits continue.
The method of CT analysis and translation to the catheter lab has only been performed in one centre with a limited personnel base. This needs to be disseminated to other staff, other centres and other nations to see if the results are reproduced.
290
Chapter 7: Standard operating policy for assessment of patients referred for ASA at LHCH
291
7.1 Introduction
Liverpool Heart and Chest Hospital (LHCH) acts as a referral point for ASA for Merseyside, North
Wales, parts of Greater Manchester, Lancashire, Scotland, West Yorkshire and Cambridge. Historic rates of performance of ASA in Liverpool have been recorded as 8-10 per year. This has increased recently to >15 per year, this is a quarter of all cases performed in the UK15. Many patients referred for assessment do not need ASA or are better managed with an alternative approach. In the period
2012-2014 only 53% of those referred for assessment entered the lab with the intention of alcohol delivery, and 43% of patients receive trans-coronary alcohol. It is therefore clear that a robust method of assessment must be put in place to assess this group of patients to ensure ASA will be of benefit. Some may have alternative cardiac pathologies, some may have co-morbidities resulting in significant symptoms, and some HOCM patients may be better treated with alternative methods.
In order to highlight those patients appropriate for ASA I have developed a pathway for assessment which now acts as the standard operating policy at LHCH. Patient selection is important and should identify some of those who would not benefit from septal reduction. We know 16% of patients in the period 2000-2011 had an unsatisfactory outcome with no change in symptoms despite resolution of the LVOT gradient. These patients suffered with a range of co-morbidity, predominantly respiratory. In these, the LVOT gradient was not the ‘rate-limiting factor’.
Appropriate patient selection may allow better, more targeted therapy for these symptomatic patients.
A significant proportion of the patients described in Chapter 2 had an unsatisfactory outcome due to presence of persisting LVOT gradient. In 18% of patients we failed to have a significant effect on the
292
LVOT gradient with ASA, and a further 5% did not receive any trans-coronary alcohol. The major reason for failure to resolve LVOT gradient was inaccurate location of iatrogenic infarct. Our pre- procedural imaging allows us to accurately identify the causes of gradients, highlight the target myocardium for infarct, and plan a route to this target before entering the catheter laboratory.
Our experience and observations over what, by international standards, is a large group of HOCM patients has also allowed us to refine our delivery of inpatient care. This has been based on the detailed description of the historical group of patients and learning during the period of study performed in Chapters 3-6. The approach to the procedure of ASA has therefore evolved.
293
7.2 Aims
To develop a standard operating policy for the assessment of HOCM patients referred for
ASA
294
7.3 Section 1: Clinical review
7.2.1 Background to HCM diagnosis
A diagnosis of genetically determined HCM in the absence of alternative causes of LVH will be made by a referring hospital in most cases. This diagnosis should be reviewed and alternative diagnoses considered. The treatment modality can vary significantly if a different diagnosis is made.
7.3.2.1 Important clinical questions
When and where was diagnosis made?
How has the disease progressed since diagnosis?
What investigations have been performed to secure the diagnosis of HCM?
Has genetic testing been performed? Detail full family pedigree.
o Genotyping is performed in all patients in whom there is a clinical benefit of predictive
screening. It is not routinely performed to guide therapy in the individual – exceptions
are when there is a high clinical suspicion of a non-sarcomeric cause of HCM (e.g.
Anderson-Fabry, Danon disease)
Have all alternative causes of LVH been ruled out? The planned set of investigations at LHCH will
rule out most alternative causes but expert help should be sought for the more complex genetic
or metabolic disorders – genetics support in clinics has now been secured on a weekly basis for
such situations.
o Significant aortic stenosis
295 o Hypertensive heart disease o Athletic adaptation o LV non-compaction (sometimes this can look like LVH on echocardiography) o Infiltrative cardiomyopathy o LVH in the context of a syndrome (e.g. Noonan’s)
. Septal reduction should still be considered in this group93 o Metabolic disorders
. Most metabolic disorders will be diagnosed during childhood, but should be
considered in younger patients referred for septal reduction therapy.
. Treatment such as enzyme replacement may negate the need for septal
reduction in this group, it is therefore critical to ensure these diagnoses have
been satisfactorily ruled out.
. Has Anderson- Fabry disease been considered?
Men: Perform spot test in clinic if not previously investigated. Consider
genetic testing if borderline results.
Women: Spot test is less reliable. Risk-stratify for likelihood of
Anderson-Fabry based on clinical details. Refer for genotyping if
suspicious
296
7.3.2 Symptom burden
Detail clinical burden of HCM in terms of:
Dyspnoea:
New York Heart Association dyspnoea class
Progression of symptoms
Pattern of dyspnoea
o Stairs / sudden change from sedentary to high output exercise
o Post prandial change
Chest pain:
Canadian Cardiovascular Society chest pain class
GTN spray (consider cessation)
Response to other anti-anginal medications?
Syncope / pre-syncope:
What burden does this place upon the patient? Work pattern (if still in employment) etc.
Investigations for arrhythmia performed? If patient suffers from syncope it is important
to clarify this is not arrhythmic, this elevates risk of sudden cardiac death.
297
7.3.3 Medication review
7.3.3.1 Current medications
Negatively inotropic medication; detail current and past history of use of beta-blockers, verapamil and disopyramide.
Clinical response?
Side effects? Tolerable vs intolerable.
Contra-indications?
Stop all medications that could be exacerbating gradient and systemic hypotension. Nitrates should be stopped and consider switching any dihydropyridine calcium channel blockers to non- dihydropyridine versions. Traditionally ACE inhibitors and angiotensin II receptor antagonists were stopped due to the fear of exacerbating LVOT gradients due to systemic hypotension, but this was not seen in the INHERIT study when Losartan and placebo were randomised to HOCM patients126.
We therefore no longer stop these medications. There is also the suggestion of some benefit in reduction of fibrosis, but this is yet to be proven in large studies126 146 147.
7.3.3.2 Medication choice and titration.
A cardio-selective beta-blocker is preferred (Bisoprolol), this is based on ESC guidelines despite the evidence for betablockade being mostly with Propanolol148 149. Verapamil SR should be the calcium channel blocker of choice. Titrate to highest tolerable dose; consider side effects and effect on HR and BP.
298
Trial of β-blockade to HR 60-70bpm
No response
BP > 100 systolic
No Yes
Progess to septal reduction therapy QTc >460msec
No Yes
Consider adding/switching to Disopyrimide Consider switching to Verapamil SR
No response No response
Progess to septal reduction therapy Progess to septal reduction therapy
Key: Introduce medication Clinical assessment Progress to septal reduction therapy
Figure 7.94: Medication choice in treatment of LVOT gradients
7.3.4 Exploration of pacing options
Right ventricular apical pacing has been shown to be of some benefit in a small proportion of patients with symptomatic LVOT obstruction in HOCM150 151. This tends to be an older, female population. The effect is limited and does not warrant device implantation as first line treatment. If the patient already has a dual chamber device (pacemaker or implantable cardioverter defibrillator) this can be used to encourage RV pacing. If the patient has a high risk of SCD and an ICD is planned this takes precedent over ASA; once the ICD is in situ RV apical pacing is trialled. It is important to ensure the RV lead is placed in the apex and not in the septum, this can exacerbate LVOT gradients.
299
The device should be interrogated to determine right ventricular apical pacing %. If this is already
>95% the settings do not need to be changed and the patient has already had a trial of RV pacing.
An echocardiographic study should be performed to maximise the effect on the LVOT gradient, exploring differing AV delays to determine the optimal settings. During this echo-pacing combined study the following steps should be followed:
1. Determine the intrinsic AV delay:
a. Pace in AAI at base rate +20bpm – switch atrial lead to unipolar to ensure pacing
spike can be seen
b. Perform 12 lead ECG – measure interval of pacing spike to earliest intrinsic QRS
spike – this is the paced PR interval.
2. Program short AV delay – 100msec – and measure paced QRS width.
3. Set AV delay to paced PR interval minus paced QRS width – this is to ensure the all of the
ventricles are captured before intrinsic depolarisation and contraction occurs – the basal
septum will then be depolarised from the RV pacing rather than intrinsic conduction tissue.
4. Perform LVOT gradient assessment.
If the LVOT gradient is significantly reduced a trial of 6-8 weeks to assess effects on symptoms is warranted. If no significant change in LVOT gradient is seen the patient should progress to the next step in assessment.
300
7.3.5 Patient information
Treating physicians should undertake a full discussion with patients with regard to the pathology and implications of a diagnosis of HCM. This should be used as an opportunity to discuss genetic testing,
SCD risk and implications for the wider family of the index patient, cascade testing can also be organised.
A frank and detailed discussion regarding all septal reduction therapies should be undertaken with the patient. The treating clinician must discuss:
Intended benefits
Patient should be aware that the primary aim of ASA is to improve symptoms and functional status.
Some observational data have suggested that prognosis can be improved46 86, but this is far from secure given the nature of the research.
Procedural detail
Techniques and rationale should be explained. Patients should be aware of the risk of complication internationally and locally. These are detailed in the patient information leaflet (provided as appendix 7.1).
301
Alternative treatments
Many patients referred for consideration of alcohol ablation do not require ablation or can be treated successfully with less invasive methods. This is based on our experience and the description from other centres in America36. When septal reduction is required the main choice to be made is whether ASA or surgical myectomy are more appropriate for the individual. The treating physician must provide information in a balanced manner, most patients will choose a percutaneous method with shorter recovery times if it is available35.
Surgical myectomy is an alternative method of septal reduction in HOCM. This has a longer history, with good outcomes in experienced surgical centres. Although symptom resolution is comparable between treatment groups in comparisons of observational studies51 52, no randomised controlled trial has ever been performed66. Patient outcomes are better in myectomy patients treated in expert centres, but there are few dedicated surgical myectomy centres available, especially in the UK. There are no publicly available myectomy surgical audit data in the UK. Mortality in non-expert centres can be up to 5%63. Patients should be made aware that when alternative cardiac pathologies are identified that are not amenable to percutaneous treatment then surgical myectomy becomes the favoured option.
Radiofrequency (RF) ablation of the basal septum has been trialled in a small number of patients81 82
152. This has been offered on a palliative or research basis only so far. Our experience with RF ablation will be detailed in Chapter 8.
302
The ‘patient journey’ should be detailed with intended hospital experience described. Generally the following patient path is accurate (there are exceptions that are dealt on an individual basis):
Total time from OPD Time required Today Clinic appointment Investigations Echocardiogram CPEX Cardiac MRI CT coronary angiography Bloods / genotyping
2 weeks MDT 2 weeks
8 weeks Pre-assessment clinic 6 weeks
10 weeks Admission 2 weeks Inpatient care Ward care 2-3 hours Catheter laboratory - procedure 1-3 hours Recovery area 30 minutes Coronary Care Unit 24 hours Cardiac MRI and echocardiogram Ward care with ECG monitoring 3-4 days 11 weeks Discharge 4-5 days total
37 weeks Follow up clinical review 26 weeks Echocardiogram CPEX Cardiac MRI
Annual review thereafer 52 weeks
303
Figure 7.95: Patient journey from assessment to follow up
7.3.6 Criteria for progression to ASA: Clinical history
Clinician must be satisfied that the pathology is genetically determined hypertrophic
cardiomyopathy in the absence of other causes of LVH
NYHA ≥III; Rare exceptions made for NYHA II interfering with desired level of activity
in otherwise well, young patients with:
CCS class ≥II. No flow restricting CAD and / or
Recurrent, disabling haemodynamic syncope / pre-syncope with documented
absence of arrhythmia during symptoms.
304
7.4 Section 2: Echocardiographic assessment
7.4.1 Resting transthoracic echocardiogram (TTE)
All patients must undergo resting TTE at LHCH prior to consideration of ASA. Referring hospital TTE reports often don’t contain all required information. TOE requires sedation, this can affect loading conditions and therefore LVOT gradients; this is misleading and therefore is not routinely used unless further information regarding mitral valve anatomy is required.
TTE should be performed at or before the first clinic meeting to document LVOT gradient. If the resting and Valsalva gradients are <50mmHg an exercise stress echocardiogram must be performed.
This can be achieved at the same sitting as CPEX testing. A full stress echocardiogram protocol is not necessary. A measurement of peak LVOT velocity at or immediately after maximal exercise must be taken. This does not need supervision from an imaging Cardiologist, and can be reported by the sonographer taking the images in conjunction with the medic supervising the CPEX.
TTE should be based on British Society of Echocardiography (BSE) standard views with particular attention to:
7.4.1.1 Septal size
Measured in PLAX view as per BSE standards. Septal reduction therapy should only be performed in those with an interventricular septum diameter in diastole of ≥17mm12.
305
7.4.1.2 Systolic anterior motion (SAM) of the mitral valve
Comment should be made on severity after viewing from several perspectives. M-mode imaging in
PLAX should be performed to grade severity. SAM is critical in the pathophysiology of increased
LVOT velocities and therefore gradients153-155.
Severity of SAM is related to the extent of LVOT flow velocity abnormality and therefore gradient. A longer SAM-septal contact leads to higher LVOT gradients. Severity is graded 0-3 (see Table 7.1)153-
155.
SAM severity scale:
Grade Observation
0 Minimal SAM that involves the MV apparatus only. Also termed ‘chordal’ SAM.
1 Clear SAM of the AMVL, but no contact with the interventricular septum
2 AMVL is in contact with IVS. Length of contact is <30% systole (measured from onset of R wave to opening of MV in diastole)
3 AMVL is in contact with the IVS. Length of contact is >30% systole.
Table 7.40: SAM severity grading scale
Examples of each grade are shown below.
Grade 0:
Minimal or no SAM visible on mmode images.
306
Figure 7.96: Example of grade 0 SAM
Grade 1:
Mild SAM of AMVL seen, no contact with septum.
307
Figure 7.97: Example of grade 1 SAM
Grade 2:
SAM with contact of AMVL with septum. Length of contact constitutes <30% of systole measured from R wave on ECG to opening of MV in diastole. In this case the length of contact is 140msec, and systole measures 530msec (140/530 x 100 = 24.1%)
308
Figure 7.98: Example of Grade 2 SAM
Grade 3:
Prolonged SAM, contact of AMVL for >30% systole. In this case for 250msec. Systole is for 400msec,
(250/400 x 100 = 62.5%)
309
Figure 7.99: Example of grade 3 SAM
7.4.1.3 LVOT gradient
Blood flow velocity should be measured by CW Doppler assessment through the aortic valve from 5 chamber and 3 chamber apical views. The envelope should be traced as AV VTI and the peak gradient quoted on reports. The highest rest peak gradient should be quoted. The pathology of LVOT gradients in HOCM relates to SAM, this causes late flow acceleration and is typically displayed as a
‘scimitar’ shape envelope (See Figure 7.7). CW assessment of flow velocities should also be performed during Valsalva manoeuvre. The peak gradient during Valsalva manoeuvre should also be reported.
310
If the peak LVOT gradient is <50mmHg exercise stress echocardiography should be performed. This can be done in conjunction with CPEX testing.
Figure 7.100: Example of 'scimitar' shaped CW envelope of LVOT obstruction
The location of flow acceleration should be sought using PW Doppler. Velocities should step up immediately distal to the SAM contact point when the pathology relates to HOCM. Velocities that do not increase until the PW box is in the subaortic area of the LVOT should raise suspicion of a subaortic ring; this is not amenable to alcohol ablation and should prompt a surgical referral.
Velocities that step up in mid cavity should be highlighted, whilst alcohol ablation can be performed in these patients the technique has to change significantly and the outcomes are less secure due to low numbers156 157.
311
Care should be taken to ensure the velocity traces acquired do not involve mitral regurgitation.
HOCM pathophysiology causes posteriorly directed MR in a large proportion of patients. It may be useful to purposefully place a CW line through the MR; this trace should be compared to the LVOT envelope shape and velocity to ensure they differ.
7.4.1.4 Diastolic function
The extent of ventricular hypertrophy can regress and diastolic function can subsequently improve if obstruction to flow is removed 91 92 158. This suggests that not all of the hypertrophy in HCM with
LVOT gradient is genetically determined and some is afterload dependent. Full assessment of diastolic function should therefore be performed as per BSE standards.
7.4.1.5 LVOT anatomical variants associated with HOCM
Care should be taken to look for presence of subaortic rings and anomalous insertion of anterior papillary muscle19. If the septal size appears to be normal but a significant gradient is present alternative causes such as abnormal MV geometry should be explored. The anterior mitral valve is often elongated in HCM as part of the phenotype135 159. A measurement can be taken in the apical 3 chamber view from hinge point to tip, in a CMR based study the AMVL length in HOCM was 26mm vs
19mm in a control group135.
A detailed assessment of all valve function should be performed as per BSE standards132. Posteriorly directed MR is associated with SAM of the AMVL and is usually not due to intrinsic MV disease that
312 requires surgical correction. Central MR associated with any valvular abnormality is unlikely to be related to HCM and should prompt detailed analysis for suitability for surgery, this often requires
TOE. Any disease that may require surgery will change the treatment plan for HOCM; septal myectomy may be required rather than ASA.
7.4.2 Criteria for progression to ASA: Echocardiographic assessment
Patients must satisfy all of the following criteria:
Septal size ≥17mm
Resting, Valsalva or exercise stress LVOT gradient of >50mmHg
Absence of alternative cause for gradient
Absence of significant valvular or other cardiac disease that may require surgery
313
7.4 Section 3: Functional testing
7.4.1 Full pulmonary function testing
Consideration must be given to the primary rate limiting factor when describing exercise restriction.
Retrospective cohorts have included significant incidence of respiratory co-morbidity in patients with removal of LVOT gradient but persisting symptoms. Half of those that did not improve symptoms after ASA with resolution of gradient at LHCH had significant lung disease (see Chapter 2).
Therefore full PFTs with transfer factor must be arranged for all patients under consideration for
ASA. Abnormal PFTs (as per Table 7.44) trigger a referral to respiratory physicians for further investigation and treatment as necessary.
Measurement Consider respiratory limitation if: FEV1 <50% predicted FVC <50% predicted KCO <60% predicted (with normal Hb) VC <60% predicted
Table 7.41: Abnormal PFT triggers for referral to Respiratory team
7.4.2 Cardio-pulmonary exercise testing
Cardiopulmonary exercise testing (CPEX) is performed in all patients referred for ASA. Standard bicycle protocol with no resting load and 10W increments per minute is adopted. The patient must reach a respiratory exchange ratio (RER) ≥1.1 to indicate a satisfactory test. A drop in pulse oximetry measured oxygen saturations is strongly indicative of respiratory disease; most cardiac disease does
314 not cause hypoxia. This should prompt clarification with arterial blood gas analysis to confirm hypoxia.
Peak VO2 should be calculated based on an average of the final 7 readings taken at 5 second intervals over the final 30 seconds. This method should also be used for O2 pulse measurements. An estimated peak VO2 based on an age and sex matched population without HCM should be calculated using the Wasserman equation (this is already part of the CPEX computer module at LHCH). A low anaerobic threshold (AT) is an indicator of cardiac rather than respiratory restriction. An AT that is achieved at <40% predicted VO2 generally indicates cardiac restriction.
