Carboplatin in combination with oral or intravenous etoposide for extra-pulmonary,

poorly-differentiated neuroendocrine carcinomas

Melissa Frizziero1, Francesca Spada2, Angela Lamarca1,£, Zoe Kordatou1, Jorge Barriuso1,3, Christina Nuttall1,

Mairéad G McNamara1,3, Richard A Hubner1, Wasat Mansoor1, Prakash Manoharan4, Nicola Fazio2, Juan W

Valle1,3.

Affiliations:

1Department of Medical Oncology, The Christie NHS Foundation Trust, Manchester, United Kingdom.

2Gastrointestinal Medical Oncology and Neuroendocrine Tumors Unit, European Institute of Oncology, Milan,

Italy.

3Division of Cancer Sciences, University of Manchester, United Kingdom.

4Department of Radiology and Nuclear Medicine, The Christie NHS Foundation Trust, Manchester, United

Kingdom.

£Part-funded by the ASCO (American Society of Clinical Oncology) Conquer Cancer Foundation Young

Investigator Award.

Short Title: Carboplatin-etoposide for extra-pulmonary neuroendocrine carcinomas

Corresponding author:

Prof Juan W. Valle

Professor and Honorary Consultant in Medical Oncology

Department of Medical Oncology

The Christie NHS Foundation Trust

550 Wilmslow Road, M20 4BX

Manchester, United Kingdom

Juan.Valle@christie.nhs.uk

Tel: +44 161 446 8106

Fax: +44 446 3468

Keywords; carboplatin-etoposide, extra-pulmonary neuroendocrine carcinoma, oral etoposide, intravenous

etoposide.

1

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

26

27

28

29

30

31

32

33

1. Abstract

Background:

Carboplatin-Etoposide (CarboEtop) is a 1st-line option for patients with advanced extra-

pulmonary (EP), poorly-differentiated (PD) neuroendocrine carcinoma (NEC). Different

schedules are used in clinical practice and randomised evidence is lacking.

Objectives:

To provide real-life outcomes of carboplatin combined with oral or intravenous (IV) etoposide

(Etop) in advanced EP-PD-NEC, from 2 specialist centres.

Methods:

Activity/efficacy/toxicity data of CarboEtop were retrospectively collected and analysed.

Results:

We identified 113 patients; median age: 65.8 years; male: 64%; gastro-entero-pancreatic

origin: 54%; stage IV: 90%; median Ki-67: 70%; median follow-up: 11.5 months. A total of

123 courses of CarboEtop (oral: 45%; IV: 55%) were administered; 106 (86%) 1st-line, 16

(13%) 2nd-line and 1 (1%) 3rd-line. Disease-control-rate: 74.5% in 1st-line and 69.2% in 2nd/3rd-

line, with no significant difference between oral- and IV-Etop in 1st-line (69.8% versus 80.8%,

p=0.237). Median progression-free survival (PFS): 6.0 and 4.5 months in 1st-line and 2nd/3rd-

line, respectively. Overall survival (OS): 11.5 and 12.5 months in 1st-line and 2nd/3rd-line,

respectively. The schedule (oral- versus IV-Etop) did not impact on 1st-line PFS (5.6 versus

6.2 months, p=0.179), although there was a trend towards shorter OS (8.9 versus 12.1

months, p=0.069). Liver metastases correlated with worse 1st-line-PFS (p=0.015) and 1st-

line-OS (p<0.001) on multivariable analysis. The commonest grade 3-4 adverse event was

myelosuppression (49%), with comparable toxicity between oral- and IV-Etop, except for

venous thromboembolism (12.5% versus 1.7%, p=0.04).

Conclusions:

2

34

35

36

37

38

39

40

41

42

43

44

45

46

47

48

49

50

51

52

53

54

55

56

57

58

59

CarboEtop for advanced EP-PD-NEC is active, effective and well-tolerated. Oral- and IV-

Etop schedules are associated with comparable toxicity; activity should be compared in

larger cohorts.

2. Introduction

Extra-pulmonary (EP) neuroendocrine carcinoma (NEC) is a poorly-differentiated (PD)

neuroendocrine neoplasm with a mitotic count >20/10 high power field (HPF) and/or a Ki-67

index >20%, as defined by the 2010 World Health Organisation (WHO) classification [1].

This diagnosis, although rare [incidence in Europe; 0.52/100,000 individuals/annum

according to the Surveillance of Rare Cancers in Europe (RARECARE) Registry

(http://www.rarecare.eu, July 20th, 2018)], has been increasingly encountered in routine

practice [2]. The gastro-entero-pancreatic (GEP) tract represents the most common site of

origin of EP-PD-NEC (up to 83% of cases), whereas the proportion of cases where a primary

tumour cannot be identified varies greatly across different studies (5-67%) [3-9].

Extra-pulmonary PD-NEC is an aggressive entity with 57-73% of patients presenting with

distant metastatic disease at diagnosis [4, 5, 10]; these patients have limited treatment

options and an extremely poor prognosis, with a median overall-survival (OS) as short as 1

month for those who receive only best supportive care, and between 9.2 and 19 months for

those who are treated with palliative chemotherapy [3, 4, 6-9, 11, 12]. The variability in

survival outcomes across different series is at least in part a reflection of the heterogeneity in

the biology of EP-PD-NEC and clinical characteristics of affected patients [13]. To guide

clinical decision making, a 5-variable prognostic score [gastro-intestinal (GI)-NEC score] was

developed and validated [14]. This score may be used to assess patient eligibility for active

treatment and for patient stratification in clinical trials.

Due to the rarity of this disease, data from large randomised trials are lacking. As a result,

clinical practice guidelines are usually extrapolated from the more extensive literature on

3

60

61

62

63

64

65

66

67

68

69

70

71

72

73

74

75

76

77

78

79

80

81

82

83

84

85

86

87

lung-NEC, based on the assumption of a biological similarity between these two entities. In

keeping with the treatment paradigm for their pulmonary counterpart, platinum-based

chemotherapy is advocated as ‘standard-of-care’ active treatment for advanced EP-PD-NEC

[15].

Although the use of cisplatin/etoposide combination is the most extensively documented in

the literature of lung- and EP-PD-NEC, carboplatin-etoposide (CarboEtop) has emerged as a

valid and equally-effective alternative [16]. In the large retrospective NORDIC study, which

included 305 patients with a diagnosis of advanced EP-NEC (252 treated with

chemotherapy), cisplatin-based and carboplatin-based regimens were associated with

comparable first-line response rates (RRs), progression-free-survival (PFS) and OS [6].

Likewise, in a meta-analysis of clinical trials comparing cisplatin-based and carboplatin-

based chemotherapy as first-line treatment for patients with small cell lung carcinoma

(SCLC), the most common variant of lung-NEC, no differences were found in activity or

efficacy outcomes between the two subgroups [17].

In the context of a rapidly disseminating disease, such as EP-PD-NEC, where patients often

present with already affected physical conditions at the beginning of the treatment,

CarboEtop is often preferred in routine practice over cisplatin-etoposide, due to its less

likelihood of causing gastro-intestinal toxicities (and possible consequent dehydration), renal

function impairment and peripheral neuropathy [17]. However, evidence on CarboEtop in

advanced EP-PD-NEC is scarce and mainly derived from small retrospective series [6, 11,

18, 19]. As a result, there is large variability in the schedules (e.g. oral or intravenous

etoposide, 3- or 4-weekly schedules) and doses of CarboEtop across different centres, and

which of them should be recommended remains unclear.

The use of oral formulations of etoposide carries undoubtable advantages for patients and

provides cost saving by reducing the frequency and duration of patient attendance to

hospital. A review of the literature, including 48 studies, suggests that patients affected by

either solid or haematological cancers prefer oral over intravenous anti-cancer treatments,

as oral formulations are perceived as more convenient and less toxic, and can be taken at

4

88

89

90

91

92

93

94

95

96

97

98

99

100

101

102

103

104

105

106

107

108

109

110

111

112

113

114

115

home and interfere less with work schedules [20]. Randomised trials comparing oral versus

intravenous etoposide in combination with cisplatin for the treatment of patients with SCLC

have shown that the two formulations of etoposide have equivalent activity and efficacy

outcomes, and a similar toxicity profile [21]. However, some pharmacokinetic studies have

shown that oral etoposide, compared to intravenous etoposide, has a higher inter- and intra-

individual variability in bioavailability, which might result in suboptimal tumour activity and

increased toxicity [22, 23], whereas others have reported equivalence in bioavailability

between the two formulations [24].

Therefore, acknowledging the difficulties in pursuing randomised comparisons between

different schedules of CarboEtop in the context of a rare disease, the current study aimed to

conduct a large, retrospective, bi-centre, data collection on treatment outcomes of

carboplatin in combination with either oral or intravenous etoposide for advanced EP-PD-

NEC, to inform clinical practice.

3. Materials and methods

This retrospective, bi-centre study was approved by The Christie NHS Foundation Trust

Audit committee (SE17/1992) and the local audit committee of The European Institution of

Oncology. Informed signed consent from individual patients was not required to be obtained.

Patient eligibility criteria for inclusion in this study were: histologically or cytologically proven

diagnosis of neuroendocrine neoplasms with a Ki-67 index ≥20%; a primary tumour from

outside the lung (extra-pulmonary) [patients with unknown primary tumours were included,

provided a lung origin was ruled out], advanced stage disease not amenable to curative

treatment, and receipt of at least one line of palliative chemotherapy with CarboEtop (any

schedule) [patients who received re-challenge with carboplatin and etoposide were allowed].

Exclusion criteria were: known well-differentiated morphology (morphology was not

5

116

117

118

119

120

121

122

123

124

125

126

127

128

129

130

131

132

133

134

135

136

137

138

139

140

141

142

systematically reviewed for all patients with unreported morphology); failed completion of at

least one cycle of CarboEtop, or date of death or last follow-up visit/contact not available.

Eligible patients were identified among those referred to and treated at two European

Neuroendocrine Tumour Society (ENETS) Centres of Excellence; The Christie National

Health Service (NHS) Foundation Trust (Manchester, United Kingdom) and The European

Institution of Oncology (Milan, Italy). All consecutive patients meeting eligibility criteria and

treated between 1st September 1996 and 28th of February 2018 were considered.

Patient demographic data, tumour pathological characteristics, treatment modalities and

outcomes (activity, efficacy and toxicity) were collected retrospectively from local medical

records. The performance status of patients was reported according to the European

Cooperative Oncology Group (ECOG) scale; where the ECOG score was not clearly stated

in the medical records, it was extrapolated either by conversion from a different scale (e.g.

Karnofsky) or based on the description of clinical signs and symptoms. The choice of

CarboEtop schedule was as per clinician preference or local protocol. Dose intensity of

CarboEtop was calculated as the percentage of dose actually delivered to a patient, taking

as reference the full dose estimated to be delivered based on the number of cycles of

CarboEtop administered. Radiological response to CarboEtop was assessed by

conventional imaging and classified according the Response Evaluation Criteria in Solid

Tumours (RECIST) version 1.1 [26]. When measurements of target tumour lesions were not

provided, original scan images were reviewed by a Radiologist Specialist with expertise in

NENs for RECIST 1.1 calculation. Waterfall plots were generated by plotting variations in

marker lesions at the time of the best response achieved (compared to baseline).

Progression-free-survival (PFS), time-to-progression (TTP) and OS were defined as the time

from the beginning of CarboEtop chemotherapy to the time of radiological/clinical

progression, or death from any cause (PFS), radiological progression (TTP) or death from

any cause (OS), respectively. Adverse events (AEs) of CarboEtop were classified according

to the National Institutes of Health Common Terminology Criteria for Adverse Events

(CTCAE), version 4.03

6

143

144

145

146

147

148

149

150

151

152

153

154

155

156

157

158

159

160

161

162

163

164

165

166

167

168

169

170

[https://ctep.cancer.gov/protocoldevelopment/electronic_applications/ctc.htm. July 20th,

2018]. Myelosuppression was defined as the presence of at least one haematological AE

(anaemia, thrombocytopenia or neutropenia).

3.1. Statistical analysis

Percentages, median values, ranges and 95%-confidence intervals (95%-CIs) were

calculated for description of categorical and continuous variables, as appropriate. Pearson

chi-square, Fisher’s exact tests and Student-T test were used as required. Receiver

operating characteristic (ROC) and calculation of area under the curve (AUC) were used for

identification of Ki-67 cut-offs. Kaplan-Meier analysis was applied to estimate median values,

hazard ratio (HR) and 95%-CIs of survival outcomes. Log-rank test for equality of survivors

function, Cox-regression univariate and multivariable analyses were used to interrogate

potential correlations between survival outcomes and patient clinical-pathological

characteristics at baseline or treatment-related data (schedules, causes of treatment

discontinuation, dose intensity) in the first-line setting. Probability values (p) were considered

statistically significant at a level below 0.05. Variables reaching statistical significance in

univariate analysis were included in the multivariable analysis. Descriptive and inferential

statistical analyses were performed using the “Statistics and Data” (STATA) software

package.

4. Results

4.1. Patients characteristics

One-hundred and thirteen patients met the criteria for inclusion in this study. The median

follow-up time was 11.5 months (range; 0.4-99.7). The median age at the beginning of

CarboEtop was 65.8 years (range; 24-88). The majority of patients were male (63.7%) and

had a primary tumour from the gastro-entero-pancreatic (GEP) tract (54.0%). Small cell

7

171

172

173

174

175

176

177

178

179

180

181

182

183

184

185

186

187

188

189

190

191

192

193

194

195

196

197

198

carcinoma was the predominant histological subtype (48.7%). Five (4.4%) patients had

previous potentially-curative surgery for early stage EP-PD-NEC. Among them, 4 received

adjuvant chemotherapy [cisplatin and etoposide (2 patients), streptozotocin and capecitabine

(1 patient), cisplatin and gemcitabine (1 patient)] and 1 patient received both neoadjuvant

chemo-radiotherapy and adjuvant chemotherapy with 5-fluorouracil/leucovorin and

oxaliplatin. Additional baseline patient and tumour characteristics are presented in Table 1.

Frequency distribution of Ki-67 values is illustrated in Figure 1.

4.2. Treatment modalities of CarboEtop

Treatment modalities are illustrated in Figure 2. A total of 123 courses of CarboEtop were

administered; 106 (86.2%) in the first-line, 16 (13.0%) in the second-line and 1 (0.8%) in the

third-line setting. Ten (8.8%) patients treated with CarboEtop in first-line received re-

challenge CarboEtop in second-line.

The following schedules of CarboEtop were applied (Supplementary Table S1):

CarboEtop-1; etoposide 50 mg twice daily orally from day 1 to day 7 (inclusive) followed

by carboplatin area under the curve (AUC) 5, intravenously on day 8, every 28 days;

CarboEtop-2; etoposide 120 mg/m2 intravenously on days 1, 2, and 3, and carboplatin

AUC 5 or 6 intravenously on day 1, every 21 days;

CarboEtop-3; etoposide 100 mg/m2 intravenously on days 1, 2, and 3, and carboplatin

AUC 4 or 5 intravenously on day 1, every 21 days;

CarboEtop-4; etoposide 120 mg/m2 intravenously on day 1 and 100-150 mg twice daily

orally on days 2 and 3, and carboplatin AUC 5 intravenously on day 1, every 21 days.

A higher proportion of patients received intravenous etoposide compared to oral etoposide,

both in first-line (54.7% versus 45.3%) and second/third-line (58.8% versus 41.2%). The

median time on chemotherapy was 3.6 months (range; 0.4-9.9) in first-line and 3.2 months

(range; 0.9-6.4) in second/third-line. The median number of cycles of CarboEtop delivered

was 4 (range; 1-13) in first-line and 4 (range; 1-8) in second/third-line.

8

199

200

201

202

203

204

205

206

207

208

209

210

211

212

213

214

215

216

217

218

219

220

221

222

223

224

225

Causes of CarboEtop discontinuation included; completion of planned treatment period

(41.5%), deterioration in general condition/clinical progression (25.2%), toxicity (14.6%),

patient decision (10.6%) and death (8.1%). There were no statistically significant differences

in the frequencies of causes of treatment discontinuation between schedules with oral and

intravenous etoposide in the first-line setting (Supplementary Table S2).

The dose-intensity of CarboEtop could be calculated for 82 courses of treatment, and the

median value was 94.3% (95%-CI; 91.3-99.8), when all courses/lines of treatment were

included, 95.7% (95%-CI; 91.7-100) in first-line and 91.1% (95%-CI; 84.7-100) in

second/third-line. The median dose-intensity of CarboEtop in first-line did not differ

significantly between schedules with oral and intravenous etoposide [99.7% (95%-CI; 92.1-

100.0) versus 91.8% (95%-CI; 83-96.8), respectively; p = 0.367].

4.3. Activity outcomes of CarboEtop

Radiological response to CarboEtop could be assessed for 111 courses (90.2%) of

treatment. The most common best response (BR) achieved was partial response (PR) in

first-line (40.8%) and stable disease (SD) in second/third-line (38.45%). The median time to

BR was 2.9 months (95%-CI; 2.5-3.0) in first-line and 2.7 (95%-CI; 2.1-4.5) in second/third-

line. Further details on the activity of CarboEtop are provided in Table 2 and

Supplementary Figure S1.

Comparisons were made between activity outcomes of schedules with oral and intravenous

etoposide in first-line. No statistically significant differences were found in disease-control-

rate (DCR) [complete response (CR) + PR + SD] between oral [69.8% (95%-CI; 55.5-84.1)]

and intravenous [80.8% (95%-CI; 69.7-91.8)] etoposide (p = 0.237). Even though there was

a trend towards improved objective response rate (ORR) [CR + PR] for intravenous [57.7%

(95%-CI; 43.8-71.6)] versus oral [37.2% (95%-CI; 22.2-52.3)] etoposide, the difference did

not achieve statistical significance (p = 0.064).

Response to treatment was assessed according to the Ki-67 value (dichotomised at <55%

vs ≥55%) for patients receiving CarboEtop as a first-line treatment. No differences in ORR

9

226

227

228

229

230

231

232

233

234

235

236

237

238

239

240

241

242

243

244

245

246

247

248

249

250

251

252

253

were identified; 50.0% (Ki-67 <55%) versus 50.0% (Ki-67 ≥55%); p = 1.000. Alternative Ki-

67 cut-offs were explored with the intention of using Ki-67 as a predictive factor of response

[defined as radiological response (CR+PR) or prolonged PFS (defined as PFS above 6

months)]. A Ki-67 value of 70% seemed to be the most informative cut-off, even though

performance was poor for both ORR (AUC 0.45 (95%-CI 0.33-0.57); sensitivity 51.2%,

specificity 36.2%) and prolonged-PFS (AUC 0.52 (95%-CI 0.39-0.65); sensitivity 55.0%,

specificity 42.5%).

4.4 Toxicity profile of CarboEtop

Toxicity data of CarboEtop was available for 122 (99.2%) courses of treatment. When all

courses/lines of treatment were pooled together, the most common grade 1-2 AEs were

fatigue (74.0%), nausea/vomiting (56.9%) and myelosuppression (58.2%), whereas the most

common grade 3-4 AEs included myelosuppression (49.2%), neutropenia (30.1%) and

anaemia (17.1%). There were no statistically significant differences in grade 3-4 AEs

between schedules with oral and intravenous etoposide, with the exception of grade 3-4

venous thromboembolism which was significantly more frequent among patients receiving

oral etoposide (p = 0.045). Full data is available in Table 3.

4.5 Efficacy outcomes of CarboEtop

At the time of data cut-off (30th April 2018), 83.7%, 94.3% and 91.3% of TTP, PFS and OS

events were available, respectively. The median PFS and TTP for patients receiving

CarboEtop were 5.8 (95%-CI; 4.8-7.0) and 6.2 months (95%-CI; 5.3-7.6) respectively: 6.0

(95%-CI; 5.0-7.1) and 6.7 months (95%-CI; 5.4-7.8) respectively, in first-line, and 4.5 (95%-

CI; 2.3-10.9) and 4.5 months (95%-CI; 2.4-10.9) respectively, in second/third line (Kaplan-

Meier curves illustrated in Figure 3A). The median OS for patients receiving CarboEtop was

11.6 months (95%-CI; 9.3-13.6), 11.5 months (95%-CI; 8.9-13.6) in first-line and 12.5

months (95%-CI; 5.5-23.4) in second/third-line (Kaplan-Meier curves illustrated in Figure

3D).

