Upload
others
View
2
Download
0
Embed Size (px)
Citation preview
Ripple-AT Study: A multicenter and randomized study comparing 3D mapping techniques during atrial
tachycardia ablations
Short Title: Randomized Ripple AT
Vishal Luther PhD MRCP1, Sharad Agarwal MD2, Anthony Chow PhD FRCP3, Michael Koa-Wing PhD
MRCP1, Nuno Cortez-Dias PhD MD4, Luís Carpinteiro PhD MD4, João de Sousa PhD MD4, Richard
Balasubramaniam MD5, David Farwell MD6, Shahnaz Jamil-Copley PhD MRCP7, Neil Srinivasan PhD
MRCP3, Hakam Abbas BSc3, James Mason BSc2, Nikki Jones BSc5, George Katritsis MBChB BSc1,
Phang Boon Lim PhD MRCP1, Nicholas S. Peters MD FHRS1, Norman Qureshi PhD MRCP1, Zachary
Whinnett PhD MRCP1, Nick Linton PhD MRCP1, Prapa Kanagaratnam PhD FRCP1
Institutional Affiliations:
1. Imperial College Healthcare, London, UK;
2. Papworth Hospital, Cambridge, UK.
3. Barts Heart Centre, London, UK;
4. Hospital de Santa Maria, Lisbon, Portugal;
5. Royal Bournemouth & Christchurch Hospital, Bournemouth, UK;
6. Essex Cardiothoracic Centre, Basildon, UK;
7. Nottingham University Hospital, Nottingham, UK
Address for correspondence: Professor Prapa Kanagaratnam, Department of Cardiology, Mary
Stanford Wing, St. Marys Hospital, Imperial College Healthcare NHS Trust, London W2 1NY, United
Kingdom. Telephone: +44 (0) 203 312 3783 Email: [email protected]
This abstract received the “Eric N. Prystowsky Fellows Clinical Research Award” at Heart Rhythm
Society 2019. This abstract presentation also received a “Late Breaking Research Award” at Asia-
Pacific Heart Rhythm Society 2018.
Journal Subject Terms: Arrhythmia, Atrial Fibrillation, Electrophysiology
1
1
2
34
5
6789
101112
13
1415161718192021
22
232425
2627282930
3132
33
ABSTRACT
Background: Ripple Mapping (RM) is an alternative approach to activation mapping
of atrial tachycardia (AT) that avoids electrogram (EGM) annotation. We tested
whether RM is superior to conventional annotation based local activation time (LAT)
mapping for AT diagnosis in a randomized and multicenter study (NCT02451995).
Methods: Patients with AT were randomized to either RM or LAT mapping using the
CARTO3v4 CONFIDENSE™ system. Operators determined the diagnosis using the
assigned 3D mapping arm alone, before being permitted a single confirmatory
entrainment manuever if needed. A planned ablation lesion set was defined. The
primary endpoint was AT termination with delivery of the planned ablation lesion set.
The inability to terminate AT with this first lesion set, the use of more than one
entrainment manuever, or the need to crossover to the other mapping arm were
defined as failure to achieve the primary endpoint.
Results: 105 patients from 7 centres were recruited with 22 patients excluded due to
premature AT termination, non-inducibility or LAA thrombus. 83 patients (RM=42,
LAT=41) completed mapping and ablation within the two groups of similar
characteristics (RM vs LAT: prior ablation or cardiac surgery n=35 (83%) vs n=35
(85%) p=0.80). The primary endpoint occurred in 38/42pts (90%) in the RM group
and 29/41 (71%) in the LAT group (p=0.045). This was achieved without any
entrainment in 31/42pts (74%) with RM and 18/41pts (44%) with LAT (p=0.01). Of
those patients who failed to achieve the primary endpoint, AT termination was
achieved in 9/12pts (75%) in the LAT group following crossover to RM with
entrainment, but 0/4 pts (0%) in the RM group crossing over to LAT mapping with
entrainment (p=0.04).
Conclusion: RM is superior to LAT mapping on the CARTO3v4 CONFIDENSE™
system in guiding ablation to terminate AT with the first lesion set, and with reduced
entrainment to assist diagnosis.
Keywords: Mapping, Atrial Tachycardia, Ablation, Ripple Mapping, CARTO
2
343536
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
INTRODUCTION
The incidence of atrial tachycardia (AT) has increased with the rising numbers of
atrial fibrillation (AF) ablation.1 3D mapping and ablation for patients with
symptomatic ATs is proven to be superior to pharmacological therapy and is the
recommended first line treatment.2
The current gold standard approach to AT mapping relies on the annotation of local
activation time (LAT) of each intra-cardiac electrogram (EGM) collected within a pre-
specified window of interest (WOI) based on the tachycardia cycle length (TCL).3
This approach can be prone to error, especially in areas of low voltage related to
prior ablation, surgery or myopathy where EGM annotation can be challenging.4 3D
mapping systems continue to develop algorithms to overcome the challenges of LAT
mapping in these areas without addressing the fundamental limitations related to
annotation. Ripple Mapping (RM) is now an established alternative approach to
activation mapping on the CARTO3 CONFIDENSE™ platform (Biosense Webster,
Inc) that does not require electrogram annotation or a window of interest.5-7
Furthermore, as Ripple Maps can be played over a color display of bipolar voltage, it
can demonstrate how activation navigates through areas of low voltage.8 Recent
non-randomized studies have suggested that RM can improve diagnostic accuracy
compared to standard LAT mapping approaches.8, 9
In this study we prospectively randomized patients with AT to test whether RM is
superior to LAT mapping on the CARTO3 version4 (v4) CONFIDENSE™ platform in
diagnosing the mechanism of AT to guide the delivery of radiofrequency ablation.
3
6768
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
METHODS
Study designThe Ripple-AT study (clinicaltrials.gov ID: NCT02451995) was a multi-centre,
prospective, and randomized study comparing RM vs LAT guided AT ablation.
