INTERNATIONAL JOURNAL OF ONCOLOGY 32: 481-490, 2008
481
RFA strongly modulates the immune system and
anti-tumor immune responses in metastatic liver patients
CHIARA NAPOLETANO1*, FEDERICA TAURINO1*, MAURO BIFFONI2, ADRIANO DE MAJO3,
GIORGIO COSCARELLA3, FILIPPO BELLATI4, HASSAN RAHIMI1, SIMONA PAUSELLI1,
ILENIA PELLICCIOTTA1,7, JOY M. BURCHELL5, LUCIO ACHILLE GASPARI3,
LUCIA ERCOLI6, PIERO ROSSI3 and AURELIA RUGHETTI1
1Department
of Experimental Medicine, University of Rome ‘Sapienza’, Rome; 2Medical Oncology Section,
Department of Hematology, Oncology and Molecular Medicine, Istituto Superiore di Sanità, Rome; 3Department
of Surgery, University of Rome ‘Tor Vergata’, Rome; 4Department of Gynecology, Obstetrics and Perinatology,
University of Rome ‘Sapienza’, Rome, Italy; 5Breast Cancer Biology Group, King's College London School of Medicine,
Guy's Hospital, London, UK; 6Department of Public Health, University of Rome ‘Tor Vergata’, Rome, Italy
Received August 17, 2007; Accepted October 22, 2007
Abstract. Radiofrequency tumor ablation (RFA) is a
therapeutic modality for liver cancer patients inducing
localized tumor necrosis with maximal preservation of
normal liver parenchyma. We investigated the immunomodulatory effects exerted by RFA treatment in liver cancer
patients with metastatic liver lesions (13 patients) or
hepatocellular carcinoma (HCC) (4 patients). Analysis of
lymphocyte subsets by flow cytometry revealed that after
RFA, CD3+ T cells, in particular CD4+, were decreased in
metastatic cancer patients, while no change was observed in
HCC patients. Moreover, RFA induced trafficking of naïve
and memory CD62L+ T cells from circulation to tissues.
When characterizing the function of T cells, proliferative
response to PHA was strongly increased after 48 h from RFA
in metastatic cancer patients. Furthermore, T cells produced
IFN-γ in response to the tumor associated MUC1 antigen. In
contrast, humoral immune responses against tumor antigens
such as MUC1 and HCV proteins were unaffected by RFA
treatment, although increase of circulating B cells was
observed only in metastatic cancer patients. These results
_________________________________________
Correspondence to: Dr Aurelia Rughetti, Department of
Experimental Medicine, Viale Regina Elena 324, 00161 Rome,
Italy
E-mail: aurelia.rughetti@uniroma1.it
Present address: 7Department of Tumor Immunology, Catholic
University of Rome, Rome, Italy
*Contributed
equally
Key words: radiofrequency tumor ablation, anti-tumor immune
response, hepatocellular carcinoma, metastatic liver lesions,
immunotherapy, MUC1, lymphocyte subset
indicate that RFA application can exert an activating effect
on the immune system in metastatic cancer patients,
favouring trafficking of lymphocyte subsets and enhancing
tumor antigen specific cellular immune responses.
Introduction
Hepatocellular carcinoma (HCC) and carcinoma metastases
(in particular from colon carcinoma) are the most frequent
liver malignancies worldwide. Surgical resection is regarded
as the gold standard treatment, although only 10 to 20% of
patients are candidates for resection because of extensive
disease or medical co-morbidities. Recently, progress in
imaging has allowed the development of interstitial therapies
with in situ destruction of liver tumors and saving of normal
tissue such as transcatheter arterial chemoembolization
(TACE), interstitial ablative therapies, including percutaneous
alcohol injection (PEI), cryotherapy, laser, microwaves and
radiofrequency thermal ablation (RFA) (1,2).
In this scenario, RFA has received increasing interest,
both for the possibility of treating patients unsuitable for
surgical resection and for the low morbidity and mortality
related to the ablation technique itself. Moreover, RFA is an
interesting therapeutical option for liver cancer patients with
limited liver reserve as a result of chronic disease (3). This
technique is designed to induce tumor destruction by
delivering a high frequency alternating current, through an
active needle-electrode introduced into the neoplastic tissue
itself. The needle-electrode can be positioned percutaneously
(under US or TC guidance), by laparoscopy or laparotomy.
Frictional heating is caused when the ions in the tissue
attempt to follow the changing directions of the alternating
current thus causing cell death by means of coagulative
necrosis. Already at 43˚C in 30-60 sec apoptosis is seen,
cellular death occurs in a few minutes at the temperature of
50˚C; in a few seconds at 55˚C and almost instantaneously at
temperatures above 60˚C (4). Low morbidity and mortality
rates, effective tumor ablation and maximal preservation of
NAPOLETANO et al: EFFECTS OF TUMOR ABLATION BY RADIOFREQUENCY
482
Table I. Patient clinical characteristics.
