Accepted Manuscript Chimney Technique with the INCRAFT® AAA Stent Graft System to Treat Pararenal Aortic Aneurysm in Narrowed Iliac Axes Nunzio Montelione, Pasqualino Sirignano, Wassim Mansour, Martina Formiconi, Laura Capoccia, Francesco Speziale PII:
S0890-5096(17)30022-5
DOI:
10.1016/j.avsg.2017.03.199
Reference:
AVSG 3359
To appear in:
Annals of Vascular Surgery
Received Date: 9 January 2017 Revised Date:
11 March 2017
Accepted Date: 26 March 2017
Please cite this article as: Montelione N, Sirignano P, Mansour W, Formiconi M, Capoccia L, Speziale F, Chimney Technique with the INCRAFT® AAA Stent Graft System to Treat Pararenal Aortic Aneurysm in Narrowed Iliac Axes, Annals of Vascular Surgery (2017), doi: 10.1016/j.avsg.2017.03.199. 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.
ACCEPTED MANUSCRIPT 1
CHIMNEY TECHNIQUE WITH THE INCRAFT® AAA STENT GRAFT SYSTEM
2
TO TREAT PARARENAL AORTIC ANEURYSM IN NARROWED ILIAC AXES
3 4
Nunzio Montelione *, Pasqualino Sirignano, Wassim Mansour,
5
Martina Formiconi, Laura Capoccia, and Francesco Speziale
RI PT
6 Vascular and Endovascular Surgery Unit
8
Departement of Surgery “Paride Stefanini”
9
Policlinico Umberto I
10
“Sapienza” University of Rome, Italy
SC
7
11
* Corresponding Author:
15
Nunzio Montelione, MD
16
Vascular and Endovascular Surgery Division
17
Department of Surgery “Paride Stefanini”
18
“Sapienza” University of Rome
19
Policlinico Umberto I
20
Viale del Policlinico, 155
21
00161 – Rome, Italy
22
Tel/Fax: +39 06 4940532
23
Mobile: +39 3479598785
24
email:
[email protected]
25 26 27 28 29 30 31 32 33 34
EP
TE D
14
AC C
13
M AN U
12
ACCEPTED MANUSCRIPT Abstract
36
To report a case of pararenal aortic aneurysm (PAAA) with narrowed iliac arteries treated by the
37
chimney technique combined with the INCRAFT® AAA stent graft system.
38
A 75-year-old man was diagnosed with a 55mm PAAA. Patient was judged unfit for open repair
39
and for standard endovascular repair (EVAR) due to its clinical and anatomical requirements.
40
Custom-made fenestrated device (FEVAR) was excluded due to the presence of narrowed and
41
heavily calcified iliac axes. Consequently, patient underwent auspicious endovascular aneurysm
42
repair with the INCRAFT® System and chimney grafts in the superior mesenteric artery and both
43
renal arteries. At 12 months’ follow-up, patient was in good clinical condition, with preserved renal
44
function. Computed tomographic angiography revealed inceptive shrinkage of the aneurysm with
45
patency of aortic and visceral grafts even in the presence of a low-flow type Ia endoleak.
46
Chimney grafts (CGs) combined with the low profile INCRAFT® System seem to be feasible for
47
treating PAAA, in high-risk patients unfit for standard EVAR or FEVAR devices.
48 49 50
M AN U
SC
RI PT
35
Keywords: Endovascular aneurysm repair - Chimney graft - Low profile endograft - Pararenal
52
aortic aneurysm - Parallel graft
55 56 57 58 59 60 61 62 63 64 65 66 67 68
EP
54
AC C
53
TE D
51
ACCEPTED MANUSCRIPT Introduction
70
Anatomical restrictions for endovascular aneurysm repair (EVAR) are characterized by
71
“challenging neck” (non-cylindrical, extremely short and highly angulated) and “difficult access”
72
(narrowed and severely angulated iliac arteries).1,2
73
Since the first bailout application,3,4 the chimney technique has become established as expanding
74
indication for EVAR in patients with challenging neck or pararenal aortic aneurysms (PAAA),5
75
yielding promising short- and mid-term results, even in elective settings.6,7
76
Distinct arrays of endografts are now routinely utilized, coupled with chimney grafts (CG),7 but
77
only a paucity of data is available about new low profile systems and CGs, for overcoming
78
difficulties related to iliac axes.8,9
79
We report a case of a patient with PAAA electively treated by chimney technique and a new ultra-
80
low profile device INCRAFT® AAA Stent Graft system.10
SC
M AN U
81
RI PT
69
Case Report
83
A 75-year-old man was referred to our center for an asymptomatic PAAA. Patient’s medical history
84
was positive for arterial hypertension, dyslipidemia, chronic obstructive pulmonary disease (FEV1
85
<50%), coronary artery disease (previous PTCA; left ventricular ejection fraction <40%), previous
86
urothelial cell carcinoma, and tobacco abuse. Pre-operative serum creatinine was 0.9 mg/dL with an
87
estimated glomerular filtration rate (eGFR ) of 85 mL/min/1.73 m2.
88
Because the patient’s comordibities precluded open surgical repair, an endovascular approach was
89
considered.
90
Pre-operative computed tomographic angiography (CTA) revealed a saccular PAAA (Fig. 1), with a
91
maximum diameter of 55 mm and a total volume of 89,1 cm3. Aortic diameter just above the renal
92
arteries was 21 mm, inter-renal aortic diameter was 28 mm, and aortic diameters 5, 10, 15, and 20
93
mm below the lowest renal artery were 38, 43, 46, and 49 mm, respectively. The entire aorta
94
presented a relatively straight anatomy with an α angle of 18°and β angle of 4°. In addition, highly
95
calcified and narrowed iliac axes (narrowest right external iliac artery diameter: 5 mm; left: 4.6
96
mm) were present (Fig. 2).
97
A custom-made fenestrated graft (FEVAR) was excluded due to anatomical restrictions: short
98
distance between the superior mesenteric and both renal arteries’ ostia, and the absence of adequate
99
access vessels.
AC C
EP
TE D
82
100
An off-label chimney procedure was then planned. At center lumen line reconstruction, the distance
101
between the superior mesenteric artery (SMA) and the lowest renal artery (RA) was found to be 3
102
mm, while the distance from the celiac trunk to the lowest RA was 21 mm. Consequently,
ACCEPTED MANUSCRIPT considering the diameter of the iliac arteries, CGs to the SMA and both RAs were contemplated,
104
combined with a low profile abdominal aortic device INCRAFT® AAA Stent Graft system (Cordis
105
Corp, Bridgewater, NJ-USA).
106
The procedure was performed under general anesthesia in an operating theater equipped with a
107
portable fluoroscopy unit (Euroamplin Alien, Eurocolumbus – Milan, Italy).
108
After systemic heparinization, percutaneous access was achieved via duplex ultrasound guidance to
109
both common femoral arteries. Preclose technique was used with one Perclose ProGlide® Suture-
110
Mediated Closure System (Abbot Vascular, Santa Clara, CA-USA) for each groin, after which,
111
surgical exposures of the right brachial and left axillary arteries were performed.
112
Two 7F × 90-cm Flexor® Shuttle® sheaths (Cook Medical, Inc.; Bloomington, IN-USA) were
113
inserted into the left axillary artery, and a 10F × 90-cm Flexor® Shuttle® sheath (Cook Medical,
114
Inc.) into the right brachial artery. All of those sheaths were advanced into the distal descending
115
thoracic aorta and renovisceral branch cannulation was attained, employing a 5-F Chuang visceral
116
reverse curve catheter (Cook Medical, Inc.), and a Rosen wire (Cook Medical, Inc.).
117
From the right femoral access, the INCRAFT® AAA Stent Graft system (Cordis Corp) was
118
advanced up to the level of the celiac trunk over 0.035-in SS Amplatz® wire (Boston Scientific
119
Corporation; Natick, MA-USA),
120
Diameter sizing of the aortic stent graft was 30 mm, based on the mean aortic neck diameter (21
121
mm) and the mean diameter of each of the CGs, according to Lachat’s experience.6
122
Once the main body was in place, all the shuttle sheaths were advanced toward their visceral target
123
vessels and the CGs were positioned and deployed with the distal end ~2 cm into the target vessels.
124
Viabahn® stent grafts (W.L. Gore & Associates, Inc.) were used in all cases: 7x50mm for the SMA
125
and 6x50mm for both RAs.
126
Bare-metal nitinol stents were employed to endoline the interior of the covered CGs (SMA:
127
Everflex 7x60mm; RAs: Everflex 6x60 mm; Medtronic, Santa Rosa, CA-USA) and then post-
128
dilated with non-compliant balloon catheters to attain secure anchoring. Once deflated, balloon
129
catheters were left in the distal landing zone of all the CGs. During the main body stent graft
130
deployment, balloons were reinflated and tensioned to hold the CGs straight and parallel to the
131
aortic wall avoiding migration, as previously reported.6,11
132
The procedure was finalized, deploying two iliac limb extensions from the femoral access (right
133
side: 10x12 mm; left side: 10x12 mm; Cordis Corp).
134
Ballooning of the docking zone and kissing balloon technique were performed in order to achieve
135
simultaneous, complete expansion of the aortic graft and CGs, especially in the overlapping areas.
AC C
EP
TE D
M AN U
SC
RI PT
103
ACCEPTED MANUSCRIPT In the completion angiography (Fig. 3), all grafts and visceral branches were patent, the aneurysm
137
was satisfactorily excluded with the presence of a delayed low flow type Ia endoleak. New kissing
138
balloon dilatations of grafts were performed as prompt corrective treatment with a significant
139
diminution of the type Ia endoleak.
140
Procedure time was 281 minutes, radiation dose area product was 62543 mGy*cm2 and total
141
contrast amount was 155 mL.
142
Intensive care unit stay was two days; postoperative course was uneventful and patient was
143
discharged on the sixth postoperative day with double antiplatelet therapy (acetylsalicylic acid 100
144
mg/d and clopidogrel 75 mg/d) for three months. Mono-antiplatelet therapy was continued
145
indefinitely.
146
At 12 months’ follow-up patient was in good clinical condition, with preserved renal function
147
(eGFR 82mL/min/1.73m2). Scheduled CTA (Fig. 4) showed initial shrinkage of the aneurysm
148
(maximum diameter 50 mm, volume 82,6 cm3), with patency of aortic and visceral grafts even with
149
a low -flow type Ia endoleak.
150
Patient gave consent for treatment and for publication of his clinical notes.
M AN U
SC
RI PT
136
151 Discussion
153
Over the years, use of CGs has been reported to treat ruptured abdominal aortic aneurysms,
154
expanding indication for EVAR even in the absence of a suitable proximal landing zone.5
155
Nowadays, recent data confirm the efficacy of the procedure even in elective treatment of PAAAs
156
and thoracoabdominal aortic aneurysms.6,7
157
We reported the case of a patient considered unfit for open surgery in which standard EVAR was
158
excluded for lack of infrarenal neck. Custom-made FEVAR devices were not considered an option
159
because positioning of the graft requires the capability to advance and rotate a large delivery system
160
(to correctly align the fenestrations with the visceral vessel orifices), which could be impaired by
161
narrowed access vessels.12,13
162
Unfavorable iliac anatomy is a well-known contraindication for endovascular aortic aneurysm
163
repair. In 2000, the EUROSTAR registry reported iliac access problems in 13% of patients selected
164
for EVAR due to iliac tortuosity, occlusive disease, or small-caliber.14 Even if the available
165
technology has improved remarkably over the years,15 EVAR, and much more F-EVAR or B-
166
EVAR, still represent challenging procedures in patients with small, calcified, or tortuous iliac
167
vessels.16 Indeed, adequate iliac access for 18F to 20F delivery catheters is 6–7 mm and, according
168
to these diameters, about 50% of patients are excluded from EVAR treatment.17 Moreover, insertion
169
of large delivery catheters into small iliac vessels potentially increases the peril of iliac artery
AC C
EP
TE D
152
ACCEPTED MANUSCRIPT rupture with life-threatening bleeding. Some feasible solutions include highly flexible hydrophilic
171
devices, iliac conduits, ‘‘cracking and paving,’’ and alternative access sites.18
172
CGs in association with conventional endografts presenting an outer diameter ranging from 16 to
173
20F were deemed unsuitable to treat the patient. Despite the fact that some authors have described
174
the chimney technique combined with a polymer-based endograft, at that time, we had not
175
performed enough of these procedures to have sufficient experience to use those grafts in
176
combination with the chimney technique.9
177
Considering all the aforementioned reasons, an off-label chimney technique was performed by
178
placing CGs in the SMA and RAs, coupled with the low profile INCRAFT® device, to overcome
179
technical limitations due to unfavorable iliac anatomy.
180
The INCRAFT® Stent Graft System was designed to address some of the limitations of current
181
devices. To increase the applicability of EVAR to a broader range of patients such as those with
182
small or diseased access vessels, design enhancements allowed the device to be constrained within
183
an ultra-low profile delivery system with a true 14F (4.7mm) outer diameter for main-body, and
184
12.5F for iliac limbs.19
185
The election of such a novel concept and low profile device allowed for successful treatment of
186
the patient, even in the presence of a small type Ia endoleak in the completion angiography and
187
follow-up images.
188
Postoperative endoleaks after the adoption of CGs are a frequent event, occurring in as many as
189
23% of cases and represent a potential cause of aneurysmal rupture.7,20-22 However, if endoleaks are
190
addressed judiciously, by close clinical and radiological follow-up, related complications could be
191
tempered by the lack of perseverence or clinical adverse sequelae .7,20
192
The majority of the endoleaks were primarily due to “gutter ” occurrences, between the CGs and
193
aortic endografts, the presence and size of which remain the main concern that compromises (the
194
early or late) outcomes of this technique. Persistent or new-onset postoperative type Ia endoleaks
195
seem to be linked to numbers of CGs implanted and length of the new landing zones.7,20 Indeed,
196
data from the PERICLES registry showed a greater incidence of Type Ia endoleaks when more
197
than three CGs were required.7 Incidence of endoleaks is reported to be rare provided an adequate
198
proximal landing zone has been created (almost 20mm).7,20
199
In the present case, it was necessary to implant three CG’s in order to achieve a satisfactory, new
200
proximal landing zone (21mm). Considering previously reported data, by performing that
201
procedure, we averted a clinically relevant, high-flow endoleak (with a new proximal landing zone
202
of 21 mm created), but we were exposed to a higher risk of gutter occurrences (with three CGs
AC C
EP
TE D
M AN U
SC
RI PT
170
ACCEPTED MANUSCRIPT performed). Furthermore, the necessity to use three CGs required two different brachial/axillary
204
accesses to advance the guiding sheaths in order to minimize potential vessels damages11,23.
205
The best combinations of aortic and branch stent grafts to potentially minimize gutter-related
206
concerns are another important aspect for consideration. Even though both experimental24,25 and
207
clinical studies6,7,20 have been performed in an endeavor to resolve this issue, no definitive data, in
208
particular, using the new concept INCRAFT® device, are currently available. Self-expandable
209
chimney grafts were preferred because profile and flexibility. Particularly Viabahn seems to
210
represent a durable treatment for renovisceral arteries 20.
211
Despite the fact that several techniques have been reported to correct type IA endoleaks after
212
chimney graft procedures (coil embolization or “neck elongation ”),7,20 some experience seems to
213
suggest that not all postoperative endoleaks should be invasively managed. In a report by Donas
214
and co-workers, out of five perioperative type Ia endoleak, only one was treated during follow-up
215
due to continuous aneurismal sac increase23. Indeed, delayed low-flow type IA endoleaks are
216
reported as not propagating either aneurysmal enlargement or clinical adverse events in all
217
published experiences.6,11,20 For these reasons, the patient is being strictly observed in the absence
218
of further treatment.
M AN U
SC
RI PT
203
219 Conclusion
221
In the present case, the use of CGs combined with the ultra-low profile INCRAFT® System has
222
proven to be a feasible, safe, and viable approach in the treatment of PAAA.
223
In our opinion, this solution represents a viable option for treating high-risk patients with a
224
challenging iliac anatomy, unsuitable for fenestrated devices. Lengthier follow-up and greater
225
numbers of cases are required.
227 228 229 230 231 232 233 234 235 236
EP
AC C
226
TE D
220
ACCEPTED MANUSCRIPT References
238
[1] de Vries JP. The proximal neck: the remaining barrier to a complete EVAR world. Semin Vasc
239
Surg 2012;25:182-186.
240
[2] Moise MA, Woo EY, Velazquez OC, Fairman RM, Golden MA, Mitchell ME, Carpenter JP.
241
Barriers to endovascular aortic aneurysm repair: past experience and implications for device
242
development. Vasc Endovasc Surg 2006;40:197-203.
243
[3] Greenberg RK, Clair D, Srivastava S, Bhandari G, Turc A, Hampton J, Popa M, Green R, Ouriel
244
K. Should patients with challenging anatomy be offered endovascular aneurysm repair? J Vasc Surg
245
2003;38:990-996.
246
[4] Larzon T, Eliasson K, Gruber G. Top-fenestrating technique in stent-grafting of aortic diseases
247
with mid-term follow-up. J Cardiovasc Surg (Torino) 2008;49:317-322.
248
[5] Pecoraro F, Pfammatter T, Mayer D, Frauenfelder T, Papadimitriou D, Hechelhammer L, Veith
249
FJ, Lachat M, Rancic Z. Multiple periscope and chimney grafts to treat ruptured thoracoabdominal
250
and pararenal aortic aneurysms. J Endovasc Ther 2011;18:642-649.
251
[6] Lachat M, Veith FJ, Pfammatter T, Glenck M, Bettex D, Mayer D, Rancic Z, Gloekler S,
252
Pecoraro F. Chimney and periscope grafts observed over 2 years after their use to revascularize 169
253
renovisceral branches in 77 patients with complex aortic aneurysms. J Endovasc Ther 2013;20:597-
254
605.
255
[7] Donas KP, Lee JT, Lachat M, Torsello G, Veith FJ. Collected world experience about the
256
performance of the snorkel/chimney endovascular technique in the treatment of complex aortic
257
pathologies: the PERICLES registry. Ann Surg 2015;262:546-53.
258
[8] Massmann A, Serracino-Inglott F, Buecker A. Endovascular aortic repair with chimney
259
technique using ultra low-profile Ovation stent-graft for juxtarenal aneurysms having small iliac
260
vessels. Cardiovasc Intervent Radiol 2014;37:488-92.
261
[9] Malkawi AH, de Bruin JL, Loftus IM, Thompson MM. Treatment of a juxtarenal aneurysm with
262
the Nellix endovascular aneurysm sealing system and chimney stent. J Endovasc Ther 2014;21:538-
263
40.
264
[10]. Torsello G, Scheinert D, Brunkwall, Chiesa R, Coppi G, Pratesi C. Safety and effectiveness of
265
the INCRAFT AAA Stent Graft for endovascular repair of abdominal aortic aneurysms. J Vasc
266
Surg 2015;61:1-8.
267
[10] Greenberg RK. Aortic aneurysm, thoracoabdominal aneurysm, juxtarenal aneurysm,
268
fenestrated endografts, branched endografts, and endovascular aneurysm repair. Ann N Y Acad Sci
269
2006;1085:187-196.
AC C
EP
TE D
M AN U
SC
RI PT
237
ACCEPTED MANUSCRIPT [11] Montelione N, Pecoraro F, Puippe G, Chaykovska L, Rancic Z, Pfammatter T, Mayer D,
271
Amann-Vesti B, Husmann MJ, Veith FJ, Mangialardi N, Lachat M. A 12-year experience with
272
chimney and periscope grafts for treatment of type I endoleaks. J Endovasc Ther 2015;22:568-74.
273
[12] Verhoeven EL, Vourliotakis G, Bos WT, Tielliu IF, Zeebregts CJ, Prins TR, Bracale UM, van
274
den Dungen JJAM. Fenestrated stent grafting for short-necked and juxtarenal abdominal aortic
275
aneurysm: an 8-year single-centre experience. Eur J Vasc Endovasc Surg 2010;39:529-536.
276
[13] Bakoyiannis CN, Economopoulos KP, Georgopoulos S, Klonaris C, Shialarou M, Kafeza M,
277
Papalambros E. Fenestrated and branched endografts for the treatment of thoracoabdominal aortic
278
aneurysms: a systematic review. J Endovasc Ther 2010;17:201-209.
279
[14] Cuypers PW, Laheij RJ, Buth J. Which factors increase the risk of conversion to open surgery
280
following endovascular abdominal aortic aneurysm repair? The EUROSTAR collaborators. Eur J
281
Vasc Endovasc Surg 2000;20:183-189.
282
[15] Sirignano P, Speziale F, Capoccia L, Menna D, Mansour W, Montelione N, Setacci F,
283
Galzerano G, Setacci C. Iliac and femoro-popliteal arteries morphological CTA features as
284
determinants of outcome after standard EVAR procedures. J Cardiovasc Surg (Torino) 2016 Sep
285
16. In press
286
[16] Moise MA, Woo EY, Velazquez OC, Fairman RM, Golden MA, Mitchell ME, Carpenter JP.
287
Barriers to endovascular aortic aneurysm repair: past experience and implications for future device
288
development. Vasc Endovascular Surg 2006;40:197-203.
289
[17] Kristmundsson T, Sonesson B, Dias N, Malina M, Resch T. Anatomic suitability for
290
endovascular repair of abdominal aortic aneurysms and possible benefits of low profile delivery
291
systems. Vascular 2014;22:112-5.
292
[18] Hinchliffe RJ, Ivancev K, Sonesson B, Malina M. ‘‘Paving and cracking’’: an endovascular
293
technique to facilitate the introduction of aortic stent-grafts through stenosed iliac arteries. J
294
Endovasc Ther 2007;14:630-633.
295
[19] Coppi G, Njila M, Coppi G, Saitta G, Silingardi R, Pratesi C, Chiesa R, Scheinert D,
296
Brunkwall JS, Torsello G. INCRAFT Stent-Graft System: one-year outcome of the INNOVATION
297
Trial. J Cardiovasc Surg (Torino) 2014;55:51-9.
298
[20] Pecoraro F, Veith FJ, Puippe G, Amman-Vesti B, Bettex D, Rancic Z, Pfammatter T, Lachat
299
M. Mid- and longer-term follow-up of chimney and/or periscope grafts and risk factors for failure.
300
Eur J Vasc Endovasc Surg 2016;51:664-73.
301
[21] Chaudhuri A. Periscopes, snorkel and chimneys: no smoke without fire? Eur J Vasc Endovasc
302
Surg 2014;47:218-20.
AC C
EP
TE D
M AN U
SC
RI PT
270
ACCEPTED MANUSCRIPT [22] Schiro A, Antoniou GA, Ormesher D, Pichel AC, Farguharson F, Serracino-Inglott F. The
304
chimney technique in endovascular aortic aneurysm repair: late ruptures after successful single
305
renal chimney stent grafts. Ann Vasc Surg 2013;27:835-43.
306
[23] Donas KP, Pecoraro F, Torsello G, Lachat M, Austermann M, Mayer D, Panuccio G, Rancic
307
Z. Use of covered chimney stents for pararenal aortic pathologies is safe and feasible with excellent
308
patency and low incidence of endoleaks. J Vasc Surg. 2012;55:659-665.
309
[24] de Bruin JL, Yeung KK, Niepoth WW, Lely RJ, Cheung Q, de Vries A, Blankensteijn JD.
310
Geometric study of various chimney graft configurations in an in vitro juxtarenal aneurysm model.
311
J Endovasc Ther 2013;20:184-90.
312
[25] Mestres G, Uribe JP, Garcìa-Madrid C, Miret E, Alomar X, Burrel M, Riambau V. The best
313
conditions for parallel stenting during EVAR: an in vitro study. Eur J Vasc Endovasc Surg
314
2012;44:468-73.
315
M AN U
SC
RI PT
303
Figure Legend
317
Figure 1. - Multiplanar reconstruction of preoperative computed tomographic angiography (CTA)
318
showing pararenal aortic aneurysm;
319
Figure 2. - Three-dimensional reconstruction of preoperative CTA underlying the presence of
320
narrowed and heavy calcified iliac arteries;
321
Figure 3. - Completion angiography showing patency of all grafts and visceral branches,
322
satisfactory exclusion of the aneurysm with the presence of a low flow type Ia endoleak (white
323
arrows);
324
Figure 4. - Three-dimensional reconstruction of 12 months’ CTA showing the good aneurysm
325
exclusion.
AC C
EP
TE D
316
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT