Purpose: Gemcitabine is a promising adjuvant treatment for patients with resected pancreatic adenocarcinoma and its use in combination with radiotherapy is under exploration. Human equilibrative nucleoside transporter 1 (hENT1) and human concentrative nucleoside transporter (hCNT) 1 and 3 are the major transporters responsible for 2′,2′-difluoro-2-deoxycytidine (gemcitabine) uptake into cells. The aim of this study was to determine patients' outcome according to the expression of hENT1 and hCNT3 in tumoral cells after postoperative gemcitabine-based chemoradiation regimen.

Experimental Design: We studied tumor blocks from 45 pancreatic adenocarcinoma patients treated with gemcitabine-based chemoradiation after curative resection and assessed hENT1 and hCNT3 expression using immunohistochemistry.

Results: When adjusted for the effects of lymph node ratio and tumor diameter, patients with high hENT1 expression had significantly longer disease-free survival and overall survival (OS) than patients with low expression, whereas high hCNT3 expression was only associated with longer OS. In a combined analysis, patients with two favorable prognostic factors (hENT1high/hCNT3high expression) had a longer survival (median OS, 94.8 months) than those having one (median OS, 18.7 months) or no (median OS, 12.2 months) favorable prognostic factor.

Conclusions: Pancreatic adenocarcinoma patients with a high expression of hENT1 and hCNT3 immunostaining have a significantly longer survival after adjuvant gemcitabine-based chemoradiation. These biomarkers deserve prospective evaluation in patients receiving gemcitabine-based adjuvant therapy.

Translational Relevance

Gemcitabine is the standard of care after resection of pancreatic cancer patients. Human equilibrative nucleoside transporter 1 (hENT-1) and human concentrative nucleoside transporter (hCNT) 1 and 3 are the major transporters responsible for gemcitabine uptake into cells. Our work suggests the prognostic value of hENT1 and hCNT3 protein expression in resected pancreatic cancer patients treated with gemcitabine monotherapy followed by concomitant radiotherapy and gemcitabine. These translational data suggest that the combination of hENT1 and hCNT3 expression in tumor cells play an important role in patients' clinical outcome after gemcitabine-based adjuvant therapy. To our knowledge, this is the first description of the importance of nucleoside transporters in predicting survival of pancreatic cancer patients in the adjuvant setting.

Pancreatic head adenocarcinoma has a very poor prognosis. Even after curative surgery, which concerns only 10% to 15% of patients at the time of the diagnosis, 5-year survival is only 10% to 20% and median survival is 20 to 24 months (1, 2). This poor survival rate is attributed to a high rate of local recurrence and development of distant metastases even after complete resection of the tumor (R0 surgery). Several multimodal adjuvant and neoadjuvant therapies have been developed to reduce both local and systemic recurrence of the disease and improve survival. Nevertheless, the value of such adjuvant radiotherapy and/or chemotherapy regimens remains questionable because several trials have failed to draw any firm conclusions. Most of these adjuvant treatments were based on the combination of radiation and 5-fluorouracil (25).

Gemcitabine is a 2′,2′-difluoro-2-deoxycytidine analogue that inhibits DNA replication and repair. Gemcitabine monotherapy has clinically important activity in advanced and metastatic pancreatic adenocarcinoma and for which it is now the standard of care (6). In the adjuvant setting, it was enhanced disease-free survival (DFS) and survival compared with surgery alone (7, 8) and significantly increased DFS, but not overall survival (OS), when added to 5-fluorouracil-based chemoradiotherapy in a recent RTOG trial (9).

As gemcitabine possesses radiosensitizing properties in both in vitro and in vivo models (1013), there is a rationale to combine it with radiotherapy in the neoadjuvant and adjuvant setting. Emerging data, including ours (1117), have shown a good tolerability and feasibility and this combined regimen is now further being explored in a randomized phase III trial.

Gemcitabine is a prodrug that is phosphorylated by deoxycytidine kinase to its mononucleotide in the rate-limiting step of its cellular anabolism. Subsequent nucleotide kinases convert gemcitabine monophosphate to its active metabolites, gemcitabine diphosphate and gemcitabine triphosphate (18, 19). The de novo DNA synthesis pathway is blocked through inhibition of ribonucleotide reductase by gemcitabine diphosphate (20). Permeation of gemcitabine through the plasma membrane requires specialized integral membrane nucleoside transporters proteins. Among these transporters, the major mediators of gemcitabine uptake into human cells appear to be the human equilibrative nucleoside transporter 1 (hENT1) and, to a lesser degree, the human concentrative nucleoside transporter 3 (hCNT3; refs. 2124). Cells lacking hENT1 are highly resistant to gemcitabine (23, 25) and metastatic pancreatic cancer patients with hENT1 expression had significantly longer survival after gemcitabine chemotherapy than patients affected by tumors without detectable hENT1 (26).

Based on the apparent importance of nucleoside transporters to gemcitabine activity in pancreatic cancer, we explored the prognostic value of the tumoral expression of the nucleoside transporter proteins hENT1 and hCNT3 in a cohort of subjects with resected pancreatic adenocarcinoma, who received adjuvant gemcitabine-based chemoradiotherapy.

