- Email: [email protected]
Twenty-Five Year Outcomes of the Lateral Tunnel Fontan Procedure
Accepted Manuscript Title: Twenty-Five Year Outcomes of the Lateral Tunnel Fontan Procedure Author: Thomas G Wilson, William Y Shi, Ajay J Iyengar, David S Winlaw, Rachael L Cordina, Gavin R Wheaton, Andrew Bullock, Thomas L Gentles, Robert G Weintraub, Robert N Justo, Leeanne E Grigg, Dorothy J Radford, Yves d'Udekem, The Australia and New Zealand Fontan Registry PII: DOI: Reference:
S1043-0679(17)30166-1 http://dx.doi.org/doi: 10.1053/j.semtcvs.2017.06.002 YSTCS 989
To appear in:
Seminars in Thoracic and Cardiovascular Surgery
Please cite this article as: Thomas G Wilson, William Y Shi, Ajay J Iyengar, David S Winlaw, Rachael L Cordina, Gavin R Wheaton, Andrew Bullock, Thomas L Gentles, Robert G Weintraub, Robert N Justo, Leeanne E Grigg, Dorothy J Radford, Yves d'Udekem, The Australia and New Zealand Fontan Registry, Twenty-Five Year Outcomes of the Lateral Tunnel Fontan Procedure, Seminars in Thoracic and Cardiovascular Surgery (2017), http://dx.doi.org/doi: 10.1053/j.semtcvs.2017.06.002. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. 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.
1 1
Twenty-Five Year Outcomes of the Lateral Tunnel Fontan Procedure
2 3
Authors
4
Thomas G Wilson, BSc, MD1, 2, 3, William Y Shi, MBBS, 1, 2, 3, Ajay J Iyengar, MBBS(Hons),
5
BMedSci, PhD1, 2, 3, David S Winlaw, MBBS(Hons), MD, FRACS4, 5, Rachael L Cordina,
6
MBBS, FRACP5, 6, Gavin R Wheaton, MBBS, FRACP 7, Andrew Bullock, MBBS, FRACP 8,
7
Thomas L Gentles, MBChB, FRACP9, Robert G Weintraub, MBBS, FRACP2, 10, Robert N
8
Justo, MBBS, FRACP 11, Leeanne E Grigg, MBBS, FRACP12, Dorothy J Radford, MBBS,
9
MD, FRACP13, 14, Yves d’Udekem, MD, PhD1, 2, 3, The Australia and New Zealand Fontan
10
Registry
11 12
Institutions and Affiliations
13
1. Department of Cardiac Surgery, Royal Children’s Hospital, Melbourne, Australia
14
2. Heart Research Group, Murdoch Childrens Research Institute, Melbourne, Australia
15
3. Department of Paediatrics, Faculty of Medicine, The University of Melbourne,
16 17 18
Victoria, Australia 4. The Heart Centre for Children, The Children’s Hospital at Westmead, Sydney, Australia
19
5. Department of Paediatrics, University of Sydney, Sydney, Australia
20
6. Department of Cardiology, Royal Prince Alfred Hospital, Sydney, Australia
21
7. Department of Cardiology, Women’s and Children’s Hospital, Adelaide, Australia
22
8. Children’s Cardiac Centre, Princess Margaret Hospital for Children, Perth, Australia
23
9. Greenlane Paediatric and Congenital Cardiac Service, Starship Children’s Hospital,
24 25
Auckland, New Zealand 10. Department of Cardiology, Royal Children’s Hospital, Melbourne, Australia
Page 1 of 28
2 26 27
11. Queensland Paediatric Cardiac Service, Lady Cilento Children’s Hospital, Brisbane, Queensland, Australia
28
12. Department of Cardiology, The Royal Melbourne Hospital, Melbourne, Australia
29
13. Adult Congenital Heart Unit, The Prince Charles Hospital, Brisbane, Australia
30
14. Faculty of Medicine, University of Queensland, Brisbane, Australia
31 32
Corresponding Author
33
A/Prof Yves d’Udekem, MD PhD
34
Department of Cardiac Surgery, Royal Children’s Hospital
35
Flemington Rd, Parkville VIC 3052
36
Phone: +61 3 9345 5200 Fax: +61 3 9345 6001
37
E-mail: [email protected]
38 39
Acknowledgement of Grant Support
40
The Australia and New Zealand Fontan Registry is funded by grants from the National Health
41
and Medical Research Council (NHMRC; Project Grants 1012241, 1047923, 1065794). The
42
authors acknowledge support provided to the Murdoch Childrens Research Institute by the
43
Victorian Government’s Operational Infrastructure Support Program. Dr William Shi is
44
supported by the Royal Australasian College of Surgeons Foundation for Surgery Peter King/
45
Heart Foundation Research Scholarship in addition to the University of Melbourne Viola
46
Edith Reid and the RG and AU Meade Scholarships. Yves d’Udekem is a NHMRC Clinician
47
Practitioner Fellow (1082186). The Victorian Government’s Operational Infrastructure
48
Support Program supported this research project.
49
Page 2 of 28
3 50
Conflicts of Interest
51
Yves d’Udekem is a consultant for MSD and Actelion. Andrew Bullock reports consulting
52
fees from Actelion. All other authors have nothing to disclose with regard to commercial
53
support.
54 55
Keywords
56
Fontan Procedure, Heart Ventricles Abnormalities, Heart Defects Congenital, Retrospective
57
Studies, Survival Rate, Follow-Up Studies, Disease-Free Survival
58 59
Article word count: 4,295
60
Page 3 of 28
4 61
Abstract
62
Objective(s): To characterize late outcomes of the lateral tunnel (LT) Fontan procedure.
63
Methods: The outcomes of all patients who underwent a LT Fontan procedure in Australia
64
and New Zealand were analysed. Original files were reviewed and outcomes data were
65
obtained through a binational Registry.
66
Results: Between 1980 and 2014, a total of 301 patients underwent a LT Fontan procedure
67
across 6 major centers. There were 13 hospital mortalities, 21 late deaths, 8 Fontan
68
conversions/revisions, 8 Fontan takedowns and 4 heart transplantations. Overall survival at 15
69
and 25 years was 90% (95% confidence interval [CI]:86-93%) and 80% (95% CI:69-91%),
70
respectively. Protein-losing enteropathy/plastic bronchitis was observed in 14 patients (5%).
71
Freedom from late failure at 15 and 25 years was 88% (95% CI:84-92%) and 82% (95%
72
CI:76-87%), respectively. Independent predictors of late Fontan failure were prolonged
73
pleural effusions post-Fontan operation (HR 3.06,1.05-8.95, p=0.041), age >7 years at Fontan
74
(vs. 3-5 years, HR 9.7, 2.46-38.21, p=0.001) and development of supraventricular tachycardia
75
(4.67, 2.07-10.58, p<0.001). Freedom from tachy- or bradyarrhythmias at 10 and 20 years was
76
87% (95% CI:83-91%) and 72% (95% CI:66-79%), respectively. Thromboembolic events
77
occurred in 45 patients (16%; 26 strokes), and freedom from symptomatic thromboembolism
78
at 10 and 20 years was 93% (95% CI:89-96%) and 80% (95% CI:74-86%), respectively.
79
Conclusions: Over a twenty-five-year period, the LT technique has achieved excellent late
80
survival. As this population ages, they are at increasing risk of failure and adverse events. We
81
are likely to see an increasing proportion requiring heart transplantation and late
82
reintervention.
83 84
Abstract word count: 249
Page 4 of 28
5 85
Glossary of abbreviations
86
LT = lateral tunnel
87
CI = confidence interval
88
HR = hazard ratio
89
SVT = supraventricular tachycardia
90
AP = atriopulmonary
91
ECC = extracardiac conduit
92
VAD = ventricular assist device
93
ECMO = extracorporeal membrane oxygenation
94
PLE = protein-losing enteropathy
95
PB = plastic bronchitis
96
NYHA = New York Heart Association
97
SD = standard deviation
98
IQR = interquartile range
99
DKS = Damus–Kaye–Stansel
100
HLHS = hypoplastic left heart syndrome
101 102
Page 5 of 28
6 103
Central message: Late survival after the lateral tunnel Fontan procedure is excellent. This
104
population is subjected to an ongoing risk of late failure.
105 106
Perspective statement: Late survival after the lateral tunnel Fontan procedure is excellent.
107
There is, however, an ongoing risk of failure of the Fontan circulation this population. The
108
onset of new arrhythmia predicts later occurrence of failure and it is likely that careful
109
monitoring of this population is warranted as their atrial channel might be subjected to
110
progressive dilatation.
111 112
Central picture: Competing events cumulative incidence curves for all patients (n=301).
113 114
Page 6 of 28
7 115
Introduction
116
The efficacy of the Fontan procedure in the treatment of children born with single ventricle
117
cardiac malformations continues to be unveiled as this population progresses further into
118
adulthood[1]. It is believed that refinements in surgical techniques have led to an
119
improvement in outcomes since the original atriopulmonary (AP) Fontan, which has now
120
been abandoned[2, 3]. Progressive atrial distension and the development of turbulent flow
121
within the atrial chamber predisposed recipients to tachyarrhythmias, thromboembolic events
122
and even compression of the pulmonary veins [6]. In 1988, the concept of total
123
cavopulmonary circulation was proposed by de Leval in an attempt to avoid turbulent flow
124
within the atrium, and his design of the intra-atrial lateral tunnel (LT) Fontan was widely
125
adopted[7]. In the 1990s, the extra-cardiac conduit (ECC) progressively came into vogue and
126
today is the most frequently performed type of Fontan [8, 9]. Many patients have undergone
127
a LT Fontan operation and the specific outcomes of this form of Fontan are unclear. In one
128
United States multicentre database, more than a third of the Fontan operations performed
129
between 2000 and 2009 were undertaken using the LT technique[9]. Locally, a quarter of the
130
patients with a Fontan circulation currently followed within our region are living with a LT
131
Fontan type[1, 10]. While the implementation of this technique has resulted in improved
132
survival up to early adulthood, the rate of failure of their circulation is still unclear. We sought
133
to characterize the late outcomes of the patients who have undergone a LT Fontan procedure
134
in Australia and New Zealand.
135 136
Methods
137
Ethical Review Board approvals were obtained at all participating institutions. The need
138
for specific consent was waived because of the retrospective nature of the study.
Page 7 of 28
8 139
The full design, data fields and administration of the Australia and New Zealand Fontan
140
Registry have been described previously[1, 10]. Of the 1,462 participants whose data were
141
collected in the Registry database between 1 January 1975 and 1 August 2015, all those who
142
underwent a LT Fontan were identified. Fontan conversions from atriopulmonary (AP)
143
connection and Bjork procedures were excluded.
144
Perioperative variables were extracted from the database and follow-up data were updated
145
where necessary via contact with patients’ cardiologists and obtaining their latest clinical
146
summaries and echocardiogram reports. Medical histories were screened for missing data.
147
The LT technique was gradually phased out between 1997 and 2006 because of the potential
148
theoretical advantages of the ECC, whereafter the ECC modification was solely performed
149
throughout the region. Over this time period, decision to fenestrate the Fontan was left to
150
surgical preference.
151 152
Definitions
153
Hospital mortality was defined as death during the initial hospital stay. Prolonged pleural
154
effusion was defined as effusions requiring pleural drainage persisting for longer than 30 days
155
or requiring reoperation including pleurodesis. Early failure was defined as death, Fontan
156
takedown, Fontan revision or mechanical support with a ventricular assist device (VAD) or
157
extracorporeal membrane oxygenation (ECMO) within 30 days or during the initial hospital
158
stay. Late failure was defined as death, Fontan takedown or conversion to ECC after hospital
159
discharge, heart transplantation, protein losing enteropathy (PLE), plastic bronchitis (PB) or
160
New York Heart Association (NYHA) class III/IV. Concomitant procedures were defined as
161
intracardiac procedures performed at the time of the Fontan operation, excluding atrial
162
septectomy.
Page 8 of 28
9 163
Analyses of long-term outcomes were performed only on those who were discharged from
164
hospital with an intact Fontan circulation (i.e. excluding Fontan takedowns performed during
165
the same admission as their initial operation).
166 167
Statistical Analysis
168
Values are reported as absolute value (percentage [%]) for proportions, mean (standard
169
deviation [SD]) for normally distributed data and median (interquartile range [IQR]) for non-
170
normally distributed data. Where appropriate, these data were compared between groups
171
using χ2 tests, t-tests and Wilcoxon rank-sum tests, respectively.
172
Multivariable logistic regression was used to identify independent predictors of early
173
outcomes, and Cox regression was utilised to identify predictors of early and late outcomes.
174
The backward elimination method was used in regression analyses to include all potential
175
predictors of end points in the initial models and subsequently eliminate covariates in an
176
iterative process to create a final model. Variables with at least moderate evidence against the
177
null hypothesis (P<.10) were included. Inclusion of all potentially important variables in the
178
initial model allows their joint predictive behaviour to be initially evaluated, which is
179
important given that a set of variables may exhibit predictive capability even if a subset does
180
not. The variables used in regression analyses were selected based on those found to be
181
associated with outcomes in prior studies by our group as well as based on our group’s
182
clinical experience. Due to the nonlinear relationship of age at the time of Fontan with the risk
183
of late outcomes, patients were grouped into categories based on their age at Fontan: <3 years,
184
3 to 5 years, 5 to 7 years, and >7 years, with 3 to 5 years as reference group. Similarly,
185
oxygen saturation prior to the Fontan procedure was divided into categories: <77%, 78% to
186
81%, 82% to 85%, and >86%, and comparisons were made using the 82% to 85% category as
187
a reference group. Linearity was examined by performing a likelihood ratio test comparing
188
regression models with continuous covariates against models in which the numeric variable
Page 9 of 28
10 189
had been categorically transformed. Age at Fontan operation had nonlinear associations with
190
mortality and the other end points and was analyzed as a categoric variable.
191
The time of the first onset of an arrhythmia was inserted into the multivariable Cox regression
192
model as a time-varying covariate and its associated hazard ratio (HR) calculated. In essence,
193
patient follow-up was split into two periods: before and after the onset of arrhythmia,
194
allowing the association between the development of an arrhythmia and subsequent late end-
195
points to be established. Analysis of long-term outcomes where onset of SVT was used as a
196
co-variate only occurred in those who were discharged with an intact Fontan circulation. As
197
such, the onset of SVT and its association with early failure was not studied.
198
Estimated cumulative incidence curves for each competing event post-Fontan (Alive, death,
199
transplant, takedown, conversion to ECC) were constructed using the cumulative incidence
200
function described by Scrucca et al[11]. Here, the cumulative incidence function is defined as
201
the probability of failing from cause r (r = 1, …, k, where k is the number of causes of failure)
202
up to a certain time point t such that:
203
Where λr(t) is the cause specific hazard rate and S(t) = Pr(T≥t) is the survival function. Non-
204
parametric maximum likelihood estimation of (cause-specific) cumulative incidence is:
205
.
206 207
Results
208
A total of 301 patients were identified from hospital and surgical databases. Patient
209
characteristics are detailed in Table 1.
210 211
Patient and surgical characteristics
Page 10 of 28
11 212
At the time of the Fontan operation, 71 patients (24%) underwent concomitant procedures,
213
consisting of pulmonary artery reconstruction in 21 patients, Damus–Kaye–Stansel procedure
214
in 8, atrioventricular valve repair in 9, subaortic resection in 8, semilunar valve repair in 4,
215
ventricular septal defect enlargement in 5, pulmonary artery band revision in 4, main
216
pulmonary artery ligation or division in 5, aortic arch reconstruction in 2, arterial switch
217
operation in 2, epicardial pacemaker insertion in 2 and other concomitant procedures in 11.
218 219
The majority of the LT Fontan procedures were performed across five major centres (295
220
patients), with a minority undergoing Fontan completion at smaller or international centres (6
221
patients).
222 223
Early outcomes
224
There were 13 hospital mortalities (4%). Causes of death were low cardiac output in 5
225
patients, sepsis in 3, massive pulmonary emboli in 2, hemorrhagic stroke in 2, and unknown
226
in one patient. No deaths occurred out of hospital within the first 30 days after surgery. A
227
total of 4 patients underwent Fontan takedown during the initial hospital stay, due to low
228
cardiac output syndrome in 2 patients, prolonged chylous effusions with obstruction of the LT
229
conduit in 1, and thrombosis of the Fontan circuit in one patient.
230
The median length of hospital stay was 13 days (IQR: 9-19 days). Prolonged effusions
231
occurred in 23 patients (8%), of whom 11 underwent reoperation (pleurodesis in 6, Fontan
232
takedown/revision in 4, and fenestration dilatation in one patient) and resulted in an increase
233
in the median length of stay to 58 days (IQR: 38-79).
234
In 30 patients (10%) the length of stay exceeded 30 days, and was attributable to prolonged
235
pleural effusions in 16 patients, low cardiac output syndrome in 6, persisting arrhythmias in 3,
Page 11 of 28
12 236
stroke in 2, pneumonia in 2 and pericardial tamponade in one patient. There were no
237
independent predictors for having prolonged pleural effusions.
238
Two patients required mechanical circulatory support (1 ECMO, 1 VAD) with the former
239
subsequently undergoing an early Fontan takedown on postoperative day 2, and the latter
240
dying on postoperative day 8 from persisting low cardiac output and renal failure. Early
241
failure (death, takedown, revision or ECMO/VAD) occurred in 21 patients (7%). On
242
multivariable logistic regression, there were no independent predictors of early Fontan failure.
243
Clinical evidence of perioperative cerebral infarction was observed in 20 patients (7%) during
244
the Fontan admission. These consisted of embolic stroke in 10 patients, hemorrhagic stroke in
245
3, and cerebral ischemia secondary to a low output state in 7.
246 247
Follow-up
248
After excluding patients who moved overseas and were followed internationally (8 patients),
249
hospital mortalities (13 patients) and those who had a Fontan takedown during the initial
250
hospital stay (4 patients), 276 patients remained for survival analysis. Late follow-up was
251
available in 275 patients (99.6%). Survival analysis was undertaken on 4,749 patient-years of
252
follow-up. The median duration of follow-up was 18.6 years (IQR: 15.1-21.1).
253 254
Survival
255
Overall survival at 10, 15, 20 and 25 years was 92% (95% confidence interval [CI]: 88-95%),
256
90% (95% CI: 86-93%), 86% (95% CI: 81-90%) and 80% (95% CI: 69-91%), respectively.
257
Late survival of patients discharged alive with an intact Fontan circulation (n=276) at 10, 15,
258
20 and 25 years was 97% (95% CI: 94-99%), 95% (95% CI: 92-98%), 91% (95% CI: 87-
259
95%) and 85% (95% CI: 74-96%), respectively. Competing events cumulative incidence
260
curves for all patients are displayed in Figure 1. There were 21 late deaths. The causes of
Page 12 of 28
13 261
death were low cardiac output syndrome in 4 patients, stroke in 3, intractable ventricular
262
arrhythmias in 3, massive pulmonary embolism in 2, sepsis in 2, pulmonary haemorrhage in
263
1, protein-losing enteropathy in 1, traumatic injury in 1, and unknown in 4. Independent
264
predictors of late mortality are displayed in Table 2.
265 266
Late failure and reintervention
267
Late failure occurred in 45 hospital survivors discharged with an intact Fontan circulation,
268
consisting of 21 late deaths, 4 heart transplantations, 7 conversions to ECC, 1 Fontan
269
takedown, 9 PLE/PB and 3 patients who where classified as NYHA Class III at last follow-
270
up. Freedom from late failure was 91% (95% CI: 88-95%) at 10 years, 88% (95% CI: 84-
271
92%) at 15 years, 83% (95% CI: 78-88%) at 20 years, and 82% (95% CI: 76-87%) at 25 years
272
(Figure 2). Independent predictors of late failure are displayed in Table 2. . Multivariable
273
predictors of late failure prior to backward elimination are included in the Appendix.
274
Major late reinterventions undertaken in hospital survivors discharged with an intact Fontan
275
circulation consisted of Damus-Kaye-Stansel procedure in 7 patients, Fontan circuit revision
276
in 6, Maze procedure in 6, atrioventricular valve repair or replacement in 5, pulmonary artery
277
reconstruction in 5, pleurodesis in 4, ligation of the left superior vena cava in 2, thoracic duct
278
ligation in 2, aortic root replacement (Bentall’s procedure) in 2, neo-aortic valve replacement
279
in 1, aortic valve repair in 1, aortic root reconstruction in 1, and diaphragm plication in one
280
patient. Transcatheter intervention was performed in 26 patients (9%), with radiofrequency
281
arrhythmia ablation in 11, embolization of venovenous or aortopulmonary collaterals in 6,
282
occlusion of left superior vena cava in 3, pulmonary artery angioplasty ± stent in 3,
283
fenestration dilatation in 2, and coronary artery stenting in one patient. The fenestration was
284
closed in 35 of the 148 patients (24%) who had a primary fenestration.