Centers experienced in the 3D mapping of ATs and using the CARTO3v4
CONFIDENSE™ module (which includes CARTO Ripple Mapping as a module)
volunteered for participation. All operators received formal training in the use of RM,
and undertook at least 4 consecutive RM guided AT ablations prior to commencing
the randomized study. All procedures were performed with informed consent and
institutional approval at each site was granted for this study. The data that support
the findings of the study are available from the corresponding author upon
reasonable request.
Patients were recruited from those referred for paroxysmal/persistent AT ablation.
Subjects were excluded if the documented ECG was consistent with typical cavo-
tricuspid isthmus dependent flutter. Patients were block randomized into an
unblinded 1:1 mapping design of either RM or LAT mapping to guide ablation.
Randomisation was performed via a sealed envelope.
A decapolar catheter was placed in the coronary sinus (CS) and a suitable CS
reference was selected. Burst pacing down to 200ms from different CS poles was
used to induce AT if in sinus rhythm at the start. The CS activation pattern in AT was
used to decide on the first chamber to map. CARTO3v4 CONFIDENSE™ (Biosense,
Inc). was used for mapping with a Lasso Nav (Biosense, Inc) or Pentaray Nav
(Biosense, Inc) as per operator discretion. A point density to color the entire
geometry with a color threshold of 5mm was targeted. Criteria for including points on
the map using CONFIDENSE Continuous Mapping™ included a cycle length stability
within a 5% range of the TCL, an electrode position stability within 2 mm, an LAT
stability filter at 3 milliseconds and tissue proximity to the endocardial surface. While
industry support was offered to assist with operating the CARTO system, they
provided no assistance in interpreting the activation map to which each patient was
randomized.
4
9091
9293
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
Randomized to Ripple Mapping: RM presents the entire EGM as a white moving bar on the surface geometry with the
height correlating to the EGM voltage amplitude at that time point. Every bipolar
EGM from each point is aligned relative to the reference signal, and the Ripple map
of all points is then played through time, with activation understood as the movement
of bars traverses from one area to the next.
The approach to RM undertaken by operators in this arm of the study is presented in
Table 1. Details of how this approach was developed can be seen in previous
published studies.7, 8 EGM deflections above 0.07mV were displayed as Ripple bars,
but could be adjusted down to 0.03mV to review very low voltage activation. The
bars were clipped above 0.30mV to allow easier visualisation. The map was played
as a continuous loop spanning two tachycardia cycle lengths. EGM data from the
same tachycardia cycle is presented as a Ripple Map by aligning the EGMs
according to their corresponding reference EGM from the same cycle.
In this arm of the study, during the geometry and electro-anatomical map acquisition,
the bipolar voltage map was displayed and the LAT map remained hidden to all lab
staff. The voltage display was set empirically to 0.30mV-0.30mV (such that tissue
<0.30mV was colored red, and tissue >0.30mV was colored purple). The voltage
color scale was manually reduced from 0.30mV to identify the surface voltage
threshold supporting wave-fronts of Ripple bars. In doing so, areas supporting ripple
wavefronts were displayed in purple, and the remaining non-conducting areas
without ripple bars in red (e.g. 0.15mV-0.15mV, where tissue below 0.15mV
appeared red, and tissue above 0.15mV appeared purple). Every potential ripple
activation wave front was studied to determine the AT mechanism. In complex cases
the ‘design line’ tool dialogue box was activated, which allowed the operator to
manually draw out the path of the visualised activation direction across the entire
geometry chamber (as small white arrows), in order to plan the ablation approach.
Randomized to LAT Mapping:Conventional activation mapping involves the measurement of the “local activation
time” of each EGM. This denotes the numerical difference in timing between a
component of a sampled EGM and a stable reference signal, the values of which are
plotted on the 3D map according to a rainbow color bar that spans the interval of the
mapped tachycardia. Annotation of LAT requires a timing “window of interest” (WOI)
5
121122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148149
150
151
152
153
to ensure that EGMs from the same beat are compared. A WOI is defined by
specifying a timing interval both before and after the reference point. The position of
the reference point in relation to the window is arbitrary. The color red is used to
represent the earliest measurement of LAT within the specified window, and purple
the latest. In a re-entrant tachycardia, the concept of “early” and “late” is a misnomer,
because any given area of activation will have other sites that are activated before or
after that location, and the red area of earliest activation will be adjacent to a purple
area of latest activation (i.e. early meets late).
The CARTO3v4 CONFIDENSE™ module offers several features to enhance LAT
mapping. This includes Wavefront™ annotation, an automated algorithm that
assigns timing annotation at the maximal negative distal unipolar derivative within
the time-window spanned by the corresponding bipolar EGM. In addition, there is an
algorithm to enable an automated WOI recompute, which allows the WOI and the
subsequent LAT of each projected point to be re-calculated at any time during map
acquisition. There is also a “map consistency” filter, which determines the
consistency of each measured LAT point relative to its neighbouring points, and
highlights outlier points with LAT’s very different to its surrounding neighbours for
review and deletion/re-annotation as required.
In this arm of the study, during the geometry and electro-anatomical map acquisition,
the LAT and bipolar voltage maps were displayed and the Ripple Map remained
hidden to all lab staff. The WOI was set from the start, spanning between 90-100%
of the TCL and either around the reference signal, or the surface p wave.10
Operators were able to adjust the WOI as required using the automated recalculation
function for the complete electroanatomic map. Automated annotation of activation
timing was assigned using the CONFIDENSE Wavefront™ algorithm in all cases. In
the absence of a sharp negative slope >0.03mV in unipolar signal, a grey square
was projected as the system could not automatically assign LAT. After all points
were collected, the Map Consistency filter was used as required. The static LAT map
was then visualised as a 2 color propagation map (red and blue). Additional manual
LAT re-annotation or deletion was allowed at sites where propagation remained
unclear on the map. In a re-entrant circuit, an “early meets late” algorithm could be
applied to interpolate between early and late sights on the map considered re-
entrant.