Patient populations. Patients from two Belgian multicentric phase II studies evaluating the feasibility of adjuvant gemcitabine-based combined regimen in resected pancreatic cancer patients between 2000 and 2003 were studied (15, 16). Both studies were approved by the ethics committees of each participating center. Standardized surgical procedures of pylorus-preserving duodenopancreatectomy or standard Whipple duodenopancreatectomy were done by one of two experienced surgeons in each center. During surgery, metallic or titanium clips were placed to delineate the original tumor limits for further determination of the target volume of irradiation. The eligibility criteria were similar for both studies: previous R0 resection with complete recovery from surgery within 8 weeks; stage I, II, or III (≥T2N0) histologically proven adenocarcinoma of the pancreatic head; age ≥18 years; Eastern Cooperative Oncology Group performance status (PS) ≤2; absence of previous or coexistent malignancy, adequate bone marrow, liver, and renal function; no previous chemotherapy or radiotherapy; and expected survival ≥6 months. Patients with cystic or neuroendocrine tumors were excluded. Proper documentation related to the number of the lymph nodes examined was mandatory. Abdominal computed tomography was done just before starting combined therapy to exclude the possible presence of early progressive disease. Patients were followed every 3 months by serial computed tomography and or magnetic resonance imaging and CA19.9 measurements until death to determine recurrence. Recurrence was proven by imaging.

Adjuvant treatment plan. Treatment was planned to start within 8 weeks post-surgery. Each patient was assigned to receive two cycles of gemcitabine 1,000 mg/m2 weekly as a 30 min infusion for 3 of 4 weeks. After 1 week rest, chemoradiation was started. Gemcitabine 300 mg/m2 as a 30 min infusion was given weekly for 5 consecutive weeks and administered 4 h before application of radiation (12). Patients received 40 Gy (n = 15) to 50.4 Gy (n = 30) according to the two trials (12, 13).

Immunohistochemistry. Anti-hENT1 and anti-hCNT3 monoclonal antibodies were produced and characterized as described previously (27, 28). Goat anti-mouse antibodies and horseradish peroxidase-labeled dextran polymer (DAKO EnVision+) were purchased from DAKO. All other reagents were of analytic grade and commercially available. Each formalin-fixed, paraffin-embedded pancreatic section (4-6 μm thick) was deparaffinized with three immersions in xylene baths (10 min each) followed by serial washes in graded alcohol from 100% to 50%. After rinsing in water, slides were placed in 250 mL high pH 1× DAKO target antigen retrieval solution and microwaved in TT-mega Milestone (ESBE Scientific) under control temperature and high pressure for 10 min at 100°C. After cooling in water for 6 min, slides were rinsed with water and peroxidase blocked in 3% H2O2 solution with methanol for 10 min and then washed in running water for 10 min. PBS (pH 7.2) was used for rinsing before incubation with appropriate dilutions of anti-hENT1 monoclonal antibodies. Slides with anti-hENT1 were incubated in a humidified chamber overnight at 4°C. Slides with anti-hCNT3 were incubated at room temperature for 30 min. The sections were then rinsed with PBS, immersed in buffer for 5 min, and incubated with goat anti-mouse dextran conjugate (DAKO Envision+) for 30 min followed by soaking in PBS. DAKO diaminobenzidine liquid chromagen was placed on the samples for 5 min and rinsed, after which the slides were soaked in 1% CuSO4 for another 5 min. Subsequently, the sections were rinsed, counterstained with hematoxylin, dehydrated through graded alcohol and xylene, and finally coverslipped. Negative controls were provided by omitting the primary antibodies and by using antibody preparations from which anti-hCNT3 and anti-hENT1 antibodies had been removed by incubation with an excess of the appropriate peptide.

Immunohistochemical evaluation. Quantitative scoring using light microscopy was done by a single pathologist (R.L.) blinded to clinical characteristics and outcomes. Immunohistochemical scoring was done only on invasive adenocarcinoma cells of one representative stained slide each for hENT1 and hCNT3.

Staining of hENT1 and hCNT3 protein was assigned a score from 0 to 3 based on staining intensity with 0 = no staining, 1 = weakly positive, 2 = moderately positive, and 3 = strongly positive. The percentage of adenocarcinoma cells staining at each intensity level was recorded for each specimen. A final score was determined by multiplying the intensity score and the percentage of the specimen. For example, if a specimen exhibited a staining distribution of 30% 1+ and 60% 2+, the final score is 150 (1 × 30 + 2 × 60). Therefore, the weighted scores ranged between 0 and 300.

Statistical methods. DFS was defined as the time from the date of surgery to the date of first relapse or death. OS was calculated from the date of surgery to the date of death. Data on survivors were censored at the last follow-up. The Kaplan-Meier method was used to plot DFS and OS (29) and differences based on biomarker status were evaluated with the log-rank test (30). A Cox proportional hazards univariate and multivariate model (31) were used to corroborate the association of clinical and pathologic factors and the tumoral expression of hENT1 and hCNT3 proteins with DFS and OS. Multivariate analyses used a backward stepwise procedure based on the likelihood ratio test. Data were analyzed using SPSS version 10.0 (SPSS). Statistical significance was set at P < 0.05.