I have created and stored a reporting template labelled ‘HCM’– this will give all the above required readings. The test can then be supervised by any medic and I can report the results remotely. I also created a flow chart for cause of exercise limitation according to CPEX results (see Figure 7.8)
315
FEV1<50% predicted OR Yes
FVC<50% Predicted OR Group 1 Respiratory Or Group 5 Mixed KCO <50% predicted OR VC < 50% predicted
No
Yes
O2 saturation drop ≥ 5% Group 1 Respiratory
No
Yes
RER <1.1 Group 2 Insufficient test Or Group 1 Respiratory
No
Yes
Peak VO2 >85% predicted Group 3 No significant limitation ?Hyperventilation
No
Yes
Ventilatory anaerobic threshold > 40% of predicted VO2 Group 1 Respiratory and VAT > 11ml/min/kg
No
Yes
Breathing reserve <30% Group 1 Respiratory
No
Group 4 Cardiac
Figure 7.101: Flow chart for CPEX investigation of cause of exercise restriction
Haemodynamic measurements during CPEX
Perform BP measurement at rest and at 2-minute intervals throughout the test. BP response to exercise should be noted and categorised into normal (>20mmHg elevation on exercise), blunted
(<20mmHg elevation on exercise), or abnormal (BP drop). HR response is variable in patients being assessed for ASA due to treatment with bradycardic medications. No target HR is set for these patients, peak exertion is guided by continuous RER measurements.
316
Indications for cessation of test
Exhaustion with RER ≥1.1
BP drop with symptoms of pre-syncope
Sustained VT
7.4.3 Criteria for progression to ASA: Functional testing
Patient must satisfy the following criteria:
Pulmonary function testing does not suggest respiratory restriction
CPEX testing
o Satisfactory test to exhaustion and:
o Oxygen saturations remained >95% throughout the test
o Restriction in peak VO2: <85% predicted
317
7.5 Section 4: Cardiac Magnetic Resonance Imaging:
Cardiac MRI (CMR) and echocardiography are complimentary imaging modalities in the assessment of HOCM patients when considering ASA. The major incremental benefits of CMR are providing excellent image quality to assess for anatomical variants and identifying and quantifying fibrosis with the use of gadolinium enhancement imaging.
7.5.1 Rule out phenocopies of HCM
Alternative causes of left ventricular hypertrophy (LVH) should be ruled out before progressing to
ASA. Distinctive LGE patterns can be seen in those with LVH that may point towards alternative diagnoses such as amyloidosis, sarcoidosis and Anderson-Fabry disease. Further investigation will be required if CMR shows suspicious Gadolinium kinetics or enhancement patterns.
7.5.2 Alternative anatomical abnormalities responsible for LVOT gradient
Alternative causes for increased LV pressures should be ruled out before progressing to the catheter lab. Particular attention should be made to rule out obstruction at several levels of outflow from the left heart:
Mid-left ventricle
Septal hypertrophy that is predominantly mid ventricle with aliasing of blood flow apical to the SAM- septal contact area should raise the suspicion of mid-ventricular obstruction to flow. The benefits of
ASA are less clear in this pattern or hypertrophy.
318
Subaortic region
Aliasing of flow after the SAM-septal contact area may be indicative of a subaortic ring. If the patient qualifies for septal reduction therapy myectomy should be the treatment of choice.
Aortic valve
Restricted opening of the AV should have already been highlighted on echocardiogram. Ensure the valve is not bicuspid, if the echocardiographic windows are poor this may be the only chance to rule out AV disease.
Supra-aortic region
Flow acceleration seen in the ascending aorta associated with supra-valvular stenosis (Williams syndrome) can be ruled out with standard 3-chamber views.
Mitral valve geometry
The pathology of LVOT obstruction in HOCM is critically linked to the presence of SAM of the AMVL.
Any pathology that deforms the anatomy of the MV can lead to SAM and obstruction. Any possible cause of altered MV and LV geometry must be considered prior to progression to ASA. An example is shown in Figure 7.9: In this patient a calcific mass (proposed to be a calcified paraganglioma) in the posterior AV groove pushed the mitral valve apparatus forward towards the LVOT. As a consequence the AMVL is caught in forward flow out of the LV and obstructs flow. The IVSd measurement was
319
11mm, the mild hypertrophy is likely secondary due to increased afterload as part of the obstruction process.
Figure 7.102: Example of altered MV geometry causing LVOT gradient
Panel A; Apical 3 chamber echocardiogram showing calcified mass at posterior MV annulus. SAM is also seen. Panel B: Echocardiogram short axis view at MV leaflet level. Panel C; CMR image showing the posterior mass forcing the MV apparatus towards the LVOT, both the posterior and anterior leaflets are pushed into the forward flow through the LVOT resulting in flow acceleration seen as aliasing. Panel D; LGE CMR image delineating borders of mass.
320
7.5.3 Visualise pre-existent fibrosis
A substantial proportion of patients have failure to resolve the LVOT gradient following their first alcohol ablation procedure. In the period 2000-March 2012 at LHCH 41% had persisting gradient after their first alcohol ablation procedure and 18% had persisting gradient at the end of available treatment. A critical part of evaluating this group of patients is visualisation of iatrogenic infarct. The appearance of fibrosis as part of the pathogenic process of HCM on late gadolinium enhancement images mirrors infarct. It is essential we document pre-existing scar to ensure this is not confused with infarct created by alcohol ablation.
7.5.4 CMR protocol for pre-procedural investigation
All scans are performed on a 1.5-T Siemens Magnetom Aera scanner using a 32-channel cardiac phased-array receiver coil. Images are acquired using electrocardiogram (ECG) gating during multiple short breath holds (8 to 15s). A series of TRUFI localisers are acquired.
A transverse HASTE stack through the thorax is acquired using 16 8mm slices. A Vertical Long Axis
(VLA, 2 Chamber), Horizontal Long Axis (HLA, 4 Chamber) and short axis planning images are used in planning the subsequent Steady-State Free Precession (SSFP) images.
Functional performance of the ventricles is assessed using cine SSFP images. The VLA cine is acquired first, this is used to plan the HLA cine. This is followed by multiple parallel short-axis slices (8-mm slice thickness, 2-mm gap) every 10mm covering the entire ventricle from base to apex. To perform a detailed assessment of the dynamics of mitral–septal interaction high-resolution images using 50
321 phases in the 3-chamber view are obtained; 3 slices 5mm apart are planned from the short axis SSFP images. In-plane flow studies are then performed with routine velocity encoding (VENC) of
200cm/sec, if aliasing is observed the VENC Is increased at the operator’s discretion. In the CMR performed at day 2-3 post-ASA a short axis, 3 chamber and 4 chamber Short Tau Inversion Recovery
(STIR) images were also acquired. A ‘look locker’ series acquiring a single-slice multi-phase sequence with a shared inversion prepulse to assess for the optimal inversion time is then performed. This also serves to examine Gadolinium kinetics for amyloidosis. Delayed-enhancement images were obtained
8-10 minutes after injection of Gadavist® gadolinium-diethylenetriamine pentaacetic acid
(gadobutrol) 0.1mmol/kg (Bayer®; Germany) using a segmented inversion recovery prepared fast gradient echo sequence and single shot phase swap sequence. The prescription for this sequence was identical to the short-axis cine sequence, 4 chamber and 3-chamber views to ensure image registration.
I have exported the study cards from the Siemens MRI scanner module with the planned images for the three types of scans performed: ‘HCM pre-ASA’, ‘HCM immediate post ablation’ and ‘HCM 6 month scan’. The protocols cards are copied below:
7.5.4.1 HCM pre-ASA trufi_loc_multi_iPAT trufi_loc_multi_iPAT@c haste_ tra
2_chamber_plan shortaxis_plan
4-chamber_plan
322
2ch_cine sa_check
4ch_cine sa_stack_cine
3ch_3_view rvot
LVOT_cor flow_LVOT_though_plane_venc_200 flow_LVOT_in_plane_venc150
TI-Scout
4ch_stack_de_high-res sa_stack_de_high-res
2ch_de_high-res
3ch_de_high-res
SA_single_shot_phase_swap
7.5.4.2 HCM Immediate Post Ablation trufi_loc_multi_iPAT trufi_loc_multi_iPAT@c haste_ tra
2_chamber_plan shortaxis_plan
4-chamber_plan
2ch_cine sa_check
323
4ch_cine
RVOT sag sa_stack_cine
LVOT_Sag_50_phases_3_view
STIR 3 Chamber Views
STIR 4ch
STIR SA to cover basal slices & oedema flow_LVOT_though_plane_venc_200 flow_LVOT_in_plane_venc150
Dynamic_rest TI-Scout 3Ch 3 slices & 4Ch
Early DE 4ch_stack_de_high-res
Early sa_stack_de_high-res to cover oedema
4ch_stack_de_high-res 2ch_de_high-res sa_stack_de_high-res 3ch_de_high-res
SA_single_shot_phase_swap
DE_single_shot_4Ch_phase_swap
DE_single_shot_2Ch_phase_swap
DE_single_shot_3Ch_phase_swap
7.5.4.3 HCM 6 month scan trufi_loc_multi_iPAT trufi_loc_multi_iPAT@c haste_ tra
2_chamber_plan
324 shortaxis_plan
4-chamber_plan
2ch_cine sa_check
4ch_cine
RVOT sag sa_stack_cine
LVOT_Sag_50_phases_3_view
Selective_STIR_SA
STIR_3Ch flow_in-plane_venc_200 flow_LVOT_though_plane_venc_200
TI-Scout
4ch_stack_de_high-res sa_stack_de_high-res
2ch_de_high-res
3ch_de_high-res
SA_single_shot_phase_swap
DE_single_shot_4Ch_phase_swap
DE_single_shot_2Ch_phase_swap
DE_single_shot_3Ch_phase_swap
325
7.6 Section 5: CT coronary angiography imaging
Following the outcomes described in Chapter 6 we now perform CT coronary angiography in all patients to describe septal arterial anatomy and assist in planning ASA.
7.6.1 Protocols for CT angiography
7.6.1.1 Acquisition of images
All patients with a presenting heart rate greater than 60 beats per minute are given oral or intravenous metoprolol. Coronary CT angiography (CTA) is performed using a dual- source CT system
(‘Definition Flash’, Siemens Healthcare, Forcheim, Germany) with 2 x 128 x 0.6mm collimation, 0.28 second rotation time, a pitch of 3.2-3.4, 100 KV tube voltage and current of 320 mAs. Data acquisition is prospectively triggered at 60 % of the RR interval. This is performed either as high – pitch spiral acquisition (‘Flash spiral’) or sequentially with a table feed of 38.4mm.
All data is acquired in deep inspiration. Coronary CTA is performed using intravenous contrast enhancement. To synchronise acquisition of the coronary CTA data set to arterial enhancement, a test bolus protocol is used; 15mls contrast agent (Optiray 350, Covedian, Mallinckrodt Medical
Imaging Ireland), followed by 40mls saline solution at 6mls /s. The time to peak enhancement in the aorta is measured using a series of trans-axial scans acquired in 2s increments, with first image acquired after 12s. For CTA 60mls of contrast is injected followed by a saline flush of 40mls, both at flow rates of 6mls/s. Image acquisition is started with a delay of measured transit time plus 2s. For image reconstruction a half scan reconstruction algorithm is used which provides a temporal
326 resolution of 75ms. Reconstructed slice thickness is 0.6mm, slice increment was 0.3mm. The reconstructed images are analysed with SYNGO software via a multimodality workstation.
7.6.1.2 Protocol for analysis of images
Assessment of CT angiography is performed with a cardiac radiologist and cardiologist with an interest in heart muscle disease. The target area of myocardium is identified using a short section of systolic imaging. The target myocardium is examined to identify a segment of its arterial branch supply. This vessel is tracked (using a frame to frame single plane reformat approach) back to its artery of origin. The length of its course in epicardial fat (before insertion to muscle) and myocardium are noted. The branch pattern is described, with locations, lengths and destinations of all sub-branches detailed. Landmarks to aid identification of the chosen septal vessel in the catheter lab are highlighted. This target artery is traced in the CT software package and a multiplanar reformated coronary angiogram ‘map’ created. The coronary tree map is then rotated through horizontal and vertical planes to allow optimal visualisation and remove any overlap or foreshortening. The optimum angiographic projections are noted, these are then used as the
‘working views’ in the catheterisation laboratory (examples are shown in Appendix X). Finally, the left main stem, LAD, circumflex and RCA are then surveyed for other vessels tracking towards the septum. All potential target vessels are followed and ultimate distribution noted, to allow identification, or rejection as a potential supplementary target vessel.
A series of images and films are created from the CT scan. The cardiologist with an interest in heart muscle disease takes these to the ASA planning meeting with the interventional cardiologist performing ASA. The procedure is then planned, this includes details such as planned radiographic
327 projections, balloons and other equipment required, proposed alcohol dosing and individualised risks of procedure.
328
7.7 Section 6: Multi-disciplinary team meeting to discuss suitability for septal reduction therapy.
A multi-disciplinary team (MDT) approach is used to decide on best treatment modalities. Team members present at our MDT meetings include cardiologist with an interest in heart muscle disease, interventional cardiologist, imaging cardiologist, named cardiac surgeon for myectomy, specialist nurse, clinical geneticist and electrophysiologist. A decision regarding best treatment option is made based on the information gathered as part of the assessment protocol in sections 7.3-7.6. This MDT meeting is monthly. The operative and post-operative care for these patients is complex and a clear, individualised plan should be specified.
329
7.8 Conclusion
Patient selection in ASA is important. A clear operating pathway has allowed LHCH to process these patients in a more uniform manner, ensuring all anatomical and functional data is collected to inform the MDT decision making process. This is now in routine use in the Inherited Cardiac
Conditions clinics.
330
Chapter 8: Radiofrequency ablation of the interventricular septum to treat outflow tract gradients in hypertrophic obstructive cardiomyopathy: A novel use of CARTOSound® technology to guide ablation
331
8.1 Introduction
The ability to perform ASA to treat LVOT gradients in HOCM depends on suitable septal arterial anatomy. In some series operators are unable to deliver alcohol due to difficulties in 15% of patients entering the lab with the intention of alcohol delivery88. In our historical series we were unable to deliver alcohol in 5% of patients. Even with the use of pre-ASA CT guidance there were anatomical restrictions that precluded the delivery of alcohol in 4/26 (15%) patients. One of the 4/26 patients is duplicated in this series as they had also experienced a previous failed procedure.
These ‘technical failures’ generally fall in to three categories:
An inability to locate the artery supplying the target for ablation, or the inability to
instrument this artery due to unfavourable size, angles and pressures.
Direct communication between septal arteries and LV cavity resulting in contrast
extravasation into the systemic circulation.
Collateralisation to a distant vessel and territory via arcade vessels.
An alternative method to damage the basal septum that does not rely on arterial anatomy would not be subject to these limitations.
Endocardial radiofrequency (RF) ablation, widely used in the management of arrhythmias, has previously been used to ablate the hypertrophied septum of HOCM patients 81 82. Techniques have involved the visualisation of the septal hypertrophy by transoesophageal echocardiography or the use of an electroanatomic electrophysiology (EP) mapping system (CARTO®) to guide and mark
332 ablation lesions. In Chapter 4 I was able to show that Intracardiac echocardiography (ICE) provides high quality images of the SAM-septal contact point. ICE was not superior to TTE overall due to its deficiencies in visualising important structures away from the target myocardium and the inability to see contrast localisation – neither of these qualities are required if using RF ablation rather than trans-coronary alcohol. ICE therefore provides an ideal imaging modality to guide endocardial RF ablation in the LV.
ICE and CARTO technologies have inherent advantages and a technology that seamlessly integrates both exists in the form of CARTOSound® (Biosense Webster, Diamond Bar, CA, USA). The
CARTOSound® module and Soundstar™ catheter integrate real time ICE images into the CARTO® mapping system. The Soundstar™ catheter tip also contains a navigation sensor and can be visualised on CARTO maps®, and RF ablation catheter tips can be seen in the live ICE image. The
CARTO technology allows mapping of electrical conduction tissue, knowledge of the location of the
His, right and left bundles may allow us to avoid this tissue and reduce the risk of complete heart block.
333
8.2 Aims
To explore the use of CARTOSound guided endocardial RF ablation to treat LVOT gradients in
HOCM
334
8.3 Methods
8.3.1 Patient selection
I have detailed the treatment path of all patients referred to LHCH for treatment of LVOT gradient in
HOCM in Chapter 2 (2000-2011) and Chapter 6 (2012-2014). Percutaneous septal reduction in the form of ASA has been available at LHCH since 2000. ASA has been the primary method of septal reduction, with myectomy performed in cases where additional cardiac pathology needed addressing (mitral valve surgery, coronary artery bypass grafting). Septal myectomy was also used in cases where ASA had failed. Four patients who failed ASA or who could not receive alcohol to the target area were deemed to be satisfactory surgical candidates and underwent septal myectomy.
Despite best attempts at standard treatments including medication optimisation, ASA, RV apical pacing optimisation in those with pre-existing devices and surgical myectomy there remained 6 patients who needed an alternative form of treatment. One patient was deemed unfit for RF ablation due to advancing age and cerebrovascular disease. This patient subsequently died within 1 month of assessment (Stroke was listed as cause of death). The remaining 5 could not undergo myectomy because of surgical risk (2) and patient choice (3) (see Table 8.45). These 5 patients underwent CARTOsound® guided RF ablation; 80% female, mean age 59.20 (44-79) years (Table
8.1.). All had a resting or exercise provoked LVOT gradient >50mmHg associated with significant
SAM. All patients were suffering with NYHA class 3 dyspnoea.
.
335
ID Gender Age Resting (provoked) IVSd ASA attempts Reason myectomy gradient (mmHg) (mm) (doses delivered) not performed
1 F 47 80 (105) 17 2 (1) Patient choice
2 M 59 15 (55) 17 4 (2) Patient choice
3 F 48 35 (85) 18 1 (0) Patient choice
4 F 70 84 (84) 21 3 (1) Operative risk
5 F 64 128 (166) 23 3 (1) Operative risk
Table 8.42: Basic demographic and echo details of patients undergoing RF ablation
8.3.1.1 Individual patient details
Patient 1
Forty four year old female with no other significant medical history. New York Heart Association
(NYHA) class 3 dyspnoea and Canadian Cardiovascular Society (CCS) class 1 chest pain. Resting LVOT gradient was 80mmHg, interventricular septum width in diastole (IVSd) measured 18mm. Severe
SAM. Peak VO2 assessed by cycle ergometer was 17.1mL/min/Kg representing 77% of a predicted value.
She underwent attempted alcohol ablation in 2011; this was unsuccessful due to the inability to instrument the chosen septal vessel to deliver alcohol (balloon shaft kinking). She had declined surgical myectomy at multiple consultations.
336
Patient 2
Fifty nine year old male with no other significant medical history. NYHA class 3 dyspnoea, no reported chest pain. Resting LVOT gradient was 14mmHg, exercise induced LVOT gradient of
55mmHg. Mild SAM at rest with severe SAM on exercise. IVSd measured 17mm. Peak VO2 measured
16.8mL/min/Kg representing 58% of a predicted value.
Alcohol ablation was attempted in 2002 and 2004 with delivery of alcohol on both occasions. A pacemaker was implanted following complete heart block after the second procedure. An initial improvement was noted following both procedures but this deteriorated after approximately 6 months. A third attempt at alcohol ablation was performed in 2005 but no artery for delivery of alcohol could be identified. A further attempt to identify an appropriate artery using CT coronary angiography guidance was attempted in 2012. Although a vessel travelling to the target septum could be seen this vented directly into the LV cavity on contrast injection. He had declined surgical myectomy at multiple consultations.
Patient 3
Forty seven year old female. She had a background history of type 2 diabetes on oral hypoglycaemic agents, and used regular inhaled salbutamol for asthma. NYHA class 3 dyspnoea. Resting LVOT gradient 35mmHg, rising to 85mmHg with Valsalva manoeuvre. IVSd 17mm. Peak VO2 measured
15.6mL/min/Kg representing 73% of a predicted value.
337
Alcohol ablation was attempted in 2012. The septal vessel travelling to the target myocardium connected to the LV cavity, contrast was seen to flow to the LV cavity on injection as in Patient 2. No other vessel supplied target myocardium. She had declined surgical myectomy at multiple consultations.
Patient 4
Sixty nine year old female with significant coronary artery disease; previous PCI to the right coronary and left main stem arteries. She had a recurrent normocytic anaemia of unknown cause requiring regular transfusions. Her local haematology team were satisfied that there was no malignancy. Mild
COPD on regular inhalers. NYHA class 3 dyspnoea (even when haemoglobin normal), CCS class 2 chest pain with no coronary revascularisation targets. Resting LVOT gradient 84mmHg, IVSd 23mm.
Peak VO2 measured 14.0mL/min/Kg, representing 64% of a predicted value.
She underwent attempted alcohol ablation in 2007, no alcohol was delivered as a vessel supplying the target myocardium could not be found. Repeat attempt at alcohol ablation in 2012 following identification of a septal artery originating from the diagonal branch using CT angiography was again unsuccessful. This artery could not be safely accessed with a balloon due to a tortuous arterial course. Her surgical operative risk was prohibitively high.
Patient 5
Sixty-six year old female. Obese with BMI 44 kg/m2. NYHA class 3 dyspnoea. CCS class 1 chest pain.
Resting LVOT gradient 124mmHg, severe SAM, IVSd 20mm. Peak VO2 measured 12.4mL/min/Kg
338 representing 89% of a predicted value. She also had episodes of paroxysmal AF progressing to persistent AF; this required DC cardioversion prior to RF ablation.
Her first attempt at alcohol ablation was in 2012. Myocardial contrast localised to the SAM septal contact point and alcohol was injected. An initial symptomatic improvement was noted up to 6 months, followed by return of LVOT gradient and symptoms. Cardiac magnetic resonance (CMR) studies highlighted a new SAM-septal contact point at a more basal area of the septum. Repeat attempts to access tiny septal arteries shown to be supplying the target myocardium in 2013 were unsuccessful. Her surgical operative risk was prohibitively high.
8.3.2 Research permissions
Appropriate permissions to perform RF ablation in these patients were granted by the research and development board of Liverpool Heart and Chest Hospital (see Appendix 8.1). Procedures were only performed in patients who could not receive effective ASA and were not candidates for surgical myectomy. All procedures were performed after detailed discussions and with full informed consent
(see Appendix 8.2 for patient information leaflet).
339
8.3.3 Patient assessment:
8.3.3.1 Imaging assessments
Patients underwent resting ± exercise stress echocardiography for LVOT gradient assessment pre- and 6 months post-procedure (Phillips ie33 scanner, Phillips S5-1 probe). Exercise stress echocardiography was performed as part of cardiopulmonary exercise testing at peak exercise in those that had a resting and Valsalva induced gradient <50mmHg.
Cardiac magnetic resonance (CMR) was performed prior to and 6 months after RF ablation.
Examinations were performed using a 1.5-Tesla scanner (Magneton AERA; Siemens, Medical
Imaging, Erlangen, Germany). Left ventricular volumes, ejection fraction and LV mass were determined using CMRtools (CVIS, London, UK). For late gadolinium enhancement (LGE) imaging, 0.1 mmol/kg gadolinium-diethylenetriamine pentaacetic acid (Gadovist, Bayer Schering, Berlin,
Germany) was administered intravenously and standard breath-hold inversion recovery imaging was performed. LGE short axis images were analysed with ImageJ software using previously validated full width half maximum thresholding method100.
8.3.3.2 Functional assessments
Cardiopulmonary exercise testing was performed using a bicycle ergometer protocol with 10W minutely increments. Euroqol EQ5D-5L quality of life questionnaires were completed in outpatient clinics at first consultation and prior to knowledge of haemodynamic and imaging outcome at follow up at 6 months.
340
8.3.4 Procedural details
All procedures were performed by the same first operator, I acted as first assistant. A proctor was present for the first two cases as per stipulations in our research agreements (see section 8.3.4).
All procedures were performed under general anaesthesia using propofol and atracurium followed by isoflurane titrated to bispectral index monitoring. Vasoconstrictors were used as required and hemodynamic monitoring was with invasive arterial blood pressure only.
The SoundStar™ catheter was inserted via the right femoral vein and manipulated into the right ventricle (RV) inlet. Multiple images were acquired with the probe in different positions and orientations. In each view the structures that can be seen are delineated with precision, this was then stored and the catheter and beam of ultrasound was moved. The same process was repeated until we had coverage of the relevant structures. Endocardial borders, papillary muscles, aortic cusps and coronary ostia were delineated and transferred into the CARTO® system (Figure 8.1). The
CARTOsound console then constructs a multiplanar reformat map by synthesis of the individual image records and combines this with live ICE images.
341
Figure 8.103: ICE delineated LV endocardial contours transferred to CARTO
The ICE probe is situated in the RV inlet and rotated to face the interventricular septum. Several planes of ultrasound are used to create a complete image of the LV (RV and aorta shells are also created). Panel A shows the LV CARTO map being constructed. Panel B shows the corresponding endocardial ICE contours (green lines, i-iv) used to create the CARTO map
The structural borders were manually contoured and recorded in the traditional method at the end of diastole. A new high density, multiplanar CARTO map was then created of the multiple regions of contact of the anterior mitral valve leaflet and the hypertrophied septum (SAM-septal contact map -
Figure 8.2). Due to physiological restrictions (i.e. contact only during systole) this map was created in systole. The SAM contact map was superimposed on the LV shell and was the target for RF energy delivery.
342
Figure 8.104: SAM-septal contact area mapping
Panel A shows the LV, Aorta and SAM-septal contact maps in the CARTO shell. The ICE probe is located in the RV inlet and directed towards the interventricular septum. The plane of ultrasound can be seen on the CARTO image to allow the operator to know exactly what level of LV and SAM contact is imaged. Panel B shows the corresponding ICE image. The green line is annotated by the operator to mark the line of SAM-septal contact in various ICE planes, creating a full SAM-septal contact map (pink). Panels A and B show a very basal area of SAM-septal contact. The ICE probe is realigned and the process is repeated (panels C & D, and Panels E & F) with each contour adding to
343 the SAM-septal CARTO map. Once the SAM-septal contact map is completed, we can accurately estimate the area of SAM septal contact in systole, in this case 3.2cm2 (Panel G).
Quadripolar diagnostic EP catheters were also used to allow RV apical pacing and demonstrate the location of the His bundle (Figure 8.3). Retrograde aortic access was used to enter the LV. After introduction of the ablation catheter to the LV, intravenous heparin was administered to keep activated clotting time >200 secs. The left bundle branch, left anterior and posterior fascicles were directly mapped and annotated on the CARTO shell and their positions noted in relation to the SAM- septal contact area (Figure 8.3). The conduction was mapped in an attempt to avoid damage and need for permanent pacing.
Figure 8.105: CARTO mapping of conduction system
Markers denoting His bundle (yellow), left anterior and posterior fascicles (white) and ablation lesions (red) are seen in the CARTO image (panels A & B). MAP catheter pressure over the left
344 bundle branch within the pink SAM-septal contact area leads to development of left bundle branch block (panel C).
Radiofrequency energy was delivered via Navistar and THERMOCOOL ® catheters (Biosense
Webster, Diamond Bar, CA, USA). Catheter stability was initially difficult due to the dynamic septum and turbulent LVOT blood flow. As ablation progressed and wall motion abnormalities were induced, it was easier to achieve stability. Anterograde access (from initial femoral venous access and then from the RA to LA via a trans-sepal puncture and finally by advancing the catheter across the MV) was attempted in one patient in an attempt to improve catheter stability, this was unsuccessful and difficult to manipulate. Retrograde aortic access was more stable and therefore adopted for this and other procedures. THERMOCOOL ® SMARTTOUCH catheters were used for patients 3-5 to allow estimation of endocardial contact. Using a combination of CARTO and intracardiac echo navigation radiofrequency energy lesions were placed over the SAM-septal contact area (Figure 8.4). If possible, specialised conduction tissue (e.g. left bundle/fascicles) was avoided unless it coursed through the central SAM-septal contact area.
345
Figure 8.106: RF burns over SAM-septal contact area
Panels A and B show an RAO projection of the process of RF energy delivery over the SAM septal contact area. Ablation is also performed around the perimeter of the SAM map to compensate for the potential inaccuracy associated with a SAM map created in systole. Panel C demonstrates the automatic ‘tracking' of the RF catheter tip on the live ICE screen (green halo). The papillary muscles can also be seen in the ICE images and marked on CARTO. Panels D & E show the final RAO and LAO projections of RF delivery. The medial displacement of the ablation lesions compared to the SAM map (panel E) is a function of systole vs. diastole-acquired points.
346
Stable contact forces of >10g/>2 bars were sought before RF application in patients 3-5 (when using a SmartTouch catheter). Catheter stability was improved with the onset of rapid right ventricular apical pacing in patient 5. Radiofrequency ablation powers of 50-60W limited to temperatures of
60°C were used with saline irrigation at 30 ml/min. Pending catheter stability 2-minute application times were used per lesion. A mean of 33.60 (28-42) minutes of RF energy were applied (Table 8.2).
The procedure was declared complete when we had achieved total coverage of SAM-septal contact area and could observe basal septal akinesia. Resolution of LVOT gradient could not be used as this paradoxically increases immediately after RF ablation. Myocardial oedema could be seen up to
10mm from the endocardial LV surface at cessation of energy delivery. The mean resting SAM-septal contact area under general anaesthetic was 2.1cm2 (0-3.2cm2). The average ablated area was
14.6cm2 (7.5-23.1cm2), representing 9.9% of total LV endocardial surface (5.6-14.2%).
Anticoagulation was reversed at the end of the case using intravenous protamine.
347
ID LV access Catheter Power RF SAM- Ablation % LV Additional (W) time contact area endocardial techniques (mins) area (cm2) (cm2) surface ablated
1 Retrograde Navistar B 50 28 1.8 8.6 8.3 - aortic curve
2 Retrograde Navistar D 50 32 0 7.5 5.6 - aortic curve 3 Retrograde SmartTouch 60 30 2.5 13.4 8.2 - aortic and D curve T/S access
4 Retrograde SmartTouch 60 42 2.8 18.2 14.2 Agilus aortic D curve steerable sheath
5 Retrograde SmartTouch 60 36 3.2 23.1 13.5 Rapid RV aortic D curve pacing
Table 8.43: Procedural details
Retrograde aortic access was the preferred access route in all patients, the attempt at transseptal access in patient 3 was unsuccessful due to unstable catheter position. SAM-septal contact areas are on table and under GA, patient 2 had SAM-septal contact on exercise testing pre-procedure.
8.3.4.1 Difference to previous RF ablation in HOCM techniques
Previous series have reported the use of endocardial RF ablation of the septum to treat LVOT obstruction in HCM. These techniques have usually involved the use of TOE to provide imaging guidance and have not incorporated any echocardiographic images on to the EP mapping systems such as CARTO. The use of the CARTOsound mapping has the ability to provide greater detail of the true target of SAM-septal contact. The EP catheters used were similar to previous procedures and the amount of energy used was comparable.
348
349
8.4 Results
8.4.1 Procedural complications
Patient 4 underwent successful RF ablation but developed a significant retroperitoneal haemorrhage and cardiovascular collapse following sheath removal. An Agilus steerable sheath was used to manipulate the ablation catheter in the LV – this was an 8.5Fr catheter inserted via a 10Fr sheath.
Manual pressure was applied for 30 minutes after the reversal of heparin. The patient deteriorated approximately 60 minutes after sheath removal. An echocardiogram showed no pericardial effusion with LVOT gradients of 80mmHg (similar to pre-procedure). Urgent surgical repair of the right femoral artery was initially successful but a secondary bleed within 24 hours lead to recurrent hypotension and subsequent mesenteric ischaemia. This caused multiorgan failure and the patient died 3 days after the procedure.
One patient developed a paradoxical increase in LVOT gradient immediately following ablation resulting in pulmonary oedema. Reintubation, intravenous dexamethasone and right ventricular apical pacing were employed to reduce the gradient. She recovered well and was discharged. As the left bundle branch was seen to pass through the SAM-septal contact area in this patient, peri- procedural left bundle branch block with subsequent late complete heart block at 6 months post procedure was also seen. She showed significant improvement symptomatically and haemodynamically at 6 month assessment, with clear hypokinesia of the basal septal myocardium.
The course of the left bundle branch also fell within the SAM-septal contact area in one further patient. LBBB was induced on LV conduction tissue mapping (and marked on CARTO shell) and did
350 not recover before ablation was eventually delivered in this area. LBBB persisted to latest follow up with no evidence of AV block on ambulatory ECG monitoring.
8.4.2 Symptomatic resolution
All four surviving patients reported an improvement in symptoms at 6 months. Three patients had improved from NYHA class 3 to 2 and one patient had improved from NYHA class 3 to 1. Chest pain was present in 2 patients prior to RF ablation, this resolved in both.
Parameter Pre-RF Post-RF NYHA class 3 1.75 Rest LVOT gradient (mmHg) 64.2 (±50.6) 12.5 (±2.5) Provoked LVOT gradient (mmHg) 93.5 (±30.8) 23.2 (±8.3) Presence of severe SAM 4/5 0/4 IVSd (mm) 18.2 (±1.9) 16.7 (±2.5) Left atrial diameter (mm) 48.8 (±6.5) 44.8 (±8.3)
Peak VO2 (mL/min/Kg) 15.4 (±2.2) 16.5 (±5.2) Exercise time (secs) 558 (±130) 730 (±64) EQ5D-5L health score (0-100) 44 (±18.9) 70 (±3.5)
Table 8.44: Measured parameters pre- and post-RF ablation
8.4.3 Echocardiographic parameters
An echocardiogram was performed at 6 months in all surviving patients. Average peak resting gradients improved from 64.25 (±50.60)mmHg to 12.25 (±2.50) mmHg (Figure 8.5). Valsalva or
351 exercise induced gradient improved from 93.50 (±30.88)mmHg to 23.25 (±8.30)mmHg. Visual estimation of SAM improved in all patients. All patients demonstrating rest or exercise SAM septal contact did not demonstrate septal contact post RF ablation. Basal septal diameter in diastole reduced from 18.25 (±1.89)mm to 16.75 (±2.50)mm. Basal septal hypokinesia was seen in all. LA size reduced from 48.75 (±6.50)mm to 44.75 (±8.30)mm. LV diameter did not change in diastole (47.25
±8.30 to 47.25 ±8.50mm) or systole (30.75 ±4.86 to 29.5 ±4.51mm). The grade of SAM associated
MR reduced in two patients, one from moderate to mild, one from mild to none. Two patients with mild MR did not change post procedure.
352
Figure 8.107: LVOT gradient and exercise capacity change post RF ablation
Pre- and post-RF ablation resting LVOT gradient, provoked LVOT gradient, and %predicted peak VO2.
Solid lines represent a mean value of the 4 surviving patients. The dotted lines represent individual patient values.
353
8.4.4 Exercise capacity
Cardiopulmonary exercise test was performed prior to and 6 months after ablation in 3 patients.
Total exercise time increased from 558 (±129.87) to 730 (±63.64) secs. Pre-procedural peak VO2 measured 15.48 (±2.27) mL/min/Kg, this improved to 16.53 (±5.16) mL/min/Kg.
8.4.5 Quality of life questionnaire
EQ5D-5L quality of life index value increased from 0.57 (±0.17) to 0.65 (±0.18). Health score improved from 44 (±18.93) to 70 (±3.54).
8.4.6 CMR imaging
Two patients had non-MRI safe cardiac rhythm management devices in situ that precluded CMR imaging. Late gadolinium enhancement (LGE) could be seen in the basal septum in both patients who underwent CMR scanning 6 months after the procedure (see Figure 8.6). LGE was seen up to a maximum depth of 8mm from the LV endocardial surface. This was similar to the depth of tissue oedema visible on ICE during the procedure. Scar measured 6.3 and 2.2g respectively, representing
2.4 and 1.1% of total LV mass. LV mass prior to ablation measured 198 and 259g respectively. This reduced to 160 and 236g, representing a 19 and 9% reduction in total LV mass.
354
Figure 8.108: CMR imaging post RF ablation
Panel A shows a CMR 3 chamber image taken at day 2 following ablation. Microvascular obstruction (MVO) is highlighted in the basal septum on this early gadolinium enhanced image. Panel B shows a 6-month 3-chamber late gadolinium enhancement (LGE) image with LGE in the basal septum. Panels C and D are short axis LGE images one cut (8mm) below the LVOT, at the level of the body of the MV. LGE can be seen in the LV endocardium in the target area.
355
8.5 Discussion
There remain a small proportion of HOCM patients that cannot receive septal reduction with the traditional methods of ASA or surgical myectomy. Some of the technical limitations to ASA are difficult to overcome. Individual centres will vary in their availability of septal reduction techniques, and in North America surgical myectomy remaining the gold standard when performed in high volume centres 61. Outcomes from myectomy in high volume centres are excellent with good procedural success and low mortality. This is not reflected in centres with a low volume of patients63
160. Myectomy is major surgery and carries risks, especially in those with co-morbidities. Patients will often choose percutaneous solutions if they are on offer35. Patients who do not respond to medications and cannot have ASA or myectomy therefore need an alternative method of septal reduction. RF ablation can deliver effective and accurate infarction to the target septum. This method was effective in reducing gradients and improving symptoms in this preliminary group of patients who could not be treated with traditional methods.
Most ASA series present outcome data as mean values, this can hide individual failure. In series with individualised results up to a third of patients have an unsatisfactory outcome as measured by symptom and LVOT gradient parameters 161. These results are in part due to the constraints of septal arterial anatomy. If alcohol induced infarction is not at the SAM-septal contact area, LV haemodynamics will not change significantly111. In addition to those with inaccurate location of infarct and poor outcome we must consider the 5-15% of patients who cannot receive alcohol due to an inability to locate an appropriate septal vessel88. There would be a clear advantage of a percutaneous septal reduction technique that afforded accurate targeting of myocardial damage, independent of arterial anatomy. If RF ablation were to be pursued and proven to be effective in a
356 larger population of HOCM patients it is possible that it could be considered alongside more traditional methods of septal reduction.
Radiofrequency septal ablation in HOCM has been shown to be feasible 81 82, however, the novel use of CARTOSound® technology in our patients defines the ablation target with previously unparalleled accuracy. The live ICE image and ability to incorporate this information into the electroanatomic mapping system displays the SAM-septal contact target clearly. Although the volume of tissue damage appears small compared to other forms of septal reduction, the accuracy seems to be sufficient to affect SAM and thus reduce LVOT gradients.
The required size of infarct to produce symptom resolution in non-surgical septal reduction is not known. CMR studies have highlighted an alcohol induced scar of up to 25g post ASA, with a total LV mass reduction of 14-16g111. The average size of infarcted myocardium seen in our observational study in Chapter 3 was 8.9g. As CMR was only possible in two patients following treatment we cannot make secure conclusions about the appearance of scar, but there are signals that the damage delivered is smaller than that in ASA. The scar size seen in these two patients measured just 4.2g.
This was re-iterated by the modest reduction in septal width measured by echocardiography, a
1.5mm reduction is less marked than the change reported in our ASA series (5mm Chapter 2 and
4mm in Chapter 6). This is similar to results from other series suggesting reduction is septal size of 5-
6mm55. This modest reduction in septal width could perhaps lend itself well to those with less marked hypertrophy who are thought to be at risk of ventricular septal defect with traditional methods. A small amount of damage accurately delivered can interrupt the SAM-septal feedback mechanism and reduce LVOT gradients effectively. The location of tissue damage seems to be at least as important as the size.
357
8.5.1 Procedural risk
This small series highlights our early experience of a novel technique for septal reduction in a highly symptomatic population with advanced, obstructive cardiac failure. Failed trial of medications, alcohol ablation and unsuitability for myectomy leaves this cohort with no other traditional therapeutic option for reduction of LVOT gradients. The fragility of patients with advanced obstructive cardiac failure is evidenced by both the early death of our patient who was deemed too frail for a RF septal reduction and by our patient who died following a retroperitoneal haemorrhage.
Whilst retroperitoneal haemorrhage is a recognised risk in any percutaneous procedure162, it is likely to be poorly tolerated in these patients due to inability to cope with the acute preload reduction.
The intolerance of reduced preload is inherent in any obstructive HCM patient but may be worsened transiently following direct radiofrequency ablation of the SAM septal contact area due to tissue oedema. Myocardial oedema is well recognised in CMR studies following RFA in other procedures163.
The mechanism of cell death in RF ablation is not the same as that observed in ASA, with direct toxicity and necrosis of myocardium and coronary artery necrosis thought to be responsible99. The cell injury created by the direct thermal injury associated with RFA is thought to be mediated by injury to the sarcolemmal membrane and subsequent calcium overload. The different processes involved in RFA and ASA may explain the disproportionate tissue oedema, as the total cell necrosis by late gadolinium enhancement CMR scanning at 6 months is substantially less in RFA.
In these patients, despite successful ablation of the SAM septal contact area, we experienced no acute change in LVOT gradients when assessed invasively at the end of the procedure. The patient
358 with a paradoxical increase in LVOT gradient had most likely generated greater tissue oedema over the narrowest region of the LVOT. This was probably more apparent as a result of improving accuracy, catheter stability and contact as we refined the procedure based on evolving learning experience (she was Patient 5).
Paradoxical increase in LVOT gradients was also observed in the paediatric HOCM population treated with RF ablation (with fatal consequence) and potentially stands to be a challenge in the adoption of this technique into the mainstream 82. Peri-procedural intravenous dexamethasone has now been employed in our patients as empiric therapy to reduce this tissue oedema effect but subsequent numbers are insufficient to assess its efficacy.
Although such a small preliminary study cannot accurately assess morbidity and mortality of a novel technique, recent analyses of complications during radiofrequency ablation for ischaemic VT found a procedure-related death rate of between 0.4-3.0%, some relating to retroperitoneal haemorrhage164. This is probably the closest representation of procedural risk in our patients due to the similarities in LV based ablation techniques.
The ability of CARTO to identify specialised conduction tissue enables the operator to deliver tissue damage to the target area with the knowledge of the location of specialised conduction tissue.
While it was hoped that the technical accuracy of the CARTO map would prevent this collateral damage to this tissue, there are clearly cases where the left bundle branch courses directly through the target area of SAM-septal contact. Whilst it is accepted the development of LBBB is generally undesirable, it is a common finding after surgical myectomy. Surgical myectomy has been suggested
359 to actually improve patient outcomes, suggesting that LBBB in this context has no detrimental effect61. In order to deliver effective treatment to the SAM-septal contact area we may be forced to accept LBBB as an outcome from the procedure. The change in LV conduction and hence contraction may be relevant when considering the mechanism of resolution of gradient in 2 patients, one had
LBBB and one was RV paced at 6 month follow up. There was however clear akinesia of the basal septum on echocardiogram suggesting myocardial damage and structural change rather than the dyskinetic contraction seen in LBBB. In addition, if rapid RV apical pacing is used to stabilise catheter movement then real time electrogram visualization of Purkinje potentials is forfeited, and the operator must rely solely on the CARTO-mapped specialized conduction tree.
360
8.6 Conclusion
Radiofrequency ablation to treat LVOT gradients in HOCM using CARTOSound® shows considerable promise. The unprecedented accuracy of tissue damage interrupts the SAM-septal contact cycle and results in improvements in LVOT gradients, symptoms and quality of life in this preliminary group of patients. Further experience of this technique is required before it can be considered equivalent to standard septal reduction techniques.
361
8.7 Limitations
This is a preliminary study to explore the use of combined ICE and CARTO technology to guide RF ablation to treat LVOT gradients in HOCM. The study population was therefore small. The ability to make secure conclusions about outcome is obviously restricted by the small numbers. The follow up period was also short.
The relatively high rate of complication in this exploratory group must be noted. It is generally accepted that risks of any new procedure are elevated when learning new techniques. However lessons must be learned and procedures adapted if these technique is to be incorporated into standard clinical care.
The images used to create the CARTO shell were taken in diastole. This is the traditional method employed in electrophysiology ablation procedures. The target area for the RF energy delivery is the
SAM-septal contact area, which by definition only occurs in systole. The SAM-septal contact map was therefore acquired in a different phase of the cardiac cycle. This may have led to subtle differences in interpretation of the position of the RF catheter and subsequent energy delivery.
Although these are restrictions to the long term application of CARTOsound guided RF ablation the aim of this study was to implement and describe a new technology. Further enrolment and follow up of appropriate patients will continue.
362
Chapter 9: Conclusions
Obstruction in hypertrophic cardiomyopathy is associated with increased morbidity and mortality.
ASA using traditional methods has an unacceptably high rate of failure to resolve LVOT gradient. In our population I observed that failure to treat LVOT gradient satisfactorily was seen in 41% after one procedure and 18% after multiple procedures.
This failure rate can be partly explained by inaccurate location of the iatrogenic infarction. Using a
CMR study I was able to show that a greater proportion of the target myocardium was occupied by alcohol induced scar in those that had successful resolution of LVOT gradient versus those that had a persisting obstruction.
Intracardiac echocardiography (ICE) provides excellent image quality of the contact point of the mitral valve on the septum in HOCM, but is no better than standard transthoracic echocardiography in describing detail of the septum or other cardiac structures relevant to ASA. ICE cannot see myocardial contrast well and therefore cannot be used to guide ASA alone.
Computed tomography (CT) angiography can visualise small septal arteries. Most septal arteries provide supply to more than one territory, often a vessel will perfuse both left and right ventricular septum. As these are different environments with different pressures this has implications for run off of any fluid injected, this will tend to follow a path towards the lower pressure right ventricular cavity and alcohol will be lost to the systemic circulation rather than directed at target myocardium in the left ventricular septum. The ability to merge angiographic images with structural detail allows description of the path of arteries to guide alcohol injection in ASA. The use of CT planning improved the success rate of ASA after one procedure from 59% to 85%. We observed less RBBB (13% vs 62%) due to improved targeting of the LV septum, confirmed by CMR.
363
Patient selection in ASA is important. A standard operating procedure for assessment and treatment of HOCM patients is now part of routine clinical care.
Some patients cannot receive trans-coronary alcohol due to arterial anatomical restrictions. Direct endocardial radiofrequency ablation of the interventricular septum with merged ICE/CARTO electrophysiology mapping system guidance was explored. This effectively treated LVOT obstruction in 4/5 patients, with correlating symptom relief.
364
References
1. Elliott P, McKenna WJ. Hypertrophic cardiomyopathy. Lancet 2004;363(9424):1881-91. 2. Maron MS, Olivotto I, Betocchi S, et al. Effect of left ventricular outflow tract obstruction on clinical outcome in hypertrophic cardiomyopathy. NEnglJMed 2003;348(4):295-303. 3. Maron MS, Olivotto I, Zenovich AG, et al. Hypertrophic cardiomyopathy is predominantly a disease of left ventricular outflow tract obstruction. Circulation 2006;114(21):2232-39. 4. Woo A, Williams WG, Choi R, et al. Clinical and echocardiographic determinants of long-term survival after surgical myectomy in obstructive hypertrophic cardiomyopathy. Circulation 2005;111(16):2033-41. 5. Nagueh SF, Ommen SR, Lakkis NM, et al. Comparison of ethanol septal reduction therapy with surgical myectomy for the treatment of hypertrophic obstructive cardiomyopathy. JAmCollCardiol 2001;38(6):1701-06. 6. Qin JX, Shiota T, Lever HM, et al. Outcome of patients with hypertrophic obstructive cardiomyopathy after percutaneous transluminal septal myocardial ablation and septal myectomy surgery. JAmCollCardiol 2001;38(7):1994-2000. 7. Come PC, Riley MF. Hypertrophic cardiomyopathy. Disappearance of auscultatory, carotid pulse, and echocardiographic manifestations of obstruction following myocardial infarction. Chest 1982;82(4):451-54. 8. Sigwart U, M G, M P, et al. Ventricular wall motion during balloon occlusion. In: Sigwart U, PH H, eds. Ventricular wall motionNew York (NY), 1983:206-10. 9. Brugada P, de SH, Smeets JL, et al. Transcoronary chemical ablation of ventricular tachycardia. Circulation 1989;79(3):475-82. 10. Sigwart U. Non-surgical myocardial reduction for hypertrophic obstructive cardiomyopathy. Lancet 1995;346(8969):211-14. 11. Vecht JA, Dave R, Vecht RJ. Alcohol septal ablation: The first patient in 1994. British Journal of Cardiology 2006;13(1):62-64. 12. Elliott PM, Anastasakis A, Borger MA, et al. 2014 ESC Guidelines on diagnosis and management of hypertrophic cardiomyopathy: The Task Force for the Diagnosis and Management of Hypertrophic Cardiomyopathy of the European Society of Cardiology (ESC). Eur Heart J 2014. 13. Gersh BJ, Maron BJ, Bonow RO, et al. 2011 ACCF/AHA Guideline for the Diagnosis and Treatment of Hypertrophic Cardiomyopathy: Executive Summary: A Report of the American College of Cardiology Foundation/American Heart Association Task Force on Practice Guidelines. Circulation 2011. 14. Ludman P. British Cardiovascular Intervention Society Audit returns 2011, 2012:251. 15. Ludman P. British Cardiovascular Intervention Society Audit returns - Adult intervention 2014. 2015. 16. Sherrid MV, Chu CK, Delia E, et al. An echocardiographic study of the fluid mechanics of obstruction in hypertrophic cardiomyopathy. JAmCollCardiol 1993;22(3):816-25. 17. Levine RA, Schwammenthal E, Song JK. Diastolic leading to systolic anterior motion: new technology reveals physiology. J Am Coll Cardiol 2014;64(19):1996-9. 18. Maron BJ, Maron MS, Wigle ED, et al. The 50-year history, controversy, and clinical implications of left ventricular outflow tract obstruction in hypertrophic cardiomyopathy from idiopathic hypertrophic subaortic stenosis to hypertrophic cardiomyopathy: from idiopathic hypertrophic subaortic stenosis to hypertrophic cardiomyopathy. JAmCollCardiol 2009;54(3):191-200. 19. Maron BJ, Nishimura RA, Danielson GK. Pitfalls in clinical recognition and a novel operative approach for hypertrophic cardiomyopathy with severe outflow obstruction due to anomalous papillary muscle. Circulation 1998;98(23):2505-08.
365
20. Ommen SR, Nishimura RA. What causes outflow tract obstruction in hypertrophic cardiomyopathy? Heart 2009;95(21):1725-26. 21. Mazur W, Nagueh SF, Lakkis NM, et al. Regression of left ventricular hypertrophy after nonsurgical septal reduction therapy for hypertrophic obstructive cardiomyopathy. Circulation 2001;103(11):1492-96. 22. Lakkis N, Plana JC, Nagueh S, et al. Efficacy of nonsurgical septal reduction therapy in symptomatic patients with obstructive hypertrophic cardiomyopathy and provocable gradients. AmJCardiol 2001;88(5):583-86. 23. Maron BJ, McKenna WJ, Danielson GK, et al. American College of Cardiology/European Society of Cardiology clinical expert consensus document on hypertrophic cardiomyopathy. A report of the American College of Cardiology Foundation Task Force on Clinical Expert Consensus Documents and the European Society of Cardiology Committee for Practice Guidelines. JAmCollCardiol 2003;42(9):1687-713. 24. Parakh N, Bhargava B. Golden jubilee of hypertrophic cardiomyopathy: is alcohol septal ablation the gold standard? CardiovascRevascMed 2009;10(3):172-78. 25. Seggewiss H. Medical therapy versus interventional therapy in hypertropic obstructive cardiomyopathy. CurrControl Trials CardiovascMed 2000;1(2):115-19. 26. Pellikka PA, Oh JK, Bailey KR, et al. Dynamic intraventricular obstruction during dobutamine stress echocardiography. A new observation. Circulation 1992;86(5):1429-32. 27. Gietzen FH, Leuner CJ, Obergassel L, et al. Role of transcoronary ablation of septal hypertrophy in patients with hypertrophic cardiomyopathy, New York Heart Association functional class III or IV, and outflow obstruction only under provocable conditions. Circulation 2002;106(4):454-59. 28. Holmes DR, Jr., Valeti US, Nishimura RA. Alcohol septal ablation for hypertrophic cardiomyopathy: indications and technique. CatheterCardiovascInterv 2005;66(3):375-89. 29. Rigopoulos AG, Seggewiss H. A decade of percutaneous septal ablation in hypertrophic cardiomyopathy. CircJ 2010;75(1):28-37. 30. Non-surgical reduction of the myocardial septum: NICE interventional procedure guidance, 2004. 31. van der Lee C, Scholzel B, ten Berg JM, et al. Usefulness of clinical, echocardiographic, and procedural characteristics to predict outcome after percutaneous transluminal septal myocardial ablation. AmJCardiol 2008;101(9):1315-20. 32. Cuoco FA, Spencer WH, III, Fernandes VL, et al. Implantable cardioverter-defibrillator therapy for primary prevention of sudden death after alcohol septal ablation of hypertrophic cardiomyopathy. JAmCollCardiol 2008;52(21):1718-23. 33. Lawrenz T, Obergassel L, Lieder F, et al. Transcoronary ablation of septal hypertrophy does not alter ICD intervention rates in high risk patients with hypertrophic obstructive cardiomyopathy. Pacing ClinElectrophysiol 2005;28(4):295-300. 34. Noseworthy PA, Rosenberg MA, Fifer MA, et al. Ventricular arrhythmia following alcohol septal ablation for obstructive hypertrophic cardiomyopathy. AmJCardiol 2009;104(1):128-32. 35. Fifer MA. Controversies in cardiovascular medicine. Most fully informed patients choose septal ablation over septal myectomy. Circulation 2007;116(2):207-16. 36. Rothman RD, Baggish AL, O'Callaghan C, et al. Management strategy in 249 consecutive patients with obstructive hypertrophic cardiomyopathy referred to a dedicated program. AmJCardiol 2012;110(8):1169-74. 37. Cuisset T, Franceschi F, Prevot S, et al. Transradial approach and subclavian wired temporary pacemaker to increase safety of alcohol septal ablation for treatment of obstructive hypertrophic cardiomyopathy: the TRASA trial. ArchCardiovascDis 2011;104(8-9):444-49. 38. Fernandes VL, Nielsen C, Nagueh SF, et al. Follow-up of alcohol septal ablation for symptomatic hypertrophic obstructive cardiomyopathy the Baylor and Medical University of South Carolina experience 1996 to 2007. JACCCardiovascInterv 2008;1(5):561-70.
366
39. Gietzen FH, Leuner CJ, Raute-Kreinsen U, et al. Acute and long-term results after transcoronary ablation of septal hypertrophy (TASH). Catheter interventional treatment for hypertrophic obstructive cardiomyopathy. EurHeart J 1999;20(18):1342-54. 40. Veselka J, Duchonova R, Palenickova J, et al. Impact of ethanol dosing on the long-term outcome of alcohol septal ablation for obstructive hypertrophic cardiomyopathy: a single-center prospective, and randomized study. CircJ 2006;70(12):1550-52. 41. Veselka J, Tomasov P, Zemanek D. Long-term effects of varying alcohol dosing in percutaneous septal ablation for obstructive hypertrophic cardiomyopathy: a randomized study with a follow-up up to 11 years. CanJCardiol 2011;27(6):763-67. 42. Nagueh SF, Lakkis NM, He ZX, et al. Role of myocardial contrast echocardiography during nonsurgical septal reduction therapy for hypertrophic obstructive cardiomyopathy. JAmCollCardiol 1998;32(1):225-29. 43. Faber L, Seggewiss H, Fassbender D, et al. [Percutaneous transluminal septal myocardial ablation in hypertrophic obstructive cardiomyopathy: acute results in 66 patients with reference to myocardial contrast echocardiography]. ZKardiol 1998;87(3):191-201. 44. Veselka J, Jensen MK, Liebregts M, et al. Long-term clinical outcome after alcohol septal ablation for obstructive hypertrophic cardiomyopathy: results from the Euro-ASA registry. Eur Heart J 2016;37(19):1517-23. 45. Jensen MK, C P, Horstkotte D, et al. Personal correspondence with Dr M Jensen, Rigshospitalet, Copenhagen: Alcohol septal ablation in patients with hypertrophic obstructive cardiomyopathy: Low incidence of sudden cardiac death and reduced risk profile., 2012. 46. Sorajja P, Ommen SR, Holmes DR, Jr., et al. Survival after alcohol septal ablation for obstructive hypertrophic cardiomyopathy. Circulation 2012;126(20):2374-80. 47. Veselka J, Tomašov P, Januška J, et al. Obstruction after alcohol septal ablation is associated with cardiovascular mortality events. Heart 2016;102(22):1793-96. 48. Ball W, Ivanov J, Rakowski H, et al. Long-term survival in patients with resting obstructive hypertrophic cardiomyopathy comparison of conservative versus invasive treatment. JAmCollCardiol 2011;58(22):2313-21. 49. Klopotowski M, Chojnowska L, Malek LA, et al. The risk of non-sustained ventricular tachycardia after percutaneous alcohol septal ablation in patients with hypertrophic obstructive cardiomyopathy. ClinResCardiol 2010;99(5):285-92. 50. Maron BJ, Spirito P, Shen WK, et al. Implantable cardioverter-defibrillators and prevention of sudden cardiac death in hypertrophic cardiomyopathy. JAMA 2007;298(4):405-12. 51. Agarwal S, Tuzcu EM, Desai MY, et al. Updated meta-analysis of septal alcohol ablation versus myectomy for hypertrophic cardiomyopathy. JAmCollCardiol 2010;55(8):823-34. 52. Leonardi RA, Kransdorf EP, Simel DL, et al. Meta-analyses of septal reduction therapies for obstructive hypertrophic cardiomyopathy: comparative rates of overall mortality and sudden cardiac death after treatment. CircCardiovascInterv 2010;3(2):97-104. 53. Knight C, Kurbaan AS, Seggewiss H, et al. Nonsurgical septal reduction for hypertrophic obstructive cardiomyopathy: outcome in the first series of patients. Circulation 1997;95(8):2075-81. 54. Alam M, Dokainish H, Lakkis N. Alcohol septal ablation for hypertrophic obstructive cardiomyopathy: a systematic review of published studies. JIntervCardiol 2006;19(4):319-27. 55. Nagueh SF, Groves BM, Schwartz L, et al. Alcohol septal ablation for the treatment of hypertrophic obstructive cardiomyopathy a multicenter north american registry. JAmCollCardiol 2011;58(22):2322-28. 56. Jensen MK, Almaas VM, Jacobsson L, et al. Long-term outcome of percutaneous transluminal septal myocardial ablation in hypertrophic obstructive cardiomyopathy: a Scandinavian multicenter study. CircCardiovascInterv 2011;4(3):256-65. 57. Seggewiss H, Rigopoulos A, Welge D, et al. Long-term follow-up after percutaneous septal ablation in hypertrophic obstructive cardiomyopathy. ClinResCardiol 2007;96(12):856-63.
367
58. Kwon DH, Kapadia SR, Tuzcu EM, et al. Long-term outcomes in high-risk symptomatic patients with hypertrophic cardiomyopathy undergoing alcohol septal ablation. JACCCardiovascInterv 2008;1(4):432-38. 59. Lyne JC, Kilpatrick T, Duncan A, et al. Long-term follow-up of the first patients to undergo transcatheter alcohol septal ablation. Cardiology 2010;116(3):168-73. 60. Jensen MK, Havndrup O, Hassager C, et al. Survival and sudden cardiac death after septal ablation for hypertrophic obstructive cardiomyopathy. ScandCardiovascJ 2011;45(3):153-60. 61. Dearani JA, Ommen SR, Gersh BJ, et al. Surgery insight: Septal myectomy for obstructive hypertrophic cardiomyopathy--the Mayo Clinic experience. NatClinPractCardiovascMed 2007;4(9):503-12. 62. Maron BJ, Yacoub M, Dearani JA. Controversies in cardiovascular medicine. Benefits of surgery in obstructive hypertrophic cardiomyopathy: bring septal myectomy back for European patients. EurHeart J 2011;32(9):1055-58. 63. Naidu SS, Y. W. Comparative In Hospital Safety and Cost of Alcohol Septal Ablation versus Isolated Surgical Myectomy for Symptomatic Hypertrophic Obstructive Cardiomyopathy: Insights from the Multi Center PREMIER Database. TCT Oral abtracts session 2011 2012. 64. Iacovoni A, Spirito P, Simon C, et al. A contemporary European experience with surgical septal myectomy in hypertrophic cardiomyopathy. EurHeart J 2012;33(16):2080-87. 65. Alam M, Dokainish H, Lakkis NM. Hypertrophic obstructive cardiomyopathy-alcohol septal ablation vs. myectomy: a meta-analysis. EurHeart J 2009;30(9):1080-87. 66. Olivotto I, Ommen SR, Maron MS, et al. Surgical myectomy versus alcohol septal ablation for obstructive hypertrophic cardiomyopathy. Will there ever be a randomized trial? JAmCollCardiol 2007;50(9):831-34. 67. Chang SM, Lakkis NM, Franklin J, et al. Predictors of outcome after alcohol septal ablation therapy in patients with hypertrophic obstructive cardiomyopathy. Circulation 2004;109(7):824-27. 68. Sorajja P, Binder J, Nishimura RA, et al. Predictors of an optimal clinical outcome with alcohol septal ablation for obstructive hypertrophic cardiomyopathy. CatheterCardiovascInterv 2012. 69. Harris KM, Spirito P, Maron MS, et al. Prevalence, clinical profile, and significance of left ventricular remodeling in the end-stage phase of hypertrophic cardiomyopathy. Circulation 2006;114(3):216-25. 70. Faber L, Welge D, Fassbender D, et al. One-year follow-up of percutaneous septal ablation for symptomatic hypertrophic obstructive cardiomyopathy in 312 patients: predictors of hemodynamic and clinical response. ClinResCardiol 2007;96(12):864-73. 71. Chen AA, Palacios IF, Mela T, et al. Acute predictors of subacute complete heart block after alcohol septal ablation for obstructive hypertrophic cardiomyopathy. AmJCardiol 2006;97(2):264-69. 72. Veselka J, Krejci J, Tomasov P, et al. Outcome of patients after alcohol septal ablation with permanent pacemaker implanted for periprocedural complete heart block. Int J Cardiol 2014;171(2):e37-8. 73. Wykrzykowska JJ, Kwaku K, Wylie J, et al. Delayed occurrence of unheralded phase IV complete heart block after ethanol septal ablation for symmetric hypertrophic obstructive cardiomyopathy. Pacing Clin Electrophysiol 2006;29(6):674-8. 74. Poyet R, Quilici J, Cuisset T, et al. Delayed occurrence of complete heart block after ethanol septal ablation for hypertrophic obstructive cardiomyopathy. Int J Cardiol 2011;147(2):e32- 4. 75. Matos GF, Hammadeh R, Francois C, et al. Controlled myocardial infarction induced by intracoronary injection of n-butyl cyanoacrylatein dogs: a feasibility study. CatheterCardiovascInterv 2005;66(2):244-53.
368
76. Oto A, Aytemir K, Okutucu S, et al. Cyanoacrylate for septal ablation in hypertrophic cardiomyopathy. JIntervCardiol 2011;24(1):77-84. 77. Gross CM, Schulz-Menger J, Kramer J, et al. Percutaneous transluminal septal artery ablation using polyvinyl alcohol foam particles for septal hypertrophy in patients with hypertrophic obstructive cardiomyopathy: acute and 3-year outcomes. JEndovascTher 2004;11(6):705-11. 78. Lanzino G, Kanaan Y, Perrini P, et al. Emerging concepts in the treatment of intracranial aneurysms: stents, coated coils, and liquid embolic agents. Neurosurgery 2005;57(3):449-59. 79. Durand E, Mousseaux E, Coste P, et al. Non-surgical septal myocardial reduction by coil embolization for hypertrophic obstructive cardiomyopathy: early and 6 months follow-up. EurHeart J 2008;29(3):348-55. 80. Kuhn H, Lawrenz T, Lieder F, et al. Survival after transcoronary ablation of septal hypertrophy in hypertrophic obstructive cardiomyopathy (TASH): a 10 year experience. ClinResCardiol 2008;97(4):234-43. 81. Lawrenz T, Borchert B, Leuner C, et al. Endocardial radiofrequency ablation for hypertrophic obstructive cardiomyopathy: acute results and 6 months' follow-up in 19 patients. JAmCollCardiol 2011;57(5):572-76. 82. Sreeram N, Emmel M, de Giovanni JV. Percutaneous radiofrequency septal reduction for hypertrophic obstructive cardiomyopathy in children. JAmCollCardiol 2011;58(24):2501-10. 83. Kimmelstiel CD, Maron BJ. Role of percutaneous septal ablation in hypertrophic obstructive cardiomyopathy. Circulation 2004;109(4):452-56. 84. Yacoub MH. Surgical versus alcohol septal ablation for hypertrophic obstructive cardiomyopathy: the pendulum swings. Circulation 2005;112(4):450-52. 85. Hess OM, Sigwart U. New treatment strategies for hypertrophic obstructive cardiomyopathy: alcohol ablation of the septum: the new gold standard? JAmCollCardiol 2004;44(10):2054- 55. 86. Jensen MK, Prinz C, Horstkotte D, et al. Alcohol septal ablation in patients with hypertrophic obstructive cardiomyopathy: low incidence of sudden cardiac death and reduced risk profile. Heart 2013;99(14):1012-17. 87. Pedone C, Vijayakumar M, Ligthart JM, et al. Intracardiac echocardiography guidance during percutaneous transluminal septal myocardial ablation in patients with obstructive hypertrophic cardiomyopathy. IntJCardiovascIntervent 2005;7(3):134-37. 88. Chan W, Williams L, Kotowycz MA, et al. Angiographic and echocardiographic correlates of suitable septal perforators for alcohol septal ablation in hypertrophic obstructive cardiomyopathy. Can J Cardiol 2014;30(8):912-9. 89. Miller MF. Search and find ... a PC database. J ahima 1991;62(12):66-73. 90. ten Cate FJ, Soliman OI, Michels M, et al. Long-term outcome of alcohol septal ablation in patients with obstructive hypertrophic cardiomyopathy: a word of caution. CircHeart Fail 2010;3(3):362-69. 91. Nagueh SF, Lakkis NM, Middleton KJ, et al. Changes in left ventricular diastolic function 6 months after nonsurgical septal reduction therapy for hypertrophic obstructive cardiomyopathy. Circulation 1999;99(3):344-47. 92. Amano Y, Takayama M, Kumita S, et al. MR imaging evaluation of regional, remote, and global effects of percutaneous transluminal septal myocardial ablation in hypertrophic obstructive cardiomyopathy. JComputAssistTomogr 2007;31(4):600-04. 93. Maekawa Y, Jinzaki M, Tsuruta H, et al. Multidetector computed tomography-guided percutaneous transluminal septal myocardial ablation in a Noonan syndrome patient with hypertrophic obstructive cardiomyopathy. Int J Cardiol 2014;172(1):e79-81. 94. Geske JB, Sorajja P, Ommen SR, et al. Variability of left ventricular outflow tract gradient during cardiac catheterization in patients with hypertrophic cardiomyopathy. JACC Cardiovasc Interv 2011;4(6):704-9.
369
95. Bois JP, Adams JC, Kumar G, et al. Relation Between Temperature Extremes and Symptom Exacerbation in Patients With Hypertrophic Cardiomyopathy. Am J Cardiol 2015. 96. Thomas N, Unsworth B, Ferenczi EA, et al. Intraobserver variability in grading severity of repeated identical cases of mitral regurgitation. Am Heart J 2008;156(6):1089-94. 97. Angelini P. The "1st septal unit" in hypertrophic obstructive cardiomyopathy: a newly recognized anatomo-functional entity, identified during recent alcohol septal ablation experience. TexHeart InstJ 2007;34(3):336-46. 98. van den Wijngaard JP, van Horssen P, ter Wee R, et al. Organization and collateralization of a subendocardial plexus in end-stage human heart failure. Am J Physiol Heart Circ Physiol 2010;298(1):H158-62. 99. Baggish AL, Smith RN, Palacios I, et al. Pathological effects of alcohol septal ablation for hypertrophic obstructive cardiomyopathy. Heart 2006;92(12):1773-8. 100. Flett AS, Hasleton J, Cook C, et al. Evaluation of techniques for the quantification of myocardial scar of differing etiology using cardiac magnetic resonance. JACC Cardiovasc Imaging 2011;4(2):150-6. 101. Hoey ET, Elassaly M, Ganeshan A, et al. The role of magnetic resonance imaging in hypertrophic cardiomyopathy. Quant Imaging Med Surg 2014;4(5):397-406. 102. Silva C, Moon JC, Elkington AG, et al. Myocardial late gadolinium enhancement in specific cardiomyopathies by cardiovascular magnetic resonance: a preliminary experience. J Cardiovasc Med (Hagerstown) 2007;8(12):1076-9. 103. O'Hanlon R, Pennell DJ. Cardiovascular magnetic resonance in the evaluation of hypertrophic and infiltrative cardiomyopathies. Heart Fail Clin 2009;5(3):369-87, vi. 104. Herzog B, Greenwood J, Plein S. European Society of Cardiology Cardiovascular Magnetic Resonance Working Group - CMR pocket guide, 2013. 105. Amado LC, Gerber BL, Gupta SN, et al. Accurate and objective infarct sizing by contrast- enhanced magnetic resonance imaging in a canine myocardial infarction model. J Am Coll Cardiol 2004;44(12):2383-9. 106. Kim RJ, Chen EL, Lima JA, et al. Myocardial Gd-DTPA kinetics determine MRI contrast enhancement and reflect the extent and severity of myocardial injury after acute reperfused infarction. Circulation 1996;94(12):3318-26. 107. de Roos A, Doornbos J, van der Wall EE, et al. MR imaging of acute myocardial infarction: value of Gd-DTPA. AJR Am J Roentgenol 1988;150(3):531-4. 108. Wagner A, Mahrholdt H, Holly TA, et al. Contrast-enhanced MRI and routine single photon emission computed tomography (SPECT) perfusion imaging for detection of subendocardial myocardial infarcts: an imaging study. Lancet 2003;361(9355):374-9. 109. van Dockum WG, Beek AM, ten Cate FJ, et al. Early onset and progression of left ventricular remodeling after alcohol septal ablation in hypertrophic obstructive cardiomyopathy. Circulation 2005;111(19):2503-08. 110. McCann GP, van Dockum WG, Beek AM, et al. Extent of myocardial infarction and reverse remodeling assessed by cardiac magnetic resonance in patients with and without right bundle branch block following alcohol septal ablation for obstructive hypertrophic cardiomyopathy. AmJCardiol 2007;99(4):563-67. 111. van Dockum WG, ten Cate FJ, ten Berg JM, et al. Myocardial infarction after percutaneous transluminal septal myocardial ablation in hypertrophic obstructive cardiomyopathy: evaluation by contrast-enhanced magnetic resonance imaging. JAmCollCardiol 2004;43(1):27-34. 112. Sohns C, Sossalla S, Schmitto JD, et al. Visualization of transcoronary ablation of septal hypertrophy in patients with hypertrophic obstructive cardiomyopathy: a comparison between cardiac MRI, invasive measurements and echocardiography. ClinResCardiol 2010;99(6):359-68.
370
113. Maron MS. Clinical utility of cardiovascular magnetic resonance in hypertrophic cardiomyopathy. J Cardiovasc Magn Reson 2012;14:13. 114. Jivraj N, Phinikaridou A, Shah AM, et al. Molecular imaging of myocardial infarction. Basic Res Cardiol 2014;109(1):397. 115. Hsu LY, Natanzon A, Kellman P, et al. Quantitative myocardial infarction on delayed enhancement MRI. Part I: Animal validation of an automated feature analysis and combined thresholding infarct sizing algorithm. J Magn Reson Imaging 2006;23(3):298-308. 116. McAlindon E, Pufulete M, Lawton C, et al. Quantification of infarct size and myocardium at risk: evaluation of different techniques and its implications. Eur Heart J Cardiovasc Imaging 2015;16(7):738-46. 117. Lisanti C, Carlin C, Banks KP, et al. Normal MRI appearance and motion-related phenomena of CSF. AJR Am J Roentgenol 2007;188(3):716-25. 118. Sievers B, Elliott MD, Hurwitz LM, et al. Rapid detection of myocardial infarction by subsecond, free-breathing delayed contrast-enhancement cardiovascular magnetic resonance. Circulation 2007;115(2):236-44. 119. Olivotto I, Maron MS, Autore C, et al. Assessment and significance of left ventricular mass by cardiovascular magnetic resonance in hypertrophic cardiomyopathy. J Am Coll Cardiol 2008;52(7):559-66. 120. Cerqueira MD, Weissman NJ, Dilsizian V, et al. Standardized myocardial segmentation and nomenclature for tomographic imaging of the heart. A statement for healthcare professionals from the Cardiac Imaging Committee of the Council on Clinical Cardiology of the American Heart Association. Circulation 2002;105(4):539-42. 121. Geske JB, Konecny T, Ommen SR, et al. Surgical myectomy improves pulmonary hypertension in obstructive hypertrophic cardiomyopathy. Eur Heart J 2014;35(30):2032-9. 122. Altarabsheh SE, Dearani JA, Burkhart HM, et al. Outcome of septal myectomy for obstructive hypertrophic cardiomyopathy in children and young adults. Ann Thorac Surg 2013;95(2):663- 9; discussion 69. 123. Orme NM, Sorajja P, Dearani JA, et al. Comparison of surgical septal myectomy to medical therapy alone in patients with hypertrophic cardiomyopathy and syncope. Am J Cardiol 2013;111(3):388-92. 124. Valeti US, Nishimura RA, Holmes DR, et al. Comparison of surgical septal myectomy and alcohol septal ablation with cardiac magnetic resonance imaging in patients with hypertrophic obstructive cardiomyopathy. JAmCollCardiol 2007;49(3):350-57. 125. de Leval MR, Kilner P, Gewillig M, et al. Total cavopulmonary connection: a logical alternative to atriopulmonary connection for complex Fontan operations. Experimental studies and early clinical experience. J Thorac Cardiovasc Surg 1988;96(5):682-95. 126. Axelsson A, Iversen K, Vejlstrup N, et al. Efficacy and safety of the angiotensin II receptor blocker losartan for hypertrophic cardiomyopathy: the INHERIT randomised, double-blind, placebo-controlled trial. Lancet Diabetes Endocrinol 2015;3(2):123-31. 127. Yuan J, Qiao S, Zhang Y, et al. Follow-up by cardiac magnetic resonance imaging in patients with hypertrophic cardiomyopathy who underwent percutaneous ventricular septal ablation. Am J Cardiol 2010;106(10):1487-91. 128. Ponnuthurai FA, van Gaal WJ, Burchell A, et al. Safety and feasibility of day case patent foramen ovale (PFO) closure facilitated by intracardiac echocardiography. IntJCardiol 2009;131(3):438-40. 129. MacDonald ST, Newton JD, Ormerod OJ. Intracardiac echocardiography off piste? Closure of the left atrial appendage using ICE and local anesthesia. CatheterCardiovascInterv 2011;77(1):124-27. 130. Rusu MC, Petrescu CI, Niculescu V, et al. Considerations on the left papillary muscles microangioarchitecture. Rom J Morphol Embryol 2006;47(1):63-6.
371
131. Wallace EL, Thompson JJ, Faulkner MW, et al. Septal perforator anatomy and variability of perfusion bed by myocardial contrast echocardiography: a study of hypertrophic cardiomyopathy patients undergoing alcohol septal ablation. J Interv Cardiol 2013;26(6):604- 12. 132. Gill Wharton (Lead Author) RSC, Jane Allen, Hollie Brewerton, Richard Jones, Prathap Kanagala, Guy Lloyd, Navroz Masani, Thomas Mathew, David Oxborough, Bushra Rana, Julie Sandoval, Richard Wheeler.
. Secondary
133. Reis RL, Bolton MR, King JF, et al. Anterion-superior displacement of papillary muscles producing obstruction and mitral regurgitation in idiopathic hypertrophic subaortic stenosis. Operative relief by posterior-superior realignment of papillary muscles following ventricular septal myectomy. Circulation 1974;50(2 Suppl):Ii181-8. 134. Kaul S. Myocardial contrast echocardiography: a 25-year retrospective. Circulation 2008;118(3):291-308. 135. Maron MS, Olivotto I, Harrigan C, et al. Mitral valve abnormalities identified by cardiovascular magnetic resonance represent a primary phenotypic expression of hypertrophic cardiomyopathy. Circulation 2011;124(1):40-47. 136. Chest pain of recent onset: Assessment and diagnosis of recent onset chest pain or discomfort of suspected cardiac origin: National Institute of Clinical Excellence, 2010. 137. Singh M, Edwards WD, Holmes DR, Jr., et al. Anatomy of the first septal perforating artery: a study with implications for ablation therapy for hypertrophic cardiomyopathy. Mayo ClinProc 2001;76(8):799-802. 138. Hosseinpour AR, Anderson RH, Ho SY. The anatomy of the septal perforating arteries in normal and congenitally malformed hearts. JThoracCardiovascSurg 2001;121(6):1046-52. 139. Choi D, Dardano J, Naidu SS. Alcohol septal ablation through an anomalous right coronary septal perforator: first report and discussion. J Invasive Cardiol 2009;21(6):e106-9. 140. Raphael CE, Cooper R, Parker KH, et al. Mechanisms of Myocardial Ischemia in Hypertrophic Cardiomyopathy: Insights From Wave Intensity Analysis and Magnetic Resonance. J Am Coll Cardiol 2016;68(15):1651-60. 141. Nagata Y, Konno T, Fujino N, et al. Right ventricular hypertrophy is associated with cardiovascular events in hypertrophic cardiomyopathy: evidence from study with magnetic resonance imaging. Can J Cardiol 2015;31(6):702-8. 142. Ansari A. Anatomy and clinical significance of ventricular Thebesian veins. Clin Anat 2001;14(2):102-10. 143. Gilbert BW, Pollick C, Adelman AG, et al. Hypertrophic cardiomyopathy: subclassification by m mode echocardiography. Am J Cardiol 1980;45(4):861-72. 144. Veselka J, Lawrenz T, Stellbrink C, et al. Early outcomes of alcohol septal ablation for hypertrophic obstructive cardiomyopathy: A European multicenter and multinational study. Catheter Cardiovasc Interv 2014;84(1):101-7. 145. Cooper RM, Shahzad A, McShane J, et al. Alcohol Septal Ablation for Hypertrophic Obstructive Cardiomyopathy: Safe and Apparently Efficacious But Does Reporting of Aggregate Outcomes Hide Less-Favorable Results, Experienced by a Substantial Proportion of Patients? J Invasive Cardiol 2015;27(7):301-8. 146. Penicka M, Gregor P, Kerekes R, et al. The effects of candesartan on left ventricular hypertrophy and function in nonobstructive hypertrophic cardiomyopathy: a pilot, randomized study. J Mol Diagn 2009;11(1):35-41. 147. Marian AJ. Experimental therapies in hypertrophic cardiomyopathy. J Cardiovasc Transl Res 2009;2(4):483-92.
372
148. Adelman AG, Shah PM, Gramiak R, et al. Long-term propranolol therapy in muscular subaortic stenosis. Br Heart J 1970;32(6):804-11. 149. Stenson RE, Flamm MD, Jr., Harrison DC, et al. Hypertrophic subaortic stenosis. Clinical and hemodynamic effects of long-term propranolol therapy. Am J Cardiol 1973;31(6):763-73. 150. Gadler F, Linde C, Daubert C, et al. Significant improvement of quality of life following atrioventricular synchronous pacing in patients with hypertrophic obstructive cardiomyopathy. Data from 1 year of follow-up. PIC study group. Pacing In Cardiomyopathy. EurHeart J 1999;20(14):1044-50. 151. Maron BJ, Nishimura RA, McKenna WJ, et al. Assessment of permanent dual-chamber pacing as a treatment for drug-refractory symptomatic patients with obstructive hypertrophic cardiomyopathy. A randomized, double-blind, crossover study (M-PATHY). Circulation 1999;99(22):2927-33. 152. Riedlbauchova L, Janousek J, Veselka J. Ablation of hypertrophic septum using radiofrequency energy: an alternative for gradient reduction in patient with hypertrophic obstructive cardiomyopathy? J Invasive Cardiol 2013;25(6):E128-32. 153. Shah PM, Taylor RD, Wong M. Abnormal mitral valve coaptation in hypertrophic obstructive cardiomyopathy: proposed role in systolic anterior motion of mitral valve. AmJCardiol 1981;48(2):258-62. 154. Jiang L, Levine RA, King ME, et al. An integrated mechanism for systolic anterior motion of the mitral valve in hypertrophic cardiomyopathy based on echocardiographic observations. Am Heart J 1987;113(3):633-44. 155. Sherrid MV, Chaudhry FA, Swistel DG. Obstructive hypertrophic cardiomyopathy: echocardiography, pathophysiology, and the continuing evolution of surgery for obstruction. Ann Thorac Surg 2003;75(2):620-32. 156. Seggewiss H, Faber L. Percutaneous septal ablation for hypertrophic cardiomyopathy and mid- ventricular obstruction. Eur J Echocardiogr 2000;1(4):277-80. 157. Tengiz I, Ercan E, Alioglu E, et al. Percutaneous septal ablation for left mid-ventricular obstructive hypertrophic cardiomyopathy: a case report. BMC Cardiovasc Disord 2006;6:15. 158. Nagueh SF, Lakkis NM, Middleton KJ, et al. Changes in left ventricular filling and left atrial function six months after nonsurgical septal reduction therapy for hypertrophic obstructive cardiomyopathy. JAmCollCardiol 1999;34(4):1123-28. 159. Reant P, Captur G, Mirabel M, et al. Abnormal septal convexity into the left ventricle occurs in subclinical hypertrophic cardiomyopathy. J Cardiovasc Magn Reson 2015;17(1):64. 160. Kim LK, Swaminathan RV, Looser P, et al. Hospital volume outcomes after septal myectomy and alcohol septal ablation for treatment of obstructive hypertrophic cardiomyopathy: Us nationwide inpatient database, 2003-2011. JAMA Cardiology 2016;1(3):324-32. 161. Steggerda RC, Balt JC, Damman K, et al. Predictors of outcome after alcohol septal ablation in patients with hypertrophic obstructive cardiomyopathy. Special interest for the septal coronary anatomy. Neth Heart J 2013;21(11):504-9. 162. Mamas MA, Anderson SG, Carr M, et al. Baseline bleeding risk and arterial access site practice in relation to procedural outcomes after percutaneous coronary intervention. J Am Coll Cardiol 2014;64(15):1554-64. 163. Arujuna A, Karim R, Caulfield D, et al. Acute pulmonary vein isolation is achieved by a combination of reversible and irreversible atrial injury after catheter ablation: evidence from magnetic resonance imaging. Circ Arrhythm Electrophysiol 2012;5(4):691-700. 164. Modi S, Skanes AC. Complex problems require complex solutions...but may result in other complex problems. Heart Rhythm 2011;8(11):1667-8.
373
374
Appendices:
Appendix 2.1: Database documenter
A facility of Microsoft Access is the ability to export a document to Microsoft Word detailing tables and fields recorded. A summary of the data fields collected is presented below.
C:\Users\Robert Cooper\Desktop\HOCM 23 August 2012 databases\RetrospectiveHOCMDataCollection 060812.accdb
Table: tblBasicDemographics Page: 1
Columns
Name Type Size
HOCMID Long Integer 4
CTNumber Text 18
PatientInitials Text 4
DOB Date/Time 8
EthnicOrigin Text 20
Gender Text 255
SourceReferral Text 255
Field1 Text 255
Field2 Text 255
375
C:\Users\Robert Cooper\Desktop\HOCM 23 August 2012 databases\RetrospectiveHOCMDataCollection 060812.accdb
Table: tblChronicDisease Page: 2
Columns
Name Type Size
HOCMID Long Integer 4
DateOfObs Date/Time 8
DM Text 255
COPD Text 20
OtherLungDisease Text 255
ExtraCardiacArteriopathy Text 20
PreviousHOCMAblation Text 20
PreviousPCI Text 255
PreviousCardiacSurgery Text 255
DiastolicHeartFailure Text 255
Comments Text 255
Comments2 Text 255
C:\Users\Robert Cooper\Desktop\HOCM 23 August 2012 databases\RetrospectiveHOCMDataCollection 060812.accdb
Table: tblCPEx Page: 3
Columns
376
Name Type Size
CPEXID Long Integer 4
HOCMID Long Integer 4
DateOfCPEx Date/Time 8
Method Text 20
RER Decimal 16
PeakVO2 Decimal 16
PeakWork Long Integer 4
PeakHR Long Integer 4
ExDurationSecs Long Integer 4
Rhythm Text 20
VentEctopy Text 255
Ischaemia Text 255
BPResponse Text 255
Field1 Text 255
C:\Users\Robert Cooper\Desktop\HOCM 23 August 2012 databases\RetrospectiveHOCMDataCollection 060812.accdb
Table: tblCRMDImplant Page: 4
Columns
Name Type Size
CRMDID Long Integer 4
377
HOCMID Long Integer 4
Device Text 255
DateOfImplant Date/Time 8
Indication Text 255
PacingNYHAImprovement Long Integer 4
PacingGradientReduction Long Integer 4
Comments Text 255
C:\Users\Robert Cooper\Desktop\HOCM 23 August 2012 databases\RetrospectiveHOCMDataCollection 060812.accdb
Table: tblCRMDVentArrhythmiaDetection Page: 5
Columns
Name Type Size
CRMDID Long Integer 4
HOCMID Long Integer 4
Device Text 30
DateOfObservationPacingReliance Date/Time 8
VT Text 255
LengthOfLongestVT(Beats) Long Integer 4
NumberEpisodesVT Long Integer 4
ATPDelivered Text 50
378
DateFirstATP Date/Time 8
NumberATPDelivered Long Integer 4
ShockDelivered Text 50
DateFirstShock Date/Time 8
NumberShocksDelivered Long Integer 4
InappropriateShock Text 20
Comments Text 255
Comments2 Text 255
C:\Users\Robert Cooper\Desktop\HOCM 23 August 2012 databases\RetrospectiveHOCMDataCollection 060812.accdb
Table: tblCRMDVentArrhythmiaDetection Page: 6
Fields:
NumberEpisodesVT Ascending
NumberShocksDelivered 1
Clustered: False
DistinctCount: 2
Foreign: False
IgnoreNulls: False
Name: NumberShocksDelivered
Primary: False
379
Required: False
Unique: False
Fields:
NumberShocksDelivered Ascending
C:\Users\Robert Cooper\Desktop\HOCM 23
August 2012 databases\RetrospectiveHOCMDataCollection 060812.accdb
Table: tblECG Page: 7
Columns
Name Type Size
ECGID Long Integer 4
DateOfECG Date/Time 8
HOCMID Long Integer 4
Rhythm Text 40
VRate Long Integer 4
PR Long Integer 4
QRS Long Integer 4
BBB Text 10
Field1 Text 255
380
C:\Users\Robert Cooper\Desktop\HOCM 23 August 2012 databases\RetrospectiveHOCMDataCollection 060812.accdb
Table: tblEcho1 Page: 8
Columns
Name Type Size
EchoID Long Integer 4
HOCMID Long Integer 4
DateEcho Date/Time 8
EchoRestMaxGrad Long Integer 4
EchoValsalvaMaxGrad Long Integer 4
EchoStressMaxGrad Long Integer 4
StressMethod Text 20
IVSd Long Integer 4
LVPWd Long Integer 4
LVIDd Long Integer 4
LVIDs Long Integer 4
LA Long Integer 4
SAM Text 255
MR Text 255
Comments Text 255
C:\Users\Robert Cooper\Desktop\HOCM 23 August 2012
381 databases\RetrospectiveHOCMDataCollection 060812.accdb
Table: tblEcho2 Page: 10
Columns
Name Type Size
EchoID Long Integer 4
HOCMID Long Integer 4
OtherHOCMFinding Text 40
E Long Integer 4
A Long Integer 4
E:A Memo -
E'Med Decimal 16
E'Lat Decimal 16
E:E'Med Decimal 16
E:E'Lat Decimal 16
IVRT Memo -
Field1 Text 255
C:\Users\Robert Cooper\Desktop\HOCM 23 August 2012 databases\RetrospectiveHOCMDataCollection 060812.accdb
Table: tblEvents Page: 11
Columns
Name Type Size
382
EventID Long Integer 4
HOCMID Long Integer 4
DateOfEvent Date/Time 8
ProcedureRelated Text 18
Event Text 255
Comments Text 255
Field1 Text 255
C:\Users\Robert Cooper\Desktop\HOCM 23 August 2012 databases\RetrospectiveHOCMDataCollection 060812.accdb
Table: tblObsStatus Page: 12
Columns
Name Type Size
HOCMID Long Integer 4
ObsStatusID Text 255
DateOfObs Date/Time 8
CardiacRhythm Text 255
NYHA Text 255
CCS Text 255
SyncopeSinceLastObs Text 255
FHHOCMSinceLastObs Text 255
BetaBlocker Text 255
383
CalciumChannelBlocker Text 255
Disopyrimide Text 255
Amiodarone Text 255
Comments Text 255
C:\Users\Robert Cooper\Desktop\HOCM 23 August 2012 databases\RetrospectiveHOCMDataCollection 060812.accdb
Table: tblPatientDetails Page: 13
Columns
Name Type Size
HOCMID Long Integer 4
Surname Text 50
Forename Text 50
ContactAddress Text 255
ContactPhone Memo -
C:\Users\Robert Cooper\Desktop\HOCM 23 August 2012 databases\RetrospectiveHOCMDataCollection 060812.accdb
Table: tblProcedure Page: 14
Columns
Name Type Size
ProcID Long Integer 4
384
HOCMID Long Integer 4
DateProc Date/Time 8
PatientProcedureNumber Long Integer 4
MultipleProcedures Text 255
TPWSite Text 10
RestGradientPreInjection Long Integer 4
PostEctopicGradientPreInjection Long Integer 4
PharmacologicalStressGradientPreInjection Long Integer 4
RestGradientPostInjection Long Integer 4
PostEctopicGradientPostInjection Long Integer 4
PharmacologicalStressGradientPostInjection Long Integer 4
AlcoholDelivered Text 20
ReasonAlcoholNotUsed Text 30
MCEUsed Text 20
CKMBRelease Long Integer 4
OtherCardiacMarker Text 10
OtherCardiacMarkerRelease Long Integer 4
OtherCardiacMarkerUpperNorma Long Integer 4
LOS Long Integer 4
Comments Text 255
C:\Users\Robert Cooper\Desktop\HOCM 23 August 2012
385 databases\RetrospectiveHOCMDataCollection 060812.accdb
Table: tblSeptalInjection Page: 16
Columns
Name Type Size
SeptalID Long Integer 4
ProcID Long Integer 4
OriginOfSeptal Text 10
ArteryInjected Text 10
ReasonNoAlcoholInjected Text 255
VolumeInjectedMl Decimal 16
ArterySize Text 20
Comments Text 50
Appendix 2.2: Service evaluation registry form NSRT at LHCH
Service Evaluation Registration Project Plan Purpose of document The purpose of this document is to define how the audit has been selected, approved, designed and how it is to be undertaken (methodology) Title Non-surgical septal reduction therapy: Current practice and future potential
Clinical outcome from alcohol septal ablation at Liverpool Heart and Chest Hospital 2001-2011
Directorate Cardiology
386
Clinical Research
Service line Cardiology
(ward /department)
Project team Name Contact details
Project lead Robert Cooper Bleep: 2731
Rod Stables Email: [email protected]
Number: 1209
Email: [email protected]
Adeel Shahzad Bleep: 2725
Email: [email protected]
Number / bleep: Other staff members Email: directly involved in carrying out Number / bleep: the project Email:
Number / bleep:
Email:
Clinical audit James McShane officer(s) facilitating
Involvement of key stakeholders Roles / accountability for project findings and to enact change / communication)
Name Title Their role in project
Robert Cooper Research Fellow Lead data collection and analysis. Preparation of data for publication and presentation
387
Rod Stables Research supervisor Supervision. Operator on all alcohol ablation procedures
James McShane Clinical information analyst Help with data analysis
Involvement of patients /carers Please Please
(Indicate with a X) X X
None □ Engaged patients /carers in the □
project e.g. Topic selection, design
Project takes into account patient priorities x Patient /carer role is part of the □ and patient defined outcomes project team directly undertaking
the project
Preparation and Planning
Background/rationale: (The reason behind doing this project)
Hypertrophic Cardiomyopathy (HCM) is an inherited disease characterised by otherwise unexplained hypertrophy of the myocardium. It is transmitted in an autosomal dominant pattern with variable penetrance, with an estimated prevalence of 1 in 5001. Whilst the distribution of hypertrophy can be varied, involvement of the basal interventricular septum is common. The classical description of asymmetrical basal hypertrophy narrowing the left ventricular outflow tract (LVOT) contributes to the pathology underlying LVOT obstruction (LVOTO). The prevalence of LVOTO in HCM is 20-30% at
388 rest2 and up to 70% with provocation3. LVOTO is associated with greater levels of dyspnoea, a greater incidence of stroke and higher mortality2.
Early cases of symptomatic hypertrophic obstructive cardiomyopathy (HOCM) were treated surgically, with good results 4-6. Whilst surgical techniques have advanced and mortality in expert centres has improved, the principle of septal myectomy through direct visualisation and incision remains intact.
The desire for a less invasive option with lower morbidity and faster recovery times were all drivers for the development of a non-surgical solution. This could also provide an option for those with high surgical risk precluding them from myectomy. Case reports of myocardial infarction had been reported to cause resolution of the clinical signs associated with HOCM7, creating a localised infarction percutaneously could therefore alter LVOT haemodynamics. Early work in this field began in
1983, when inflation of an angioplasty balloon in a septal artery was noted to cause reduction in
LVOT gradients8. The gradients returned when blood supply was restored. A more definitive solution for inducing infarct was therefore required. Some success had been reported with the creation of an infarct using trans-coronary alcohol delivery to remove a ventricular arrhythmogenic substrate9; this was therefore adapted to target the basal septum. The pioneering non-surgical septal reduction therapy (NSRT) was performed with trans-coronary alcohol septal injection by Sigwart in 1994. A 68- year-old female with HOCM who failed to respond to medical therapy and pacing underwent alcohol septal ablation (ASA) with a good result. Her case and 2 others were reported in 199510. She remained well 10 years later11. An alternative to surgery was now available.
The safety and efficacy of NSRT can be improved with fastidious patient selection and procedural planning. The 2011 ACCF / AHA Guidelines for the Diagnosis and Treatment of HCM stipulate that there must be a subaortic gradient of at least 50mmHg, at rest or with provocation 12. The 2003
ACC/ESC consensus document on HCM first set this gradient requirement 13, whilst others have suggested a resting gradient of >30mmHg and a rise with provocation to >50mmHg14;15 or
>100mmHg16 will suffice. It is important that provocation should be physiological rather than pharmacological, as dobutamine can produce subaortic gradients in the normal heart17. Treating a symptomatic patient with no resting pressure difference, but with exercise induced gradient can have a good outcome 18. Symptom burden is part of selection criteria. Patients must have limiting symptoms that are refractory to medical therapy 13. Dyspnoea classed as NYHA grade III-IV, and to a
389 lesser extent chest pain and syncope are indicators for treatment. Medical therapy uses negative inotropes and can include beta-blockers, calcium channel blockers and disopyramide. Patients should have hypertrophy of the basal interventricular septum thought to be responsible for causing a gradient. Some quote interventricular width of >18mm with clear protrusion into the LV cavity19, and others >15mm20. The 2011 ACCF/AHA guidelines simply state that septal thickness is “sufficient to perform the procedure safely and effectively in the judgement of the individual operator”12.
The National Institute of Clinical Excellence in the UK states that NSRT can be performed in HOCM patients with symptoms refractory to medical therapy. They make a requirement this should only be performed in specialist units with clinicians who have adequate training 21. This was re-enforced in the
ACCF/ACC guidelines stating an operator must have experience of at least 20 procedures or work in a centre with a cumulative procedural volume of 50 patients12. The learning curve in ASA can require the performance of at least 40 procedures 22.
The evidence base for the performance of alcohol septal ablation is limited to case series. The incidence of HOCM patients requiring invasive treatment is relatively low. This hinders the ability to power a randomised controlled trial of ablation versus septal myectomy surgery to detect superiority in terms of survival. There will likely never be such a trial. Our cohort of patients is the largest in the
UK and should be reported to add to the literature base. A brief summary of the evidence to date is provided below.
Survival:
HCM patients with resting LVOTO have a higher mortality than matched HCM patients without a gradient2. Recent case series have suggested that removing the gradient may have a beneficial effect on survival. Of 173 ASA patients, with mean age of 64 years, followed up for a median time of 5.7 years, survival was no different to that of the general, non-HCM population23. The survival was identical to those treated with surgical myectomy in the same time period. Residual LVOTO was a predictor of mortality in this series.
390
The largest series of ASA patients studied for survival (n=465) over a mean of 8.4 years showed a 1,
5 and 10-year survival of 99, 94, and 90%. This compares favourably to that of the age and sex matched general, non-HCM population figures of 99, 93, and 84% respectively24.
A further study comparing survival following invasive treatment with ASA versus conservative management found a benefit with ASA. This was explained by non-cardiac death25.
These recent series will go some way to reassuring operators that removing the LVOT gradient with
ASA has a beneficial effect on long term outcome.
Risk of ventricular arrhythmia:
The medium term risk of ventricular arrhythmia has been studied with conflicting results. Some observational studies provide the reassurance that the risk of sudden cardiac death (SCD) is no higher than HCM patients without an iatrogenic myocardial infarction26-28 . A registry of all ASA patients with ICDs for primary prevention highlighted the annual incidence of appropriate ICD discharge as 2.8%29. A comprehensive registry of 465 ASA patients in Denmark and Germany identified the SCD risk over 8.4 years as 0.5% per annum, with 16 sudden deaths and 3 appropriate
ICD discharges (representing 2.7% appropriate ICD discharge rate)24. The majority of ICD series identify an appropriate discharge rate of 2.5-3% per annum post-ASA.
The multi-centre HCM ICD registry in North America highlighted an increased risk of appropriate ICD activation post-ASA. A 4-fold increase in appropriate ICD treatment was observed with event rates of
10.3% per year compared to 2.6% per year in those treated with myectomy30. The ASA group, however, consisted of only 17 patients and 4 events, the low numbers may contribute to an eye- catching difference in this report.
Meta-analyses have not indicated any difference between ASA and septal myectomy in the medium term incidence of SCD or all-cause mortality31;32. The overall trend in SCD and VT risk in reported series appears to favour no clear increase in arrhythmia post ASA.
Symptom and gradient resolution:
The first series of 18 patients to undergo ASA was reported in 199733. This was promising and showed that the procedure had potential to reduce LVOT gradients, reduce symptom burden and
391 increase exercise tolerance. A larger series of patients were reported in 1999, 50 in total, were followed up for 7 months34, showing a clear improvement in LVOT gradients, interventricular septal size, VO2MAX and pulmonary artery pressures. Perhaps more importantly the change in patient symptoms was reproduced, with NYHA status improving from a mean of 3 to 1.7.
Following these breakthrough studies multiple early patient series were reported, and summarised in a systematic review collated in 200635. In total 2959 patients were summarised from 42 published studies, although the authors accept some were duplicated by involvement in more than one report. A baseline assessment revealed a mean age of 53.5 years, an NYHA class of 2.9 despite medical therapy, and peak LVOT gradients of 65mmHg at rest and 125mmHg on provocation. Mean follow up was 12.7 months. Early mortality (defined as within 30 days) was reported to be 1.5%, with late all- cause mortality at 0.5% (0-9.3%). NYHA status improved significantly to 1.2 (p<0.001). A reduction in angina burden was also seen; Canadian Cardiovascular Society score reduced from 1.9 to 0.4
(p<0.001). Echocardiographic gradients reduced to 15mmHg at rest and 31mmHg with provocation
(both p<0.001). Repeat procedures were required in 7% of patients.
A 2011 report from the Multicentre North American registry detailed procedures in 874 patients. All had a resting gradient of ≥30mmHg or provocable gradient of ≥60mmHg with advanced symptoms of exertional dyspnoea and/or angina despite medical therapy36. Mean follow up duration was 26 months. Seventy-eight per cent of patients suffered NYHA class III or IV dyspnoea prior to the procedure, with a gradient at rest measuring 70mmHg, and 99mmHg on provocation. Following ASA
72.5% of patients were classified NYHA class I, 23% NYHA class II, 3.9% class III and 0.65% class
IV. Repeat procedure was required in 12.8% of patients. A mean of peak resting gradients post procedure reduced to 35mmHg. The Scandinavian multicentre study included 279 patients with similar results, with NYHA III/IV breathlessness reduced from 94% to 21%, and outflow tract gradients falling from 58mmHg to 12mmHg. Those persisting with NYHA III/IV breathlessness had a high prevalence of co-morbidities including chronic obstructive pulmonary disease and valve disease37.
A number of series of medium term results are now available, with time periods ranging from 25-141 months26;34;36-45 (see Table 2.). Improvement in dyspnoea is observed in most, with a clear trend towards lower LVOT gradients. The procedure does not provide uniform improvement for all, however, with large recent series displaying a mean post procedure resting gradient of 35mmHg46 (a
392 gradient some would claim justifies repeat treatment), and persistent NYHA class III dyspnoea in
21%37. There is therefore scope for improvement in the treatment and subsequent outcomes for HCM patients with LVOTO refractory to medical therapy.
Complete
Please Does the project meet any of the following? (Indicate with a X only those that apply) X
The topic is concerned with high cost, high volume or high risk to staff or patients / service users □
There is evidence of a quality problem (e.g. patient complaints, high complication rates, adverse □ outcomes or poor symptom control)
There is evidence of wide variation in practice □
There is good evidence available to inform best practice (e.g. systematic reviews or professional □ guidelines)
The problem is measurable against the literature x
Evaluation of the problem is likely to improve healthcare outcomes as well as process x improvements
Evaluation of the problem is likely to have economic and efficiency benefits □
The topic is of a key professional or clinical interest x
393
Reliable sources of data are readily available for data collection purposes □
Data can be collected within a reasonable time frame x
The problem concerned is amenable to change □
The topic is pertinent to national or local initiatives or priorities x
The topic lends itself to a service evaluation, a different process is not deemed more appropriate □ (e.g. research)
There is scope for improvement and potential benefits of undertaking this evaluation □
Project aims & objectives: (The goal you wish the project to achieve, what you need the project to tell you/what you hope to identify)
Complete
The purpose of this project is to:
Report clinical and echocardiographic outcomes in 90 patients taken to the cardiac catheterisation laboratory with the intention of delivering alcohol. Detail safety of the procedure to inform local consent policies and produce a patient information leaflet. Provide a base for comparison to prospective patients undergoing alcohol ablation with improved methods
Evidence base: (What does the literature tell you about this topic?)
394
List references
Please see background section. Appropriate papers are referenced as per BMJ group methods.
Methodology to measure performance robust data collection methods / audit design
Project type (Indicate with a X those that apply) Retrospective Prospecti Observatio Preval
studyx ve study x nal study x ence
study
□
Multi-disciplinary x Uni-
disciplin
ary □
Case selection /sample size (define cases you are selecting /over what time frame? E.g. isolated 1st time CABG Alcohol septal ablation patients treated between between .../../... and .../.../....)? Make sure adequately significant, representative, clinically relevant, unbiased, to provide 2001-2011. meaningful results) 90 appropriate patients have been identified.
395
Data source – what sources are you using? Utility of Case note review pre-existing data, electronic sources Tomcat and pacing clinic notes
Xcelera echocardiographic reports
Data collection / collation – how /who proforma Data will be collected using a relational Microsoft design (attach) data entry? Please ensure that you detail Access database with many linked tables and the exact personal data being used for the service fields. A copy of this can be attached as an evaluation electronic file upon request.
Data analysis (detail the packages to be used, e.g. SPSS, MS
Excel, MS Access consider how to present results -use most Microsoft Access appropriate manner for audience, use of graphical illustration)
Microsoft Excel
SPSS
Stats Direct
396
Clinical Quality Department support? None □ registration Advice Proforma Identifying
(Indicate with a X those that apply) only only □ design □ cases □
Extracting raw data from a clinical database □ Analytical
support x
(You need to specify all variables required from the clinical
database as the data will be extracted once only)
Data handling and storage following recommendations in the Caldicott Report the Data Protection
Act, and other legislation regarding the security of data
Transferring the data
Personal data should not be transferred via the internet Paper records □ as this is not a secure system.
Computer record x
(Note – Patient/user identifiable data must only transferred by e-mail if the outgoing message is encrypted to NHS standards. Either by using the Portable media device □ Trust’s email encryption tool or by using the secure
NHS network ie @nhs.net)
Is the data being transferred outside the UK?
Yes □
397
No x
If yes, please provide detail on the steps taken to
secure the data:
Sharing the data Will the data be shared with other organisations?
Yes □
No x
If yes – please provide further detail, including
the recipient
organisation’s Data Protection Act registration
number:
Storage of the data Paper records
How long will the data be stored and are there any To be kept on site, locked in safe place □ physical precautions in place? Computer records
Filed on network server, restrict access to project
team x
Portable media device
Data stored on an encrypted (USB or email) device
□
398
Retention and destruction of the data Paper records
Data should only be retained for the project purpose and Shredded in confidential manner □ destroyed once analysis complete and report written Computer records
Deleted from server and recycle bin x
Portable media device
Deleted from device □
Email correspondence deleted □
DECLARATION
I agree that I will arrange to securely receive, store, and process the data requested as described above and that it will be destroyed in a confidential manner when no longer needed.
I confirm that I undertook my mandatory IG training on: ___5__/___8___/__2011___
I confirm that I understand my personal responsibilities towards Information Governance particularly regarding confidentiality and data protection
I confirm that I understand my legal obligations to comply with the requirements of information legislation, in particular the Data Protection Act 1998 and to act in accordance with Trust Information Governance Policies
I confirm that I understand the requirement to meet the standards set out in my terms of employment and that failure to meet the standards set out in the confidentiality and access to information clause could result in
399 disciplinary action, termination of a contract, dismissal and in some cases criminal charges
Name: Robert Cooper Designation: Research Fellow
Signature: _____R M Cooper______Date: 15/8/11
Work plan: insert anticipated start dates for each stage
Project time scales (from start to finish) Date
Preparation and planning start: (Registration, proforma design) __15__/___8__/__11__
Collecting data: __25__/___8__/__11__
Inputting / analysing data: (data input, extracting and analysis) __25__/___8__/__11__
Report writing: (methodology, conclusion & action plan if recommendations) __1___/___2__/__12__
Project presentation Details
Report completion by _1___/__12____/___12___
Where are you presenting? LHCH clinical audit session
Data will be submitted in abstract form to BCS and ESC meetings
When Dates pending
What format PPT, poster
400
Service evaluation project approval
This section to be completed by RAE Officer
Approval – required for priority levels 3-5 if not previously approved in the audit forward plan
Priority level
1 External ‘Must-do’ (AFP) □
2 Internal ‘Must-do’ (AFP) □
3 High local priority □
4 Medium local priority □
5 Low local priority □
Support level
1 Full facilitation □
Moderate support – review □
2 design, practical assistance
Minimal support – registration □ 3 and advice only
Key outcome level
A Assurance □
This section to be completed by Service line lead B Safety improvement □ Name of Service Line Clinical Lead: Date approved:C Clinical and /or cost effective □
D Pt. experience improvement □ 401
Registered on clinical audit database: Audit ID number ______Code of Conduct for data extractions from clinical databases completed □ N/A □ Reference List
(1) Elliott P, McKenna WJ. Hypertrophic cardiomyopathy. Lancet 2004; 363(9424):1881-1891.
(2) Maron MS, Olivotto I, Betocchi S, Casey SA, Lesser JR, Losi MA et al. Effect of left ventricular
outflow tract obstruction on clinical outcome in hypertrophic cardiomyopathy. N Engl J Med
2003; 348(4):295-303.
(3) Maron MS, Olivotto I, Zenovich AG, Link MS, Pandian NG, Kuvin JT et al. Hypertrophic
cardiomyopathy is predominantly a disease of left ventricular outflow tract obstruction.
Circulation 2006; 114(21):2232-2239.
(4) Woo A, Williams WG, Choi R, Wigle ED, Rozenblyum E, Fedwick K et al. Clinical and
echocardiographic determinants of long-term survival after surgical myectomy in obstructive
hypertrophic cardiomyopathy. Circulation 2005; 111(16):2033-2041.
(5) Nagueh SF, Ommen SR, Lakkis NM, Killip D, Zoghbi WA, Schaff HV et al. Comparison of
ethanol septal reduction therapy with surgical myectomy for the treatment of hypertrophic
obstructive cardiomyopathy. J Am Coll Cardiol 2001; 38(6):1701-1706.
(6) Qin JX, Shiota T, Lever HM, Kapadia SR, Sitges M, Rubin DN et al. Outcome of patients with
hypertrophic obstructive cardiomyopathy after percutaneous transluminal septal myocardial
ablation and septal myectomy surgery. J Am Coll Cardiol 2001; 38(7):1994-2000.
(7) Come PC, Riley MF. Hypertrophic cardiomyopathy. Disappearance of auscultatory, carotid
pulse, and echocardiographic manifestations of obstruction following myocardial infarction.
Chest 1982; 82(4):451-454.
402
(8) Sigwart U, Grbic M, Payot M, Essinger A, Sadeghi H. Ventricular wall motion during balloon
occlusion. Sigwart U, Heintzen PH, editors. Ventricular wall motion.New York (NY) , 206-210.
1983.
Ref Type: Generic
(9) Brugada P, de SH, Smeets JL, Wellens HJ. Transcoronary chemical ablation of ventricular
tachycardia. Circulation 1989; 79(3):475-482.
(10) Sigwart U. Non-surgical myocardial reduction for hypertrophic obstructive cardiomyopathy.
Lancet 1995; 346(8969):211-214.
(11) Vecht JA, Dave R, Vecht RJ. Alcohol septal ablation: The first patient in 1994. British Journal
of Cardiology 2006; 13(1):62-64.
(12) Gersh BJ, Maron BJ, Bonow RO, Dearani JA, Fifer MA, Link MS et al. 2011 ACCF/AHA
Guideline for the Diagnosis and Treatment of Hypertrophic Cardiomyopathy: Executive
Summary: A Report of the American College of Cardiology Foundation/American Heart
Association Task Force on Practice Guidelines. Circulation 2011.
(13) Maron BJ, McKenna WJ, Danielson GK, Kappenberger LJ, Kuhn HJ, Seidman CE et al.
American College of Cardiology/European Society of Cardiology clinical expert consensus
document on hypertrophic cardiomyopathy. A report of the American College of Cardiology
Foundation Task Force on Clinical Expert Consensus Documents and the European Society of
Cardiology Committee for Practice Guidelines. J Am Coll Cardiol 2003; 42(9):1687-1713.
(14) Lakkis N, Plana JC, Nagueh S, Killip D, Roberts R, Spencer WH, III. Efficacy of nonsurgical
septal reduction therapy in symptomatic patients with obstructive hypertrophic
cardiomyopathy and provocable gradients. Am J Cardiol 2001; 88(5):583-586.
403
(15) Parakh N, Bhargava B. Golden jubilee of hypertrophic cardiomyopathy: is alcohol septal
ablation the gold standard? Cardiovasc Revasc Med 2009; 10(3):172-178.
(16) Seggewiss H. Medical therapy versus interventional therapy in hypertropic obstructive
cardiomyopathy. Curr Control Trials Cardiovasc Med 2000; 1(2):115-119.
(17) Pellikka PA, Oh JK, Bailey KR, Nichols BA, Monahan KH, Tajik AJ. Dynamic intraventricular
obstruction during dobutamine stress echocardiography. A new observation. Circulation
1992; 86(5):1429-1432.
(18) Gietzen FH, Leuner CJ, Obergassel L, Strunk-Mueller C, Kuhn H. Role of transcoronary
ablation of septal hypertrophy in patients with hypertrophic cardiomyopathy, New York
Heart Association functional class III or IV, and outflow obstruction only under provocable
conditions. Circulation 2002; 106(4):454-459.
(19) Holmes DR, Jr., Valeti US, Nishimura RA. Alcohol septal ablation for hypertrophic
cardiomyopathy: indications and technique. Catheter Cardiovasc Interv 2005; 66(3):375-389.
(20) Rigopoulos AG, Seggewiss H. A decade of percutaneous septal ablation in hypertrophic
cardiomyopathy. Circ J 2010; 75(1):28-37.
(21) Non-surgical reduction of the myocardial septum: NICE interventional procedure guidance.
2004.
Ref Type: Generic
(22) van der Lee C, Scholzel B, ten Berg JM, Geleijnse ML, Idzerda HH, van Domburg RT et al.
Usefulness of clinical, echocardiographic, and procedural characteristics to predict outcome
after percutaneous transluminal septal myocardial ablation. Am J Cardiol 2008; 101(9):1315-
1320.
404
(23) Sorajja P, Ommen SR, Holmes DR, Jr., Dearani JA, Rihal CS, Gersh BJ et al. Survival after
alcohol septal ablation for obstructive hypertrophic cardiomyopathy. Circulation 2012;
126(20):2374-2380.
(24) Jensen MK, Prinz C, Horstkotte D, van Buuren F, Bitter F, Faber L et al. Personal
correspondence with Dr M Jensen, Rigshospitalet, Copenhagen: Alcohol septal ablation in
patients with hypertrophic obstructive cardiomyopathy: Low incidence of sudden cardiac
death and reduced risk profile. 2012.
Ref Type: Personal Communication
(25) Ball W, Ivanov J, Rakowski H, Wigle ED, Linghorne M, Ralph-Edwards A et al. Long-term
survival in patients with resting obstructive hypertrophic cardiomyopathy comparison of
conservative versus invasive treatment. J Am Coll Cardiol 2011; 58(22):2313-2321.
(26) Jensen MK, Havndrup O, Hassager C, Helqvist S, Kelbaek H, Jorgensen E et al. Survival and
sudden cardiac death after septal ablation for hypertrophic obstructive cardiomyopathy.
Scand Cardiovasc J 2011; 45(3):153-160.
(27) Lawrenz T, Obergassel L, Lieder F, Leuner C, Strunk-Mueller C, Meyer Z, V et al.
Transcoronary ablation of septal hypertrophy does not alter ICD intervention rates in high
risk patients with hypertrophic obstructive cardiomyopathy. Pacing Clin Electrophysiol 2005;
28(4):295-300.
(28) Klopotowski M, Chojnowska L, Malek LA, Maczynska R, Kukula K, Demkow M et al. The risk
of non-sustained ventricular tachycardia after percutaneous alcohol septal ablation in
patients with hypertrophic obstructive cardiomyopathy. Clin Res Cardiol 2010; 99(5):285-
292.
405
(29) Cuoco FA, Spencer WH, III, Fernandes VL, Nielsen CD, Nagueh S, Sturdivant JL et al.
Implantable cardioverter-defibrillator therapy for primary prevention of sudden death after
alcohol septal ablation of hypertrophic cardiomyopathy. J Am Coll Cardiol 2008;
52(21):1718-1723.
(30) Maron BJ, Spirito P, Shen WK, Haas TS, Formisano F, Link MS et al. Implantable cardioverter-
defibrillators and prevention of sudden cardiac death in hypertrophic cardiomyopathy.
JAMA 2007; 298(4):405-412.