10

254

255

256

257

258

259

260

261

262

263

264

265

266

267

268

269

270

271

272

273

274

275

276

277

278

279

280

281

Univariate (univ) and multivariable (multiv) Cox-regression analyses for PFS and OS in the

first-line setting, according to baseline clinical-pathological characteristics and CarboEtop

schedules are presented in Table 4. The presence of liver metastases was the only factor

independently associated with worse PFS (multiv-Cox-p = 0.015) and OS (multiv-Cox-p =

0.002) (Kaplan-Meier curves illustrated in Figure 3B and 3E). Noticeably, the presence of

lung metastases correlated with shorter OS on multivariable analysis, even though findings

did not reach statistical significance (multiv-Cox-p = 0.051). No statistically significant

differences were found in PFS (univ-Cox p = 0.179) nor in OS (univ-Cox p = 0.069) between

schedules with oral and intravenous etoposide, even though a trend towards prolonged OS

was identified in favour of intravenous formulations (Kaplan-Meier curves illustrated in

Figure 3C and 3F).

4.6. CarboEtop re-challenge

Ten of the patients who received CarboEtop as first-line treatment had a re-challenge with

CarboEtop as second-line. All patients who were offered re-challenge with CarboEtop had a

first-line PFS >6 months [median PFS 11.5 months (95%-CI; 7.0-15.4)], which is suggestive

of sensitivity to platinum, and 90.0% had achieved radiological objective response (CR+PR)

in first-line. Median time from completion of first-line chemotherapy to initiation of re-

challenge with CarboEtop was 9.8 months (95% CI 6.5-21.6). Results obtained by re-

challenge with CarboEtop seemed less favourable than those achieved with the same

regimen in the first-line setting with regard to both median PFS [4.4 months (95%-CI; 1.6-

10.9); Log-rank-p = 0.566] and ORR [33.3%; p = 0.200].

5. Discussion

This study presents activity, efficacy and toxicity data from one of the largest pooled series

of patients with advanced EP-PD-NEC treated with CarboEtop in the published literature. In

11

282

283

284

285

286

287

288

289

290

291

292

293

294

295

296

297

298

299

300

301

302

303

304

305

306

307

308

309

the NORDIC study, 67 patients were treated with CarboEtop and a further 56 patients

received a triple-drug combination including carboplatin, etoposide and vincristine or other

drugs [6]. In two large cohorts of patients with PD or ‘highly aggressive’ EP-NEC, one

French and one Japanese, among patients receiving first-line palliative chemotherapy, those

treated with CarboEtop were 39 and 4, respectively [11, 12]. By far, only two other small

retrospective studies by Di Meglio et al. [18] and Imai et al. [19] have specifically looked at

treatment outcomes of CarboEtop in patients with advanced EP-NEC and included 20 and

19 patients, respectively.

Although results from other studies on the use of platinum-based chemotherapy in advanced

EP-PD-NEC are rather heterogeneous, due to differences in patient selection criteria,

treatment protocols and population size (ORR; 31-67%, DCR; 62-79%, median PFS/TTP; 4-

9 months, median OS; 7.3-20 months) [3, 4, 6-8, 11, 12, 18, 19, 27], first-line activity and

efficacy outcomes of CarboEtop in the present study were in line with data from the current

literature. In particular, first-line RRs were very close to those of CarboEtop reported in the

two studies by Di Meglio et al. [18] and Imai et al. [19], whereas survival outcomes closely

mirrored those of CarboEtop in the NORDIC study [6]. This suggests that, although

retrospective in nature, the results of the present study are reliable and provide further

support to the use of CarboEtop as a valid strategy for patients with advanced, previously

untreated EP-PD-NEC, with comparable outcomes to cisplatin/etoposide.

The formulation of etoposide, oral versus intravenous, did not significantly impact on

response rates nor on survival outcomes, even though there was a trend toward prolonged

OS and improved ORR in the intravenous etoposide subgroup, which should be explored

further in larger series. There are two possible explanations for this observation; 1) the

difference in ORR and OS might not have achieved statistical significance because of the

insufficient sample size of the two subgroups or 2) there is a still unidentified confounding

factor leading to imbalance between the two subgroups which misleadingly drives better

outcomes in the subgroup receiving intravenous etoposide. A recently published post-hoc

analysis of the NORDIC trial reported no statistically significant differences in PFS [oral-

12

310

311

312

313

314

315

316

317

318

319

320

321

322

323

324

325

326

327

328

329

330

331

332

333

334

335

336

337

etoposide; 5.4 months vs 24h-intravenous-etoposide; 3.8 months vs 5h-intravenous-

etoposide; 5.6 months] or OS [oral-etoposide; 11.3 months vs 24h-intravenous-etoposide;

14.5 months vs 5h-intravenous-etoposide; 11 months] between patients receiving oral and

intravenous etoposide in combination with cisplatin or carboplatin [27]. However, the large

disproportion in sample size between the oral-etoposide subgroup (33 patients) and the

intravenous-etoposide subgroup (203 patients) might have undermined the robustness of

these results. Furthermore, the proportion of patients receiving CarboEtop and activity data

of oral versus intravenous formulations of etoposide are not provided. Controversial results

were reported by Dorroh et al. on a large retrospective cohort of patients (n=300) with

advanced small cell lung cancer (SCLC) treated with oral versus or intravenous formulations

of etoposide in combination with a platinum derivative [28]. The study showed a statistically

significant 2.2 month OS advantage in favour the ‘oral-etoposide only’ subgroup compared

to the ‘any intravenous-etoposide’ subgroup. However, schedules and doses of

platinum/etoposide used were not specified, and the efficacy and pharmacokinetic of

platinum/etoposide chemotherapy may be different in patients with PD-NEC of lung origin.

The present study is the second in the published literature, to date, describing activity and

efficacy data of platinum-etoposide chemotherapy in the second-line setting of EP-PD-NEC.

The other study included 23 patients and reported a median PFS and OS of 1.9 months and

5 months, respectively [11]. The present study reports second-line CarboEtop outcomes

consistent with those of other second-line chemotherapy regimens used in clinical practice

for patients with advanced EP-PD-NEC [5-fluourouracil-based, temozolomide-based,

docetaxel-based, topotecan] (ORR; 0-31%, median PFS; 2.1-6 months, median OS; 3.2-22

months) [6, 12, 29-32], and indicates that CarboEtop is an active and effective option also in

the second-line setting, where appropriate. Unfortunately, when CarboEtop is used following

a first-line chemotherapy with the same regimen (re-challenge), the outcomes are less

encouraging than those achieved by the same patients in the first-line setting, despite

adequate patient selection based on PFS >6 months and achievement of PR or CR in first-

line.

13

338

339

340

341

342

343

344

345

346

347

348

349

350

351

352

353

354

355

356

357

358

359

360

361

362

363

364

365

Noteworthy, in the present study no comparisons were sought between first- and second-line

treatment outcomes of CarboEtop, as the small sample size of the second-line subgroup

could have affected the reliability of the results of such comparisons. A further potential bias

could be the enrichment of patients with better outcomes in the second-line subgroup, as

shown by the longer median OS compared to the first-line subgroup.

In the present study, correlations of survival outcomes with other patient-, tumour- and

treatment-related factors (univariate and multivariable analyses) were explored only in the

first-line setting to avoid potential biases related to the less favourable prognosis of patients

receiving second-line treatment, and to allow indirect comparisons with other studies in

advanced EP-PD-NEC, nearly all of which included only patients who did not have any

previous treatment [3, 4, 6-8, 11, 19].

The present study reports that patients with liver metastases have significantly poorer first-

line survival outcomes compared to patients with no liver involvement, which is in line with

the findings of a Japanese study including 41 patients with advanced GEP-NEC treated with

platinum-based chemotherapy [7]. Likewise, in a large multicentre European study of high-

grade GEP-NECs (313 patients) of any disease stage [14], the presence of liver metastases

was an independent negative prognostic factor for OS. Molecular/genetic studies and

comparisons with series of patients receiving first-line non-platinum-based chemotherapy or

best supportive care might help clarify whether liver metastatic disease has only a prognostic

significance or is suggestive of a tumour biology characterised by a reduced sensitivity to

platinum-based chemotherapy.

The lack of impact of the ECOG performance status on survival outcomes may be explained

by the disproportion in sample size between the comparator subgroups (ECOG≥2; 17.7%

versus ECOG=0-1; 80.7%), which may have hampered the emergence of potential

statistically significant differences. However, some degree of inconsistency in classifying

patient performance status according to the ECOG scale across clinicians or researchers

who have collected data in retrospect, especially in cases where the ECOG score had to be

estimated based on the clinical description, should be taken into consideration and regarded

14

366

367

368

369

370

371

372

373

374

375

376

377

378

379

380

381

382

383

384

385

386

387

388

389

390

391

392

393

to as a potential limitation of the study. In contrast with findings from the NORDIC study [6],

where a Ki-67 threshold of 55% was able to discriminate between poor and good responders

to platinum-based chemotherapy, as well as between shorter and longer survivors, no

significant differences in first-line RRs or survival outcomes were observed in the present

series of patients with EP-PD-NECs between cases with a Ki-67 <55% and ≥55%,

suggesting that CarboEtop is similarly active and effective in both subgroups. A

consideration worth mentioning is that, in the NORDIC trial the morphology was not a

selection criteria, therefore, the potential inclusion of well-differentiated G3-NENs, which

have a lower Ki-67 (usually towards the lowest extreme of the spectrum of G3-NENs), a

more favourable prognosis and a lower sensitivity to platinum-etoposide chemotherapy

compared to PD-G3-NECs [4], might have positively influenced the outcomes of the

subgroup with a Ki-67 <55%. In line with the findings of the present study, a large French

prospective cohort including 253 patients with EP-PD-NEC of all stages and specifically

excluding well-differentiated G3-NENs, showed that a Ki-67 of <55% versus ≥55% has no

prognostic significance [12].

Furthermore, despite efforts in identifying alternative cut-offs, in the present study the

performance of Ki-67 remains limited for predictive purposes. In other words, the role of Ki-

67 (either as a binomial or a continuous variable) as an isolated prognostic/predictive factor

for advanced EP-PD-NEC could not be confirmed. This corroborates the results of the

previously mentioned large retrospective study of patients with GEP-NECs [14], showing that

a risk-stratification score combining multiple prognostic factors has a stronger ability to

discriminate between better and worse outcomes, than Ki-67 alone.

A limitation of this study is that the tumour morphology and Ki-67 were not systematically

assessed at the time of study entry (although cases where the morphology was reported as

“well differentiated” were excluded), and a pathological review of the samples was pursued

only in selected cases. Therefore, an accurate estimation of the proportion of patients with

well differentiated neuroendocrine tumours as per 2017 World Health Organisation

15

394

395

396

397

398

399

400

401

402

403

404

405

406

407

408

409

410

411

412

413

414

415

416

417

418

419

420

Classification [25] (if any) could not be provided, and a comparison of treatment outcomes

between the poorly and well differentiated subgroups could not be pursued.

In keeping with other studies of CarboEtop in lung-NEC and EP-NEC [17-19], grade 3-4 AEs

observed in the present study were most commonly haematological, with neutropenia being

the most frequent, whereas non-haematological grade 3-4 AEs occurred at a low frequency,

consistent with an overall manageable toxicity profile. However, consideration should be

given to the possible inconsistency and incompleteness in reporting AEs, which is inherent in

the retrospective nature of the study.

Noticeably, dose-intensity data of CarboEtop were missing for 33% of courses of treatment;

this represents an attrition bias due to the fact that information on doses of chemotherapy

could be retrieved only for patients with a more recent diagnosis (in contrast with toxicity

data which could be extracted from medical records for nearly all patients).

It has also to be acknowledged that, as the majority of patients who received CarboEtop as

second-line treatment had already been treated with the same regimen in first-line (10 out of

16 patients), in order to increase the sample size of and gather as much information as

possible on the second-line subgroup, survival and toxicity outcomes of CarboEtop were

calculated for courses of chemotherapy instead of for individual patients, and this might have

introduced potential biases.

Finally, despite the effort to retrieve information from all consecutive patients over a 22 year

period, potential selection biases (e.g. missed inclusion of eligible patients) related to the

retrospective design of the study cannot be completely excluded. In conclusion, the present

study provides further corroboration for the use of CarboEtop as an active, effective and

well-tolerated alternative to other platinum-based chemotherapy regimens for the treatment

of advanced EP-PD-NEC, both in first- and second-line. It also suggests that oral etoposide

can be substituted for intravenous etoposide, in order to minimise the frequency and length

of patient attendance to hospital, advocating patient preferences and potentially cutting

down on health economic-related expenses for nursing staff, materials and facilities, with no

or minimal impact on survival or additional toxicities. To investigate whether the differences

16

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

observed in treatment outcomes in patients receiving schedules using oral and intravenous

etoposide are only due to bias related to the retrospective nature of the data collection, or

reflect actual discrepancies in the activity and efficacy of the two schedules, participation by

other centres, or alternatively a pooled/meta-analysis including other studies investigating

both formulations of etoposide, will be sought to further increase the study population size,

and allow more reliable conclusions.

6. Statements

6.1 Statement of Ethics

This study was conducted ethically in accordance with the World Medical Association

Declaration of Helsinki, and was approved by The Christie NHS Foundation Trust Audit

committee (SE17/1992) and the local audit committee of The European Institution of

Oncology. Informed signed consent from individual patients was not required to be obtained.

6.2 Disclosure Statement

The authors have declared no conflict of interest.

6.3 . Funding Sources

This study was partially funded by the American Society of Clinical Oncology (ASCO)

Conquer Cancer Foundation Young Investigator Award.

6.4 Author contribution

Melissa Frizziero, Francesca Spada; identification of eligible patients, data collection, and

manuscript writing.

Angela Lamarca; idea conception and study design, identification of eligible patients,

statistical analysis, review and proof-reading, and approval of the final version.

17

449

450

451

452

453

454

455

456

457

458

459

460

461

462

463

464

465

466

467

468

469

470

471

472

473

474

475

476

Nicola Fazio, Juan W Valle; idea conception and study design, review and proof-reading,

and approval of the final version.

Zoe Kordatou, Jorge Barriuso, Christina Nuttall, Mairéad G McNamara, Richard A

Hubner, Wasat Mansoor, Prakash Manoharan; review, proof-reading, and approval of the

final version.

7. References1 Bosman T CF, Hruban RH, Theise ND: WHO Classification of Tumours of the Digestive

System. 4th ed. Lyon, IARC press, 2010.

2 Dasari A, Shen C, Halperin D, Zhao B, Zhou S, Xu Y, Shih T, Yao JC: Trends in the

Incidence, Prevalence, and Survival Outcomes in Patients With Neuroendocrine Tumors in the United

States. JAMA Oncol 2017;3:1335-1342.

3 Moertel CG, Kvols LK, O'Connell MJ, Rubin J: Treatment of neuroendocrine carcinomas with

combined etoposide and cisplatin. Evidence of major therapeutic activity in the anaplastic variants of

these neoplasms. Cancer 1991;68:227-232.

4 Heetfeld M, Chougnet CN, Olsen IH, Rinke A, Borbath I, Crespo G, Barriuso J, Pavel M,

O'Toole D, Walter T, other Knowledge Network m: Characteristics and treatment of patients with G3

gastroenteropancreatic neuroendocrine neoplasms. Endocr Relat Cancer 2015;22:657-664.

5 Sorbye H, Strosberg J, Baudin E, Klimstra DS, Yao JC: Gastroenteropancreatic high-grade

neuroendocrine carcinoma. Cancer 2014;120:2814-2823.

6 Sorbye H, Welin S, Langer SW, Vestermark LW, Holt N, Osterlund P, Dueland S, Hofsli E,

Guren MG, Ohrling K, Birkemeyer E, Thiis-Evensen E, Biagini M, Gronbaek H, Soveri LM, Olsen IH,

Federspiel B, Assmus J, Janson ET, Knigge U: Predictive and prognostic factors for treatment and

survival in 305 patients with advanced gastrointestinal neuroendocrine carcinoma (WHO G3): the

NORDIC NEC study. Ann Oncol 2013;24:152-160.

7 Terashima T, Morizane C, Hiraoka N, Tsuda H, Tamura T, Shimada Y, Kaneko S, Kushima

R, Ueno H, Kondo S, Ikeda M, Okusaka T: Comparison of chemotherapeutic treatment outcomes of

advanced extrapulmonary neuroendocrine carcinomas and advanced small-cell lung carcinoma.

Neuroendocrinology 2012;96:324-332.

8 Mitry E, Baudin E, Ducreux M, Sabourin JC, Rufie P, Aparicio T, Aparicio T, Lasser P, Elias

D, Duvillard P, Schlumberger M, Rougier P: Treatment of poorly differentiated neuroendocrine

tumours with etoposide and cisplatin. Br J Cancer 1999;81:1351-1355.

9 Hainsworth JD, Spigel DR, Litchy S, Greco FA: Phase II trial of paclitaxel, carboplatin, and

etoposide in advanced poorly differentiated neuroendocrine carcinoma: a Minnie Pearl Cancer

Research Network Study. J Clin Oncol 2006;24:3548-3554.

18

477

478

479

480

481

482

483

484

485

486

487

488

489

490

491

492

493

494

495

496

497

498

499

500

501

502

503

504

505

506

507

508

509

510

511

512

10 Garcia-Carbonero R, Capdevila J, Crespo-Herrero G, Diaz-Perez JA, Martinez Del Prado MP,

Alonso Orduna V, Sevilla-Garcia I, Villabona-Artero C, Beguiristain-Gomez A, Llanos-Munoz M,

Marazuela M, Alvarez-Escola C, Castellano D, Vilar E, Jimenez-Fonseca P, Teule A, Sastre-Valera J,

Benavent-Vinuelas M, Monleon A, Salazar R: Incidence, patterns of care and prognostic factors for

outcome of gastroenteropancreatic neuroendocrine tumors (GEP-NETs): results from the National

Cancer Registry of Spain (RGETNE). Ann Oncol 2010;21:1794-1803.

11 Yamaguchi T, Machida N, Morizane C, Kasuga A, Takahashi H, Sudo K, Nishina T,

Tobimatsu K, Ishido K, Furuse J, Boku N, Okusaka T: Multicenter retrospective analysis of systemic

chemotherapy for advanced neuroendocrine carcinoma of the digestive system. Cancer Sci

2014;105:1176-1181.

12 Walter T, Tougeron D, Baudin E, Le Malicot K, Lecomte T, Malka D, Hentic O, Manfredi S,

Bonnet I, Guimbaud R, Coriat R, Lepere C, Desauw C, Thirot-Bidault A, Dahan L, Roquin G, Aparicio

T, Legoux JL, Lombard-Bohas C, Scoazec JY, Lepage C, Cadiot G, investigators C: Poorly

differentiated gastro-entero-pancreatic neuroendocrine carcinomas: Are they really heterogeneous?

Insights from the FFCD-GTE national cohort. Eur J Cancer 2017;79:158-165.

13 Fazio N, Milione M: Heterogeneity of grade 3 gastroenteropancreatic neuroendocrine

carcinomas: New insights and treatment implications. Cancer Treat Rev 2016;50:61-67.

14 Lamarca A, Walter T, Pavel M, Borbath I, Freis P, Nunez B, Childs A, McNamara MG, Hubner

RA, Garcia-Carbonero R, Meyer T, Valle JW, Barriuso J: Design and Validation of the GI-NEC Score

to Prognosticate Overall Survival in Patients With High-Grade Gastrointestinal Neuroendocrine

Carcinomas. J Natl Cancer Inst 2017;109

15 Garcia-Carbonero R, Sorbye H, Baudin E, Raymond E, Wiedenmann B, Niederle B,

Sedlackova E, Toumpanakis C, Anlauf M, Cwikla JB, Caplin M, O'Toole D, Perren A, Vienna

Consensus Conference p: ENETS Consensus Guidelines for High-Grade Gastroenteropancreatic

Neuroendocrine Tumors and Neuroendocrine Carcinomas. Neuroendocrinology 2016;103:186-194.

16 Fazio N, Spada F, Giovannini M: Chemotherapy in gastroenteropancreatic (GEP)

neuroendocrine carcinomas (NEC): a critical view. Cancer Treat Rev 2013;39:270-274.

17 Rossi A, Di Maio M, Chiodini P, Rudd RM, Okamoto H, Skarlos DV, Fruh M, Qian W, Tamura

T, Samantas E, Shibata T, Perrone F, Gallo C, Gridelli C, Martelli O, Lee SM: Carboplatin- or

cisplatin-based chemotherapy in first-line treatment of small-cell lung cancer: the COCIS meta-

analysis of individual patient data. J Clin Oncol 2012;30:1692-1698.

18 Di Meglio G, Massacesi C, Radice D: Carboplatin with etoposide in patients with

extrapulmonary "aggressive" neuroendocrine carcinoma. J Clin Oncol 2010;Suppl:e13072,

conference abstract.

19 Imai H, Shirota H, Okita A, Komine K, Saijo K, Takahashi M, Takahashi S, Takahashi M,

Shimodaira H, Ishioka C: Efficacy and Safety of Carboplatin and Etoposide Combination

Chemotherapy for Extrapulmonary Neuroendocrine Carcinoma: A Retrospective Case Series.

Chemotherapy 2016;61:111-116.

19

513

514

515

516

517

518

519

520

521

522

523

524

525

526

527

528

529

530

531

532

533

534

535

536

537

538

539

540

541

542

543

544

545

546

547

548

549

550

20 Eek D, Krohe M, Mazar I, Horsfield A, Pompilus F, Friebe R, Shields AL: Patient-reported

preferences for oral versus intravenous administration for the treatment of cancer: a review of the

literature. Patient Prefer Adherence 2016;10:1609-1621.

21 Rezonja R, Knez L, Cufer T, Mrhar A: Oral treatment with etoposide in small cell lung cancer -

dilemmas and solutions. Radiol Oncol 2013;47:1-13.

22 Hande K, Messenger M, Wagner J, Krozely M, Kaul S: Inter- and intrapatient variability in

etoposide kinetics with oral and intravenous drug administration. Clin Cancer Res 1999;5:2742-2747.

23 Toffoli G, Corona G, Sorio R, Robieux I, Basso B, Colussi AM, Boiocchi M: Population

pharmacokinetics and pharmacodynamics of oral etoposide. Br J Clin Pharmacol 2001;52:511-519.

24 Kroschinsky FP, Friedrichsen K, Mueller J, Pursche S, Haenel M, Prondzinsky R, Ehninger G,

Schleyer E: Pharmacokinetic comparison of oral and intravenous etoposide in patients treated with

the CHOEP-regimen for malignant lymphomas. Cancer Chemother Pharmacol 2008;61:785-790.

25 Lloyd RV OR, Klöppel G, Rosai J. Who Classification of Tumours of Endocrine Organs. 4th

ed. Lyon, IARC press, 2017.

26 Eisenhauer EA, Therasse P, Bogaerts J, Schwartz LH, Sargent D, Ford R, Dancey J, Arbuck

S, Gwyther S, Mooney M, Rubinstein L, Shankar L, Dodd L, Kaplan R, Lacombe D, Verweij J: New

response evaluation criteria in solid tumours: revised RECIST guideline (version 1.1). Eur J Cancer

2009;45:228-247.

27 Ali AS, Gronberg M, Langer SW, Ladekarl M, Hjortland GO, Vestermark LW, Osterlund P,

Welin S, Gronbaek H, Knigge U, Sorbye H, Janson ET: Intravenous versus oral etoposide: efficacy

and correlation to clinical outcome in patients with high-grade metastatic gastroenteropancreatic

neuroendocrine neoplasms (WHO G3). Med Oncol 2018;35:47.

28 Dorroh AS, Siegle ER, Govindarajan R. Oral (O) versus intravenous (IV) etoposide and

platinum in the treatment of extensive-stage small cell lung cancer (SCLC). Journal of Clinical

Oncology 2013 31:15_suppl, 7597-7597 2013.

29 Hentic O, Hammel P, Couvelard A, Rebours V, Zappa M, Palazzo M, Maire F, Goujon G,

Gillet A, Levy P, Ruszniewski P: FOLFIRI regimen: an effective second-line chemotherapy after failure

of etoposide-platinum combination in patients with neuroendocrine carcinomas grade 3. Endocr Relat

Cancer 2012;19:751-757.

30 Hadoux J, Malka D, Planchard D, Scoazec JY, Caramella C, Guigay J, Boige V, Leboulleux

S, Burtin P, Berdelou A, Loriot Y, Duvillard P, Chougnet CN, Deandreis D, Schlumberger M, Borget I,

Ducreux M, Baudin E: Post-first-line FOLFOX chemotherapy for grade 3 neuroendocrine carcinoma.

Endocr Relat Cancer 2015;22:289-298.

31 Welin S, Sorbye H, Sebjornsen S, Knappskog S, Busch C, Oberg K: Clinical effect of

temozolomide-based chemotherapy in poorly differentiated endocrine carcinoma after progression on

first-line chemotherapy. Cancer 2011;117:4617-4622.

32 Olsen IH, Knigge U, Federspiel B, Hansen CP, Skov A, Kjaer A, Langer SW: Topotecan

monotherapy in heavily pretreated patients with progressive advanced stage neuroendocrine

carcinomas. J Cancer 2014;5:628-632.

20

551

552

553

554

555

556

557

558

559

560

561

562

563

564

565

566

567

568

569

570

571

572

573

574

575

576

577

578

579

580

581

582

583

584

585

586

587

588

589

590

21

02

46

810

1214

1618

Per

cent

(%)

30 40 50 60 70 80 90 100

Ki-67 (%)

Figure 1. Frequency distribution of Ki-67 values assessed on diagnostic pathological samples

591

22

592593

23

594595596

Characteristic Variable N % Characteristic Variable N %

Gender Distant met

Female 41 36.3% Yes 102 90.3%

Male 72 63.7% No 11 9.7%

ECOG PS Number of met sites

0 9 8.0% Median 2

1 82 72.6% Range 1-6

≥2 20 17.7%

N.A. 2 1.8%

ACE-27 score Commonest met sites (>1%)

0 71 62.8% Distant lymph nodes 58 56.7%

1 13 11.5% Liver 54 52.9%

≥2 6 5.3% Bones 21 20.6%

N.A. 3 2.6% Lung 20 19.6%

Peritoneum/omentum 12 11.8%

PT site Adrenal gland 7 6.9%

Foregut 27 23.9% Subcutaneous tissue 4 3.9%

Pancreas 11 9.7% Brain 2 2.0%

Hindgut 19 16.8%

UKP 26 23.0% Ki-67

Bladder 10 8.8% Median 70

Biliary tract 4 3.5% Range 30-100

Cervix/Ovary 7 6.2% 95%-CI 60-80

Prostate 7 6.2% Mean (+/-SD) 67.8 (+/- 20.0)

Others 2 1.8% <55 26 23.0%

≥55 67 59.3%

G.E.P. 61 54.0% N.A. 20 17.7%

G.U. 24 21.2%

Morphological subtype NSE (ug/L)*

Small cell 55 48.7% Median 19.6

Large cell 20 17.7% Range 6.4-122.6

Others 2 1.8% 95%-CI 13.1-41.4

N.A. 36 31.9%

24

Table 1. Baseline characteristics of patients with advanced extra-pulmonary, poorly differentiated NEC and their tumours

597

598

599

600

601

602

Outcome Variable Any lineN=111 (90.2%)

First-lineN=98 (89.9%)

Second/third lineN=13 (92.9%)

Response*[number of courses (%)]

CR 9 (8.1%) 7 (7.1%) 2 (15.4%)

PR 42 (37.8%) 40 (40.8%) 2 (15.4%)

SD 31 (27.9%) 26 (26.5%) 5 (38.5%)

PD 29 (26.1%) 25 (25.5%) 4 (30.8%)

DCR 82 (73.43%) 73 (74.5%) 9 (69.2%)

ORR 51 (45.9%) 47 (47.9%) 4 (30.8%)

Median time to BR[months]

All BR 2.86 (95%CI; 2.5-3.0) 2.87 (95%CI; 2.5-3.0) 2.71 (95%CI; 2.1-4.5)

if CR 3.02 (95%CI; 1.9-4.4) 3.02 (95%CI; 1.8-4.6) 3.47 (95%CI; 2.5-4.4)

if PR 3.04 (95%CI; 2.5-4.2) 3.04 (95%CI; 2.6-4.1) 2.95 (95%CI; 1.2-4.7)

if SD 2.87 (95%CI; 2.5-3.3) 2.86 (95%CI; 2.4-3.3) 3.01 (95%CI; 1.6-4.6)

if PD 2.45 (95%CI; 2.3-2.7) 2.45 (95%CI; 2.3-2.9) 2.38 (95%CI; 1.9-14.1)

if CR/PR 3.04 (95%CI; 2.6-3.8) 3.04 (95%CI; 2.6-3.7) 3.47 (95%CI; 1.2-4.7)

Median variation**[%]

Any lineN=64 (77.1%)

First-lineN=52 (75.4%)

Second/third lineN=12 (85.7%)

All BR -19.1 (95%CI; -41.1- -0.8) -27.9 (95%CI; -43.3- -0.512) -15.2 (95%CI; -76.7- 14.8)

if PR -48.4 (95%CI; -58.4- -44.0) -46.5 (95%CI; -58.0- -43.6) -64.6 (95%CI; -80.0- -49.2)

if SD -7.3 (95%CI; -16.8-4.3) -7.3 (95%CI; -18.8-4.1) -4.7 (95%CI; -21.2-17.7)

if PD +26.9 (95%CI; 15.0-37.9) +28.8 (95%CI; 15.2-42.1) +8.8 (95%CI;-13.9-68.8)

PD due to new lesions 8 5 3

25

Table 2. Activity data of CarboEtop in the first-line treatment of patients with advanced extra-pulmonary, poorly differentiated NEC

NEC = neuroendocrine carcinoma; CarboEtop = carboplatin/etoposide chemotherapy; N = number of courses of CarboEtop; * = response evaluated as per RECIST 1.1. criteria; ** = percentage of variation in the sum of the sizes of “marker lesions” as defined by RECIST 1.1 criteria. CR = complete response; PR = partial response; SD = stable disease; PD = progressive disease; DCR = disease control rate (CR + PR + SD); ORR = objective response rate (CR + PR); BR = best response; 95%CI = 95% Confidence Interval.

603

604

605

606

607

608

609

610

26

611612

613

Univariate Cox Regression for PFS

Multivariable Cox Regression for PFS

Univariate Cox Regression for OS

Multivariable Cox Regression for OS

Variable Category HR (95%-CI) p-value HR (95%-CI) p-value HR (95%-CI) p-value HR (95%-CI) p-value

Age (start of CarboEtop)

Continuous variable 1.01 (0.98-1.02) 0.804 1.01 (0.99-1.02) 0.541

Gender Male (vs Female) 1.04 (0.69-1.57) 0.837 1.04 (0.69-1.57) 0.848

Morphology Small cell 1 (Ref) - 1 (Ref) -

Large cell 1.18 (0.69-2.02) 0.542 1.31 (0.76-2.28) 0.336

Others 1.09 (0.15-8.07) 0.928 2.73 (0.64-11.70) 0.175

Ki-67 Continuous variable 1.01 (0.99-1.02) 0.511 1.01 (0.99-1.02) 0.517

Ki-67 ≥55% (vs <55%) 1.29 (0.78-2.11) 0.317 0.99 (0.60-1.62) 0.958

ECOG PS >2 (vs 0/1) 1.35 (0.79-2.28) 0.265 1.47 (0.86-2.50) 0.156

ACE-27 score (binary) None/Mild 1 (Ref) - 1 (Ref) -

Moderate/Severe 0.96 (0.57-1.63) 0.886 1.39 (0.81-2.36) 0.229

PT site GEP/UKP 1 (Ref) - 1 (Ref) -

Others 1.05 (0.65-1.69) 0.852 1.04 (0.65-1.67) 0.870

Previous curative resection Yes (vs. No) 1.16 (0.47-2.87) 0.750 0.97 (0.35-2.66) 0.955

Baseline NSE (ug/L) Continuous variable 1.01 (0.99-1.03) 0.083 1.01 (0.99-1.03) 0.263

Stage IV* Yes (vs. No) 1.8 (0.93-3.48 0.080 1.59 (0.62-3.08) 0.164

Presence of distant met* Yes (vs. No) 2.28 (1.05-4.99) 0.038 1.47 (0.62-3.48) 0.387 1.51 (0.72-3.13) 0.273

N distant met sites Continuous variable 1.24 (1.07-1.44) 0.005 1.14 (0.94-1.38) 0.180 1.16 (0.98-1.37) 0.094

Liver Yes (vs. No) 2.01 (1.33-3.03) 0.001 1.71 (1.11-2.63) 0.015 2.01 (1.32-3.06) 0.001 1.88 (1.28-2.76) 0.002

Lung Yes (vs. No) 1.65 (0.98-2.77) 0.059 1.80 (1.05-3.09) 0.032 1.72 (1.00-2.95) 0.050

Distant lymph nodes Yes (vs. No) 1.14 (0.77-1.71) 0.511 0.72 (0.48-1.09) 0.122

Peritoneum/omentum Yes (vs. No) 1.02 (0.67-1.83) 0.950 0.94 (0.49-1.76) 0.837

CarboEtop schedule IV Etop (vs. Oral Etop) 0.76 (0.51-1.14) 0.179 0.68 (0.45-1.03) 0.069

CarboEtop dose intensity

Continuous variable 0.99 (0.98-1.02) 0.978 0.99 (0.97-1.01) 0.232

27

Table 4. Univariate and multivariable analyses of factors effecting PFS and OS in patients with extra-pulmonary, poorly differentiated NEC treated with CarboEtop in the first-line setting

614

615

28