Journal Pre-proof

First Trimester Cesarean Scar Pregnancy: A Comparative Analysis of Treatment

Options from the International Registry

Andrea KAELIN AGTEN, MD, Davor JURKOVIC, MD, Ilan TIMOR-TRITSCH, MD,
Nia JONES, MD, Susanne JOHNSON, MD, Ana MONTEAGUDO, MD, Judith
HUIRNE, MD, Jonah FLEISHER, MD, MPH, Ron MAYMON, MD, Tania HERRERA,
MD, Federico PREFUMO, MD, Stephen CONTAG, MD, Marcos CORDOBA, MD,
Gwendolin MANEGOLD-BRAUER, MD, on behalf of the CSP COLLABORATIVE
NETWORK

PII: S0002-9378(23)00758-5
DOI: https://doi.org/10.1016/j.ajog.2023.10.028
Reference: YMOB 15312

To appear in: American Journal of Obstetrics and Gynecology

Received Date: 26 July 2023

Revised Date: 8 October 2023
Accepted Date: 9 October 2023

Please cite this article as: KAELIN AGTEN A, JURKOVIC D, TIMOR-TRITSCH I, JONES N, JOHNSON
S, MONTEAGUDO A, HUIRNE J, FLEISHER J, MAYMON R, HERRERA T, PREFUMO F, CONTAG
S, CORDOBA M, MANEGOLD-BRAUER G, on behalf of the CSP COLLABORATIVE NETWORK,
First Trimester Cesarean Scar Pregnancy: A Comparative Analysis of Treatment Options from
the International Registry, American Journal of Obstetrics and Gynecology (2023), doi: https://
doi.org/10.1016/j.ajog.2023.10.028.

This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition
of a cover page and metadata, and formatting for readability, but it is not yet the definitive version of
record. This version will undergo additional copyediting, typesetting and review before it is published
in its final form, but we are providing this version to give early visibility of the article. Please note that,
during the production process, errors may be discovered which could affect the content, and all legal
disclaimers that apply to the journal pertain.

© 2023 The Author(s). Published by Elsevier Inc.

 1

1 First Trimester Cesarean Scar Pregnancy: A Comparative Analysis of

2 Treatment Options from the International Registry

4 Andrea KAELIN AGTEN, MD1; Davor JURKOVIC, MD2; Ilan TIMOR-TRITSCH, MD3;

5 Nia JONES, MD4; Susanne JOHNSON MD5; Ana MONTEAGUDO, MD6; Judith

6 HUIRNE, MD7; Jonah FLEISHER MD, MPH8; Ron MAYMON, MD9; Tania

7 HERRERA, MD10; Federico PREFUMO, MD11; Stephen CONTAG, MD12; Marcos

8 CORDOBA MD13; Gwendolin MANEGOLD-BRAUER, MD14 on behalf of the CSP

f
oo
9 COLLABORATIVE NETWORK

r
10

11
-p
The authors report no conflict of interest.
re
12
lP

13 The study was supported with unrestricted grants from the “Bangerter-Rhyner
na

14 Stiftung” and the “Freiwillige Akademische Gesellschaft Basel”. The funding

15 institutions did not have any role in conduction of the study or manuscript
ur

16 preparation.
Jo

17

18 Affiliations

19 1
Liverpool Women’s Hospital NHS Foundation Trust, United Kingdom
2
20 University College London, United Kingdom
3
21 Hackensack Meridian School of Medicine, New Jersey, USA
4
22 University of Nottingham, United Kingdom
5
23 Princess Anne Hospital, University Hospitals Southampton, United Kingdom
6
24 Icahn School of Medicine Mount Sinai, New York, USA
7
25 Amsterdam UMC, Netherlands
 2

8
26 University of Illinois at Chicago, USA
9
27 Shamir Medical Center (Assaf Harofeh), Israel
10
28 Pacifica Salud Hospital, Panama
11
29 Department of Clinical and Experimental Sciences, University of Brescia, Brescia,

30 Italy
12
31 University of Minnesota, Division of Maternal Fetal Medicine, Minneapolis, USA
13
32 Corewell Health West, Grand Rapids, Michigan, USA
14
33 University of Basel, Switzerland

f
oo
34

r
35 Corresponding author -p
re
36 Dr Andrea Kaelin Agten, Liverpool Women’s Hospital NHS Foundation Trust, Fetal
lP

37 Medicine Unit, Crown Street, Liverpool L8 7SS, United Kingdom

na

38 Email: a.kaelin@gmail.com, Cell phone: +447491321102

ur

39 Word count abstract: 420

40 Word count main document: 4289

Jo
 3

41 CSP collaborative network

42 AGOSTINI, Aubert15, AJJAWI, Sajida4, ARDABILI, Sara16, BARTELS, Helena16,

43 BOHILTEA, Roxana17, BRITTAIN, Gayle18, BUONOMO, Francesca19, BURN,

44 Sabrina12, BRUNNSCHWEILER Elena14, CHANTRAINE, Frédéric20, CHIPETA,

45 Hlupekile21, COUTINHO, Conrado Milani22, DE ALMEIDA FIORILLO, Clarice19, DE

46 BRAUD, Lucrezia Viola2, DEBRAS, Elodie23, DOGRA MARWAHA, Poojan24,

47 EDWARDS, Philipa4, EL HAIEG, Dalia25, ELNAMOURY, Mohamed26, EREME,

f
48 Keemi8, FARRÀS, Alba27, FERNANDEZ, Herve23, FRATELLI, Nicola11, GAL-

oo
49 KOCHAV, Maayan9, GEORG, Alexia Viegas28, GUANDALINI Fabiola11, GUTAJ,

r
50 Paweł29, HELMY-BADER, Samir 30
-p
, HIGUERAS, Teresa 27
, HODEL, Marcus16,

51 JOHNS, Jemma31, KAMEL, Rasha32, Laure NOEL20, MIQUEL, Laura15, NEGM,

re
52 Sherif32, NIETO-CALVACHE, Albaro33, PARACHA, Ayesha30, PATEISKY, Petra30,
lP

53 ROBERTSON, Louise1, ROSS, Jackie31, SADEK, Somayya25, SCHOETZAU Andy14,

na

54 SHARMA, Mona34, VERBERKT, Carry7, WENDER-OZEGOWSKA, Ewa29

ur

55
Jo

56 Affiliations of the CSP collaborative network:

15
57 Hôpitaux de Marseille, la Conception Hospital, Assistance Publique, Marseille,

58 France
16
59 Cantonal Hospital of Lucerne, Department of Gynecology and Obstetrics, Lucerne,

60 Switzerland
17
61 University Emergency Hospital Bucharest, Bucharest, Romania

62 18
St George’s University London, London, United Kingdom
19
63 Institute for Maternal and Child Health, IRCCS Burlo Garofolo, Trieste, Italy
20
64 Centre Hospitalier Universitaire (CHU) de Liège, Hopital Citadelle, Liège, Belgium
 4

