Clinical Investigation: Gastrointestinal Cancer
Intraoperative Radiotherapy for Pancreatic Cancer:
30-Year Experience in a Single Institution in Japan
Keiichi Jingu, M.D.,* Takaya Tanabe, M.D.,* Kenji Nemoto, M.D.,y
Hisanori Ariga, M.D.,* Rei Umezawa, M.D.,* Yoshihiro Ogawa, M.D.,*
Ken Takeda, M.D.,* Masashi Koto, M.D.,* Toshiyuki Sugawara, M.D.,*
Masaki Kubozono, M.D.,* Eiji Shimizu, M.D.,* Keiko Abe, M.D.,*
and Shogo Yamada, M.D.*
*Department of Radiation Oncology, Tohoku University School of Medicine, Sendai, Japan; andyDepartment of Radiation
Oncology, Yamagata University School of Medicine, Yamagata, Japan
Received May 20, 2011, and in revised form Jan 5, 2012. Accepted for publication Jan 6, 2012
The objective was to analyze
retrospectively the results of
for localized pancreatic
cancer in the past three
decades and analyze prog-
nostic factors. Eighty-three
patients underwent gross
total resection, and 109
underwent only biopsy or
palliative resection. Two-
year local control and overall
survival rates were 71.0%
and 16.9%, respectively.
Overall survival rate was
decade by decade. Multivar-
iate analysis showed that
degree of resection and
adjuvant chemotherapy had
significant impacts on
Purpose: To analyze retrospectively the results of intraoperative radiotherapy (IORT) with or
without external beam radiotherapy (? EBRT) for localized pancreatic cancer in the past three
decades and to analyze prognostic factors by multivariate analysis.
Methods and Materials: Records for 322 patients with pancreatic cancer treated by IORT ?
EBRT in Tohoku University Hospital between 1980 and 2009 were reviewed. One hundred
ninety-two patients who had no distant organ metastases or dissemination at the time of lapa-
rotomy were enrolled in the present study.
Results: Eighty-three patients underwent gross total resection (R0: 48 patients, R1: 35 patients),
(LC) and overall survival (OS) rates were 71.0% and 16.9%, respectively. Comparison of the
results for each decade showed that OS was significantly improved decade by decade (2-year:
25.0% vs. 18.8% vs. 4.2%, p < 0.001). Multivariate analysis showed that degree of resection
Z 1.54, p Z 0.028) had significant impacts on OS. Late gastrointestinal morbidity of Common
Terminology Criteria for Adverse Events version 3.0 grade 4 or 5 was observed in four patients.
chemotherapy had significant impacts on OS. ? 2012 Elsevier Inc.
Keywords: Pancreatic cancer, Intraoperative radiotherapy, 30-year experience, Prognostic factor,
Reprint requests to: Keiichi Jingu, M.D., Ph.D., Department of Radi-
ation Oncology, Tohoku University School of Medicine, Seiryo-chou 1-1,
Aoba-ku, Sendai 980-8574, Japan. Tel: (þ81) 22-717-7312; Fax: (þ81)
22-717-7316; E-mail: email@example.com
Presented at the 52nd Annual Meeting of the American Society for
Radiation Oncology (ASTRO), 31 October, 2010e04 November, 2010, in
San Diego, USA.
Conflict of interest: none.
Int J Radiation Oncol Biol Phys, Vol. 83, No. 4, pp. e507ee511, 2012
0360-3016/$ - see front matter ? 2012 Elsevier Inc. All rights reserved.
International Journal of
Despite important improvements made in the fields of surgery,
chemotherapy, and radiation therapy, pancreatic cancer remains
one of the most lethal malignancies with overall survival (OS) of
only approximately 5% at 5 years (1).
Intraoperative radiation therapy (IORT), the delivery of
a single large dose of irradiation at the time of surgery, was
developed to administer higher doses of irradiation while dis-
placing or shielding adjacent normal tissue structures from radi-
ation exposure. Results of several studies published over the past
30 years have shown favorable effects of IORT for pancreatic
cancer (2, 3) but are limited with regard to any clear interpretation.
Despite many institutional approaches to IORT for pancreatic
cancer, randomized data remain sparse. A possible beneficial
impact of IORT on survival remains controversial, because most
studies have not indicated a benefit. At Tohoku University
Hospital, IORT has been used for treatment of patients with
pancreatic cancer since 1977. We were interested in reviewing our
institutional long-term experience and in determining the impact
of IORT on early and late postoperative outcomes after surgery.
Methods and Materials
We retrospectively reviewed data for 322 consecutive patients who
underwent IORT for pancreatic cancer. Of those 322 patients, 192
patients had no dissemination or distant metastases at the time of
laparotomy. The characteristics of those 192 patients are shown in
Table 1. The mean age at the start of treatment was 62.8 years.
Pathologically or clinically under laparotomy (in patients whom
could only biopsy) was performed; one patient was classified as
Stage IA (Union for International Cancer Control (UICC) 2002),
16 patients were classified as Stage IB, 29 patients were classified
as Stage IIA, 65 patients were classified as Stage IIB, and 81
patients were classified as Stage III. The median serum concen-
tration of carbohydrate antigen 19-9 (CA19-9) before surgery þ
IORT was 305.5 U/mL (range, 0.1e99,999 U/mL).
The operations were performed in the Department of Surgery of
Tohoku University Hospital. All pathology and operative reports
were reviewed to determine the extent of resection. Margin
status was reported as R0 for no residual tumor, R1 for micro-
scopic residual tumor <1 mm from the margin (if quantitative
margin data were available), and R2 for macroscopic residual
tumor, consistent with American Joint Committee on Cancer
IORT was performed under laparotomy with a linear accelerator
(Varian Medical Systems, Palo Alto, CA) using a 6-20 MeV
electron beam for the tumor or tumor bed with a cylinder size of
46-12 cm in a single fraction. The IORT field was designed to
cover the primary tumor or tumor bed with a 1-cm margin by at
least 80% of the prescribed dose. The median dose of IORT was
25 Gy (range, 20e30 Gy).
External beam radiotherapy
External beam radiotherapy (EBRT) was performed mainly for
patients with R2 resection. EBRT was performed with 10-MV X-
ray using 3 or 4 beams for the primary tumor or primary tumor
bed and regional lymph nodes. The daily fractional dose of
radiotherapy was 2.0 Gy, administered 5 days a week. The median
dose of adjuvant EBRT was 40 Gy (range, 14e50 Gy).
During external beam radiotherapy, a 5-fluorouracil (5-FU)-based
regimen before 2000 and gemcitabine at 1000 mg/body weekly for
3 weeks with a 1-week rest were administrated concurrently. As
additional chemotherapy, 5-fluorouracil-based regimen or oral
tegafur/uracil was administrated between 1980 and 1999, and
gemcitabine has been administrated since 2000 at 1,000 mg/m2
weekly for 3 weeks with a 1-week rest after discharge from our
hospital. The patients who underwent additional chemotherapy
continued receiving the chemotherapy were until significant
disease progression or patient’s refusal.
The patients treated before the mid-1990s were followed up by
ultrasound (US) and serum tumor marker (CA19-9 and/or carci-
noembryonicantigen [CEA])testevery month forthe first12 months
after surgery and/or IORT and thereafter every 2 months. In those
patients, computed tomography (CT) or magnetic resonance imaging
(MRI) was performed for patients who showed abnormalities in
examinations or who showed symptoms. After the mid-1990s, all
months after surgery IORTand thereafter every 2 or 3 months.
Toxicity was graded according to the Common Terminology
Criteria for Adverse Events (version 3.0). An adverse effect at
more than 90 days after completion of IORT was defined as a late
The present retrospective study protocol was reviewed and
approved by Tohoku University Institutional Review Board
Categorical variables were compared using the chi-square test. The
Kruskal-Wallis test was used to compare continuous variables.
Local control (LC), defined as the absence of significant tumor
growth at the primary site of disease on interval CTor US, and OS
were calculated by the Kaplan-Meier method, from the date of
surgery þ IORT until the date of local recurrence or death,
respectively, or up to the date of last follow-up. Differences
between survival curves in the subsets of patients were analyzed
using the log-rank test. Prognostic factors included in the Cox
proportional-hazards regression model were age (?64 vs. <64),
performance status (0e1 vs. 2e3), pathological stage (IAeIIAvs.
IIBeIII), CA19-9 (?300 U/mL vs. <300 U/mL), degree of
resection (R0e1 vs. R2), adjuvant chemotherapy (yes vs. no),
Jingu et al. International Journal of Radiation Oncology ? Biology ? Physics
EBRT (yes vs. no), and treatment period (1980e1989 vs.
using the SPSS 11.0 software package (SPSS, Chicago, IL).
Eighty-three patients underwent gross total removal, and 109
patients underwent only biopsy or palliative resection (R0: 48
patients, R1: 35 patients, R2: 109 patients). Fifty-five patients
underwent additional EBRT, and 124 patients underwent addi-
tional chemotherapy. Comparison of the patients’ characteristics
in each decade showed that there were significant differences in
the degree of resection, pathological stage, and additional
chemotherapy (Table 1). The median follow-up period for all
patients was 10.7 months (range, 0.1e250.3 months) and that for
the 23 surviving patients was 37.5 months (range, 0.7e192). At
the time of the analysis, 166 patients (86.5%) had disease recur-
rence, and 35 patients (18.2%) had local failure. The 2-year and
5-year OS rates in all of the 192 patients were 16.9% (95%
confidence interval [CI] Z 11.4%e22.3%) and 8.6% (95% CI Z
4.4%e12.8%), respectively, with a median survival period of 16.0
months. The 2- and 5-year LC rates were 71.0% (95% CI Z
60.7%e81.3%) and 63.9% (95% CI Z 41.0%e86.3%), respec-
tively (Fig. 1). Comparison of the results for each decade
(2000e2009 vs. 1990e1999 vs. 1980e1989) showed that OS was
significantly improved decade by decade (2-year: 25.0% vs.
18.8% vs. 4.2%, p < 0.001; Fig. 2). In patients in whom R0
resection could be performed, OS in the past 2 decades was also
significantly higher than that in patients between 1980 and 1989
(2-year: 38.8% vs. 42.9% vs. 0%, p Z 0.002 and p Z 0.009,
respectively). In patients in whom only palliative resection or
biopsy was performed, OS in the past decade was significantly
higher than that in patients between 1980 and 1989, and OS in the
past decade tended to be higher than that in patients between 1990
and 1999 (2-year: 11.5% vs. 4.8% vs. 5.3%, p Z 0.02 and
p Z 0.09, respectively). In patients in whom R1 resection could
patients who were enrolled in the present study (n Z 192).
Local control rate and overall survival rate for all
Age (mean ? SD)
Primary lesion (n)
Stage (UICC 2002) (n)
Degree of resection (n)
Radiotherapy method (n)
IORT þ EBRT
Additional chemotherapy (n)
62.8 ? 9.4 64.5 ? 8.3 61.4 ? 9.963.3 ? 9.6 0.263
40 146 42
Abbreviations: CA19-9 Z carbohydrate antigen 19-9; EBRT Z external beam radiotherapy; IORT Z intraoperative radiation therapy; UICC Z Union
for International Cancer Control.
Volume 83 ? Number 4 ? 2012Intraoperative radiotherapy for pancreatic cancer e509
be performed, there was no significant difference in OS (2-year:
23.4% vs. 25.0% vs. 0%).
Late gastrointestinal morbidity of grade 4 or 5 was observed in
4 patients (2.1%). There was no grade 4 or 5 toxicity in patients
treated with IORT þ EBRT.
Multivariate analysis showed that degree of resection (R0e1
vs. R2, hazard ratio [HR] Z 1.97, 95% CI Z 1.33e2.93,
p Z 0.001) and additional chemotherapy (yes vs. no, HR Z 1.54,
95% CI Z 1.05e2.25, p Z 0.028) had significant impacts on OS
(Table 2). Pathological stage (HR Z 1.44, 95% CI Z 0.93e2.25,
CI Z 0.49e1.03, p Z 0.068) tended to be prognostic factors.
Pancreatic cancer is one of the most difficult malignancies to cure.
Despite progress in surgery, radiotherapy, and chemotherapy, OS
between 1999 and 2005 improved only slightly from that between
1975 and 1977 and that between 1984 and 1986 (1). Prognosis of
localized pancreatic cancer treated with IORT in our institution
has improved significantly decade by decade despite little change
in the technique used for IORT. On the basis of differences in
patients’ characteristics in the three decades and the results of
multivariate analysis, the improvement in survival is thought to be
due to not only the frequency of additional chemotherapy and
improvement in diagnostic imaging technology for early detection
but also the introduction of gemcitabine and improvement in
surgical technique and perioperative care, which were included as
factors in the treatment period in multivariate analysis.
IORT allows a theoretical increase in the radiation therapeutic
index to the tumor compared to adjacent organs at risk (OAR) for
the following three reasons: (1) The biological effectiveness of
a single high dose of radiation is greater than that of the same dose
administered by conventional fractionated radiation. (2) The
radiation is directed exactly on the area with an increased risk of
tumor relapse or on the tumor. (3) Irradiation of dose-limiting
OAR, such as the choledochus, small intestine, stomach, and
spinal cord, can reduce through manual mobilization of OAR from
the irradiation area, through the use of appropriate lead protections
or through the proper use of an electron beam with energies
sufficient to limit the radiation to deep structures (5).
Alfieri et al. (6) noted increased local control with the addition
of IORT in resected pancreatic cancer, and they found that 5-year
LC was 58.4% in the surgery with IORT group vs. 29.8% in the
surgery alone group (p < 0.01). In the present study also, excellent
local control for pancreatic cancer with little severe late toxicity
was achieved by using IORT. However, LC before 1990 might not
be accurate because resolution of CT and MRI during that period
was not good, and patients before 1990 might have died from
distant metastases and/or dissemination or perioperative compli-
cations before local recurrence. Actually, the rate of local recur-
rence in patients between 2000 and 2009 was significantly higher
than that in patients between 1980 and 1989 (2/48 vs. 17/64,
2-year LC, 77.1% vs. 65.9%, p Z 0.027; Figure 3).
In most of the past studies using IORT for pancreatic cancer,
IORT did not increase the risks associated with surgery (7, 8). In
1990e1999 vs. 2000e2009). Overall survival was significantly
improved decade by decade (2-year: 25.0% vs. 18.8% vs. 4.2%,
log-rank test: p < 0.001).
Overall survival rate for each decade (1980e1989 vs.
Multivariate analysis of prognostic factors for
p valueHR (95%CI)
Stage (UICC 2002)
Degree of resection
Abbreviations: CI Z confidence interval; HR Z hazard ratio;
UICC Z Union for International Cancer Control.
1990e1999 vs. 2000e2009). Local control rate in patients
between 1980 and 1989 was significantly better than that in
patients between 2000 and 2009 (log-rank test: p Z 0.027).
Local control rate for each decade (1980e1989 vs.
Jingu et al. International Journal of Radiation Oncology ? Biology ? Physics
addition, a recent multi-institutional case series, which included Download full-text
the largest number of patients treated with IORT to date, revealed
a low frequency of late adverse events after IORT (3.3%) (9).
These were late grade 3e4 gastrointestinal events including
colitis, ileus, and bleeding. In the present study, 4 patients showed
grade 4 or higher toxicities. Those for 4 patients, except for one
patient in whom ileus occurred, had severe gastrointestinal
bleeding. Although Ogawa et al. (9) reported that a dose of 25 Gy
appears to be appropriate for IORT, those 4 patients in our series
were irradiated with 20 or 25 Gy even without EBRT. When IORT
is performed, radiation oncologists and surgeons are convinced
that the OAR have been manually moved out of irradiation area,
but great care is needed because a high dose is sometimes irra-
diated to the intestinum, especially the second and third portions
of the duodenum, in patients with unresectable pancreas head
cancer. Mitsunaga et al. (10) reported that 67% of long-term
survivors after IORT showed extrahepatic portal vein occlusion;
however, none of the patients who survived for more than 3 years
in our series showed extrahepatic portal vein occlusion. One
patient had portal vein thrombosis only 6 months after surgery þ
Some reports have suggested a benefit of IORT for survival
(11, 12); however, recently, two propensity score analyses
showed that there was no benefit of IORT (7, 8). IORT seems to
prolong not OS but LC. On the other hand, many studies have
suggested a significant benefit of additional chemotherapy for
OS (13, 14). Even with IORT, additional chemotherapy signifi-
cantly prolonged OS (2-year, 48.0% vs. 34.8%, p Z 0.0011) (9).
Because local control was not correlated with survival due to
distant metastasis and/or dissemination in patients with pancre-
atic cancer, additional chemotherapy, especially a gemcitabine-
based regimen, is considered to be essential (15). The results
of our multivariate analysis also showed that only additional
chemotherapy and degree of resection were independent prog-
nostic factors for OS. Therefore, future IORT might assume an
auxiliary role of resection and additional chemotherapy for
survival and play an important role for preservation of quality of
life by preventing local recurrence in patients with pancreatic
cancer. This consists with the results of several past studies in
patients treated with surgery and IORT (4, 8). New drugs such as
gemcitabine (15), taxanes (16), and platinum compounds (17)
may contribute to further improvement in the prognosis of this
disease, not only as antiproliferative drugs but also as potent
EBRT had no correlation with local control and overall
survival in any of the patients in the present study. In some past
studies, IORT þ EBRT showed effectiveness, especially for
unresectable pancreatic cancer (18, 19); however, in the present
study, additional EBRT had no benefit even in patients who
underwent only palliative resection or biopsy. Thus, the effec-
tiveness of additional EBRT is still controversial, and a large
prospective study is required. New radiotherapy techniques have
recently been developed (e.g., IMRT, particle beam). Because
pancreatic cancer is known to be radioresistant due to the hypoxic
environment in the tumor (20), stereotactic radiotherapy and
carbon ion radiotherapy, rather than IORT, are promising.
In conclusion, our results indicate that IORT has enabled
excellent local control for pancreatic cancer with little severe late
toxicity and that OS has improved with each decade. Multivariate
analysis suggests that IORT combined with additional chemo-
therapy has a survival benefit in comparison with IORT without
1. Jemal A, Siegel R, Ward E, et al. Cancer statistics, 2010. CA Cancer
J Clin 2010;60:277e300.
2. Manabe T, Baba N, Ono K, et al. Radical pancreatectomy with
intraoperative radiation therapy for pancreatic head cancer. In: Abe M,
Takahashi M, editors. Intraoperative radiation therapy. Proceedings of
the 3rd International Symposium on Intraoperative Radiation Therapy.
New York: Pergamon Press; 1991. p. 249e250.
following pancreaticoduodenectomy. Ann Surg 1993;218:54e60.
4. Greene FL, Page DL, Fleming ID, et al. AJCC cancer staging manual.
6th ed. New York: Springer-Verlag; 2002.
5. Kinsella TJ, Sindelar WF. Intraoperative radiotherapy for pancreatic
6. Alfieri S, Morganti AG, Di Giorgio A, et al. Improved survival and
local control after intraoperative radiation therapy and postoperative
radiotherapy: A multivariate analysis of 46 patients undergoing
surgery for pancreatic head cancer. Arch Surg 2001;136:343e347.
7. Showalter TN, Rao AS, Rani Anne P, et al. Does intraoperative
radiation therapy improve local tumor control in patients undergoing
propensity score analysis. Ann Surg Oncol 2009;16:2116e2222.
8. Schwarz RE, Smith DD, Keny H, et al. Impact of intraoperative
radiation on postoperative and disease-specific outcome after pan-
creatoduodenectomy for adenocarcinoma: A propensity score anal-
ysis. Am J Clin Oncol 2003;26:16e21.
9. Ogawa K, Karasawa K, Ito Y, et al. Intraoperative radiotherapy for
resected pancreatic cancer: A multi-institutional retrospective analysis
of 210 patients. Int J Radiat Oncol Biol Phys 2010;77:734e742.
10. Mitsunaga S, Kinoshita T, Kawashima M, et al. Extrahepatic portal
vein occlusion without recurrence after pancreaticoduodenectomy and
intraoperative radiation therapy. Int J Radiat Oncol Biol Phys 2006;64:
11. Ozaki H, Kinoshita T, Kosuge T, et al. Long-term survival after
multimodality treatment for resectable pancreatic cancer. Int J Pan-
12. Reni M, Panucci MG, Ferreri AJ, et al. Effect on local control and
survival of electron beam intraoperative irradiation for resectable
pancreatic adenocarcinoma. Int J Radiat Oncol Biol Phys 2001;50:
13. Neoptolemos JP, Stocken DD, Friess H, et al. A randomized trial of
chemoradiotherapy and chemotherapy after resection of pancreatic
cancer. N Engl J Med 2004;350:1200e1210.
14. Neoptolemos JP, Stocken DD, Tudur Smith C, et al. Adjuvant
5-fluorouracil and folinic acid vs observation for pancreatic cancer:
Composite data from the ESPAC-1 and -3(v1) trials. Br J Cancer
15. Sultana A, Tudur Smith C, Cunningham D, et al. Meta-analyses of
chemotherapy for locally advanced and metastatic pancreatic cancer.
J Clin Oncol 2007;25:2607e2615.
16. Turrini O, Ychou M, Moureau-Zabotto L, et al. Neoadjuvant
docetaxel-based chemoradiation for resectable adenocarcinoma of the
pancreas: New neoadjuvant regimen was safe and provided an inter-
esting pathologic response. Eur J Surg Oncol 2010;36:987e992.
17. Boz G, De Paoli A, Roncandin M, et al. Radiation therapy combined
with chemotherapy for inoperable pancreatic carcinoma. Tumori 1991;
18. Abe M, Shibamoto Y, Takahashi M, et al. Intraoperative radiotherapy
in carcinoma of the stomach and pancreas. World J Surg 1987;11:
19. Kawamura M, Kataoka M, Fujii T, et al. Electron beam intraoperative
radiation therapy (EBIORT) for localized pancreatic carcinoma. Int
J Radiat Oncol Biol Phys 1992;23:751e757.
20. Coleman CN. Conference summary: The ninth international confer-
ence on the Chemical Modifiers of Cancer Treatment. Br J Cancer
Volume 83 ? Number 4 ? 2012 Intraoperative radiotherapy for pancreatic cancer e511