–––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––
Patienta
Gender
Age
Primary tumor
Tumor nodules
Nodule diameterb
Time of RFA treatmentc
––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––
P1
F
67
Breast
1
1.8
24
P2
F
61
Breast
1
1.2
12
P3
F
74
Breast
3
2.8/1.2/2.1
26
P4
F
53
Breast
4
3/3/1.5/1.5
36
P5
M
70
Pancreas
3
4/2.5/1
30
P6
M
53
Pancreas
1
5
46
P7
M
70
Stomach
3
1.2/2/2.3
25
P8
M
77
Colon
1
6
46
P9
F
63
Colon
3
1.8/2/2
26
P10
M
78
Colon
1
5
24
P11
M
76
Colon
1
5
36
P12
F
63
Colon
2
2/3.6
46
P13
F
54
Breast
4
1.5/1/3/3
36
PH1
PH2
PH3
PH4
M
M
M
M
60
74
73
73
HCC (HCV+/HBV+)
HCC (HCV+)
HCC (Alcohol)
HCC (HCV+)
2
2
2
1
2/3
3/3
2.5/2
2
16
25
21
7
–––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––
aP, Metastatic cancer patient; PH, HCC patient; bDiameters are expressed as cm; cTime in minutes of the all RFA applications.
–––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––
normal liver parenchyma are the main advantages of RFA
therapy for the management of liver lesions (5). In fact, it is
considered an efficacious therapy as a bridge to orthotopic
liver transplantation in HCC patients (6). Moreover, recent
studies have suggested a survival benefit for patients with
colorectal liver metastases undergoing RFA compared with
historical controls undergoing chemotherapy alone and have
shown that number and size of tumor nodules and serum
marker levels can be considered predictors of survival after
RFA treatment (7).
In animal models RFA has been shown to mimic a
pathological stimulus inducing strong inflammatory signals:
T cells are present at the site of the ablated lesions and
release of the necrotic tumor tissue activates the immune
system against the tumor (8,9). Recent reports have also
shown that the in vitro proliferative response to tumor
extracts is amplified in HCC patients following RFA (10,11),
although RFA impacts on important parameters of biological
tumor behaviour such as metastatic spreading and viral load
are still unclear (12). So far, little is known about the effect
of RFA treatment on the immune response in metastatic
cancer patients.
We report that in metastatic cancer patients thermal
ablation by radiofrequency induces a consistent modulation
of circulating immune effector cells, conversely to that
observed in HCC patients. However, in both group of
patients, RFA treatment induced a redistribution of both
naïve and memory CD62L+ T cells. We studied the specific
anti-tumor T cell response using as target antigen the MUC1
glycoprotein. MUC1 is a tumor-associated antigen, commonly
overexpressed in breast, colon and pancreatic cancers. Cancer
transformation undergoes aberrant O-glycosylation thus
generating tumor-associated carbohydrate epitopes such as
GalNAc-O-S/T (Tn) and Gal-GalNAc-O-S/T (T) (13) on the
peptide core (tandem repeat, TR) of its extracellular
domain. We report the first evidence that in metastatic cancer
patients T cell response against the MUC1 antigen was
increased. These results may contribute to better define the
immunological aftermath induced by RFA and its possible
relevance in designing combinatorial immuno-radio-surgical
therapies for the treatment of liver cancer lesions.
Materials and methods
Patient characteristics. Seventeen patients with mean age
of 67 (range 53-77 years) were enrolled in this study: 13
metastatic liver cancer patients and 4 hepatocellular
carcinoma (HCC) patients (Table I). Among metastatic liver
cancer patients, breast cancer was the primary tumor for 5
patients with a total of 13 liver metastasis, colorectal
carcinoma for other 5 with 8 liver metastasis, pancreatic
carcinoma for 2 patients with 4 metastasis, gastric cancer
for 1 patient with 3 metastases. Mean diameter of the 28
metastases was 3.2 cm (range: 1.2-6 cm). HCC patients were
all cirrhotic, as a consequence of HCV(n=2) and HCV/HBV
(n=1) infection and of alcohol abuse (n=1) respectively
and had a total of 7 HCC nodules. Mean diameter of the 7
HCC nodules was 2.5 cm (range: 2-3 cm).
RFA treatment. All patients signed informed consent for the
treatment. RFA procedures were performed percutaneously
except in one (P11), 6 under general anaesthesia and 11
INTERNATIONAL JOURNAL OF ONCOLOGY 32: 481-490, 2008
used to deliver 0˚C saline through the electrodes at 10-25
ml/min to maintain a tip temperature of 20-25˚C. In 3
patients (1 breast cancer and 2 colorectal cancer with 2 and
1 metastatic nodules, respectively) treatment was repeated
in order to complete necrosis.
Table II. Summary of the circulating immune cell populations
assesed in the study by flow cytometry.
–––––––––––––––––––––––––––––––––––––––––––––––––
Target population
Predominant reactivity
–––––––––––––––––––––––––––––––––––––––––––––––––
CD3-/CD19+
CD3+
CD3+/CD4+
CD3+/CD4+/CD45RA+
CD3+/CD4+/CD45RACD3+/CD8+
CD3+/CD4-/CD45RA+
CD3+/CD4-/CD45RACD3-/CD16+/CD56+
483
B cells
T cells
T cell subset
Naïve CD4+ cells
Memory CD4+ cells
T cell subset
Naïve CD8+ cells
Memory CD8+ cells
NK cells
Flow cytometry. Peripheral blood mononuclear cells
(PBMCs) were isolated by Ficoll-Hypaque (1.077 g/ml,
Pharmacia, LKB, Sweden) density gradient centrifugation
from heparinized blood obtained at different time points
before (day 0) and after RFA (day 2 and 7) and frozen
immediately. The phenotype of circulating PBMCs was
defined using the Multistaining Kits (Becton-Dickinson,
San, Diego, CA). Samples were analyzed using a FACScalibur
(Becton-Dickinson) and analysis performed employing the
CellQuest software (BD Bioscience). Specific lymphocyte
populations were gated by SSC and FSC and expression of
specific markers (Table II) and results plotted as a percentage
of positive cells.
–––––––––––––––––––––––––––––––––––––––––––––––––
under conscious sedation with Midazolam and Fentanyl.
Local anaesthesia was achieved by using 2% lidocaine. All
procedures were performed under ultrasound (US) guidance
using Hitachi Spazio EUB-404 with B-314 probe for single
needle or Ca-11 probe for cluster needle and AL-33S probe
for RFA during laparotomy. The procedures were performed
using the RF Generator Cool-Tip System (Series 3;
Radionics, MA). The single RF electrode was 17 gauge, with
internal dual channels for chilled water. The clustered
electrode comprised three needles spaced 0.5 cm apart. A
peristaltic pump (Watson-Marlow, Wilmington, MA) was
Evaluation of IFN-γ production to MUC1 tumor-associated
antigen by ELISPOT assay. MUC1 glycopeptides carrying
O-linked GalNAc or Gal-GalNAc (Tn and T carbohydrate
epitopes, respectively) were chemically synthesized (14)
and kindly provided by Professor H. Paulsen (Hamburg
University, Germany) (Table III). PBMCs were grown in
RPMI-1640 (Hyclone) with AB Human Serum (HS; Sigma)
and IL-2 (20 UI/ml; R&D System, Germany) and MUC1
glycopeptide (25 μg/ml). As control antigens, Tetanus
Toxoid (TT) antigen and C. albicans extracts (kindly
Table III. Tumor-associated MUC1 based glycopeptides.
–––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––
MUC1
AHG
A
H
G
V
T
S
A
P
D
T
R
P
A
P
G S
T
A P
P
naked
peptide
––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––
A4
A
H
G
V
T
S
A
P
D
T
R
P
A
P
G S
T
A P
P
|
Gal-GalNAc
Tumor
A4GN
A
H
G
V
T
associated
S
A
P
D
T
R
P
A
P
G
S
T
A
P
P
A
P
E
S
R
P
A
P
G
S
T
A
P
P
P
E
S
R
P
A
P
G
S
T
A
P
A
|
MUC1
GalNAc
glycopeptides
A4M1
A
H
G
V
T
S
|
Gal-GalNAc
A4M3
A
H
G
V
T
S
A
|
Gal-GalNAc
––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––
MUC1 glycopeptides based on the tandem repeat sequence of MUC1 carrying the tumor associated O-linked carbohydrate epitopes GalNAc (Tn
antigen) or Gal-GalNAc (T antigen). Amino acid substitutions corresponding to alternative sequence of MUC1 TR are indicated in bold.
––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––
484
NAPOLETANO et al: EFFECTS OF TUMOR ABLATION BY RADIOFREQUENCY
Figure 1. Scatterplots comparing the proportions of CD3+ lymphocyte populations just before RFA treatment (day 0), and after 2 and 7 days from RFA in liver
metastatic patients (A) and HCC patients (B). Values express percentage of blood circulating CD3+ T cells, CD3+/CD4+ and CD3+/CD8+ lymphocytes in first,
second and third row, respectively. Significant statistical differences were found in the distribution of CD3+ T cells following RFA in metastatic liver patients
(p<0.005) by two-tail ANOVA test and in particular statistic significance was found between day 0 vs. day 2 groups and therein day 2 vs. day 7 as assayed
by Bonferroni test. Similar effects were observed in the CD3+/CD4+ subset in metastatic cancer patients (day 0 vs. day2, p<0.01; day 2 vs. day 7, p<0.01). No
significant changes were observed in the HCC patients (B).
provided by Dr Toro Santucci, IIS, Rome, Italy) were used
at 10 μg/ml. After a week of culture, T cells were boosted
with peptide pulsed autologous irradiated PBMCs (1:1)
and IFN-γ response to specific antigens was evaluated by
ELISPOT. Plates (Millipore, Bedford, MA) were coated
with 100 μl of MoAb anti-human IFN-γ (10 μg/ml; clone
NIB42, Pharmingen, San Diego, CA), (overnight at 4˚C).
After blocking with PBS-1% bovine serum albumin (BSA)
(Sigma) (200 μl/well), PBMCs were plated in quadruplicate
(5x104 cells/well) for 24 h. Cells were removed and plates
washed and IFN-γ production was detected with biotinylated
anti-IFN-γ mAb (clone 4S.B3, Pharmingen) (2 h RT, 5 μg/ml,
50 μl/well), revealed with streptavidin-alkaline phosphatase
(Pharmingen) (2 h; 1:2,000; 50 μl/well) and chromogen
substrate (50 μl/well). Spots were counted using the ImmunoSpot Image Analyzer.
Proliferative response to PHA. PBMCs, isolated at day 0,
day 2 and day 7 from RFA were seeded at 5x105 cells/ml
(200 μl/well) in a 96-well round bottom plate (Corning
Incorporated, NY) with phytohemoagglutinin (PHA; Sigma)
5 μg/ml in RPMI-5% HS. After 4 days, proliferative response
was analysed by [3H]-thymidine incorporation (ICN, Costa
Mesa, CA) (1 μCi/well). The Proliferation Index (P.I.) was
INTERNATIONAL JOURNAL OF ONCOLOGY 32: 481-490, 2008
Figure 2. Scatterplots comparing circulating NK cells (CD45 +/CD3 -/
CD56+/CD16+) during RFA treatment (day 0, 2 and 7) in liver metastatic
patients (closed circle) and HCC patients (open square). The percentage of
circulating NK cells was higher in metastatic patients than HCC during the
study period (p<0.05). Following RFA application an increase of circulating
NK cells was observed in metastatic liver patients, while no change was
observed in HCC patients.
calculated by dividing the average cpm obtained for PHA
stimulated PBMCs cell samples by the average cpm obtained
for PBMCs alone.