Patient population, adjuvant therapy, and clinical outcome. A total of 45 patients were studied. They were 23 men and 22 women, median PS 0 (range, 0-1), and median age 58 years (range, 34-83). There were 34 patients with tumor-node-metastasis stage III and 18 patients with stage II. The median delay between surgery and the beginning of adjuvant therapy was 45 days (range, 24-74). Patients' pretreatment characteristics are described in Table 1. Adjuvant chemoradiation regimens were well tolerated and completed in 43 of 45 (95%) of the patients. WHO grade III/IV hematologic toxicities was noticed in 10 of 45 (21%) patients and nonhematologic grade III/IV in 3 of 45 (7%) patients. Only 2 patients required complication-related hospitalization after adjuvant treatment. Median follow-up was 21.9 months (range, 3.3-107.4) after surgery. Overall, median DFS and OS were 13.0 months (range, 1-107.4) and 21.9 months (range, 3.3-107.4), respectively. At the time of the last follow-up, 30 patients had died due to disease recurrence (distant n = 17, local n = 8, and local + distant n = 5). Fifteen patients were still alive at the last visit (range, 24.9-107.4 months).

Table 1.

Baseline patient characteristics

Sex  
    Male 23 
    Female 22 
Age, y, median (range) 56 (34-83) 
Eastern Cooperative Oncology Group PS, median (range) 0 (0-1) 
Tumor-node-metastasis classification (n 
    T1N1b 
    T2N0 
    T2N1 
    T2N1a 
    T2N1b 
    T3N0 10 
    T3N1 
    T3N1a 11 
    T3N1b 
    T4N1b 
Lymph node ratio, median (range) 0.2 (0-1) 
CA19-9 (IU/mL) at diagnosis, median (range) 49 (0.8-3327) 
Delay between surgery and start of adjuvant RCT, d, median (range) 47 (24-74) 
Survival, mo, median (range) 21.9 (3.3-107.4) 
DFS, mo, median (range) 13 (1.0-107.4) 
Sex  
    Male 23 
    Female 22 
Age, y, median (range) 56 (34-83) 
Eastern Cooperative Oncology Group PS, median (range) 0 (0-1) 
Tumor-node-metastasis classification (n 
    T1N1b 
    T2N0 
    T2N1 
    T2N1a 
    T2N1b 
    T3N0 10 
    T3N1 
    T3N1a 11 
    T3N1b 
    T4N1b 
Lymph node ratio, median (range) 0.2 (0-1) 
CA19-9 (IU/mL) at diagnosis, median (range) 49 (0.8-3327) 
Delay between surgery and start of adjuvant RCT, d, median (range) 47 (24-74) 
Survival, mo, median (range) 21.9 (3.3-107.4) 
DFS, mo, median (range) 13 (1.0-107.4) 

Immunostaining of hENT1 and hCNT3. The immunohistochemical hENT1 staining was present within the islets of Langerhans cells and lymphocytes, as reported previously, and used as internal positive controls. Within carcinoma cells, hENT1 protein and hCNT3 staining were localized to the cytoplasm (Fig. 1). In 45 tumor samples, 7 (16%) samples had uniformly detectable hENT1 immunostaining (intensity scores of 1+ and/or 2+ and/or 3+) with no regions lacking hENT1, and the remaining 38 samples possessed a proportion (range, 11-100%) of adenocarcinoma cells without detectable hENT1 (intensity score of 0). In contrast, each of these 45 tumors had detectable hCNT3 immunostaining of at least 1+ intensity. hCNT3 staining intensity was heterogeneous, both within some tumors, and in comparison between individual tumors. We dichotomized hENT1 staining based on the median value of the staining score. By this criterion, the cutpoint was 80, and patients were divided in two groups: (a) low hENT1 expression (staining score <80) and (b) high hENT1 expression (staining score ≥80; Fig. 1). For hCNT3, the cutoff value was 150. Accordingly, we define two groups: (a) low hCNT3 expression (staining score <150) and (b) high hCNT3 expression(staining score ≥150; Fig. 1).

Fig. 1.

Immunohistochemical staining of pancreatic adenocarcinoma for hENT1 and hCNT3 showing strong hENT1 (A) and hCNT3 (B) staining, low hCNT3 staining (C), and pancreatic adenocarcinoma cells lacking hENT1 (D), whereas adjacent lymphocytes show staining and provide positive internal control (original magnification, ×400).

Fig. 1.

Immunohistochemical staining of pancreatic adenocarcinoma for hENT1 and hCNT3 showing strong hENT1 (A) and hCNT3 (B) staining, low hCNT3 staining (C), and pancreatic adenocarcinoma cells lacking hENT1 (D), whereas adjacent lymphocytes show staining and provide positive internal control (original magnification, ×400).

Close modal

Correlation between hENT1 and hCNT3 expression and patients' outcome. The survival curves according to tumoral hENT1 and hCNT3 expression are shown in Fig. 2. Both OS and DFS correlated with hENT1 expression. The median OS was 13.3 months [95% confidence interval (95% CI), 8.5-16.7] and 3-year survival was 19.2% (95% CI, 11.5-26.9) for patients with low hENT1 expression, whereas the median OS was not yet reached during the follow-up period for the patients with high hENT1 expression [hazard ratio (HR) for death, 4.31; 95% CI, 1.95-9.52; P = 0.0001] and 3-year survival was 68.4% (95% CI, 57.7-79.1; P = 0.0007). The median DFS was 8.4 months (95% CI, 4.6-12.1) for patients with low expression and 46.8 months (95% CI, 30.3-93.9) for patients with high expression (HR for disease recurrence, 4.36; 95% CI, 2.01-9.44; P = 0.0001).

Fig. 2.

DFS (A) and OS (B) according to hENT1 expression (hENT1highn = 19 and hENT1lown = 26). DFS (C) and OS (D) according to hCNT3 expression (hCNT3highn = 22 and hCNT3lown = 23; n = 45).

Fig. 2.

DFS (A) and OS (B) according to hENT1 expression (hENT1highn = 19 and hENT1lown = 26). DFS (C) and OS (D) according to hCNT3 expression (hCNT3highn = 22 and hCNT3lown = 23; n = 45).

Close modal

In the same way, patients with low hCNT3 expression showed a higher risk of disease recurrence (median DFS, 8.6 months; 95% CI, 5.6-11.6) than patients with high hCNT3 expression (median DFS, 23.5 months; 95% CI, 12.4-76.3; HR for DFS, 2.27; 95% CI, 1.12-4.65; P = 0.02) and a significant difference was observed between groups for OS [median OS, not yet reached during the follow-up period versus 12.6 months (95% CI, 7.7-17.5), respectively] and 3-year survival [54.6% (95% CI, 43.9-65.1) versus 26.1% (95% CI, 15.0-35.2; P = 0.028)] according to hCNT3 expression (HR for death, 2.59; 95% CI, 1.24-5.43; P = 0.003). Among clinicopathologic factors, tumor size was the lone to be correlated with DFS and OS (Table 2). In a multivariate Cox proportional model that adjusted for the effect of lymph node metastasis, tumor size, Eastern Cooperative Oncology Group PS, age, CA19-9 level and lymph node ratio, and low hENT1 and hCNT3 expression were independent risk factors for death, whereas hENT1 was the only independent predictor for disease recurrence (Table 3).

Table 2.

Univariate analysis of OS and DFS

VariablesOS, HR (95% CI)PDFS, HR (95% CI)P
Age, y     
    <56 (n = 22)    
    ≥56 (n = 23) 1.23 (0.61-1.47) 0.31 1.19 (0.69-1.27) 0.18 
Sex     
    Female   
    Male 1.21 (0.76-1.43) 0.45 1.12 (0.84-1.52) 0.51 
Eastern Cooperative Oncology Group PS     
    0 (n = 40)   
    1 (n = 5) 1.09 (0.76-1.62) 0.63 1.15 (0.69-1.74) 0.72 
CA19-9 at diagnosis     
    <49 (n = 23)   
    ≥49 (n = 22) 1.82 (0.86-3.85) 0.14 1.57 (0.78-3.18) 0.23 
Tumor diameter (cm)     
    <2.5 (n = 24)   
    ≥2.5 (n = 21) 1.88 (0.94-3.73) 0.07 1.81 (0.87-3.70) 0.10 
Lymph node metastasis     
    No (n = 13)   
    Yes (n = 32) 1.80 (0.73-4.46) 0.19 1.59 (0.71-3.57) 0.26 
Lymph node ratio     
    <0.2 (n = 22)   
    ≥0.2 (n = 23) 2.01 (0.94-4.24) 0.06 1.81 (0.89-3.64) 0.11 
hENT1 expression     
    High (n = 19)   
    Low (n = 26) 3.88 (1.78-8.92) 0.0007 3.55 (1.65-7.63) 0.005 
hCNT3 expression     
    High (n = 22)   
    Low (n = 23) 3.08 (1.42-6.67) 0.0028 2.27 (1.12-4.65) 0.02 
VariablesOS, HR (95% CI)PDFS, HR (95% CI)P
Age, y     
    <56 (n = 22)    
    ≥56 (n = 23) 1.23 (0.61-1.47) 0.31 1.19 (0.69-1.27) 0.18 
Sex     
    Female   
    Male 1.21 (0.76-1.43) 0.45 1.12 (0.84-1.52) 0.51 
Eastern Cooperative Oncology Group PS     
    0 (n = 40)   
    1 (n = 5) 1.09 (0.76-1.62) 0.63 1.15 (0.69-1.74) 0.72 
CA19-9 at diagnosis     
    <49 (n = 23)   
    ≥49 (n = 22) 1.82 (0.86-3.85) 0.14 1.57 (0.78-3.18) 0.23 
Tumor diameter (cm)     
    <2.5 (n = 24)   
    ≥2.5 (n = 21) 1.88 (0.94-3.73) 0.07 1.81 (0.87-3.70) 0.10 
Lymph node metastasis     
    No (n = 13)   
    Yes (n = 32) 1.80 (0.73-4.46) 0.19 1.59 (0.71-3.57) 0.26 
Lymph node ratio     
    <0.2 (n = 22)   
    ≥0.2 (n = 23) 2.01 (0.94-4.24) 0.06 1.81 (0.89-3.64) 0.11 
hENT1 expression     
    High (n = 19)   
    Low (n = 26) 3.88 (1.78-8.92) 0.0007 3.55 (1.65-7.63) 0.005 
hCNT3 expression     
    High (n = 22)   
    Low (n = 23) 3.08 (1.42-6.67) 0.0028 2.27 (1.12-4.65) 0.02 
Table 3.

