- Email: [email protected]
Electrical parameters with His-bundle pacing: Considerations for automated programming
Journal Pre-proof Electrical parameters with His bundle pacing: considerations for automated programming Neasa Starr, MD, Nicolas Dayal, MD, Giulia Domenichini, MD, Carine Stettler, RN, Haran Burri, MD PII:
S1547-5271(19)30721-0
DOI:
https://doi.org/10.1016/j.hrthm.2019.07.035
Reference:
HRTHM 8103
To appear in:
Heart Rhythm
Received Date: 26 June 2019
Please cite this article as: Starr N, Dayal N, Domenichini G, Stettler C, Burri H, Electrical parameters with His bundle pacing: considerations for automated programming, Heart Rhythm (2019), doi: https:// doi.org/10.1016/j.hrthm.2019.07.035. This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition of a cover page and metadata, and formatting for readability, but it is not yet the definitive version of record. This version will undergo additional copyediting, typesetting and review before it is published in its final form, but we are providing this version to give early visibility of the article. Please note that, during the production process, errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain. © 2019 Published by Elsevier Inc. on behalf of Heart Rhythm Society.
Electrical parameters with His bundle pacing: considerations for automated programming
Short title: Considerations for automated His pacing programming
Neasa Starr, MD, Nicolas Dayal, MD, Giulia Domenichini, MD, Carine Stettler, RN, Haran Burri, MD
From: Cardiology Department, University Hospital of Geneva, Geneva, Switzerland
Address for correspondence: Prof Haran Burri Cardiac pacing Unit, Cardiology Department University Hospital of Geneva Rue Gabrielle Perret Gentil 4 1205 Geneva, Switzerland e-mail: [email protected] phone: +41 22 372 72 00 Word count: Abstract 227 words. Total (including abstract, text, references tables, figure legends) 4609 words. Disclosures: H.B. has received speaker fees and research grants from Abbott, Biotronik, Boston Scientific, Medtronic and Microport. N.D has received fellowship support and speaker fees from Medtronic. Funding: None.
1
1
ABSTRACT
2
Background: Programming of His bundle pacing may be challenging because
3
current implantable pulse generators are not specifically designed for this pacing
4
modality.
5
Objectives: Our aim was to evaluate electrical parameters in order to propose
6
preset programming options with different configurations.
7
Methods: Data were collected from 50 patients with His pacing leads connected
8
to various ports (atrial, right ventricular or left ventricular) of pacemakers and
9
defibrillators during a detailed device interrogation which included capture
10
thresholds with various pacing vectors, measurement of timing intervals, and
11
performance of automatic threshold algorithms.
12
Results: His bundle-pacing thresholds were significantly lower during unipolar
13
pacing compared to bipolar and extended bipolar polarities. However, current
14
drain was offset due to lower impedance. The His pace - right ventricular sensed
15
intervals were measured at 40-150ms (mean 85±25ms) with the longest delays
16
in patients with uncorrected right bundle branch block and selective His capture.
17
This has implications for ventricular safety pacing windows (which were
18
inactivated without evidence of crosstalk) and delays to minimize unnecessary
19
ventricular backup pacing (which was also affected by refractory periods). The
20
measured intervals also impacted the performance of automatic threshold
21
algorithms, which performed differently depending on which port the His lead
22
was connected to and did not distinguish between His and myocardial capture.
23
Conclusion: Our report provides data which could serve to configure automated
24
programming settings to simplify management of His bundle pacing.
25
Keywords: His bundle pacing; capture thresholds; programming; timing; optimization.
2
26 27
INTRODUCTION
28
His bundle pacing (HBP) is becoming increasingly adopted as it preserves
29
physiological conduction along the His-purkinje system, unlike traditional right
30
ventricular (RV) pacing which results in dysynchrony and a gradual decline in
31
left ventricular (LV) function.1 HBP has not only been used in lieu of RV pacing,2
32
but also in lieu of biventricular pacing for cardiac resynchronization therapy
33
(CRT)3, 4 and for His-optimized CRT (HOT-CRT).5, 6
34
There are currently no implantable pulse generators (IPGs) that include
35
timing cycles specific to HBP, and depending on which port the His lead is
36
connected to, different advanced settings need to be considered to safely
37
optimize programming.5 Another consideration is battery longevity, as capture
38
thresholds with HBP are often higher than with RV pacing.7 Pacing polarity has
39
been shown to significantly affect LV thresholds with CRT8 but there are few data
40
regarding HBP,9 and no information on whether automatic capture management
41
algorithms yield valid results for HBP. As a precautionary measure or in the
42
instance of implantable cardioverter defibrillators (ICDs), backup ventricular
43
leads may be implanted and can result in unnecessary pacing due to restrictions
44
of timing cycles with current devices. For example, with His leads connected to
45
the atrial (A) port, if sensing from the backup RV lead occurs within 110ms of His
46
pacing, ventricular safety pacing (VSP) will automatically be delivered. To avoid
47
this unnecessary current drain, the feature must be inactivated, but may expose
48
the patient to consequences of crosstalk.
49
The aim of this study was to evaluate various electrical parameters in
50
order to facilitate management, optimize programming for HBP and prolong
3
51
battery longevity whilst minimizing risk to the patient. To do this we
52
investigated effect of pacing polarity on HBP thresholds, measured His pace-RV
53
sensed (HP-RVS) intervals, evaluated risk of crosstalk due to inactivation of VSP,
54
evaluated accuracy of automatic threshold measurement algorithms and efficacy
55
of prolonging atrioventricular (AV) and interventricular (VV) delays to avoid
56
unnecessary ventricular backup pacing.
57 58
METHODS
59
We included consecutive patients followed-up at the device clinic of the
60
University Hospital of Geneva, Switzerland, implanted with a Medtronic
61
(Minneapolis, MN, USA) 3830 lead on the His bundle, connected to a Medtronic
62
IPG. HBP was confirmed by transitions in QRS morphology during threshold tests
63
according to current definitions.10 The study was approved by the institutional
64
ethics committee and all patients gave informed consent to participate.
65
The device check consisted of a His capture threshold measurement
66
starting at 8V in each available polarity at a pulse width of 0.4ms and 1ms, and
67
decrementing output in 0.25V steps (except for the initial steps which are only
68
possible between 8V, 6V and 5.5V) while recording a 12-lead ECG. Unipolar
69
pacing is only available in pacemakers in Medtronic devices. An extended bipolar
70
configuration was tested in CRT devices with the His lead connected to the LV
71
port.
72
(unipolar/bipolar/extended bipolar configurations, each tested at 0.4ms and
73
1ms). The test was used to determine the various types of His capture and their
74
respective thresholds i.e. non-selective His capture (NSHC), selective His capture
75
(SHC) and correction of bundle branch block (BBB), using a standard 12-lead
These
patients
with
a
CRT
pacemaker
had
6
datasets
4
76
ECG. Pacing impedance was measured for each polarity, and current drain at
77
threshold output calculated by V/R.
78
To analyze the accuracy of the automatic threshold measurement
79
algorithms in determining the His threshold, the stored values collected by the
80
device were compared to the in-office thresholds obtained at the same pulse
81
width and same pacing polarity. Generally, the automatic threshold value was
82
obtained at 1 a:m the same day as the device check. An automatic threshold
83
within 0.5V of the in-office value was deemed accurate.
84
For each patient with the His lead connected to the A or LV port and a
85
backup RV lead, the His paced-RV Sensed (HP-RVS) interval was measured using
86
digital calipers on the Medtronic programmer at 100mm/s sweep speed, for each
87
form of His capture Data on percentages of pacing from the His and ventricular
88
leads was recorded, in order to assess the effectiveness of programming pacing
89
intervals to minimize ventricular pacing.
90
For the patients with the His lead on the A channel, a test for cross-talk
91
was completed by increasing the output to maximum on the His lead (8V at 1ms
92
in unipolar) and programming maximum sensitivity for the RV channel (0.15mV
93
or 0.45 mV depending on whether the IPG was a defibrillator or a pacemaker).
94
Using these ‘worst-case’ scenarios, evidence of RV sensing of the after-potential
95
of the high output His pacing spike was evaluated by analyzing the timing of the
96
RVS event on the marker channel compared with the RV electrogram (crosstalk
97
being deduced when the RVS marker occurred earlier than the potential of the
98
captured RV electrogram).
99 100 5
101
Statistical analysis
102
Analysis was performed used the SPSS v.25 (Armonk, NY) software. Data
103
showed a normal distribution according to histogram analysis and the
104
Kolmogorov-Smirnoff test. Data were reported as mean±SD unless specified
105
otherwise. Differences between groups was assessed by the Student’s t-test. A P
106
value <0.05 was considered significant.
107 108
RESULTS
109
In total 50 patients were included in the study (see table 1). All ICD leads
110
were implanted at the apex, and all RV pacing leads on the interventricular mid-
111
septum. A backup ventricular lead was present in 45/50 patients, which allowed
112
measurement of HP-RVS in these patients. The time from implantation to device
113
follow-up was 3.2±4.1 months.
114 115
ECG findings
116
Overall at 8V/1ms, NSHC was present in 47 patients and SHC in three
117
patients. Of the patients with NSHC, a transition to SHC at a lower output was
118
observed in 24 patients and to myocardial capture in 23 patients (i.e SHC was
119
present in 27/50 patients overall). An extended bipolar pacing mode was tested
120
in the 20 patients with the His lead connected to the LV port. Of these, 5 (25%)
121
patients (all implanted with pacemakers) had evidence of anodal capture from
122
the RV lead (occurring at a threshold of 4.2±0.9V/1ms and at 5.6±1.5V/0.4ms).
123
An example of anodal capture in the extended bipolar mode is shown in figure 1.
124 125
Capture thresholds, lead impedance and current drain at different pacing polarities 6
126
Comparison of electrical parameters between unipolar and bipolar pacing
127
configurations are shown in table 2. A separate analysis was performed for the
128
12 CRT-P patients with the His lead connected to the LV port, in whom unipolar,
129
bipolar and extended bipolar pacing configurations (His lead tip cathode to RV
130
ring anode) were directly comparable. Even though unipolar pacing
131
configurations resulted in significantly lower capture thresholds compared to
132
bipolar and extended bipolar pacing, the lower impedances offset the calculated
133
current drain at threshold output. There were no significant differences in
134
electrical parameters between bipolar and extended bipolar configurations.
135 136
HP-RVS delay
137
Results from the 45 patients with a His lead in either the A or LV port are
138
shown in figure 2 (separate values were recorded in a same patient during NSHC
139
and SHC when pacing at different outputs – with a total of 65 different
140
measurements). We analyzed results based upon presence of RBBB (12 patients
141
had underlying RBBB in intrinsic rhythm and three of the six pacemaker-
142
dependant patients had paced QRS morphologies showing uncorrected RBBB).
143
The intervals ranged from 40 to 150ms (mean 85±25ms), with the longest
144
intervals in patients with SHC and uncorrected RBBB (P<0.01 for all
145
comparisons). With a reduction in pacing output, 10 patients with RBBB had
146
NSHC which transitioned to SHC. The HP-RVS interval subsequently lengthened
147
from 89±32ms to 115±25ms (P=0.001) in these patients (and did not change
148
significantly in the five patients with RBBB who transitioned from NSHC to
149
myocardial capture only). In patients without RBBB, the HP-RVS intervals were
7
150
significantly shorter during NSHC than during SHC (71±18ms vs. 88±12ms
151
respectively P=0.005).
152 153 154
Crosstalk and VSP
155
No evidence of crosstalk was observed during in-office testing in any of
156
the 25 patients with the His lead in the atrial port. Of these patients, 10 had
157
complete heart block at the time of testing, of whom six were due to AV node
158
ablations. None of the patients reported symptoms suggestive of crosstalk such
159
as syncope or malaise.
160 161
Percentages of unnecessary ventricular pacing
162
Of the 25 patients with a His lead connected to the atrial port, 18 had a
163
paced AV interval programmed to 180ms with the intent to suppress
164
unnecessary RV pacing (otherwise, three patients were programmed to VVIR
165
mode and four with a short AV interval for HOT-CRT). The AV interval of 180ms
166
successfully reduced the percentage of RV pacing after HBP to 0.8 ±0.8 % (range
167
0.1-2.8%).
168
In eight patients with the His lead connected to the LV port of a CRT
169
device, the interventricular (VV) interval was programmed to 80ms (the
170
maximum programmable value in Medtronic devices) in an attempt to avoid
171
unnecessary RV pacing. All these patients had 100% RV pacing following His
172
paced events, despite the fact that 2 patients had a HP-RVS interval less than the
173
programmed 80ms (40ms and 70ms).
174 8
175
Automatic threshold measurement algorithms
176
The automatic capture management algorithm was programmed to
177
“monitor” in the 25 patients with the His lead on the atrial port. At follow-up,
178
only one patient had automatic threshold readings for the His lead, the result of
179
which were accurate when compared to the in-office threshold. Of note, this
180
patient had a HP-RVS interval of 140ms and SHC with an uncorrected RBBB.
181
Another observation was that when the His lead was connected to the A port, not
182
only were the automatic thresholds not available for the His lead, but they were
183
also unavailable for all the other leads connected to the IPG.
184
For the five patients with the His lead on the RV port, each had myocardial
185
capture as the lowest capture form. Of these patients, four had automatic
186
threshold measurements within 0.5V of the in-office myocardial capture
187
threshold. The fifth patient had an automatic threshold of >2.5V/0.4ms with an
188
in-office myocardial threshold of 2.75V/0.4ms (an exact threshold value is not
189
given in these instances).
190
Regarding the 20 patients in the LV group, 17 had automatic threshold
191
measurements which were programmed to “monitor”. Of these patients, 10 had
192
reported values reflective of myocardial thresholds and seven reflective of SHC.
193
All automatic threshold values were within 0.5V of in-office readings.
194 195 196
DISCUSSION
197 198
The main findings of our study are that 1) As expected, unipolar pacing
199
configuration yielded lower capture thresholds compared to bipolar and 9
200
extended bipolar configurations, but the impact on calculated current drain was
201
mitigated because of reduced pacing impedance 2) HP-RVS delays vary between
202
40 and 150ms (average 85±25ms) with the longest delays associated with SHC
203
and uncorrected RBBB 3) The risk of AV crosstalk with His leads plugged in the
204
atrial port is low and no adverse effects were observed with inactivation of VSP
205
4) Unnecessary RV pacing was prevented in patients with the His lead in the
206
atrial port by programming a paced AV interval of 180ms, but was unable to be
207
avoided when the His lead was connected to the LV port 5) Automatic threshold
208
measurement algorithms are dependent upon the port the His lead is connected
209
to and do not distinguish between His and myocardial capture.
210
As previously shown with LV leads,8 capture thresholds and lead
211
impedances with His leads are significantly lower with unipolar compared to
212
bipolar pacing polarities. However in contradiction with data from Su et al.9 we
213
found that unipolar thresholds and impedances are also lower compared to the
214
extended bipolar configuration in the subset of our patients with the His lead
215
connected to the LV port of CRT devices. Therefore, due to lower pacing
216
impedance, the calculated current drain was in fact higher with unipolar pacing.
217
However, an advantage with unipolar pacing is a visible pacing spike, which
218
facilitates identification of HBP on ECG tracings. Ideally, pulse generators should
219
automatically calculate battery longevity based upon programmed output, taking
220
into account the pacing impedance (as is already the case with some models).
221
An observation we describe for the first time with HBP in an extended
222
bipolar pacing configuration is anodal capture from the RV lead (well described
223
with CRT). This may create confusion during threshold testing, as it results in an
224
additional transition in QRS morphology. 10
225
The HP-RVS delay is of major importance for device programming. This
226
delay will depend upon a number of factors, such as SHC vs NSHC, correction of
227
RBBB, and possibly RV lead position (see figure 2). In patients with SHC, it is the
228
sum of conduction duration through the His-Purkinje system (usually 40-50ms)
229
and intra-myocardial conduction from the Purkinje exit to the lead implantation
230
site. In patients with NSHC, the HP-RVS delay will depend on the shorter of either
231
this pathway or intra-myocardial conduction from the His lead to the RV lead.