Detection of anti-MUC1 and anti-HCV circulating human
antibodies. Serum samples were collected at day 0, 2 and 7
from tumor ablation and stored frozen at -80˚C until use. For
detection of MUC1 antibodies EIA/RIA plates (Corning
Incorporated) were precoated overnight with streptavidin
(Sigma) (1 μg/ml), then blocked with PBS-5% BSA. Biotinylated MUC1 60 mer (1 μg/ml) was added for 2 h; 100 μl
patient sera (1:100) were added to the wells in triplicate.
Peroxidase-conjugated α -human IgG (Dako, Denmark)
(1:1000) and OPD-substrate (Sigma) were used to reveal
reaction; absorbance was measured at 492 nm. Anti-HCV
Igs levels were detected in HCC patients sera at the same
time-point as above by chemiluminescent microparticle
immunoassay (Architect kit, Abbot Diagnostic, Germany)
following the manifacturer's instructions (15). Results are
expressed as a ratio between the Relative Light Units (RLU)
of the sample and the RLU of the standard control. Values
above 1 were considered positive.
Statistical analysis. Descriptive statistics (average and
standard deviation) were used to describe the groups of data.
Repeated measures ANOVA with post-test for matched pairs
was used to test for overall differences among the groups,
followed by Bonferroni's significant difference adjustment
for multiple comparisons.
Results
Modulation of circulating CD3+ T cell subsets and NK cells
in liver cancer patients following RFA. Effects of thermal
ablation by radiofrequency on T cells were characterized,
studying circulating CD3 T cell subsets (CD4+ and CD8+) in
liver metastatic cancer patients before (day 0) and after RFA
application (day 2 and day 7) (Fig. 1A). Also effects of such
485
treatment on circulating lymphocyte subsets were monitored
in the group of HCC patients (Fig. 1B).
At day 0, levels of circulating CD3+ T cells were similar
in both groups of patients. Following RFA, a significant
modulation of CD3+ T cells was observed in patients with
liver metastasis (p<0.005). In particular, circulating CD3+ T
cells significantly decreased 2 days following RFA treatment
(p<0.01) and original baseline values were recovered only at
day 7 (p<0.05) (Fig. 1A), whereas in HCC patients CD3+ cell
count remained unmodified (p=0.78) (Fig. 1B). The reduction
observed in CD3+ T cells in metastatic patients was mainly
due to a modification of the CD4+ T cell subset (p<0.001).
A statistically significant decrease was observed at day 2
(p<0.01) vs. day 0 and vs. day 7 in CD3 + /CD4 + cells.
Moreover, a consistent but not significant decrease was
observed at day 7 in the CD3+/CD8+ cells (p=0.29). In HCC
patients, both CD4+ and CD8+ counts remained unaffected
by RFA treatment (Fig. 1B).
The effects of RFA on the natural immune effector cells
were also studied (Fig. 2). In metastatic cancer patients, NK
cells showed a weak increase, although not significant,
while in HCC patients this was not observed. Moreover, the
percentage of circulating NK cells was statistically higher in
metastatic as compared to HCC patients at each time point
(p<0.05) (Fig. 2). Thus, RFA treatment seemed to strongly
modulate the T cell subsets in patients with metastatic liver
lesions.
RFA modulates CD62L+ T cells in liver cancer patients.
Since RFA modulates circulating CD4 and CD8 T cells,
the redistribution of these T cell subsets was analyzed on
the basis of naïve (CD45RA +) and memory (CD45RA -)
phenotype and on the basis of CD62L expression, a marker
associated with the migratory phenotype of T cells (16)
(Fig. 3). CD62L+ T cells (CD4 and CD8) of both naïve and
memory subsets showed a gradual decline following RFA
treatment in the two groups of patients, with statistical
significance for naïve T cells in the metastatic cancer
patient group (CD4: p<0.05; CD8: p<0.05) (Fig.3A). In the
memory T cell compartment, a general decline of T cell
count was detected, however, this was significant only for
the CD62L+CD4+ cell subset (p<0.01).
When analyzing the CD62L- T cells (Fig. 3B), naïve T
cells were significantly increased in metastatic cancer
patients after 7 days (CD4, p<0.01; CD8, p<0.05), while no
change was observed in HCC patients. In both groups of
patients, CD62L - memory T cells did not appear to be
modulated by RFA.
RFA treatment modulates proliferative T cell response and
enhances anti-MUC1 cellular response in metastatic cancer
patients. The effects exerted by RFA on the general
performace of the immune system of metastatic cancer
patients were investigated by evaluating PBMC proliferation
to strong aspecific stimuli such as PHA before (day 0)
and after (day 2 and 7) tumor ablation (Fig. 4A). A strong
increase in T cell response to PHA was observed at day 2
following RFA, while at day 7 proliferation returned to the
baseline values. The proliferation pattern was consistent
in all the patients tested, despite the heterogeneity of
486
NAPOLETANO et al: EFFECTS OF TUMOR ABLATION BY RADIOFREQUENCY
Figure 3. Effects of RFA on CD62L T cells of both naïve and memory T cell subsets. Cytofluorimetric analysis was performed on circulating lymphocytes
following RFA treatment (day 0, day 2 and day 7) in HCC patients (open squares) and in metastatic liver patients (open triangles). T cells were
simultaneously characterised for CD62L, CD4 and CD45RA expression. (A) Shows the circulating levels of CD62L+ CD4+ and CD8+ T cells of both naïve
(CD45RA+) and memory (CD45RA-) subsets, while in (B) results regarding CD62L- T cells are reported. Statistical significance in redistribution of T cell
subsets following RFA was evaluated by ANOVA test.
proliferative responses observed at day 0 (ranging from 15
up to 97 P.I.).
To study whether the stimulatory effect observed on the
aspecific immune response could imply any influence on the
anti-tumor one, the cellular response to the MUC1 tumor
epithelial antigen and in particular to tumor-associated
glycoepitopes of MUC1, was evaluated as IFN-γ production
in ELISPOT assay. PBMCs before and 7 day after RFA were
stimulated using a panel of MUC1 tumor-associated glycopeptides (20 μg/ml, 20 mer) (Table III). The employed
glycopeptides mimicked the glycopeptide structures found
on the tumor-associated MUC1 molecules: A4, A4M1 and
A4M2 glycopeptides carried the T carbohydrate residue
(Gal-GalNAc), while the A4GN glycopeptide carried the Tn
glycoepitope (GalNAc). Moreover, the A4M1 and A4M3
showed an alternative MUC1 peptide backbone as found
in vivo (14). The AHG peptide corresponded to the
unglycosylated MUC1 sequence. One week from priming, T
cells were boosted with the specific antigen and IFN- γ
secretion was evaluated by ELISPOT assay (Fig. 4B).
Following RFA treatment, IFN- γ production to MUC1
glycopeptides markedly increased. A response to the
unglycosylated MUC1 sequence (AHG peptide) was also
observed. The IFN-γ response appeared to be independent
upon the specific amino acid substitution occurring in the TR
sequences in vivo (A4M1 and A4M3 vs. A4 glycopeptides).
IFN-γ response to Tetanus Toxoid and C. albicans antigens
did not significantly increase.
RFA and B-cell mediated response. B cell-mediated response
is a key point in the memory immune response following
inflammation and can be an important parameter in tumor
INTERNATIONAL JOURNAL OF ONCOLOGY 32: 481-490, 2008
487
HCV IgG titers were not significantly modified by RFA
treatment.
Discussion
Figure 4. Immunostimulatory effect of RFA on anti-tumor immune
response. (A) Proliferative assay of circulating PBLs before and after
RFA treatment (day 0, open histogram; day 2, black histogram; day 7
striped histogram) following PHA stimulation for 72 h in metastatic liver
patients (breast, P1, P2, P4, P13; stomach, P7; colon, P12). Proliferative
response was evaluated as [3H]-thymidine incorporation and results plotted
as Proliferative Index (P.I). (B) Elispot assay for evaluating the production of
IFN-γ after stimulation with specific tumor-associated MUC1 glycopeptides
before and after RFA in metastatic liver patients P1 (breast) and P9 (colon).
MUC1 based antigens were AHG, 20 mer corresponding to the
unglycosylated TR, Gal-GalNAc (A4) and GalNAc (A4GN) carrying glycopeptides. A4M3 and A4M1 are MUC1 Gal-GalNAc carrying glycopeptides
corresponding to peptide TR variants. As control antigens, TT and C.
albicans extracts were used. PBLs were stimulated with the specific antigen
and after seven days, 105 cells/well were plated and stimulated overnight in
millipore plates coated with anti-IFN-γ antibody in triplicate for each
condition. Results were evaluated as IFN-γ spot forming unit (sfu) by
ELISPOT analyser and plotted as histograms (white and black histograms
for before and after RFA treatment, respectively).
immunity. In metastatic cancer patients, RFA induced an
increase of the circulating B cell subset (p<0.05), particularly
in the 48 h following RFA (p<0.05) (Fig. 5A). Conversely,
no B cell increment was observed in HCC patients, although
baseline B cells were similar in both groups of patients
(Fig. 5B). The B cell response was also analyzed as means of
circulating IgGs specific for MUC1 tumor antigen in
metastatic patients as well as anti-HCV IgGs in HCC patients
(Table IV). In both groups of patients, anti-MUC1 or anti-
Different loco-regional therapies have been developed for
unresectable liver tumors aimed at achieving local tumor
control and prolonging patient survival. RFA and other
ablative treatments have an increasing application for liver
cancers showing high benefit for clinical management. In
particular, RFA has a definitive role for HCC treatment both
as curative intent and as a bridge to liver transplantation;
moreover, it is the most widely used tumor ablative
technique for treatment of colorectal liver metastasis (17-19).
RFA destroys tumoral tissues by necrotic and apoptotic
mechanisms (4) and these pathological events may influence
the overall and anti-tumor host immune responses (20).
Results reported herein show that indeed the RFA application
may enhance the immune system reactivity in liver cancer
patients, in particular in those with metastatic lesions.
Several reports suggest that in animal models, RFA
treatment could provide activatory signals and become a
source of tumor antigens for the immune system (8,9,21). In
humans, RFA treatment can exert modulatory effects on
the immune system (22) and could boost lymphocyte
proliferation and production of inflammatory cytokines in
response to tumor extracts in HCC patients (10,11). Herein,
we show for the first time that in metastatic cancer patients,
RFA treatment can amplify the specific T cell response
against an epithelial tumor antigen such as the MUC1 Oglycoprotein. MUC1 tumor antigen is target of humoral and
cellular immune response in cancer patients and is
considered an optimal target for cancer immunotherapy (23).
In the change to malignancy, MUC1 is overexpressed and
aberrantly glycosylated: novel cancer-associated carbohydrate moieties (such as T and Tn) exist on the extracellular domain thus generating MUC1 glycoepitopes that are
targets of a tumor specific immune response both humoral
and cellular (24-27). RFA treatment amplifies T cell
mediated IFN-γ response towards these tumor-associated
glycoepitopes, while the specific immune response to recall
antigens is not significantly increased. The enhancement of
the anti-tumor immune response could be a bystander effect
induced by the overall inflammatory response generated by
the tumor ablation procedure. However, the MUC1 specific
immune reaction is still sustained after a week from RFA,
while the proliferative response to PHA, regarded as a
general indicator of immune activation, was augmented after
48 h and declined to baseline values after a week from RFA.
These results suggest that in metastatic cancer patients,
thermal ablation of the tumor mass may induce an antigen
specific immune response that could be fostered independently by the overall inflammatory response.
No previous report exists on the influence of RFA on the
humoral anti-tumor immune response. Other ablative tumor
treatments have been shown to induce an anti-tumor IgM
response (28). We characterized the humoral immune
response monitoring the B cell pattern during RFA and the
specific IgGs titers against the MUC1 tumor antigen and
HCV capsidic proteins in metastatic and HCC patients,
488
NAPOLETANO et al: EFFECTS OF TUMOR ABLATION BY RADIOFREQUENCY
Figure 5. Effects of RFA treatment on circulating B cells. Circulating B cells (CD45+/CD3-/CD19+) were analysed by flow cytometry at different time points
of the RFA protocol in metastatic liver patients (A) and in HCC patients (B). Only in metastatic cancer patients, RFA treatment significantly modulates B cell
counts as evaluated by ANOVA test (p<0.05), as showed by increase of B cell level at day 2 (p<0.05).
Table IV. Anti-tumor humoral response in liver cancer patients undergoing tumor ablation.
–––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––
Metastatic patients
Primary tumor
Anti-MUC1 circulating IgGsa
–––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––
Pre-RFA
Post-RFA
––––––––––––––––––––––––––––––––––––––
P1
Breast
0.69
0.80
P2
Breast
0.96
1.31
P3
Breast
0.53
0.50
P5
Pancreas
0.42
0.40
P6
Pancreas
0.54
0.61
P13
Breast
1.06
0.93
P8
Colon
0.67
0.75
P9
Colon
0.43
0.67
P10
Colon
0.48
0.76
P12
Colon
0.48
0.56
–––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––
HCC patients
Primary tumor
Anti-HCV circulating Igsb
–––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––
Pre-RFA
Post-RFA
––––––––––––––––––––––––––––––––––––––
PH1
HCC (HBV+/HCV+)
12.53
13.04
PH2
HCC (HCV+)
15.68
15.14
PH3
HCC (HCV-)
0.98
0.94
PH4
HCC (HCV+)
16.06
15.94
–––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––
aO.D.
values read at 492 nm wavelength; bArbitrary unit of Architect system anti-HCV test, cut-off values <1.
––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––
respectively. We found an increment of circulating B cells
following RFA in metastatic cancer patients, while no
significant increase of anti-MUC1 or anti-HCV (in HCC
patients) IgG titers was detected. Several reasons could
account for this discrepancy between the B cell modulation
and IgG titers e.g. release of antigens from the necrotic tumor
extracts could consume circulating antibodies or the antibody
determination could be too close to the stimulus. Moreover,
the immune system of cancer patients undergoing RFA is
already impaired by the establisment of immunosuppressive
networks that contribute to the progression of liver tumor and
to the metastatic process (29). Removal of this immunosuppressive network is needed to potentiate an active antitumor immune response (30). In mouse models, RFA
INTERNATIONAL JOURNAL OF ONCOLOGY 32: 481-490, 2008
treatment can induce a protective immune response towards
an exogenous antigen such ovalbumin only when combined
with the removal of Treg through the CTLA-4 blockade (31).
In humans, RFA has been shown to be more effective than
chemotherapy for the clinical treatment of metastatic cancer
patients with minimal liver lesions (32). On the other hand,
RFA treatment does not show any advantage in long-term
survival compared to surgery in metastatic cancer patients (33).
The in vivo evidence suggests that the inflammatory noxa
engendered by thermal ablation (both RFA and cryoablation) appears to be a robust stimulus for the enhancement
of a specific immune response, releasing tumor antigens and
supplying strong ‘danger signals’ such as HSPs and release
of inflammatory cytokines (21), although not sufficient to
overcome tumor immunosuppression.