285
Page 13 of 28
14 286
Arrhythmia
287
A total of 38 patients (14%) developed SVT, and freedom from SVT at 10, 20 and 25 years
288
was 90% (95% CI: 86-93%), 76% (95% CI: 70-82%) and 65% (95% CI: 55-74%). Freedom
289
from tachy- or bradyarrhythmias at 10 and 20 years was 87% (95% CI: 83-91%) and 72%
290
(95% CI: 66-79%), respectively. On multivariable Cox regression analysis, only right atrial
291
isomerism was independently associated with development of SVT (HR 3.41, 95% CI: 1.35-
292
8.64, p=0.009).
293
Pacemakers were implanted in 41 patients (14%). Five patients had a pacemaker in situ at the
294
time of Fontan, and the remaining 36 patients had a pacemaker inserted at the time of Fontan
295
(2 patients), during the Fontan hospital stay (5 patients) or during follow-up (29 patients).
296
Amongst hospital survivors discharged with an intact Fontan circulation, freedom from
297
pacemaker implantation at 10, 20 and 25 years were 89% (95% CI: 85-93%), 84% (95% CI:
298
79-89%) and 82% (95% CI: 76-88%) respectively
299 300
Anticoagulation and thromboembolic events
301
Of the 242 surviving patients with an intact LT Fontan at last follow-up, 203 patients (84%)
302
were on anticoagulation therapy or antiplatelet agents, consisting of warfarin in 108 patients
303
(45%), aspirin in 89 (37%), low-molecular-weight heparins in 2 (0.8%) and a combination of
304
agents in 4 (1.6%). No anticoagulation or antiplatelet agent was utilized in 29 patients (12%),
305
and the method of anticoagulation was unknown in 10 (4%).
306
Symptomatic thromboembolism occurred in 45 patients (16%) during follow-up. These
307
included 26 strokes, 9 pulmonary emboli, 7 transient ischemic attacks, 1 peripheral embolus,
308
1 renal embolus and 1 paradoxical embolic myocardial infarction. Thirteen of the patients
309
who suffered a symptomatic thromboembolic (13/45, 29%) event had a history of SVT. A
310
further 22 patients had a thrombus detected within the conduit or central venous circulation on
Page 14 of 28
15 311
routine echocardiography during follow-up. Freedom from symptomatic thromboembolism at
312
10, 20 and 25 years was 93% (95% CI: 89-96%), 80% (95% CI: 74-86%), and 80% (95% CI:
313
74-86%), respectively. On multivariable Cox regression analysis, pre-Fontan common
314
atrioventricular valve regurgitation ≥ moderate (HR 9.34, 3.80-22.93, p<0.001) and
315
development of SVT (HR 2.67, 95% CI: 1.06-6.71, p=0.04) were associated with
316
thromboembolism. Age < 3 years at the time of Fontan was associated with a reduced
317
likelihood of thromboembolism (HR 0.37, 95% CI: 0.15-0.88, p=0.02). Ten of the 45 patients
318
who experienced a symptomatic thromboembolic event during follow-up (22%) had a
319
documented history of tachyarrhythmia prior to the event.
320 321
Discussion
322
More than twenty-five years after its original description, the LT modification of the
323
originally-described Fontan procedure has demonstrated is superiority over the former AP
324
Fontan. The twenty-five-year survival of the patients discharged from hospital with an intact
325
LT circulation has reached a remarkable 85% within our population.
326
Estimates of late survival from the Mayo Clinic of 84%, 70% and 39% at 10, 20 and 30 years
327
after the LT Fontan, respectively, are less optimistic than our current series. They have
328
suggested that this may have been related to their large number of patients with heterotaxy
329
syndromes and the incorporation of larger amounts of native atrial tissue into surgical repair,
330
rendering their patients more susceptible to some of the complications observed after the
331
atriopulmonary connection[12]. Smaller patient numbers beyond 20-years in both instances
332
are also likely to contribute to the observed discrepancies in survival estimates. Our series is a
333
historical group of patients. One should be aware that, today it is likely that the lateral tunnel
334
procedure would be reserved to very small patients because it avoids the need for
335
implantation of bulky adult-size conduit. In many centers the lateral tunnel procedure would
Page 15 of 28
16 336
therefore be reserved to patients with more severe condition such as those with hypoplastic
337
left heart syndrome (HLHS) which were present in only very small numbers in our series. It is
338
therefore possible that contemporary results of the lateral tunnel would be worse than those
339
presented here.
340
Despite excellent survival rates, this population seems to be subjected to a slow but consistent
341
failure of their Fontan circulation. We have now demonstrated that none of the three main
342
types of Fontan circulation have experienced a sudden drop in their survival as was initially
343
expected, however, there still seems to be constant rate of attrition over time[13]. The exact
344
reasons for the failure of the Fontan circulation remain to be elucidated, and therefore the best
345
way to prevent the decline is still unclear [14-16]. One may be surprised to notice a
346
discrepancy between the large number of patients with a failed Fontan circulation and the
347
small number of those reaching heart transplantation. This phenomenon has been observed
348
worldwide, and is attributed to the complexity of undertaking transplantation in these patients,
349
the risk to undergo the procedure and inequality of access to transplantation[12, 17, 18]. The
350
lateral tunnel technique was primarily introduced in order to decrease the rate of arrhythmias
351
developing after Fontan completion, and it seems that it has achieved this goal [7]. We have
352
previously demonstrated that the incidence of arrhythmias was significantly decreased by the
353
introduction of the LT technique, which is again supported by our 20-year overall freedom
354
from arrhythmias of around 70% of patients [2].
355
Although ongoing failure and late onset of arrhythmia were not unexpected findings, we can
356
now demonstrate that that late failure is associated with the eventual occurrence of
357
arrhythmias. The onset of arrhythmias have been shown to predict mortality in other Fontan
358
populations around the world [19]. However, it must be acknowledged that the development
359
of arrhythmias may indeed be just one manifestation of a failing Fontan.
Page 16 of 28
17 360
We may face a similar phenomenon that was earlier observed in the patients with an
361
atriopulmonary connection, albeit at a later stage. In the LT Fontan, half of the venous circuit
362
draining the blood from the inferior vena cava to the pulmonary arteries is made of native
363
atrial tissue. With time, this atrial tissue may dilate and this stretch of the atrial wall, in
364
conjunction with the long suture lines, may form triggers for supra-ventricular
365
tachyarrhythmias. Atrial stretching may occur at a later stage than in patients with an
366
atriopulmonary connection as the wall stress is reduced by the elimination of turbulent flow
367
within the atrium. Serial imaging of the LT conduit may be useful to aid risk stratification,
368
however, this is yet to be fully investigated.
369
The incidence of thromboembolic events, and the number of massive pulmonary emboli,
370
seems to be increased compared to what we have observed at the same time-points in our
371
ECC cohort. The reasons for this are unclear, although the increased turbulence within a
372
dilated atrial channel may predispose to later thrombus formation [20]. This is supported by
373
the additional number of patients in the current study who were noted to have thrombus
374
present within the LT Fontan circuit.
375 376
Limitations
377
The incidence of arrhythmias, particularly those that were intermittent or transient in nature,
378
may be underestimated due to the limited availability of Holter monitoring or serial
379
electrocardiography data.
380
Due to the low number of patients with HLHS in this study, we were unable to draw any
381
significant association between HLHS morphology and an increased risk of adverse events,
382
failure or mortality.
Page 17 of 28
18 383
Another important limitation is the inability to accurately assess NYHA status retrospectively
384
from medical records. The numbers derived for late Fontan failure may therefore not reflect
385
the true burden of symptoms and non-cardiac morbidities.
386
Prolonged pleural effusions were defined in this study as lasting greater than 30 days, a
387
definition which differs from that of other institutions that have used greater than 14 days.
388
This is definition is based on how the data has been collected in the Registry.
389 390
Conclusion
391
Over a twenty-five-year period, the LT technique has achieved excellent late survival. As this
392
population ages, they face an increasing risk of failure and adverse events, including
393
arrhythmia and thromboembolism. We are likely to see a growing number of these survivors
394
requiring heart transplantation and late reintervention. Efforts must be made to develop
395
treatment strategies for the increasing number of Fontan patients that will continue to
396
experience attrition over time.
397 398
Acknowledgements
399
The authors thank the Murdoch Childrens Research Institute for infrastructure support. The
400
authors also acknowledge the Fontan Registry management and research assistants for their
401
invaluable support in the creation and maintenance of the Registry, as well as support in data
402
gathering for this paper, as well as Belinda Bortone for administrative support. The authors
403
acknowledge support provided to the Murdoch Childrens Research Institute by the Victorian
404
Government's Operational Infrastructure Support Program.
405
Page 18 of 28
19 406
Disclosures
407
Yves d’Udekem is a consultant for MSD and Actelion. Andrew Bullock reports consulting
408
fees from Actelion. All other authors have nothing to disclose with regard to commercial
409
support.
Page 19 of 28
20 410
References
411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457
[1] Iyengar AJ, Winlaw DS, Galati JC, Gentles TL, Weintraub RG, Justo RN, et al. The Australia and New Zealand Fontan Registry: description and initial results from the first population-based Fontan registry. Internal medicine journal. 2014;44:148-55. [2] d'Udekem Y, Iyengar AJ, Cochrane AD, Grigg LE, Ramsay JM, Wheaton GR, et al. The Fontan procedure: contemporary techniques have improved long-term outcomes. Circulation. 2007;116:I157-64. [3] Fontan F, Baudet E. Surgical repair of tricuspid atresia. Thorax. 1971;26:240-8. [4] Kreutzer J, Keane JF, Lock JE, Walsh EP, Jonas RA, Castaneda AR, et al. Conversion of modified Fontan procedure to lateral atrial tunnel cavopulmonary anastomosis. The journal of thoracic and cardiovascular surgery. 1996;111:1169-76. [5] Quinton E, Nightingale P, Hudsmith L, Thorne S, Marshall H, Clift P, et al. Prevalence of atrial tachyarrhythmia in adults after Fontan operation. Heart. 2015;101:1672-7. [6] Lardo AC, Webber SA, Friehs I, del Nido PJ, Cape EG. Fluid dynamic comparison of intra-atrial and extracardiac total cavopulmonary connections. The journal of thoracic and cardiovascular surgery. 1999;117:697-704. [7] de Leval MR, Kilner P, Gewillig M, Bull C. Total cavopulmonary connection: a logical alternative to atriopulmonary connection for complex Fontan operations. Experimental studies and early clinical experience. The journal of thoracic and cardiovascular surgery.1988;96(5):682-95. [8] Iyengar AJ, Winlaw DS, Galati JC, Wheaton GR, Gentles TL, Grigg LE, et al. The extracardiac conduit Fontan procedure in Australia and New Zealand: hypoplastic left heart syndrome predicts worse early and late outcomes. European journal of cardio-thoracic surgery : official journal of the European Association for Cardio-thoracic Surgery. 2014;46:465-73. [9] Stewart RD, Pasquali SK, Jacobs JP, Benjamin DK, Jaggers J, Cheng J, et al. Contemporary Fontan operation: association between early outcome and type of cavopulmonary connection. The Annals of thoracic surgery. 2012;93:1254-60. [10] d'Udekem Y, Iyengar AJ, Galati JC, Forsdick V, Weintraub RG, Wheaton GR, et al. Redefining expectations of long-term survival after the Fontan procedure: twenty-five years of follow-up from the entire population of Australia and New Zealand. Circulation. 2014;130:S32-S8. [11] Scrucca L, Santucci A, Aversa F. Competing risk analysis using R: an easy guide for clinicians. Bone Marrow Transplant. 2007;40:381-7. [12] Pundi KN, Johnson JN, Dearani JA, Pundi KN, Li Z, Hinck CA, et al. 40-Year FollowUp After the Fontan Operation: Long-Term Outcomes of 1,052 Patients. Journal of the American College of Cardiology. 2015;66:1700-10. [13] Schilling C, Dalziel K, Nunn R, Du Plessis K, Shi WY, Celermajer D, et al. The Fontan epidemic: Population projections from the Australia and New Zealand Fontan Registry. International journal of cardiology. 2016;219:14-9. [14] de Leval MR, Deanfield JE. Four decades of Fontan palliation. Nat Rev Cardiol. 2010;7:520-7. [15] Rychik J. The Relentless Effects of the Fontan Paradox. Semin Thorac Cardiovasc Surg Pediatr Card Surg Annu. 2016;19:37-43. [16] Ohuchi H. Adult patients with Fontan circulation: What we know and how to manage adults with Fontan circulation? J Cardiol (2016), http://dx.doi.org/10.1016/j.jjcc.2016.04.001. [17] Shi WY, Yong MS, McGiffin DC, Jain P, Ruygrok PN, Marasco SF, et al. Heart transplantation in Fontan patients across Australia and New Zealand. Heart. 2016;102:1120-6.
Page 20 of 28
21 458 459 460 461 462 463 464 465 466 467 468
[18] Iyengar AJ, Sharma VJ, Weintraub RG, Shipp A, Brizard CP, d'Udekem Y, et al. Surgical strategies to facilitate heart transplantation in children after failed univentricular palliations: the role of advanced intraoperative surgical preparation. European Journal of Cardio-Thoracic Surgery. 2014;46:480-5. [19] Giannakoulas G, Dimopoulos K, Yuksel S, Inuzuka R, Pijuan-Domenech A, Hussain W, Tay EL, et al. Atrial tachyarrhythmias late after Fontan operation are related to increase in mortality and hospitalization. International journal of cardiology. 2012;157:221-6. [20] Iyengar AJ, Winlaw DS, Grigg LE, Celermajer DS, d’Udekem Y. No difference between aspirin and warfarin after extracardiac Fontan in a propensity score analysis of 475 patients. European Journal of Cardio-Thoracic Surgery. 2016; doi:10.1093/ejcts/ezw159.
469 470
Page 21 of 28
22 471
Figure 1: Competing events cumulative incidence curves for all patients (n=301). Late
472
survival after the lateral tunnel Fontan is excellent with an overall survival of 82% at 20
473
years.
474
Figure 2: Freedom from Fontan failure for those surviving to hospital discharge with an
475
intact Fontan circulation (n=276). Late Fontan failure was defined as death, takedown,
476
transplantation, Fontan conversion/revision, NYHA class III/IV, protein-losing enteropathy or
477
plastic bronchitis.
478 479
Page 22 of 28
23 480
Table 1: Patient characteristics Characteristics Male, n (%)
Total ( n = 301) 165 (55%)
Anatomical comorbidities, n (%) Dextrocardia/Mesocardia
24 (8%)
Isomerism/Heterotaxy
21 (7%)
Non-cardiac anomaly
42 (14%)
Ventricular morphology, n (%) Left
188 (63%)
Right
81 (27%)
Biventricular/Indeterminate
32 (11%)
Primary morphological diagnosis, n (%) TA
81 (27%)
DILV
65 (22%)
DORV
50 (17%)
CAVC
28 (9%)
PA-IVS
19 (6%)
ccTGA
18 (6%)
TGA
12 (4%)
HLHS
9 (3%)
PA-VSD
6 (2%)
Other
13 (4%)
Pre-Fontan procedures Number of prior palliations, mean (SD) Prior aortic arch intervention, n (%)
2 (1) 39 (13%)
Page 23 of 28
24 Prior pulmonary artery banding, n (%)
90 (30%)
Prior staging with BCPS, n (%)
120 (40%)
Bilateral BCPS, n (%) Age at BCPS in years, median (IQR) Atrioventricular valve repair/replacement, n (%) Pulmonary artery reconstruction or angioplasty, n (%)
7 (2%) 1.4 (0.8-2.8) 7 (2%) 35 (12%)
Pre-Fontan haemodynamics Oxygen saturation (%), mean (SD)
81 (7)
Pulmonary artery pressure (mmHg), mean (SD)
12 (4)
Aortopulmonary or venovenous collaterals, n (%)
25 (8%)
Atrio-venous malformations, n (%)
7 (2%)
Atrioventricular valve regurgitation ≥ moderate, n (%)
12 (4%)
Fontan operative characteristics Age at Fontan in years, median (IQR)
3.8 (2.8 – 5.9)
Concomitant procedure, n (%)
77 (26%)
Concomitant pulmonary artery reconstruction, n (%)
20 (6%)
Concomitant atrioventricular valve repair/replacement, n (%)
8 (3%)
Fenestration, n (%)
156 (52%)
481
TA: tricuspid atresia; DILV: double-inlet left ventricle; DORV: double-outlet right ventricle;
482
CAVC: common atrioventricular canal; PA-IVS: pulmonary atresia with intact ventricular
483
septum; ccTGA: congenitally-corrected transposition of the great arteries; TGA: transposition
484
of the great arteries; HLHS: hypoplastic left heart syndrome; PA-VSD: pulmonary atresia
485
with ventricular septal defect; SD: standard deviation; BCPS: bidirectional cavopulmonary
486
shunt; IQR: interquartile range; mmHg: millimeters of mercury.
487
Page 24 of 28
25 488
Table 2: Results of multivariable Cox regression for late outcomes Predictors of late mortality Hazard
95% CI
95% CI
ratio
Lower
Upper
2.84
1.06
7.60
0.04
6.29
1.30
30.45
0.02
8.11
2.97
22.16
<0.001
Variable
p-value
Male Pre-Fontan common atrioventricular valve regurgitation ≥ moderate Prolonged pleural effusions
Predictors of late Fontan failure Hazard
95% CI
95% CI
Variable
p-value Ratio
Lower
Upper
Development of SVT#
4.68
2.07
10.58
<0.001
Age >7 (vs. 3-5) at Fontan
9.69
2.46
38.21
0.001
Prolonged pleural effusions
3.06
1.05
8.95
0.04
489
#
490
CI: confidence interval; SVT: supraventricular tachycardia.
Development of SVT was analyzed as a time-dependent covariate.
491 492
Page 25 of 28
26 493
Appendix 1. Multivariable predictors of late mortality prior to backward elimination. HR
95.0% CI for HR
p value
Lower
Upper
8.982
2.026
39.822
.004
Development of SVT
4.973
1.449
17.063
.011
Pre-Fontan atrioventricular
9.834
1.438
67.267
.020
Prior PA banding
4.019
1.091
14.806
.037
Male
2.698
.844
8.617
.094
Prior aortic arch intervention
.288
.042
1.970
0.20
Age 5-7 (vs. 3-5)
.543
.093
3.177
0.50
Prior BCPS
.613
.147
2.562
0.50
Fenestration at Fontan
.675
.212
2.151
0.51
LV dominance
1.240
.481
3.196
0.66
Age < 3 (vs. 3-5)
.742
.155
3.546
0.71
Age > 7 (vs. 3-5)
1.315
.292
5.912
0.72
Era 1990-1999 (vs. pre-1990)
1.143
.248
5.275
0.86
Era 2000-2014 (vs.pre-1990)
.000
n/a
n/a
0.99
HLHS
.000
n/a
n/a
0.99
Prolonged post-operative pleural effusions
valve regurgitation ≥ moderate
494 495
Page 26 of 28
27 496
Appendix 2. Multivariable predictors of late failure prior to backward elimination. HR
95.0% CI for HR
Sig.
Lower
Upper
Development of SVT
5.294
2.271
12.338
<0.0001
Prolonged effusions
3.510
1.131
10.890
.030
Age > 7 (vs. 3-5)
2.648
.867
8.089
.087
Pre Fontan AV regurg
2.507
.448
14.032
0.30
Fenestration
.629
.258
1.532
0.31
HLHS
3.524
.302
41.175
0.32
Age 5-7 (vs. 3-5)
1.618
.494
5.300
0.43
Male
1.360
.633
2.922
0.43
Era 2000-2015 (vs. pre 1990)
.379
.031
4.672
0.45
Era 1990-1999 (vs. pre 1990)
.725
.267
1.968
0.53
LV dominance
1.228
.627
2.403
0.55
Prior BCPS
1.339
.512
3.500
0.55
Age < 3 (vs. 3-5)
1.152
.358
3.703
0.81
Prior aortic arch
.882
.221
3.525
0.86
Prior PA banding
.999
.360
2.771
1.00
497
SVT: supraventricular tachycardia; PA: pulmonary artery; BCPS: bidirectional
498
cavopulmonary shunt; LV: left ventricle; HLHS: hypoplastic left heart syndrome.