Multivariate analysis of pancreatic cancer recurrence and mortality risk

VariableOS, HR (95% CI)PDFS, HR (95% CI)P
hENT1 expression     
    High (n = 19)   
    Low (n = 26) 3.42 (1.44-8.81) 0.005 3.17 (1.43-6.73) 0.004 
hCNT3 expression     
    High (n = 22)   
    Low (n = 23) 2.65 (1.19-5.87) 0.017 2.09 (0.99-4.42) .052 
Lymph node ratio     
    <0.2 (n = 22)   
    ≤0.2 (n = 23) 1.49 (0.67-3.31) 0.33 1.42 (0.68-2.94) 0.35 
Tumor size (cm)     
    <2.5 (n = 23)   
    ≥2.5 (n = 22) 1.48 (0.69-3.19) 0.31 1.64 (0.79-3.28) 0.18 
Lymph node metastasis     
    No (n = 13)   
    Yes (n = 32) 1.31 (0.45-3.80) 0.62 1.24 (0.48-3.15) 0.66 
CA19-9 (IU/mL)     
    <49 (n = 23)   
    ≥49 (n = 22) 1.89 (0.88-4.06) 0.10 1.86 (0.89-3.88) 0.09 
Eastern Cooperative Oncology Group PS     
    0 (n = 40)   
    1 (n = 1) 1.31 (0.28-6.21) 0.73 1.57 (0.48-5.18) 0.45 
Age, y     
    <56 (n = 22)   
    ≥56 (n = 23) 1.75 (0.84-2.21) 0.43 1.81 (0.55-2.38) 0.39 
VariableOS, HR (95% CI)PDFS, HR (95% CI)P
hENT1 expression     
    High (n = 19)   
    Low (n = 26) 3.42 (1.44-8.81) 0.005 3.17 (1.43-6.73) 0.004 
hCNT3 expression     
    High (n = 22)   
    Low (n = 23) 2.65 (1.19-5.87) 0.017 2.09 (0.99-4.42) .052 
Lymph node ratio     
    <0.2 (n = 22)   
    ≤0.2 (n = 23) 1.49 (0.67-3.31) 0.33 1.42 (0.68-2.94) 0.35 
Tumor size (cm)     
    <2.5 (n = 23)   
    ≥2.5 (n = 22) 1.48 (0.69-3.19) 0.31 1.64 (0.79-3.28) 0.18 
Lymph node metastasis     
    No (n = 13)   
    Yes (n = 32) 1.31 (0.45-3.80) 0.62 1.24 (0.48-3.15) 0.66 
CA19-9 (IU/mL)     
    <49 (n = 23)   
    ≥49 (n = 22) 1.89 (0.88-4.06) 0.10 1.86 (0.89-3.88) 0.09 
Eastern Cooperative Oncology Group PS     
    0 (n = 40)   
    1 (n = 1) 1.31 (0.28-6.21) 0.73 1.57 (0.48-5.18) 0.45 
Age, y     
    <56 (n = 22)   
    ≥56 (n = 23) 1.75 (0.84-2.21) 0.43 1.81 (0.55-2.38) 0.39 

When we combined status of both hENT1 and hCNT3, for patients with two favorable prognostic factors (n = 15, both high hENT1 and hCNT3 expression), median OS was not reached compared with a median OS of 18.7 months (95% CI, 6.5-30.9) for patients with one favorable prognostic factor (n = 19, high hENT1 or hCNT3 expression; P = 0.0016) and 12.2 months (95% CI, 8.3-16.1) for patients with no favorable prognostic factors (n = 11; P < 0.0001; Fig. 3). The 3-year survival was 81.1% (95% CI, 71.4-90.8), 24.9% (95% CI, 14.1-36.7), and 23.1% (95% CI, 11.5-34.7; P < 0.0001), respectively, for the three groups.

Fig. 3.

OS according to the number of favorable prognostic factors. Two favorable prognostic factors (n = 15): median survival not reached; one favorable prognostic (n = 19): median survival, 18.7 mo (95% CI, 6.5-30.9); zero prognostic factor (n = 11): median survival, 12.2 months (95% CI, 8.3-16.1).

Fig. 3.

OS according to the number of favorable prognostic factors. Two favorable prognostic factors (n = 15): median survival not reached; one favorable prognostic (n = 19): median survival, 18.7 mo (95% CI, 6.5-30.9); zero prognostic factor (n = 11): median survival, 12.2 months (95% CI, 8.3-16.1).

Close modal

Gemcitabine chemotherapy is the current standard of care for palliation of advanced pancreatic cancer (3). In adjuvant setting, however, no single approach has achieved widespread acceptance. The activity of gemcitabine in advanced pancreatic cancer, and its radiosensitizing properties, led us to evaluate gemcitabine in combination with chemoradiotherapy to improve patients' outcomes after pancreatic adenocarcinoma resection. Chemoradiotherapy remains common in North American centers based on data from the Gastrointestinal Tumor Study Group and the RTOG 9074 studies (9, 32), whereas, in Europe, chemotherapy is the standard based on the results of the European Study Group for Pancreatic Cancer Trial 1 (4, 5) and CONKO-001 study (7, 8).

We previously conducted two phase II multicentric studies evaluating (15, 16) adjuvant chemoradiotherapy with gemcitabine for completely resected pancreatic adenocarcinoma. Both regimens shared similar designs, were highly feasible, and produced limited toxicity. We observed potential benefit for a subset of these patients, as 30% were long-term survivors, and the median OS was 21.9 months despite a high percentage of nodal metastases (71%) and T3 or T4 disease (73%). The strategy of using gemcitabine monotherapy before beginning concurrent gemcitabine and radiotherapy spared those patients with early progression or difficulty in tolerating treatment toxicities of concurrent chemoradiotherapy.

Identification of a subset of patients who derive benefit from adjuvant therapy would further improve the risk: benefit ratio of adjuvant pancreas cancer treatment. Because gemcitabine remains the backbone of systemic therapy for pancreatic cancer, we evaluated two molecular biomarkers of potential relevance to gemcitabine sensitivity. In the present study, we analyzed tumor samples from 45 participants in our adjuvant pancreatic cancer chemoradiation studies to assess the prognostic value of the gemcitabine transporter proteins, hENT1 and hCNT3. Using immunohistochemistry, we found that patients with low tumor hENT1 and hCNT3 had a shorter DFS and OS times when compared with patients with high hENT1 or hCNT3 expression. In our multivariate analysis, expression of hENT1 was the only significant independent marker of DFS, whereas hENT1 and hCNT3 were the only independent determinants of OS. hENT1 and hCNT3 status provided better prognostic information than tumor-node-metastasis classification, tumor diameter, lymph node status, ratio of positive nodes, and serum CA19.9 levels; furthermore, those patients with tumors exhibiting both high hENT1 and hCNT3 expression had particularly favorable outcome, with a mean survival exceeding 90 months. Although hENT1 and hCNT3 abundance within pancreatic tumor cells are prognostic markers for survival in our study, prospective molecular correlative studies should be done in trials where patients are randomized to either gemcitabine or a nonnucleoside therapy to prove whether hENT1 and hCNT3 are really predictive for benefit from adjuvant gemcitabine based-therapy and not simply a marker of prognosis.

These results are concordant with previous reports of the prognostic value of hENT1 in gemcitabine-treated pancreas cancer (26, 33, 34). The absence of detectable hENT1 in metastatic pancreatic adenocarcinoma treated with gemcitabine was associated with a shorter OS compared with patients for whom hENT1 was present (26). Giovannetti et al. showed in a retrospective cohort prolonged OS, DFS, and time to progression for patients with high levels of hENT1 mRNA (34).

In aggregate, the consistency of the preclinical data (2125) with these clinical data strongly suggest that pancreatic adenocarcinoma cells are resistant to gemcitabine if they lack the transporter proteins required for efficient gemcitabine cellular entry. Although the majority of data support the central role of hENT1 in clinical gemcitabine sensitivity, the present study also suggests that hCNT3 may also meaningfully contribute to gemcitabine permeation in pancreatic adenocarcinoma. Furthermore, the observation that those patients with high expression of both hENT1 and hCNT3 had significantly better outcomes than those patients with high expression of only one transporter suggests that multiple transporters may incrementally increase cellular sensitivity; this has been reported in preclinical models, where introduction of hCNT1 by stable transfection increased the gemcitabine sensitivity of a pancreatic cell line producing high levels of hENT1 (24). The emerging understanding of the cellular regulation of hENT1 and hCNT3 activities (3537) also raises the possibility of circumventing transporter-mediated drug resistance by up-regulating nucleoside transporter activity.

Gemcitabine is a potent radiosensitizer of human pancreatic tumor cells (38). Mechanisms of gemcitabine radiosensitization included dATP depletion (39, 40), accumulation of cells in the S phase of the cell cycle (41), and increased DNA residual damage. It is unknown whether nucleoside transporters are related to the ability of gemcitabine to radiosensitize pancreatic cancer cells. Furthermore, clarification for the relative contributions of gemcitabine induction chemotherapy and the chemoradiation therapy is needed and will deserve future comparison.

We conclude that the expression of hENT1 and hCNT3 provides independent prognostic information in pancreatic carcinoma patients treated with adjuvant gemcitabine-based chemoradiotherapy. Whether these assays provide sufficient predictive information to guide treatment decisions requires prospective evaluation in randomized clinical trials. However, the consistency and strength of the accumulating preclinical and translational data suggest that nucleoside transporters play an important role in clinical outcomes after gemcitabine adjuvant chemotherapy for pancreatic cancer.

No potential conflicts of interest were disclosed.

Grant support: Fonds Erasme and Alberta Cancer Board.

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.

Note: R. Maréchal is a research fellow of the Fonds Erasme.

We thank Cheryl Santos for expert assistance.

1
Yeo CJ, Cameron JL, Lillemoe KD, et al. Pancreaticoduodenectomy for cancer of the head of the pancreas. 201 patients.
Ann Surg
1995
;
221
:
721
–31.
2
Klinkenbijl JH, Jeekel J, Samouhd T, et al. Adjuvant radiotherapy and 5-fluorouracil after curative resection for the cancer of the pancreas and periampullary region. Phase III trial of the EORTC Gastrointestinal Tract Cancer Cooperative Group.
Ann Surg
1999
;
230
:
776
–7.
3
Smeenk HG, van Eijck CH, Hop WC, et al. Long-term survival and metastatic pattern of pancreatic and periampullary cancer after adjuvant chemoradiation or observation: long-term results of EORTC trial 40891.
Ann Surg
2007
;
246
:
734
–40.
4
Neoptolemos JP, Dunn JA, Stocken DD, et al. Adjuvant chemoradiotherapy and chemotherapy in resectable pancreatic cancer: a randomised controlled trial.
Lancet
2001
;
358
:
1576
–85.
5
Neoptolemos JP, Stocken D, Freiss H, et al. A randomised trial of chemoradiotherapy and chemotherapy after resection of pancreatic cancer.
N Engl J Med
2004
;
350
:
1200
–10.
6
Burris HA, Moore MJ, Andersen J, et al. Improvements in survival and in clinical benefit with gemcitabine as first-line therapy for patients with advanced pancreas cancer. A randomized trial.
J Clin Oncol
1997
;
15
:
2403
–13.
7
Oettle H, Post S, Neuhaus P, et al. Adjuvant chemotherapy with gemcitabine vs observation in patients undergoing curative-intent resection of pancreatic cancer: a randomized controlled trial.
JAMA
2007
;
297
:
267
–77.
8
Neuhaus P, Riess H, Post S, et al. CONKO-001: final results of the randomized, prospective, multicenter phase III trial of adjuvant chemotherapy with gemcitabine versus observation in patients with resected pancreatic cancer (PC).
J Clin Oncol
2008
;
26
Suppl 15:
abstract 4504
.
9
Regine WF, Winter KA, Abrams RA, et al. Fluorouracil vs gemcitabine chemotherapy before and after fluorouracil-based chemoradiation following resection of pancreatic adenocarcinoma: a randomized controlled trial.
JAMA
2008
;
299
:
1019
–26.
10
Lawrence TS, Change EY, Hohn TM, et al. Radiosensitization of pancreatic cancer cells by 2′,2′-difluoro-2-deoxycytidine.
Int J Radiat Oncol Biol Phys
1996
;
34
:
867
–72.
11
Milas L, Fujii T, Hunter NR, et al. Enhancement of tumor radioresponse in vivo by gemcitabine.
Cancer Res
1999
;
59
:
107
–14.
12
Mason KA, Milas L, Hunter NR, et al. Maximizing therapeutic gain with gemcitabine and fractionated radiation.
Int J Radiat Oncol Biol Phys
1999
;
44
:
1125
–35.
13
Blackstock AW, Bernard SA, Richards F, et al. Phase I trial of twice-weekly gemcitabine and concurrent radiation in patients with advanced pancreatic cancer.
J Clin Oncol
1999
;
17
:
2208
–12.
14
Murphy JD, Adusumilli S, Kent A, et al. Full-dose gemcitabine and concurrent radiotherapy for unresectable pancreatic cancer.
Int J Radiat Oncol Biol Phys
2007
;
68
:
801
–8.
15
Van Laethem JL, Demols A, Gay F, et al. Postoperative adjuvant gemcitabine and concurrent radiation after curative resection of pancreatic head carcinoma: a phase II study.
Int J Radiat Oncol Biol Phys
2003
;
56
:
974
–80.
16
Demols A, Peeters M, Polus M, et al. Adjuvant gemcitabine and concurrent continuous radiation (45 Gy) for resected pancreatic head carcinoma: a multicenter Belgian phase II study.
Int J Radiat Oncol Biol Phys
2005
;
62
:
1351
–6.
17
Van Laethem JL, Van Cutsem E, Hammel P, et al. Adjuvant chemotherapy alone versus chemoradiation after curative resection for pancreatic cancer: feasibility results of a randomised EORTC/FFCD/GERCOR phase II/III study (40013/22012/0304).
J Clin Oncol
2008
;
26
Suppl 15:
abstract 4514
.
18
Huang P, Plunkett W. Induction of apoptosis by gemcitabine.
Semin Oncol
1995
;
15
:
2403
–13.
19
Ruiz van Haperen VW, Veerman G, Vermorken JB, et al. 2′,2′-Difluoro-deoxycytidine (gemcitabine) incorporation into RNA and DNA of tumour cell lines.
Biochem Pharmacol
1993
;
46
:
762
–6.
20
Plunkett W, Huang P, Searcy CE, et al. Gemcitabine: preclinical pharmacology and mechanisms of action.
Semin Oncol
1996
;
23
:
3
–15.
21
Mackey JR, Yao SY, Smith KM, et al. Gemcitabine transport in Xenopus oocytes expressing recombinant plasma membrane mammalian nucleoside transporters.
J Natl Cancer Inst
1999
;
91
:
1876
–81.
22
Ritzel MW, Ng AM, Yao SY, et al. Recent molecular advances in studies of the concentrative nucleoside transporter (CNT): identification and characterization of novel human and mouse proteins (hCNT3 and mCNT3) broadly selective for purine and pyrimidine nucleosides.
Mol Membr Biol
2001
;
18
:
65
–72.
23
Mackey JR, Mani RS, Selner M, et al. Functional nucleoside transporters are required for gemcitabine influx and manifestation of toxicity in cancer cell lines.
Cancer Res
1998
;
58
:
4349
–57.
24
Garcia-Manteiga J, Molina-Arcas M, Casado FJ, et al. Nucleoside transporter profiles in human pancreatic cancer cells: role of hCNT1 in 2′,2′-difluorodeoxycitidine-induced cytotoxicity.
Clin Cancer Res
2003
;
9
:
5000
–8.
25
Nakano Y, Tanno S, Koizumi K, et al. Gemcitabine chemoresistance and molecular markers associated with gemcitabine transport and metabolism in human pancreatic cancer cells.
Br J Cancer
2007
;
96
:
457
–63.
26
Spratlin J, Sangha R, Glubrecht D, et al. The absence of human equilibrative nucleoside transporter 1 is associated with reduced survival in patients with gemcitabine-treated pancreas adenocarcinoma.
Clin Cancer Res
2004
;
10
:
6956
–61.
27
Dabbagh L, Coupland RW, Cass CE, et al. Immunohistochemical variation of human nucleoside transporters in primary breast cancer.
Biochem Biophys Res Commun
2001
;
280
:
951
–9.
28
Mackey JR, Jennings LL, Clarke ML, et al. Immunohistochemical variation of human nucleoside transporters in primary breast cancers.
Clin Cancer Res
2002
;
8
:
110
–6.
29
Kaplan EL, Meier P. Nonparametric estimation from incomplete observations.
J Am Stat Assoc
1958
;
53
:
457
–81.
30
Peto R, Pike MC, Armitage P, et al. Design and analysis of randomised clinical trials requiring prolonged observation of each patient.
Br J Cancer
1997
;
35
:
1
–39.
31
Cox DR. Regression models and life tables.
J R Stat Soc
1972
;
34
:
187
–220.
32
Douglass H. Further evidence of effective adjuvant combined radiation and chemotherapy following curative resection of pancreatic cancer.
Cancer
1987
;
59
:
2006
–10.
33
Farrell JJ, Garcia M, Lai R, et al. Human ENT1 is predictive of response to gemcitabine treatment in patients with pancreatic cancer: results from the RTOG 9704 prospective randomized trial.
Pancreas
2007
;
35
:
401
–2.
34
Giovannetti E, Del Tacca M, Mey V, et al. Transcription analysis of human equilibrative nucleoside transporter-1 predicts survival in pancreas cancer patients treated with gemcitabine.
Cancer Res
2006
;
66
:
3928
–35.
35
Fernández Calotti P, Galmarini CM, Cañones C, et al. Modulation of the human equilibrative nucleoside transporter1 (hENT1) activity by IL-4 and PMA in B cells from chronic lymphocytic leukemia.
Biochem Pharmacol
2008
;
75
:
857
–65.
36
Abdulla P, Coe IR. Characterization and functional analysis of the promoter for the human equilibrative nucleoside transporter gene, hENT1.
Nucleosides Nucleotides Nucleic Acids
2007
;
26
:
99
–110.
37
Eltzschig HK, Abdulla P, Hoffman E, et al. HIF-1-dependent repression of equilibrative nucleoside transporter (ENT) in hypoxia.
J Exp Med
2005
;
202
:
1493
–505.
38
Lawrence TS, Chang EY, Hahn TM, et al. Radiosensitization of pancreatic cancer cells by 2′,2′-difluoro-deoxycytidine.
Int J Radiat Oncol Biol Phys
1996
;
34
:
867
–72.
39
Shewach DS, Hahn TM, Chang E, et al. Metabolism of 2′,2′-difluoro-2′-deoxycytidine and radiation sensitization of human colon carcinoma cells.
Cancer Res
1994
;
54
:
3218
–23.
40
Robinson BW, Shewach DS. Radiosensitization by gemcitabine in p53 wild-type and mutant MCF-7 breast carcinoma cell lines.
Clin Cancer Res
2001
;
7
:
2581
–89.
41
Pauwels B, Korst AEC, Pattyn GGO, et al. Cell cycle effect of gemcitabine and its role in the radiosensitizing mechanism in vitro.
Int J Radiat Oncol Biol Phys
2003
;
57
:
1075
–83.