Abstract

Background

Sorafenib is an oral anticancer agent targeting Ras-dependent signaling and angiogenic pathways. A phase I trial demonstrated that the combination of gemcitabine and sorafenib was well tolerated and had activity in advanced pancreatic cancer (APC) patients. The BAYPAN study was a multicentric, placebo-controlled, double-blind, randomized phase III trial comparing gemcitabine/sorafenib and gemcitabine/placebo in the treatment of APC.

Patients and methods

The patient eligibility criteria were locally advanced or metastatic pancreatic adenocarcinoma, no prior therapy for advanced disease and a performance status of zero to two. The primary end point was progression-free survival (PFS). The patients received gemcitabine 1000 mg/m2 i.v., weekly seven times followed by 1 rest week, then weekly three times every 4 weeks plus sorafenib 200 mg or placebo, two tablets p.o., twice daily continuously.

Results

Between December 2006 and September 2009, 104 patients were enrolled on the study (52 pts in each arm) and 102 patients were treated. The median and the 6-month PFS were 5.7 months and 48% for gemcitabine/placebo and 3.8 months and 33% for gemcitabine/sorafenib (P = 0.902, stratified log-rank test), respectively. The median overall survivals were 9.2 and 8 months, respectively (P = 0.231, log-rank test). The overall response rates were similar (19 and 23%, respectively).

Conclusion

The addition of sorafenib to gemcitabine does not improve PFS in APC patients.

introduction

Pancreatic cancer is a devastating disease, with the highest mortality rate among solid tumors. On diagnosis, patients usually have an unresectable, locally advanced or metastatic stage and the overall 5-year survival rate is less than 5% [1]. Gemcitabine has been considered the standard therapy for advanced pancreatic cancer (APC) for the past 15 years. In a comparative study with 5-fluorouracil (5-FU), gemcitabine was demonstrated to increase median overall survival (OS, from 4.4 to 5.6 months) and clinical benefit response. However, gemcitabine showed a limited objective response rate (5.4%) and the 12-month survival rate was only 18% [2]. Since this landmark study, a large number of gemcitabine-based combinations have been compared with gemcitabine alone and the results obtained were disappointing. Thus, although meta-analyses have suggested a potential survival advantage [3, 4], randomized trials evaluating the addition of a second cytotoxic (such as a platinum compound or a fluoropyrimidine) to gemcitabine did not demonstrate a clear superiority to gemcitabine alone [5–11]. Similarly, a combination of gemcitabine with monoclonal antibody-based targeted therapies, such as cetuximab [12] or bevacizumab [13], failed to substantially improve survival. Only the association of gemcitabine with erlotinib, a small tyrosine kinase inhibitor, targeting epidermal growth factor receptor (EGFR), has been demonstrated to significantly prolong patient survival, leading to its approval by regulatory agencies [14]. However, the advantage for the experimental arm is considered by many not clinically relevant, justifying the need for identifying novel agents with more significant activity.

Sorafenib is a multitargeted kinase inhibitor of vascular endothelial growth factor (VEGF) receptors (VEGFR) 2 and 3, platelet-derived growth factor (PDGF) receptor and Ras-accelerated fibrosarcoma (RAF) kinase, with registration in advanced hepatocellular and renal cell carcinoma. The potential mechanisms of action include inhibition of angiogenesis by targeting the VEGF and PDGF pathways and perturbation of cellular proliferation, differentiation and survival by disrupting the Kirsten rat sarcoma (KRAS)-B rapidly accelerated fibrosarcoma (BRAF)-mitogen-activated protein/extracellular signal-regulated kinase kinase/extracellular signal-regulated kinase (MEK/ERK) pathway [15]. Owing to the high prevalence of KRAS mutation and therefore to the associated KRAS-pathway activation, the latter activity seems particularly relevant in pancreatic cancer. In pancreatic cancer cell lines [16], sorafenib showed antitumor activity and its combination with gemcitabine was found to be not antagonistic in preclinical xenograft models [17]. Finally, a phase I study revealed that the combination of gemcitabine and sorafenib is safe and achieves a relatively promising stabilization rate of 56.5% in an expanded cohort of 23 cases of APC [18]. Taken together, these data provided the rationale to evaluate this regimen in a randomized phase III trial.

