Pancreatic resections after chemoradiotherapy for locally advanced ductal adenocarcinoma: analysis of perioperative outcome and survival.
ABSTRACT The most accepted treatment for locally advanced pancreatic cancer is chemoradiotherapy. However, indications to and results of pancreatic resections after chemoradiation are not yet defined.
From June 1999 to December 2003, 28 patients with locally advanced pancreatic cancer (group 1) were enrolled for institutional trials of gemcitabine-based chemoradiotherapy. Tumors were stratified as unresectable or borderline resectable according to the pattern of vascular involvement at pretreatment computed tomographic scan. Patients with partial response or stable disease and in-range Ca19-9 were surgically explored. Perioperative outcome and survival of group 1 were compared with 44 patients primary resected for localized cancer with or without adjuvant treatment in the same time period (group 2).
Only one unresectable tumor was successfully resected compared to 7 out of 18 (39%) that were borderline resectable. Operations after chemoradiation were 1 hour longer and postoperative stays 5 days longer, but transfusion rate, morbidity, and mortality were not significantly different. Median survival was 15.4 months for group 1 (>21 for resected vs. 10 for not resected, P < 0.01) and 14 months for group 2. In both groups, a disease-free survival beyond 24 months was recorded only among patients resected with negative margins.
The conversion of an unresectable cancer to a resectable one is a rare event. On the contrary, the resection of a borderline resectable tumor was successfully accomplished in one-third of cases. Chemoradiotherapy did not increase the operative risk, but the interventions were more technically demanding and required a longer postoperative stay. Patients resected after chemoradiation for a locally advanced tumor had at least the same survival as those primary resected for a localized one. Only R0 resections in both groups gave the chance of disease-free survival longer than 24 months.
- [show abstract] [hide abstract]
ABSTRACT: A prospective, controlled double-blind study involving a substantial number of patients suggests that 5-fluorouracil (5-F.U.) significantly augments the effectiveness of radiation therapy for locally unresectable carcinoma of the stomach, pancreas, and large bowel. It is also possible that rarely this therapy may be curative. This approach should not be advocated as routine treatment since the vast majority of these patients still die of their cancer; and, if the present results are not spurious, the method offers only a few extra months of life. These results should, however, serve as stimulus and foundation for continued study of augmented radiation therapy in the treatment of advanced gastrointestinal cancer.The Lancet 11/1969; 2(7626):865-7. · 39.06 Impact Factor
- [show abstract] [hide abstract]
ABSTRACT: One-hundred-ninety-four eligible and evaluable patients with histologically confirmed locally unresectable adenocarcinoma of the pancreas were randomly assigned to therapy with high-dose (6000 rads) radiation therapy alone, to moderate-dose (4000 rads) radiation + 5-fluorouracil (5-FU), and to high-dose radiation plus 5-FU. Median survival with radiation alone was only 5 1/2 months from date of diagnosis. Both 5-FU-containing treatment regimens produced a highly significant survival improvement when compared with radiation alone. Survival differences between 4000 rads plus 5-FU and 6000 rads plus 5-FU were not significant with an overall median survival of ten months. Significant prognostic variables, in addition to treatment, were pretreatment performance status and pretreatment CEA level. The toxic reactions related to the treatment are discussed.Cancer 10/1981; · 5.20 Impact Factor
- [show abstract] [hide abstract]
ABSTRACT: One hundred ninety-one patients with pathologically confirmed, locally unresectable adenocarcinoma of the stomach (57 patients) and pancreas (91 patients), were randomly allocated to therapy with 5-fluorouracil (5-FU) alone, 600 mg/m2 intravenously (IV) once weekly, or radiation therapy, 4,000 rad, plus adjuvant 5-FU, 600 mg/m2 IV, the first three days of radiotherapy, then follow-up maintenance 5-FU, 600 mg/m2, weekly. Forty-three patients (22%) could not be analyzed because of ineligibility or cancellation, thus 148 patients were evaluable. The median survival time was similar for both treatment programs and for both types of primary carcinoma, and was as follows: gastric primary carcinoma, 5-FU, 9.3 months; 5-FU plus radiotherapy, 8.2 months; pancreatic primary carcinoma, 5-FU, 8.2 months; 5-FU plus radiotherapy, 8.3 months. Substantially more toxicity was experienced by patients treated with the combined modality arm than by those patients receiving 5-FU alone. Severe or worse toxicity experienced by patients with gastric primary carcinoma treated by 5-FU was 19%, and the combined modality arm was 31%. The toxicity experienced by patients with pancreatic primary carcinoma treated with 5-FU was 27%, and the combined modality arm was 51%. Significant prognostic variables included: weight loss in stomach-cancer patients; and performance status, degree of anaplasia, and reduced appetite in pancreas-cancer patients.Journal of Clinical Oncology 04/1985; 3(3):373-8. · 18.04 Impact Factor
Pancreatic Resections after Chemoradiotherapy for Locally
Advanced Ductal Adenocarcinoma: Analysis of Perioperative
Outcome and Survival
Paolo Massucco, MD,1Lorenzo Capussotti, MD,1Antonella Magnino, MD,2
Elisa Sperti, MD,2Marco Gatti, MD,3Andrea Muratore, MD,1Enrico Sgotto, MD,1
Pietro Gabriele, MD,3and Massimo Aglietta, MD2
1Unit of Surgical Oncology, Institute for Research and Cure of Cancer, 10060 Candiolo, Italy
2Unit of Medical Oncology, Institute for Research and Cure of Cancer, 10060 Candiolo, Italy
3Unit of Radiation Therapy, Institute for Research and Cure of Cancer, 10060 Candiolo, Italy
Background: The most accepted treatment for locally advanced pancreatic cancer is
chemoradiotherapy. However, indications to and results of pancreatic resections after
chemoradiation are not yet defined.
Methods: From June 1999 to December 2003, 28 patients with locally advanced pancreatic
cancer (group 1) were enrolled for institutional trials of gemcitabine-based chemoradiother-
apy. Tumors were stratified as unresectable or borderline resectable according to the pattern of
vascular involvement at pretreatment computed tomographic scan. Patients with partial re-
sponse or stable disease and in-range Ca19-9 were surgically explored. Perioperative outcome
and survival of group 1 were compared with 44 patients primary resected for localized cancer
with or without adjuvant treatment in the same time period (group 2).
Results: Only one unresectable tumor was successfully resected compared to 7 out of 18
(39%) that were borderline resectable. Operations after chemoradiation were 1 hour longer
and postoperative stays 5 days longer, but transfusion rate, morbidity, and mortality were not
significantly different. Median survival was 15.4 months for group 1 (>21 for resected vs. 10
for not resected, P < 0.01) and 14 months for group 2. In both groups, a disease-free survival
beyond 24 months was recorded only among patients resected with negative margins.
Conclusions: The conversion of an unresectable cancer to a resectable one is a rare event.
On the contrary, the resection of a borderline resectable tumor was successfully accomplished
in one-third of cases. Chemoradiotherapy did not increase the operative risk, but the inter-
ventions were more technically demanding and required a longer postoperative stay. Patients
resected after chemoradiation for a locally advanced tumor had at least the same survival as
those primary resected for a localized one. Only R0 resections in both groups gave the chance
of disease-free survival longer than 24 months.
Key Words: Pancreatic neoplasm—Combined modality therapy—Surgery—Locally advanced
Pancreatic cancer continues to be the most lethal
gastrointestinal neoplasm owing to the advanced
stage at diagnosis and the lack of effective treatment.
Pancreatic resection is the only potentially curative
therapy, but unfortunately, fewer than 20% of cases
are amenable to operation at presentation. Up to
40% of all cases have no evidence of distant metas-
tases but are borderline resectable or unresectable
because of the local spread of disease (locally ad-
vanced neoplasm). On the basis of a few randomized
Received October 17, 2005; accepted April 05, 2006; published
Address correspondence and reprint requests to: Paolo Mas-
succo, MD; E-mail: email@example.com
Published by Springer Science+Business Media, Inc. ? 2006 The Society of
Surgical Oncology, Inc.
Annals of Surgical Oncology (? 2006)
trials, the most accepted treatment for this condition
is chemoradiotherapy.1–4In the last decade, many
regimens have been proposed with encouraging re-
sults.5–18However, the indications, if any, and the
results of pancreatic resections after chemoradio-
therapy are not yet clear.
Since 1999, patients presenting at our institution
with a locally advanced pancreatic adenocarcinoma
were enrolled in a phase II chemoradiation treatment
protocol. The results of this study have been reported
elsewere.19In this report, we present the perioperative
and long-term outcomes of patients resected after
resections consecutively performed during the same
PATIENTS AND METHODS
Patient Population and Treatments
From June 1999 to December 2003, 28 patients
with locally advanced pancreatic cancer (group 1)
were enrolled for institutional phase II trials of che-
moradiotherapy. In all cases, cytological diagnosis of
ductal adenocarcinoma was obtained by computed
tomography (CT)-guided fine needle aspiration.
Determination of local and distant spread was based
on a thin-section enhanced multidetector CT scan,
with vascular reconstruction when needed. Patients
with extrapancreatic disease or with massive retro-
peritoneal nodal involvement (i.e., outside the resec-
tion field) were not eligible. The tumor was defined as
locally advanced in the presence of (1) abutment (lack
of a fat plane between the tumor and the arterial wall)
or encasement of celiac axis (CA) or superior mes-
enteric artery (SMA) (cT4) or (2) involvement of
superior mesenteric-portal vein axis with either ste-
nosis or thrombosis of the vessel (cT3v+). Patients
were then stratified as (1) unresectable (in the pres-
ence of vein thrombosis and/or arterial encasement)
or (2) borderline resectable (in the presence of vein
stenosis and/or arterial abutment).
Ca19.9 serum level was recorded at registration.
The treatment protocol of the first 23 patients has
been previously detailed.19In brief, patients received
three-dimensional conformal radiotherapy with high-
energy X-rays; the clinical target volume (CTV) in-
cluded the entire lesion and the nodal areas at risk for
a total dose of 45 Gy (1.8 Gy/day). Concurrent che-
motherapy consisted of gemcitabine administered as a
60-minute intravenous infusion each Monday and
Thursdayfor5weeks. Gemcitabine was given within2
hours before radiation treatment at a dose of 100 mg/
m2twice weekly in the first 15 patients. The dose was
reduced to 50 mg/m2twice weekly in the subsequent
eight patients because of the toxicity observed and the
difficulty in giving the planned dose. The last five
patients were enrolled in a novel institutional protocol
and received two courses of induction gemcitabine-
and oxaliplatin-based chemotherapy before chemo-
radiation. Eligible patients had to sign an informed
consent form before enrollment into the trials.
Response Assessment and Indication for Operation
The therapeutic response was evaluated according
to responce evaluation criteria in solid tumours
(RECIST) criteria by a CT scan performed 6–8 weeks
after the end of treatment. Surgical exploration was
scheduled in patients who experienced downsizing of
the mass or had a stable disease with in-range Ca19-9.
In the absence of distant progression at laparotomy,
a pancreatic resection (Whipple procedure or distal
splenopancreatectomy) with extended lymphadenec-
tomy was performed as previously described.20In
summary, the lymphadenectomy field included the
nodal groups of hepatic pedicle, hepatic artery, CA,
SMA on its right and anterior aspect, and the pre-
aortic retroperitoneal tissue between the CA and
inferior mesenteric artery. A total pancreatectomy
was performed if a free pancreatic resection margin
was not obtained from intraoperative frozen section
examination. When an adhesion between the tumor
and portal/mesenteric vein wall precluding the iden-
tification of a clear cleavage plane was encountered
during the dissection, a vein resection was performed
without any attempt to directly document the sup-
posed vein infiltration during the operation. Patho-
International Union against Cancer (UICC) classifi-
cation system.21Resection margins, including the
retroperitoneal and uncinate surfaces, were checked
for neoplastic invasion. Nodal metastases were clas-
sified as regional or extraregional if found in the
preaortic nodal group.
Analysis of Results
The perioperative and long-term outcomes of
group 1 patients were compared with a group of 44
patients who in the same period underwent a primary
resection for localized ductal adenocarcinoma with or
without adjuvant gemcitabine-based chemotherapy
or chemoradiation (group 2). The adjuvant therapy
has been given since the year 2002 in patients who
P. MASSUCCO ET AL.
Ann. Surg. Oncol. (? 2006)
were able to start the treatment within 2 months of
operation. Mortality, morbidity, operative time,
transfusion rate, and hospital stay were recorded. All
patients were followed by abdominal ultrasound and
CA19-9 dosage every 4 months for the first 2 years
and every 6 months thereafter. A thoracoabdominal
CT scan was performed yearly. Additional tests (i.e.,
cytology on ascites or pleural effusion) were per-
formed when a recurrence was suspected. Data were
prospectively collected on a dedicated database.
Survival was computed from the date of treatment
initiation, i.e., the starting date of medical treatment
for group 1 and the date of operation for group 2.
The site of recurrence was recorded as local (any mass
in the resection field), peritoneal, or distant. Follow-
up was updated at June 2005.
mean ± standard deviation and compared with a
two-tailed t-test. Categoric variables were compared
with K-square or Fisher?s exact test as appropriate.
Survival curves were estimated with the Kaplan-
Meier method and compared by the log-rank test.
Overall survival was computed considering as events
all the deaths whatever the cause, including postop-
erative deaths. The events considered for disease-free
survival computation were all deaths and all recur-
rences. An analysis of prognostic factors for survival
was performed including all resected patients from
groups 1 and 2 and considering as covariates sex, age
>70 years, type of operation (partial vs. total pan-
createctomy), perioperative transfusions, grading
(G1-2 vs. G3), T stage (T1-2 vs. T3), N stage (positive
vs. negative), presence of neoplastic emboli, presence
of perineural invasion, R status, neoadjuvant therapy
(resection after neoadjuvant therapy vs. primary
resection), and any form of chemotherapy (neoadju-
vant or adjuvant therapy vs. primary resection alone).
P < 0.05 was considered significant in all cases.
Patients? characteristics are summarized in Table 1.
Group 1 patients were significantly younger but with
a mean age difference of less than 5 years. Group 2
tumors had a significantly smaller diameter at staging
CT scan, none had cT3v+ or cT4 vascular involve-
ment, and there was a trend toward a lower mean
Ca19.9 level at presentation.
All group 1 tumors had a cT3v+ or cT4 vascular
involvement. Isolated vein involvement was found
in five cases (four stenosis, one thrombosis). Of
the 23 patients with arterial involvement, a vessel
encasement was evident in eight (SMA in two, CT in
four, and both in two) and the absence of a fat plane
with the vessel wall in 15 (SMA in three, CT in six,
and both in six). In all but three of these 23 patients,
vein involvement was present as well. In five patients,
preaortic enlarged lymph nodes were detected in the
field of the planned lymphadenectomy (between CT
and inferior mesenteric artery). As for resectability
status, 10 patients were unresectable (three for vein
thrombosis and seven for arterial encasement) and 18
were borderline resectable, four of whom had isolated
Toxicity and Response
All 28 group 1 patients were assessable for toxicity.
A detailed report of toxicities registered during our
first institutional protocol has been previously re-
ported.19In summary, at the gemcitabine dose of 100
mg/m2, three of the 15 patients had to stop the
treatment because of toxicities >G3. One patient
needed hospitalization for gastrointestinal toxicity,
and three required a temporary interruption because
of hematological and gastrointestinal toxicity. The
median dose received was 65% of the planned dose.
At the 50 mg/m2dose, all patients completed the
treatment; none needed hospitalization. The median
dose received was 80% of the planned dose; no one
needed dose reduction. Among the five patients
treated with the new protocol, all completed the
induction chemotherapy course but two required
interruption of chemotherapy during radiation. No
deaths or life-threatening toxic reactions were directly
attributable to chemoradiation.
Of the 23 patients who were assessable for re-
sponse, we observed a partial response in five (22%),
TABLE 1. Comparison of groups
62.7 ± 5.6
4.3 ± 1.3
66.5 ± 8.7
2.2 ± 0.6
Size (cm) <0.01
1,053 ± 2,869
742 ± 1,510
are mean ± standarddeviationand range,unless
PANCREATIC RESECTION AFTER CHEMORADIATION
Ann. Surg. Oncol. (? 2006)
stable disease in 14 (61%), and disease progression in
four (17%). All partial responders had a borderline
resectable tumor at presentation. Among the four
patients who had disease progression, three devel-
oped distant metastases and one, a portal thrombosis.
The latter was the only case of evolution from bor-
derline resectability to unresectability because of local
progression. A reduction of serum Ca19.9 was ob-
served in 15 of 21 (71%) patients with an elevated
level at presentation. Normalization (complete mar-
ker response) was evidenced in five patients, three
with partial response and two with stable disease.
Eight patients (29%) were scheduled for operation,
five with a partial response and three with stable
disease but in-range Ca19.9 at restaging (two for
normalization and one in-range at presentation). All
patients explored were successfully resected. Consid-
ering the pretreatment resectability status, seven of
the 18 (39%) borderline resectable cases underwent a
pancreatic resection, including all four cases with
isolated vein stenosis. Only one unresectable patient
(for a limited vein thrombosis) was brought to
resection. Of the 11 borderline resectable patients
who did not undergo operation, one died of sepsis
related to hepatic abscesses before surgical evalua-
tion, four were not available for response, four had
disease progression, and two had stable disease but
were not explored because of stomach infiltration in
one case and increasing serum marker in a head
neoplasm with adhesion to the CA, SMA, and portal
vein in the other.
Surgery and Pathology
The operations performed for group 1 patients
were four duodenopancreatectomies and four total
pancreatectomies. In three cases, a segmental resec-
tion of the superior mesenteric vein was necessary.
One patient died 2 months after operation due to
failure to thrive after a biliary leak treated radiolog-
ically. Complications were registered in four cases:
one biliary leak, one pancreatic fistula, and two de-
layed gastric emptying. No reoperations were neces-
In the same period, group 2 patients underwent 28
pancreatoduodenectomies, 13 total pancreatectomies,
and three distal splenopancreatectomies. Resection of
the mesenteric vein confluence was needed in one
patient. Two patients died in the postoperative peri-
od. One patient was reexplored for hemoperitoneum
associated to a pancreatic fistula and died of septic
shock. Another patient died of systemic candidiasis
without evidence ofabdominal complications.
Morbidity was 36% (16/44). The more frequent
complications were pancreatic fistula (five cases) and
delayed gastric emptying (three cases). Reoperations
were performed in four patients: for pancreatic fistula
in two cases, leak from gastroenteric anastomosis in
one, and segmental colonic necrosis in one.
The analysis of perioperative outcome was per-
formed excluding the three distal pancreatectomies
from group 2, leaving 41 patients with a head cancer
for comparison with group 1 resected cases. Mortal-
ity, morbidity, and transfusion rate were not signifi-
cantly different between the two groups. However,
operative time was 1 hour longer and hospital stay
was 5 days longer for group 2. This comparison is
reported in Table 2.
Final pathology of group 1 patients revealed a
partial pathological response (residual microscopic
cancer foci) in two cases and pT3 ductal adenocar-
cinoma in six cases. No one had a complete patho-
logical response. Regional nodal involvement was
found in two patients (one and two positive nodes,
respectively). Among the patients with enlarged pre-
aortic nodes at presentation, two were resected and
were node-negative at final pathology. No one had
extraregional nodal involvement. The retroperitoneal
margin was positive in one case.
Group 2 patients had a significantly higher node-
positivity rate. The mean number of positive nodes
was 4.5. Clinically unsuspected extraregional positive
nodes (stage IV) were found in five cases. There was a
trend toward a higher rate of retroperitoneal margin
positivity. The pathological features are presented in
Median overall survival and disease-free survival
were 15.4 and 10 months for group 1 and 14 and 8
months for group 2, respectively. Group 1 patients
who underwent resection had significantly longer
TABLE 2. Perioperative outcome
Group 1 resected
Operation type4 Whipple,
6.6 ± 0.5
28 Whipple, 13 total n.s.
5.5 ± 1.0
P. MASSUCCO ET AL.
Ann. Surg. Oncol. (? 2006)
overall survival and disease-free survival compared to
those not resected (P < 0.01). This difference per-
sisted even when only borderline resectable cases were
left for comparison (P < 0.01). Group 2 patients
who received adjuvant treatment had twofold the
survival of untreated patients even if statistical sig-
nificance was not reached (P = 0.17). The detailed
survival figures are shown in Table 4. Figure 1 shows
the survival curves for group 1 patients stratified as
resected and not resected compared to group 2
At the time of analysis, 18 out of 20 group 1 pa-
tients who did not undergo operation died of disease
within 24 months, one died of sepsis after the end of
treatment, and one is alive with peritoneal progres-
sion at 14 months. Operative deaths excluded, among
the seven group 1 resected patients, two died of dis-
ease, whereas five are alive, three with disease recur-
rence and two disease-free at 42 and 17 months.
Among the 42 group 2 primary resected patients, 25
died of disease, eight are alive with disease recur-
rence, and nine are alive and disease-free with a
median follow up of 14 months. No one with positive
extraregional nodes was alive beyond 18 months.
Patterns of recurrence were similar in the two
resected groups. Among group 1 resected patients,
two experienced a local relapse (one isolated and one
local and peritoneal), one peritoneal and two distant.
In group 2 patients, the site of first recurrence was
documented in 26 cases: 12 local (four isolated and
eight with distant metastases), one peritoneal, and 13
distant (two with peritoneal disease).
The univariate analysis of overall survival for all
group 1 and 2 resected patients identified as prog-
nostic factors ‘‘any form of chemotherapy’’ (median
20.6 months for treated vs. 10.9 for untreated
patients, P = 0.05) and R status (median 22.3
months for R0 vs. 9.5 for R1, P < 0.02). R status
was the sole prognostic factor for disease-free sur-
vival (median 8.7 months for R0 vs. 5.5 for R1,
P = 0.04). R status was significantly related to local
recurrence rate (67% for R1 patients vs. 20% for R0,
P < 0.01). No one with retroperitoneal positive
margins survived beyond 24 months.
At presentation, a substantial number of pancre-
atic cancers are already locally advanced. They are a
challenging problem regarding type of treatment,
relative role of chemotherapy and radiotherapy, and
indications for surgical resection. Studies on chemo-
radiation in this clinical setting are largely heteroge-
neous and difficult to compare due to the lack of a
TABLE 3. Pathological findings
Group 1 resected
TABLE 4. Long-term outcome
Resection (n = 8)
(n = 20)
(n = 15)
No adjuvant treatment
(n = 29)
aTime to progression.
OS, overall survival; DFS, disease-free survival.
Prop. Cumulative Survival (Kaplan-Meier)
0 12 2436 48 60
Prop. Cumulative Survival
Group 1 resected
Group 1 not resected
FIG. 1. Overall survival of group 1 patients, stratified as resected
and not resected, compared to group 2.
PANCREATIC RESECTION AFTER CHEMORADIATION
Ann. Surg. Oncol. (? 2006)
clear definition of what is locally advanced or
localized and to the different criteria applied to de-
clare a patient resectable or unresectable. Moreover,
data about the chance to render resectable an unre-
sectable tumor are conflicting and indications for
operation after chemoradiation are not standardized.
A retrospective series of surgically staged unresec-
table tumors from Memorial Sloan Kettering Cancer
Center22showed that it is virtually impossible to
obtain enough downsizing to permit the resection of a
tumor deemed unresectable at laparotomy. They
could resect only one out of 87 patients treated with
various chemoradiation protocols. On the other
hand, a recent paper from Boston University23
reported the resection of all six patients deemed ‘‘not
resectable by conventional Whipple resection’’ during
a pretreatment laparotomy. Other centers have re-
ported a resectability rate ranging 5–43%.5–7,9,13,15,18
This huge difference may be explained by many fac-
tors, such as the staging modalities adopted (surgical
or radiological), inclusion criteria, and the effective-
ness of different therapeutic protocols; but it revolves
mainly around two issues: the definition of locally
advanced cancer in each institution and the indica-
tions for surgical exploration based on radiological
The definition of the patient population is crucial
for evaluating the potential role of surgery for locally
advanced tumors. We applied rigorous radiological
criteria to clearly define our study population. In
agreement with other centers,24,25we consider a high-
quality CT scan as an objective, reproducible, and
reliable technique for pretreatment staging, identifi-
cation of subsets of patients on the basis of vascular
involvement, and definition of resectability status. In
our opinion, a portal thrombosis or a major artery
(SMA or CA) encasement should be regarded as
unresectability criteria. Whereas patients showing an
arterial abutment or in whom a vein resection can be
anticipated from narrowing of the vein lumen should
be classified as marginally resectable. In the present
experience, we succeeded in resecting 39% of bor-
derline resectable tumors while, on the contrary, pa-
tients with truly unresectable neoplasms rarely
became resectable. Our data clearly show that strat-
ification of locally advanced neoplasms according to
resectability status (truly unresectable or borderline
resectable) is mandatory to evaluate the results of
chemoradiation protocols and should be routinely
The inclusion of four patients with an isolated vein
stenosis among the locally advanced tumors was
dictated by the observation that, in our previous
experience, patients requiring a vein resection were
less likely to receive a curative operation, due to the
higher rate of positive retroperitoneal margins in
comparison with patients operated without a vein
resection. The same finding has been recently
reported in a larger series from the M. D. Anderson
Cancer Center.26This prompted us to change our
therapeutic strategy and resulted in a lower rate of
vein resections in the group of primary resected pa-
tients in comparison to our previous reported series
(1/41 vs. 22/100). Our current attitude is to consider
for a program of resection and adjuvant treatment
only patients with intrapancreatic cancers or abutting
at most on the vein axis and to treat with chemora-
diation more advanced stages, including patients with
an isolated vein stenosis, in order to obtain a lower
rate of positive margins.
The second debatable issue concerning the role of
operation in locally advanced cancer is the indication
for surgical exploration on the basis of the restaging
CT scan. In the Memorial Sloan Kettering series,
only patients showing a partial response allowing
anticipation of a curative resection were referred for
surgery.22Contrarily, at Duke University, all patients
without portal thrombosis or arterial encasement
were explored.9Those authors could resect many
patients showing arterial abutment on the CA or
SMA with a low rate of positive margins at final
pathology. We adopted an intermediate policy, con-
sidering for resection not only partial responders at
restaging but also cases of stable disease with nor-
malization of serum markers as a surrogate for
pathological response. In these cases, we did not find
any sign of disease progression precluding resection
The possibility of rendering resectable a locally
advanced cancer by chemoradiation relies on the
capability of downstaging the lesion and sterilizing
the perivascular neoplastic tissue. According to pre-
vious reports,27–29our data show an evident nodal
downstaging in terms of percentage of node-positive
cases, number of positive nodes for case, and type of
node group involvement (regional vs. extraregional)
when locally advanced were compared with localized
tumors. This resulted in a significant lower final
UICC stage for group 1 resected patients despite the
more advanced clinical stage at presentation. More-
over, we observed a trend toward a higher rate of
curative resections (i.e., with negative microscopic
margins) after chemoradiation in comparison to pri-
mary resected cases. Although it is not yet proved if
chemoradiation increases the percentage of negative
30this is a crucial point
P. MASSUCCO ET AL.
Ann. Surg. Oncol. (? 2006)
because margin status is the most powerful predictor
of survival in many series, including the present one.
Therefore, indications for resection should be based
on the estimate of the grade of pathological response.
Unfortunately, the current generation CT is unable to
distinguish between neoplastic and fibrotic tissue. So,
if, on the one hand, CT scan should be considered
adequate to stage patients at presentation, on the
other hand, it is unreliable for evaluating the response
to treatment. One of our patients with a pathological
partial response had stable disease on CT (data not
shown). Data from Duke University show that
posttreatment CT scans overestimate the unresecta-
bility status, thus precluding the chance of resection
in 20% of patients.31We clearly need new radiologi-
cal tools capable of estimating the pathological re-
sponse after combined modality treatment. CT-
positron emission tomography or dynamic RM could
give useful information in the near future. In the
meantime, lacking reliable biological or radiological
restaging criteria, it may be that a more aggressive
policy than the one we adopted (i.e., to explore all
patients without disease progression) could improve
the rate of resections.
As for perioperative results, our data conflict with
other published series. In 1997, Spitz et al.32reported
the postoperative results of a group of 41 localized
tumors resected after chemoradiation compared with
19 primarily resected. They stated that operation
after preoperative treatment was not associated with
increased blood loss or hospital stay. Our experience
is somewhat different. Pancreatectomy after chemo-
radiation shares the same difficulties of resection in
the setting of chronic pancreatitis. In particular, we
encountered in each case a sclerotic reaction that
rendered problematic finding cleavage planes around
major vessels and, moreover, having a good intra-
operative determination of neoplastic spread. These
findings may be due to the more advanced stage of
our patients or the different protocol of treatment.
Besides, even if morbidity was not substantially
higher than in our previous report, patients required
a longer postoperative stay because they are more
fragile compared with those primarily resected and
need a longer time to resume performance status
Survival data were encouraging. Patients resected
after chemoradiation had significantly longer survival
than those not resected. Although they represent two
groups of patients not directly comparable because
the resected group includes less advanced tumors
with better response to treatment, we think that this
survival advantage can be attributed to the direct
impact of resection on the natural history of the
disease. In 2000, Snady et al.33compared a group of
patients treated with chemoradiation with or without
resection to a group of less advanced cancers treated
with resection with or without adjuvant therapy.
They stated that for the first time the negative impact
of a more advanced stage was reversed by changing
the therapeutic strategy. This finding resulted, on the
one hand, from the extraordinary survival of patients
not resected after their chemoradiation protocol
(median 21 months) and, on the other hand, from the
suboptimal survival of their patients resected and
treated with adjuvant therapy (median 16 months). In
our opinion, these data should not be used to propose
chemoradiation as initial treatment for every stage of
pancreatic cancer and to minimize the role of surgery.
We made the same type of comparison. Our data
show that group 1 resected patients had at least the
same prognosis as group 2 treated with adjuvant
chemotherapy. On the contrary, group 2 patients who
had not had adjuvant therapy had the same median
survival as patients treated with chemoradiotherapy
alone. The former finding could be interpreted as a
confirmation of the potential role of preoperative
chemoradiation in reversing the expected negative
impact of a more advanced clinical stage, whereas the
latter underscores the fact that surgery alone is not an
adequate treatment for localized cancer. However,
the most prominent considerations drawn from our
data are the association of resection with improved
survival and the importance of negative margins. In
fact, disease-free survival beyond 24 months was
possible only among patients resected in both groups.
However, patients resected with positive margins had
the same prognosis as patients treated with chemo-
In conclusion, when
according to the pattern of vascular involvement, the
conversion of a truly unresectable cancer to a
resectable one was a rare event with our current
chemoradiotherapy regimes. On the contrary, the
resection of a borderline resectable tumor after
chemoradiation was accomplished in one-third of
these patients with negative margins in most cases.
Resections after chemoradiation in this group of pa-
tients did not increase the overall morbidity but were
more technically demanding than primary resections
for localized cancers, and patients required a longer
postoperative stay. Patients resected after chemora-
diation for a locally advanced tumor had at least the
same survival as patients primarily resected for a
localized cancer. However, in both groups, disease-
free survival beyond 24 months was possible only for
PANCREATIC RESECTION AFTER CHEMORADIATION
Ann. Surg. Oncol. (? 2006)
patients resected with negative margins. The potential
of chemoradiation to increase the rate of microscopic
curative resections for borderline resectable locally
advanced tumors and the criteria for deciding which
patients to refer for surgical exploration need further
1. Moertel CG, Childs DS, Reitemeier RJ, et al. Combined
5-fluorouracil and supervoltage radiation therapy of locally
unresectable gastrointestinal cancer. Lancet 1969; 2:865–7.
2. Moertel CG, Frytak S, Hahn RG, et al. Therapy of locally
unresectable pancreatic carcinoma: a randomised comparison
of high dose (6000 rads) radiation alone, moderate dose radi-
ation (400 rads + 5-fluorouracil), and high dose radiation + 5-
fluorouracil: Gastrointestinal Tumour Study Group. Cancer
3. Klaassen DJ, MacIntyre JM, Catton GE, Engstrom PF,
Moertel CG. Treatment of locally unresectable cancer of the
stomach and pancreas: a randomized comparison of 5-fluoro-
uracil alone with radiation plus concurrent and maintenance 5-
fluorouracil – an Eastern Cooperative Oncology Group study.
J Clin Oncol 1985; 3:373–8.
4. Gastrointestinal Tumor Study GroupTreatment of locally
unresectable carcinoma of the pancreas: comparison of com-
bined-modality therapy (chemotherapy plus radiotherapy) to
chemotherapy alone. J Natl Cancer Inst 1988; 80:751–5.
5. Kamthan AG, Morris JC, Dalton J, et al. Combined modality
therapy for stage II and stage III pancreatic carcinoma. J Clin
Oncol 1997; 15:2920–7.
6. Todd KE, Gloor B, Lane JS, Isacoff WH, Reber HA. Resec-
tion of locally advanced pancreatic cancer after downstaging
with continuous-infusion 5-fluorouracil, mitomycin-C, leu-
covorin, and dipyridamole. J Gastrointest Surg 1998; 2:159–66.
7. Bajetta E, Di Bartolomeo M, Stani SC, et al. Chemoradio-
therapy as preoperative treatment in locally advanced unre-
sectable pancreatic cancer patients: results of a feasibility
study. Int J Radiat Oncol Biol Phys 1999; 45:285–9.
8. Kornek GV, Schratter-Sehn A, Marczell A, et al. Treatment of
unresectable, locally advanced pancreatic adenocarcinoma
with combined radiochemotherapy with 5-fluorouracil, leuc-
overin and cisplatin. Br J Cancer 2000; 82:98–103.
9. White RR, Hurwitz HI, Morse MA, et al. Neoadjuvant
chemoradiation for localized adenocarcinoma of the pancreas.
Ann Surg Oncol 2001; 8:758–65.
10. Mehta VK, Poen JC, Ford JM, et al. Protracted venous infu-
sion 5-fluorouracil with concomitant radiotherapy compared
with bolus 5-fluorouracil for unresectable pancreatic cancer.
Am J Clin Oncol 2001; 24:155–9.
11. Safran H, Moore T, Iannitti D, et al. Paclitaxel and concurrent
radiation for locally advanced pancreatic cancer. Int J Radiat
Oncol Biol Phys 2001; 49:1275–9.
12. Blackstock AW, Melin SA, Butler JM, et al. Irinotecan/gem-
citabine followed by twice-weekly gemcitabine/radiation in
locally advanced pancreatic cancer. Oncology (Huntingt) 2002;
13. Crane CH, Abbruzzese JL, Evans DB, et al. Is the therapeutic
index better with gemcitabine-based chemoradiation than with
5-fluorouracil-based chemoradiation in locally advanced pan-
creatic cancer? Int J Radiat Oncol Biol Phys 2002; 52:1293–302.
14. Blackstock AW, Tepper JE, Niedwiecki D, Hollis DR, Mayer
RJ, Tempero MA. Cancer and leukemia group B (CALGB)
89805: phase II chemoradiation trial using gemcitabine in
patients with locoregional adenocarcinoma of the pancreas. Int
J Gastrointest Cancer 2003; 34:107–16.
15. Aristu J, Canon R, Pardo F, et al. Surgical resection after
preoperative chemoradiotherapy benefits selected patients with
unresectable pancreatic cancer. Am J Clin Oncol 2003; 26:30–6.
16. Rich T, Harris J, Abrams R, et al. Phase II study of external
irradiation and weekly paclitaxel for nonmetastatic, unresec-
table pancreatic cancer: RTOG-98-12. Am J Clin Oncol 2004;
17. Chung HW, Bang SM, Park SW, et al. A prospective ran-
domized study of gemcitabine with doxifluridine versus pac-
litaxel with doxifluridine in concurrent chemoradiotherapy for
locally advanced pancreatic cancer. Int J Radiat Oncol Biol
Phys 2004; 60:1494–1501.
18. Wilkowski R, Thoma M, Schauer R, Wagner A, Heinemann
V. Effect of chemoradiotherapy with gemcitabine and cisplatin
on locoregional control in patients with primary inoperable
pancreatic cancer. World J Surg 2004; 28:1011–8.
19. Magnino A, Gatti M, Massucco P, et al. Phase II trial of
primary radiation therapy and concurrent chemotherapy for
patients with locally advanced pancreatic cancer. Oncology
20. Capussotti L, Massucco P, Ribero D, Vigano ` L, Muratore A,
Calgaro M. Extended lymphadenectomy and vein resection for
pancreatic head cancer. Outcomes and implications for ther-
apy. Arch Surg 2003; 138:1316–22.
21. Sobin LH, Wittekind Ch, eds. UICC TNM Classification of
Malignant Tumours, 6th ed. New York: John Wiley & Sons,
22. Kim HJ, Czischke K, Brennan MF, Conlon KC. Does neo-
adjuvant chemoradiation downstage locally advanced pancre-
atic cancer? J Gastrointest Surg 2002; 6:763–9.
23. Wanebo HJ, Glicksman AS, Vezeridis MP, et al. Preoperative
chemotherapy, radiotherapy, and surgical resection of locally
advanced pancreatic cancer. Arch Surg 2000; 135:81–8.
24. Fuhrman GM, Charnsangavej C, Abruzzese JL, et al. Thin-
section contrast-enhanced computed tomography accurately
predicts the resectability of malignant pancreatic neoplasms.
Am J Surg 1994; 167:104–11.
25. Freeny PC, Traverso LW, Ryan JA. Diagnosis and staging of
pancreatic adenocarcinoma with dynamic computed tomog-
raphy. Am J Surg 1993; 165:600–6.
26. Tseng JF, Raut CP, Lee JE, et al. Pancreaticoduodenectomy
with vascular resection: margin status and survival duration. J
Gastrointest Surg 2004; 8:935–50.
27. Breslin TM, Hess KR, Harbison DB, et al. Neoadjuvant che-
moradiotherapy for adenocarcinoma of the pancreas: treat-
ment variables and survival duration. Ann Surg Oncol 2001;
28. Moutardier V, Magnin V, Turrini O, et al. Assessment of
pathologic response after preoperative chemoradiotherapy and
surgery in pancreatic adenocarcinoma. Int J Radiat Oncol Biol
Phys 2004; 60:437–43.
29. White RR, Xie HB, Gottfried MR, et al. Significance of his-
tological response to preoperative chemoradiotherapy for
pancreatic cancer. Ann Surg Oncol 2005; 12:214–21.
30. Pingpank JF, Hoffman JP, Ross EA, et al. Effect of preoper-
ative chemoradiotherapy on surgical margin status of resected
adenocarcinoma of the head of the pancreas. J Gastrointest
Surg 2001; 5:121–30.
31. White RR, Paulson EK, Freed KS, et al. Staging of pancreatic
cancer before and after neoadjuvant chemoradiation. J Gas-
trointest Surg 2001; 5:626–33.
32. Spitz FR, Abbruzzese JL, Lee JE, et al. Preoperative and
postoperative chemoradiation strategies in patients treated
with pancreaticoduodenectomy for adenocarcinoma of the
pancreas. J Clin Oncol 1997; 15:928–37.
33. Snady H, Bruckner H, Cooperman A, Paradiso J, Kiefer L.
Survival advantage of combined chemoradiotherapy compared
with resection as the initial treatment of patients with regional
pancreatic carcinoma. An outcomes trial. Cancer 2000; 89:314–
P. MASSUCCO ET AL.
Ann. Surg. Oncol. (? 2006)