21
65 Leeds Teaching Hospitals NHS Trust, Gynecology Acute Treatment Unit, Leeds

66 United Kingdom
22
67 Hospital das Clínicas da Faculdade de Medicina de Ribeirão Preto da

68 Universidade de São Paulo, Brasil

23
69 Hopital Bicêtre, Service Gynécologie Obstétrique, Le Kremlin-Bicêtre, France
24
70 Shri Lal Bahadur Shastri Government Medical College, Nerchowk Mandi, India
25
71 Zagazig University, Obstetrics and Gynecology department, Ash Sharqia

72 Governorate, Egypt

f
oo
26
73 Tanta University, Fetal Medicine Unit, Gharbia Governorate, Egypt

r
74 27
Vall d’Hebron Hospital Universitari, Barcelona, Spain

75 28
-p
Hospital das Clínicas da Faculdade de Medicina de Ribeirão Preto da
re
76 Universidade de São Paulo, Brasil
lP

77 29
Poznan University of Medical Sciences, Department of Reproduction, Poznań,
na

78 Poland
30
79 Medical University of Vienna Department of Obstetrics and Gynaecology, Division
ur

80 of Obstetrics and feto-maternal Medicine, Vienna, Austria

Jo

81 31
King’s College Hospital NHS Foundation Trust, London, United Kingdom
32
82 Cairo University, Maternal-Fetal Medicine Unit, Giza, Egypt
33
83 Clinica de espectro de acretismo placentario, Fundación Valle de Lili Cali, Valle del

84 Cauca, Colombia
34
85 Barking, Havering and Redbridge Hospitals NHS, London
 5

86 Condensation page

87 Tweetable statement

88 Cesarean Scar pregnancy in the first trimester can effectively be managed with

89 suction evacuation, surgical excision, and balloon treatment. The single use of

90 methotrexate is not recommended due to a higher risk of failing and complications.

91 Short title

92 CSP-registry: First trimester management

f
oo
93

r
94 AJOG at a Glance: -p
re
95 A. Why was this study conducted?
lP

96 • To study the safety and efficacy of different approaches to first trimester

na

97 management of Cesarean scar pregnancies

ur

98 B. What are the key findings?

Jo

99 • Suction evacuation, surgical excision and balloon treatment are effective

100 (>90%) in the treatment of Cesarean Scar pregnancies.

101 • The single first line use of methotrexate has a higher risk of failing and

102 complications compared to other methods and should not be recommended.

103 C. What does this study add to what is already known?

104 • Largest published set of original data on 460 Cesarean Scar pregnancy cases

105 comparing different methods from over 30 high and low volume centers

106 worldwide.

107
 6

108 Abstract

109 Background:

110 Cesarean scar pregnancy (CSP) is an iatrogenic consequence of a prior cesarean

111 delivery (CD). The gestational sac implants into a niche created by the incision of the

112 previous CD, carrying a substantial risk for major maternal complications. The aim of

113 this study is to report, analyze and compare effectiveness and safety of different

114 treatments options of CSP managed in the first trimester through a registry.

f
oo
115 Objectives:

r
116 To study the ultrasound findings, disease behavior, and management of first trimester

117 CSP.
-p
re
lP

118 Study design:

119 We created an international registry of CSP to study the ultrasound findings, disease
na

120 behavior, and management of CSP. The CSP-registry collects anonymized ultrasound
ur

121 and clinical data of individual CSP patients on a secure digital information platform.
Jo

122 Cases were uploaded by 31 participating centers in 19 countries. In this study we only

123 included live and failing CSP (with or without positive fetal heart beat) who received

124 active treatment (medical or surgical) before 12+6 weeks’ gestation in order to

125 evaluate effectiveness and safety of the different management options. Patients

126 managed expectantly were not included in this study and will be reported separately.

127 Treatment was classified as successful if it led to a complete resolution of pregnancy

128 without the need for any additional medical intervention.

129 Results:
 7

130 Between 29 August 2018 and 28 February 2023, we recorded 460 patients with CSP

131 (281 live, 179 failing CSP) who fulfilled the inclusion/exclusion and were registered.

132 270/460 (58.7%) patients were managed surgically, 123/460 (26.7%) underwent

133 medical management, 46/460 (10%) underwent balloon management and 21/460

134 (4.6%) underwent other, less frequently used treatment options. Suction evacuation

135 was very effective, with 202/221(91.5%, 95%CI 87.8-95.2) success rate whilst

136 systemic methotrexate was least effective with only 38/64 (59.4%, 95% CI 48.4-70.4)

137 patients not requiring additional treatment. Overall, surgical treatment of CSP was

f
oo
138 successful in 236/258 (91.5%, 95% CI 88.4-94.5) and the complication rate was

r
139 24/258 (9.3%, 95% CI 6.6-11.9).
-p
140 Conclusion:
re
lP

141 CSP in the first trimester of pregnancy can be managed effectively in more than 90%

142 of cases with either suction evacuation, balloon treatment or surgical excision. The
na

143 effectiveness of all treatment options decreases with advancing gestational age CSP
ur

144 should be treated as early as possible after confirmation of the diagnosis. Local
Jo

145 medical treatment with KCl or methotrexate is less efficient in and has higher rates of

146 complications compared to the other treatment options. Systemic methotrexate has a

147 substantial risk of failing and a higher complication rate and should not be

148 recommended as first line treatment.

149

150 Key words: management, balloon treatment, suction evacuation, surgical excision,

151 methotrexate, complications, KCL, hysterectomy, hemorrhage

 8

152 Introduction

153 Cesarean scar pregnancies (CSP) can develop as consequence of a prior cesarean

154 delivery (CD) with implantation of the placenta in the niche created by the incision of

155 the previous CD1. They represent a rising clinical concern stemming from an

156 increasing rate of CDs2. In recent years, nearly one third of babies were delivered by

157 CD in developed countries3. The true incidence of CSP in not known but estimates

158 range between 1 in 1800-1 in 2656 pregnancies4.

159 Due to placental implantation in the uterine defect with absent decidua and partial loss

f
oo
160 of the myometrium and distal uterine vascular network, CSPs are strongly associated

r
161 with serious complications. Among these are severe hemorrhage, preterm labor,

162
-p
uterine rupture and implantation disorders referred to as placenta accreta spectrum.
re
5,6
163 .Major hemorrhage, which can occur in all three trimester scan lead to loss of fertility,
lP

164 hysterectomy, and even death7,8.

na

165 Further serious complications include uterine rupture, second trimester morbidly

166 adherent placenta, uterine rupture, severe hemorrhage, and preterm labor.
ur

167 A systematic review showed complications rates as high as 44%, mainly due to missed
Jo

168 diagnoses or potentially inappropriate treatments that increase risk of massive

169 hemorrhage2. CSPs can be best diagnosed by ultrasound between 5 to 9 weeks’

170 gestation9. Diagnosis is primarily based on transvaginal ultrasound 9,10. Depending on

171 the gestational age, some ultrasound signs are easier to detect than others, which

172 poses an ongoing diagnostic dilemma11. Due to the rarity of CSP, patients are

173 counseled based upon case series containing diverse procedures and techniques12.

174 There is no national and/or international agreement upon management strategies and

175 there is a huge variety of different approaches for pregnancy interruption12,13. Although

176 publications exist on the evaluation of different management options to treat CSP,
 9

177 individual practitioners and institutions persist in using very different methods at their

178 discretion. In short, treatments are mostly based upon individual clinicians’ experience,

179 expertise, and the resources available.

180 There is little evidence regarding effectiveness and safety of various treatment

181 methods. A recent systematic review, based on multiple small single center case

182 series, showed that there were more than 30 methods of treating a CSP2. To study

183 the diagnosis, the natural history and the management of CSPs in more detail, a group

184 of experts set up the international registry for CSP (www.csp-registry.com)12. The aim

f
oo
185 of this study was to investigate safety and effectiveness of different treatment options

r
186 in the management of first trimester CSP up to 12+6 weeks gestation, based upon

187 data from the International CSP registry.

-p
re
188
lP

189 Material & Methods

na

190 The CSP registry is an international research platform for data on CSP formed by an

191 international network of collaborators referred to as the “CSP collaborative network”.

ur
Jo

192 The project was approved by the NHS Health Research Authority (HRA) and Health

193 and Care Research Wales (HCRW), IRAS project ID: 246295. Ethical approval was

194 further sought for each participating center according to local regulations. The registry

195 is funded with unrestricted grants from the Freiwillige Akademische Gesellschaft Basel

196 (FAG), Basel, Switzerland and the Bangerter-Rhyner Stiftung”, Basel, Switzerland”.

197

198 The CSP registry was set up in 2018 to study the diagnosis, natural history, and

199 management of CSP pregnancies. The registry collects structured, retrospective,

200 individual, coded data from CSP patients including demographic data, personal

201 medical history, and previous pregnancies. Data on ultrasound findings, management
 10

202 and outcome are also recorded. An anonymized panoramic ultrasound image showing

203 the pregnancy, uterus and endocervical canal in longitudinal section is uploaded for

204 each case and reviewed by the board. Only coded, non-identifiable data is recorded

205 and stored in a secure data base (Adjumed, Zurich). A sample of the collection form

206 is provided in the supplemental material.

207 Nine public hospitals, 21 academic institutions and 2 private clinics contributed to the

208 registry until data closure. For this analysis we had 31 clinics contributing of which

f
209 there were 21 small volume clinics contributing less than 20 cases each (median 9

oo
210 cases, range 1-19) and 10 large volume clinics contributing 20 or more cases each

r
211 -p
(median 37 cases, range 23-95 cases). The small volume clinics contributed in total

212 33.1% and the large volume clinics contributed 66.9% of the cases that were analysed
re
213 for the purpose of this study.
lP

214 In all centers, the examiners declared that they had high level of experience and
na

215 expertise in gynecological ultrasound and additionally went through an onboarding

ur

216 process where the important features of CSP were discussed. A transvaginal
Jo

217 ultrasound scan was performed to examine the location of the pregnancy and the

218 presence of embryonic/fetal cardiac activity. We accepted the commonly used

219 diagnostic criteria of CSP based on current literature, including a gestational sac

220 located anteriorly at the level of the internal os covering the visible or presumed site

221 of the previous lower uterine segment Cesarean section scar or niche and presence
6,9,14
222 of peri-trophoblastic or peri-placental vascularity on color Doppler examination .

223 We have collected a number of different ultrasound variables of the population

224 including gestational age, presence of a heartbeat, crown rump length, gestational sac

225 diameter, placental location, presence of placental lacunae, residual myometrial

226 thickness (RMT), adjacent myometrial thickness (AMT), subjective degree of

 11

227 vascularization and type of CSP (details can be viewed in supplemental material). CSP

228 is defined as type 1 if >50% of the gestational sac protrudes towards the uterine

229 cavity/cervical canal. CSP is defined as type 2 when the placenta implanted into a

230 deficient or dehiscent scar and the protrusion of the gestational sac is ≤50%. Residual

231 Myometrium Thickness (RMT) is the measurement of the residual myometrium

232 between the gestational sac and the serosa in a sagittal view. Adjacent Myometrium

233 Thickness (AMT) is the measurement of the myometrium next to the gestational sac

234 in a sagittal view (Figure 1).

f
oo
235

r
236 For the purpose of this study, we included all CSP from the start of the registry until

237
-p
data download on the 28.2.2023. We included both live and failing CSP (with or without
re
238 a positive heartbeat) that underwent surgical, medical or balloon catheter treatment
lP

239 before 12+6 weeks’ gestation to evaluate effectiveness and safety of the different
na

240 methods used (Figure 2). Expectantly managed cases of CSP were excluded for this

241 study. We collected data on complications with each treatment method. Adverse
ur

242 events which occurred as a result of procedures were considered as complications.

Jo

243 The questionnaire contained queries about treatment complications and maternal

244 morbidity (including blood loss ≥1000ml or requiring transfusion, surgical injury,

245 thrombotic event, sepsis, renal failure, cerebrovascular event, pulmonary edema,

246 maternal death and a free text box for clinics to add additional complications or

247 morbidities). We have collected Clavien-Dindo grade II, III and IV complications15.

248 Blood loss above 1000ml were counted as grade II, RPOC, bladder injury and deep

249 vein thrombosis were counted as grade III and sepsis were counted as grade IV.

250 Safety was defined as the inverse of the complication rate.

 12

251 Effectiveness was defined as the success rate. Treatments were considered

252 successful if the CSP resolved fully with no need for any additional medical or surgical

253 intervention. We compared ultrasound features of both live and failing CSP.

254

255 Statistical analysis

256 All descriptive statistical analysis was performed in the Department of Biostatistics,

257 University of Basel, Switzerland. The data is presented as counts and proportions for

258 categorical data or median and min, max for ordinal or metric data. P-values

f
oo
259 correspond to Mann–Whitney U test (for medians) and in case of categorical variables

r
260 Chi-squared or exact Fisher test were used depending on variable numbers. All

261
-p
evaluations were done using the statistical software R version 4.1.3.
re
262
lP

263 Results
na

264 Since the start of the CSP-registry in 2018 until data download on the 28.2.2023, 460

265 cases of CSP (281 live, 179 failing CSP) were recorded that either underwent surgical,
ur

266 medical or balloon catheter treatment before 12+6 weeks gestation (Fig. 2). A
Jo

267 summary of demographic and clinical data of the study population can be found in

268 Table 1.

269

270 Ultrasound features of CSP

271 Ultrasound features comparing live and failing CSP can be found in Table 2. There

272 was no difference in adjacent myometrial thickness (AMT) measurement or residual

273 myometrial thickness (RMT) between live and failed CSPs (p=0.55 and p=0.17

274 respectively) in our cohort.

 13

275 The majority of both live and failing CSPs were type 2 on scan. The crown rump length

276 (CRL) measured slightly larger in live CSP, p<0.001. However, there was no difference

277 in mean gestational sac diameter, p= 0.27. There was no statistically significant

278 difference in median gestational age between live CSP and failing CSP, p=0.14.

279

280 Surgical management (including suction evacuation, surgical resection, or primary

281 hysterectomy) was undertaken in 270/460 (58.7%), 123/460 (26.7%) underwent

282 medical management, 46/460 (10%) had balloon management and 21/460 (4.6%)

f
oo
283 underwent other, rarer treatment options, such as mifepristone/misoprostol or uterine

r
284 artery embolization. A summary of effectiveness and safety of the different treatment

285 options is shown in Table 3.

-p
re
286
lP

287 Surgical Management of CSP

na

288 The 270 cases underwent surgical management includes suction evacuation, surgical
ur

289 excision (via laparoscopy and laparotomy) and primary hysterectomy of CSP. A total
Jo

290 of 20 patients had a hysterectomy, 12 of which were primary and 8 secondary following

291 failure of the initial treatment. The secondary hysterectomy cases were analysed in

292 the subgroups of their first line treatments. To assess overall effectiveness and safety

293 of surgical treatment we assessed the 12 cases of primary hysterectomy separately.

294 Overall, surgical treatment of CSP was successful in 236/258 (91.5%, 95% CI 88.4-

295 94.5) and the complication rate was 24/258 (9.3%, 95% CI 6.6-11.9). The vast majority

296 of cases had a single second line treatment. There were five cases with multiple

297 treatments, all of which were in patients with a blood loss >1000ml.
 14

298 The most frequently used surgical treatment in both live and failed CSP was suction

299 evacuation with or without ultrasound-guidance. In 43 (15.3%) suction evacuation was

300 combined with prophylactic placement of a temporarily Shirodkar cerclage. Suction

301 evacuation was primary treatment of choice in 122/281 (43.4%) of live CSPs and

302 99/179 (55.3%) of failing CSPs. The presence of cardiac activity did not affect the

303 effectiveness of this treatment with 10/111 (9.0%) live CSPs and 9/87 (10.3%) failing

304 CSPs requiring additional treatment (p=0.75). Additional treatments specified were

305 laparotomy in three cases, repeat suction evacuation in ten cases, emergency

f
oo
306 hysterectomy with a blood loss of 3500ml in one case, systemic methotrexate in three

r
307 cases, balloon treatment and uterine artery embolization in one case each. Two

308
-p
patients a third-line of treatment: one patient had single balloon placement and uterine
re
309 artery embolization (UAE) as hemostatic measures. As these failed laparotomy and
lP

310 hysterectomy had to be performed. Overall, suction evacuation had relatively low
na

311 complication rate (Table 3). Ten of 221 (4.5%) suction evacuation cases had a

312 clinically significant blood loss (defined as loss of ≥1000ml), of which 5/99 (5.0%) were
ur

313 in failing CSP and 5/122 (4.1%) in live CSP between 7+0 and 12+4 weeks gestation.
Jo

314 Six out of 221 (3.1%) cases had retained products of conception (tissue from the

315 conception persisting after the pregnancy has ended), one case a deep vein

316 thrombosis (0.5%), one case sepsis (0.5%) and one bladder injury (0.5%) at 11+5

317 weeks.

318 There were 37 cases of surgical excision of the CSP in our dataset, which accounted

319 for 13.7% of surgical treatments utilized: 19/37 (51.3%) cases were live CSP and

320 18/37 (48.6%) failing CSP. Five (13.5%) were hysteroscopic resections, 15 (40.5%)

321 laparoscopical excisions and 17 (45.9%) laparotomies. Two cases that were started

322 as hysteroscopic resections and one case of laparoscopic excision needed converting
 15

323 to laparotomy and hysterectomy. The other cases did not require any additional

324 treatment. The overall complication rate of surgical excision was 13.5% (Table 3). In

325 terms of complications of surgical excision, two cases (5.4%) resulted in a clinically

326 significant blood loss (≥1000ml). There was one case of maternal sepsis (2.7%), one

327 case of Asherman syndrome (2.7%) and one case of broad ligament hematoma

328 (2.7%).

329 We recorded 12 cases of primary hysterectomy for CSP in the first trimester, of which

f
330 8 (66.7%) were between 10+0 and 12+6 weeks’ gestation, two were at 7 weeks’, one

oo
331 at 8 weeks’, and another one at 9 weeks’. The 12 hysterectomies accounted for 4.4%

r
332 -p
of first choice surgical treatment of CSP. Six (50%) cases (all after 10+0 weeks) were

333 done on asymptomatic cases in patients with an incidental finding of a CSP on scan.
re
334 The six symptomatic patients presented with heavy vaginal bleeding and abdominal
lP

335 pain. Three of the 12 (25%) cases were failed CSP and 9/12 (75.0%) cases were live
na

336 CSP. The median maternal age was 37 (31-59) and the median parity was 2.5. None
ur

337 of the 12 cases needed any additional treatment. However, there was a substantial
Jo

338 number of complications, including clinically significant bleeding (≥1000ml) in 5/12

339 (41.6%), and four (33%) bladder injuries.

340 Medical treatment

341 A total of 123 CSPs received medical management including 59 patients who had local

342 gestational sac injection and 64 who had systemic methotrexate. The treatment was

343 moderately efficient with 82/123 (66.6%) success rate and 16/123 (13.0%) had

344 significant complications.

345 Among 59 cases of local gestational sac injection of potassium chloride (KCL) (15/59

346 cases) or methotrexate (44/59 cases), there were 7 (11.8%) in whom this was
 16

347 combined with systemic methotrexate. Forty-seven out of 59 (79.6%) were live CSP

348 and 12/59 (20.4%) cases were failing CSPs. One-quarter (25.5%) of cases managed

349 with gestational sac injection needed additional treatment (12/44, 27.2% methotrexate

350 cases and 3/15 (20.0%) KCL cases). Six of 59 cases (10.1%) underwent suction

351 evacuation, one (1.6%) underwent balloon treatment, and eight (13.5%) underwent

352 uterine artery embolization. In two cases the suction evacuation was combined with

353 UAE.Three of 59 (5.1%) cases had clinically significant hemorrhage (≥1000ml), one

354 (1.6%) was subsequently diagnosed with Asherman syndrome, and in one case

f
oo
355 (1.6%) it took >150 days for hCG to return to pre-pregnancy level (median 62 days

r
356 (11-155 days)).
-p
re
357 We collected 64 cases of systemic methotrexate treatment, of which 33 (51.6%) were

358 in live CSP and 31 (48.3%) in failing CSP. Additional treatment was necessary in 30/64
lP

359 (46.9%). Five (7.8%) cases underwent surgical excision, five (7.8%) cases underwent
na

360 suction evacuation, two (3.1%) cases resulted in hysterectomy, six (9.3%) cases had
ur

361 uterine artery embolization and eight (12.5%) had a second course of methotrexate.
Jo

362 In one case surgical excision was combined with UAE and in another case firstly

363 suction evacuation and secondly surgical excision was needed. There were two

364 (3.1%) cases of maternal sepsis, one case (1.5%) of gestational trophoblastic

365 neoplasia, three (4.6%) cases of clinically significant hemorrhage (≥1000ml) and one

366 (1.5%) case of a prolonged inhomogeneous mass (RPOC) reported. Four (6.2%)

367 cases reported methotrexate toxicity.

368 Balloon catheter treatment

369 We collected 46 cases of balloon catheter treatment, 42/46 (91.3%) of which were live

370 CSPs and the remaining four of which were failing pregnancies. In 11 (23.9%) cases
 17

371 the treatment was combined with systemic methotrexate. Overall, balloon

372 management of CSP had a high success rate with 4/46 (8.7%) of cases requiring

373 additional treatment. Two patients requiring emergency hysterectomy for delayed

374 hemorrhage within the first 28 days after removal of the balloon (one at 22 days and

375 one at 27 days). One (2.1%) patient required uterine artery embolization for ‘arterio-

376 venous malformation’ and another had curettage. In five cases a transiently increased

377 blood flow within the uterine myometrium (enhanced myometrial vascularity (EMV)

378 was visible on follow-up scans, of which two were combined with an area of

f
oo
379 inhomogeneous mass (RPOC). Three of the five EMV disappeared spontaneously,

r
380 one required uterine artery embolization as this turned out to be an ‘arteriovenous

381
-p
malformation’ and one was suspected to be retained products of conception and
re
382 required surgical removal.
lP

383 Large volume vs small volume CSP clinics

na

384 There was no statistically significant difference in gestational age at treatment between
ur

385 small volume clinics (median 7.49 weeks, 5-12 weeks) and large volume clinics
Jo

386 (median 7.49, range 4-12 weeks), p=0.44. There was a difference in choice of first-

387 line treatment between small and large-volume clinics (p<0.001) and also in regard to

388 primary hysterectomy. 10/12 (83.3%) primary hysterectomies for CSP were done in

389 small-volume clinics, whereas only two were done at advanced gestation (12 weeks)

390 in large-volume clinics. Otherwise, both small and large-volume clinics reported on

391 suction evacuation (39.4% vs 57.6%), surgical excision (10.9% vs 7.6%,), local

392 medical treatment (13.1% vs 14.1%), and balloon catheter (12.4% vs 10.1%), systemic

393 methotrexate (24.1% vs 10.6%). Effectiveness was lower in small-volume clinics as

394 compared to high volume clinics with 27/152 (17.7%, 95th CI 12.1-23.4) compared to

395 40/308 (12.9%, 95th CI 9.5-16.5) requiring second line treatment (p<0.001) as a
 18

396 substantially higher proportion of cases were managed with systemic methotrexate.

397 There was also no statistically significant difference in regard to complications with

398 14/152 (9.2%, 95th CI 5.2-13.2) complications reported in small-volume clinics and

399 30/308 (9.7%, 95th CI 6.8-12.7) complications in large volume clinics, p=0.86.

400

401 Effectiveness and safety of treatment depending on gestational age

402 An overview on the treatments used at different gestational ages can be found in

f
oo
403 Figure 3. The percentage of cases requiring additional treatment in relation to

gestational age can be found in Figure 4. There was a negative correlation between

r
404

405
-p
the success of treatment and gestational age, p=0.11, r=-0.156. There was also a
re
406 moderate significant correlation between complications of treatment and gestational
lP

407 age (p<0.01, r=0.261). (Fig. 5)

na

408 Comment
ur

409 Principal findings

Jo

410 The results of our study showed that suction evacuation, surgical excision and double

411 balloon catheter all have a high success rate for the treatment of first trimester CSP.

412 Hysterectomy rates were higher in patients who underwent surgical excision and

413 treatment with balloon catheter compared to suction evacuation. Although this may be

414 a chance finding bearing in mind that the overall number of hysterectomy was low, our

415 findings indicate that close monitoring of hysterectomy rates following different

416 treatment options is needed. Local and systemic medical management have been less

417 effective than surgery with a higher proportion of patients requiring additional

418 treatment.
 19

419 Results in the context of what is known

420 Surgical management

421 Several smaller studies have investigated the effectiveness and safety of suction

422 evacuation for CSP. Harb et al’s UK cohort study published in 2018 reported on the

423 surgical and medical management of 92 patients with CSP 7. Comparable to our

424 findings, their success rate for surgical management was 96% (54/56), with 5.6% of

425 surgically managed cases requiring additional treatment. Our additional treatment rate

f
oo
426 was 8.5%. They reported a higher complication rate of 36% (20/56) compared to our

427 study rate of 9.6% (24/258) 7. Our complications rate was lower as we only included

r
428 -p
blood loss above 1000ml and major maternal morbidity as opposed to Harb et al.
re
429 where any bleeding complication regardless of amount was considered a
lP

430 complication. Maymon et al. published a systematic review and meta-analysis of 22

431 studies involving 374 patients with CSP16. The review reported a success rate of
na

16
432 92.2% for suction evacuation treatment, with low complication rate , similar to our
ur

433 study. A previously published systematic review by Ilan Timor-Tritsch based on

Jo

434 smaller case series reported a complication rate of dilatation and curettage of 62.9%,

435 this higher complication rate being attributed principally to bleeding complications . We

436 suspect that there may also be publication bias based on over reporting of cases that

437 had severe blood loss or other complications.

438 This same systematic review also reported higher complication rates from surgical
17
439 excision of 28.6% (n=14) , although the details of the complications were not listed

440 in this paper to allow direct comparison of complications. Verberkt et al. utilized

441 different surgical treatments depending on type of CSP and they reported high
 20

442 success rate for both suction curettage and laparoscopic niche resection with a low

443 complication rate 18.

444 Medical management

445 In the small UK cohort study, the success rate of medical treatment was 46% (7/15)

446 and the complication rate was 9/15 (60%) 7. Our success rate, based on larger

447 numbers, was slightly higher with 38/64 (59.3%) being sufficiently treated with medical

448 management alone. Our complication rate was substantially lower, however, at 13%

f
oo
449 (16/123). The differences in success and complication rates can be partially explained

450 by the fact that in study by Harb et al., 14/15 patients received systemic methotrexate

r
451
-p
as the primary treatment and only one case had local gestational sac injection,
re
452 whereas in our cohort only 52% (64/159) had systemic methotrexate and 48%
lP

453 (59/123) had local gestational sac injection. If we compared their medical complication

454 rate with our systemic methotrexate data then we would similarly report a substantially
na

455 lower success and higher complication rate. A recently published systematic review
ur

456 and meta-analysis of methotrexate treatment of CSP containing six articles with a
Jo

457 sample size of 600 individuals showed a success rate of 90.7% (95% CI: 86.7-93.5%),

458 and complication of 9% (95% CI: 6.3-12.8%), but reported a large heterogeneity of the

459 studies 19. We perceive that the small case numbers and failure to report the route of

460 methotrexate administration in some of the studies and/or systematic review may

461 account for the differences in success and complication rates, with local route for

462 methotrexate demonstrating the best effectiveness. In contrast, a previously published

463 systematic review by Timor-Tritsch reported a complication rate of systemic

464 methotrexate of 62.1% (n=87) 20.

 21

465 Several studies have suggested that local gestational sac injection is an effective

466 treatment option for early first trimester CSP. Timor-Tritsch published 19 cases treated

467 with intra-gestational sac injection of methotrexate and systemic methotrexate, none

468 of the cases had any complications and all were successfully treated 10.

469 Balloon catheter treatment

470 Timor-Tritsch et al. reported the first 60 patients with CSP who underwent double

471 balloon treatment, with a success rate of 85%, complete resolution of CSP and no

f
20

oo
472 recurrence during the follow-up period . However, in our group there were two

473 emergency hysterectomies (4.3%). The procedure was well-tolerated, with no major

r
474 -p
complications reported. This was confirmed on a subsequent publication with 38
re
475 patients who underwent double balloon treatment, which reported a success rate of
lP

476 98.8%, with only one patient requiring additional treatment due to incomplete

477 resolution of CSP. A retrospective cohort study of Type 1 CSP (n=18) managed with
na

478 double balloon treatment published by Kus et al. showed low morbidity and high
ur

479 treatment success 21. Our data support the findings of previous studies that the double
Jo

480 balloon treatment is an efficient treatment method for CSP, however needs close

481 follow-up and we will continue to monitor its safety in the future.

482 More than 30 different treatment options have been described in the literature in the

483 treatment of CSP including combined treatments and different methods of local

484 injections. However, their role in first line treatment seem to be limited since we found

485 only 21 cases that were not classified into the treatment options that we analysed. For

486 further analysis of these treatment options more cases are needed.

487 Gestational age and risk of complications / additional treatment

 22

488 Several studies have suggested that the risk of complications in the treatment of CSP

489 is higher when the gestational age of the pregnancy is advanced. In a previous study

490 of 15 patients, all those managed at gestational ages of < 8 weeks had complete

491 resolution of CSP, while those with gestational ages of ≥ 8 weeks had a higher risk of

492 incomplete resolution. A recently published systematic review and meta-analysis

493 including 36 studies (724 patients with CSP) reported an overall adverse outcome

494 complicated 5.9 % (95 % CI 3.5-9.0) of CSP diagnosed ≤9 weeks and 32.4 % (95 %

495 CI 15.7-51.8) of those diagnosed > 9 weeks 22.

f
oo
496 The risk of complications increasing with advancing gestational age in CSP is thought

r
497 -p
to be due to several factors. Firstly, the placenta tends to become more deeply

498 implanted in the scar tissue as the pregnancy advances, which can increase the risk
re
499 of bleeding and cause incomplete resolution during treatment. Secondly, the size of
lP

500 the gestational sac also increases with advancing gestational age, which can make it
na

501 more difficult to completely evacuate the sac during treatment 22.
ur

502 Research implications

Jo

503 To further analyze effectiveness and safety for individual treatments according to

504 gestational week or for potential differences in effectiveness and safety between type

505 I and type II CSP further growth of the registry as more cases are collected.

506 Strength and limitations

507 The strength of our study is the unique study population, representing multiple centers

508 around the world with different treatment approaches advocated, each of centers

509 choosing their preferred treatment. The results therefore are unbiased toward a

510 specific treatment modality based on an individual clinician, center or country bias and

511 results can be considered generalizable. Furthermore, it is the first study to directly
 23

512 compare various treatments within a large patient population using primary data,

513 permitting clinicians to use these direct comparisons of effectiveness and

514 complications when discussing options with their patients. The registry also allows for

515 participation of a much larger number of centers compared to a prospective study as

516 required for the study on rare entities such as CSP.

517 The study has the limitations of data sets being dependent on accurate record keeping

518 within each contributing center, as is the case in all retrospective and registry-based

f
519 data collections. A registry allows international participation from centers who have an

oo
520 interest in the condition which poses a risk of unrecognizing early CSP or over-

r
521 -p
reporting of complications in centers which manage CSP rarely.
re
522 Clinical implications and Conclusions
lP

523 CSP in the first trimester of pregnancy can be managed effectively in more than 90%
na

524 of cases with either suction evacuation, balloon treatment or surgical excision. Surgical

525 excision requires an experienced surgeon due to risk of emergency hysterectomy and
ur

526 double balloon catheter patients require close follow-up. The effectiveness of all
Jo

527 treatment options decreases with advancing gestational age CSP should be treated

528 as early as possible after confirmation of the diagnosis. Local medical treatment with

529 KCl or methotrexate is less efficient in and has higher rates of complications compared

530 to the other treatment options. Systemic methotrexate has a substantial risk of failing

531 and a higher complication rate and should not be recommended as first line treatment.
 24

532 Acknowledgement

533 We especially thank the Professor Basky Thilaganathan from St George’s University

534 Hospitals NHS Foundation who initially supported the CSP registry team as the chief

535 investigator but chose not to participate as author of this study.

f
r oo
-p
re
lP
na
ur
Jo
 25

536 Literature

537 1. Timor-Tritsch IE, Monteagudo A, Cali G, El Refaey H, Kaelin Agten A, Arslan AA. Easy
538 sonographic differential diagnosis between intrauterine pregnancy and cesarean delivery scar
539 pregnancy in the early first trimester. Am J Obstet Gynecol. 2016;215(2):225.e1-7.
540 doi:10.1016/j.ajog.2016.02.028

541 2. Timor-Tritsch IE, Monteagudo A. Unforeseen consequences of the increasing rate of cesarean
542 deliveries: early placenta accreta and cesarean scar pregnancy. A review. Am J Obstet
543 Gynecol. 2012;207(1):14-29. doi:10.1016/J.AJOG.2012.03.007