patients and methods

patients

The following were the patient eligibility criteria for the study: locally advanced or metastatic pancreatic adenocarcinoma; age >18 years an Eastern Cooperative Oncology Group (ECOG) performance status (PS) of zero to two; a life expectancy of >12 weeks; adequate bone marrow (granulocytes > 1.5 × 109/l, platelets > 100 × 109/l), renal (creatinine < 1.5 mg/dl or creatinine clearance > 60 ml/min) and hepatic [total bilirubin < 1.5× upper limit of normal, aspartate transaminases/alanine transaminases <2× upper limit of normal or <5× upper limit of normal in case of liver metastases, alkaline phosphatases <2× upper limit of normal] functions and at least one measurable lesion according to Response Evaluation Criteria in Solid Tumors (RECIST) 1.0. Prior chemotherapy for metastatic disease was not permitted. Previous adjuvant or neoadjuvant chemotherapy or radiochemotherapy was allowed if it had been given more than 6 months before enrollment. The key exclusion criteria included brain metastases, intestinal obstruction, history of inflammatory bowel disease or extended small bowel resection, peripheral neuropathy ≥ grade 2 [(CTCAE) version 3.0 classification], major surgery or radiotherapy within 28 days before registration, any severe and uncontrolled disease, serious infections and any other unstable medical condition which could compromise participation in the study. Previous history of treatment with inhibitors of RAS, RAF, MEK, AKT, mamalian target of rapamycin, farnesyl transferase or angiogenesis (such as bevacizumab) and concomitant use of any CYP3A4 inductors were not allowed. This protocol was approved by an independent ethics committee and all the patients had to give written informed consent according to all required guidelines (EudraCT number: 2006-000240-11; NCT00541021). A summary of the protocol is presented as supplementary material (S1), available at Annals of Oncology online.

treatment

Gemcitabine was given intravenously at a dose of 1000 mg/m2 over 30 min on days 1, 8, 15, 22, 29, 36 and 42, with rest on day 49 (for cycle 1) and on days 1, 8, 15, 29, 36 and 42 with rest on days 22 and 49 (cycles 2 and 3) of 8-week cycles. Two tablets of placebo or sorafenib 200 mg were given twice daily (i.e. 800 mg/day of sorafenib) continuously. Gemcitabine/sorafenib or gemcitabine/placebo was administered for a maximum of three cycles, i.e. 24 weeks. In case of objective response or disease stabilization after 24 weeks, continuation of sorafenib or placebo alone until progression, intolerance or patient refusal was allowed at the investigator's discretion.

dose adjustments

Adverse effects were graded according to the National Cancer Institute Common Terminology Criteria for Adverse Events (NCI-CTCAE) version 3.0. Gemcitabine and sorafenib were administered and dose-adapted according to the observed toxic effects. In case of permanent discontinuation of gemcitabine or sorafenib/placebo due to intolerance, a single-agent treatment was allowed to be continued.

study evaluations

Pretreatment evaluations included a complete medical history, physical examination, weight, Karnofsky Performance status (KPS), pain evaluation by using visual analog scale, complete blood cell count (CBC) with differential, chemistry panel, prothrombin time/International normalized ratio, urinalysis, carbohydrate antigen 19-9, pregnancy test (in women of childbearing potential), electrocardiogram and computed tomography scan of the abdomen and the chest. Physical examination including weight, KPS and pain evaluation, CBC and serum chemistries was carried out weekly. Liver function tests were done every 2 weeks. Tumor assessments were done every 8 weeks. Patients were evaluated for response according to RECIST 1.0 every 8 weeks [19]. Compliance with sorafenib administration was monitored through the reports given by the subjects - and pill counts given by pharmacists.

statistical design and analysis

The BAYPAN study was an academic study sponsored by Institut Paoli-Calmettes, Marseille, France. Study design, data collection and analysis, and interpretation of the results were exclusively carried out by the participating teams. Bayer Company was only providing the experimental drug. The primary end point of this study was progression-free survival (PFS), as measured from randomization until time of disease progression (according to RECIST 1.0 criteria) or death from any cause. Patients were randomly assigned with equal probability to gemcitabine plus sorafenib or placebo, using a minimization procedure based on the following parameters: disease extent (locally advanced versus metastatic) and ECOG performance. The primary objective was to detect a constant PFS hazard reduction of 50% [hazard ratio (HR) = 0.5]. Assuming constant hazard rates, this effect corresponds to double the PFS median and to an increase in the 6-month PFS rate from 20% in the control arm (gemcitabine/placebo) [5, 7, 8, 11] to 45% in the experimental arm (gemcitabine/sorafenib). The power of the trial was 80% and the alpha-risk was 5%. Under these assumptions, 104 patients (52 in each arm) were needed. The primary efficacy analysis population was the intent-to-treat population, defined as all patients randomly assigned, irrespective of whether the assigned treatment was actually received. For efficacy analyses, patients were grouped according to the treatment assigned at randomization.

