The JI
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     
 

The Journal of Immunology, 2008, 181, 6595 -6603
Copyright © 2008 by The American Association of Immunologists, Inc.
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow Request Permissions
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Tassi, E.
Right arrow Articles by Protti, M. P.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Tassi, E.
Right arrow Articles by Protti, M. P.

Carcinoembryonic Antigen-Specific but Not Antiviral CD4+ T Cell Immunity Is Impaired in Pancreatic Carcinoma Patients1

Elena Tassi*,||, Francesca Gavazzi{dagger},#, Luca Albarello, Vladimir Senyukov*,||, Renato Longhi**, Paolo Dellabona§,||, Claudio Doglioni{ddagger},¶, Marco Braga{dagger},{ddagger},#, Valerio Di Carlo{dagger},{ddagger},# and Maria Pia Protti2,*,||

* Tumor Immunology Unit, {dagger} Pancreas Unit, {ddagger} Università Vita-Salute San Raffaele, § Experimental Immunology Unit, Unità Operativa Anatomia Patologica, || Cancer Immunotherapy and Gene Therapy Program, and # Department of Surgery, Scientific Institute H. San Raffaele, Milan, Italy; and ** Consiglio Nazionale Ricerche-Istituto di Chimica del Riconoscimento Molecolare, Milan, Italy


   Abstract
Top
 Abstract
Introduction
 Materials and Methods
 Results
 Discussion
 Disclosures
 References
 
Pancreatic carcinoma is a very aggressive disease with dismal prognosis. Although evidences for tumor-specific T cell immunity exist, factors related to tumor microenvironment and the presence of immunosuppressive cytokines in patients’ sera have been related to its aggressive behavior. Carcinoembryonic Ag (CEA) is overexpressed in 80–90% of pancreatic carcinomas and contains epitopes recognized by CD4+ T cells. The aim of this study was to evaluate the extent of cancer-immune surveillance and immune suppression in pancreatic carcinoma patients by comparing the anti-CEA and antiviral CD4+ T cell immunity. CD4+ T cells from 23 normal donors and 44 patients undergoing surgical resection were tested for recognition of peptides corresponding to CEA and viral naturally processed promiscuous epitopes by proliferation and cytokine release assays. Anti-CEA CD4+ T cell immunity was present in a significantly higher number of normal donors than pancreatic cancer patients. Importantly, whereas CD4+ T cells from normal donors produced mainly GM-CSF and IFN-{gamma}, CD4+ T cells from the patients produced mainly IL-5, demonstrating a skew toward a Th2 type. On the contrary, the extent of antiviral CD4+ T cell immunity was comparable between the two groups and showed a Th1 type. The immunohistochemical analysis of tumor-infiltrating lymphocytes showed a significantly higher number of GATA-3+ compared with T-bet+ lymphoid cells, supporting a Th2 skew also at the tumor site. Collectively, these results demonstrate that Th2-immune deviation in pancreatic cancer is not generalized but tumor related and suggests that the skew might be possibly due to factor(s) present at the tumor site.


   Introduction
Top
 Abstract
 Introduction
Materials and Methods
 Results
 Discussion
 Disclosures
 References
 
Pancreatic cancer is a very aggressive disease and the fourth cause of cancer mortality (1, 2). Clinical manifestations develop late and in most cases the disease is metastatic at diagnosis, with a mean life expectancy of 3–6 mo. Surgery is the most effective treatment but only 15–20% of patients present with resectable tumors. These patients have a mean life expectancy of 15–18 mo and a few of them survive significantly longer (1). Chemotherapy and radiotherapy offer limited benefit (1).

Pancreatic carcinoma has a rather peculiar tumor microenvironment consisting of fibroblasts, pancreatic stellate cells, endothelial cells, immune and endocrine cells, and perineural spreading that is believed to play an active role in disease progression and aggressiveness (3).

Evidences for involvement of adaptive immunity in pancreatic cancer are the presence of immune infiltrates at the tumor site consisting of memory CD4+ and CD8+ T cells (4, 5) and of tumor-reactive T cells in bone marrow and blood of pancreatic cancer patients (6, 7, 8, 9, 10). However, mechanisms of immune evasion adopted by the tumor (11, 12, 13, 14, 15) as well as released factors and immune regulatory cells present at the tumor site possibly affect the efficacy of antitumor immunity in pancreatic cancer (16, 17, 18, 19, 20).

The adaptive immune response mediated by effector CD4+ Th cells is highly heterogeneous and based on the development of distinct subsets characterized by different cytokine production. At least four subsets have been described: Th1, Th2, Th17, and Th3/T regulatory (Tr)3 cells. Th1 cells produce IFN-{gamma} and are mainly devoted to protection against intracellular microbes and tumors and Th2 produce IL-4, IL-5, and IL-13 and are involved in the protection against gastrointestinal nematodes and are responsible for allergic disorders (21, 22). Th17 produce IL-17 and are expanded in autoimmune diseases (23). Th3/Tr produce TGF-β and/or IL-10 and exert immunosuppressive function (24). In antitumor immunity, a role for both Th1 and Th2 types has been shown in animal models (25, 26, 27). In humans, Th2-type responses have been associated with disease progression in melanoma and renal cell carcinoma (28, 29, 30, 31).

Previous studies have shown a predominance of Th2 cytokines in the sera of pancreatic cancer patients and hypothesized that this predominance would contribute to a general Th2 skewed immune phenotype (13, 32).

The carcinoembryonic Ag (CEA) is a 180-kDa glycoprotein overexpressed in almost all colorectal cancers and in a large percentage of non-small cell lung, breast, and pancreatic cancers (33, 34). Others (35, 36, 37) and we (38, 39) previously reported that CEA contains naturally processed epitopes recognized by CD4+ T cells from normal donors and colon and lung cancer patients.

In the present study, we addressed the questions whether in pancreatic cancer patients tumor Ag-specific CD4+ T cells are indeed polarized toward a Th2 type and whether this is generalized or related to local mechanisms by comparing the anti-CEA with antiviral CD4+ T cell immunity.

