ORIGINAL ARTICLES

Prolonged Survival of Initially Unresectable Hepatic

Colorectal Cancer Patients Treated With Hepatic
Arterial Infusion of Oxaliplatin Followed by
Radical Surgery of Metastases
Diane Goéré, MD,* Isabelle Deshaies, MD,* Thierry de Baere, MD, PhD,† Valérie Boige, MD,‡
David Malka, MD, PhD,‡ Frédéric Dumont, MD,* Clarisse Dromain, MD,† Michel Ducreux, MD, PhD,‡
and Dominique Elias, MD, PhD*

ing on their clinical status. Nevertheless, there is increasing evidence

Purpose: The aim of this study was to analyze the impact of hepatic arterial
that the frontier between curatively intended treatment and palliative
infusion (HAI) of oxaliplatin with systemic 5-Fluorouracil and leucovorin on
therapy can sometimes overlap in initially unresectable patients
patients with isolated unresectable liver metastases.
whose disease becomes resectable after a good response to chemo-
Patients and Methods: A total of 87 patients treated in our hospital with
therapy.1,5– 8
HAI of oxaliplatin with systemic 5-Fluorouracil and leucovorin for isolated
Hepatic arterial infusion (HAI) of chemotherapy has been
unresectable colorectal liver metastases from May 1999 to May 2007 were
investigated in the palliative setting mostly for patients presenting
extracted from a prospective database and analyzed. The resectability rate,
with metastases exclusively or dominantly affecting the liver and
perioperative findings, postoperative outcomes, and long-term follow-up
preserved hepatic function.9 –12 Two specific principles underlie the
were evaluated.
rationale for using this more invasive route. The first is that hepatic
Results: HAI was delivered after failure of previous systemic chemotherapy
metastases are predominantly vascularized by the hepatic artery
in 69 patients (79%). The main criterion for unresectability was massive liver
compared with the mixed arterial and portal blood supply of the
involvement (86% of patients). Most patients had synchronous (85%),
surrounding parenchyma. Arterial infusion, therefore, increases in-
bilateral metastases (89%). The median number of HAI courses was 8
trametastatic concentration of chemotherapy. The second principle
(0 –25). About 31 patients experienced technical catheter-related problems,
is to reinforce local activity and decrease systemic toxicity. That is
which were responsible for withdrawal of HAI in only 7 patients (8%).
why, the drugs used in HAI yield a better therapeutic index due to
Finally, a total of 23 patients (26%) were operated on, and resection or
high first-pass hepatic extraction and high systemic clearance.
radiofrequency ablation was performed in 21 patients (24%). No postoper-
Two previous studies demonstrated the efficacy of HAI of
ative mortality was observed and the morbidity rate was 35%. Five-year
oxaliplatin associated with systemic 5-fluorouracil and leucovorin
overall survival was 56% in the surgery group versus none in the nonsurgery
infusion (modified LV5FU2 protocol). The overall response rate was
group (P ⬍ 0.0001). After a median follow-up of 63 months, intrahepatic
64% in a first-line setting and 62% in a salvage setting (failure of at
recurrence occurred in 10 patients among the 23 operated patients.
least one line of chemotherapy).12,13 However, the rate of resection
Conclusions: HAI of oxaliplatin with systemic 5-Fluorouracil and leucov-
of initially unresectable CRCLM after oxaliplatin HAI and systemic
orin offers a second chance to remove initially unresectable isolated colo-
modified LV5FU2 was not assessed.
rectal liver metastases in 24% of patients, and appears to be more efficient
The goal of the present study was to evaluate the rate of
when performed as first-line therapy. Long-term overall survival can be
resectability after HAI of oxaliplatin associated with systemic mod-
obtained with this approach.
ified LV5FU2 in patients with liver metastases, initially treated
(Ann Surg 2010;251: 686 – 691) palliatively and to analyze the immediate and long-term results of
surgery.

PATIENTS AND METHODS

T he liver is the most common site of metastases, affecting as

many as 60% of patients with colorectal primary cancer. Meta-
static spread is confined to the liver in approximately one-third of
cases. Liver resection and tumor ablation are currently the only
available curative treatments of colorectal cancer liver metastases
(CRCLM) with 5-year survival ranging from 28% to 45%.1– 4
Unfortunately, curatively intended treatment can only be offered to
10% to 20% of the patients. The remaining majority of patients
receive palliative chemotherapy or the best supportive care depend-
From the Departments of *Surgical Oncology, †Radiology, and ‡Medical Oncol-
ogy, Institut Gustave Roussy, Villejuif, Cedex, France.
Reprints: Diane Goéré, MD, Department of Surgical Oncology, Institut Gustave
Roussy, 39 Rue Camille Desmoulins, 94805 Villejuif-Cedex, France. E-mail:
goere@igr.fr.
Copyright © 2010 by Lippincott Williams & Wilkins
ISSN: 0003-4932/10/25104-0686
DOI: 10.1097/SLA.0b013e3181d35983
Inclusion Criteria
Entries from May 1999 to May 2007 in an institutional
prospectively established database were reviewed. Inclusion criteria
were unresectable CRCLM and treatment with HAI of oxaliplatin
and systemic modified LV5FU2 administered in our institution.
Unresectability criteria were as follows: an insufficient estimated
future liver remnant (FLR), vascular invasion, progression under
chemotherapy, or an unresectable extrahepatic lesion. The presence
of resectable and limited extrahepatic metastases at one site, such as
a solitary pulmonary metastasis, was not a contraindication for this
kind of treatment.
Patients had to be younger than 75 years of age, have a WHO
performance status ⱕ2, satisfactory hematological results (WBC ⱖ
4 ⫻ 109/L, neutrophils ⱖ1.5 ⫻ 109/L, and platelets ⱖ100 ⫻ 109/L),
satisfactory biochemical results (serum creatinine ⬍130 ␮mol/L and
bilirubin ⬍35 ␮mol/L), and measurable disease (hepatic lesion
diameter ⱖ20 mm). Patients with peripheral neuropathy, heart

686 | www.annalsofsurgery.com Annals of Surgery • Volume 251, Number 4, April 2010

Annals of Surgery • Volume 251, Number 4, April 2010
failure, respiratory insufficiency, severe coronary artery disease, or
failure of other organs likely to contraindicate treatment were not
included.
Data were collected prospectively for patients who underwent
surgery and retrospectively for patients treated with chemotherapy
alone.
HAI Protocol
Patients with unresectable CRCLM, without extrahepatic dis-
ease, received HAI of oxaliplatin after failure of a first-line of
systemic chemotherapy, or as first-line therapy during prospective
studies. A catheter connected to a subcutaneous chamber (Celsite
T202F, 6,5F, B Braun) was installed either during open surgery or
interventional radiology to exclusively perfuse the hepatic artery and
liver parenchyma. Any additional left hepatic artery originating from
the left gastric artery or right hepatic artery originating from the
superior mesenteric artery was ligated or embolized. Isolated liver
perfusion was first verified with fluorescein during surgery or
standard angiography during interventional radiology. An isotope
perfusion study was also performed before each chemotherapy
infusion. The chemotherapy protocol consisted of an initial HAI
bolus of 100 mg/m2 oxaliplatin, 200 mg/m2 intravenous (i.v.)
leucovorin, and 400 mg/m2 i.v. 5FU over a 2-hour period followed
by an infusion of 2400 mg/m2 i.v. 5FU over a 2-day period
(modified LV5FU2 protocol). Chemotherapy was repeated every 2
weeks, in our institution. Chemotherapy doses could be modified in
the event of local or systemic toxicity. The HAI treatment was
stopped if major toxicity, serious technical catheter-related prob-
lems, or disease progression occurred.
Surgical Techniques
Chemotherapy was stopped at least 4 weeks prior to surgery.
Preoperative evaluation included thoracic and abdominal computed
tomography (CT) scans for all patients. Since 2003, contrast-en-
hanced ultrasound has also been performed in most cases. The
positron emission tomography scan and MRI were added when
considered necessary. In addition, liver function was assessed in all
patients by the indocyanin green clearance (ICG) test, as described
by Makuuchi et al.14 Patients were considered to have resectable
disease when the tumor response allowed resection and/or radiofre-
quency (RF) ablation of all CRCLM seen on imaging (R0 resection),
with a sufficient FLR. Hepatic resections were performed by 2
senior surgeons. Bleeding was controlled using intermittent pedicle
clamping and low central venous pressure. Parenchyma dissection
was carried out using the Kelly crush and bipolar dissection. Radio-
frequency ablation was preferred over extensive anatomic resection
for small centrally located lesions (⬍2.5 cm), to preserve hepatic
parenchyma in these heavily treated patients.15 Previously detect-
able liver metastases that were no longer visible after chemotherapy
(or “missing metastases”) were resected if they could be precisely
localized and if the estimated FLR was sufficient.15 A major hepa-
tectomy was defined as resection of more than 3 hepatic segments.
Follow-Up
Patients on chemotherapy were followed up after each che-
motherapy course and tumor response was evaluated every 3
months. Tumor response to chemotherapy was evaluated according
to RECIST criteria16 by abdominal CT scan or, if not assessable on
CT scan, by liver MRI or contrast-enhanced ultrasound. Resectabil-
ity was then reevaluated by a multidisciplinary team composed of
medical oncologists, surgeons, radiologists, and radiotherapists.
Operated patients were followed up every 3 months and
screened for postoperative complications or tumor recurrence by
clinical evaluation, abdominal CT scan, and tumor marker determi-
© 2010 Lippincott Williams & Wilkins
Colorectal Liver Metastases Initially Unresectable
nation (carcinoembryonic antigen and CA 19.9) during the first 2
years, then every 6 months thereafter.
Statistical Evaluation
The statistical analysis was performed using the ␹2 test for
quantitative values and the Student t test for qualitative values. A P
value of less than 0.05 was considered statistically significant.
Survival was determined using the Kaplan-Meier method. Overall
survival was calculated from the date of the diagnosis of CRCLM
for all patients. In addition, overall survival and progression-free
survival were calculated from resection of CRCLM, for operated
patients. All statistical calculations were performed using StatView
software.
RESULTS
Patient Characteristics
A total of 87 patients with unresectable isolated CRCLM
were included in the final analysis. Of total, 60 patients were treated
between May 1999 and December 2003, and 27 between January
2004 and May 2007. Their demographic characteristics and clinical
features are described in Table 1.
The patients had advanced liver disease with synchronous
metastases in 74 (85%) and bilateral, centrohepatic, or unilateral
lesions after a previous lobectomy for metastases in 78 patients
(89%). The median number of lesions was 7 (1– 60) and the median
size of the largest lesion was 50 mm (10 –150). Eighteen patients
(20.6%) had previously been operated on for CRCLM. The main
reason for unresectability was an insufficient estimated FLR in 75
patients (86%). Other reasons for palliative treatment alone were
vascular invasion (6 patients), progression under chemotherapy (4
patients), and the appearance of unresectable extrahepatic metastasis
(2 patients).
Hepatic Arterial Infusion and Systemic
Chemotherapy
All 87 patients were fitted with an intraarterial catheter
connected to a subcutaneous chamber. The catheter was installed
during open surgery in 47 patients and during interventional radiol-
ogy in the remaining 40 patients. This last technique has been used
by our team since 2002 and is now preferred over open surgery if
patients do not require surgery for another reason.
About 79% of patients had received prior treatment with
either systemic oxaliplatin or irinotecan, thus explaining the long
TABLE 1. Demographic Characteristics and Clinical
Features of 87 Patients
Characteristics Values
Sex
Male 44
Female 43
Age
Median (range) yr 57 (32–77)
Primary tumor
Colon 62
Rectum 23
2 synchronous sites 2
CEA—median (range) ␮mol 102 (1–44,130)
Previous chemotherapy
Chemonaive 19
Previous chemotherapy 68
www.annalsofsurgery.com | 687
Goéré et al Annals of Surgery • Volume 251, Number 4, April 2010

interval between the diagnosis of CRCLM and the beginning of HAI

(median interval of 220 days 关13–1161兴). The median number of TABLE 3. Preoperative and Operative Findings in 23
HAI courses completed was 8 (0 –25), leading to a response rate of Patients
55% in this heavily treated group. Characteristics Values
Technical complications occurred in 31 patients but the prob-
Interval between diagnosis and surgery
lem was solved and HAI was resumed in 24 out of 31 patients.