232
In patients who have the His lead connected to the atrial port, RVS will
233
usually occur during the VSP window (95-110ms depending upon the
234
manufacturer), leading to unnecessary RV pacing. VSP may be inactivated, but
235
this potentially exposes the patient to asystole in case of crosstalk, which was
236
however absent in all our patients tested in “worst case” scenarios, but who had
237
a post atrial pacing ventricular blanking period (PAVB) set to a non-
238
programmable values of 30ms in the devices used in this study. Alternative
239
algorithms to VSP also exist e.g. Boston Scientific devices (Marlborough, MA,
240
USA) rely upon retriggerable noise windows.
241
In order to avoid unnecessary RV pacing in patients with a His lead in the
242
atrial port, a paced AV interval of 180ms was found to be effective (with an
243
average of <1% of RV pacing, corresponding most probably to cycles with loss of
244
His capture). It is therefore unnecessary to program unduly long AV intervals,
245
which also carry the risk of pacing in the vulnerable period of the T-wave in case
246
of R-wave undersensing. The main issue was found to be in patients with the His
247
lead in the LV port in whom RV pacing was delivered 100% of the time. This was
248
not only due to the maximum interventricular delay being limited to 80ms, but
249
also to ventricular blanking (which was set by default to 200-230ms in the 11
250
devices in our study) and committed RV pacing after delivery of pacing by the LV
251
channel. A solution in this configuration would be to design refractory periods
252
able to consider RVS events which fall within 30-150ms of His paced events to be
253
His capture, allowing inhibition of RV pacing. Blanking during the first 30ms
254
following His pacing would avoid crosstalk (by analogy with the PAVB) and a
255
refractory period after 150ms would avoid T-wave oversensing. A schematic
256
representation of the proposed timing intervals is shown in figure 3.
257
The HP–RVS intervals also impacted the automatic capture management
258
algorithms. In patients with the His lead connected to the atrial port, an RVS
259
event falling within 110ms of pacing from the atrial channel will abort the
260
threshold test. Short intervals will also prevent automatic threshold testing of RV
261
and LV leads in these patients. The RV capture management algorithm of
262
Medtronic devices considers any RVS event that occurs in the 110ms window
263
following pacing to be V capture (i.e. it does not specifically detect the evoked
264
potential). Although this was not tested in our patient population, patients with
265
SHBC and uncorrected RBBB may have erroneous diagnosis of noncapture by the
266
algorithm due to long pace-sense intervals (falling outside the 110ms window).
267
In addition, as the AV interval of the test cycle is shortened to 10ms (to avoid AV
268
conduction), detection of a far-field P-wave on the His lead may result in
269
erroneous diagnosis of capture. The LV capture management algorithm of
270
Medtronic devices relies upon interventricular conduction delay and accurately
271
determined thresholds for loss of myocardial or His capture (whichever was
272
lowest) in all cases. Anodal capture in case of an extended bipolar configuration
273
of the His lead may however confound the results (as the local RV electrogram
274
occurs directly after pacing from the His lead). Devising an algorithm capable of 12
275
accurately measuring His capture is likely to be challenging. As recently
276
described, morphological analysis of near-field electrograms may be an option to
277
distinguish NSHC from SHC.11 Analysis of far-field electrogram morphology may
278
be helpful for distinguishing NSHC from myocardial capture, but this needs to be
279
evaluated. Capture threshold algorithms which yield accurate results in different
280
pacing configurations will obviously be of great interest, not only to adapt pacing
281
output and preserve battery longevity, but also to ensure patient safety. Changes
282
in thresholds with HBP may be even more unpredictable than with standard
283
ventricular pacing, underlining the need for this feature.
284
Many of the issues described in our report occur in the context of backup
285
ventricular leads. Although their utility in HBP has been questioned by
286
experienced implanters, they are unavoidable in patients with ICDs.
287
Furthermore, as requirement for lead revisions of His leads (e.g. due to loss of
288
capture) are frequent (close to 7% at 5 years7), safety is a concern, especially in
289
pacing-dependant patients. Backup ventricular leads are also useful in patients
290
planned for atrioventricular node ablation (which was indicated in over a
291
quarter of our patients) due to the risk of lead dislodgement or increase in His
292
capture thresholds.12 Backup ventricular leads also provide adequate sensing,
293
which may be an issue with His leads (with oversensing of A/His potentials or V
294
undersensing5). In addition, backup V leads also allow programming of lower
295
safety margins of the His lead, prolonging battery life, and provide the option of
296
HOT-CRT. Currently, the FDA only approves the 3830 lead for HBP when it is
297
connected to the ventricular port of single- and dual-chamber pacemakers.
298
However, clinical requirements extend beyond these configurations, and
299
hopefully approval will also be granted with a wider perspective in the future. 13
300 301 302 303
Study limitations
304
The sample size is relatively limited, especially in the subgroup of patients
305
with the His lead connected to the RV port. These patients were included to
306
evaluate the accuracy of the automatic threshold algorithm, but none of these
307
patients had SHC. Also, we were not able to test sensing issues in these patients
308
(e.g. A or His oversensing) as these are carefully evaluated at implantation at our
309
centre and only patients without such issues have His leads connected to the RV
310
port. We only included Medtronic devices in our analysis, so results may not
311
apply to other manufacturers (e.g. for automatic threshold algorithms).
312
Nevertheless, most of our data such as capture thresholds with different pacing
313
polarities and HP-RVS intervals should be applicable to all manufacturers.
314 315
CONCLUSIONS
316
Our results provide a framework for developing automated programming
317
settings which are tailed to meet the needs of HBP (see table 3). Incorporation of
318
predefined settings in IPGs for the different configurations of HBP would greatly
319
simplify programming and optimization of this therapy.
320 321
Acknowledgments: The authors would like to thank Mr Todd Sheldon
322
(Medtronic) for his answers to our technical queries.
323 324 14
325
References
326 327
1.
Lee MA, Dae MW, Langberg JJ, et al. Effects of long-term right ventricular
328
apical pacing on left ventricular perfusion, innervation, function and
329
histology. Journal of the American College of Cardiology 1994;24:225-
330
232.
331
2.
Abdelrahman M, Subzposh FA, Beer D, et al. Clinical Outcomes of His
332
Bundle Pacing Compared to Right Ventricular Pacing. Journal of the
333
American College of Cardiology 2018;71:2319-2330.
334
3.
Barba-Pichardo R, Manovel Sanchez A, Fernandez-Gomez JM, Morina-
335
Vazquez P, Venegas-Gamero J, Herrera-Carranza M. Ventricular
336
resynchronization therapy by direct His-bundle pacing using an internal
337
cardioverter defibrillator. Europace : European pacing, arrhythmias, and
338
cardiac electrophysiology : journal of the working groups on cardiac
339
pacing, arrhythmias, and cardiac cellular electrophysiology of the
340
European Society of Cardiology 2013;15:83-88.
341
4.
Sharma PS, Dandamudi G, Herweg B, et al. Permanent His-bundle pacing
342
as an alternative to biventricular pacing for cardiac resynchronization
343
therapy: A multicenter experience. Heart rhythm 2018;15:413-420.
344
5.
Burri H, Keene D, Whinnett Z, Zanon F, Vijayaraman P. Device
345
Programming for His Bundle Pacing. Circulation Arrhythmia and
346
electrophysiology 2019;12:e006816.
347
6.
Vijayaraman P, Herweg B, Ellenbogen KA, Gajek J. His-Optimized Cardiac
348
Resynchronization Therapy to Maximize Electrical Resynchronization.
349
Circulation Arrhythmia and electrophysiology 2019;12:e006934. 15
350
7.
Vijayaraman P, Naperkowski A, Subzposh FA, et al. Permanent His-bundle
351
pacing: Long-term lead performance and clinical outcomes. Heart rhythm
352
2018;15:696-702.