499
Page 27 of 28
28 500
Page 28 of 28
S1043-0679(17)30166-1 http://dx.doi.org/doi: 10.1053/j.semtcvs.2017.06.002 YSTCS 989
To appear in:
Seminars in Thoracic and Cardiovascular Surgery
Please cite this article as: Thomas G Wilson, William Y Shi, Ajay J Iyengar, David S Winlaw, Rachael L Cordina, Gavin R Wheaton, Andrew Bullock, Thomas L Gentles, Robert G Weintraub, Robert N Justo, Leeanne E Grigg, Dorothy J Radford, Yves d'Udekem, The Australia and New Zealand Fontan Registry, Twenty-Five Year Outcomes of the Lateral Tunnel Fontan Procedure, Seminars in Thoracic and Cardiovascular Surgery (2017), http://dx.doi.org/doi: 10.1053/j.semtcvs.2017.06.002. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. 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.
1 1
Twenty-Five Year Outcomes of the Lateral Tunnel Fontan Procedure
2 3
Authors
4
Thomas G Wilson, BSc, MD1, 2, 3, William Y Shi, MBBS, 1, 2, 3, Ajay J Iyengar, MBBS(Hons),
5
BMedSci, PhD1, 2, 3, David S Winlaw, MBBS(Hons), MD, FRACS4, 5, Rachael L Cordina,
6
MBBS, FRACP5, 6, Gavin R Wheaton, MBBS, FRACP 7, Andrew Bullock, MBBS, FRACP 8,
7
Thomas L Gentles, MBChB, FRACP9, Robert G Weintraub, MBBS, FRACP2, 10, Robert N
8
Justo, MBBS, FRACP 11, Leeanne E Grigg, MBBS, FRACP12, Dorothy J Radford, MBBS,
9
MD, FRACP13, 14, Yves d’Udekem, MD, PhD1, 2, 3, The Australia and New Zealand Fontan
10
Registry
11 12
Institutions and Affiliations
13
1. Department of Cardiac Surgery, Royal Children’s Hospital, Melbourne, Australia
14
2. Heart Research Group, Murdoch Childrens Research Institute, Melbourne, Australia
15
3. Department of Paediatrics, Faculty of Medicine, The University of Melbourne,
16 17 18
Victoria, Australia 4. The Heart Centre for Children, The Children’s Hospital at Westmead, Sydney, Australia
19
5. Department of Paediatrics, University of Sydney, Sydney, Australia
20
6. Department of Cardiology, Royal Prince Alfred Hospital, Sydney, Australia
21
7. Department of Cardiology, Women’s and Children’s Hospital, Adelaide, Australia
22
8. Children’s Cardiac Centre, Princess Margaret Hospital for Children, Perth, Australia
23
9. Greenlane Paediatric and Congenital Cardiac Service, Starship Children’s Hospital,
24 25
Auckland, New Zealand 10. Department of Cardiology, Royal Children’s Hospital, Melbourne, Australia
Page 1 of 28
2 26 27
11. Queensland Paediatric Cardiac Service, Lady Cilento Children’s Hospital, Brisbane, Queensland, Australia
28
12. Department of Cardiology, The Royal Melbourne Hospital, Melbourne, Australia
29
13. Adult Congenital Heart Unit, The Prince Charles Hospital, Brisbane, Australia
30
14. Faculty of Medicine, University of Queensland, Brisbane, Australia
31 32
Corresponding Author
33
A/Prof Yves d’Udekem, MD PhD
34
Department of Cardiac Surgery, Royal Children’s Hospital
35
Flemington Rd, Parkville VIC 3052
36
Phone: +61 3 9345 5200 Fax: +61 3 9345 6001
37
E-mail: [email protected]
38 39
Acknowledgement of Grant Support
40
The Australia and New Zealand Fontan Registry is funded by grants from the National Health
41
and Medical Research Council (NHMRC; Project Grants 1012241, 1047923, 1065794). The
42
authors acknowledge support provided to the Murdoch Childrens Research Institute by the
43
Victorian Government’s Operational Infrastructure Support Program. Dr William Shi is
44
supported by the Royal Australasian College of Surgeons Foundation for Surgery Peter King/
45
Heart Foundation Research Scholarship in addition to the University of Melbourne Viola
46
Edith Reid and the RG and AU Meade Scholarships. Yves d’Udekem is a NHMRC Clinician
47
Practitioner Fellow (1082186). The Victorian Government’s Operational Infrastructure
48
Support Program supported this research project.
49
Page 2 of 28
3 50
Conflicts of Interest
51
Yves d’Udekem is a consultant for MSD and Actelion. Andrew Bullock reports consulting
52
fees from Actelion. All other authors have nothing to disclose with regard to commercial
53
support.
54 55
Keywords
56
Fontan Procedure, Heart Ventricles Abnormalities, Heart Defects Congenital, Retrospective
57
Studies, Survival Rate, Follow-Up Studies, Disease-Free Survival
58 59
Article word count: 4,295
60
Page 3 of 28
4 61
Abstract
62
Objective(s): To characterize late outcomes of the lateral tunnel (LT) Fontan procedure.
63
Methods: The outcomes of all patients who underwent a LT Fontan procedure in Australia
64
and New Zealand were analysed. Original files were reviewed and outcomes data were
65
obtained through a binational Registry.
66
Results: Between 1980 and 2014, a total of 301 patients underwent a LT Fontan procedure
67
across 6 major centers. There were 13 hospital mortalities, 21 late deaths, 8 Fontan
68
conversions/revisions, 8 Fontan takedowns and 4 heart transplantations. Overall survival at 15
69
and 25 years was 90% (95% confidence interval [CI]:86-93%) and 80% (95% CI:69-91%),
70
respectively. Protein-losing enteropathy/plastic bronchitis was observed in 14 patients (5%).
71
Freedom from late failure at 15 and 25 years was 88% (95% CI:84-92%) and 82% (95%
72
CI:76-87%), respectively. Independent predictors of late Fontan failure were prolonged
73
pleural effusions post-Fontan operation (HR 3.06,1.05-8.95, p=0.041), age >7 years at Fontan
74
(vs. 3-5 years, HR 9.7, 2.46-38.21, p=0.001) and development of supraventricular tachycardia
75
(4.67, 2.07-10.58, p<0.001). Freedom from tachy- or bradyarrhythmias at 10 and 20 years was
76
87% (95% CI:83-91%) and 72% (95% CI:66-79%), respectively. Thromboembolic events
77
occurred in 45 patients (16%; 26 strokes), and freedom from symptomatic thromboembolism
78
at 10 and 20 years was 93% (95% CI:89-96%) and 80% (95% CI:74-86%), respectively.
79
Conclusions: Over a twenty-five-year period, the LT technique has achieved excellent late
80
survival. As this population ages, they are at increasing risk of failure and adverse events. We
81
are likely to see an increasing proportion requiring heart transplantation and late
82
reintervention.
83 84
Abstract word count: 249
Page 4 of 28
5 85
Glossary of abbreviations
86
LT = lateral tunnel
87
CI = confidence interval
88
HR = hazard ratio
89
SVT = supraventricular tachycardia
90
AP = atriopulmonary
91
ECC = extracardiac conduit
92
VAD = ventricular assist device
93
ECMO = extracorporeal membrane oxygenation
94
PLE = protein-losing enteropathy
95
PB = plastic bronchitis
96
NYHA = New York Heart Association
97
SD = standard deviation
98
IQR = interquartile range
99
DKS = Damus–Kaye–Stansel
100
HLHS = hypoplastic left heart syndrome
101 102
Page 5 of 28
6 103
Central message: Late survival after the lateral tunnel Fontan procedure is excellent. This
104
population is subjected to an ongoing risk of late failure.
105 106
Perspective statement: Late survival after the lateral tunnel Fontan procedure is excellent.
107
There is, however, an ongoing risk of failure of the Fontan circulation this population. The
108
onset of new arrhythmia predicts later occurrence of failure and it is likely that careful
109
monitoring of this population is warranted as their atrial channel might be subjected to
110
progressive dilatation.
111 112
Central picture: Competing events cumulative incidence curves for all patients (n=301).
113 114
Page 6 of 28
7 115
Introduction
116
The efficacy of the Fontan procedure in the treatment of children born with single ventricle
117
cardiac malformations continues to be unveiled as this population progresses further into
118
adulthood[1]. It is believed that refinements in surgical techniques have led to an
119
improvement in outcomes since the original atriopulmonary (AP) Fontan, which has now
120
been abandoned[2, 3]. Progressive atrial distension and the development of turbulent flow
121
within the atrial chamber predisposed recipients to tachyarrhythmias, thromboembolic events
122
and even compression of the pulmonary veins [6]. In 1988, the concept of total
123
cavopulmonary circulation was proposed by de Leval in an attempt to avoid turbulent flow
124
within the atrium, and his design of the intra-atrial lateral tunnel (LT) Fontan was widely
125
adopted[7]. In the 1990s, the extra-cardiac conduit (ECC) progressively came into vogue and
126
today is the most frequently performed type of Fontan [8, 9]. Many patients have undergone
127
a LT Fontan operation and the specific outcomes of this form of Fontan are unclear. In one
128
United States multicentre database, more than a third of the Fontan operations performed
129
between 2000 and 2009 were undertaken using the LT technique[9]. Locally, a quarter of the
130
patients with a Fontan circulation currently followed within our region are living with a LT
131
Fontan type[1, 10]. While the implementation of this technique has resulted in improved
132
survival up to early adulthood, the rate of failure of their circulation is still unclear. We sought
133
to characterize the late outcomes of the patients who have undergone a LT Fontan procedure
134
in Australia and New Zealand.
135 136
Methods
137
Ethical Review Board approvals were obtained at all participating institutions. The need
138
for specific consent was waived because of the retrospective nature of the study.
Page 7 of 28
8 139
The full design, data fields and administration of the Australia and New Zealand Fontan
140
Registry have been described previously[1, 10]. Of the 1,462 participants whose data were
141
collected in the Registry database between 1 January 1975 and 1 August 2015, all those who
142
underwent a LT Fontan were identified. Fontan conversions from atriopulmonary (AP)
143
connection and Bjork procedures were excluded.