A key point for the activation and maintenance of an
efficacious and long-lasting immune response is the
trafficking of immune cells between tumor site, lymphoid
organs and systemic circulation. Herein, we describe that
RFA exerted immunomodulatory effects on distinct effector
immune cells in metastatic cancer patients, in particular in
the redistribution of circulating T cells; similar event was
observed in pancreatic cancer patients undergoing RFA (data
not shown). No modulation effect seems to occur in HCC
patients: the chronic condition underlying tumor development could account for an overall weakening of NK and T
cell-mediated immunity (34,35). However, it is noteworthy
that in all the liver cancer patients studied a significant
and stable decrease of circulating CD62L+ T cells of both
memory and naïve subsets was observed suggesting the
migration of these cells into the tissues. Any strategy for the
activation of an efficacious immune response is aimed at
enhancing the anti-tumor immunity and to prime and expand
the repertoire of naïve anti-tumor immune cells that become
activated only when the tumor antigen is presented in
secondary lymphoid organs over a sufficiently long period of
time (36). Transient, but sufficient amount of antigens,
antigen presenting cell (APCs) activation, adequate cytokine
environment may dictate the delicate balance needed for an
efficacious immune response (37,38). Indeed, tumor ablation
by radiofrequency provides these conditions in vivo. The
necrotic/apoptotic events induced by RFA can supply an
adequate inflammatory environment to activate APCs and
generate a massive and transient release of tumor antigens
that could be presented by the activated APC to T cells
recruited from the circulation. On the other hand, massive
antigen release could overload the immune system thus
producing exhaustion of responder T cells. The inflammatory
environment could also switch the balance from activation of
anti-tumor to regulatory immune cells (30). Clinical
parameters, such the number of treated lesions, the total
tumor mass, the efficacy of necrosis induced may have great
relevance in stirring the balance between activation and
suppression of the anti-tumor immune response. Moreover,
the baseline immune performance of patients may well be
another restriction point for the activation of productive
immunity.
In conclusion, we showed that RFA can activate a
specific immune response to the MUC1 tumor antigen in
metastatic liver patients and analysis of the immune cell
489
repertoire suggests that RFA may activate different immunological responses, implying an important clinical impact for
the use of this ablative technique. The bystander biological
effects produced by RFA on the immune system (strong
activation signal, recruitment of immune cells, release of
tumor antigens) and the applicability of this procedure for
treatment of solid tumors of different origins (39), make
tumor ablation an appealing procedure when designing
combined therapeutical approaches for the treatment of
cancer disease.
Acknowledgments
We are most grateful to Dr H. Paulsen (Hamburg University,
Germany) for providing the MUC1 glycopeptides; C. albicans
extracts were the generous gift of Dr Santoro Tucci (ISS,
Rome, Italy). We also thank Dr D. Lilli for the analysis of
anti-HCV IgG and both Dr D. Lazzereschi and Dr H.
Snelgrove for their invaluable help in revising the manuscript
(University of Rome ‘Sapienza’, Italy). Chiara Napoletano
is supported by ‘Consorzio Interuniversitario Trapianti
d'Organo’, Italy. This work was funded by EU Consortia
(QLRT-1999-00217; QLRT-2001-02010), Ministero dell'
Università e della Ricerca, Progetto Ateneo and Facoltà
(Università di Roma ‘Sapienza’).
References
1. Blum HE: Treatment of hepatocellular carcinoma. Best Pract
Res Clin Gastroenterol 19: 129-145, 2005.
2. Tateishi R, Shiina S, Teratani T, et al: Percutaneous radiofrequency ablation for hepatocellular carcinoma. An analysis
of 1000 cases. Cancer 103: 1201-1209, 2005.
3. Goldberg SN, Gazelle GS, Compton CC, Mueller PR and
Tanabe KK: Treatment of intrahepatic malignancy with radiofrequency ablation: radiologic-pathologic correlation. Cancer
88: 2452-2463, 2000.
4. McGahan JP, Brock JM, Tesluk H, Gu WZ, Schneider P and
Browning PD: Hepatic ablation with use of radiofrequency
electrocautery in the animal model. J Vasc Interv Radiol 3:
291-297, 1992.
5. Ng KK and Poon RT: Radiofrequency ablation for malignant
liver tumor. Surg Oncol 14: 41-52, 2005.
6. Lu DS, Yu NC, Raman SS, et al: Percutaneous radiofrequency
ablation of hepatocellular carcinoma as a bridge to liver
transplantation. Hepatology 41: 1130-1137, 2005.
7. Berber E, Pelley R and Siperstein AE: Predictors of survival
after radiofrequency thermal ablation of colorectal cancer
metastases to the liver: a prospective study. J Clin Oncol 23:
1358-1364, 2005.
8. Den Brok MH, Sutmuller RP, van der Voort R, Bennink EJ,
Figdor CG, Ruers TJ and Adema GJ: In situ tumor ablation
creates an antigen source for the generation of antitumor
immunity. Cancer Res 64: 4024-4029, 2004.
9. Wissniowski TT, Hansler J, Neureiter D, et al: Activation of
tumor-specific T lymphocytes by radio-frequency ablation of
the VX2 hepatoma in rabbits. Cancer Res 63: 6496-6500,
2003.
10. Zerbini A, Pilli M, Penna A, et al: Radiofrequency thermal
ablation of hepatocellular carcinoma liver nodules can activate
and enhance tumor-specific T-cell responses. Cancer Res 66:
1139-1146, 2006.
11. Hansler J, Wissniowski TT, Schuppan D, Witte A, Bernatik T,
Hahn EG and Strobel D: Activation and dramatically increased
cytolytic activity of tumor specific T lymphocytes after radiofrequency ablation in patients with hepatocellular carcinoma
and colorectal liver metastases. World J Gastroenterol 12:
3716-3721, 2006.
12. Livraghi T, Lazzaroni S, Meloni F, et al: Risk of tumour seeding
after percutaneous radiofrequency ablation for hepatocellular
carcinoma. Br J Surg 92: 856-858, 2005.
490
NAPOLETANO et al: EFFECTS OF TUMOR ABLATION BY RADIOFREQUENCY
13. Taylor-Papadimitriou J, Burchell JM, Plunkett T, Graham R,
Correa I, Miles D and Smith M: MUC1 and the immunobiology
of cancer. J Mammary Gland Biol Neoplasia 7: 209-221, 2002.
14. Karsten U, Diotel C, Klich G, Paulsen H, Goletz S, Muller S
and Hanisch FG: Enhanced binding of antibodies to the DTR
motif of MUC1 tandem repeat peptide is mediated by sitespecific glycosylation. Cancer Res 15: 2541-2549, 1998.
15. Bukh J, Miller RH and Purcell RH: Biology and genetic
heterogeneity of hepatitis C virus. Clin Exp Rheumatol 13: 3-7,
1995.
16. Hengel RL, Thaker V, Pavlick MV, et al: Cutting edge: LSelectin (CD62L) expression distinguishes small resting
memory CD4+ T cells that preferentially respond to recall
antigen. J Immunol 170: 28-32, 2003.
17. Poston GJ: Radiofrequency ablation of colorectal liver
metastases: where we really going? J Clin Oncol 23: 1342-1344,
2005.
18. Oshowo A, Gillams AR, Harrison E, Lees WR and Taylor I:
Comparison of resection and radiofrequency ablation for treatment of solitary colorectal liver metastases. Br J Surg 90: 12401243, 2003.
19. Livraghi T, Solbiati L, Meloni F, et al: Percutaneous radiofrequency ablation of liver metastases in potential candidates
for resection: the ‘test-of-time approach’. Cancer 97: 30273035, 2003.
20. Bartholomae WC, Rininsland FH, Eisenberg JC, Boehm BO,
Lehmann PV and Tary-Lehmann M: T cell immunity induced by
live, necrotic and apoptotic tumor cells. J Immunol 73: 1012-1022,
2004.
21. Yang WL, Nair DG, Makizumi R, Gallos G, Ye X, Sharma RR
and Ravikumar TS: Heat shock protein 70 is induced in mouse
human colon tumor xenografts after sublethal radiofrequency
ablation. Ann Surg Oncol 11: 399-406, 2004.
22. Rughetti A, Rahimi H, Rossi P, et al: Modulation of blood
circulating immune cells by radiofrequency tumor ablation. J
Exp Clin Cancer Res 22: 247-250, 2003.
23. Acres B and Limacher JM: MUC1 as a target antigen for cancer
immunotherapy. Expert Rev Vaccines 4: 493-502, 2005.
24. Xu Y, Gendler S and Franco A: Designer glycopeptides for
cytotoxic T cell-based elimination of carcinomas. J Exp Med
199: 707-716, 2004.
25. Tarp MA, Sorensen AL, Mandel U, Paulsen H, Burchell J,
Taylor-Papadimitriou and Clausen H: Identification of a novel
cancer-specific immunodominant glycopeptide epitope in the
MUC1 tandem repeat. Glycobiology 17: 197-209, 2007.
26. Rughetti A, Turchi V, Ghetti CA, et al: Human B-cell immune
response to the polymorphic epithelial mucin. Cancer Res 53:
2457-2459, 1993.
27. Von Mensdorff-Pouilly S, Verstraeten AA, Kenemans P, et al:
Survival in early breast cancer patients is favorably influenced
by a natural humoral immune response to polymorphic epithelial
mucin. J Clin Oncol 18: 574-583, 2000.
28. Ravindranath MH, Wood TF, Soh D, et al: Cryosurgical ablation
of liver tumors in colon cancer patients increases the serum total
ganglioside level and then selectively augments antiganglioside
IgM. Cryobiology 45: 10-21, 2002.
29. Ryungsa K, Manabu E and Kazuaki T: Tumor driven evolution
of immunosuppressive networks during malignant progression.
Cancer Res 66: 5527-5536, 2006.
30. Zou W: Immunosuppressive networks in the tumour environment
and their therapeutic relevance. Nat Rev Cancer 5: 263-274,
2005.
31. Den Brok MH, Sutmuller RP, Nierkens S, et al: Efficient loading
of dendritic cells following cryo and radiofrequency ablation
in combination with immune modulation induces anti-tumour
immunity. Br J Cancer 95: 896-905, 2006.
32. Gillams AR and Lees WR: Survival after percutaneous, imageguided, thermal ablation of hepatic metastases from colorectal
cancer. Dis Colon Rectum 43: 656-661, 2000.
33. Abdalla EK, Vauthey JN, Ellis LM, et al: Recurrence and
outcomes following hepatic resection, radiofrequency ablation,
and combined resection/ablation for colorectal liver metastases.
Ann Surg 239: 818-827, 2004.
34. Golden-Mason L and Rosen HR: Natural killer cells: primary
target for hepatitis C virus immune evasion strategies? Liver
Transpl 12: 363-372, 2006.
35. Gremion C and Cerny A: Hepatitis C virus and the immune
system: a concise review. Rev Med Virol 15: 235-268, 2005.
36. Ochesenbein AF: Immunological ignorance of solid tumors.
Springer Semin Immunol 27: 19-35, 2005.
37. Zinkernagel RM and Hengartner H: Regulation of the immune
response by antigen. Science 293: 251, 2001.
38. Blais ME, Louis I and Perreault C: T-cell development: an
extrathymic perspective. Immunol Rev 209: 103-114, 2006.
39. Rose SC, Thistlethwaite PA, Sewell PE and Vance RB: Lung
cancer and radiofrequency ablation. J Vasc Interv Radiol 7:
927-951, 2006.
G. Coscarella