The secondary end points included adverse events (AEs), objective response rate, OS and clinical benefit rate. AEs were evaluated according to the CTCAE v3.0 scale. Only patients for whom the treatment had been initiated were analyzed for tolerance and AEs. Responses were classified as complete response or partial response as per the RECIST1.0 criteria. OS was measured from randomization until death from any cause. Clinical benefit assessment was adapted from [2] and included weight, KPS and pain.

Survival probabilities were estimated using the Kaplan–Meier method and were compared using a two-sided log-rank test stratified according to the prognostic variables: age (< or ≥70) and extent of the disease (locally advanced or metastatic) with a p-value threshold of 0.05. A univariate Cox proportional hazards regression was used to estimate the treatment arm effect (HR) and its confidence interval. Rates and proportions were compared using Fisher's exact test, when appropriate, or a chi-2 approximation. SAS 9.2 (SAS Institute, Cary, NC) and R software, version 2.13.1. (http://CRAN.R-project.org) were used for all the statistical analyses. All the analyses were based on the study database frozen on 24 April 2011.

results

patient characteristics

Between December 2006 and September 2009, 104 patients (Figure 1) were randomly assigned (52 to each arm) across nine participating French hospitals, 102 of which actually received treatment (52 in the gemcitabine/placebo and 50 in gemcitabine/sorafenib; two patients assigned to gemcitabine/sorafenib did not start treatment: one patient died before starting treatment and the other one was lost to follow-up). The patient characteristics are listed in Table 1. The treatment arms were well balanced in terms of age, sex, performance status and median baseline CA19-9 levels, and nearly 80% of the patients had metastatic disease in each group. Among them, there was no difference between the arms in terms of metastatic site.

Table 1.

Patient characteristics

 Gemcitabine/placebo (N = 52) Gemcitabine/sorafenib (N = 52) 
Age (years) 
 Median 64 61 
 Range 40–82 42–85 
Gender 
 Male 32 (62%) 30 (58%) 
 Female 20 (38%) 22 (42%) 
ECOG-WHO PS 
 0 18 (35%) 16 (34%) 
 1 30 (59%) 26 (55%) 
 2 3 (6%) 5 (11%) 
 NA 
Stage 
 Locally advanced 12 (23%) 9 (17%) 
 Metastatic 40 (77%) 43 (83%) 
 Metastatic sites 
  Liver 34 33 
  Lung 15 11 
  Lymph node 11 
  Peritoneum 
  Bone 
CA 19–9 (UI/l) 
 Median 424 471 
 Range (1.3–3300) (1.2–21 500) 
 Gemcitabine/placebo (N = 52) Gemcitabine/sorafenib (N = 52) 
Age (years) 
 Median 64 61 
 Range 40–82 42–85 
Gender 
 Male 32 (62%) 30 (58%) 
 Female 20 (38%) 22 (42%) 
ECOG-WHO PS 
 0 18 (35%) 16 (34%) 
 1 30 (59%) 26 (55%) 
 2 3 (6%) 5 (11%) 
 NA 
Stage 
 Locally advanced 12 (23%) 9 (17%) 
 Metastatic 40 (77%) 43 (83%) 
 Metastatic sites 
  Liver 34 33 
  Lung 15 11 
  Lymph node 11 
  Peritoneum 
  Bone 
CA 19–9 (UI/l) 
 Median 424 471 
 Range (1.3–3300) (1.2–21 500) 

Abbreviations: ECOG, Eastern Cooperative Oncology Group; WHO, World Health Organization; CA 19-9, carbohydrate antigen 19-9.

Figure 1.

CONSORT diagram.

Figure 1.

CONSORT diagram.

treatment administration

In the gemcitabine/placebo arm, a median number of three cycles (1–3) of gemcitabine and three cycles of placebo (1–8) were delivered, whereas the patients in the gemcitabine/sorafenib arm received a median number of two cycles of gemcitabine (0–3) and two cycles of sorafenib (0–8). Premature study discontinuation seemed more common (Figure 1) and very often related to progression or nontreatment-related death in gemcitabine/sorafenib patients. AEs leading to early treatment discontinuation were uncommon and their frequency seemed identical in both the arms. However, there was a nonsignificant trend toward a more frequent reduction of gemcitabine dosage in the gemcitabine/sorafenib arm (Table 2).

Table 2.

Drug exposure

 Gemcitabine/placebo (N = 52) Gemcitabine/sorafenib (N = 50aP Value 
Gemcitabine 
 Transient interruption 
  Yes 20 (38%) 18 (36%) = 0.95 
  No 32 (62%) 32 (64%) 
 Dosage reduction 
  Yes 9 (17%) 17 (33%) = 0.11 
  No 43 (83%) 35 (67%) 
Sorafenib/placebo 
 Transient interruption 
  Yes 16 (31%) 22 (44%) = 0.26 
  No 35 (67%) 28 (56%) 
  NA 1 (2%)  
 Dosage reduction 
  Yes 5 (10%) 3 (6%) = 0.50 
  No 46 (88%) 47 (94%) 
  NA 1 (2%)  
 Gemcitabine/placebo (N = 52) Gemcitabine/sorafenib (N = 50aP Value 
Gemcitabine 
 Transient interruption 
  Yes 20 (38%) 18 (36%) = 0.95 
  No 32 (62%) 32 (64%) 
 Dosage reduction 
  Yes 9 (17%) 17 (33%) = 0.11 
  No 43 (83%) 35 (67%) 
Sorafenib/placebo 
 Transient interruption 
  Yes 16 (31%) 22 (44%) = 0.26 
  No 35 (67%) 28 (56%) 
  NA 1 (2%)  
 Dosage reduction 
  Yes 5 (10%) 3 (6%) = 0.50 
  No 46 (88%) 47 (94%) 
  NA 1 (2%)  

aIn this group, two patients did not start treatment and were not evaluable for this analysis. NA, not available.

efficacy analysis

The median follow-up was 27.7 months. As for the primary end point, no statistically significant difference in PFS was observed between the study arms (P = 0.902, stratified log-rank test; Figure 2). The median PFS was 5.7 months (95% CI 3.7–7.5) in the gemcitabine/placebo arm versus 3.8 months (95% CI 3.1–6) in the gemcitabine/sorafenib arm. The 6-month PFS rate was 48% (95% CI 36% to 64%) in the gemcitabine/placebo arm and 33% (95% CI 22% to 49%) in the gemcitabine/sorafenib arm. The resulting HR was 1.04 (95% CI 0.697–1.545; P = 0.854, Cox proportional hazards regression) for sorafenib versus placebo.

Figure 2.

Progression-free survival (PFS) by treatment arm. *There were only 51 patients initially at risk in the gemcitabine/sorafenib arm, because one patient randomized in this arm did not undergo any tumor evaluation.

Figure 2.

Progression-free survival (PFS) by treatment arm. *There were only 51 patients initially at risk in the gemcitabine/sorafenib arm, because one patient randomized in this arm did not undergo any tumor evaluation.

There was no statistically significant difference in the median OS between the study arms. The Kaplan–Meier curves for OS by the treatment arm are shown in Figure 3. The median OS was 9.2 months (95% CI 7.7–11.6) for gemcitabine/placebo and 8 months (95% CI 6–10.8) for gemcitabine/sorafenib (P = 0.231, log-rank test).The resulting HR was 1.27 for sorafenib versus placebo (95% CI 0.837–1.932; P = 0.261, Cox proportional hazards regression).

Figure 3.

Overall survival (OS) by treatment arm.

Figure 3.

Overall survival (OS) by treatment arm.

Response rates were similar in both groups: 23% of gemcitabine/sorafenib and 19% of gemcitabine/placebo experienced an objective response. Stable disease occurred in 42% and 52% of patients in the sorafenib and placebo arms, respectively. There were no significant differences between the arms in terms of clinical benefit (pain, KPS, weight): 13 and 23% experienced a positive response in gemcitabine/sorafenib and gemcitabine/placebo, respectively (P = 0.29, chi-square test).

adverse events

Grade 3–5 AEs were observed in 44 patients (88%) in the gemcitabine/sorafenib arm and in 41 patients (79%) in the gemcitabine/placebo arm. Serious AE (SAE) occurred in 33 patients (66%) in the sorafenib arm and in 31 patients (60%) in the placebo arm. SAE led to transient or definitive treatment discontinuation in 17 cases in the sorafenib arm and in 14 patients in the placebo arm. In the gemcitabine/sorafenib arm, there were 12 ultimately fatal serious AEs, only one of them was possibly treatment related (cerebrovascular accident), whereas only six were observed (none being considered being treatment-related) in the gemcitabine/placebo arm.

The most frequent AEs in both the treatment arms are summarized in Table 3. Anemia was more common in the gemcitabine/placebo arm (15%), maybe reflecting longer treatment duration in the control group. Conversely and as expected, skin toxicity (mainly including hand-foot syndromes, skin rash, pruritis) alopecia and stomatitis were higher in the gemcitabine/sorafenib arm. There was no significant difference in infectious, cardiac, venous thrombosis, pulmonary embolism, cerebrovascular or hypertension events.

Table 3.

Adverse events

  Any grade
 
Grade 3–4
 
Gemcitabine/placebo (= 52) Gemcitabine/sorafenib (= 50) Gemcitabine/placebo (= 52) Gemcitabine/sorafenib (N = 50) 
Anemia (%) 50 20** 15 2* 
Neutropenia (%) 48 32 29 26 
Febrile neutropenia (%) 
Thrombocytopenia (%) 38 40 12 
Cardiac events (%) 15 
Venous thrombosis (%) 
Pulmonary embolism (%) 
Cererbrovascular event (%) 
Hypertension (%) 
Skin toxicity (%) 38 58 18** 
Alopecia (%) 32** 10** 
Stomatitis (%) 30** 
Diarrhea (%) 29 49 
  Any grade
 
Grade 3–4
 
Gemcitabine/placebo (= 52) Gemcitabine/sorafenib (= 50) Gemcitabine/placebo (= 52) Gemcitabine/sorafenib (N = 50) 
Anemia (%) 50 20** 15 2* 
Neutropenia (%) 48 32 29 26 
Febrile neutropenia (%) 
Thrombocytopenia (%) 38 40 12 
Cardiac events (%) 15 
Venous thrombosis (%) 
Pulmonary embolism (%) 
Cererbrovascular event (%) 
Hypertension (%) 
Skin toxicity (%) 38 58 18** 
Alopecia (%) 32** 10** 
Stomatitis (%) 30** 
Diarrhea (%) 29 49 

*P < 0.05; **< 0.01.

discussion

This double-blind randomized phase III trial failed to demonstrate any benefit for gemcitabine/sorafenib over gemcitabine/placebo. The combination was confirmed to be feasible without unexpected and significant alteration of the tolerance profile of gemcitabine alone, but did not provide a superior PFS, the primary end point of this study, and did not improve response rate, OS and clinical benefit. AEs were slightly more frequent, and there was a nonsignificant trend toward a more frequent chemotherapy dosage reduction in the gemcitabine/sorafenib arm. However, toxicity-related treatment interruptions were comparable between the arms. Thus, it is unlikely that tolerance issues could explain the lack of impact of sorafenib.

Disappointing results were also recently reported in a noncomparative phase II randomized study that enrolled 74 metastatic pancreatic cancer patients and evaluating gemcitabine/sorafenib versus sorafenib [20]. In this trial, only one objective response was observed in the combination arm, whereas no patient responded to sorafenib. The median PFS and the OS were very short in both the arms (2.9 and 6.5 months, respectively, in the gemcitabine/sorafenib group; 2.3 and 4.4 months, respectively, in the sorafenib group). Finally, a randomized phase II study, only recently reported as an abstract, compared gemcitabine-cisplatin doublets with and without sorafenib and failed to identify any improvement in sorafenib-treated patients [21].

The therapeutic combination tested in our study was based on a strong conceptual rationale: cancer with a dismal prognosis, weak therapeutic gold standard, high unmet need, high frequency of KRAS mutation, combination of the gold standard and a well-known drug targeting angiogenesis and KRAS-driven BRAF/MEK/ERK pathway. However, preclinical data supporting the activity of gemcitabine associated with sorafenib in pancreatic cancer were relatively limited and preliminary [16, 17]. In addition, only disease stabilizations but no objective response were registered in APC patients during the phase I study by Siu et al. Nevertheless, preclinical models are known to be far from perfectly predicting clinical activity, especially in pancreatic cancer, and small-sized, noncomparative, single-arm phase II studies have been commonly misleading in pancreatic cancer, supporting the need for a comparative trial.

A questionable aspect of this study was the nature of the primary end point, since PFS has not been validated as a surrogate for OS in pancreatic cancer. However, it would have been very unlikely to observe any relevant OS impact without any detectable improvement in PFS. In addition, OS results were also reported and were clearly negative. With only 104 patients planned to be enrolled, this study was designed to only capture huge differences in PFS and had only limited power. Thus, the BAYPAN study might be regarded as a comparative randomized phase II, rather than an authentic phase III, study. However, we wanted to detect only clinically meaningful survival difference, and building a larger sample-size trial to identify a more subtle effect would have been questionable. Of note, there was no trend in favor of the experimental arm in any end point to be analyzed. Rather, there was a nonsignificant numerical advantage favoring the control arm, making it unlikely that a more powered design could have identified a clinically significant benefit for sorafenib.

A potential bias of this study, also shared by many studies initiated at this time, may relate to the relative heterogeneity of the patient population, since both locally advanced and metastatic patients were eligible, as well as patients with variable performance status (from 0 to 2). It was recently reported at a meeting of experts on planning clinical trials in pancreatic cancer that these very different prognosis-groups should not be included in the same clinical trial [22]. However, the randomization was carried out by a minimization method, making these characteristics well-balanced between both the arms and thus, unlikely to significantly affect the results.

Our results are also in line with recent investigations evaluating other targeted therapies in combination with gemcitabine. Thus, Kindler et al. recently reported a phase III randomized study comparing gemcitabine/bevacizumab and gemcitabine/placebo: response rates were similar and no significant survival advantage was observed for the experimental arm [13]. Bevacizumab was also tested in combination with gemcitabine/erlotinib in a randomized, double-blind, placebo-controlled, phase III trial and failed to significantly improve survival, in spite of marginally increasing PFS [23]. Axitinib, an oral selective inhibitor of VEGFR1, 2 and 3, was also associated with gemcitabine and compared with gemcitabine/placebo in the phase III study [24]. Once again, no significant advantage was observed for the axitinib/gemcitabine arm. These results, taken together with those reported in the present study, further confirm that currently available VEGF/VEGFR-targeted antiangiogenic strategies have no impact in APC, at least in an unselected patient population. This may be related to the relative poor density of intratumor vasculature, contrasting with the highly fibrotic stroma component, and making this tumor type independent from conventional angiogenesis. Another potential interest of sorafenib was related to its ability to inhibit BRAF and therefore to disrupt the KRAS/BRAF/MEK/ERK proliferative pathway, which is supposed to be commonly hyperactivated, as a consequence of the highly frequent KRAS mutation in this tumor type. The failure of sorafenib to improve outcome in the present study may suggest that sorafenib is not a major BRAF inhibitor, as already suggested in other settings [25]. Indeed, sorafenib was shown to have limited antitumor activity in BRAF-mutated melanoma [26], whereas this disease was recently demonstrated to be exquisitely sensitive to vemurafenib, a more recent and potent BRAF-inhibitor [27]. However, the lack of detectable clinical activity in this study does not necessarily demonstrate that sorafenib is a poor inhibitor of BRAF since other pathways may be upregulated which may circumvent BRAF inhibition, particularly in predominantly KRAS-mutated tumors, such as pancreatic cancer. Consistent with this hypothesis, it was recently reported that ATP-competitive RAF inhibition, through transactivation of the noninhibited member of CRAF-CRAF homodimers or CRAF-BRAF heterodimers, leads to increased signaling through the RAF-MEK-ERK pathway in BRAF wild-type tumors, resulting in paradoxical increased tumor growth [28, 29]. Consistent with this hypothesis, data from very recent studies evaluating vemurafenib in BRAF-mutated melanoma have shown an increase in the development of RAS-mutated cutaneous squamous-cell carcinomas [30, 31]. This may be of particular relevance, in view of the trend in favor of the placebo arm that was noted in the present study. Finally, other signaling pathways such as PI3K/AKT, frequently activated in pancreatic cancer [32, 33], may also counteract BRAF inhibition by sorafenib.

Was gemcitabine alone the best cytotoxic backbone to associate with targeted therapies including sorafenib in APC? When the BAYPAN trial was designed, it was clearly the standard regimen to test in this setting. However, combining a promising targeted therapy with a marginally active cytotoxic drug could lead to marginal improvement only. Conroy et al. recently reported that FOLFIRINOX polychemotherapy (5-FU, oxaliplatin and irinotecan) drastically improves response and survival in good PS metastatic pancreatic cancer, compared with gemcitabine alone [34]. Whether future targeted therapies will be able to be combined with this more aggressive and potentially toxic regimen remains to be investigated. In conclusion, this randomized phase III trial failed to demonstrate a superior PFS with the combination of gemcitabine/sorafenib over gemcitabine/placebo. Currently, sorafenib has no place in the management of APC.

disclosure

The authors have declared no conflicts of interest.

acknowledgements

We thank study coordinator Dr Agnès Boyer-Chamard for her contribution. Presented, in part, at the 47th Annual Meeting of the American Society of Clinical Oncology (ASCO), June 2011, Chicago, IL, USA.

references

1
Hidalgo
M
Pancreatic cancer
N Engl J Med
 , 
2010
, vol. 
362
 (pg. 
1605
-
1617
)
2
Burris
HA
Moore
MJ
Andersen
J
, et al.  . 
Improvements in survival and clinical benefit with gemcitabine as first- line therapy for patients with advanced pancreas cancer: a randomized trial
J Clin Oncol
 , 
1997
, vol. 
15
 (pg. 
2403
-
2413
)
3
Heinemann
V
Boeck
S
Hinke
A
, et al.  . 
Meta-analysis of randomized trials: evaluation of benefit from gemcitabine-based combination chemotherapy applied in advanced pancreatic cancer
BMC Cancer
 , 
2008
, vol. 
8
 pg. 
82
 
4
Sultana
A
Smith
CT
Cunningham
D
, et al.  . 
Meta-analyses of chemotherapy for locally advanced and metastatic pancreatic cancer
J Clin Oncol
 , 
2007
, vol. 
25
 (pg. 
2607
-
2615
)
5
Herrmann
R
Bodoky
G
Ruhstaller
T
, et al.  . 
Gemcitabine plus capecitabine compared with gemcitabine alone in advanced pancreatic cancer: a randomized, multicenter, phase III trial of the Swiss Group for Clinical Cancer Research and the Central European Cooperative Oncology Group
J Clin Oncol
 , 
2007
, vol. 
25
 (pg. 
2212
-
2217
)
6
Stathopoulos
GP
Syrigos
K
Aravantinos
G
, et al.  . 
A multicentre phase III trial comparing irinotecan-gemcitabine (IG) with gemcitabine (G) monotherapy as first-line treatment in patients with locally advanced or metastatic pancreatic cancer
Br J Cancer
 , 
2006
, vol. 
95
 (pg. 
587
-
592
)
7
Heinemann
V
Quietzsch
D
Gieseler
F
, et al.  . 
Randomized phase III trial of gemcitabine plus cisplatin compared with gemcitabine alone in advanced pancreatic cancer
J Clin Oncol
 , 
2006
, vol. 
24
 (pg. 
3946
-
3952
)
8
Louvet
C
Labianca
R
Hammel
P
, et al.  . 
Gemcitabine in combination with oxaliplatin compared with gemcitabine alone in locally advanced or metastatic pancreatic cancer: results of a GERCOR and GISCAD phase III trial
J Clin Oncol
 , 
2005
, vol. 
23
 (pg. 
3509
-
3516
)
9
Di Costanzo
F
Carlini
P
Doni
L
, et al.  . 
Gemcitabine with or without continuous infusion 5-FU in advanced pancreatic cancer: a randomised phase II trial of the Italian oncology group for clinical research (GOIRC)
Br J Cancer
 , 
2005
, vol. 
93
 (pg. 
185
-
189
)
10
Rocha Lima
CM
Green
MR
Rotche
R
, et al.  . 
Irinotecan plus gemcitabine results in no survival advantage compared with gemcitabine monotherapy in patients with locally advanced or metastatic pancreatic cancer despite increased tumor response rate
J Clin Oncol
 , 
2004
, vol. 
22
 (pg. 
3776
-
3883
)
11
Cunningham
D
Chau
I
Stocken
DD
, et al.  . 
Phase III randomized comparison of gemcitabine versus gemcitabine plus capecitabine in patients with advanced pancreatic cancer
J Clin Oncol
 , 
2009
, vol. 
27
 (pg. 
5513
-
5518
)
12
Philip
PA
Benedetti
J
Corless
CL
, et al.  . 
Phase III study comparing gemcitabine plus cetuximab versus gemcitabine in patients with advanced pancreatic adenocarcinoma: Southwest Oncology Group-Directed Intergroup Trial S0205
J Clin Oncol
 , 
2010
, vol. 
28
 (pg. 
3605
-
3610
)
13
Kindler
HL
Niedzwiecki
D
Hollis
D
, et al.  . 
Gemcitabine plus bevacizumab compared with gemcitabine plus placebo in patients with advanced pancreatic cancer: phase III trial of the Cancer and Leukemia Group B (CALGB 80303)
J Clin Oncol
 , 
2010
, vol. 
28
 (pg. 
3617
-
3622
)
14
Moore
MJ
Goldstein
D
Hamm
J
, et al.  . 
Erlotinib plus gemcitabine compared with gemcitabine alone in patients with advanced pancreatic cancer. A phase III trial of the National Cancer Institute of Canada Clinical Trials Group
J Clin Oncol
 , 
2007
, vol. 
25
 (pg. 
1960
-
1966
)
15
Wilhelm
SM
Carter
C
Tang
L
, et al.  . 
BAY 43–9006 exhibits broad spectrum oral antitumor activity and targets the RAF/MEK/ERK pathway and receptor tyrosine kinases involved in tumor progression and angiogenesis
Cancer Res
 , 
2004
, vol. 
64
 (pg. 
7099
-
7109
)
16
Ulivi
P
Arienti
C
Amadori
D
, et al.  . 
Role of RAF/MEK/ERK pathway, p-STAT-3 and Mcl-1 in sorafenib activity in human pancreatic cancer cell lines
J Cell Physiol
 , 
2009
, vol. 
220
 (pg. 
214
-
221
)
17
Vincent
P
Zhang
X
Chen
C
, et al.  . 
Chemotherapy with the raf kinase inhibitor BAY 43–9006 in combination with irinotecan, vinorelbine, or gemcitabine is well tolerated and efficacious in preclinical xenograft models
Proc Am Soc Clin Oncol
 , 
2002
, vol. 
21
 pg. 
2002
 
18
Siu
LL
Awada
A
Takimoto
CH
, et al.  . 
Phase I trial of sorafenib and gemcitabine in advanced solid tumors with an expanded cohort in advanced pancreatic cancer
Clin Cancer Res
 , 
2006
, vol. 
12
 (pg. 
144
-
151
)
19
Therasse
P
Arbuck
SG
Eisenhauer
EA
, et al.  . 
New guidelines to evaluate the response to treatment in solid tumors
J Natl Cancer Inst
 , 
2000
, vol. 
92
 (pg. 
205
-
216
)
20
El-Khoueiry
AB
Ramanathan
RK
Yang
DY
, et al.  . 
A randomized phase II of gemcitabine and sorafenib versus sorafenib alone in patients with metastatic pancreatic cancer
Invest New Drugs
 , 
2011
21
Berardi
R
Sobrero
A
Labianca
R
, et al.  . 
Sorafenib does not improve the results of chemotherapy in advanced pancreatic cancer: a GISCAD randomised phase II trial
Ann Oncol
 , 
2010
, vol. 
21
 (pg. 
viii225
-
viii249
)
22
Philip
PA
Mooney
M
Jaffe
D
, et al.  . 
Consensus report of the national cancer institute clinical trials planning meeting on pancreas cancer treatment
J Clin Oncol
 , 
2009
, vol. 
27
 (pg. 
5660
-
5669
)
23
Van Cutsem
E
Vervenne
WL
Bennouna
J
, et al.  . 
Phase III trial of bevacizumab in combination with gemcitabine and erlotinib in patients with metastatic pancreatic cancer
J Clin Oncol
 , 
2009
, vol. 
27
 (pg. 
2231
-
2237
)
24
Kindler
HL
Ioka
T
Richel
DJ
, et al.  . 
Axitinib plus gemcitabine versus placebo plus gemcitabine in patients with advanced pancreatic adenocarcinoma: a double-blind randomised phase 3 study
Lancet Oncol
 , 
2011
, vol. 
12
 (pg. 
256
-
262
)
25
Ott
PA
Hamilton
A
Min
C
, et al.  . 
A phase II trial of sorafenib in metastatic melanoma with tissue correlates
PLoS One
 , 
2010
, vol. 
5
 pg. 
e15588
 
26
Ott
PA
Hamilton
A
Min
C
, et al.  . 
A phase II trial of sorafenib in metastatic melanoma with tissue correlates
PLoS One
 , vol. 
5
 pg. 
e15588
 
27
Chapman
PB
Hauschild
A
Robert
C
, et al.  . 
Improved survival with vemurafenib in melanoma with BRAF V600E mutation
N Engl J Med
 , vol. 
364
 (pg. 
2507
-
2516
)
28
Hatzivassiliou
G
Song
K
Yen
I
, et al.  . 
RAF inhibitors prime wild-type RAF to activate the MAPK pathway and enhance growth
Nature
 , 
2010
, vol. 
464
 (pg. 
431
-
435
)
29
Poulikakos
PI
Zhang
C
Bollag
G
, et al.  . 
RAF inhibitors transactivate RAF dimers and ERK signalling in cells with wild-type BRAF
Nature
 , vol. 
464
 (pg. 
427
-
430
)
30
Su
F
Viros
A
Milagre
C
, et al.  . 
RAS mutations in cutaneous squamous-cell carcinomas in patients treated with BRAF Inhibitors
N Engl J Med
 , 
2012
, vol. 
366
 (pg. 
207
-
215
)
31
Weeraratna
AT
RAF around the Edges “The Paradox of BRAF Inhibitors”
N Engl J Med
 , 
2012
, vol. 
366
 (pg. 
271
-
273
)
32
Yamamoto
S
Tomita
Y
Hoshida
Y
, et al.  . 
Prognostic significance of activated Akt expression in pancreatic ductal adenocarcinoma
Clin Cancer Res
 , 
2004
, vol. 
10
 (pg. 
2846
-
2850
)
33
Edling
CE
Selvaggi
F
Buus
R
, et al.  . 
Key role of phosphoinositide 3-kinase class IB in pancreatic cancer
Clin Cancer Res
 , vol. 
16
 (pg. 
4928
-
4937
)
34
Conroy
T
Desseigne
F
Ychou
M
, et al.  . 
FOLFIRINOX versus gemcitabine for metastatic pancreatic cancer
N Engl J Med
 , 
2011
, vol. 
364
 (pg. 
1817
-
1825
)