144
Perioperative variables were extracted from the database and follow-up data were updated
145
where necessary via contact with patients’ cardiologists and obtaining their latest clinical
146
summaries and echocardiogram reports. Medical histories were screened for missing data.
147
The LT technique was gradually phased out between 1997 and 2006 because of the potential
148
theoretical advantages of the ECC, whereafter the ECC modification was solely performed
149
throughout the region. Over this time period, decision to fenestrate the Fontan was left to
150
surgical preference.
151 152
Definitions
153
Hospital mortality was defined as death during the initial hospital stay. Prolonged pleural
154
effusion was defined as effusions requiring pleural drainage persisting for longer than 30 days
155
or requiring reoperation including pleurodesis. Early failure was defined as death, Fontan
156
takedown, Fontan revision or mechanical support with a ventricular assist device (VAD) or
157
extracorporeal membrane oxygenation (ECMO) within 30 days or during the initial hospital
158
stay. Late failure was defined as death, Fontan takedown or conversion to ECC after hospital
159
discharge, heart transplantation, protein losing enteropathy (PLE), plastic bronchitis (PB) or
160
New York Heart Association (NYHA) class III/IV. Concomitant procedures were defined as
161
intracardiac procedures performed at the time of the Fontan operation, excluding atrial
162
septectomy.
Page 8 of 28
9 163
Analyses of long-term outcomes were performed only on those who were discharged from
164
hospital with an intact Fontan circulation (i.e. excluding Fontan takedowns performed during
165
the same admission as their initial operation).
166 167
Statistical Analysis
168
Values are reported as absolute value (percentage [%]) for proportions, mean (standard
169
deviation [SD]) for normally distributed data and median (interquartile range [IQR]) for non-
170
normally distributed data. Where appropriate, these data were compared between groups
171
using χ2 tests, t-tests and Wilcoxon rank-sum tests, respectively.
172
Multivariable logistic regression was used to identify independent predictors of early
173
outcomes, and Cox regression was utilised to identify predictors of early and late outcomes.
174
The backward elimination method was used in regression analyses to include all potential
175
predictors of end points in the initial models and subsequently eliminate covariates in an
176
iterative process to create a final model. Variables with at least moderate evidence against the
177
null hypothesis (P<.10) were included. Inclusion of all potentially important variables in the
178
initial model allows their joint predictive behaviour to be initially evaluated, which is
179
important given that a set of variables may exhibit predictive capability even if a subset does
180
not. The variables used in regression analyses were selected based on those found to be
181
associated with outcomes in prior studies by our group as well as based on our group’s
182
clinical experience. Due to the nonlinear relationship of age at the time of Fontan with the risk
183
of late outcomes, patients were grouped into categories based on their age at Fontan: <3 years,
184
3 to 5 years, 5 to 7 years, and >7 years, with 3 to 5 years as reference group. Similarly,
185
oxygen saturation prior to the Fontan procedure was divided into categories: <77%, 78% to
186
81%, 82% to 85%, and >86%, and comparisons were made using the 82% to 85% category as
187
a reference group. Linearity was examined by performing a likelihood ratio test comparing
188
regression models with continuous covariates against models in which the numeric variable
Page 9 of 28
10 189
had been categorically transformed. Age at Fontan operation had nonlinear associations with
190
mortality and the other end points and was analyzed as a categoric variable.
191
The time of the first onset of an arrhythmia was inserted into the multivariable Cox regression
192
model as a time-varying covariate and its associated hazard ratio (HR) calculated. In essence,
193
patient follow-up was split into two periods: before and after the onset of arrhythmia,
194
allowing the association between the development of an arrhythmia and subsequent late end-
195
points to be established. Analysis of long-term outcomes where onset of SVT was used as a
196
co-variate only occurred in those who were discharged with an intact Fontan circulation. As
197
such, the onset of SVT and its association with early failure was not studied.
198
Estimated cumulative incidence curves for each competing event post-Fontan (Alive, death,
199
transplant, takedown, conversion to ECC) were constructed using the cumulative incidence
200
function described by Scrucca et al[11]. Here, the cumulative incidence function is defined as
201
the probability of failing from cause r (r = 1, …, k, where k is the number of causes of failure)
202
up to a certain time point t such that:
203
Where λr(t) is the cause specific hazard rate and S(t) = Pr(T≥t) is the survival function. Non-
204
parametric maximum likelihood estimation of (cause-specific) cumulative incidence is:
205
.
206 207
Results
208
A total of 301 patients were identified from hospital and surgical databases. Patient
209
characteristics are detailed in Table 1.
210 211
Patient and surgical characteristics
Page 10 of 28
11 212
At the time of the Fontan operation, 71 patients (24%) underwent concomitant procedures,
213
consisting of pulmonary artery reconstruction in 21 patients, Damus–Kaye–Stansel procedure
214
in 8, atrioventricular valve repair in 9, subaortic resection in 8, semilunar valve repair in 4,
215
ventricular septal defect enlargement in 5, pulmonary artery band revision in 4, main
216
pulmonary artery ligation or division in 5, aortic arch reconstruction in 2, arterial switch
217
operation in 2, epicardial pacemaker insertion in 2 and other concomitant procedures in 11.
218 219
The majority of the LT Fontan procedures were performed across five major centres (295
220
patients), with a minority undergoing Fontan completion at smaller or international centres (6
221
patients).
222 223
Early outcomes
224
There were 13 hospital mortalities (4%). Causes of death were low cardiac output in 5
225
patients, sepsis in 3, massive pulmonary emboli in 2, hemorrhagic stroke in 2, and unknown
226
in one patient. No deaths occurred out of hospital within the first 30 days after surgery. A
227
total of 4 patients underwent Fontan takedown during the initial hospital stay, due to low
228
cardiac output syndrome in 2 patients, prolonged chylous effusions with obstruction of the LT
229
conduit in 1, and thrombosis of the Fontan circuit in one patient.
230
The median length of hospital stay was 13 days (IQR: 9-19 days). Prolonged effusions
231
occurred in 23 patients (8%), of whom 11 underwent reoperation (pleurodesis in 6, Fontan
232
takedown/revision in 4, and fenestration dilatation in one patient) and resulted in an increase
233
in the median length of stay to 58 days (IQR: 38-79).
234
In 30 patients (10%) the length of stay exceeded 30 days, and was attributable to prolonged
235
pleural effusions in 16 patients, low cardiac output syndrome in 6, persisting arrhythmias in 3,
Page 11 of 28
12 236
stroke in 2, pneumonia in 2 and pericardial tamponade in one patient. There were no
237
independent predictors for having prolonged pleural effusions.
238
Two patients required mechanical circulatory support (1 ECMO, 1 VAD) with the former
239
subsequently undergoing an early Fontan takedown on postoperative day 2, and the latter
240
dying on postoperative day 8 from persisting low cardiac output and renal failure. Early
241
failure (death, takedown, revision or ECMO/VAD) occurred in 21 patients (7%). On
242
multivariable logistic regression, there were no independent predictors of early Fontan failure.
243
Clinical evidence of perioperative cerebral infarction was observed in 20 patients (7%) during
244
the Fontan admission. These consisted of embolic stroke in 10 patients, hemorrhagic stroke in
245
3, and cerebral ischemia secondary to a low output state in 7.
246 247
Follow-up
248
After excluding patients who moved overseas and were followed internationally (8 patients),
249
hospital mortalities (13 patients) and those who had a Fontan takedown during the initial
250
hospital stay (4 patients), 276 patients remained for survival analysis. Late follow-up was
251
available in 275 patients (99.6%). Survival analysis was undertaken on 4,749 patient-years of
252
follow-up. The median duration of follow-up was 18.6 years (IQR: 15.1-21.1).
253 254
Survival
255
Overall survival at 10, 15, 20 and 25 years was 92% (95% confidence interval [CI]: 88-95%),
256
90% (95% CI: 86-93%), 86% (95% CI: 81-90%) and 80% (95% CI: 69-91%), respectively.
257
Late survival of patients discharged alive with an intact Fontan circulation (n=276) at 10, 15,
258
20 and 25 years was 97% (95% CI: 94-99%), 95% (95% CI: 92-98%), 91% (95% CI: 87-
259
95%) and 85% (95% CI: 74-96%), respectively. Competing events cumulative incidence
260
curves for all patients are displayed in Figure 1. There were 21 late deaths. The causes of
Page 12 of 28
13 261
death were low cardiac output syndrome in 4 patients, stroke in 3, intractable ventricular
262
arrhythmias in 3, massive pulmonary embolism in 2, sepsis in 2, pulmonary haemorrhage in
263
1, protein-losing enteropathy in 1, traumatic injury in 1, and unknown in 4. Independent
264
predictors of late mortality are displayed in Table 2.
265 266
Late failure and reintervention
267
Late failure occurred in 45 hospital survivors discharged with an intact Fontan circulation,
268
consisting of 21 late deaths, 4 heart transplantations, 7 conversions to ECC, 1 Fontan
269
takedown, 9 PLE/PB and 3 patients who where classified as NYHA Class III at last follow-
270
up. Freedom from late failure was 91% (95% CI: 88-95%) at 10 years, 88% (95% CI: 84-
271
92%) at 15 years, 83% (95% CI: 78-88%) at 20 years, and 82% (95% CI: 76-87%) at 25 years
272
(Figure 2). Independent predictors of late failure are displayed in Table 2. . Multivariable
273
predictors of late failure prior to backward elimination are included in the Appendix.
274
Major late reinterventions undertaken in hospital survivors discharged with an intact Fontan
275
circulation consisted of Damus-Kaye-Stansel procedure in 7 patients, Fontan circuit revision
276
in 6, Maze procedure in 6, atrioventricular valve repair or replacement in 5, pulmonary artery
277
reconstruction in 5, pleurodesis in 4, ligation of the left superior vena cava in 2, thoracic duct
278
ligation in 2, aortic root replacement (Bentall’s procedure) in 2, neo-aortic valve replacement
279
in 1, aortic valve repair in 1, aortic root reconstruction in 1, and diaphragm plication in one
280
patient. Transcatheter intervention was performed in 26 patients (9%), with radiofrequency
281
arrhythmia ablation in 11, embolization of venovenous or aortopulmonary collaterals in 6,
282
occlusion of left superior vena cava in 3, pulmonary artery angioplasty ± stent in 3,
283
fenestration dilatation in 2, and coronary artery stenting in one patient. The fenestration was
284
closed in 35 of the 148 patients (24%) who had a primary fenestration.
285
Page 13 of 28
14 286
Arrhythmia
287
A total of 38 patients (14%) developed SVT, and freedom from SVT at 10, 20 and 25 years
288
was 90% (95% CI: 86-93%), 76% (95% CI: 70-82%) and 65% (95% CI: 55-74%). Freedom
289
from tachy- or bradyarrhythmias at 10 and 20 years was 87% (95% CI: 83-91%) and 72%
290
(95% CI: 66-79%), respectively. On multivariable Cox regression analysis, only right atrial
291
isomerism was independently associated with development of SVT (HR 3.41, 95% CI: 1.35-
292
8.64, p=0.009).
293
Pacemakers were implanted in 41 patients (14%). Five patients had a pacemaker in situ at the
294
time of Fontan, and the remaining 36 patients had a pacemaker inserted at the time of Fontan
295
(2 patients), during the Fontan hospital stay (5 patients) or during follow-up (29 patients).
296
Amongst hospital survivors discharged with an intact Fontan circulation, freedom from
297
pacemaker implantation at 10, 20 and 25 years were 89% (95% CI: 85-93%), 84% (95% CI:
298
79-89%) and 82% (95% CI: 76-88%) respectively
299 300
Anticoagulation and thromboembolic events
301
Of the 242 surviving patients with an intact LT Fontan at last follow-up, 203 patients (84%)
302
were on anticoagulation therapy or antiplatelet agents, consisting of warfarin in 108 patients
303
(45%), aspirin in 89 (37%), low-molecular-weight heparins in 2 (0.8%) and a combination of
304
agents in 4 (1.6%). No anticoagulation or antiplatelet agent was utilized in 29 patients (12%),
305
and the method of anticoagulation was unknown in 10 (4%).
306
Symptomatic thromboembolism occurred in 45 patients (16%) during follow-up. These
307
included 26 strokes, 9 pulmonary emboli, 7 transient ischemic attacks, 1 peripheral embolus,
308
1 renal embolus and 1 paradoxical embolic myocardial infarction. Thirteen of the patients
309
who suffered a symptomatic thromboembolic (13/45, 29%) event had a history of SVT. A
310
further 22 patients had a thrombus detected within the conduit or central venous circulation on
Page 14 of 28
15 311
routine echocardiography during follow-up. Freedom from symptomatic thromboembolism at
312
10, 20 and 25 years was 93% (95% CI: 89-96%), 80% (95% CI: 74-86%), and 80% (95% CI:
313
74-86%), respectively. On multivariable Cox regression analysis, pre-Fontan common
314
atrioventricular valve regurgitation ≥ moderate (HR 9.34, 3.80-22.93, p<0.001) and
315
development of SVT (HR 2.67, 95% CI: 1.06-6.71, p=0.04) were associated with
316
thromboembolism. Age < 3 years at the time of Fontan was associated with a reduced
317
likelihood of thromboembolism (HR 0.37, 95% CI: 0.15-0.88, p=0.02). Ten of the 45 patients
318
who experienced a symptomatic thromboembolic event during follow-up (22%) had a
319
documented history of tachyarrhythmia prior to the event.
320 321
Discussion
322
More than twenty-five years after its original description, the LT modification of the
323
originally-described Fontan procedure has demonstrated is superiority over the former AP
324
Fontan. The twenty-five-year survival of the patients discharged from hospital with an intact
325
LT circulation has reached a remarkable 85% within our population.
326
Estimates of late survival from the Mayo Clinic of 84%, 70% and 39% at 10, 20 and 30 years
327
after the LT Fontan, respectively, are less optimistic than our current series. They have
328
suggested that this may have been related to their large number of patients with heterotaxy
329
syndromes and the incorporation of larger amounts of native atrial tissue into surgical repair,
330
rendering their patients more susceptible to some of the complications observed after the
331
atriopulmonary connection[12]. Smaller patient numbers beyond 20-years in both instances
332
are also likely to contribute to the observed discrepancies in survival estimates. Our series is a
333
historical group of patients. One should be aware that, today it is likely that the lateral tunnel
334
procedure would be reserved to very small patients because it avoids the need for
335
implantation of bulky adult-size conduit. In many centers the lateral tunnel procedure would
Page 15 of 28
16 336
therefore be reserved to patients with more severe condition such as those with hypoplastic
337
left heart syndrome (HLHS) which were present in only very small numbers in our series. It is
338
therefore possible that contemporary results of the lateral tunnel would be worse than those
339
presented here.
340
Despite excellent survival rates, this population seems to be subjected to a slow but consistent
341
failure of their Fontan circulation. We have now demonstrated that none of the three main
342
types of Fontan circulation have experienced a sudden drop in their survival as was initially
343
expected, however, there still seems to be constant rate of attrition over time[13]. The exact
344
reasons for the failure of the Fontan circulation remain to be elucidated, and therefore the best
345
way to prevent the decline is still unclear [14-16]. One may be surprised to notice a
346
discrepancy between the large number of patients with a failed Fontan circulation and the
347
small number of those reaching heart transplantation. This phenomenon has been observed
348
worldwide, and is attributed to the complexity of undertaking transplantation in these patients,
349
the risk to undergo the procedure and inequality of access to transplantation[12, 17, 18]. The
350
lateral tunnel technique was primarily introduced in order to decrease the rate of arrhythmias
351
developing after Fontan completion, and it seems that it has achieved this goal [7]. We have
352
previously demonstrated that the incidence of arrhythmias was significantly decreased by the
353
introduction of the LT technique, which is again supported by our 20-year overall freedom
354
from arrhythmias of around 70% of patients [2].
355
Although ongoing failure and late onset of arrhythmia were not unexpected findings, we can
356
now demonstrate that that late failure is associated with the eventual occurrence of
357
arrhythmias. The onset of arrhythmias have been shown to predict mortality in other Fontan
358
populations around the world [19]. However, it must be acknowledged that the development
359
of arrhythmias may indeed be just one manifestation of a failing Fontan.
Page 16 of 28
17 360
We may face a similar phenomenon that was earlier observed in the patients with an
361
atriopulmonary connection, albeit at a later stage. In the LT Fontan, half of the venous circuit
362
draining the blood from the inferior vena cava to the pulmonary arteries is made of native
363
atrial tissue. With time, this atrial tissue may dilate and this stretch of the atrial wall, in
364
conjunction with the long suture lines, may form triggers for supra-ventricular
365
tachyarrhythmias. Atrial stretching may occur at a later stage than in patients with an
366
atriopulmonary connection as the wall stress is reduced by the elimination of turbulent flow
367
within the atrium. Serial imaging of the LT conduit may be useful to aid risk stratification,
368
however, this is yet to be fully investigated.
369
The incidence of thromboembolic events, and the number of massive pulmonary emboli,
370
seems to be increased compared to what we have observed at the same time-points in our
371
ECC cohort. The reasons for this are unclear, although the increased turbulence within a
372
dilated atrial channel may predispose to later thrombus formation [20]. This is supported by
373
the additional number of patients in the current study who were noted to have thrombus
374
present within the LT Fontan circuit.
375 376
Limitations
377
The incidence of arrhythmias, particularly those that were intermittent or transient in nature,
378
may be underestimated due to the limited availability of Holter monitoring or serial
379
electrocardiography data.
380
Due to the low number of patients with HLHS in this study, we were unable to draw any
381
significant association between HLHS morphology and an increased risk of adverse events,
382
failure or mortality.
Page 17 of 28
18 383
Another important limitation is the inability to accurately assess NYHA status retrospectively
384
from medical records. The numbers derived for late Fontan failure may therefore not reflect
385
the true burden of symptoms and non-cardiac morbidities.
386
Prolonged pleural effusions were defined in this study as lasting greater than 30 days, a
387
definition which differs from that of other institutions that have used greater than 14 days.
388
This is definition is based on how the data has been collected in the Registry.
389 390
Conclusion
391
Over a twenty-five-year period, the LT technique has achieved excellent late survival. As this
392
population ages, they face an increasing risk of failure and adverse events, including
393
arrhythmia and thromboembolism. We are likely to see a growing number of these survivors
394
requiring heart transplantation and late reintervention. Efforts must be made to develop
395
treatment strategies for the increasing number of Fontan patients that will continue to
396
experience attrition over time.
397 398
Acknowledgements
399
The authors thank the Murdoch Childrens Research Institute for infrastructure support. The
400
authors also acknowledge the Fontan Registry management and research assistants for their
401
invaluable support in the creation and maintenance of the Registry, as well as support in data
402
gathering for this paper, as well as Belinda Bortone for administrative support. The authors
403
acknowledge support provided to the Murdoch Childrens Research Institute by the Victorian
404
Government's Operational Infrastructure Support Program.
405
Page 18 of 28
19 406
Disclosures
407
Yves d’Udekem is a consultant for MSD and Actelion. Andrew Bullock reports consulting
408
fees from Actelion. All other authors have nothing to disclose with regard to commercial
409
support.
Page 19 of 28
20 410
References
411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457
[1] Iyengar AJ, Winlaw DS, Galati JC, Gentles TL, Weintraub RG, Justo RN, et al. The Australia and New Zealand Fontan Registry: description and initial results from the first population-based Fontan registry. Internal medicine journal. 2014;44:148-55. [2] d'Udekem Y, Iyengar AJ, Cochrane AD, Grigg LE, Ramsay JM, Wheaton GR, et al. The Fontan procedure: contemporary techniques have improved long-term outcomes. Circulation. 2007;116:I157-64. [3] Fontan F, Baudet E. Surgical repair of tricuspid atresia. Thorax. 1971;26:240-8. [4] Kreutzer J, Keane JF, Lock JE, Walsh EP, Jonas RA, Castaneda AR, et al. Conversion of modified Fontan procedure to lateral atrial tunnel cavopulmonary anastomosis. The journal of thoracic and cardiovascular surgery. 1996;111:1169-76. [5] Quinton E, Nightingale P, Hudsmith L, Thorne S, Marshall H, Clift P, et al. Prevalence of atrial tachyarrhythmia in adults after Fontan operation. Heart. 2015;101:1672-7. [6] Lardo AC, Webber SA, Friehs I, del Nido PJ, Cape EG. Fluid dynamic comparison of intra-atrial and extracardiac total cavopulmonary connections. The journal of thoracic and cardiovascular surgery. 1999;117:697-704. [7] de Leval MR, Kilner P, Gewillig M, Bull C. Total cavopulmonary connection: a logical alternative to atriopulmonary connection for complex Fontan operations. Experimental studies and early clinical experience. The journal of thoracic and cardiovascular surgery.1988;96(5):682-95. [8] Iyengar AJ, Winlaw DS, Galati JC, Wheaton GR, Gentles TL, Grigg LE, et al. The extracardiac conduit Fontan procedure in Australia and New Zealand: hypoplastic left heart syndrome predicts worse early and late outcomes. European journal of cardio-thoracic surgery : official journal of the European Association for Cardio-thoracic Surgery. 2014;46:465-73. [9] Stewart RD, Pasquali SK, Jacobs JP, Benjamin DK, Jaggers J, Cheng J, et al. Contemporary Fontan operation: association between early outcome and type of cavopulmonary connection. The Annals of thoracic surgery. 2012;93:1254-60. [10] d'Udekem Y, Iyengar AJ, Galati JC, Forsdick V, Weintraub RG, Wheaton GR, et al. Redefining expectations of long-term survival after the Fontan procedure: twenty-five years of follow-up from the entire population of Australia and New Zealand. Circulation. 2014;130:S32-S8. [11] Scrucca L, Santucci A, Aversa F. Competing risk analysis using R: an easy guide for clinicians. Bone Marrow Transplant. 2007;40:381-7. [12] Pundi KN, Johnson JN, Dearani JA, Pundi KN, Li Z, Hinck CA, et al. 40-Year FollowUp After the Fontan Operation: Long-Term Outcomes of 1,052 Patients. Journal of the American College of Cardiology. 2015;66:1700-10. [13] Schilling C, Dalziel K, Nunn R, Du Plessis K, Shi WY, Celermajer D, et al. The Fontan epidemic: Population projections from the Australia and New Zealand Fontan Registry. International journal of cardiology. 2016;219:14-9. [14] de Leval MR, Deanfield JE. Four decades of Fontan palliation. Nat Rev Cardiol. 2010;7:520-7. [15] Rychik J. The Relentless Effects of the Fontan Paradox. Semin Thorac Cardiovasc Surg Pediatr Card Surg Annu. 2016;19:37-43. [16] Ohuchi H. Adult patients with Fontan circulation: What we know and how to manage adults with Fontan circulation? J Cardiol (2016), http://dx.doi.org/10.1016/j.jjcc.2016.04.001. [17] Shi WY, Yong MS, McGiffin DC, Jain P, Ruygrok PN, Marasco SF, et al. Heart transplantation in Fontan patients across Australia and New Zealand. Heart. 2016;102:1120-6.
Page 20 of 28
21 458 459 460 461 462 463 464 465 466 467 468
[18] Iyengar AJ, Sharma VJ, Weintraub RG, Shipp A, Brizard CP, d'Udekem Y, et al. Surgical strategies to facilitate heart transplantation in children after failed univentricular palliations: the role of advanced intraoperative surgical preparation. European Journal of Cardio-Thoracic Surgery. 2014;46:480-5. [19] Giannakoulas G, Dimopoulos K, Yuksel S, Inuzuka R, Pijuan-Domenech A, Hussain W, Tay EL, et al. Atrial tachyarrhythmias late after Fontan operation are related to increase in mortality and hospitalization. International journal of cardiology. 2012;157:221-6. [20] Iyengar AJ, Winlaw DS, Grigg LE, Celermajer DS, d’Udekem Y. No difference between aspirin and warfarin after extracardiac Fontan in a propensity score analysis of 475 patients. European Journal of Cardio-Thoracic Surgery. 2016; doi:10.1093/ejcts/ezw159.
469 470
Page 21 of 28
22 471
Figure 1: Competing events cumulative incidence curves for all patients (n=301). Late
472
survival after the lateral tunnel Fontan is excellent with an overall survival of 82% at 20
473
years.
474
Figure 2: Freedom from Fontan failure for those surviving to hospital discharge with an
475
intact Fontan circulation (n=276). Late Fontan failure was defined as death, takedown,
476
transplantation, Fontan conversion/revision, NYHA class III/IV, protein-losing enteropathy or
477
plastic bronchitis.
478 479
Page 22 of 28
23 480
Table 1: Patient characteristics Characteristics Male, n (%)
Total ( n = 301) 165 (55%)
Anatomical comorbidities, n (%) Dextrocardia/Mesocardia
24 (8%)
Isomerism/Heterotaxy
21 (7%)
Non-cardiac anomaly
42 (14%)
Ventricular morphology, n (%) Left
188 (63%)
Right
81 (27%)
Biventricular/Indeterminate
32 (11%)
Primary morphological diagnosis, n (%) TA
81 (27%)
DILV
65 (22%)
DORV
50 (17%)
CAVC
28 (9%)
PA-IVS
19 (6%)
ccTGA
18 (6%)
TGA
12 (4%)
HLHS
9 (3%)
PA-VSD
6 (2%)
Other
13 (4%)
Pre-Fontan procedures Number of prior palliations, mean (SD) Prior aortic arch intervention, n (%)
2 (1) 39 (13%)
Page 23 of 28
24 Prior pulmonary artery banding, n (%)
90 (30%)
Prior staging with BCPS, n (%)
120 (40%)
Bilateral BCPS, n (%) Age at BCPS in years, median (IQR) Atrioventricular valve repair/replacement, n (%) Pulmonary artery reconstruction or angioplasty, n (%)
7 (2%) 1.4 (0.8-2.8) 7 (2%) 35 (12%)
Pre-Fontan haemodynamics Oxygen saturation (%), mean (SD)
81 (7)
Pulmonary artery pressure (mmHg), mean (SD)
12 (4)
Aortopulmonary or venovenous collaterals, n (%)
25 (8%)
Atrio-venous malformations, n (%)
7 (2%)
Atrioventricular valve regurgitation ≥ moderate, n (%)
12 (4%)
Fontan operative characteristics Age at Fontan in years, median (IQR)
3.8 (2.8 – 5.9)
Concomitant procedure, n (%)
77 (26%)
Concomitant pulmonary artery reconstruction, n (%)
20 (6%)
Concomitant atrioventricular valve repair/replacement, n (%)
8 (3%)
Fenestration, n (%)
156 (52%)
481
TA: tricuspid atresia; DILV: double-inlet left ventricle; DORV: double-outlet right ventricle;
482
CAVC: common atrioventricular canal; PA-IVS: pulmonary atresia with intact ventricular
483
septum; ccTGA: congenitally-corrected transposition of the great arteries; TGA: transposition
484
of the great arteries; HLHS: hypoplastic left heart syndrome; PA-VSD: pulmonary atresia
485
with ventricular septal defect; SD: standard deviation; BCPS: bidirectional cavopulmonary
486
shunt; IQR: interquartile range; mmHg: millimeters of mercury.
487
Page 24 of 28
25 488
Table 2: Results of multivariable Cox regression for late outcomes Predictors of late mortality Hazard
95% CI
95% CI
ratio
Lower
Upper
2.84
1.06
7.60
0.04
6.29
1.30
30.45
0.02
8.11
2.97
22.16
<0.001
Variable
p-value
Male Pre-Fontan common atrioventricular valve regurgitation ≥ moderate Prolonged pleural effusions
Predictors of late Fontan failure Hazard
95% CI
95% CI
Variable
p-value Ratio
Lower
Upper
Development of SVT#
4.68
2.07
10.58
<0.001
Age >7 (vs. 3-5) at Fontan
9.69
2.46
38.21
0.001
Prolonged pleural effusions
3.06
1.05
8.95
0.04
489
#
490
CI: confidence interval; SVT: supraventricular tachycardia.
Development of SVT was analyzed as a time-dependent covariate.
491 492
Page 25 of 28
26 493
Appendix 1. Multivariable predictors of late mortality prior to backward elimination. HR
95.0% CI for HR
p value
Lower
Upper
8.982
2.026
39.822
.004
Development of SVT
4.973
1.449
17.063
.011
Pre-Fontan atrioventricular
9.834
1.438
67.267
.020
Prior PA banding
4.019
1.091
14.806
.037
Male
2.698
.844
8.617
.094
Prior aortic arch intervention
.288
.042
1.970
0.20
Age 5-7 (vs. 3-5)
.543
.093
3.177
0.50
Prior BCPS
.613
.147
2.562
0.50
Fenestration at Fontan
.675
.212
2.151
0.51
LV dominance
1.240
.481
3.196
0.66
Age < 3 (vs. 3-5)
.742
.155
3.546
0.71
Age > 7 (vs. 3-5)
1.315
.292
5.912
0.72
Era 1990-1999 (vs. pre-1990)
1.143
.248
5.275
0.86
Era 2000-2014 (vs.pre-1990)
.000
n/a
n/a
0.99
HLHS
.000
n/a
n/a
0.99
Prolonged post-operative pleural effusions
valve regurgitation ≥ moderate
494 495
Page 26 of 28
27 496
Appendix 2. Multivariable predictors of late failure prior to backward elimination. HR
95.0% CI for HR
Sig.
Lower
Upper
Development of SVT
5.294
2.271
12.338
<0.0001
Prolonged effusions
3.510
1.131
10.890
.030
Age > 7 (vs. 3-5)
2.648
.867
8.089
.087
Pre Fontan AV regurg
2.507
.448
14.032
0.30
Fenestration
.629
.258
1.532
0.31
HLHS
3.524
.302
41.175
0.32
Age 5-7 (vs. 3-5)
1.618
.494
5.300
0.43
Male
1.360
.633
2.922
0.43
Era 2000-2015 (vs. pre 1990)
.379
.031
4.672
0.45
Era 1990-1999 (vs. pre 1990)
.725
.267
1.968
0.53
LV dominance
1.228
.627
2.403
0.55
Prior BCPS
1.339
.512
3.500
0.55
Age < 3 (vs. 3-5)
1.152
.358
3.703
0.81
Prior aortic arch
.882
.221
3.525
0.86
Prior PA banding
.999
.360
2.771
1.00
497
SVT: supraventricular tachycardia; PA: pulmonary artery; BCPS: bidirectional
498
cavopulmonary shunt; LV: left ventricle; HLHS: hypoplastic left heart syndrome.
499
Page 27 of 28
28 500
Page 28 of 28