Median (range) mo 14.7 (7.5–41.2)
Therefore, the treatment was stopped for technical reasons in only 7
patients (8%). The other reasons for discontinuation of HAI were No. preoperative HAI courses
mainly disease progression in 48%, chemotherapy-related systemic Median (range) 9 (4–20)
toxicity in 21%, surgery in 21%, and a poor status in 7%. One Previous liver procedure
patient developed liver insufficiency with portal hypertension after None 18
12 courses of HAI. All chemotherapy was stopped and hepatic One 4
function was restored, with an ICG 15-minute retention rate decreas- Two 1
ing from 44% to 6.5% and normalization of all signs of portal Preoperative portal embolization
hypertension over a 2-month period. We also noted that 3 clinically Yes 5
asymptomatic patients had an elevated preoperative ICG 15-minute
No 18
retention rate that returned to normal after 1 to 4 months of
Type of surgery
chemotherapy.
Major hepatectomy 10
Resectability Rate and Postoperative Outcomes Minor hepatectomy 4
After a median number of 9 (4 –20) HAI courses, 23 patients Radiofrequency ablation alone 7
were operated on with a curative intent, which represents 26% of the Exploration without resection 2
initial group. This subgroup of patients did not differ significantly Peroperative transfusion
from the nonoperated group in terms of demographic characteristics
Yes 1
(data not shown) and CRCLM characteristics (Table 2). However, 2
No 22
characteristics were significantly different between these 2 groups.
First, HAI was the initial treatment in 43.5% of operated patients (10
out of 23) and 14% of the nonoperated patients (9 out of 64; P ⫽
0.004). Second, the catheter was installed during open surgery in TABLE 4. Postoperative Findings in 23 Operated Patients
74% of the operated patients (17 out of 23) and in 47% of the
nonoperated patients (30 out of 64; P ⫽ 0.01). No. Patients
All visible metastases were resected or ablated in 21 patients, General complication
but disease proved to be unresectable at surgical exploration in 2 Pleural effusion 4
patients so that potentially curative surgery was ultimately per- Atelectasis 2
formed in 24% of the initial group. The resection rate was 53% (10 Urinary infection 1
out of 19) in patients who received HAI in first line and 19% (13 out
Liver complication
of 68) in patients who received HAI after failure of prior systemic
chemotherapy (P ⫽ 0.008). A major hepatectomy was performed in Biliary leak 1
10 patients and 7 patients exclusively underwent multiple RF abla- Biliary stenosis 1
tions. Ten patients had a combination of hepatic resection (n ⫽ 8 Infected collection 1
major, n ⫽ 2 minor) and RF ablation. Pathologic samples were Noninfected collection 1
available for 16 patients. Complete tumor necrosis was found in 3 Liver insufficiency 0
out of 16 (18.7%) cases and more than 80% of necrosis in another Percutaneous drainage 4
2 cases (preoperative and operative findings are detailed in Table 3). Reoperation 0
No mortality occurred during the first 3 postoperative months 90-d total morbidity 8
and the morbidity rate was 35%. No patients were reoperated on but 90-d mortality 0
4 required percutaneous drainage (1 for a pleural effusion, 2 for a
hepatic collection, and 1 for biliary stenosis). Of note, no liver
failure occurred (Table 4).
Adjuvant treatment, HAI (n ⫽ 10) and systemic chemother-
apy (n ⫽ 4), was administered to 61% of the operated patients. Long-Term Follow-Up
Median follow-up was 74 months (23–117) for the entire
group and 63 months (17–93) for the 23 surgical patients.
TABLE 2. Comparison Between CRCLM Characteristics of Three-year overall survival was 73% for the 23 surgical patients
Nonoperated Patients and Operated Patients versus 16% for the nonsurgical patients, and 5-year overall
survival was 56% and 0%, respectively (Figs. 1, 2; P ⬍ 0.0001).
Nonoperated Operated
CRCLM Characteristics N ⴝ 64 N ⴝ 23 P
In the 23 surgical patients, median overall survival was 41.9
months, and 1- and 2-year disease-free survival rates were 28%
Synchronous 54 20 0.96 and 10%, respectively (Fig. 2). About 10 patients developed
Recurrence 14 5 0.78 liver-only recurrences and 11 patients developed systemic or
Bilateral 59 19 0.37 combined systemic and intrahepatic recurrences. Five patients
Number—median (range) 8 (1–60) 5 (1–38) 0.22 whose recurrence arose postoperatively underwent subsequent
Size—median (range) mm 50 (10–150) 53 (24–140) 0.88 curatively intended procedures, 1 for a liver-only recurrence, 1
for an extrahepatic-only recurrence, and 3 for systemic recur-
CRCLM indicates colorectal cancer liver metastasis.
rences.

688 | www.annalsofsurgery.com © 2010 Lippincott Williams & Wilkins

Annals of Surgery • Volume 251, Number 4, April 2010
FIGURE 1. Overall survival of nonoperated (n ⫽ 64) and op-
erated (n ⫽ 23) patients calculated from the date of diagno-
sis of liver metastases.
FIGURE 2. Overall survival and disease-free survival for the
23 operated patients calculated from the date of surgery.
DISCUSSION
HAI of oxaliplatin and systemic LV5FU2 were able to
transform isolated unresectable CRCLM into resectable lesions
in 26% of the cases. After surgical exploration, 21/23 patients
underwent resection or RF ablation of metastases, leading to
potentially curative therapy in 24% of the initial group. As a
result, prolonged survival was achieved in the surgical group
compared with that in the nonsurgical group, with 5-year survival
attaining 56%.
The administration of HAI with oxaliplatin associated with
systemic 5-FU is decided during a multidisciplinary meeting,
after failure of first-line systemic chemotherapy, or during first-
line therapy in a part of a prospective study, in patients who fulfill
the following inclusion criteria: patients with no extrahepatic
disease (or only one site accessible to resection), a good perfor-
mance status, and who receive chemotherapy in our institute.
Therefore, few patients (approximately 10%) can benefit from
this treatment, because of rigid inclusion criteria and due to
progress in chemotherapy during the last years. Regarding the
distribution of patients as a function of time, HAI was more
frequently administered up to 2004, when targeted therapies
appeared to be efficient and were preferentially administered.
Currently, patients with unresectable CRCLM receive HAI with
oxaliplatin associated with systemic 5-FU and targeted therapy,
during first-line therapy as part of a prospective study, or after
failure of systemic chemotherapy. HAI is also administered to
enhance response to chemotherapy, when CRCLM is still unre-
sectable but potentially convertable after chemotherapy. In addi-
tion, after failure of a first-line, HAI with oxaliplatin should be
© 2010 Lippincott Williams & Wilkins
Colorectal Liver Metastases Initially Unresectable
used, preferentially in patients whose tumors harbor a KRAS
mutation and who cannot receive cetuximab or panitumumab.
The 2 main drawbacks of HAI reported in the literature are its
invasiveness, which is associated with a non-negligible technical
failure rate and direct hepatotoxicity related to a high concentration
of chemotherapy. HAI is certainly a more invasive technique than
the standard peripheral approach for delivering chemotherapy. Some
authors have reported a high technical failure rate. In a randomized
controlled EORTC study including 290 patients, comparing HAI of
5FU to systemic 5FU, 50 patients never started HAI and 39 had to
stop before the completion of the sixth course due to catheter failure.
The median number of courses received was 2.17 In our study,
technical complications were responsible for discontinuation of HAI
in only 7 patients, representing 8% of the group, and the median
number of courses received was 8. However, our complication rate
of 36% was not dissuasive. The 4 most frequent complications were
catheter-related thrombosis, catheter displacement, extrahepatic per-
fusion, and infection, accounting for 84% of all complications. Two
of these complications, ie, catheter-related thrombosis and infection,
can also occur with the standard implantable central venous access.
Furthermore, the complications were mostly minor and relatively
easy to resolve with simple maneuvers. However, it must be em-
phasized that patients need to be closely monitored and a skilled
technical team is mandatory if HAI is to be performed safely and
efficiently.
The second disadvantage associated with HAI is hepatotox-
icity. Most studies on HAI used FUDR for intrahepatic perfusion.
Hepatotoxicity was mostly reported with this agent. Biliary sclerosis
is the feared complication, and it can occur in as many as 18% to
29% of patients receiving intrahepatic FUDR.10,18,19 Studies on HAI
with oxaliplatin are scarce, but such biliary toxicity does not seem to
occur. In our study, only one patient developed liver insufficiency
with portal hypertension, which was reversible upon discontinuation
of HAI. Unfortunately, no liver biopsy specimen was obtained
during this episode of disturbed liver function. We suppose that
it was related to reversible hepatic sinusoidal obstruction, as
described with systemic oxaliplatin.20 Nonetheless, hepatotoxic-
ity was moderate, rapidly reversible, and did not result in post-
operative liver failure.
An aspect that merits discussion is the significant difference
in the resectability rate according to whether HAI was performed as
first-line therapy or after failure of chemotherapy. More than half of
the patients who had received first-line HAI were operated on,
which is a very high proportion. Indeed, in a recent study that
evaluated the long-term outcome of such patients treated with
first-line FOLFOXIRI,8 19% underwent a radical resection and
achieved a 5-year survival rate of 42%. Disease in patients who
respond to chemotherapy straight-away, is probably more likely to
become resectable. The choice of first-line HAI therefore, seems
to be crucial for enhancing the likelihood for resection.
For the 78% of patients with advanced disease who had
previously failed on systemic chemotherapy (including failure on
systemic oxaliplatin), the 19% resection rate obtained after HAI of
oxaliplatin, and systemic LV5FU2 is good. Even if we attempt to
define clear nonresectability criteria, the resectability status is some-
what subjective, which makes it difficult to compare this rate with
others. Nevertheless, our study can be favorably compared with
recent studies. Adam et al reported a 13% rate for secondary
resectability in patients who were down-staged following various
chemotherapy regimens (mainly oxaliplatin based).1 He subse-
quently reported a 7% resectability rate after cetuximab rescue
therapy for patients refractory to conventional systemic chemother-
apy.5 Meric et al obtained a 6% resection rate after HAI of FUDR.21
The conclusions concerning resectability and postoperative out-
www.annalsofsurgery.com | 689
Goéré et al
comes in our study are however clear and solid. It was conducted in
a single center, with the same multidisciplinary team evaluating
patients from the beginning right through to the end. As the resect-
ability status is to some extent subjective, this is an important point
to consider when conducting and evaluating studies on this issue.
For example, in the study on hepatic resection after cetuximab
rescue for CRCLM, Adam et al described that they included 133
patients from their own hospital and subsequently resected 6.7% of
these patients, while 88.8% of the 18 patients who came from other
centers underwent subsequent resection. We wonder whether all
external centers applied the same unresectability criteria as
Adam’s team.5 Nevertheless, the surgeons had to be “aggressive”
to completely remove all visible metastases, as shown by all the
procedures performed before or during surgery, including portal
vein embolization, RF ablations with biliary cooling, and tran-
sradiofrequency hepatectomies, which are not a standard proce-
dure for patients with straightforward resectable disease. Our
policy is to treat all the visible lesions to achieve an R0 resection.
However, more than half of the patients underwent a minor
hepatectomy and/or RF. Indeed, in these patients with initially
unresectable lesions, limited resections and/or RF allowed us to
treat most often bilobar residual CRCLM, which had dramati-
cally responded to chemotherapy, leaving a sufficient volume of
remnant liver, which is not possible with a major hepatectomy
and ablation of lesions located in the other hemi-liver.
The postoperative outcomes were simple with no 90-day
postoperative mortality and a 90-day morbidity rate of 35%. Other
studies conducted on heavily treated patients have reported similar
results.5,21–23 In a study published in 2008, there was no difference
in postoperative mortality between the group treated with neoadju-
vant HAI and the group without neoadjuvant therapy.22 Major
hepatic surgery can obviously be performed in heavily treated
patients with acceptable postoperative outcomes. The only precau-
tions required are a good evaluation of liver function, which was
performed using the ICG test, and an interval of at least 4 weeks
between the end of HAI and surgery.
Considering the extent of hepatic involvement in the 23
operated patients, long-term survival after HAI treatment and sub-
sequent surgery is encouraging with a 5-year survival of 56%. On
the other hand, both extra- and intrahepatic recurrences remain high.
Other studies on secondary resection for initially unresectable
CRCLM have reported similar high recurrence rates. In studies by
Adam et al1 and Meric et al,21 80% and 83% of the patients relapsed,
respectively. There are some options for dealing with this high
recurrence rate. First, the patient must be followed up carefully and
each recurrence must be analyzed by a multidisciplinary team for
potential resection. In our study, 5 patients underwent subsequent
resections or ablations for metastases. In a study of Adam et al, more
than half of the patients who relapsed in the liver were operated on.1
Second, adjuvant chemotherapy must be carefully assessed in this
selected subgroup of patients. Adjuvant HAI seems to be interesting
for hepatic recurrences, as shown in a previous study.15 Finally, the
weak point of the regimen presented here is the lack of a strong
systemic chemotherapeutic agent. Adding stronger systemic chemo-
therapy to HAI (irinotecan and/or targeted therapies) might lower
the failure rate.
CONCLUSION
HAI with oxaliplatin and systemic modified LV5FU2 offers a
real chance for resection to a substantial subgroup of patients with
initially unresectable liver-only CRCLM, and appears to be more
efficient when performed as first-line therapy. With this ap-
proach, long-term overall survival can be achieved. However, the
690 | www.annalsofsurgery.com
Annals of Surgery • Volume 251, Number 4, April 2010
rate of postoperative recurrences is still high. Chemotherapy
intensification, notably with targeted therapy, is definitely a
strategy worth exploring.
ACKNOWLEDGMENT
The authors thank Lorna Saint Ange for editing.
REFERENCES
1. Adam R, Delvart V, Pascal G, et al. Rescue surgery for unresectable
colorectal liver metastases downstaged by chemotherapy. Ann Surg. 2004;
240:644 – 657.
2. Malik HZ, Prasad KR, Halazun KJ, et al. Preoperative prognostic score for
predicting survival after hepatic resection for colorectal liver metastases. Ann
Surg. 2007;246:806 – 814.
3. Nordlinger B, Guiguet M, Vaillant JC, et al. Surgical resection of colorectal
carcinoma metastases to the liver: a prognostic scoring system to improve
case selection, based on 1568 patients. Association francaise de chirurgie.
Cancer. 1996;77:1254 –1262.
4. Fong Y, Fortner J, Sun RL, et al. Clinical score for predicting recurrence after
hepatic resection for metastatic colorectal cancer: analysis of 1001 consecu-
tive cases. Ann Surg. 1999;230:309 –318.
5. Adam R, Aloia T, Levi F, et al. Hepatic resection after rescue cetuximab
treatment for colorectal liver metastases previously refractory to conventional
systemic therapy. J Clin Oncol. 2007;25:4593– 4602.
6. Leonard GD, Brenner B, Kemeny N. Neoadjuvant chemotherapy before liver
resection for patients with unresectable liver metastases from colorectal
carcinoma. J Clin Oncol. 2005;23:2038 –2048.
7. Masi G, Loupakis F, Pollina L, et al. Long-term outcome of initially
unresectable metastatic colorectal cancer patients treated with 5-fluorouracil/
leucovorin, oxaliplatin, and irinotecan (FOLFOXIRI) followed by radical
surgery of metastases. Ann Surg. 2009;249:420 – 425.
8. Falcone A, Ricci S, Brunetti I, et al. Phase III trial of infusional fluorouracil,
leucovorin, oxaliplatin, and irinotecan (FOLFOXIRI) compared with infu-
sional fluorouracil, leucovorin, and irinotecan (FOLFIRI) as first-line treat-
ment for metastatic colorectal cancer: the Gruppo Oncologico Nord Ovest.
J Clin Oncol. 2007;25:1670 –1676.
9. Kemeny N, Niedzwiecki D, Hollis D, et al. Hepatic arterial infusion versus
systemic therapy for hepatic metastases from colorectal cancer: a randomized
trial of efficacy, quality of life, and molecular markers (CALGB 9481). J Clin
Oncol. 2006;4:1395–1403.
10. Rougier P, Laplanche A, Huguier M, et al. Hepatic arterial infusion of
floxuridine in patients with liver metastases from colorectal carcinoma:
long-term results of a prospective randomized control trial. J Clin Oncol.
1992;10:1112–1118.
11. Link KH, Pillasch J, Formentini A, et al. Downstaging by regional chemo-
therapy of non-resectable isolated colorectal liver metastases. Eur J Surg
Oncol. 1999;25:381–388.
12. Boige V, Malka D, Elias D, et al. Hepatic arterial infusion of oxaliplatin and
intravenous LV5FU2 in unresectable liver metastases from colorectal cancer
after systemic chemotherapy failure. Ann Surg Oncol. 2007;14:3188 –3194.
13. Ducreux M, Ychou M, Laplanche A, et al. Hepatic arterial oxaliplatin
infusion plus intravenous chemotherapy in colorectal cancer with inoperable
hepatic metastases: a trial of the gastrointestinal group of the Federation
Nationale des Centres de Lutte Contre le Cancer. J Clin Oncol. 2005;23:
4881– 4887.
14. Miyagawa S, Makuuchi M, Kawasaki S, et al. Criteria for safe hepatic
resection. Am J Surg. 1995;169:589 –594.
15. Elias D, Goere D, Boige V, et al. Outcome of posthepatectomy-missing
colorectal liver metastases after complete response to chemotherapy: impact
of adjuvant intra-arterial hepatic oxaliplatin. Ann Surg Oncol. 2007;14:3188 –
3194.
16. Therasse P, Arbuck SG, Eisenhauer EA, et al. New guidelines to evaluate
the response to treatment in solid tumors. European Organization for
Research and Treatment of Cancer, National Cancer Institute of the United
States, National Cancer Institute of Canada. J Natl Cancer Inst. 2000;92:
205–216.
17. Kerr D, McArdle CS, Ledermann J, et al. Intrahepatic arterial versus intra-
venous fluorouracil and folinic acid for colorectal cancer liver metastases: a
multicentre randomized trial. Lancet. 2003;361:368 –373.
18. Onaitis M, Morse M, Hurwitz H, et al. Adjuvant hepatic arterial chemother-
apy following metastasectomy in patients with isolated liver metastases. Ann
Surg. 2003;237:782–789.
19. Martin R, Edwards MJ, McMaster KM. Morbidity of adjuvant hepatic arterial
© 2010 Lippincott Williams & Wilkins
Annals of Surgery • Volume 251, Number 4, April 2010
infusion pump chemotherapy in the management of colorectal cancer meta-
static to the liver. Am J Surg. 2004;188:714 –721.
20. Rubbia-Brandt L, Audard V, Sartoretti P, et al. Severe hepatic sinusoidal
obstruction associated with oxaliplatin-based chemotherapy in patients with
metastatic colorectal cancer. Ann Oncol. 2004;15:460 – 466.
21. Meric F, Patt YZ, Curley SA, et al. Surgery after downstaging of unresectable hepatic
tumors with intra-arterial chemotherapy. Ann Surg Oncol. 2000;7:490–495.
© 2010 Lippincott Williams & Wilkins
Colorectal Liver Metastases Initially Unresectable
22. Pulitano C, Arru M, Catena M, et al. Results of preoperative hepatic arterial
infusion chemotherapy in patients undergoing liver resection for colorectal
liver metastases. Ann Surg Oncol. 2008;15:1661–1669.
23. Tanaka K, Shimada H, Kubota K, et al. Effectiveness of prehepatectomy
intra-arterial chemotherapy for multiple bilobar colorectal cancer metastases
to the liver: a clinicopathologic study of peritumoral vasculobiliary invasion.
Surgery. 2005;137:156 –164.
www.annalsofsurgery.com | 691