353
8.
Burri H, Schrage M, Morani G, et al. Effect of lead design and pacing vector
354
on electrical parameters of quadripolar coronary sinus leads: the RALLY-
355
X4 study. Pacing and clinical electrophysiology : PACE 2019;in press.
356
9.
Su L, Xu L, Wu SJ, Huang WJ. Pacing and sensing optimization of
357
permanent His-bundle pacing in cardiac resynchronization
358
therapy/implantable cardioverter defibrillators patients: value of
359
integrated bipolar configuration. Europace : European pacing,
360
arrhythmias, and cardiac electrophysiology : journal of the working
361
groups on cardiac pacing, arrhythmias, and cardiac cellular
362
electrophysiology of the European Society of Cardiology 2016;18:1399-
363
1405.
364
10.
Vijayaraman P, Dandamudi G, Zanon F, et al. Permanent His bundle
365
pacing: Recommendations from a Multicenter His Bundle Pacing
366
Collaborative Working Group for standardization of definitions, implant
367
measurements, and follow-up. Heart rhythm 2018;15:460-468.
368
11.
Saini A, Serafini NJ, Campbell S, et al. Novel Method for Assessment of His
369
Bundle Pacing Morphology Using Near Field and Far Field Device
370
Electrograms. Circulation Arrhythmia and electrophysiology
371
2019;12:e006878.
372
12.
Vijayaraman P, Subzposh FA, Naperkowski A. Atrioventricular node
373
ablation and His bundle pacing. Europace : European pacing, arrhythmias,
374
and cardiac electrophysiology : journal of the working groups on cardiac 16
375
pacing, arrhythmias, and cardiac cellular electrophysiology of the
376
European Society of Cardiology 2017;19:iv10-iv16.
377 378 379
17
380 381
Table 1. Patient Demographics:
Patient demographics (n=50) Male/female Age (years) Port of His lead connection (A/LV/RV) Device - DDD PM - CRT-P - CRT-D Indication for device implantation - AVB II - AVB III - Ablate and pace of AF - Slow AF - CRT indication with RBBB - Painfull LBBB Left ventricular ejection fraction Intrinsic QRS duration (ms) Intrinsic QRS morphology: - Normal - RBBB - LBBB - IVCD - Pacemaker dependent Hypertension Chronic renal failure Ischemic heart disease Permanent AF Diabetes
33/17 71±12 25/20/5
382 383
Values
and
384
eGFR<50ml/min/1.73m2.
385
IVCD=interventricular conduction delay; LBBB=left bundle branch block; LV=left ventricular;
386
RBBB=right bundle branch block; RV= right ventricular.
represent
numbers
of
A=atrial;
patients
23 15 12 10 14 13 6 6 1 0.52±0.12 119±28 18 13 3 10 6 24 17 19 25 12
AF=atrial
mean±SD. fibrillation;
Renal
failure
defined
AVB=atrioventricular
as
block;
387
18
388
Table 2. His capture thresholds, pacing impedance, and calculated current drain
389
at threshold output. There were no significant differences between bipolar and
390
extended bipolar configurations. Current drain calculated at capture threshold
391
voltage output using I=V/R. All comparisons are for paired analyses.
392 393
394
Total population
His lead connected to LV port of
(n=50)
CRT-P (n=12)
Unipolar
Bipolar
Unipolar
Bipolar
Extended
(n=38)
(n=50)
(n=12)
(n=12)
bipolar (n=12)
Threshold @0.4ms (V)
1.9±1.5
2.4±1.9
†
1.4±0.9
1.9±1.4
Threshold @1ms (V)
1.5±1.2
1.9±1.5
†
1.1±0.7
1.4±1.0
Impedance (Ω)
318±52
473±7
†
309±69
454±92
414±92
Current drain @0.4ms (mA)
6.2±5.0
5.4±4.5
‡
4.8±3.7
4.5±3.9
4.5±3.7
Current drain @1ms (mA)
4.8±4.0
4.0±3.2
‡
3.8±2.8
3.2±2.8
†
‡
Compared to unipolar: P<0.001 P<0.01
§
‡
§
†
‡
‡
1.9±1.3
§
1.3±0.8
†
§
§
3.2±2.2
P<0.05
395 396
19
397
Table 3. Framework for predefined settings for His bundle pacing. Parameter Common to all configurations Pacing polarity Unipolar and bipolar (also for A and LV ports of ICDs)
Comment
Lower thresholds with unipolar and visible pacing spike. However, current drain may be offset by lower impedance. Estimation of remaining Based upon pacing output Immediate calculation upon battery longevity and impedance, and history reprogramming to chose of percentage of pacing. configuration with least drain. Automatic threshold Analysis of EGM morphology To identify SHC (near-field algorithm EGM) or distinguish NSHC from myocardial capture (far-field EGM). His lead connected to the atrial port Sensing Inactivate Ventricular sensing ensured by ventricular lead. Ventricular safety Inactivate Crosstalk prevented by pacing “PAVB” of 30ms. Use retriggerable noise window (currently used for Boston Scientific devices). “Atrioventricular” Paced interval of 180ms Maximum delay of HP – RVS interval interval was 150ms. Automatic capture Detection of VS within 40Should be no interlock with threshold algorithm 150ms window after HP RV and LV automatic capture algorithms. His lead connected to the right ventricular port Post “AS” and “AP” Individually programmable To avoid A oversensing by the His lead. blanking period for a range of values in dual chamber and biventricular devices Automatic capture Extension of the detection Maximum delay of HP – RVS algorithm window beyond 110ms interval was 150ms in patients with uncorrected RBBB. His lead connected to the left ventricular port Interventricular interval Programmable up to >150ms To avoid unnecessary backup RV pacing “Interventricular” Shortened to 30ms refractory period Ventricular refractory Initiate 150ms after HP Avoid unnecessary backup period RV pacing (while protecting against T-wave oversensing) 20
Ventricular tachyarrhythmia counters
RVS within 150ms of HP not counted
Automatic threshold algorithm
Based upon HP-RVS conduction delay
Avoid misdiagnosis of tachyarrhythmia or lead dysfunction (e.g. due to algorithm comparing nearfield with far-field counts). Current Medtronic LVCM algorithm performs well (although does not distinguish between His and myocardial capture).
398 399
AS=atrial sens; AP=atrial pace; EGM=electrogram; HP= His paced; LV=left ventricle;
400
LVCM=left ventricular capture management; PAVB: post atrial pace ventricular
401
blanking; RBBB: right bundle branch block; RV: right ventricle; RVS: right ventricular
402
sense; SHC=selective His capture; VP=ventricular pace; VS=ventricular sense. Quotation
403
marks are used in instances where the parameter does not truly reflect the designation.
404
21
405
Figure 1. Illustration of anodal capture. QRS morphology during pacing from
406
ventricular lead implanted in the interventricular mid-septum (left panel) and
407
extended bipolar pacing at decrementing outputs (other four panels) from the
408
His lead connected to the left ventricular port of a biventricular pacemaker. The
409
transitions are readily seen in lead V1. RV=right ventricle; A=anodal capture;
410
C+=with correction of right bundle branch block; C- without correction of right
411
bundle branch block; NS=non-selective His capture; S= selective His capture
412
22
413
Figure 2. Schematic representation of effect of type of His capture on delays
414
between pacing from the His lead and sensing from the right ventricular lead
415
(values are mean±SD). The interval is longest in case of uncorrected right bundle
416
branch block (RBBB) with selective His capture (SHC) and is shortened by non-
417
selective His capture (NSHC) in this instance.
418 419
420 421 422 423 23
424
Figure 3. Proposed timing intervals for His pacing with His leads connected to
425
the atrial or left ventricular ports (parameters with the His lead connected to the
426
right ventricular port would follow standard settings and are not shown). AH=
427
atrio-His interval; AP=atrial pace marker; ARP=atrial refractory period; AS=
428
atrial sense marker; HP=His pace maker; HV=His-ventricular interval ;
429
PVARP=post-ventricular atrial refractory period; VA=ventriculo-atrial interval;
430
VH=ventriculo-His interval; VP=ventricular pace marker; VRP=ventricular
431
refractory period; VS= ventricular sense marker.
432
433
24
S1547-5271(19)30721-0
DOI:
https://doi.org/10.1016/j.hrthm.2019.07.035
Reference:
HRTHM 8103
To appear in:
Heart Rhythm
Received Date: 26 June 2019
Please cite this article as: Starr N, Dayal N, Domenichini G, Stettler C, Burri H, Electrical parameters with His bundle pacing: considerations for automated programming, Heart Rhythm (2019), doi: https:// doi.org/10.1016/j.hrthm.2019.07.035. This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition of a cover page and metadata, and formatting for readability, but it is not yet the definitive version of record. This version will undergo additional copyediting, typesetting and review before it is published in its final form, but we are providing this version to give early visibility of the article. Please note that, during the production process, errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain. © 2019 Published by Elsevier Inc. on behalf of Heart Rhythm Society.
Electrical parameters with His bundle pacing: considerations for automated programming
Short title: Considerations for automated His pacing programming
Neasa Starr, MD, Nicolas Dayal, MD, Giulia Domenichini, MD, Carine Stettler, RN, Haran Burri, MD
From: Cardiology Department, University Hospital of Geneva, Geneva, Switzerland
Address for correspondence: Prof Haran Burri Cardiac pacing Unit, Cardiology Department University Hospital of Geneva Rue Gabrielle Perret Gentil 4 1205 Geneva, Switzerland e-mail: [email protected] phone: +41 22 372 72 00 Word count: Abstract 227 words. Total (including abstract, text, references tables, figure legends) 4609 words. Disclosures: H.B. has received speaker fees and research grants from Abbott, Biotronik, Boston Scientific, Medtronic and Microport. N.D has received fellowship support and speaker fees from Medtronic. Funding: None.
1
1
ABSTRACT
2
Background: Programming of His bundle pacing may be challenging because
3
current implantable pulse generators are not specifically designed for this pacing
4
modality.
5
Objectives: Our aim was to evaluate electrical parameters in order to propose
6
preset programming options with different configurations.
7
Methods: Data were collected from 50 patients with His pacing leads connected
8
to various ports (atrial, right ventricular or left ventricular) of pacemakers and
9
defibrillators during a detailed device interrogation which included capture
10
thresholds with various pacing vectors, measurement of timing intervals, and
11
performance of automatic threshold algorithms.
12
Results: His bundle-pacing thresholds were significantly lower during unipolar
13
pacing compared to bipolar and extended bipolar polarities. However, current
14
drain was offset due to lower impedance. The His pace - right ventricular sensed
15
intervals were measured at 40-150ms (mean 85±25ms) with the longest delays
16
in patients with uncorrected right bundle branch block and selective His capture.
17
This has implications for ventricular safety pacing windows (which were
18
inactivated without evidence of crosstalk) and delays to minimize unnecessary
19
ventricular backup pacing (which was also affected by refractory periods). The
20
measured intervals also impacted the performance of automatic threshold
21
algorithms, which performed differently depending on which port the His lead
22
was connected to and did not distinguish between His and myocardial capture.
23
Conclusion: Our report provides data which could serve to configure automated
24
programming settings to simplify management of His bundle pacing.
25
Keywords: His bundle pacing; capture thresholds; programming; timing; optimization.
2
26 27
INTRODUCTION
28
His bundle pacing (HBP) is becoming increasingly adopted as it preserves
29
physiological conduction along the His-purkinje system, unlike traditional right
30
ventricular (RV) pacing which results in dysynchrony and a gradual decline in
31
left ventricular (LV) function.1 HBP has not only been used in lieu of RV pacing,2
32
but also in lieu of biventricular pacing for cardiac resynchronization therapy
33
(CRT)3, 4 and for His-optimized CRT (HOT-CRT).5, 6
34
There are currently no implantable pulse generators (IPGs) that include
35
timing cycles specific to HBP, and depending on which port the His lead is
36
connected to, different advanced settings need to be considered to safely
37
optimize programming.5 Another consideration is battery longevity, as capture
38
thresholds with HBP are often higher than with RV pacing.7 Pacing polarity has
39
been shown to significantly affect LV thresholds with CRT8 but there are few data
40
regarding HBP,9 and no information on whether automatic capture management
41
algorithms yield valid results for HBP. As a precautionary measure or in the
42
instance of implantable cardioverter defibrillators (ICDs), backup ventricular
43
leads may be implanted and can result in unnecessary pacing due to restrictions
44
of timing cycles with current devices. For example, with His leads connected to
45
the atrial (A) port, if sensing from the backup RV lead occurs within 110ms of His
46
pacing, ventricular safety pacing (VSP) will automatically be delivered. To avoid
47
this unnecessary current drain, the feature must be inactivated, but may expose
48
the patient to consequences of crosstalk.
49
The aim of this study was to evaluate various electrical parameters in
50
order to facilitate management, optimize programming for HBP and prolong
3
51
battery longevity whilst minimizing risk to the patient. To do this we
52
investigated effect of pacing polarity on HBP thresholds, measured His pace-RV
53
sensed (HP-RVS) intervals, evaluated risk of crosstalk due to inactivation of VSP,
54
evaluated accuracy of automatic threshold measurement algorithms and efficacy
55
of prolonging atrioventricular (AV) and interventricular (VV) delays to avoid
56
unnecessary ventricular backup pacing.
57 58
METHODS
59
We included consecutive patients followed-up at the device clinic of the
60
University Hospital of Geneva, Switzerland, implanted with a Medtronic
61
(Minneapolis, MN, USA) 3830 lead on the His bundle, connected to a Medtronic
62
IPG. HBP was confirmed by transitions in QRS morphology during threshold tests
63
according to current definitions.10 The study was approved by the institutional
64
ethics committee and all patients gave informed consent to participate.
65
The device check consisted of a His capture threshold measurement
66
starting at 8V in each available polarity at a pulse width of 0.4ms and 1ms, and
67
decrementing output in 0.25V steps (except for the initial steps which are only
68
possible between 8V, 6V and 5.5V) while recording a 12-lead ECG. Unipolar
69
pacing is only available in pacemakers in Medtronic devices. An extended bipolar
70
configuration was tested in CRT devices with the His lead connected to the LV
71
port.
72
(unipolar/bipolar/extended bipolar configurations, each tested at 0.4ms and
73
1ms). The test was used to determine the various types of His capture and their
74
respective thresholds i.e. non-selective His capture (NSHC), selective His capture
75
(SHC) and correction of bundle branch block (BBB), using a standard 12-lead
These
patients
with
a
CRT
pacemaker
had
6
datasets
4
76
ECG. Pacing impedance was measured for each polarity, and current drain at
77
threshold output calculated by V/R.
78
To analyze the accuracy of the automatic threshold measurement
79
algorithms in determining the His threshold, the stored values collected by the
80
device were compared to the in-office thresholds obtained at the same pulse
81
width and same pacing polarity. Generally, the automatic threshold value was
82
obtained at 1 a:m the same day as the device check. An automatic threshold
83
within 0.5V of the in-office value was deemed accurate.
84
For each patient with the His lead connected to the A or LV port and a
85
backup RV lead, the His paced-RV Sensed (HP-RVS) interval was measured using
86
digital calipers on the Medtronic programmer at 100mm/s sweep speed, for each
87
form of His capture Data on percentages of pacing from the His and ventricular
88
leads was recorded, in order to assess the effectiveness of programming pacing
89
intervals to minimize ventricular pacing.
90
For the patients with the His lead on the A channel, a test for cross-talk
91
was completed by increasing the output to maximum on the His lead (8V at 1ms
92
in unipolar) and programming maximum sensitivity for the RV channel (0.15mV
93
or 0.45 mV depending on whether the IPG was a defibrillator or a pacemaker).
94
Using these ‘worst-case’ scenarios, evidence of RV sensing of the after-potential
95
of the high output His pacing spike was evaluated by analyzing the timing of the
96
RVS event on the marker channel compared with the RV electrogram (crosstalk
97
being deduced when the RVS marker occurred earlier than the potential of the
98
captured RV electrogram).
99 100 5
101
Statistical analysis
102
Analysis was performed used the SPSS v.25 (Armonk, NY) software. Data
103
showed a normal distribution according to histogram analysis and the
104
Kolmogorov-Smirnoff test. Data were reported as mean±SD unless specified
105
otherwise. Differences between groups was assessed by the Student’s t-test. A P
106
value <0.05 was considered significant.
107 108
RESULTS
109
In total 50 patients were included in the study (see table 1). All ICD leads
110
were implanted at the apex, and all RV pacing leads on the interventricular mid-
111
septum. A backup ventricular lead was present in 45/50 patients, which allowed
112
measurement of HP-RVS in these patients. The time from implantation to device
113
follow-up was 3.2±4.1 months.
114 115
ECG findings
116
Overall at 8V/1ms, NSHC was present in 47 patients and SHC in three
117
patients. Of the patients with NSHC, a transition to SHC at a lower output was
118
observed in 24 patients and to myocardial capture in 23 patients (i.e SHC was
119
present in 27/50 patients overall). An extended bipolar pacing mode was tested
120
in the 20 patients with the His lead connected to the LV port. Of these, 5 (25%)
121
patients (all implanted with pacemakers) had evidence of anodal capture from
122
the RV lead (occurring at a threshold of 4.2±0.9V/1ms and at 5.6±1.5V/0.4ms).
123
An example of anodal capture in the extended bipolar mode is shown in figure 1.
124 125
Capture thresholds, lead impedance and current drain at different pacing polarities 6
126
Comparison of electrical parameters between unipolar and bipolar pacing
127
configurations are shown in table 2. A separate analysis was performed for the
128
12 CRT-P patients with the His lead connected to the LV port, in whom unipolar,
129
bipolar and extended bipolar pacing configurations (His lead tip cathode to RV
130
ring anode) were directly comparable. Even though unipolar pacing
131
configurations resulted in significantly lower capture thresholds compared to
132
bipolar and extended bipolar pacing, the lower impedances offset the calculated
133
current drain at threshold output. There were no significant differences in
134
electrical parameters between bipolar and extended bipolar configurations.
135 136
HP-RVS delay
137
Results from the 45 patients with a His lead in either the A or LV port are
138
shown in figure 2 (separate values were recorded in a same patient during NSHC
139
and SHC when pacing at different outputs – with a total of 65 different
140
measurements). We analyzed results based upon presence of RBBB (12 patients
141
had underlying RBBB in intrinsic rhythm and three of the six pacemaker-
142
dependant patients had paced QRS morphologies showing uncorrected RBBB).
143
The intervals ranged from 40 to 150ms (mean 85±25ms), with the longest
144
intervals in patients with SHC and uncorrected RBBB (P<0.01 for all
145
comparisons). With a reduction in pacing output, 10 patients with RBBB had
146
NSHC which transitioned to SHC. The HP-RVS interval subsequently lengthened
147
from 89±32ms to 115±25ms (P=0.001) in these patients (and did not change
148
significantly in the five patients with RBBB who transitioned from NSHC to
149
myocardial capture only). In patients without RBBB, the HP-RVS intervals were
7
150
significantly shorter during NSHC than during SHC (71±18ms vs. 88±12ms
151
respectively P=0.005).
152 153 154
Crosstalk and VSP
155
No evidence of crosstalk was observed during in-office testing in any of
156
the 25 patients with the His lead in the atrial port. Of these patients, 10 had
157
complete heart block at the time of testing, of whom six were due to AV node
158
ablations. None of the patients reported symptoms suggestive of crosstalk such
159
as syncope or malaise.
160 161
Percentages of unnecessary ventricular pacing
162
Of the 25 patients with a His lead connected to the atrial port, 18 had a
163
paced AV interval programmed to 180ms with the intent to suppress
164
unnecessary RV pacing (otherwise, three patients were programmed to VVIR
165
mode and four with a short AV interval for HOT-CRT). The AV interval of 180ms
166
successfully reduced the percentage of RV pacing after HBP to 0.8 ±0.8 % (range
167
0.1-2.8%).
168
In eight patients with the His lead connected to the LV port of a CRT
169
device, the interventricular (VV) interval was programmed to 80ms (the
170
maximum programmable value in Medtronic devices) in an attempt to avoid
171
unnecessary RV pacing. All these patients had 100% RV pacing following His
172
paced events, despite the fact that 2 patients had a HP-RVS interval less than the
173
programmed 80ms (40ms and 70ms).
174 8
175
Automatic threshold measurement algorithms
176
The automatic capture management algorithm was programmed to
177
“monitor” in the 25 patients with the His lead on the atrial port. At follow-up,
178
only one patient had automatic threshold readings for the His lead, the result of
179
which were accurate when compared to the in-office threshold. Of note, this
180
patient had a HP-RVS interval of 140ms and SHC with an uncorrected RBBB.
181
Another observation was that when the His lead was connected to the A port, not
182
only were the automatic thresholds not available for the His lead, but they were
183
also unavailable for all the other leads connected to the IPG.
184
For the five patients with the His lead on the RV port, each had myocardial
185
capture as the lowest capture form. Of these patients, four had automatic
186
threshold measurements within 0.5V of the in-office myocardial capture
187
threshold. The fifth patient had an automatic threshold of >2.5V/0.4ms with an
188
in-office myocardial threshold of 2.75V/0.4ms (an exact threshold value is not
189
given in these instances).
190
Regarding the 20 patients in the LV group, 17 had automatic threshold
191
measurements which were programmed to “monitor”. Of these patients, 10 had
192
reported values reflective of myocardial thresholds and seven reflective of SHC.
193
All automatic threshold values were within 0.5V of in-office readings.
194 195 196
DISCUSSION
197 198
The main findings of our study are that 1) As expected, unipolar pacing
199
configuration yielded lower capture thresholds compared to bipolar and 9
200
extended bipolar configurations, but the impact on calculated current drain was
201
mitigated because of reduced pacing impedance 2) HP-RVS delays vary between
202
40 and 150ms (average 85±25ms) with the longest delays associated with SHC
203
and uncorrected RBBB 3) The risk of AV crosstalk with His leads plugged in the
204
atrial port is low and no adverse effects were observed with inactivation of VSP
205
4) Unnecessary RV pacing was prevented in patients with the His lead in the
206
atrial port by programming a paced AV interval of 180ms, but was unable to be
207
avoided when the His lead was connected to the LV port 5) Automatic threshold
208
measurement algorithms are dependent upon the port the His lead is connected
209
to and do not distinguish between His and myocardial capture.
210
As previously shown with LV leads,8 capture thresholds and lead
211
impedances with His leads are significantly lower with unipolar compared to
212
bipolar pacing polarities. However in contradiction with data from Su et al.9 we
213
found that unipolar thresholds and impedances are also lower compared to the
214
extended bipolar configuration in the subset of our patients with the His lead
215
connected to the LV port of CRT devices. Therefore, due to lower pacing
216
impedance, the calculated current drain was in fact higher with unipolar pacing.
217
However, an advantage with unipolar pacing is a visible pacing spike, which
218
facilitates identification of HBP on ECG tracings. Ideally, pulse generators should
219
automatically calculate battery longevity based upon programmed output, taking
220
into account the pacing impedance (as is already the case with some models).
221
An observation we describe for the first time with HBP in an extended
222
bipolar pacing configuration is anodal capture from the RV lead (well described
223
with CRT). This may create confusion during threshold testing, as it results in an
224
additional transition in QRS morphology. 10
225
The HP-RVS delay is of major importance for device programming. This
226
delay will depend upon a number of factors, such as SHC vs NSHC, correction of
227
RBBB, and possibly RV lead position (see figure 2). In patients with SHC, it is the
228
sum of conduction duration through the His-Purkinje system (usually 40-50ms)
229
and intra-myocardial conduction from the Purkinje exit to the lead implantation
230
site. In patients with NSHC, the HP-RVS delay will depend on the shorter of either
231
this pathway or intra-myocardial conduction from the His lead to the RV lead.
232
In patients who have the His lead connected to the atrial port, RVS will
233
usually occur during the VSP window (95-110ms depending upon the
234
manufacturer), leading to unnecessary RV pacing. VSP may be inactivated, but
235
this potentially exposes the patient to asystole in case of crosstalk, which was
236
however absent in all our patients tested in “worst case” scenarios, but who had
237
a post atrial pacing ventricular blanking period (PAVB) set to a non-
238
programmable values of 30ms in the devices used in this study. Alternative
239
algorithms to VSP also exist e.g. Boston Scientific devices (Marlborough, MA,
240
USA) rely upon retriggerable noise windows.
241
In order to avoid unnecessary RV pacing in patients with a His lead in the
242
atrial port, a paced AV interval of 180ms was found to be effective (with an
243
average of <1% of RV pacing, corresponding most probably to cycles with loss of
244
His capture). It is therefore unnecessary to program unduly long AV intervals,
245
which also carry the risk of pacing in the vulnerable period of the T-wave in case
246
of R-wave undersensing. The main issue was found to be in patients with the His
247
lead in the LV port in whom RV pacing was delivered 100% of the time. This was
248
not only due to the maximum interventricular delay being limited to 80ms, but
249
also to ventricular blanking (which was set by default to 200-230ms in the 11
250
devices in our study) and committed RV pacing after delivery of pacing by the LV
251
channel. A solution in this configuration would be to design refractory periods
252
able to consider RVS events which fall within 30-150ms of His paced events to be
253
His capture, allowing inhibition of RV pacing. Blanking during the first 30ms
254
following His pacing would avoid crosstalk (by analogy with the PAVB) and a
255
refractory period after 150ms would avoid T-wave oversensing. A schematic
256
representation of the proposed timing intervals is shown in figure 3.
257
The HP–RVS intervals also impacted the automatic capture management
258
algorithms. In patients with the His lead connected to the atrial port, an RVS
259
event falling within 110ms of pacing from the atrial channel will abort the
260
threshold test. Short intervals will also prevent automatic threshold testing of RV
261
and LV leads in these patients. The RV capture management algorithm of
262
Medtronic devices considers any RVS event that occurs in the 110ms window
263
following pacing to be V capture (i.e. it does not specifically detect the evoked
264
potential). Although this was not tested in our patient population, patients with
265
SHBC and uncorrected RBBB may have erroneous diagnosis of noncapture by the
266
algorithm due to long pace-sense intervals (falling outside the 110ms window).
267
In addition, as the AV interval of the test cycle is shortened to 10ms (to avoid AV
268
conduction), detection of a far-field P-wave on the His lead may result in
269
erroneous diagnosis of capture. The LV capture management algorithm of
270
Medtronic devices relies upon interventricular conduction delay and accurately
271
determined thresholds for loss of myocardial or His capture (whichever was
272
lowest) in all cases. Anodal capture in case of an extended bipolar configuration
273
of the His lead may however confound the results (as the local RV electrogram
274
occurs directly after pacing from the His lead). Devising an algorithm capable of 12
275
accurately measuring His capture is likely to be challenging. As recently
276
described, morphological analysis of near-field electrograms may be an option to
277
distinguish NSHC from SHC.11 Analysis of far-field electrogram morphology may
278
be helpful for distinguishing NSHC from myocardial capture, but this needs to be
279
evaluated. Capture threshold algorithms which yield accurate results in different
280
pacing configurations will obviously be of great interest, not only to adapt pacing
281
output and preserve battery longevity, but also to ensure patient safety. Changes
282
in thresholds with HBP may be even more unpredictable than with standard
283
ventricular pacing, underlining the need for this feature.
284
Many of the issues described in our report occur in the context of backup
285
ventricular leads. Although their utility in HBP has been questioned by
286
experienced implanters, they are unavoidable in patients with ICDs.
287
Furthermore, as requirement for lead revisions of His leads (e.g. due to loss of
288
capture) are frequent (close to 7% at 5 years7), safety is a concern, especially in
289
pacing-dependant patients. Backup ventricular leads are also useful in patients
290
planned for atrioventricular node ablation (which was indicated in over a
291
quarter of our patients) due to the risk of lead dislodgement or increase in His
292
capture thresholds.12 Backup ventricular leads also provide adequate sensing,
293
which may be an issue with His leads (with oversensing of A/His potentials or V
294
undersensing5). In addition, backup V leads also allow programming of lower
295
safety margins of the His lead, prolonging battery life, and provide the option of
296
HOT-CRT. Currently, the FDA only approves the 3830 lead for HBP when it is
297
connected to the ventricular port of single- and dual-chamber pacemakers.
298
However, clinical requirements extend beyond these configurations, and
299
hopefully approval will also be granted with a wider perspective in the future. 13
300 301 302 303
Study limitations
304
The sample size is relatively limited, especially in the subgroup of patients
305
with the His lead connected to the RV port. These patients were included to
306
evaluate the accuracy of the automatic threshold algorithm, but none of these
307
patients had SHC. Also, we were not able to test sensing issues in these patients
308
(e.g. A or His oversensing) as these are carefully evaluated at implantation at our
309
centre and only patients without such issues have His leads connected to the RV
310
port. We only included Medtronic devices in our analysis, so results may not
311
apply to other manufacturers (e.g. for automatic threshold algorithms).
312
Nevertheless, most of our data such as capture thresholds with different pacing
313
polarities and HP-RVS intervals should be applicable to all manufacturers.
314 315
CONCLUSIONS
316
Our results provide a framework for developing automated programming
317
settings which are tailed to meet the needs of HBP (see table 3). Incorporation of
318
predefined settings in IPGs for the different configurations of HBP would greatly
319
simplify programming and optimization of this therapy.
320 321
Acknowledgments: The authors would like to thank Mr Todd Sheldon
322
(Medtronic) for his answers to our technical queries.
323 324 14
325
References
326 327
1.
Lee MA, Dae MW, Langberg JJ, et al. Effects of long-term right ventricular
328
apical pacing on left ventricular perfusion, innervation, function and
329
histology. Journal of the American College of Cardiology 1994;24:225-
330
232.
331
2.
Abdelrahman M, Subzposh FA, Beer D, et al. Clinical Outcomes of His
332
Bundle Pacing Compared to Right Ventricular Pacing. Journal of the
333
American College of Cardiology 2018;71:2319-2330.
334
3.
Barba-Pichardo R, Manovel Sanchez A, Fernandez-Gomez JM, Morina-
335
Vazquez P, Venegas-Gamero J, Herrera-Carranza M. Ventricular
336
resynchronization therapy by direct His-bundle pacing using an internal
337
cardioverter defibrillator. Europace : European pacing, arrhythmias, and
338
cardiac electrophysiology : journal of the working groups on cardiac
339
pacing, arrhythmias, and cardiac cellular electrophysiology of the
340
European Society of Cardiology 2013;15:83-88.
341
4.
Sharma PS, Dandamudi G, Herweg B, et al. Permanent His-bundle pacing
342
as an alternative to biventricular pacing for cardiac resynchronization
343
therapy: A multicenter experience. Heart rhythm 2018;15:413-420.
344
5.
Burri H, Keene D, Whinnett Z, Zanon F, Vijayaraman P. Device
345
Programming for His Bundle Pacing. Circulation Arrhythmia and
346
electrophysiology 2019;12:e006816.
347
6.
Vijayaraman P, Herweg B, Ellenbogen KA, Gajek J. His-Optimized Cardiac
348
Resynchronization Therapy to Maximize Electrical Resynchronization.
349
Circulation Arrhythmia and electrophysiology 2019;12:e006934. 15
350
7.
Vijayaraman P, Naperkowski A, Subzposh FA, et al. Permanent His-bundle
351
pacing: Long-term lead performance and clinical outcomes. Heart rhythm
352
2018;15:696-702.
353
8.
Burri H, Schrage M, Morani G, et al. Effect of lead design and pacing vector
354
on electrical parameters of quadripolar coronary sinus leads: the RALLY-
355
X4 study. Pacing and clinical electrophysiology : PACE 2019;in press.
356
9.
Su L, Xu L, Wu SJ, Huang WJ. Pacing and sensing optimization of
357
permanent His-bundle pacing in cardiac resynchronization
358
therapy/implantable cardioverter defibrillators patients: value of
359
integrated bipolar configuration. Europace : European pacing,
360
arrhythmias, and cardiac electrophysiology : journal of the working
361
groups on cardiac pacing, arrhythmias, and cardiac cellular
362
electrophysiology of the European Society of Cardiology 2016;18:1399-
363
1405.
364
10.
Vijayaraman P, Dandamudi G, Zanon F, et al. Permanent His bundle
365
pacing: Recommendations from a Multicenter His Bundle Pacing
366
Collaborative Working Group for standardization of definitions, implant
367
measurements, and follow-up. Heart rhythm 2018;15:460-468.
368
11.
Saini A, Serafini NJ, Campbell S, et al. Novel Method for Assessment of His
369
Bundle Pacing Morphology Using Near Field and Far Field Device
370
Electrograms. Circulation Arrhythmia and electrophysiology
371
2019;12:e006878.
372
12.
Vijayaraman P, Subzposh FA, Naperkowski A. Atrioventricular node
373
ablation and His bundle pacing. Europace : European pacing, arrhythmias,
374
and cardiac electrophysiology : journal of the working groups on cardiac 16
375
pacing, arrhythmias, and cardiac cellular electrophysiology of the
376
European Society of Cardiology 2017;19:iv10-iv16.
377 378 379
17
380 381
Table 1. Patient Demographics:
Patient demographics (n=50) Male/female Age (years) Port of His lead connection (A/LV/RV) Device - DDD PM - CRT-P - CRT-D Indication for device implantation - AVB II - AVB III - Ablate and pace of AF - Slow AF - CRT indication with RBBB - Painfull LBBB Left ventricular ejection fraction Intrinsic QRS duration (ms) Intrinsic QRS morphology: - Normal - RBBB - LBBB - IVCD - Pacemaker dependent Hypertension Chronic renal failure Ischemic heart disease Permanent AF Diabetes
33/17 71±12 25/20/5
382 383
Values
and
384
eGFR<50ml/min/1.73m2.
385
IVCD=interventricular conduction delay; LBBB=left bundle branch block; LV=left ventricular;
386
RBBB=right bundle branch block; RV= right ventricular.
represent
numbers
of
A=atrial;
patients
23 15 12 10 14 13 6 6 1 0.52±0.12 119±28 18 13 3 10 6 24 17 19 25 12
AF=atrial
mean±SD. fibrillation;
Renal
failure
defined
AVB=atrioventricular
as
block;
387
18
388
Table 2. His capture thresholds, pacing impedance, and calculated current drain
389
at threshold output. There were no significant differences between bipolar and
390
extended bipolar configurations. Current drain calculated at capture threshold
391
voltage output using I=V/R. All comparisons are for paired analyses.
392 393
394
Total population
His lead connected to LV port of
(n=50)
CRT-P (n=12)
Unipolar
Bipolar
Unipolar
Bipolar
Extended
(n=38)
(n=50)
(n=12)
(n=12)
bipolar (n=12)
Threshold @0.4ms (V)
1.9±1.5
2.4±1.9
†
1.4±0.9
1.9±1.4
Threshold @1ms (V)
1.5±1.2
1.9±1.5
†
1.1±0.7
1.4±1.0
Impedance (Ω)
318±52
473±7
†
309±69
454±92
414±92
Current drain @0.4ms (mA)
6.2±5.0
5.4±4.5
‡
4.8±3.7
4.5±3.9
4.5±3.7
Current drain @1ms (mA)
4.8±4.0
4.0±3.2
‡
3.8±2.8
3.2±2.8
†
‡
Compared to unipolar: P<0.001 P<0.01
§
‡
§
†
‡
‡
1.9±1.3
§
1.3±0.8
†
§
§
3.2±2.2
P<0.05
395 396
19
397
Table 3. Framework for predefined settings for His bundle pacing. Parameter Common to all configurations Pacing polarity Unipolar and bipolar (also for A and LV ports of ICDs)
Comment
Lower thresholds with unipolar and visible pacing spike. However, current drain may be offset by lower impedance. Estimation of remaining Based upon pacing output Immediate calculation upon battery longevity and impedance, and history reprogramming to chose of percentage of pacing. configuration with least drain. Automatic threshold Analysis of EGM morphology To identify SHC (near-field algorithm EGM) or distinguish NSHC from myocardial capture (far-field EGM). His lead connected to the atrial port Sensing Inactivate Ventricular sensing ensured by ventricular lead. Ventricular safety Inactivate Crosstalk prevented by pacing “PAVB” of 30ms. Use retriggerable noise window (currently used for Boston Scientific devices). “Atrioventricular” Paced interval of 180ms Maximum delay of HP – RVS interval interval was 150ms. Automatic capture Detection of VS within 40Should be no interlock with threshold algorithm 150ms window after HP RV and LV automatic capture algorithms. His lead connected to the right ventricular port Post “AS” and “AP” Individually programmable To avoid A oversensing by the His lead. blanking period for a range of values in dual chamber and biventricular devices Automatic capture Extension of the detection Maximum delay of HP – RVS algorithm window beyond 110ms interval was 150ms in patients with uncorrected RBBB. His lead connected to the left ventricular port Interventricular interval Programmable up to >150ms To avoid unnecessary backup RV pacing “Interventricular” Shortened to 30ms refractory period Ventricular refractory Initiate 150ms after HP Avoid unnecessary backup period RV pacing (while protecting against T-wave oversensing) 20
Ventricular tachyarrhythmia counters
RVS within 150ms of HP not counted
Automatic threshold algorithm
Based upon HP-RVS conduction delay
Avoid misdiagnosis of tachyarrhythmia or lead dysfunction (e.g. due to algorithm comparing nearfield with far-field counts). Current Medtronic LVCM algorithm performs well (although does not distinguish between His and myocardial capture).
398 399
AS=atrial sens; AP=atrial pace; EGM=electrogram; HP= His paced; LV=left ventricle;
400
LVCM=left ventricular capture management; PAVB: post atrial pace ventricular
401
blanking; RBBB: right bundle branch block; RV: right ventricle; RVS: right ventricular
402
sense; SHC=selective His capture; VP=ventricular pace; VS=ventricular sense. Quotation
403
marks are used in instances where the parameter does not truly reflect the designation.
404
21
405
Figure 1. Illustration of anodal capture. QRS morphology during pacing from
406
ventricular lead implanted in the interventricular mid-septum (left panel) and
407
extended bipolar pacing at decrementing outputs (other four panels) from the
408
His lead connected to the left ventricular port of a biventricular pacemaker. The
409
transitions are readily seen in lead V1. RV=right ventricle; A=anodal capture;
410
C+=with correction of right bundle branch block; C- without correction of right
411
bundle branch block; NS=non-selective His capture; S= selective His capture
412
22
413
Figure 2. Schematic representation of effect of type of His capture on delays
414
between pacing from the His lead and sensing from the right ventricular lead
415
(values are mean±SD). The interval is longest in case of uncorrected right bundle
416
branch block (RBBB) with selective His capture (SHC) and is shortened by non-
417
selective His capture (NSHC) in this instance.
418 419
420 421 422 423 23
424
Figure 3. Proposed timing intervals for His pacing with His leads connected to
425
the atrial or left ventricular ports (parameters with the His lead connected to the
426
right ventricular port would follow standard settings and are not shown). AH=
427
atrio-His interval; AP=atrial pace marker; ARP=atrial refractory period; AS=
428
atrial sense marker; HP=His pace maker; HV=His-ventricular interval ;
429
PVARP=post-ventricular atrial refractory period; VA=ventriculo-atrial interval;
430
VH=ventriculo-His interval; VP=ventricular pace marker; VRP=ventricular
431
refractory period; VS= ventricular sense marker.
432
433
24