Imaoka et al.

BMC Cancer (2020) 20:946

https://doi.org/10.1186/s12885-020-07462-4

RESEARCH ARTICLE Open Access

Clinical outcomes of chemotherapy in

patients with undifferentiated carcinoma of
the pancreas: a retrospective multicenter
cohort study
Hiroshi Imaoka1* , Masafumi Ikeda1, Kosuke Maehara2, Kumiko Umemoto3, Masato Ozaka4, Satoshi Kobayashi5,
Takeshi Terashima6, Hiroto Inoue7, Chihiro Sakaguchi8, Kunihiro Tsuji9, Kazuhiko Shioji10, Keiya Okamura11,
Yasuyuki Kawamoto12, Rei Suzuki13, Hirofumi Shirakawa14, Hiroaki Nagano15, Makoto Ueno5,
Chigusa Morizane2 and Junji Furuse16

Abstract
Background: Undifferentiated carcinoma (UC) of the pancreas is a rare subtype of pancreatic cancer. Although UC
has been considered a highly aggressive malignancy, no clinical studies have addressed the efficacy of
chemotherapy for unresectable UC. Therefore, we conducted multicenter retrospective study to investigate the
efficacy of chemotherapy in patients with UC of the pancreas.
Methods: This multicenter retrospective cohort study was conducted at 17 institutions in Japan between January
2007 and December 2017. A total of 50 patients treated with chemotherapy were analyzed.
Results: The median overall survival (OS) in UC patients treated with chemotherapy was 4.08 months. The details of
first-line chemotherapy were as follows: gemcitabine (n = 24), S-1 (n = 12), gemcitabine plus nab-paclitaxel (n = 6),
and other treatment (n = 8). The median progression-free survival (PFS) was 1.61 months in the gemcitabine group,
2.96 months in the S-1 group, and 4.60 months in the gemcitabine plus nab-paclitaxel group. Gemcitabine plus
nab-paclitaxel significantly improved PFS compared with gemcitabine (p = 0.014). The objective response rate (ORR)
was 4.2% in the gemcitabine group, 0.0% in the S-1 group, and 33.3% in the gemcitabine plus nab-paclitaxel group.
Gemcitabine plus nab-paclitaxel also showed a significantly higher ORR compared with both gemcitabine and S-1
(gemcitabine plus nab-paclitaxel vs. gemcitabine: p = 0.033; gemcitabine plus nab-paclitaxel vs. S-1: p = 0.034). A
paclitaxel-containing first-line regimen significantly improved OS compared with a non-paclitaxel-containing
regimen (6.94 months vs. 3.75 months, respectively; p = 0.041). After adjustment, use of a paclitaxel-containing
regimen in any line was still an independent predictor of OS (hazard ratio for OS, 0.221; 95% confidence interval,
0.076–0.647; p = 0.006) in multiple imputation by chained equation.
(Continued on next page)

* Correspondence: [email protected]
1
Department of Hepatobiliary and Pancreatic Oncology, National Cancer
Center Hospital East, 6-5-1, Kashiwanoha, Kashiwa, Chiba 277-8577, Japan
Full list of author information is available at the end of the article

© The Author(s). 2020 Open Access This article is licensed under a Creative Commons Attribution 4.0 International License,
which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give
appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if
changes were made. The images or other third party material in this article are included in the article's Creative Commons
licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons
licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain
permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.
The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the
data made available in this article, unless otherwise stated in a credit line to the data.
Imaoka et al. BMC Cancer (2020) 20:946 Page 2 of 11

(Continued from previous page)

Conclusions: The results of the present study indicate that a paclitaxel-containing regimen would offer relatively
longer survival, and it is considered a reasonable option for treating patients with unresectable UC.
Keywords: Undifferentiated carcinoma, Pancreatic cancer, Anaplastic carcinoma, Chemotherapy, Osteoclast-like
giant cells, Paclitaxel, Gemcitabine, Predictor