544 3. Betrán AP, Ye J, Moller AB, Zhang J, Gülmezoglu AM, Torloni MR. The Increasing Trend in
545 Caesarean Section Rates: Global, Regional and National Estimates: 1990-2014. PLoS One.
546 2016;11(2):e0148343. doi:10.1371/journal.pone.0148343

f
547 4. Society for Maternal-Fetal Medicine (SMFM) M, Miller R, Gyamfi-Bannerman C, Publications

oo
548 Committee. Electronic address: pubs@smfm.org. Society for Maternal-Fetal Medicine Consult
549 Series #63: Cesarean scar ectopic pregnancy. Am J Obstet Gynecol. 2022;227(3):B9-B20.

r
550 doi:10.1016/j.ajog.2022.06.024

551 5.
-p
Timor-Tritsch IE, Monteagudo A, Cali G, et al. Cesarean scar pregnancy is a precursor of
552 morbidly adherent placenta. Ultrasound Obstet Gynecol. 2014;44(3):346-353.
re
553 doi:10.1002/UOG.13426
lP

554 6. Kaelin Agten A, Cali G, Monteagudo A, Oviedo J, Ramos J, Timor-Tritsch I. The clinical
555 outcome of cesarean scar pregnancies implanted “on the scar” versus “in the niche.” Am J
556 Obstet Gynecol. 2017;216(5):510.e1-510.e6. doi:10.1016/j.ajog.2017.01.019
na

557 7. Harb HM, Knight M, Bottomley C, et al. Caesarean scar pregnancy in the UK: a national cohort
558 study. BJOG. 2018;125(13):1663-1670. doi:10.1111/1471-0528.15255
ur

559 8. Ben Nagi J, Ofili-Yebovi D, Marsh M, Jurkovic D. First-trimester cesarean scar pregnancy
Jo

560 evolving into placenta previa/accreta at term. J Ultrasound Med. 2005;24(11):1569-1573.

561 doi:10.7863/jum.2005.24.11.1569

562 9. Jordans IPM, Verberkt C, Leeuw RA, et al. Definition and sonographic reporting system for
563 Cesarean scar pregnancy in early pregnancy: modified Delphi method. Ultrasound Obstet
564 Gynecol. Published online November 14, 2021. doi:10.1002/UOG.24815

565 10. Timor-Tritsch IE, Monteagudo A, Santos R, Tsymbal T, Pineda G, Arslan AA. The diagnosis,
566 treatment, and follow-up of cesarean scar pregnancy. Am J Obstet Gynecol.
567 2012;207(1):44.e1-13. doi:10.1016/j.ajog.2012.04.018

568 11. Jurkovic D, Hillaby K, Woelfer B, Lawrence A, Salim R, Elson CJ. First-trimester diagnosis and
569 management of pregnancies implanted into the lower uterine segment Cesarean section
570 scar. Ultrasound Obstet Gynecol. 2003;21(3):220-227. doi:10.1002/uog.56

571 12. Agten AK, Monteagudo A, Timor-Tritsch IE, Thilaganathan B. Cesarean Scar Pregnancy
572 Registry: an international research platform. Ultrasound Obstet Gynecol. 2020;55(4):438-440.
573 doi:10.1002/uog.21952
 26

574 13. Miller R, Timor-Tritsch IE, Gyamfi-Bannerman C. Society for Maternal-Fetal Medicine (SMFM)
575 Consult Series #49: Cesarean scar pregnancy. Am J Obstet Gynecol. 2020;222(5):B2-B14.
576 doi:10.1016/J.AJOG.2020.01.030

577 14. Timor-Tritsch IE, Monteagudo A, Calì G, D’Antonio F, Kaelin Agten A. Cesarean Scar
578 Pregnancy: Diagnosis and Pathogenesis. Obstet Gynecol Clin North Am. 2019;46(4):797-811.
579 doi:10.1016/J.OGC.2019.07.009

580 15. Dindo D, Demartines N, Clavien PA. Classification of surgical complications: a new proposal
581 with evaluation in a cohort of 6336 patients and results of a survey. Ann Surg.
582 2004;240(2):205-213. doi:10.1097/01.sla.0000133083.54934.ae

583 16. Maymon R, Halperin R, Mendlovic S, et al. Ectopic pregnancies in Caesarean section scars: the
584 8 year experience of one medical centre. Hum Reprod. 2004;19(2):278-284.
585 doi:10.1093/HUMREP/DEH060

f
586 17. Timor-Tritsch IE, Monteagudo A, Cali G, et al. Cesarean scar pregnancy and early placenta

oo
587 accreta share common histology. Ultrasound Obstet Gynecol. 2014;43(4):383-395.
588 doi:10.1002/uog.13282

r
589
590
18. -p
Verberkt C, Lemmers M, de Leeuw RA, et al. Effectiveness, complications, and reproductive
outcomes after cesarean scar pregnancy management: a retrospective cohort study. AJOG
re
591 global reports. 2023;3(1):100143. doi:10.1016/j.xagr.2022.100143

592 19. Salari N, Kazeminia M, Shohaimi S, Nankali AAD, Mohammadi M. Evaluation of treatment of
lP

593 previous cesarean scar pregnancy with methotrexate: a systematic review and meta-analysis.
594 Reprod Biol Endocrinol. 2020;18(1):108. doi:10.1186/s12958-020-00666-0
na

595 20. Timor-Tritsch IE, Khatib N, Monteagudo A, Ramos J, Berg R, Kovács S. Cesarean scar
596 pregnancies: experience of 60 cases. J Ultrasound Med. 2015;34(4):601-610.
597 doi:10.7863/ULTRA.34.4.601
ur

598 21. Kus LH, Veade AE, Eisenberg DL, Dicke JM, Kelly JC, Dickison SM. Maternal Morbidity After
Jo

599 Double Balloon Catheter Management of Cesarean Scar and Cervical Pregnancies. Obstetrics
600 and gynecology. 2022;140(6):993-995. doi:10.1097/AOG.0000000000004977

601 22. Timor-Tritsch I, Buca D, Di Mascio D, et al. Outcome of cesarean scar pregnancy according to
602 gestational age at diagnosis: A systematic review and meta-analysis. Eur J Obstet Gynecol
603 Reprod Biol. 2021;258:53-59. doi:10.1016/J.EJOGRB.2020.11.036

604
 27

605 Table 1: Demographics of the study population

Maternal demographics Median [Min, N=460

Max] or (%)

Age (years) 35.0 [18.0;59.0] 460

Ethnicity 426

• White 241 (56.6%)

• Middle Eastern 75 (17.6%)
• Asian 36 (8.5%)
• Afro-Caribbean 38 (8.9%)

f
• Hispanic 20 (4.7%)

oo
• mixed, other 16 (3.8%)
BMI 26.7 [16.7;54.6] 291

r
Smoking -p
• never or gave up before pregnancy 294 (85.7%)
343
re
• current 49 (14.3%)
Parity 2 [1;11] 458
lP

Conception 445
• Spontaneous 418 (93.9%)
• Artificial conception with embryo placement 25 (5.6%)
na

• Artificial conception with ovarian stimulation 2 (0.5%)

•
Symptomatic at diagnosis 362 (65.8%) 458
• Vaginal bleeding
ur

173 (37.8%)
• Abdominal pain 29 (6.33%)
• Incidental finding 185 (40.4%)
Jo

• Combination
plXWplacement of pain and bleeding
placement 66 (14.4%)
• Other symptoms 5 (1.1%)
Number of previous CS 2 [1;8] 459

606

607 Table 1. Demographics of the study population. BMI = Body-Mass-Index (kg/m2).

608 Median and Min, Max or percentage (%).

 28

609 Table 2. Ultrasound findings of live and failing CSP at diagnosis

Live CSP Failing CSP n=179 p-value n

n=281

Gestational age at 8+2 7+5 [5+2;12+6] 0.14 457

treatment (wks) [3+6;12+6]

CRL at diagnosis 7.0 [1.0;74.0] 4.0 [7.0;95.0] <0.001 331

(mm)

Gestational sac 18.0 [1.0;71.0] 17.0 [1.0;81.0] 0.27 356

f
oo
diameter (mm)

r
Site of implantation 0.038 448

Type 1 104 (37.5%)

-p
82 (48.0%)
re
Type 2 173 (62.5%) 89 (52.0%)
lP

Residual myometrial 2.8 [0.3;10.5] 3.00 [0.5;12.9] 0.17 316

thickness (mm)
na

Adjacent myometrial 13.0 [0.4;41.6] 13.0 [0.5;30.0] 0.55 217

ur

thickness (mm)
Jo

Enhanced myometrial 149 (66.5%) 74 (52.9%) 0.013 364

vascularity in the area

of the scar

610

611 Table 2: CSP is defined as type 1 if >50% of the gestational sac protrudes towards

612 the uterine cavity/cervical canal. CSP is defined as type 2 when the placenta

613 implanted into a deficient or dehiscent scar and the protrusion of the gestational sac

614 is ≤50%. Residual Myometrium Thickness (RMT) is the measurement of the residual

615 myometrium between the gestational sac and the serosa in a sagittal view. Adjacent
 29

616 Myometrium Thickness (AMT) is the measurement of the myometrium next to the

617 gestational sac in a sagittal view.

f
roo
-p
re
lP
na
ur
Jo
 30

618 Table 3. Success rates and complications of different management options for

619 CSP in the first trimester

Success rate (effectiveness) Complication rate (safety) Type of complication

n (%) [95th CI] n (%) [95th CI] (n)

Suction 202/221 91.5% [87.8-95.2] 19/221 8.5% [5.9-11.2] Hemorrhage** (10);

evacuation RPOC (6); thrombosis

(1); sepsis (1); bladder

injury (1)

f
Surgical 34/37 91.8% [83.8-99.9] 5/37 13.5% [0-29.1] Hemorrhage** (2);

oo
excision Sepsis (1); Asherman

r
(1); broad ligament
-p hematoma (1)
re
Balloon catheter 42/46 91.3% [83.5-99.1] 4/46* 8.7% [2.4-14.9] Hemorrhage** (2);
lP

treatment EMV (2)

Local 44/59 74.5% [64.1-85.1] 5/59 9.5% [1.6-15.4] Hemorrhage** (3);

na

gestational sac Asherman (1); delayed

injection resorption >150 days

ur

(1)
Jo

Systemic 38/64 59.4% [48.4-70.4] 11/64 23.9% [8.5-25.9] Hemorrhage** (3);

methotrexate methotrexate toxcicity

(4); Sepsis (2),

RPOC (1); GTN (1)

Primary 12/12 100% 9/12 75.0% [74.8- Hemorrhage** (5)

Hysterectomy 75.3] bladder injury (4)

620

621 Table 3: Complication rate and additional treatment rate of the different treatment

622 options. Details of the complications of each treatment method can be found in the

623 results section. *Only 2/5 enhanced myometrial vascularity (EMV) that required
 31

624 treatment were included in the complication rate. 3/5 EMV showed spontaneous

625 resolution and were therefore not considered a complication. The other 2

626 complications were delayed hemorrhage after balloon treatment Number of cases and

627 percentage (%). [95th CI] 95% confidence interval. Abbreviations: RPOC: retained

628 products of conception. GTN: gestational trophoblastic neoplasia. **Hemorrhage

629 refers to a blood loss>1000 ml.

f
r oo
-p
re
lP
na
ur
Jo
 32

630 Figure legends

631 Figure 1: CSP Type 1 and CSP Type 2

632 Schematic drawing and typical gray scale ultrasound images of CSP Type I (A) and

633 CSP Type II (B) showing the measurements of AMT and RMT. CSP type 1: if >50%

634 of the gestational sac protrudes towards the uterine cavity/cervical canal. CSP is type

635 2: the protrusion of the gestational sac into the cavity is ≤50%. Residual Myometrium

636 Thickness (RMT): measurement of the residual myometrium between the gestational

f
oo
637 sac and the serosa in a sagittal view. Adjacent Myometrium Thickness (AMT):

638 measurement of the myometrium next to the gestational sac in a sagittal view.

r
639 Figure 2: Cases included in study
-p
re
640 Summary of included cases. All cases that underwent expectant management or were
lP

641 managed ≥13+0 weeks were excluded and only cases that underwent surgical,
na

642 medical or balloon catheter treatment ≤12+6 weeks gestation were included. *Cases
ur

643 where no objective measurement of gestational age (i.e. crown-rump length or

644 gestational sac diameter) was possible were excluded as we could not accurately
Jo

645 determine them to be ≤12+6 weeks.

646 Figure 3. Number of cases treated at different gestational ages.

647 Number of cases treated with medical treatment, surgical treatment and Double

648 Balloon at different gestational ages.

649 Figure 4. Need for additional treatment at different gestational ages in live CSP.

650 Percentage of live CSP cases requiring additional treatment after different types of

651 first–line management depending on gestational age (effectiveness) for surgical,

652 medical and double balloon treatment.

 33

653 Figure 5. Safety of first line treatment at different gestational ages in live CSP.

654 Percentage of cases of live CSP having complications depending on gestational age

655 (Safety) for surgical, medical and double balloon treatment.

656

f
r oo
-p
re
lP
na
ur
Jo
Jo
ur
na
lP
re
-p
ro
of
Jo
ur
na
lP
re
-p
ro
of
Jo
ur
na
lP
re
-p
ro
of
Jo
ur
na
lP
re
-p
ro
of
Jo
ur
na
lP
re
-p
ro
of