6
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
DiagnosisThe AT diagnosis was considered macro-re-entrant if it travelled continuously around
the mapped atrial chamber (where head met tail), whilst small loop circuits were
contained within a single plane (i.e. anterior wall). Both focal/localised re-entrant
circuits appeared to emanate from a small focus, spreading radially away and often
involving low voltage and fractionated EGMs at the break-out site.
Ablation:A diagnosis was made using the mapping arm assigned and a strategy for ablation
was planned. Power controlled (25–35 W) radiofrequency energy was delivered
(Stockert ۛ 70 RF generator, Biosense Webster) through an irrigated ablation catheter
(ThermoCool SmartTouch, Biosense Webster) at the putative target for AT ablation.
EndpointsThis study protocol was applied only to the first map collected (i.e. AT1). The primary
endpoint of the study was “AT termination with the first ablation set”. The “first
ablation set” encompassed all the ablation lesions delivered to target the AT based
on the studied activation map. For example, if the mapping approach suggested LA
roof dependency, the “first ablation set” would have been a roof line(s) and AT
termination following this ablation set alone would achieve the primary endpoint. If
AT1 changed to AT2 with ablation (defined as a sustained change in CS activation or
cycle length), this was also considered to have met the primary endpoint.
Cases where the tachycardia 1) terminated to sinus rhythm or 2) degenerated to
atrial fibrillation/alternating ATs, before mapping was completed, were excluded from
analysis. The objective of this study was to assess the diagnostic efficacy in the
acute setting only, therefore no long-term data was collected.
EntrainmentIn this study protocol, operators were asked to make a diagnosis with RM or LAT
alone, and entrainment was restricted in order to allow a fair comparison of the two
mapping approaches in isolation. Thus, if the operator was confident of the AT
diagnosis and ablation strategy from the 3D map, no entrainment was performed. A
single entrainment manuever was allowed where the operator had some diagnostic
uncertainty, in order to validate the diagnosis before ablation. For example, if the
mapping approach suggested peri-mitral macro-re-entry, entrainment from CS
proximal and CS distal would have been permissible as a single confirmatory
7
187188
189
190
191
192
193194
195
196
197
198199
200
201
202
203
204
205
206
207
208
209
210
211212
213
214
215
216
217
218
219
entrainment manuever. If the diagnosis remained uncertain, or if ablation failed to
terminate AT, this was considered a failure to meet the primary endpoint of the
study. Crossover to the other mapping arm of the study and additional entrainment
mapping was open as per operator discretion.
StatisticsA prior non-randomized study demonstrated a diagnostic yield using RM compared
to LAT maps of 90% vs. 65%.9 Assuming the same effect size, in order to detect a
difference between arms with 80% power at the 5% two-tailed significance level, this
required at least 40 patients in each group. A sample size of at least 80 patients was
targeted.
Categorical variables were expressed as percentages. Continuous variables were
expressed as mean ± 1 standard deviation for parametric data and/or median (lower
quartile – upper quartile) for non-normal data. Categorical data were analysed using
either a Fishers Exact Test or Chi Squared test where appropriate. Unpaired data
were analysed using a Student’s T-test for parametric data and Mann Whitney U test
for non-normal data. A two-sided p value was determined where applicable and a
value of p≤0.05 was considered significant.
8
220
221
222
223
224225
226
227
228
229
230
231
232
233
234
235
236
RESULTS
Figure 1 summarises the study design and number of patients in each arm. A total of
105 patients were recruited from 7 centers (6 UK; 1 Portugal) participating in this
study. Of these patients, 22 were excluded. This was due premature termination of
AT whilst mapping (n=6), AT non-inducibility (n=9), or degeneration to alternating
ATs/atrial fibrillation prior to the delivery of ablation (n=4). Left atrial appendage
thrombus on TOE seen at the start of the procedure (n=3) resulted in immediate
procedural cessation.
The remaining 83 patients completed mapping and AT ablation as per the assigned
randomized arm, and these were used for subsequent analysis. This included 42
patients in the RM group and 41 patients in the LAT group.
The baseline demographics, prior ablation history, and mapping details between the
two groups were similar, and is summarised in Table 2. (RM vs LAT: (65±9yrs) vs
(65±10yrs) p=0.78; prior atrial ablation n=32 (76%) vs n=33 (80%) p=0.83 – this
included prior pulmonary vein isolation n=30 (71%) vs n=29 (71%); linear
lesions/complex fractionated atrial electrogram ablation n=20 (48%) vs n=12 (29%)
p=0.13; prior right atrial ablation only n=2 (5%) vs n=4 (10%) p=0.65. Patients with
prior cardiac surgery without ablation included (RM vs LAT) n=3 (7%) vs n=2 (5%)
p=0.66. Therefore, the total number of patients with potential iatrogenic scar causing
AT (prior ablation or cardiac surgery) included RM n=35 (83%) vs LAT n=35 (85%)
p=0.80.
The median AT cycle lengths were similar between groups (RM vs LAT: 267ms (LQ
240, UQ 298) vs 260ms (LQ 231, UQ 278) p=0.20). Multipolar catheters (Lasso and
Pentaray) were used for mapping in almost all cases (a linear Smarttouch catheter
was used in the non-coronary cusp in a focal AT) with similar median point densities
between groups (RM vs LAT: 2681points (LQ 1722, UQ 3647) vs 2219 points (LQ
1453, UQ 3232) p=0.44. The Pentaray was the most preferred mapping catheter
(RM vs LAT: n=27 (64%) vs n=26 (63%) p=0.93).
Figure 2 summarises the key endpoints of the study. The primary endpoint of
termination of AT with the first ablation set occurred in 38/42pts (90%) in the RM
group and 29/41 (71%) in the LAT group (p=0.045). This was achieved with the 3D
9
237238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
map alone, i.e. without entrainment in 31/42pts (74%) in the RM group and 18/41pts
(44%) in the LAT group (p=0.01). AT1 changed to AT2 with ablation in 6pts in both
groups, and the AT2 map confirmed a different mechanism and ablation target in all
cases. 1pt in each group failed to terminate with ablation where entrainment
confirmed the same diagnosis as the map. Where the primary endpoint was not met
(i.e. those without a diagnosis based on the 3D mapping method, or where the first
ablation set failed to terminate AT), AT termination was achieved in n=9/12 (75%)
patients crossing from LAT to a combination of RM and entrainment mapping, and in
n=0/4 (0%) patients crossing from RM to a combination of LAT and entrainment
mapping (p=0.04).
Where the primary endpoint was met, left sided ATs were predominantly identified in
both groups (RM vs LAT: n=31/38 (82%) vs n=24/29 (83%) p=0.90). Macro/small
loop re-entrant mechanisms were more common in both groups (RM vs LAT: 27/38
(71%) vs 22/29 (76%) p=0.87; with the remaining diagnosed as focal/localized re-
entrant (RM vs LAT: 11/38 (29%) vs 7/29 (24%)).
There were no procedural complications in either groups in this study.
Figure 3 (and supplementary video 1) illustrates a case of a patient randomized to
Ripple Mapping. A complex LA circuit circumnavigating around an island of probable
scar <0.15mV was observed on the anterior wall, determined by the absence of
Ripple wave-fronts through it. A second circuit was also observed around the mitral
annulus (supplementary video 1). Both these circuits were dependent on a narrow
and slowly conducting isthmus between the inferior border of this island and the
mitral annulus, and transecting this isthmus terminated tachycardia. This case
demonstrates how Ripple Mapping is used to define putative scar on the map by
functional assessment, and then exploiting the concurrent display of activation and
scar to define the optimal ablation site.
There were 4/42 cases which failed to meet the primary endpoint in the RM group.
One case was peri-mitral (confirmed with entrainment) where MI block could not be
achieved. One case terminated with ablation after further diagnostic mapping. One
case targeted a septal source without effect. The last case involved extensive low
voltage substrate/scar from multiple prior ablations where a roof substrate was
targeted without effect.
10
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
Figures 4, 5, 6 & 7 (and corresponding supplementary videos 2, 3, 4) highlight cases
of patients randomized to LAT mapping that did not meet the primary endpoint. The
cases in Figures 4 and 5 are from patients with prior surgical Atrial Septal Defect
repair, and illustrate issues related to LAT color interpolation. In Figure 4, several
early and late sites colored red and purple are seen on the final LAT display. As
these early and late sites were in close proximity, the operator applied the “early
meets late” tool to interpolate the colors between these sites and produce a uniform
pattern. When propagation was played, the operator observed continuous rotation
around scar on the anterolateral wall suggestive of small loop re-entry. However,
split potentials were identified along the circuit, implying wave-front collision, not
observed on the propagation display. Given this uncertainty, the operator crossed
over to RM. The Ripple Map demonstrated that a line of conduction block prevented
small loop re-entry. Thus, the appearance of re-entry was false, and a consequence
of over-interpolation creating the impression of re-entry.
In figure 5, the full color coded spectrum was observed within a small area
collocating with an area of low voltage consistent with the lateral surgical cannulation
site. The propagation map demonstrated wavefront turning around this site, and
sampled EGMs were fractionated, leading the operator to again consider small loop
re-entry. This site was ablated without effect. The patient crossed over to RM, where
no rotational activity was seen, rather splitting of activation on either side of this
region of probable scar. Post procedure, a band of false color interpolation spanning
the full rainbow spectrum was appreciated on the LAT map between the apparent
early and late sites on the map. This created the appearance of a slowly moving
backward wave-front and this false appearance of wavefront turning on the
propagation display. As RM does not require a WOI and does not interpolate, this
error was avoided.
Figure 6 depicts an iatrogenic AT post extensive AF ablation where the LAT WOI
had been set equally around the CS reference EGM. The subsequent activation
pattern appeared focal in origin from within the LAA. Given the absence of EGMs of
interest, this diagnosis was uncertain, and entrainment revealed this site was outside
the circuit. RM did not demonstrate focal activation from the LAA, rather activation
breakout from probable scar near the posterior floor. Post procedure, after multiple
arbitrary post hoc adjustments of the LAT WOI, it revealed a similar activation
11
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
pattern as seen with RM. This case highlights how setting the WOI is arbitrary and
can be misleading in complex cases. As RM does not require a WOI, this limitation
can be avoided.
Despite all attempts to optimise, the LAT map in figure 7 was considered
uninterpretable, with multiple early and late sites. This was mapped from a patient
with prior mitral valvuloplasty and extensive low voltage atrial tissue. EGMs sampled
within the map were frequently long duration and multicomponent. Despite being in
close proximity, sites have been labelled as early and late due to overlapping EGMs
spanning a large portion of the set WOI. This case highlights the challenge of having
to annotate a single LAT to represent multi-component fractionated EGMs. Ripple
Mapping does not annotate, rather it presents all EGM components in its entirety.
Following crossover to RM, a breakout of activation from an extensive area of low
voltage along the posterior roof was appreciated and a decision to perform a
posterior box lesion resulted in AT termination.
12
334
335
336
337
338
339
340
341
342
343
344
345
346
347
DISCUSSION
This study is the first prospective, multicenter and randomized study comparing 3D
activation mapping techniques. We show that operators experienced with LAT
mapping on the CARTO3v4 CONFIDENSE™ platform had a higher rate of AT
termination using Ripple Mapping, and achieved this with less reliance on
entrainment support.
The effectiveness of mapping tachycardia activation using LAT is proven. Published
studies document ~85% success rates in AT termination using this approach1. LAT
mapping has seen recent advances, including automated annotation to the
maximum negative unipolar EGM derivative within the period of bipolar activation,
and its incorporation with high point density. However, most of these published
studies have combined LAT with entrainment mapping such that the efficacy of LAT
mapping in isolation remains unknown. Furthermore, these studies did not report
whether ablation had been delivered at multiple incorrect sites prior to eventual AT
termination. This is the first study to measure the efficacy of LAT mapping in
isolation, without entrainment, and following the delivery of only the first ablation set.
This study demonstrated acute AT termination with first lesion set in 71% with LAT
mapping, and in only 44% without entrainment. The figures highlight three sources
for error in relation to LAT mapping that likely explain this ~30% failure rate,
including: (1) incorrect color interpolation; (2) window of interest errors and (3) mis-
annotation.
Interpolation algorithms assign the average activation time between mapped points
in order to display an interpretable propagation pattern on the assumption that
activation is uniform; however, these estimates of timing can be misleading,
especially in areas of conduction delay or block, as seen in Figure 4. “Backward
wave-fronts” are a specific interpolation error observed in macro-re-entrant circuits,
and caused the problems demonstrated in Figure 5. These occur at sites where
“early” and “late” do not quite meet, and interpolation of colors between these
apparent early and late sites occur, resulting in a slowly moving wave-front in
reverse to the true direction of activation.
13
12
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
A windows of interest (WOI) is required to ensure that EGM LATs from the same
cycle are compared. However, the process of setting this window is arbitrary, with
different color coded activation patterns generated depending on the setting, as
shown in Figure 6. Whilst methods to standardise this approach around the surface
p-wave have been considered, they are only applicable in non-focal mechanisms,
which is not known at the start of the case.10 Furthermore, in diseased tissue with
prolonged conduction times, no matter how the window is set, being limited to a
single cycle can lead to very late activating sites being erroneously displayed as
early in the WOI with respect to the next cycle of activation, resulting in more than
one early site on the map.11 Ripple Mapping is the only contact-mapping system
which currently reviews more than a single tachycardia cycle length. The CARTO3v4
LAT maps used in this study present their WOI as a color bar. A WOI color wheel
has also been proposed to solve the challenges of setting a WOI. Rotation of this
wheel has the equivalent effect of sliding the WOI without causing a full map re-
compute. In principle, this can avoid some errors related to the WOI as considered
above. However, several studies have reported on a high prevalence of small
pseudo-re-entrant circuits from continuously rotating the wheel, some of which were
misdiagnosed as localized/small loop re-entry and inappropriately ablated.12, 13
Mis-annotation of LAT can lead to a complete change in the color coded pattern. The
advent of high point density acquisition with algorithms that filter out points with
inconsistent timings in relation to neighbouring LAT measurements have reduced
annotation errors. However, these errors remain prevalent, particularly in areas of
low voltage containing multicomponent EGMs as in Figure 7. Manual checking and
re-annotation of LAT when >2000points are collected is time consuming and
impractical during clinical procedures.
Whilst all 3D-mapping systems continue to develop algorithms to overcome these
limitations to LAT mapping, Ripple Mapping offers a completely alternative activation
mapping approach. RM presents activation information without the need for
annotation of activation time or setting of a window of interest, and does not
interpolate between unmapped sites.5 Patients randomized to RM in this study
achieved AT termination in 90% with the first ablation set (p=0.045).
14
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
This superiority was partly attributable to avoiding these errors associated with LAT
mapping. It was also consequent to a unique means of studying activation in areas
of low voltage and scar. There remains no consensus on a voltage parameter to
differentiate between active and non-conducting tissue (i.e. either true scar from
fibrosis or areas of functional block dependant on the wave-front direction and atrial
rate) using endocardial mapping. LAT maps apply an arbitrary pre-set voltage
threshold to display scar, and display areas below this threshold as grey tags to
blank the color-coded map.14 With RM, displaying activation wave-fronts on a colored
voltage map enables a novel approach to defining this voltage parameter that
differentiates electrically active myocardium from non-conducting tissue (voltage
thresholding).8 The example in Figure 3 is a case where only the common isthmus of
the dual-loop tachycardia needed to be ablated. This was possible because of the
simultaneous display of activating and non-conducting myocardium during
tachycardia that helped determine the optimal site for ablation. This can be
particularly helpful in peri-mitral tachycardias in which conventional mitral isthmus
lines can be avoided.
Entrainment enhances our electrophysiological understanding of the AT circuit prior
to ablation, and this study does not advocate entrainment avoidance.15 However,
entrainment does have limitations in areas of low voltage due to pacing latency, non-
capture, and can cause degeneration to AF.16 In this study, patients randomized to
RM underwent significantly less entrainment than those in the LAT group (p=0.01),
as operators felt more confident to ablate the AT based on the Ripple Map alone.
Entrainment appears to be essential to LAT mapping to help overcome the core
limitations of this approach, whilst with RM it was more supportive by confirming a
diagnosis.
RM looks very different to conventional LAT activation maps. There is a learning
curve, and each operator (who had extensive experience in LAT mapping and
entrainment) received 2-3hrs of formal training in RM on a workstation with example
cases. This was followed by 4 consecutive clinical cases where operators made a
diagnosis with RM first, with the improved diagnostic efficacy of RM already apparent
at that stage.9 These operators did not need technical support from industry
representatives to make these diagnoses. We would consider similar training to be
essential for other operators aiming to replicate the results seen in this study.
15
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
Whilst this study was specific to mapping EGMs within the atria, RM is also suited to
mapping EGMs within the ventricle. We and others have shown how RM can be
used to follow late potentials through channels of slow conduction that might support
re-entry.17, 18 Ripple Mapping to study VT and substrate during sinus rhythm has a
very different workflow due to the main challenge being differentiating local and far-
field EGMs.
Limitations:The objective of this study was to assess the diagnostic efficacy of each mapping
technique in the acute setting, and does not consider the best approach to achieve
long-term freedom from AT. The analysis of long-term outcomes of patients
randomised to each mapping arm would require a different protocol that prohibited
crossover to the other mapping arm during the entire case, and mandated post-
ablation inducibility testing (which was performed at operator discretion in this
protocol) and treatment of any subsequent ATs using the same mapping approach.
These cases were all performed on the CARTO3v4 CONFIDENSE™ platform, and
some of the findings might not apply to other mapping software.
Approximately 20% of patients recruited were not included in the study analysis due
to tachycardia termination before mapping was started/completed. We excluded
these patients as the ATs were not stable and therefore our endpoint of acute
termination during ablation would not have been robust. The best approach to
achieve acute success and long-term benefit in this group of patients needs further
study.
16
96
97
98
99
100
101
102103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
CONCLUSION
This prospective, randomized and multi-center study demonstrates that Ripple
Mapping is superior to LAT mapping on the CARTO3v4 CONFIDENSE™ platform in
achieving acute atrial tachycardia termination using the first delivered ablation set,
with reduced reliance on entrainment to assist diagnosis.
Sources of Funding: The study was funded by a clinical study grant from Biosense Webster
and a British Heart Foundation Clinical Research Training Fellowship award (no. FS/15/12/31239).
Disclosures: Imperial Innovations holds Intellectual Property relating to Ripple Mapping on behalf
of PK and NL, who have also received royalties from Biosense Webster. PK, NL, SJC and VL have
received consulting fees with respect to Ripple Mapping from Biosense Webster. The remaining
authors have nothing to disclose.
17
12
3
4
5
6
78
9
10
11
12
131415
REFERENCES
1. Chae, S.; Oral, H.; Good, E.; Dey, S.; Wimmer, A.; Crawford, T.; Wells, D.; Sarrazin, J. F.; Chalfoun, N.; Kuhne, M.; Fortino, J.; Huether, E.; Lemerand, T.; Pelosi, F.; Bogun, F.; Morady, F.; Chugh, A., Atrial tachycardia after circumferential pulmonary vein ablation of atrial fibrillation: mechanistic insights, results of catheter ablation, and risk factors for recurrence. J Am Coll Cardiol 2007, 50, 1781-7.2. Zhang, X. D.; Gu, J.; Jiang, W. F.; Zhao, L.; Zhou, L.; Wang, Y. L.; Liu, Y. G.; Liu, X., Optimal rhythm-control strategy for recurrent atrial tachycardia after catheter ablation of persistent atrial fibrillation: a randomized clinical trial. Eur Heart J 2014, 35, 1327-34.3. Jais, P.; Shah, D. C.; Haissaguerre, M.; Hocini, M.; Peng, J. T.; Takahashi, A.; Garrigue, S.; Le Metayer, P.; Clementy, J., Mapping and ablation of left atrial flutters. Circulation 2000, 101, 2928-34.4. Del Carpio Munoz, F.; Buescher, T. L.; Asirvatham, S. J., Teaching points with 3-dimensional mapping of cardiac arrhythmias: teaching point 3: when early is not early. Circ Arrhythm Electrophysiol 2011, 4, e11-4.5. Linton, N. W.; Koa-Wing, M.; Francis, D. P.; Kojodjojo, P.; Lim, P. B.; Salukhe, T. V.; Whinnett, Z.; Davies, D. W.; Peters, N. S.; O'Neill, M. D.; Kanagaratnam, P., Cardiac ripple mapping: a novel three-dimensional visualization method for use with electroanatomic mapping of cardiac arrhythmias. Heart Rhythm 2009, 6, 1754-62.6. Jamil-Copley, S.; Linton, N.; Koa-Wing, M.; Kojodjojo, P.; Lim, P. B.; Malcolme-Lawes, L.; Whinnett, Z.; Wright, I.; Davies, W.; Peters, N.; Francis, D. P.; Kanagaratnam, P., Application of ripple mapping with an electroanatomic mapping system for diagnosis of atrial tachycardias. J Cardiovasc Electrophysiol 2013, 24, 1361-9.7. Koa-Wing, M.; Nakagawa, H.; Luther, V.; Jamil-Copley, S.; Linton, N.; Sandler, B.; Qureshi, N.; Peters, N. S.; Davies, D. W.; Francis, D. P.; Jackman, W.; Kanagaratnam, P., A diagnostic algorithm to optimize data collection and interpretation of Ripple Maps in atrial tachycardias. Int J Cardiol 2015, 199, 391-400.8. Luther, V.; Linton, N. W.; Koa-Wing, M.; Lim, P. B.; Jamil-Copley, S.; Qureshi, N.; Ng, F. S.; Hayat, S.; Whinnett, Z.; Davies, D. W.; Peters, N. S.; Kanagaratnam, P., A Prospective Study of Ripple Mapping in Atrial Tachycardias: A Novel Approach to Interpreting Activation in Low-Voltage Areas. Circ Arrhythm Electrophysiol 2016, 9, e003582.9. Luther, V.; Cortez-Dias, N.; Carpinteiro, L.; de Sousa, J.; Balasubramaniam, R.; Agarwal, S.; Farwell, D.; Sopher, M.; Babu, G.; Till, R.; Jones, N.; Tan, S.; Chow, A.; Lowe, M.; Lane, J.; Pappachan, N.; Linton, N.; Kanagaratnam, P., Ripple mapping: Initial multicenter experience of an intuitive approach to overcoming the limitations of 3D activation mapping. J Cardiovasc Electrophysiol 2017, 28, 1285-1294.10. De Ponti, R.; Verlato, R.; Bertaglia, E.; Del Greco, M.; Fusco, A.; Bottoni, N.; Drago, F.; Sciarra, L.; Ometto, R.; Mantovan, R.; Salerno-Uriarte, J. A., Treatment of macro-re-entrant atrial tachycardia based on electroanatomic mapping: identification and ablation of the mid-diastolic isthmus. Europace 2007, 9, 449-57.11. Ju, W.; Yang, B.; Chen, H.; Zhang, F.; Gu, K.; Yu, J.; Li, M.; Yang, G.; Cao, K.; Chen, M., Mapping of focal atrial tachycardia with an uninterpretable activation map after extensive atrial ablation: tricks and tips. Circ Arrhythm Electrophysiol 2014, 7, 598-604.12. Luther, V.; Sikkel, M.; Bennett, N.; Guerrero, F.; Leong, K.; Qureshi, N.; Ng, F. S.; Hayat, S. A.; Sohaib, S. M.; Malcolme-Lawes, L.; Lim, E.; Wright, I.; Koa-Wing, M.; Lefroy, D. C.; Linton, N. W.; Whinnett, Z.; Kanagaratnam, P.; Davies, D. W.; Peters, N. S.; Lim, P. B., Visualizing Localized Reentry With Ultra-High Density Mapping in Iatrogenic Atrial Tachycardia: Beware Pseudo-Reentry. Circ Arrhythm Electrophysiol 2017, 10.
18
1617
1819202122232425262728293031323334353637383940414243444546474849505152535455565758596061626364
13. Bradfield, J. S.; Huang, W.; Tung, R.; Buch, E.; Okhovat, J. P.; Fujimura, O.; Boyle, N. G.; Gornbein, J.; Ellenbogen, K. A.; Shivkumar, K., Tissue voltage discordance during tachycardia versus sinus rhythm: implications for catheter ablation. Heart Rhythm 2013, 10, 800-4.14. Kalman, J. M.; VanHare, G. F.; Olgin, J. E.; Saxon, L. A.; Stark, S. I.; Lesh, M. D., Ablation of 'incisional' reentrant atrial tachycardia complicating surgery for congenital heart disease. Use of entrainment to define a critical isthmus of conduction. Circulation 1996, 93, 502-12.15. Pathik, B.; Lee, G.; Nalliah, C.; Joseph, S.; Morton, J. B.; Sparks, P. B.; Sanders, P.; Kistler, P. M.; Kalman, J. M., Entrainment and high-density three-dimensional mapping in right atrial macroreentry provide critical complementary information: Entrainment may unmask "visual reentry" as passive. Heart Rhythm 2017, 14, 1541-1549.16. Vollmann, D.; Stevenson, W. G.; Luthje, L.; Sohns, C.; John, R. M.; Zabel, M.; Michaud, G. F., Misleading long post-pacing interval after entrainment of typical atrial flutter from the cavotricuspid isthmus. J Am Coll Cardiol 2012, 59, 819-24.17. Luther, V.; Linton, N. W.; Jamil-Copley, S.; Koa-Wing, M.; Lim, P. B.; Qureshi, N.; Ng, F. S.; Hayat, S.; Whinnett, Z.; Davies, D. W.; Peters, N. S.; Kanagaratnam, P., A Prospective Study of Ripple Mapping the Post-Infarct Ventricular Scar to Guide Substrate Ablation for Ventricular Tachycardia. Circ Arrhythm Electrophysiol 2016, 9.18. Xie, S.; Kubala, M.; Liang, J. J.; Yang, J.; Desjardins, B.; Santangeli, P.; van der Geest, R. J.; Schaller, R.; Riley, M.; Supple, G.; Frankel, D. S.; Callans, D.; Pac, E. Z.; Marchlinski, F.; Nazarian, S., Utility of ripple mapping for identification of slow conduction channels during ventricular tachycardia ablation in the setting of arrhythmogenic right ventricular cardiomyopathy. J Cardiovasc Electrophysiol 2019, 30, 366-373.
19
65666768697071727374757677787980818283848586
87
88
89
Ripple Mapping of an Atrial Tachycardia – Diagnostic Steps
Mapping • Multipolar mapping - Lasso/PentaRay
• Colour threshold 5mm: point density = no grey areas (usually >2000points).
• Only “points projected” displayed (visualisation set-up)
• Annular points tagged and mitral/tricuspid annulus removed
Ripple Setup • Ripple Preferences: Show bars above - 0.07mV; Clip bars above - 0.30mV
• Selected point viewer: 2 cycle lengths played
Surface Voltage thresholding • Bipolar voltage map displayed - set empirically at to 0.3mV – 0.3mV on custom settings
• Play Ripple Map and reduce voltage limits in 0.05mV steps (i.e. 0.25mV – 0.25mV; continued down to 0.05mV – 0.05mV as required) until no ripple wave-fronts in red areas (non-activating tissue)
• Ripple bar wave-fronts should only visible in purple areas
Identifying mechanism • Study ripple activation in small patches of geometry. Use design lines with arrowheads to mark these wave-fronts throughout the entire chamber
• Follow the arrows backwards to identify focal source or re-entrant circuit.
• Reduce “Show bars above” to 0.03mV - study activation in areas of interest, to locate 1) earliest bar of focal source or 2) narrowest isthmus of re-entrant circuit.
Table 1: Protocol for Atrial Tachycardia diagnosis using Ripple Mapping
20
90
91
Table 2: Baseline demographics and mapping details
RM - Ripple Mapping; LAT – Local Activation Time; RA – right atrium; LA – left atrium; PVI – pulmonary vein isolation; CFAE – complex fractionated atrial electrogram; AT – atrial tachycardia; denotes 3 missing values; denotes 4 missing values.
21
Baseline CharacteristicsAssigned Group
p valueRM LAT
No of Patients 42 41
Prior atrial ablation or cardiac surgery (%) 83 85 1.00
Age (yrs) 65±9 65±10 0.78
Prior atrial ablation (%)(RA or LA) 76 80 0.80
Prior PVI (%) 71 71 NA
Prior LA substrate ablation (e.g. lines/CFAE) (%) 48 29 0.12
Prior RA ablation only (%) 5 10 0.43
Prior cardiac surgery without ablation (%) 7 5 1.00
AT cycle length/ ms, median 267* 260* 0.2
Pentaray use (%) 64 63 0.93
Collected Points, median 2681* 2219 0.44
92
9394
95
96
97
98
99
100
101
102
103
104
105106
107
108
109
110
111
112
113114115
116
Figure Legends:
Figure 1: Study design and patient numbers
Figure 2: Results of acute ablation outcomes.
Figure 3: Randomized to Ripple Mapping. Voltage thresholding defines the critical
isthmus (See also supplementary video 1).
Patient with an idiopathic left AT (262ms). (A). Upper panel - bipolar voltage map set
to 0.30mV– 0.30mV. Tissue with voltage <0.30mV is displayed in red, and >0.30mV
in purple. A large area colored red was seen on the anterior wall. Wave-fronts of
ripple bars (yellow circle) were seen in areas of the voltage map colored red. (B) The
surface voltage limits were reduced until 0.15mV-0.15mV (i.e. red=<0.15mV;
purple=>0.15mV) – here no Ripple bars were seen in areas colored red. At this
voltage setting, an island of red tissue was seen on the anterior wall with ripple bars
circumnavigating clockwise around it (marked out by white arrows). Middle panel –
Ripple markings were sampled around this island, and their corresponding EGMs
spanned the ATCL (3 consecutive EGM cycles shown). Shown in the accompanying
video is a second circuit travelling counter-clockwise around the mitral annulus. Both
circuits were dependent on a narrow isthmus between the inferior border of this
island and the mitral annulus. Lower panel - transecting this critical isthmus with
ablation terminated tachycardia.
Figure 4: Randomized to LAT mapping. Interpolation creates the false impression of
small loop re-entry (See also supplementary video 2).
Patient with a prior surgical ASD repair mapped in AT (260ms) in the right atrium. A)
LAT map (modified AP) with the WOI set either side of the reference signal (-130ms
to +130ms). More than one region of early and late were seen (labelled). B) The
“early meets late” tool was applied (80%-default) and color interpolation between
these early and late sites were filled (dark red). Scar was highlighted as areas with
22
117118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
peak bipolar voltage amplitude ≤0.03mV and colored grey. All the colors of the LAT
spectrum were seen to progress around a patch of scar on the anterolateral wall
suggestive of small loop re-entry. However, split potentials were identified along the
circuit, implying wave-front collision in the circuit. C) The Ripple Map demonstrated
that a line of conduction block prevented small loop re-entry around this patch of
scar, as evident by a line of double potentials along it (double yellow lines). In fact,
ripple bars rotated around this line of block, creating a “larger loop” re-entry. The
pattern of activation is depicted by white design lines, and can be viewed in the
corresponding video. Entrainment supported this observation – the post pacing
interval was long within the interpolated false small loop circuit seen by the LAT map,
and shortened moving superiorly into larger loop circuit seen with the Ripple Map.
Figure 5: Randomized to LAT mapping. Interpolation creates a false backward
wavefront (See also supplementary video 3).
Another patient with prior surgical ASD repair was mapped in AT (270ms) in the right
atrium. A) The operator observed the full rainbow color spectrum around the lateral
RA corresponding to B) an area of low voltage <0.30mV on the bipolar voltage map
and considered this small loop re-entry based on C) the presence of long and
fractionated EGMs within the circuit and the appearance of wavefront
turning/curvature (although not a complete rotation) on the corresponding LAT
propagation map (see video). However, ablation at this site (red circular VisiTag
disks) was ineffective. D) There was no evidence of wavefront curvature on the
Ripple Map, rather splitting of wavefronts on either side of this region of scar. E) A
band of false color interpolation spanning the full rainbow color spectrum was
evident between the early and late sites on the LAT map (labelled) that created the
appearance of a slowly moving back ward wave-front and the impression of wave-
front turning.
Figure 6: Randomized to LAT mapping. Varying LAT activation patterns when
changing the window of interest (See also supplementary video 4).
23
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
Patient with post AF ablation (PVI, Roof + Mitral isthmus lines, CFAE ablation) AT
(258ms). Upper panel: The WOI was set either side of the reference signal to span
90% of the TCL. The LAT map, seen in AP (A) and PA (B) depicts earliest focal
activation (red) within the left atrial appendage. C) The corresponding voltage map
(PA) demonstrates extensive substrate <0.1mV (colored red). Unlike the LAT map,
RM revealed a breakout of activation from probable scar near the posterior floor
(yellow star), considered the exit of a region of slow conduction supporting a macro-
re-entrant circuit around the mitral annulus (white design lines). D) Adjustment of the
LAT WOI to (-18ms) to (+201ms) revealed the same breakout activation pattern (red
isochrone collocating with yellow star) as suggested by RM, but considered the
mechanism focal rather than part of a macro-re-entrant circuit. AT terminated in the
mid coronary sinus (purple star, snapshot catheter projection applied), presumably
along an epicardial connection of a mitral annular circuit.
Figure 7: Randomized to LAT mapping: Uninterpretable activation pattern
Patient with post-surgical (mitral valvuloplasty) left AT (230ms). The activation
pattern (PA) was uninterpretable with multiple “early” and “late” sites (labelled). The
EGMs from 2 immediately adjacent points are presented. These bipolar EGMs are
low voltage and multicomponent, and span a large portion of the TCL. Their
corresponding unipolar EGMs are also displayed. Despite being in immediate
proximity, the system has annotated one (left) as “early” and the other (right) as
“late” with respect to the set WOI, based on the sharpest unipolar dV/dT within this
long period of bipolar activation. Repeated occurrences of this event have resulted in
this uninterpretable appearance.
24
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
Figure 1
25
201202
203
204
205
206
207
Figure 2
26
208209
210
211
Figure 3
27
212
213
214
215
Figure 4
28
216
217
218
219
Figure 5:
29
220
221
222
223
224
225
Figure 6
30
226
227
228
Figure 7
31
229
230
231
232
233
234
235