Background (Fig. 1). The inclusion criteria were histopathologically

Pancreatic cancer (PC) is one of the deadliest cancers and diagnosed UC of the pancreas (including UC with
the fourth leading cause of cancer death in the United States. osteoclast-like giant cells [UC-OGCs]), recurrent/meta-
It has been estimated that, in 2020, approximately 47,050 pa- static or locally advanced disease, and treated with
tients will die of this disease [1]. Global data showed that chemotherapy. The study protocol was approved by the
448,000 patients were diagnosed with PC in 2017, with a 2.3- institutional review boards of the participating
times increase in the number of incident cases and deaths institutions.
from 1990 to 2017 [2]. Despite efforts to develop new treat-
ments [3, 4], the prognosis of PC patients remains poor, and Patient evaluation
the morbidity is increasing. Undifferentiated carcinoma (UC) Data regarding clinical and laboratory features, histo-
of the pancreas, also known as anaplastic carcinoma of the logical findings, treatment, and outcome measures were
pancreas, is a rare subtype of PC and accounts for 0.3–7% of collected retrospectively. Histological findings were
malignant neoplasms of the pancreas [5–7]. UC is an epithe- based on pathology reports and classified into the fol-
lial neoplasm displaying no particular differentiation such as lowing subtypes: anaplastic type, sarcomatoid type, carci-
glandular formation, mucin production, or keratinization. nosarcoma, UC-OGCs, and not otherwise specified
One population-based study reported that the median age (NOS) [11].
at diagnosis of UC was 67 years, with a slight male predom- OS was measured from the date of start of first-line
inance (57.5%) [6]. These characteristics are similar to those treatment to the date of death from any cause. OS for
of PC. On the other hand, UC of the pancreas has been con- patients who were lost to follow-up was censored at the
sidered more aggressive than PC, and median overall survival last date they were known to be alive. PFS was measured
(OS) does not exceed 6 months [6, 7]. Clark et al. reported a from the date of start of treatment to the date of first
population-based study comparing patients with UC and PC documented disease progression or the date of death
[6]. The median OS was 11 months in the PC group and 3 from any cause. PFS was censored at the time of the last
months in the UC group, and it was significantly shorter in follow-up if there was no documentation of disease pro-
the UC group than in the PC group (hazard ratio [HR], 1.9; gression or death. Tumor response was based on the
95% confidence interval [CI], 1.7–2.1). Paal et al. reported 35 best overall response throughout the entire course of the
patients with UC, and their median OS was 5.2 months [8]. observation period. The Response Evaluation Criteria in
However, these data were based primarily on surgical series Solid Tumors (RECIST) version 1.1 were used to assess
or registry data. Considering that it has been reported that tumor responses [12].
approximately 80% of PC patients were diagnosed at unre-
sectable stages [9, 10], the majority of UC cases potentially
have metastases. However, previous reports have rarely men- Statistical analysis
tioned unresectable stage disease, and clinical and treatment Univariate analysis was performed using the chi-squared
data for patients with unresectable UC are lacking. Thus, test for categorical variables. The Kaplan-Meier method
identification of effective chemotherapy regimens for UC is was used to estimate the time-to-event distribution, and
crucial for improving the prognosis of patients with UC. p-values were calculated using the log-rank test. HRs
Therefore, clinical and treatment data of patients with were calculated using the Cox proportional hazards
unresectable UC were retrospectively collected. The aim of model. Values of p < 0.05 were considered statistically
this multicenter retrospective cohort study was to investigate significant, and all p-values are two-sided. To analyze
the efficacy of chemotherapy in patients with UC of the predictors of OS in patients with unresectable UC, a
pancreas. multivariate Cox proportional hazards model was used,
including predictors (p < 0.10) on univariate analysis and
clinically relevant variables (Eastern Cooperative Oncol-
Methods ogy Group performance status [ECOG PS], extent of dis-
Study design ease, and histological subtype). Due to the retrospective
This retrospective study was conducted at 17 institutions analysis, covariate data were often missing. Thus, mul-
in Japan between January 2007 and December 2017 tiple imputation was performed by multiple imputation
Imaoka et al. BMC Cancer (2020) 20:946 Page 3 of 11

Fig. 1 Selection of patients for the study

by chained equation (MICE) to avoid potential bias [13]. Overall treatment efficacy
A Cox proportional hazards model was performed in For all patients treated with chemotherapy, the median
complete-case analysis and MICE. Data were analyzed OS was 4.08 months, and the 12-month OS rate was
using STATA version 15.1 (StataCorp, College Station, 16.3% (Fig. 2a). The median PFS in patients receiving
TX, USA) and R version 3.6.1 (http://www.r-project.org/). first-line treatment and in those receiving second-line
treatment was 1.84 months and 3.19 months, respectively
Results (Fig. 2b). For the cumulative total of 65 treatments, the
Patient characteristics median PFS was 2.01 months (Fig. 2c). The individual
A cumulative total of 65 treatments were given to 50 UC PFS for each is shown in Fig. 3, and their objective re-
patients between January 2007 and December 2017. The sponse rate (ORR) was 10.8%.
baseline characteristics of the patients with UC are shown
in Table 1. ECOG PS was 0 in 13 patients (26.0%), 1 in 31 Efficacy of first-line treatment
patients (62.0%), and ≥ 2 in 4 patients (8.0%). The median The details of chemotherapy in first-line treatment were
treatment line was 1 (range 1–3). A total of 13 patients re- as follows: gemcitabine (n = 24), S-1 (n = 12), gemcita-
ceived second-line treatment, and 2 patients received bine plus nab-paclitaxel (n = 6), and other treatment
third-line treatment. The details of chemotherapy in any (n = 8). The median OS in the first-line gemcitabine
treatment line were as follows: gemcitabine (n = 27), S-1 group, S-1 group, and gemcitabine plus nab-paclitaxel
(n = 18), gemcitabine plus nab-paclitaxel (n = 9), FOLFIRI- group was 2.70 months, 8.16 months, and 6.77 months,
NOX (n = 4), gemcitabine plus S-1 (n = 2), paclitaxel (n = respectively (Fig. 4a). The median PFS in the first-line
1), and other treatment (n = 4). gemcitabine group, S-1 group, and gemcitabine plus
Imaoka et al. BMC Cancer (2020) 20:946 Page 4 of 11

Table 1 Patient baseline characteristics

All patients
(n = 50) Missing
Sex
Male (%) 34 (68.0)
Female (%) 16 (32.0)
Age (y)
Median (range) 69 (41–83)
ECOG PS
0 (%) 13 (26.0) 2 (4.0)
1 (%) 31 (62.0)
≥ 2 (%) 4 (8.0)
Prior surgical resection 25 (50.0)
Tumor location
Head (%) 24 (48.0)
Body-Tail (%) 26 (52.0)
Tumor size (cm)
Median (range) 4.5 (2.0–18.0) 1 (2.0)
Extent of disease
Locally advanced (%) 6 (12.0)
Metastatic (%) 44 (88.0)
Measurable metastatic sites
Liver (%) 26 (52.0)
Lymph node (%) 20 (40.0)
Lung (%) 4 (8.0)
Peritoneal (%) 12 (24.0)
LDH, U/L
Median (range) 205 (128–909) 2 (4.0)
CRP, mg/L
Median (range) 13 (0–178) 2 (4.0)
CEA, ng/mL
Median (range) 3.0 (0.7–64.1) 1 (2.0)
CA19–9, U/mL
Median (range) 35.7 (1.0–43,645) 1 (2.0)
Histological subtype
Anaplastic type (%) 16 (32.0)
Sarcomatoid type (%) 4 (8.0)
Undifferentiated carcinoma with OGCs (%) 11 (22.0)
NOS (%) 19 (38.0)
ECOG PS Eastern Cooperative Oncology Group performance status, LDH lactate dehydrogenase, CRP C-reactive protein, CEA carcinoembryonic antigen, CA19–9
carbohydrate antigen 19–9, OGCs osteoclast-like giant cells, NOS not otherwise specified

nab-paclitaxel group was 1.61 months, 2.96 months, plus nab-paclitaxel significantly improved PFS com-
and 4.60 months, respectively (Fig. 4b). Tumor re- pared with gemcitabine (p = 0.014), and it showed
sponses in first-line treatment are shown in Table 2. significantly higher ORR compared with both gemci-
There was no significant difference in OS among tabine and S-1 (gemcitabine plus nab-paclitaxel vs.
UC patients treated with gemcitabine, S-1, and gem- gemcitabine: p = 0.033; gemcitabine plus nab-
citabine plus nab-paclitaxel. However, gemcitabine paclitaxel vs. S-1: p = 0.034).
Imaoka et al. BMC Cancer (2020) 20:946 Page 5 of 11

Fig. 2 Kaplan-Meier curves of overall survival for all patients with undifferentiated carcinoma of the pancreas (a), progression-free survival with
first-line and second-line treatments (b), and progression-free survival for the cumulative total of 65 treatments (c). OS, overall survival; PFS,
progression-free survival

Efficacy of paclitaxel-containing regimens 0.004) (Fig. 5b). ORR was significantly higher with a
Two different paclitaxel-containing regimens (gemcita- paclitaxel-containing regimen than with a non-
bine plus nab-paclitaxel (n = 6) and paclitaxel monother- paclitaxel-containing regimen (40.0% vs. 5.5%, respect-
apy (n = 1)) were used as first-line treatment. A ively; p = 0.001). In addition, one complete response was
paclitaxel-containing first-line regimen significantly im- observed in a patient treated with paclitaxel monother-
proved OS compared with a non-paclitaxel-containing apy, and this patient achieved long survival.
regimen (6.94 months vs. 3.75 months, respectively; p = The results of the Cox proportional hazards model for
0.041) (Fig. 5a). For the cumulative total of 65 treat- predicting OS in patients treated with chemotherapy are
ments, a paclitaxel-containing regimen significantly im- shown in Table 3. In the univariate Cox proportional
proved PFS compared with a non-paclitaxel-containing hazards model, use of a paclitaxel-containing regimen in
regimen (4.60 months vs. 1.81 months, respectively; p = any line and absence of liver metastasis were significant

Fig. 3 Individual progression-free survival of patients treated with chemotherapy

Imaoka et al. BMC Cancer (2020) 20:946 Page 6 of 11

Fig. 4 Kaplan-Meier curves of overall survival in the first-line gemcitabine group, S-1 group, and gemcitabine plus nab-paclitaxel group (a); and progression-free
survival in the first-line gemcitabine group, S-1 group, and gemcitabine plus nab-paclitaxel group (b). OS, overall survival; PFS, progression-free survival

Table 2 Tumor response in each line of treatment

Total number CR PR SD PD NE ORR DCR
1st-line treatment
All patients 50 1 4 12 27 6 10.0% 34.0%
Gemcitabine 24 0 1 5 15 3 4.2% 25.0%
S-1 12 0 0 5 6 1 0.0% 41.7%
Gemcitabine plus nab-paclitaxel 6 0 2 1 2 1 33.3% 50.0%
Gemcitabine plus S-1 2 0 0 0 2 0
FOLFIRINOX 2 0 1 0 1 0
S-1 plus radiation 1 0 0 1 0 0
Paclitaxel 1 1 0 0 0 0
S-1 plus cisplatin 1 0 0 0 1 0
Gemcitabine plus radiation 1 0 0 0 0 1
2nd-line treatment
All patients 13 0 1 4 4 4 7.7% 38.5%
S-1 6 0 0 3 1 2 0.0% 50.0%
Gemcitabine 2 0 0 0 1 1
Gemcitabine plus nab-paclitaxel 2 0 0 0 1 1
FOLFIRINOX 2 0 1 1 0 0
FOLFOX 1 0 0 0 1 0
3rd-line treatment
All patients 2 0 1 1 0 0
Gemcitabine 1 0 0 1 0 0
Gemcitabine plus nab-paclitaxel 1 0 1 0 0 0
Cumulative total
All patients 65 1 6 17 31 10 10.8% 36.9%
Gemcitabine 27 0 1 6 16 4 3.7% 25.9%
S-1 18 0 0 8 7 3 0.0% 44.4%
Gemcitabine plus nab-paclitaxel 9 0 3 1 3 2 33.3% 44.4%
CR complete response, PR partial response, SD stable disease, PD progressive disease, NE not evaluable, PFS progression-free survival, CI confidence interval, NR
not reached
Imaoka et al. BMC Cancer (2020) 20:946 Page 7 of 11

Fig. 5 Kaplan-Meier curves comparing overall survival in patients treated with a first-line paclitaxel-containing regimen and a non-paclitaxel-
containing regimen (a), and comparing progression-free survival in patients treated with a paclitaxel-containing regimen and a non-paclitaxel-
containing regimen in any line (b). OS, overall survival; PFS, progression-free survival

factors associated with OS. In the multivariate Cox pro- limited response, with median PFS and an ORR of 1.61
portional hazards model, use of a paclitaxel-containing months and 3.7%, respectively. Most patients with unre-
regimen in any line was still an independent predictor of sectable UC had lower ECOG PS at the time of diagno-
OS (HR for OS, 0.221; 95% CI, 0.076–0.647; p = 0.006) sis, and the majority of patients received only one line of
in MICE. treatment. Although gemcitabine monotherapy still re-
mains a therapeutic option for frail and elderly patients
Discussion with PC [16–18], the benefit of gemcitabine monother-
Using a retrospective cohort design, the present study apy may be limited for UC patients. On the other hand,
examined the efficacy of chemotherapy in patients with paclitaxel-containing regimens may have a relatively high
UC of the pancreas. The most frequently used first-line anti-tumor effect in UC. Paclitaxel has shown activity in
treatment regimens were gemcitabine, S-1, and gemcita- anaplastic carcinoma of the thyroid [19, 20] and sarcoma
bine plus nab-paclitaxel. Although there was no signifi- (e.g. angiosarcoma and Kaposi’s sarcoma) [21–24]. It has
cant difference in OS among these first-line regimens, been reported that UC of the pancreas expressed
gemcitabine plus nab-paclitaxel significantly improved epithelial-mesenchymal transition (EMT) markers (e.g.
PFS compared with gemcitabine, and it showed a signifi- Slug, Twist, Zeb1), as in anaplastic carcinoma of the thy-
cantly higher ORR compared with both gemcitabine and roid and sarcoma [8, 25]. Drug sensitivity of UC of the
S-1. In addition, one complete response was observed in pancreas may be similar to these neoplasms because
a patient treated with paclitaxel. A paclitaxel-containing they show similar pathological features and expression
first-line regimen significantly improved OS compared of EMT markers.
with a non-paclitaxel-containing regimen. After adjust- EMT is a complex process by which epithelial cells
ment, use of a paclitaxel-containing regimen in any line lose their cell polarity and cell-cell adhesion, and they
was still an independent predictor of OS. All these ob- gain migratory and invasive properties to mesenchymal
servations indicate that a paclitaxel-containing regimen cells. Accumulating evidence indicates that EMT plays a
is a reasonable option for treatment of patients with crucial role in cancer-related events, including cancer in-
unresectable UC of the pancreas. vasion and metastasis. EMT is known to be associated
The present study showed that most UC patients were with a poor prognosis in various cancers, and this fact
diagnosed at the age of 60–70 years, with a slight male may contribute to the aggressive clinical course of UC.
predominance. These clinical features were similar to On the other hand, the mechanism of the effect of pacli-
those of PC [2], but median OS for UC patients treated taxel on UC is unclear. Paclitaxel, including nab-
with chemotherapy did not exceed 5 months. In recent paclitaxel, is a widely used chemotherapy drug for vari-
phase 3 trials for metastatic PC [14, 15], the median OS ous cancers, including PC [26–28]. However, many stud-
reached approximately 1 year. The present study clearly ies have reported that the EMT is associated with
showed that UC was refractory to chemotherapy. Of the acquired resistance to chemotherapy drugs [29–31], in-
chemotherapeutic regimens used for UC, gemcitabine cluding paclitaxel [32]. The EMT is known as a hetero-
monotherapy was the most frequently used regimen for geneous phenomenon and the progression of cancer
unresectable UC in the present study; it provided a varies depending on the EMT phenotype [33]. Mattiolo
Imaoka et al. BMC Cancer (2020) 20:946 Page 8 of 11

Table 3 Predictors of survival in patients treated with chemotherapy

Complete-case analysis (n = 44) Multiple imputation by chained equation (n = 50)
Unadjusted HR Adjusted HR Unadjusted HR Adjusted HR
Estimate 95% CI P Estimate 95% CI P Estimate 95% CI P Estimate 95% CI P
Sex
Male 1.000 1.000
(Reference)
Female 0.537 0.255 — 1.131 0.102 0.606 0.303 — 1.213 0.157
Age, y
< 65 1.000 1.000 1.000 1.000
(Reference)
≥ 65 1.982 0.981 — 4.004 0.057 3.404 1.472 — 7.875 0.004 1.877 0.981 — 3.591 0.057 2.242 1.119 0.023
—
4.494
ECOG PS
0 (Reference) 1.000 1.000 1.000 1.000
≥1 1.384 0.626 — 3.060 0.422 1.709 0.752 — 3.885 0.201 1.292 0.638 — 2.619 0.477 1.293 0.627 0.486
—
2.667
Prior surgical resection
No (Reference) 1.000 1.000
Yes 0.798 0.414 — 1.537 0.500 0.760 0.413 — 1.397 0.377
Tumor location
Head 1.000 1.000
(Reference)
Body-Tail 1.124 0.574 — 2.200 0.733 1.153 0.619 — 2.145 0.654
Extent of disease
Locally 1.000 1.000 1.000 1.000
advanced
(Reference)
Metastatic 0.820 0.316 — 2.132 0.685 0.630 0.210 — 1.891 0.410 0.820 0.319 — 2.107 0.681 0.606 0.204 0.369
—
1.805
Location of metastases
Liver 2.173 1.111 — 4.248 0.023 2.373 1.086 — 5.184 0.030 1.919 1.034 — 3.559 0.039 1.721 0.856 0.127
—
3.457
Lymph node 0.805 0.413 — 1.570 0.525 0.831 0.442 — 1.563 0.566
Peritoneal 0.865 0.406 — 1.842 0.707 0.933 0.458 — 1.901 0.849
LDH, U/L
≤ 250 1.000 1.000
(Reference)
> 250 0.715 0.324 — 1.576 0.405 0.829 0.407 — 1.686 0.604
CRP, mg/L
≤ 10 1.000 1.000
(Reference)
> 10 1.624 0.846 — 3.116 0.145 1.650 0.898 — 3.031 0.106
CEA, ng/mL
≤ 5.0 1.000 1.000
(Reference)
> 5.0 1.239 0.600 — 2.559 0.562 1.337 0.675 — 2.650 0.405
CA19–9, U/mL
Imaoka et al. BMC Cancer (2020) 20:946 Page 9 of 11

Table 3 Predictors of survival in patients treated with chemotherapy (Continued)

Complete-case analysis (n = 44) Multiple imputation by chained equation (n = 50)
Unadjusted HR Adjusted HR Unadjusted HR Adjusted HR
Estimate 95% CI P Estimate 95% CI P Estimate 95% CI P Estimate 95% CI P
≤ 37.0 1.000 1.000
(Reference)
> 37.0 1.032 0.537 — 1.981 0.926 1.143 0.620 — 2.107 0.669
Histological subtype
UC without 1.000 1.000 1.000 1.000
OGCs
(Reference)
UC with OGCs 0.935 0.385 — 2.268 0.882 2.372 0.848 — 6.635 0.100 0.857 0.378 — 1.943 0.713 0.826 0.354 0.660
—
1.930
Use of paclitaxel-containing regimen in any line
No (Reference) 1.000 1.000 1.000 1.000
Yes 0.216 0.076 — 0.620 0.004 0.181 0.062 — 0.534 0.002 0.218 0.077 — 0.621 0.004 0.221 0.076 0.006
—
0.647
HR hazard ratio, CI confidence interval, ECOG PS Eastern Cooperative Oncology Group performance status, LDH lactate dehydrogenase, CRP C-reactive protein, CEA
carcinoembryonic antigen, CA19–9 carbohydrate antigen 19–9, UC undifferentiated carcinoma, OGCs osteoclast-like giant cells

et al. reported that the EMT was also expressed in UC, Future directions in research on UC will lead to identi-
both with and without OGCs [23]. Ishida et al. catego- fication of biomarkers for therapeutic stratification, such
rized UC into 2 subgroups based on the expression pat- as microsatellite instability-high in various types of can-
terns of EMT markers and E-cadherin. They suggested cers [34] and EGFR in lung cancer [35]. Currently, there
that these differences in EMT phenotypes may have an is no established treatment specific to UC. Thus, most
impact on the prognosis of UC [24]. Given these find- UC patients were treated by chemotherapy in accord-
ings, the EMT status may have an impact on the re- ance with PC, but the result of the present study showed
sponse to paclitaxel in UC. However, further elucidation limited survival benefit of chemotherapy. Presently, pre-
is required to understand differences in drug responses. cision medicine, tailoring treatment based on an individ-
In addition to the efficacy of paclitaxel-containing regi- ual’s genetics, lifestyle, and environment, has emerged.
mens, a multivariate Cox proportional hazard model Development of technologies can provide a comprehen-
showed that age ≥ 65 years was an independent predictor sive view of an individual patient’s cancer [36, 37], which
of OS in patients treated with chemotherapy. Funda- can impact real-time clinical decision-making. For UC
mentally, chemotherapy provides a modest survival patients, precision medicine has not been well estab-
benefit in patients with unresectable UC. However, some lished. However, newer technologies [38] can unveil
patients had a good response to chemotherapy and various potential predictive biomarkers for possible de-
achieved relatively long survival. In such a situation, velopment of new treatments and precision medicine.
there is a critical need to identify high-risk patients and This study has some limitations. The first limitation is
select patients who will potentially benefit from treat- the fact that the present analysis was a retrospective
ment based on predictors. For example, for patients who study that lacked adequate statistical power due to the
are not expected to respond to chemotherapy, it is pos- small sample size. Therefore, the study should be con-
sible to avoid highly invasive treatments and focus on sidered only an exploratory investigation. However, UC
quality of life. By predicting the chemotherapeutic re- of the pancreas is a rare malignant neoplasm, and this
sponse, it makes a significant contribution to the selec- could complicate the recruiting for and completion of
tion of treatment for UC. Age is a widely accepted clinical trials for UC of the pancreas. The retrospective
prognostic factor for PC, and Clark et al. reported that design and relatively small sample size limit the strength
age was an independent prognostic factor for survival in of this study, and the effects of chemotherapy need to be
UC (HR per 10 years, 1.1; 95% CI, 1.04–1.2) in their evaluated in a larger patient cohort. However, the result
population-based study [6]. This report supports the is not negligible because the results will benefit patients
present findings. It should be noted that the survival with UC for whom it has been difficult to establish
benefit of chemotherapy for UC may be limited in pa- therapeutic strategies. The second limitation is missing
tients aged ≥65 years. values. Due to the retrospective nature of this study,
Imaoka et al. BMC Cancer (2020) 20:946 Page 10 of 11

missing data were unavoidable, which may lead to bias Competing interests
and loss of information in the study [39]. It may under- Yasuyuki Kawamoto has received speaking honoraria from Taiho
Pharmaceutical and Lilly. Makoto Ueno has received research funding from
mine the value of such a small-sized study for a rare dis- Taiho Pharmaceutical and Yakult Honsha, and speaking honoraria from Taiho
ease. Thus, multiple imputation was used to account for Pharmaceutical and Yakult Honsha. The other authors have no conflict of
missing values. The prognostic factors obtained by mul- interest.

tiple imputation may be useful in decision-making for Author details

the treatment of UC of the pancreas. The final limitation 1
Department of Hepatobiliary and Pancreatic Oncology, National Cancer
is the small number of patients treated with FOLFIRI- Center Hospital East, 6-5-1, Kashiwanoha, Kashiwa, Chiba 277-8577, Japan.
2
Department of Hepatobiliary and Pancreatic Oncology, National Cancer
NOX. Of the 4 patients treated with FOLFIRINOX, two Center Hospital, Tokyo, Japan. 3Department of Clinical Oncology, St.Marianna
had partial response. FOLFIRINOX may be potentially University School of Medicine, Kawasaki, Japan. 4Department of
effective for UC of the pancreas. However, even so, UC Gastroenterological Medicine, Cancer Institute Hospital Japanese Foundation
for Cancer Research, Tokyo, Japan. 5Department of Gastroenterology,
patients are often in poor condition. Thus, less invasive Hepatobiliary and Pancreatic Medical Oncology Division, Kanagawa Cancer
treatment than FOLFIRINOX, gemcitabine plus nab- Center, Yokohama, Japan. 6Department of Gastroenterology, Kanazawa
paclitaxel, is a reasonable choice for UC of the pancreas University Hospital, Kanazawa, Japan. 7Division of Gastrointestinal Oncology,
Shizuoka Cancer Center, Shizuoka, Japan. 8Department of Gastroenterology,
[16]. Shikoku Cancer Center, Matsuyama, Japan. 9Department of Gastroenterology,
Ishikawa Prefectural Central Hospital, Kanazawa, Japan. 10Department of
Conclusions Internal Medicine, Niigata Cancer Center Hospital, Niigata, Japan. 11Division
of Pancreato-Biliary Section, Department of Gastroenterology, JA Sapporo
The results of the present retrospective multicenter co- Kohsei Hospital, Sapporo, Japan. 12Division of Cancer Center, Hokkaido
hort study show that paclitaxel-containing regimens University Hospital, Sapporo, Japan. 13Department of Gastroenterology,
would offer relatively longer survival, and they are con- Fukushima Medical University School of Medicine, Fukushima, Japan.
14
Department of Hepato-Biliary-Pancreatic Surgery, Tochigi Cancer Center,
sidered a reasonable option for treating patients with Utsunomiya, Japan. 15Department of Gastroenterological, Breast and
unresectable UC. Endocrine Surgery, Yamaguchi University Graduate School of Medicine, Ube,
Japan. 16Department of Medical Oncology, Kyorin University Faculty of
Abbreviations Medicine, Tokyo, Japan.
CI: Confidence interval; ECOG PS: Eastern Cooperative Oncology Group
performance status; HR: Hazard ratio; MICE: Multiple imputation by chained Received: 15 July 2020 Accepted: 25 September 2020
equation; NOS: Not otherwise specified; ORR: Objective response rate;
OS: Overall survival; PC: Pancreatic cancer; PFS: Progression-free survival; RECI
ST: Response Evaluation Criteria in Solid Tumors; UC: Undifferentiated
References
carcinoma; UC-OGCs: Undifferentiated carcinoma with osteoclast-like giant
1. Siegel RL, Miller KD, Jemal A. Cancer statistics, 2020. CA Cancer J Clin. 2020;
cells
70(1):7–30.
2. Pourshams A, Sepanlou SG, Ikuta KS, Bisignano C, Safiri S, Roshandel G,
Acknowledgements
Sharif M, Khatibian M, Fitzmaurice C, Nixon MR, et al. The global, regional,
This work was conducted by the Japan Observational Study Committee of and national burden of pancreatic cancer and its attributable risk factors in
Hepatobiliary and Pancreatic Oncology (JOSC-HBP). We would like to thank 195 countries and territories, 1990–2017: a systematic analysis for the global
all the members of JOSC-HBP. burden of disease study 2017. Lancet Gastroenterol Hepatol. 2019;4(12):934–
47.
Authors’ contributions 3. Chien W, Sudo M, Ding LW, Sun QY, Wuensche P, Lee KL, Hattori N, Garg M,
Conception and design: HI and MI. Acquisition of data: HI, KM, KU, MO, SK, Xu L, Zheng Y, et al. Functional genome-wide screening identifies targets
TT, HI, CS, KT, KS, KO, YK, RS, HS, and HN. Analysis and interpretation of data: and pathways sensitizing pancreatic Cancer cells to Dasatinib. J Cancer.
HI, MI, SK, MU, CM, and JF. Writing, review, and revision of the manuscript: 2018;9(24):4762–73.
HI, MI, KM, KU, MO, SK, TT, HI, CS, KT, KS, KO, YK, RS, HS, HN, MU, CM, and JF. 4. Chien W, Ding LW, Sun QY, Torres-Fernandez LA, Tan SZ, Xiao J, Lim SL,
Study supervision: MI, MU, CM, and JF. All authors have read and approved Garg M, Lee KL, Kitajima S, et al. Selective inhibition of unfolded protein
the manuscript. response induces apoptosis in pancreatic cancer cells. Oncotarget. 2014;
5(13):4881–94.
Funding 5. Tschang TP, Garza-Garza R, Kissane JM. Pleomorphic carcinoma of the
This work was supported in part by the National Cancer Center Research and pancreas: an analysis of 15 cases. Cancer. 1977;39(5):2114–26.
Development Fund (29-A-3), which plays a role of scientific research on 6. Clark CJ, Graham RP, Arun JS, Harmsen WS, Reid-Lombardo KM. Clinical
multi-institutional trials to establish new standard treatment of solid tumors outcomes for anaplastic pancreatic cancer: a population-based study. J Am
in adults, from the Ministry of Health, Labour, and Welfare of Japan. Coll Surg. 2012;215(5):627–34.
7. Strobel O, Hartwig W, Bergmann F, Hinz U, Hackert T, Grenacher L,
Availability of data and materials Schneider L, Fritz S, Gaida MM, Buchler MW, et al. Anaplastic pancreatic
The datasets used and/or analyzed during the current study are available cancer: presentation, surgical management, and outcome. Surgery. 2011;
from the corresponding author on reasonable request. 149(2):200–8.
8. Paal E, Thompson LD, Frommelt RA, Przygodzki RM, Heffess CS. A
Ethics approval and consent to participate clinicopathologic and immunohistochemical study of 35 anaplastic
This study was approved by the Institutional Review Board of National carcinomas of the pancreas with a review of the literature. Ann Diagn
Cancer Center (reference number, 2018–242). Approval for the review of Pathol. 2001;5(3):129–40.
hospital records was obtained from the Institutional Review Board of 9. Latenstein AEJ, van der Geest LGM, Bonsing BA, Groot Koerkamp B, Haj
National Cancer Center and the patients’ informed consent was waived Mohammad N, de Hingh I, de Meijer VE, Molenaar IQ, van Santvoort HC,
given the retrospective nature of the study. van Tienhoven G, et al. Nationwide trends in incidence, treatment and
survival of pancreatic ductal adenocarcinoma. Eur J Cancer. 2020;125:83–93.
Consent for publication 10. Vincent A, Herman J, Schulick R, Hruban RH, Goggins M. Pancreatic cancer.
Not applicable. Lancet. 2011;378(9791):607–20.
Imaoka et al. BMC Cancer (2020) 20:946 Page 11 of 11

11. Hruban RH, Adsay NV, Esposito I, Fukushima N, Furukawa T, Klöppel G, 30. Dongre A, Weinberg RA. New insights into the mechanisms of epithelial-
Maitra A, Notohara K, Offerhaus GJA, Ohike N, et al. Pancreatic ductal mesenchymal transition and implications for cancer. Nat Rev Mol Cell Biol.
adenocarcinoma. In: World Health Organization Classification of Tumours 2019;20(2):69–84.
Digestive System Tumours. 5th edn.: IARC Press; 2019. p. 322–32. 31. Huang J, Li H, Ren G. Epithelial-mesenchymal transition and drug resistance
12. Eisenhauer EA, Therasse P, Bogaerts J, Schwartz LH, Sargent D, Ford R, in breast cancer (review). Int J Oncol. 2015;47(3):840–8.
Dancey J, Arbuck S, Gwyther S, Mooney M, et al. New response evaluation 32. Yang Q, Huang J, Wu Q, Cai Y, Zhu L, Lu X, Chen S, Chen C, Wang Z.
criteria in solid tumours: revised RECIST guideline (version 1.1). Eur J Cancer. Acquisition of epithelial-mesenchymal transition is associated with Skp2
2009;45(2):228–47. expression in paclitaxel-resistant breast cancer cells. Br J Cancer. 2014;110(8):
13. White IR, Royston P, Wood AM. Multiple imputation using chained 1958–67.
equations: issues and guidance for practice. Stat Med. 2011;30(4):377–99. 33. Saitoh M. Involvement of partial EMT in cancer progression. J Biochem.
14. Conroy T, Desseigne F, Ychou M, Bouche O, Guimbaud R, Becouarn Y, 2018;164(4):257–64.
Adenis A, Raoul JL, Gourgou-Bourgade S, de la Fouchardiere C, et al. 34. Diaz L, Marabelle A, Kim TW, Geva R, Van Cutsem E, André T, Ascierto PA,
FOLFIRINOX versus gemcitabine for metastatic pancreatic cancer. N Engl J Maio M, Delord J-P, Gottfried M, et al. 386PEfficacy of pembrolizumab in
Med. 2011;364(19):1817–25. phase 2 KEYNOTE-164 and KEYNOTE-158 studies of microsatellite instability
15. Von Hoff DD, Ervin T, Arena FP, Chiorean EG, Infante J, Moore M, Seay T, high cancers. Ann Oncol. 2017;28(suppl_5):v128–9.
Tjulandin SA, Ma WW, Saleh MN, et al. Increased survival in pancreatic 35. Rosell R, Moran T, Queralt C, Porta R, Cardenal F, Camps C, Majem M, Lopez-
cancer with nab-paclitaxel plus gemcitabine. N Engl J Med. 2013;369(18): Vivanco G, Isla D, Provencio M et al: Screening for epidermal growth factor
1691–703. receptor mutations in lung cancer. N Engl J Med 2009, 361(10):958-967. doi:
16. Ducreux M, Cuhna AS, Caramella C, Hollebecque A, Burtin P, Goere D, https://doi.org/10.1056/NEJMoa0904554 Epub 0902009 Aug 0904519.
Seufferlein T, Haustermans K, Van Laethem JL, Conroy T, et al. Cancer of the 36. Mardis ER, Wilson RK. Cancer genome sequencing: a review. Hum Mol
pancreas: ESMO clinical practice guidelines for diagnosis, treatment and Genet. 2009;18(R2):R163–8.
follow-up. Ann Oncol. 2015;26(Suppl 5):v56–68. 37. Aziz N, Zhao Q, Bry L, Driscoll DK, Funke B, Gibson JS, Grody WW, Hegde
17. Network. NCC: Clinical Practice Guidelines in Oncology: Pancreatic MR, Hoeltge GA, Leonard DG, et al. College of American Pathologists'
Adenocarcinoma. Version 1.2020. 2020. laboratory standards for next-generation sequencing clinical tests. Arch
18. Imaoka H, Kou T, Tanaka M, Egawa S, Mizuno N, Hijioka S, Hara K, Yazumi S, Pathol Lab Medi. 2015;139(4):481–93.
Shimizu Y, Yamao K. Clinical outcome of elderly patients with unresectable 38. Imaoka H, Sasaki M, Hashimoto Y, Watanabe K, Ikeda M. New Era of
pancreatic cancer treated with gemcitabine plus S-1, S-1 alone, or Endoscopic Ultrasound-Guided Tissue Acquisition: Next-Generation
gemcitabine alone: subgroup analysis of a randomised phase III trial, GEST Sequencing by Endoscopic Ultrasound-Guided Sampling for Pancreatic
study. Eur J Cancer. 2016;54:96–103. Cancer. J Clin Med. 2019;8(8):1173.
19. Ain KB, Egorin MJ, DeSimone PA. Treatment of anaplastic thyroid carcinoma 39. Sterne JA, White IR, Carlin JB, Spratt M, Royston P, Kenward MG, Wood AM,
with paclitaxel: phase 2 trial using ninety-six-hour infusion. Collaborative Carpenter JR. Multiple imputation for missing data in epidemiological and
anaplastic thyroid Cancer health intervention trials (CATCHIT) group. clinical research: potential and pitfalls. Bmj. 2009;338:b2393.
Thyroid. 2000;10(7):587–94.
20. Gallardo E, Medina J, Sanchez JC, Viudez A, Grande E, Porras I, Ramon YCT, Publisher’s Note
Trigo J, Iglesias L, Capdevila J. SEOM clinical guideline thyroid cancer (2019). Springer Nature remains neutral with regard to jurisdictional claims in
Clin Transl Oncol. 2020;22(2):223–35. published maps and institutional affiliations.
21. Reid E, Suneja G, Ambinder RF, Ard K, Baiocchi R, Barta SK, Carchman E,
Cohen A, Crysler OV, Gupta N, et al. AIDS-Related Kaposi Sarcoma, Version 2.
2019, NCCN Clinical Practice Guidelines in Oncology. J Nat Comprehensive
Cancer Network. 2019;17(2):171–89.
22. Linch M, Miah AB, Thway K, Judson IR, Benson C. Systemic treatment of
soft-tissue sarcoma-gold standard and novel therapies. Nat Rev Clin Oncol.
2014;11(4):187–202.
23. Mattiolo P, Fiadone G, Paolino G, Chatterjee D, Bernasconi R, Piccoli P,
Parolini C, El Aidi M, Sperandio N, Malleo G, et al. Epithelial-mesenchymal
transition in undifferentiated carcinoma of the pancreas with and without
osteoclast-like giant cells. Virchows Arch. 2020. https://doi.org/10.1007/
s00428-020-02889-3. (Online ahead of print).
24. Ishida K, Yamashita R, Osakabe M, Uesugi N, Yamada N, Nitta H, Fujishima F,
Motoi F, Suzuki H, Shimamura H, et al. Expression of epithelial-Mesenchymal
transition proteins in pancreatic anaplastic (undifferentiated) carcinoma.
Pancreas. 2019;48(1):36–42.
25. Hoorens A, Prenzel K, Lemoine NR, Kloppel G. Undifferentiated carcinoma of
the pancreas: analysis of intermediate filament profile and Ki-ras mutations
provides evidence of a ductal origin. J Pathol. 1998;185(1):53–60.
26. Lee JH, Kim C, Sethi G, Ahn KS. Brassinin inhibits STAT3 signaling pathway
through modulation of PIAS-3 and SOCS-3 expression and sensitizes human
lung cancer xenograft in nude mice to paclitaxel. Oncotarget. 2015;6(8):
6386–405.
27. Jung YY, Shanmugam MK, Narula AS, Kim C, Lee JH, Namjoshi OA, Blough
BE, Sethi G, Ahn KS. Oxymatrine Attenuates Tumor Growth and Deactivates
STAT5 Signaling in a Lung Cancer Xenograft Model. Cancers (Basel). 2019;
11(1):49.
28. Hwang ST, Kim C, Lee JH, Chinnathambi A, Alharbi SA, Shair OHM, Sethi G,
Ahn KS. Cycloastragenol can negate constitutive STAT3 activation and
promote paclitaxel-induced apoptosis in human gastric cancer cells.
Phytomedicine. 2019;59:152907.
29. De Craene B, Berx G. Regulatory networks defining EMT during cancer
initiation and progression. Nat Rev Cancer. 2013;13(2):97–110.