(31) Agarwal S, Tuzcu EM, Desai MY, Smedira N, Lever HM, Lytle BW et al. Updated meta-analysis
of septal alcohol ablation versus myectomy for hypertrophic cardiomyopathy. J Am Coll
Cardiol 2010; 55(8):823-834.
(32) Leonardi RA, Kransdorf EP, Simel DL, Wang A. Meta-analyses of septal reduction therapies
for obstructive hypertrophic cardiomyopathy: comparative rates of overall mortality and
sudden cardiac death after treatment. Circ Cardiovasc Interv 2010; 3(2):97-104.
(33) Knight C, Kurbaan AS, Seggewiss H, Henein M, Gunning M, Harrington D et al. Nonsurgical
septal reduction for hypertrophic obstructive cardiomyopathy: outcome in the first series of
patients. Circulation 1997; 95(8):2075-2081.
(34) Gietzen FH, Leuner CJ, Raute-Kreinsen U, Dellmann A, Hegselmann J, Strunk-Mueller C et al.
Acute and long-term results after transcoronary ablation of septal hypertrophy (TASH).
Catheter interventional treatment for hypertrophic obstructive cardiomyopathy. Eur Heart J
1999; 20(18):1342-1354.
(35) Alam M, Dokainish H, Lakkis N. Alcohol septal ablation for hypertrophic obstructive
cardiomyopathy: a systematic review of published studies. J Interv Cardiol 2006; 19(4):319-
327.
406
(36) Nagueh SF, Groves BM, Schwartz L, Smith KM, Wang A, Bach RG et al. Alcohol septal ablation
for the treatment of hypertrophic obstructive cardiomyopathy a multicenter north american
registry. J Am Coll Cardiol 2011; 58(22):2322-2328.
(37) Jensen MK, Almaas VM, Jacobsson L, Hansen PR, Havndrup O, Aakhus S et al. Long-term
outcome of percutaneous transluminal septal myocardial ablation in hypertrophic
obstructive cardiomyopathy: a Scandinavian multicenter study. Circ Cardiovasc Interv 2011;
4(3):256-265.
(38) Kwon DH, Kapadia SR, Tuzcu EM, Halley CM, Gorodeski EZ, Curtin RJ et al. Long-term
outcomes in high-risk symptomatic patients with hypertrophic cardiomyopathy undergoing
alcohol septal ablation. JACC Cardiovasc Interv 2008; 1(4):432-438.
(39) Fernandes VL, Nielsen C, Nagueh SF, Herrin AE, Slifka C, Franklin J et al. Follow-up of alcohol
septal ablation for symptomatic hypertrophic obstructive cardiomyopathy the Baylor and
Medical University of South Carolina experience 1996 to 2007. JACC Cardiovasc Interv 2008;
1(5):561-570.
(40) Seggewiss H, Rigopoulos A, Welge D, Ziemssen P, Faber L. Long-term follow-up after
percutaneous septal ablation in hypertrophic obstructive cardiomyopathy. Clin Res Cardiol
2007; 96(12):856-863.
(41) Lyne JC, Kilpatrick T, Duncan A, Knight CJ, Sigwart U, Fox KM. Long-term follow-up of the first
patients to undergo transcatheter alcohol septal ablation. Cardiology 2010; 116(3):168-173.
(42) Sorajja P, Valeti U, Nishimura RA, Ommen SR, Rihal CS, Gersh BJ et al. Outcome of alcohol
septal ablation for obstructive hypertrophic cardiomyopathy. Circulation 2008; 118(2):131-
139.
407
(43) Kuhn H, Lawrenz T, Lieder F, Leuner C, Strunk-Mueller C, Obergassel L et al. Survival after
transcoronary ablation of septal hypertrophy in hypertrophic obstructive cardiomyopathy
(TASH): a 10 year experience. Clin Res Cardiol 2008; 97(4):234-243.
(44) Klopotowski M, Chojnowska L, Malek LA, Maczynska R, Kukula K, Demkow M et al. The risk
of non-sustained ventricular tachycardia after percutaneous alcohol septal ablation in
patients with hypertrophic obstructive cardiomyopathy. Clin Res Cardiol 2010; 99(5):285-
292.
(45) Veselka J, Tomasov P, Zemanek D. Long-term effects of varying alcohol dosing in
percutaneous septal ablation for obstructive hypertrophic cardiomyopathy: a randomized
study with a follow-up up to 11 years. Can J Cardiol 2011; 27(6):763-767.
(46) Nagueh SF, Groves BM, Schwartz L, Smith KM, Wang A, Bach RG et al. Alcohol septal ablation
for the treatment of hypertrophic obstructive cardiomyopathy a multicenter north american
registry. J Am Coll Cardiol 2011; 58(22):2322-2328.
Appendix 4.1: Intracardiac Echocardiography in LHCH permissions
408
Our Ref: CAEG/November 2012
19th November 2012
Dr R Stables Dr R Cooper
Consultant Cardiologist Research SpR
LHCH LHCH
Dear Rod / Rob,
Re: Use of radiofrequency ablation (RF) in the treatment of left ventricular outflow tract (LVOT) gradients in hypertrophic cardiomyopathy (HCM).
Many thanks for attending the Clinical Audit & Effectiveness Group meeting on 15th November 2012 and discussing this new technology.
On behalf of the Trust we are happy for you to continue with the programme to introduce this new technology. Operators approved for using this new technology include Dr M Hall, Dr S Modi and Dr D Todd, fully supported by both of you in fully interpreting intracardiac echocardiography merged with our CARTO systems.
Your application and subsequent discussion at the meeting demonstrates that you have the necessary governance arrangements in place for consent, training and audit for the use of this device.
This was approved subject to the following:
409
1) A report to the CAEG in 9 months, outlining initial experience including complications
2) Please provide the hospital number of the first planned case to the CAEG Administrator
Best wishes for its successful development, sent on behalf of CAEG members.
Yours sincerely
Janet Deane (CAEG Administrator)
Senior Research, Audit and Effectiveness Officer
Care Pathways Coordinator
c.c. Cath Barton Dr G Russell Dr R Perry General Manager, CCM Medical Director Deputy Medical Director
Dr D Todd Dr S Modi Dr M Hall
Service Line Clinical Lead Cons Cardiologist / EPS Cons Cardiologist / EPS
Karen Wafer Ann-Marie Francis
Catheter Lab Manager SLM
Dr M Ledson Dr M Jackson
410
Chair, CAEG Exec Dir. Research and Informatics
Appendix 4.2: Patient information leaflet for the use of intracardiac echocardiography
411
412
Appendix 6.1: Permissions to use CT angiography prior to ASA at LHCH
Our Ref: CAEG/November 2012
19th November 2012
Dr R Stables Dr R Cooper
Consultant Cardiologist Research SpR
LHCH LHCH
Dear Rod / Rob,
Re: Use of radiofrequency ablation (RF) in the treatment of left ventricular outflow tract (LVOT) gradients in hypertrophic cardiomyopathy (HCM).
Many thanks for attending the Clinical Audit & Effectiveness Group meeting on 15th November 2012 and discussing this new technology.
On behalf of the Trust we are happy for you to continue with the programme to introduce this new technology. Operators approved for using this new technology include Dr M Hall, Dr S Modi and Dr
413
D Todd, fully supported by both of you in fully interpreting intracardiac echocardiography merged with our CARTO systems.
Your application and subsequent discussion at the meeting demonstrates that you have the necessary governance arrangements in place for consent, training and audit for the use of this device.
This was approved subject to the following:
1) A report to the CAEG in 9 months, outlining initial experience including complications
2) Please provide the hospital number of the first planned case to the CAEG Administrator
Best wishes for its successful development, sent on behalf of CAEG members.
Yours sincerely
Janet Deane (CAEG Administrator)
Senior Research, Audit and Effectiveness Officer
Care Pathways Coordinator
c.c. Cath Barton Dr G Russell Dr R Perry General Manager, CCM Medical Director Deputy Medical Director
Dr D Todd Dr S Modi Dr M Hall
414
Service Line Clinical Lead Cons Cardiologist / EPS Cons Cardiologist / EPS
Karen Wafer Ann-Marie Francis
Catheter Lab Manager SLM
Dr M Ledson Dr M Jackson
Chair, CAEG Exec Dir. Research and Informatics
Appendix 6.2: Example case record form for CT details prior to ASA
CT assessment of septal coronary arteries case record form
Patient sticker Anonymised
Study ID HOCM28
Date of scan xx
Radiologists SB
415
Scan Indication HOCM ablation
416
Septal 1 Septal 2 Septal 3 Septal 4 Septal 5
Septal name 1 2 3 Method of Retrograde Retrograde Retrograde identification A. Parent Artery LAD LAD LAD
A. Parent Artery IQS 2 2 2
B. Bifurcation angle 151 146 145
B. Bifurcation angle 2 2 2 IQS C. Septal artery area 0.46 0.77 1.69 1.47 3.8 2.19 / diameter (mm2/mm C. Diameter IQS 2 2 2 D. Length of septal in 2.1 4.2 5.9 14.2 0 33.2 fat/muscle (mm) D. length IQS 2 2 2 E. Length pre- 2.1 6.1 13.9 intraseptal split E. Length pre- 2 2 2 intraseptal split IQS F. Landmarks 26.2mm from 4.9mm from Myocardial LMS S1 bridge, S3 bifurcation F. Landmarks IQS 2 2 2 G. Myocardial RV RV LVBS, inferior distribution septum and RV G. Distribution IQS 2 2 2 H. Preferred N/A N/A N/A N/A L35 AP projection ostium Ca32 Cr20 H. Preferred N/A 2 projection ostium IQS I. Preferred projection N/A N/A N/A N/A R45 x distal Ca20 I. Preferred projection N/A 2 distal IQS J. Order of preference X X 1 for alcohol injection (subselective) Comments
417
RV 13.9mm
RV RV
3.1mm
12.8
Inferior SAM septum
contact
418
Appendix 7.1: Patient information leaflet for ASA at LHCH
419
420
Appendix 8.1: Permissions to use RF ablation in HOCM
Our Ref: CAEG/November 2012
19th November 2012
Dr R Stables Dr R Cooper Consultant Cardiologist Research SpR LHCH LHCH
Dear Rod / Rob, Re: Use of radiofrequency ablation (RF) in the treatment of left ventricular outflow tract (LVOT) gradients in hypertrophic cardiomyopathy (HCM).
Many thanks for attending the Clinical Audit & Effectiveness Group meeting on 15th November 2012 and discussing this new technology.
On behalf of the Trust we are happy for you to continue with the programme to introduce this new technology. Operators approved for using this new technology include Dr M Hall, Dr S Modi and Dr D Todd, fully supported by both of you in fully interpreting intracardiac echocardiography merged with our CARTO systems.
Your application and subsequent discussion at the meeting demonstrates that you have the necessary governance arrangements in place for consent, training and audit for the use of this device.
This was approved subject to the following:
1) A report to the CAEG in 9 months, outlining initial experience including complications
2) Please provide the hospital number of the first planned case to the CAEG Administrator
Best wishes for its successful development, sent on behalf of CAEG members.
421
Yours sincerely
Janet Deane (CAEG Administrator) Senior Research, Audit and Effectiveness Officer Care Pathways Coordinator c.c. Cath Barton Dr G Russell Dr R Perry General Manager, CCM Medical Director Deputy Medical Director
Dr D Todd Dr S Modi Dr M Hall Service Line Clinical Lead Cons Cardiologist / EPS Cons Cardiologist / EPS
Karen Wafer Ann-Marie Francis Catheter Lab Manager SLM
Dr M Ledson Dr M Jackson Chair, CAEG Exec Dir. Research and Informatics
422
Appendix 8.2: Patient information leaflet for RF in HOCM
423
Appendix 9: Published papers
Appendix 9.1: Current status of NSRT
Current Status of non-surgical septal reduction therapy for hypertrophic obstructive cardiomyopathy R M Cooper, A Shahzad, R H Stables Interventional Cardiology August 2013; 5; 4; 427-439
424
Appendix 9.1 Current status of non-surgical septal reduction therapy Review article current status NSRT.pdf
Appendix 9.2: Historical control ASA results
Alcohol septal ablation for hypertrophic obstructive cardiomyopathy: Safe and apparently efficacious but does reporting of aggregate outcomes hide less-favourable results, experienced by a substantial proportion of patients? R M Cooper, A Shahzad, J McShane, R H Stables Journal of Invasive Cardiology 2015;27(7)
425
Appendix 9.2 ASA for HOCM, does aggregate reporting hide leff favourable results.pdf
Appendix 9.3: ICE vs TTE
Intra-cardiac echocardiography in alcohol septal ablation: A comparative prospective study against trans- thoracic echocardiography R M Cooper, A Shahzad, J Newton, N Vejlstrup, A Axelsson, V Sharma, O Ormerod, R H Stables Echo Research and practice March 2015 2:9-17
426
Appendix 9.3 ICE vs TTE in ASA for HOCM.pdf
Appendix 9.4: CT planning for ASA improves infarct location and patient outcome
Computed Tomography angiography planning identifies the target vessel for optimum infarct location and improves clinical outcome in alcohol septal ablation for hypertrophic obstructive cardiomyopathy R M Cooper, S Binukrishnan, A Shahzad, J Hasleton, U Sigwart, R H Stables
427
Eurointervention 2016 Nov 29 Epub ahead of print
http://www.pcronline.com/eurointervention/ahead-of-print/JAA_EIJ_EIJ-D-16-00159/computed-tomography- angiography-planning-identifies-the-target-vessel-for-optimum-infarct-location-and-improves-clinical- outcome-in-alcohol-septal-ablation-for-hypertrophic-obstructive-cardiomyopathy.html
pdf not available at time of submission
Appendix 9.5: Intervention in HCM; Patient selection and emerging techniques in NSRT
Intervention in HCM; Patient selection, procedural approach and emerging techniques in alcohol septal ablation R M Cooper, A Shahzad, R H Stables Echo Research and Practice March 2015 2:R25-R35
428
Appendix 9.5 - Intervention in HCM, patient selection, procedural approach and emerginf techniques in alcohol septal ablation.pdf
Appendix 9.6: RF ablation to treat LVOT gradients in HOCM
Radiofrequency ablation of the interventricular septum to treat outflow tract gradients in hypertrophic obstructive cardiomyopathy: A novel use of CARTOsound technology to guide ablation R M Cooper, S Modi, A Shahzad, J Hasleton, J DiGiovanni, M Hall, D Todd, R H Stables
429
Europace . 2016 Jan;18(1):113-20
Appendix 9.6 RF in HOCM CARTOsound.pdf
430
Appendix 10: Permissions for reproduction:
Appendix 10.1: Permissions for review article
Permissions to reproduce content from ‘Current status of non-surgical septal reduction therapy’ from Future Medicine group
431
Appendix 10.2: Permissions for historical patient group reporting
432
Appendix 10.3: Permissions from Echo Research and Practice
ERP is an open access journal, on consulting with them no permissions to reproduce any content in my thesis were required.
Appendix 10.4: Permissions for CT guidance of ASA paper
Response from Eurointervention for request to reproduce paper in my thesis:
433
Appendix 10.5: Permissions for RF ablation in HOCM
434
Appendix 11: Required Imperial/NHLI courses attended
Introduction to statistic using SPSS – 1 credit.
7/2/13
435
Stats from scratch: An introduction to statistics – 1 credit
4/3/13
Stats for research: Intermediate statistics – 1 credit
4/3/13
Effective poster presentations – 1 credit
Wolfson Education Centre, Hammersmith
23/10/13
Appendix 12: Awards and presentations based on work from this thesis
Appendix 12.1: Awards:
Appendix 12.1.1 National Heart and Lung Institute Postgraduate Research Day 2013:
Best poster – Winner
436
Appendix 12.1.12: British Cardiovascular Intervention Society 2014
Young Investigator Award – Runner up
437
Appendix 12.2: National podium presentations:
1. Liverpool National Cath Lab study day 2012
a. Early experience of CT to guide ASA
2. Liverpool National Cath Lab study day 2013
a. CT angiography to guide ASA improves infarct location
3. Institute of Cardiovascular Medicine and Science Research symposium 2014
a. Non-Surgical Septal Reduction Therapy in Hypertrophic Obstructive Cardiomyopathy
4. British Cardiovascular Intervention Society:: Young investigator award: January 2015
a. CT angiography planning improves localisation of infarct and patient outcome in
alcohol septal ablation.
5. British society of echocardiography: Advanced imaging: HCM: May 2015
a. Emerging techniques in NSRT for ASA: The use of CT and RF ablation
6. Institute of Cardiovascular Medicine and Science Research symposium 2015
a. Non-Surgical Septal Reduction Therapy in Hypertrophic Obstructive Cardiomyopathy
7. British society of cardiovascular imaging : HCM: May 2016
a. CT planning improves infarct location and patient outcome in ASA
Appendix 12.3: International / National Poster Presentations:
438
Computed Tomography Angiography Planning Improves Localisation of Iatrogenic Infarct and Procedural
Success in Alcohol Septal Ablation for HOCM
British Cardiac Society Annual meeting 2015
Radiofrequency Ablation Of The Interventricular Septum To Treat Outflow Tract Gradients In HOCM: Novel Use
Of Cartosound Technology To Guide Ablation
British Cardiac Society Annual meeting 2015
Intra-cardiac Echocardiography (ICE) to guide Alcohol Septal Ablation (ASA) in Hypertrophic Obstructive
Cardiomyopathy (HOCM): A prospective comparison study against trans-thoracic echocardiography (TTE)
British Cardiac Society Annual meeting 2015
Computed Tomography Angiography Planning Improves Localisation of Iatrogenic Infarct and Procedural
Success in Alcohol Septal Ablation for HOCM
American Congress of Cardiology 2015
Radiofrequency Ablation Of The Interventricular Septum To Treat Outflow Tract Gradients In HOCM: Novel Use
Of Cartosound Technology To Guide Ablation
American Congress of Cardiology 2015
439
The Liverpool experience of alcohol septal ablation (ASA) in hypertrophic obstructive cardiomyopathy (HOCM):
12 year follow up
British Cardiac Society annual conference 2013
Early experience of computed tomography angiography to characterise septal vascular anatomy prior to alcohol septal ablation for hypertrophic obstructive cardiomyopathy
British Cardiac Society annual conference 2013
Computed tomography angiography to describe septal vascular anatomy prior to alcohol septal ablation (ASA) for hypertrophic obstructive cardiomyopathy.
Society for Cardiac Computed Tomography Annual Conference 2013
Computed tomography angiography to describe septal vascular anatomy prior to alcohol septal ablation (ASA) for hypertrophic obstructive cardiomyopathy.
Brazilian Society of Radiology annual Congress 2013
Medium term survival following alcohol septal ablation (ASA) for hypertrophic obstructive cardiomyopathy
(HOCM)
European Society of Cardiology Congress 2013
Alcohol septal ablation (ASA) for hypertrophic obstructive cardiomyopathy (HOCM) - follow up to 12 years
440
Transcatheter Cardiovascular Therapeutics San Francisco 2013
Computed tomography angiography provides intricate detail of septal arterial anatomy to plan alcohol septal ablation for hypertrophic obstructive cardiomyopathy
Transcatheter Cardiovascular Therapeutics San Francisco 2013
Early experience of CT angiography in planning Alcohol Septal Ablation (ASA) for Hypertrophic Obstructive
Cardiomyopathy (HOCM)
Transcatheter Cardiovascular Therapeutics San Francisco 2013
Lessons learned from CT angiography in planning for alcohol ablation for HOCM
ICMS research day 2012
441