CEA peptides corresponding to the identified epitopes were used to study the spontaneous anti-CEA response in normal donors and pancreatic carcinoma patients as compared with CD4+ T cell responses specific for influenza and EBV- derived peptides. We found that the patients compared with normal donors have impaired quantitative (i.e., reduced number of responders) and qualitative (i.e., Th2 skew) anti-CEA CD4+ T cell responses. This impairment is CEA specific because quantitative and qualitative CD4+ T immunity against viral proteins was preserved, suggesting that local more than systemic immunosuppressive/regulatory mechanisms are active in pancreatic cancer patients. In agreement, immunohistochemical analysis at the tumor site showed the presence of a significantly higher number of GATA-3+ compared with T-bet+ infiltrating lymphoid cells.


   Materials and Methods
Top
 Abstract
 Introduction
 Materials and Methods
Results
 Discussion
 Disclosures
 References
 
Subjects and tumor samples

PBMC were obtained from the blood of 24 healthy subjects (i.e., normal donors) of the H. San Raffaele Blood Bank and 46 consecutive patients undergoing surgical resection for suspected pancreatic cancer. Patients’ blood was obtained at the time of surgery. Tumor samples were obtained from surgical specimens, cut into small pieces, and frozen for RNA extraction. The Institutional Ethics Committee had approved the study protocol and informed consent was obtained from all donors before blood sampling and surgical resection.

Selection and synthesis of peptides

Sequences CEA99–111, CEA117–129, CEA177–189/355–367, CEA425–437, CEA568–582, CEA652–666, and CEA666–678 were selected based on the TEPITOPE algorithm (40) as promiscuous HLA-DR binders (i.e., able to bind to several alleles) and previously shown to contain naturally processed epitopes (35, 37, 38, 39). Sequence 307–319 from influenza hemagglutinin (HA307–319) (41) and sequence 280–294 from Epstein-Barr nuclear Ag 2 (EBNA2280–294) (42) were selected to test antiviral immunity. Peptides corresponding to selected sequences were synthesized by the stepwise solid-phase method as described previously (43). The peptides were lyophilized, reconstituted in DMSO at 10 mg/ml, and diluted in RPMI 1640 (BioWhittaker) as needed.

Short-term culture and cytokine release assay

The experiments were performed as described previously (31, 44). Briefly, CD4+ T cells were purified from total PBMC by using magnetic beads (Miltenyi Biotec) and cultured in 96-well plates at 3 x 104/well in five replicates for each condition in 200 µl/well X-VIVO 15 (BioWhittaker) supplemented with penicillin (100 U/ml; BioWhittaker), streptomycin (50 mg/ml; BioWhittaker), and 3% heat-inactivated pooled human serum (tissue culture medium; BioWhittaker) in the presence of irradiated CD4+ depleted PBMC as APC, at a CD4+:APC ratio of 1:3. Stimuli were: PHA (10 µg/ml; Sigma-Aldrich), as positive control; CD4+ T cells in the presence of the APC only, as baseline (blank); and each single peptide (10 µg/ml). At day 7, half medium from each well was removed and replenished with fresh tissue culture medium containing IL-2 (25 IU/ml; R&D Systems) without any further Ag stimulation. At day 13, 100 µl of supernatant was collected from each well for cytokine detection and the cultures were pulsed for 16 h with [3H]TdR (1 µCi/well, 6.7 Ci/mol; Amersham Biosciences). The cells were collected with a FilterMate Universal Harvester (Packard Instrument) in specific plates (Unifilter GF/C; Packard Instrument) and the thymidine incorporated was measured in a liquid scintillation counter (TopCount NXT; Packard Instrument). Supernatants collected from each well of the five replicates of each condition were pooled and used for the detection by ELISA, according to the manufacturers’ instructions, of the following cytokines: GM-CSF, IFN-{gamma}, and TGF-β (BioSource International); IL-5, IL-10, and IL-13 (Mabtech); and IL-17 (eBioscience).

RNA extraction and RT-PCR

Tumor samples (50–100 mg) were homogenized in the presence of 1 ml of TRIzol Reagent (Invitrogen) and total RNA was extracted following the manufacturer’s instruction. Single-strand cDNA was synthesized from 2 µg of total RNA using Moloney murine leukemia virus reverse transcriptase (Promega). RT-PCR was performed with TaqDNA polymerase (Promega) by using CEA-specific oligonucleotides: CEA left, 5'-AACCCAGAACCCAGTGAGTG-3' and CEA right, 5'-ATTGCTGGAAAGTCCCATTG-3'. PCR conditions were: 5 min at 94°C, 30 cycles of 30 s at 94°C, 30 s at 58°C, and 30 s at 72°C, and a final elongation step of 10 min at 72°C.

Immunohistochemical analysis

Surgical specimens were fixed in buffered formalin and embedded in paraffin. Immunohistochemistry was performed on 5-µm tissue sections. All cases were immunostained with a sensitive nonbiotin detection system (NovoLink polymer; Novocastra), with diaminobenzidine development. Heat-induced Ag retrieval was performed using 0.01 M citrate buffer (pH 6.0) or Tris-EDTA (pH 9.0) in a water bath for 30 min. The following Abs were used: anti-CD3 mAb (clone PS1; Novocastra), anti-CD4 mAb (clone 4B12; Novocastra), anti-CD8 mAb (clone 1A5; Novocastra), anti-T-bet mAb (clone 4B10; Santa Cruz Biotechnology), anti-GATA3 polyclonal Ab (AF2605; R&D Systems), and anti-FoxP3 mAb (236A/E7; Santa Cruz Biotechnology). All stainings were performed with an automatic immunostainer (Biogenex Optimax). Labeled cells were enumerated independently by two pathologists within microscopic fields at x200 (0.280 mm2) of adjacent sections.


   Results
Top
 Abstract
 Introduction
 Materials and Methods
 Results
Discussion
 Disclosures
 References
 
CD4+ T cell immunity against CEA

Twenty-three normal donors and 44 pancreatic cancer patients were studied. Characteristics of recruited patients are reported in Table I. All but one patient (patient 33) underwent resective surgery. Patients 6 and 7 were excluded from the study after recruitment because of a diagnosis of chronic pancreatitis. The majority of patients had a ductal adenocarcinoma and metastatic disease in the lymph nodes (76% of cases). More than 80% of tumors tested by RT-PCR expressed CEA, while soluble CEA above normal values was detected in the sera of 23% of the patients evaluated.


View this table:
[in this window]
[in a new window]

  		Table I. Patients’ characteristics

 
Normal donors and patients were tested for the presence of CD4+ T cells against CEA sequences described to contain promiscuous and naturally processed epitopes (35, 37, 38, 39) using an ex vivo restimulation assay, which we previously demonstrated not to favor in vitro priming (44). CD4+ T cells were cultured in several replicates in the presence or the absence of single CEA peptides or PHA as a positive control. Proliferation and cytokine release assays were then evaluated. Responders were defined as subjects with specific reactivity of CD4+ T cells in the presence of one or more CEA peptides in either one or both assays. CEA-specific CD4+ T cells were present in the blood of 65.2 and 29.5% of normal donors and pancreatic carcinoma patients, respectively. The two percentages are significantly different as from the {chi}2 test (p = 0.0108).

To evaluate the functional activity of CEA-specific CD4+ T cells, culture supernatants were tested for cytokines released (i.e., GM-CSF, IFN-{gamma}, and IL-5). IFN-{gamma} and IL-5 were chosen as prototypic cytokines released by Th1 and Th2 cells, respectively. GM-CSF was tested to detect unpolarized cells. IL-4 was not tested because of potential consumption by CD4+ T cells in culture in the assay used (44). IL-10 and IL-13 were tested in a limited number of subjects to confirm production of Th2 cytokines. Fig. 1 shows the results obtained with normal donors. CD4+ T cells produced mostly GM-CSF and/or IFN-{gamma}. IL-5 was produced in the majority of cases in association with the other two cytokines. On the contrary, CD4+ T cells from most pancreatic carcinoma patients produced IL-5 alone or in association with GM-CSF and/or IFN-{gamma} (Fig. 2). IL-10 was rarely produced while IL-13 paralleled IL-5 production (data not shown). Collectively, these results suggest an impairment of the anti-CEA CD4+ T cell immunity in pancreatic cancer.


Figure 1
View larger version (26K):
[in this window]
[in a new window]

  		FIGURE 1. CD4+ T cells against CEA in normal donors produce mainly GM-CSF and IFN-{gamma}. CD4+ T cells from 23 normal donors were tested as described in Materials and Methods with the CEA peptides. Results for the 15 responders are shown. For [3H]thymidine incorporation assays, all data from the five replicates in the absence (CD4 + APC) or in the presence of the peptides are reported. For cytokine release assays, supernatants from the five wells were pooled and determination was performed in duplicate. Samples from peptide-stimulated cells were considered positive if at least double the values from unstimulated cells and above 50 pg/ml. The mean value of the unstimulated cells was subtracted from each value of peptide-stimulated cells and data reported are means of duplicate determinations ± SD. Responses significantly higher than the blanks (i.e., CD4 + APC) are indicated as *, p < 0.05; **, 0.001 < p < 0.05; and ***, p < 0.001 (determined by unpaired, one-tailed Student’s t test). n.t., Peptide not tested.

 

Figure 2
View larger version (26K):
[in this window]
[in a new window]

  		FIGURE 2. CD4+ T cells against CEA in pancreatic cancer patients produce mainly IL-5. CD4+ T cells from 44 pancreatic cancer patients were tested as described in Materials and Methods with the CEA peptides. Results for the 13 responders are shown. For [3H]thymidine incorporation assays, all data from the five replicates in the absence (CD4 + APC) or in the presence of the peptides are reported. For cytokine release assays, supernatants from the five wells were pooled and determination was performed in duplicate. Samples from peptide-stimulated cells were considered positive if at least double the values from unstimulated cells and above 50 pg/ml. The mean value of the unstimulated cells was subtracted from each value of peptide- stimulated cells and data reported are means of duplicate determinations ± SD. Responses significantly higher than the blanks (i.e., CD4 + APC) are indicated as *, p < 0.05; **, 0.001 < p < 0.05; and ***, p < 0.001 (determined by unpaired, one-tailed Student’s t test). n.t., Peptide not tested.

 
To rule out the presence of Th17 and Th3/Tr1 CD4+ T cells, IL-17 and TGF-β release were assayed in the culture supernatant of 12 patients (subjects 11, 29, and 37–46) and 2 normal donors (subjects 14 and 21). The two cytokines were not detected in any sample (data not shown).

Comparison between anti-CEA and antiviral CD4+ T cell immunity

To verify whether the Th2 skew in anti-CEA CD4+ T cell immunity observed in pancreatic carcinoma patients was generalized or CEA specific, we tested the reactivity of CD4+ T cells against the promiscuous viral sequences HA307–319 and EBNA2280–294 in 21 patients and 12 normal donors. The results are reported in detail in Figs. 3 and 4 and summarized in Fig. 5. In normal donors, the percentages of responders to HA and EBNA2 were almost comparable to the percentage of responders to CEA (Fig. 5A, left panel). On the contrary, in pancreatic cancer patients, while the percentage of responders to CEA was lower than in normal donors the percentages of responders to HA and EBNA2 were even higher (Fig. 5A, right panel), although not significantly different (p = 0.9183 and p = 0.3022, respectively, as from {chi}2 test). Most importantly, although CEA-specific CD4+ T cells in patients produced mostly IL-5 (Fig. 2), HA- and EBNA2-specific CD4+ T cells produced mostly IFN-{gamma} and/or GM-CSF (Figs. 3 and 5B, right panel). The pattern of cytokine produced by CD4+ T cells against the three different Ags did not change substantially in normal donors (Figs. 1, 4, and 5B, left panel). When comparing Th1 (IFN-{gamma} production), Th2 (IL-5 production) and mixed Th1/Th2 (IFN-{gamma} and IL-5 production) responses in normal donors and pancreatic cancer patients, we found that normal donors had a statistically significant higher percentage of Th1 CEA-specific CD4+ T cell responses than patients (p = 0.003, determined by a one-tailed z test) and patients showed a statistically significant higher percentage of Th2 CEA-specific CD4+ T cell responses than normal donors (p = 0.005, determined by a one-tailed z test; Fig. 5C, upper panel). The percentages of Th1, Th2, and Th1/Th2 responses toward HA and EBNA2 between normal donors and patients did not differ substantially (Fig. 5C, middle and lower panels). As expected, CD4+ T cells in the presence of the positive control (i.e., PHA) produced all three cytokines in the majority of the cases in both normal donors and patients (Fig. 5B, lower panels).


Figure 3
View larger version (26K):
[in this window]
[in a new window]

  		FIGURE 3. CD4+ T cells against HA and EBNA2 in pancreatic cancer patients. CD4+ T cells from 21 pancreatic cancer patients were tested as described in Materials and Methods with the HA and EBNA2 peptides. Results for the 17 responders are shown. For [3H]thymidine incorporation assays, all data from the five replicates in the absence (CD4 + APC) or in the presence of the peptides are reported. For cytokine release assays, supernatants from the five wells were pooled and determination was performed in duplicate. Samples from peptide-stimulated cells were considered positive if at least double the values from unstimulated cells and above 50 pg/ml. The mean value of the unstimulated cells was subtracted from each value of peptide- stimulated cells and data reported are means of duplicate determinations ± SD. Responses significantly higher than the blanks (i.e., CD4 + APC) are indicated as *, p < 0.05; **, 0.001 < p < 0.05; and ***, p < 0.001 (determined by unpaired, one-tailed Student’s t test).

 

Figure 4
View larger version (31K):
[in this window]
[in a new window]

  		FIGURE 4. CD4+ T cells against HA and EBNA2 in normal donors. CD4+ T cells from 12 normal donors were tested as described in Materials and Methods with the HA and EBNA2 peptides. Results for the nine responders are shown. For [3H]thymidine incorporation assays, all data from the five replicates in the absence (CD4 + APC) or in the presence of the peptides are reported. For cytokine release assays, supernatants from the five wells were pooled and determination was performed in duplicate. Samples from peptide-stimulated cells were considered positive if at least double the values from unstimulated cells and above 50 pg/ml. The mean value of the unstimulated cells was subtracted from each value of peptide-stimulated cells and data reported are means of duplicate determinations ± SD. Responses significantly higher than the blanks (i.e., CD4 + APC) are indicated as *, p < 0.05; **, 0.001 < p < 0.05, and ***, p < 0.001 (determined by unpaired, one-tailed Student’s t test). n.t., Peptide not tested; n.d., cytokine not determined.

 

Figure 5
View larger version (28K):
[in this window]
[in a new window]

  		FIGURE 5. Anti-CEA and antiviral-specific CD4+ T cell immunity in normal donors and pancreatic cancer patients. A, Percentage of responders to CEA, HA, and EBNA2 peptides in normal donors (left) and pancreatic carcinoma patients (right). B, Percentage of responders among normal donors (left) and pancreatic carcinoma patients (right) to CEA, HA, EBNA2, and PHA in the different settings of cytokines produced alone or in combination. C, Responders are categorized based on the production of Th1 (IFN-{gamma}), Th2 (IL-5), and Th1/Th2 (IFN-{gamma} and IL-5) cytokines in response to CEA, HA, and EBNA2 peptides. Statistical differences in percentage of responders were determined by a one-tailed z test and are indicated as **, 0.001 < p < 0.05.

 
GATA-3+ lymphoid cells predominate within immune infiltrates in pancreatic cancer tissues

Immunohistochemical analysis was performed to characterize the immune infiltrate in 13 pancreatic cancer tissues (Table II). We determined the presence of CD8+ and CD4+ T cells and the expression by lymphoid cells of transcription factors T-bet and GATA-3 to distinguish between Th1 and Th2 cells, respectively (45). FoxP3 expression was also evaluated to identify Tr/recently activated effector T cells (46). We found immune infiltrates in all cases, mostly within the tumor stroma; although, as shown in Table II, the intensity differed among samples. Both CD8+ and CD4+ T cells were represented without a significant predominance of one or the other subset. When considering the Th1/Th2 polarization of tumor-infiltrating lymphocytes, we found that the number of GATA-3+ was significantly higher than the number of T-bet+ lymphoid cells (p < 0.001, determined by unpaired, one-tailed Student’s t test). FoxP3+ T cells were also present and their number was significantly higher than the number of T-bet+ cells (p < 0.02) but lower than GATA-3+ cells (p < 0.03). Interestingly, patients 39, 41, and 43, who had CEA-specific CD4+ T cells in the blood that produced discrete amounts of IFN-{gamma}, showed the highest number of T-bet+ in their infiltrate. Fig. 6 shows representative examples of immunohistochemical analyses in patients with an intense (patient 4, left panels) or a moderate infiltrate (patient 15, middle panels), and in patients with the highest IFN-{gamma} production in the blood (patient 39, right panels). Accordingly, in these last patients (subjects 39, 41, and 42), the number of T-bet+ was comparable or higher than the number of GATA-3+ cells (Table II). Cytoplasmic and nuclear GATA-3 staining was present in neoplastic epithelial cells, as also previously described (47); although in contrast to the report of Kalthoff and colleagues (48), we did not observe any FoxP3 staining in tumor cells.


View this table:
[in this window]
[in a new window]

  		Table II. Immunohistochemical analysis of immune infiltratesa

 

Figure 6
View larger version (146K):
[in this window]
[in a new window]

  		FIGURE 6. Immunohistochemical analysis of immune infiltrates. Representative staining of tumor samples from patient 4 (left panels), patient 15 (middle panels), and patient 39 (right panels) showing CD8+ (A–C)-, CD4+ (D–F)-, T-bet (G–I)-, GATA-3 (L–N)-, and FoxP3 (O–Q)-positive lymphoid cells within cancer cell stroma (original magnification, x200). Cytoplasmic and nuclear GATA-3 staining are also present in neoplastic epithelial cells. The arrows in the T-bet and FoxP3 staining help the reader by indicating the positive cells in panels with a low/moderate number of positive cells.

 

   Discussion
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
Disclosures
 References
 
In this study, we compare the spontaneous antiviral and anti-CEA CD4+ T cell immunity in pancreatic carcinoma patients and normal donors. We found Th1-type antiviral immunity present in both patients and normal donors. On the contrary, in patients compared with normal donors, we found quantitative (i.e., reduced number of responders) and qualitative impairment (i.e., Th2 type) of anti-CEA immunity. In agreement, as shown by immunohistochemical analysis, the number of GATA-3+ lymphoid cells was predominant among tumor-infiltrating lymphocytes. These results show for the first time that pancreatic cancer patients have a Th2 skew in CD4+ T cells specifically directed against a tumor-associated Ag, suggesting that in pancreatic cancer local more than systemic immune-subversive factors are responsible for the observed immune deviation.

Previous studies (6, 7, 8, 9, 10) have addressed the presence of spontaneous immune responses against tumor-associated Ags in pancreatic carcinoma patients. CD8+ and CD4+ T cells specific for SART1–3 and ART4 (7), wild-type and mutated ras (6, 8, 10), MUC-1 (9), and the autologous tumors (9) were found in variable percentages. In most cases, effectors were shown to produce IFN-{gamma} in response to their cognate Ags, suggesting a Th1 polarization. However, in the majority of reports, the release of Th2 cytokines by Ag-specific T cells was not investigated or was limited to IL-4 and IL-10. We also found 10% of patients with IFN-{gamma}-producing CD4+ T cells and do not exclude the possibility that a Th1 along with a Th2 repertoire against the same or different tumor Ags coexist in the patients. Beckhove and colleagues (9) reported higher frequencies of T cells specific for the autologous tumor in the bone marrow than in peripheral blood. They explained this discrepancy with the presence of a suppressive microenvironment at the tumor site, which might negatively affect the antitumor response measurable in the blood. On the contrary, the presence of low amounts of pancreatic tumor cells in the bone marrow could explain priming/restimulation of tumor-reactive T cells in a more favorable environment. It would be interesting to verify whether CEA-specific CD4+ T cells are also present in the bone marrow and whether these cells are endowed with the same or different effector functions compared with the ones detected in the blood.

A distinctive feature of our results is the observed impairment of the anti-CEA CD4+ T cell immunity in the patients compared with normal donors: this impairment being both quantitative and qualitative. With the experimental system used, we found proliferation and/or cytokine released by CEA-reactive CD4+ T cells in ~65% of normal donors. We previously (39) failed to detect spontaneous anti-CEA CD4+ T cells in a small cohort of seven normal subjects: important differences between the two studies are the number of subjects tested and the culture conditions used. The importance of the culture conditions in monitoring of immune responses for both background and low frequency detection has been recently reported (49). A high number of normal donors with CEA-reactive CD4+ T cells was also recently reported (50). It has been shown (51, 52) that CEA is expressed in the thymus with potential deletion of CEA-specific T cells. Nonetheless, it has been reported that negative selection in the thymus is incomplete, causing T cells with intermediate to low avidity escape into the periphery (53). These autoreactive T cells may then be kept under control by regulatory cells or by avoiding contact with the Ag, which is expressed at a low level in normal adult tissues. This second hypothesis is further supported by the possibility that CEA expressed in the thymus and in the tumor might have different posttranslational modifications, as it has been shown for MUC-1 (51). It is interesting to note that, although we confirmed CEA177–189/355–367 as immunodominant (38), other sequences (CEA99–111 and CEA568–582) were preferentially recognized by a larger number of normal donors or patients, respectively (Figs. 1 and 2). Indeed, de novo priming might have occurred in the patients with a different form of CEA and/or in the presence of different Ag-processing machineries, explaining the tendency to a preferential repertoire of the epitopes recognized within the two populations.

A Th2-immune deviation in CD4+ T cell responses against tumor-associated Ags may be due to either generalized or local factors. The intact repertoire of Th1 antiviral CD4+ T cells found here points to factor(s) present at the tumor site. Due to the likely existence of a preexisting repertoire of anti-CEA Th1 CD4+ T cells in pancreatic cancer patients, it may be speculated that during cancer development Th1 cells are recruited to the tumor site and exhausted/deleted after having exerted their effector function. At the tumor site, tumor Ag(s)-loaded dendritic cells conditioned by factors present within tumor cells or tumor stroma might favor differentiation of dendritic cells with Th2-polarizing capability. De novo priming of Th2 cytokine-secreting cells might in turn further conditioning the tumor microenvironment and contribute to fibrosis, which is a hallmark in pancreatic cancer. Indeed, it has been shown in several models that fibrogenesis is strongly linked with the development of Th2 responses through activation by Th2 cytokines of collagen synthesis by fibroblasts (54). The observed infiltration at the tumor site of GATA-3+ lymphoid cells well correlates with this interpretation. Indeed, the predominance of GATA-3+ cells also in the infiltrate of patients with no anti-CEA response in the blood (see Table II) suggests that the Th2 skew may apply to other tumor-associated Ag-specific immune responses. Moreover, the presence of CD8+ T cells in the absence or in the presence of a very scarce T-bet+ lymphoid infiltrate possibly suggests also impairment in the CD8+ T cell compartment.

It has been reported (28, 29) that successful therapeutic intervention may allow for a "resetting" of the Th1/Th2 imbalance in melanoma and renal cancer patients. Although therapy benefits and long-term disease-free survival after surgery are rare in pancreatic patients, it would be interesting in future studies to evaluate the effects of surgery and/or chemoradiotherapy on the quality of anti-CEA CD4+ T cell immunity.

Future studies are also needed to investigate the extent of the Th2 skew that we have found by investigating the CD4+ T cell response to other tumor-associated Ags, to identify factors and mechanisms responsible for the observed skew, and to devise strategies to block further development of Th2 cells and to revert the existing effectors to a Th1 type.


   Acknowledgments
 
We thank Anna Maria Paganoni for statistical analyses and Matteo Bellone for critical reading of this manuscript.


   Disclosures
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
 Disclosures
References
 
The authors have no financial conflict of interest.


   Footnotes
 
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

1 This work was supported by: the Cancer Research Institute (preclinical grant), the European Community (Dendritic cells and Novel Immunotherapies), the Italian Ministry of Health, the Italian Ministry of University and Scientific Research, and the Rich Foundation (LDP Pancreas Cancer Research Project). Back

2 Address correspondence and reprint requests to Dr. Maria Pia Protti, Tumor Immunology Unit, Cancer Immunotherapy and Gene Therapy Program, DIBIT, Scientific Institute H. San Raffaele, Via Olgettina 58, 20132 Milan. E-mail address: m.protti{at}hsr.it Back

3 Abbreviations used in this paper: Tr, T regulatory; CEA, carcinoembryonic Ag; EBNA2, Epstein-Barr nuclear Ag 2; HA, influenza hemagglutinin. Back

Received for publication April 4, 2008. Accepted for publication August 20, 2008.


   References
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
 Disclosures
 References
 

  1. Warshaw, A. L., C. Fernández-del Castillo. 1992. Pancreatic carcinoma. N. Engl. J. Med. 326: 455-465. [Medline]
  2. Jemal, A., R. Siegel, E. Ward, T. Murray, J. Xu, C. Smigal, M. J. Thun. 2006. Cancer statistics, 2006. CA Cancer J. Clin. 56: 106-130. [Abstract/Free Full Text]
  3. Kleeff, J., P. Beckhove, I. Esposito, S. Herzig, P. E. Huber, J. M. Löhr, H. Friess. 2007. Pancreatic cancer microenvironment. Int. J. Cancer 121: 699-705. [Medline]
  4. Ademmer, K., M. Ebert, F. Müller-Ostermeyer, H. Friess, M. W. Büchler, W. Schubert, P. Malfertheiner. 1998. Effector T lymphocyte subsets in human pancreatic cancer: detection of CD8+CD18+ cells and CD8+CD103+ cells by multi-epitope imaging. Clin. Exp. Immunol. 112: 21-26. [Medline]
  5. Fukunaga, A., M. Miyamoto, Y. Cho, S. Murakami, Y. Kawarada, T. Oshikiri, K. Kato, T. Kurokawa, M. Suzuoki, Y. Nakakubo, et al 2004. CD8+ tumor-infiltrating lymphocytes together with CD4+ tumor-infiltrating lymphocytes and dendritic cells improve the prognosis of patients with pancreatic adenocarcinoma. Pancreas 28: e26-e31. [Medline]
  6. Qin, H., W. Chen, M. Takahashi, M. L. Disis, D. R. Byrd, L. McCahill, K. A. Bertram, R. G. Fenton, D. J. Peace, M. A. Cheever. 1995. CD4+ T-cell immunity to mutated ras protein in pancreatic and colon cancer patients. Cancer Res. 55: 2984-2987. [Abstract/Free Full Text]
  7. Suzuki, N., Y. Maeda, S. Tanaka, N. Hida, T. Mine, K. Yamamoto, M. Oka, K. Itoh. 2002. Detection of peptide-specific cytotoxic T-lymphocyte precursors used for specific immunotherapy of pancreatic cancer. Int. J. Cancer 98: 45-50. [Medline]
  8. Shono, Y., H. Tanimura, M. Iwahashi, T. Tsunoda, M. Tani, H. Tanaka, K. Matsuda, H. Yamaue. 2003. Specific T-cell immunity against Ki-ras peptides in patients with pancreatic and colorectal cancers. Br. J. Cancer 88: 530-536. [Medline]
  9. Schmitz-Winnenthal, F. H., C. Volk, K. Z'graggen, L. Galindo, D. Nummer, Y. Ziouta, M. Bucur, J. Weitz, V. Schirrmacher, M. W. Büchler, P. Beckhove. 2005. High frequencies of functional tumor-reactive T cells in bone marrow and blood of pancreatic cancer patients. Cancer Res. 65: 10079-10087. [Abstract/Free Full Text]
  10. Kubuschok, B., F. Neumann, R. Breit, M. Sester, C. Schormann, C. Wagner, U. Sester, F. Hartmann, M. Wagner, K. Remberger, et al 2006. Naturally occurring T-cell response against mutated p21 ras oncoprotein in pancreatic cancer. Clin. Cancer Res. 12: 1365-1372. [Abstract/Free Full Text]
  11. Ungefroren, H., M. Voss, M. Jansen, C. Roeder, D. Henne-Bruns, B. Kremer, H. Kalthoff. 1998. Human pancreatic adenocarcinomas express Fas and Fas ligand yet are resistant to Fas-mediated apoptosis. Cancer Res. 58: 1741-1749. [Abstract/Free Full Text]
  12. Ungefroren, H., M. Voss, W. V. Bernstorff, A. Schmid, B. Kremer, H. Kalthoff. 1999. Immunological escape mechanisms in pancreatic carcinoma. Ann. NY Acad. Sci. 880: 243-251. [Medline]
  13. von Bernstorff, W., M. Voss, S. Freichel, A. Schmid, I. Vogel, C. Jöhnk, D. Henne-Bruns, B. Kremer, H. Kalthoff. 2001. Systemic and local immunosuppression in pancreatic cancer patients. Clin. Cancer Res. 7: 925s-932s. [Abstract/Free Full Text]
  14. Mukherjee, P., A. R. Ginardi, C. S. Madsen, T. L. Tinder, F. Jacobs, J. Parker, B. Agrawal, B. M. Longenecker, S. J. Gendler. 2001. MUC1-specific CTLs are non-functional within a pancreatic tumor microenvironment. Glycoconj. J. 18: 931-942. [Medline]
  15. Ryschich, E., T. Nötzel, U. Hinz, F. Autschbach, J. Ferguson, I. Simon, J. Weitz, B. Fröhlich, E. Klar, M. W. Büchler, J. Schmidt. 2005. Control of T-cell-mediated immune response by HLA class I in human pancreatic carcinoma. Clin. Cancer Res. 11: 498-504. [Abstract/Free Full Text]
  16. Esposito, I., M. Menicagli, N. Funel, F. Bergmann, U. Boggi, F. Mosca, G. Bevilacqua, D. Campani. 2004. Inflammatory cells contribute to the generation of an angiogenic phenotype in pancreatic ductal adenocarcinoma. J. Clin. Pathol. 57: 630-636. [Abstract/Free Full Text]
  17. Bellone, G., A. Carbone, C. Smirne, T. Scirelli, A. Buffolino, A. Novarino, A. Stacchini, O. Bertetto, G. Palestro, C. Sorio, et al 2006. Cooperative induction of a tolerogenic dendritic cell phenotype by cytokines secreted by pancreatic carcinoma cells. J. Immunol. 177: 3448-3460. [Abstract/Free Full Text]
  18. Bharadwaj, U., M. Li, R. Zhang, C. Chen, Q. Yao. 2007. Elevated interleukin-6 and G-CSF in human pancreatic cancer cell conditioned medium suppress dendritic cell differentiation and activation. Cancer Res. 67: 5479-5488. [Abstract/Free Full Text]
  19. Liyanage, U. K., T. T. Moore, H. G. Joo, Y. Tanaka, V. Herrmann, G. Doherty, J. A. Drebin, S. M. Strasberg, T. J. Eberlein, P. S. Goedegebuure, D. C. Linehan. 2002. Prevalence of regulatory T cells is increased in peripheral blood and tumor microenvironment of patients with pancreas or breast adenocarcinoma. J. Immunol. 169: 2756-2761. [Abstract/Free Full Text]
  20. Nummer, D., E. Suri-Payer, H. Schmitz-Winnenthal, A. Bonertz, L. Galindo, D. Antolovich, M. Koch, M. Büchler, J. Weitz, V. Schirrmacher, P. Beckhove. 2007. Role of tumor endothelium in CD4+CD25+ regulatory T cell infiltration of human pancreatic carcinoma. J. Natl. Cancer Inst. 99: 1188-1199. [Abstract/Free Full Text]
  21. Abbas, A. K., K. M. Murphy, A. Sher. 1996. Functional diversity of helper T lymphocytes. Nature 383: 787-793. [Medline]
  22. Romagnani, S.. 1997. The Th1/Th2 paradigm. Immunol. Today 18: 263-266. [Medline]
  23. McGeachy, M. J., D. J. Cua. 2008. Th17 cell differentiation: the long and winding road. Immunity 28: 445-453. [Medline]
  24. Cottrez, F., H. Groux. 2004. Specialization in tolerance: innate CD4+CD25+ versus acquired TR1 and TH3 regulatory T cells. Transplantation 77: S12-S15. [Medline]
  25. Hung, K., R. Hayashi, A. Lafond-Walker, C. Lowenstein, D. Pardoll, H. Levitsky. 1998. The central role of CD4+ T cells in the antitumor immune response. J. Exp. Med. 188: 2357-2368. [Abstract/Free Full Text]
  26. Nishimura, T., K. Iwakabe, M. Sekimoto, Y. Ohmi, T. Yahata, M. Nakui, T. Sato, S. Habu, H. Tashiro, M. Sato, A. Ohta. 1999. Distinct role of antigen-specific T helper type 1 (Th1) and Th2 cells in tumor eradication in vivo. J. Exp. Med. 190: 617-627. [Abstract/Free Full Text]
  27. Dranoff, G., E. Jaffee, A. Lazenby, P. Golumbek, H. Levitsky, K. Brose, V. Jackson, H. Hamada, D. Pardoll, R. C. Mulligan. 1993. Vaccination with irradiated tumor cells engineered to secrete murine granulocyte-macrophage colony-stimulating factor stimulates potent, specific, and long-lasting anti-tumor immunity. Proc. Natl. Acad. Sci. USA 90: 3539-3543. [Abstract/Free Full Text]
  28. Tatsumi, T., L. S. Kierstead, E. Ranieri, L. Gesualdo, F. P. Schena, J. H. Finke, R. M. Bukowski, J. Mueller-Berghaus, J. M. Kirkwood, W. W. Kwok, W. J. Storkus. 2002. Disease-associated bias in T helper type 1 (Th1)/Th2 CD4+ T cell responses against MAGE-6 in HLA-DRB10401+ patients with renal cell carcinoma or melanoma. J. Exp. Med. 196: 619-628. [Abstract/Free Full Text]
  29. Tatsumi, T., C. J. Herrem, W. C. Olson, J. H. Finke, R. M. Bukowski, M. S. Kinch, E. Ranieri, W. J. Storkus. 2003. Disease stage variation in CD4+ and CD8+ T-cell reactivity to the receptor tyrosine kinase EphA2 in patients with renal cell carcinoma. Cancer Res. 63: 4481-4489. [Abstract/Free Full Text]
  30. Slager, E. H., M. Borghi, C. E. van der Minne, C. A. Aarnoudse, M. J. Havenga, P. I. Schrier, S. Osanto, M. Griffioen. 2003. CD4+ Th2 cell recognition of HLA-DR-restricted epitopes derived from CAMEL: a tumor antigen translated in an alternative open reading frame. J. Immunol. 170: 1490-1497. [Abstract/Free Full Text]
  31. Marturano, J., R. Longhi, V. Russo, M. P. Protti. 2008. Endosomal proteases influence the repertoire of MAGE-A3 epitopes recognized in vivo by CD4+ T cells. Cancer Res. 68: 1555-1562. [Abstract/Free Full Text]
  32. Bellone, G., A. Turletti, E. Artusio, K. Mareschi, A. Carbone, D. Tibaudi, A. Robecchi, G. Emanuelli, U. Rodeck. 1999. Tumor-associated transforming growth factor-β and interleukin-10 contribute to a systemic Th2 immune phenotype in pancreatic carcinoma patients. Am. J. Pathol. 155: 537-547. [Abstract/Free Full Text]
  33. Berinstein, N. L.. 2002. Carcinoembryonic antigen as a target for therapeutic anticancer vaccines: a review. J. Clin. Oncol. 20: 2197-2207. [Abstract/Free Full Text]
  34. Allum, W. H., H. J. Stokes, F. Macdonald, J. W. Fielding. 1986. Demonstration of carcinoembryonic antigen (CEA) expression in normal, chronically inflamed, and malignant pancreatic tissue by immunohistochemistry. J. Clin. Pathol. 39: 610-614. [Abstract/Free Full Text]
  35. Kobayashi, H., R. Omiya, M. Ruiz, E. Huarte, P. Sarobe, J. J. Lasarte, M. Herraiz, B. Sangro, J. Prieto, F. Borras-Cuesta, E. Celis. 2002. Identification of an antigenic epitope for helper T lymphocytes from carcinoembryonic antigen. Clin. Cancer Res. 8: 3219-3225. [Abstract/Free Full Text]
  36. Ruiz, M., H. Kobayashi, J. J. Lasarte, J. Prieto, F. Borrás-Cuesta, E. Celis, P. Sarobe. 2004. Identification and characterization of a T-helper peptide from carcinoembryonic antigen. Clin. Cancer Res. 10: 2860-2867. [Abstract/Free Full Text]
  37. Shen, L., R. Schroers, J. Hammer, X. F. Huang, S. Y. Chen. 2004. Identification of a MHC class-II restricted epitope in carcinoembryonic antigen. Cancer Immunol. Immunother. 53: 391-403. [Medline]
  38. Campi, G., M. Crosti, G. Consogno, V. Facchinetti, B. M. Conti-Fine, R. Longhi, G. Casorati, P. Dellabona, M. P. Protti. 2003. CD4+ T cells from healthy subjects and colon cancer patients recognize a carcinoembryonic antigen-specific immunodominant epitope. Cancer Res. 63: 8481-8486. [Abstract/Free Full Text]
  39. Crosti, M., R. Longhi, G. Consogno, G. Melloni, P. Zannini, M. P. Protti. 2006. Identification of novel subdominant epitopes on the carcinoembryonic antigen recognized by CD4+ T cells of lung cancer patients. J. Immunol. 176: 5093-5099. [Abstract/Free Full Text]
  40. Sturniolo, T., E. Bono, J. Ding, L. Raddrizzani, O. Tuereci, U. Sahin, M. Braxenthaler, F. Gallazzi, M. P. Protti, F. Sinigaglia, J. Hammer. 1999. Generation of tissue-specific and promiscuous HLA ligand databases using DNA microarrays and virtual HLA class II matrices. Nat. Biotechnol. 17: 555-561. [Medline]
  41. Jardetzky, T. S., J. C. Gorga, R. Busch, J. Rothbard, J. L. Strominger, D. C. Wiley. 1990. Peptide binding to HLA-DR1: a peptide with most residues substituted to alanine retains MHC binding. EMBO J. 9: 1797-1803. [Medline]
  42. Omiya, R., C. Buteau, H. Kobayashi, C. V. Paya, E. Celis. 2002. Inhibition of EBV-induced lymphoproliferation by CD4+ T cells specific for an MHC class II promiscuous epitope. J. Immunol. 169: 2172-2179. [Abstract/Free Full Text]
  43. Curnis, F., A. Gasparri, A. Sacchi, R. Longhi, A. Corti. 2004. Coupling tumor necrosis factor-{alpha} with {alpha}V integrin ligands improves its antineoplastic activity. Cancer Res. 64: 565-571. [Abstract/Free Full Text]
  44. Seresini, S., M. Origoni, F. Lillo, L. Caputo, A. M. Paganoni, S. Vantini, R. Longhi, G. Taccagni, A. Ferrari, C. Doglioni, et al 2007. IFN-{gamma} produced by human papilloma virus-18 E6-specific CD4+ T cells predicts the clinical outcome after surgery in patients with high-grade cervical lesions. J. Immunol. 179: 7176-7183. [Abstract/Free Full Text]
  45. Nakayama, T., M. Yamashita. 2008. Initiation and maintenance of Th2 cell identity. Curr. Opin. Immunol. 20: 265-271. [Medline]
  46. Allan, S. E., S. Q Crome, N. K. Crellin, L. Passerini, T. S. Steiner, R. Bacchetta, M. G. Roncarolo, M. K. Levings. 2007. Activation-induced FOXP3 in human T effector cells does not suppress proliferation or cytokine production. Int. Immunol. 19: 345-354. [Abstract/Free Full Text]
  47. Gulbinas, A., P. O. Berberat, Z. Dambrauskas, T. Giese, N. Giese, F. Autschbach, J. Kleeff, S. Meuer, M. W. Büchler, H. Friess. 2006. Aberrant gata-3 expression in human pancreatic cancer. J. Histochem. Cytochem. 54: 161-169. [Abstract/Free Full Text]
  48. Hinz, S., L. Pagerols-Raluy, H. H. Oberg, O. Ammerpohl, S. Grüssel, B. Sipos, R. Grützmann, C. Pilarsky, H. Ungefroren, H. D. Saeger, et al 2007. Foxp3 expression in pancreatic carcinoma cells as a novel mechanism of immune evasion in cancer. Cancer Res. 67: 8344-8350. [Abstract/Free Full Text]
  49. Janetzki, S., K. S. Panageas, L. Ben-Porat, J. Boyer, C. M. Britten, T. M. Clay, M. Kalos, H. T. Maecker, P. Romero, J. Yuan, et al 2008. Results and harmonization guidelines from two large-scale international Elispot proficiency panels conducted by the Cancer Vaccine Consortium (CVC/SVI). Cancer Immunol. Immunother. 57: 303-315. [Medline]
  50. Pickford, W. J., A. J. Watson, R. N. Barker. 2007. Different forms of helper tolerance to carcinoembryonic antigen: ignorance and regulation. Clin. Cancer Res. 13: 4528-4537. [Abstract/Free Full Text]
  51. Cloosen, S., J. Arnold, M. Thio, G. M. Bos, B. Kyewski, W. T. Germeraad. 2007. Expression of tumor-associated differentiation antigens, MUC1 glycoforms and CEA, in human thymic epithelial cells: implications for self-tolerance and tumor therapy. Cancer Res. 67: 3919-3926. [Abstract/Free Full Text]
  52. Bos, R., S. van Duikeren, T. van Hall, P. Kaaijk, R. Taubert, B. Kyewski, L. Klein, C. J. Melief, R. Offringa. 2005. Expression of a natural tumor antigen by thymic epithelial cells impairs the tumor-protective CD4+ T-cell repertoire. Cancer Res. 65: 6443-6449. [Abstract/Free Full Text]
  53. Zehn, D., M. J. Bevan. 2006. T cells with low avidity for a tissue-restricted antigen routinely evade central and peripheral tolerance and cause autoimmunity. Immunity 25: 261-270. [Medline]
  54. Wynn, T. A.. 2004. Fibrotic disease and the TH1/TH2 paradigm. Nat. Rev. Immunol. 4: 583-594. [Medline]




This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow Request Permissions
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Tassi, E.
Right arrow Articles by Protti, M. P.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Tassi, E.
Right arrow Articles by Protti, M. P.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS