Riccardo Giampieri, Elena Maccaroni,
Rossana Berardi and
Clinica di Oncologia Medica, AO-Ospedali Riuniti, Università Politecnica delle Marche, Ancona, Italy
Despite recent advances in diagnosis and treatment options during the past decades, pancreatic cancers remain one of the deadliest tumours in western countries, with an average overall five-year survival of less than 5%. To date, surgical resection still represents the only potentially curative therapeutic option, but unfortunately, due to the lack of early symptoms and reliable screening methods for early detection, more than half of pancreatic cancer patients have metastatic or locally advanced disease at presentation. In this setting palliative chemotherapy remains the main treatment choice.1
Cornerstone of care
After the demonstration in 1997 that Gemcitabine could lead to an improvement in clinical benefit and overall survival when compared to standard 5-Fluoruracil (5-FU) infusion, this drug became the cornerstone of care for advanced pancreatic cancer patients.2
Since then, several authors have tried to improve results obtained with single agent Gemcitabine by exploring the activity of novel chemotherapy on biologically targeted agents in combination with Gemcitabine, with mostly disappointing results. Because of the shortcomings of standard treatment modalities, it is not surprising that these kind of tumours are often treated differently in the various institutions.
Thus, it is on a unifying principle that the European Society for Medical Oncology (ESMO) guidelines are conceived. However, is also pretty disappointing that, whenever asked for an update to ESMO guidelines regarding treatment of pancreatic cancer, it usually consists of just some small step forward, especially compared to the other solid malignancies.
Indeed, hypotheses generated by phase II trials, usually enrolling a population of fit patients, are often not met when applied to phase III confirmatory trials, conducted on a more heterogenous population. One notable example of this is the lack of efficacy of monoclonal antibodies, such as Bevacizumab and Cetuximab in this setting, despite the promising results of preliminary phase II studies and the results conducted in other tumour types.
The addition of Bevacizumab to Gemcitabine in first-line treatment in advanced pancreatic cancer patients does not improve survival as it is shown in the CALGB 80303 trial, with a median overall survival for the experimental arm of just 5.8 months against 5.9 months in the control arm (p=0.95), despite the results of a previous phase II trial (8.8 months of overall survival).3,4
Another monoclonal antibody, Cetuximab, in the Southwest Oncology Group–Directed Intergroup Trial S0205, also fails to show an improvement in overall survival in patients with unresectable locally advanced or metastatic pancreatic cancer (6.3 months for the combination of Gemcitabine and Cetuximab versus 5.9 months for Gemcitabine alone, p=0.23).5 Because of these results, it is not surprising that these two antibodies are not viable treatment options for metastatic or locally advanced pancreatic cancer patients.
Indeed, when talking about the use of biological agents in treatment of metastatic pancreatic cancer, only one reached a statistically significant advantage in terms of improving survival, namely Erlotinib. This oral tyrosine kinase inhibitor has proven in a recent study by Moore6 to achieve superior overall survival when used in combination with Gemcitabine than the control arm of single agent Gemcitabine (6.24 months versus 5.91 months, p=0.038). Even if the advantage is rather small – a median improvement in survival of about two weeks – the drug was approved for this indication and is also considered as a possible option in fit patients as stated in the guidelines.
Value of chemo-radiation
Other than this, the recent ESMO 2010 guidelines show only a few new notable topics. For example, data from the recent RTOG 97-04 trial7 suggests that the addition of 5-FU-based chemo-radiation after adjuvant chemotherapy with Gemcitabine in R1 resected patients improves survival up to 21.2 months (compared with 14.4 months of the control arm, p=0.001).
Another observation made by the authors of this study is that CA19.9 marker serum levels after resection are linked to a better outcome. Different overall survival of patients enrolled in RTOG 97-04, stratified in two different groups (less than or over than 180U/ml, less than or over than 90U/ml) is observed, with a benefit for a lower marker level than 180U/ml, with a hazard ratio [HR] 3.53, p<0.0001).8 The potential value of chemo-radiation is also been suggested in stage IIA patients as neoadjuvant treatment.
Moreover, intraoperative radiotherapy (IORT) is another promising new approach showing interesting results. One example is the ISIORT-Europe experience.9 In the data collected, an improvement in overall survival is shown for patients receiving both external radiotherapy (ERT) before surgery and IORT (30 months’ median overall survival).
Regarding patients with locally advanced pancreatic cancer, the ESMO guidelines state that the role of induction treatment is an interesting option. In fact, in the relatively recent GERCOR10 systematic review, encompassing about 21 studies of both induction chemotherapy, radiation therapy or chemoradiation, a slight benefit towards an improved survival for patients receiving both induction chemotherapy followed by chemo- radiation when compared with standard chemotherapy can be seen.
In particular, retrospective analysis performed on patients enrolled in GERCOR phase II and III studies suggests a benefit for induction chemotherapy followed by chemo-radiation when compared with standard chemotherapy with either 5-fluorouracil (5-FU) or Gemcitabine, with a difference in terms of overall survival of 15 months versus 10.7 months (p=.0009).11
Regarding best treatment options in patients with metastatic pancreatic cancer, it seems that the role of Gemcitabine as the mainstay of treatment has been confirmed yet another time by the results of many other studies.
Disappointing combination therapies
In fact, several other different combination candidates have been tested such as Cisplatin, Oxaliplatin, Capecitabine, 5-FU, Docetaxel and others, often with disappointing or barely satisfying results. One example is the combination of Gemcitabine and Capecitabine. The GEMCAP study is one of the first trials to suggest some kind of sinergy between the two drugs, with a reported advantage in terms of overall survival of 7.4 versus six months between the Gemcitabine + Capecitabine arm against Gemcitabine alone.
It must be said that the preliminary results have not been published in extenso and there has not been any other study showing comparable results among many other series. In particular, in a study of the Swiss Group for Clinical Cancer Research and the Central European Co-operative Oncology Group it is shown how Gemcitabine + Capecitabine does not improve either overall survival at a statistically significant level, or quality of life.12 It should also be noted that, in this study, subgroup analysis has found a statistically significant advantage for patients who have started treatment in good clinical condition (Karnofsky performance score >90). The median survival gain for this selected subgroup of patients is of 2.7 months, with no difference observed regarding an improvement of quality of life or symptoms.
The authors concluded that because of the small size of the subgroups, caution would be advised when taking into account this combination in the clinical practice. Taking these data into account, the guidelines have not pointed to this regimen as one of the preferred options until more solid data can be provided.
Another famous combination that has failed from time to time to provide any unarguable proof of efficacy is the doublet of a platinum compound and Gemcitabine.
The recent GIP-1 study, evaluating whether the combination of Gemcitabine to Cisplatin could yield any survival advantage over the use of Gemcitabine as single agent treatment, also has resulted negative. Overall survival for patients receiving the combination or the single agent treatment appeared similar, with 8.3 months of survival for the Gemcitabine alone arm versus 7.2 months for the combination (p=0.38), thus casting many doubts about the real usefulness of this regimen in everyday clinical practice.13
Another platinum compound that is usually selected in everyday practice, without any real solid proof of efficacy, is Oxaliplatin. The study E6201 comparing treatment with standard single-agent Gemcitabine versus either treatment with combination Gemcitabine-Oxaliplatin or treatment with Gemcitabine fixed-dose rate infusion has shown an equivalence in terms of overall survival and clinical benefit between the different regimens. Indeed, standard 1,000mg/m2 as a 30-minute infusion has reached a median overall survival of 4.9 months, whereas Gemcitabine fixed-dose rate has reached 6.2 months and GEMOX combination has reached 5.7 months, with either differences seen not statistically significant (GEM vs GEM FDR, p=0.04 and GEM vs GEMOX, p=0.22 with P<0.25 for statistical significance).
Moreover, no benefit in progression-free survival nor symptom control has been seen, with a poor pain control rate registered in all treatment arms. As for toxicities, Gemcitabine monotherapy has been shown to be well tolerated, with only grade 4 neutropenia reported in 33% of patients receiving treatment. This percentage has become as high as 59% in patients receiving Gemcitabine as fixed-dose rate infusion along with an increased the risk of other haematological toxicities. Patients receiving GEMOX have not achieved an increase in haematological toxicities, a higher incidence of other forms of toxicities, such as severe neuropathies, has been seen.14
One interesting solution is the one proposed by the ACCORD4/PRODIGE11 trial, presented as interim analysis at the last American Society of Clinical Oncology (ASCO) 2010 meeting. In particular, the FOLFIRINOX regimen (Oxaliplatin 85mg/m2 d1 in combination with Irinotecan 180mg/m2 d1 and folinic acid 400mg/m2 d1 followed by 5FU 400mg/m2 bolus d1 and 2,400mg/m2 46h continuous infusion biweekly) was compared with standard Gemcitabine (1g/m2 weekly for seven weeks followed by one week of rest and then weekly for three out of four weeks).
The authors found a statistically significant improvement in every endpoint explored. Overall survival was, in particular, better in FOLFIRINOX arm than in Gemcitabine arm (10.5 months vs 6.9 months, HR=0.61; 95%CI=0.46-0.81; p<0.001). The combination treatment was significantly more toxic than Gemcitabine alone, with a median relative dose-intensities of 5FU, Irinotecan, Oxaliplatin and Gemcitabine of 0.81, 0.80, 0.78 and 1.01 respectively. Authors concluded that FOLFIRINOX may become one new standard of therapy in fit patients candidate to a first-line therapy. One of the new topics that has been discussed in the guideline is the role of 2nd line chemotherapy. The CONKO 003 trial sheds some light over the advantage of a second-line regimen of 5-FU alone or in combination with Oxaliplatin in patients previously treated with Gemcitabine alone as first-line chemotherapy. Patients have been randomised to receive a chemotherapy regimen with FF regimen (5-FU 2,000mg/mq 24h pvi + folinic acid 200mg/mq d1-8-15-22) or OFF (same as FF regimen plus Oxaliplatin 85mg/mq d8-22).
Indeed, preliminary results published in ASCO 2005 have shown a rather striking difference between the two arms of treatment, with the study initially conceived as comparison between second-line chemotherapy with OFF versus best supportive care alone. The combination arm has offered a rather improved progression-free survival over the best supportive care (21 weeks vs 10 weeks, p=0.0077). Because of the poor accrual of patients (46 out of the 165 planned), mainly due to scepticism towards the best supportive care treatment arm, the protocol has been amended to change the two treatment arms into OFF+BSC vs FF+BSC. The final results of the study have shown a difference between OFF vs FF with a median PFS of 13 weeks vs nine weeks (p=0.012) and a median OS of 28 versus 13 weeks of survival respectively.
Authors have concluded that OFF can be considered as a plausible second-line option: it simply remains to discuss with the patient if an improvement of just a few weeks is acceptable in front of the increased – even if manageable – toxicities that such treatments will inevitably cause.15
To conclude, the landscape of treatment options in pancreatic cancer remains rather bleak and this remains particularly true for metastatic patients. Improvements of therapy in these patients usually consists of the consolidation of the few therapeutic regimens at our disposal (either Gemcitabine or 5-FU based) rather than the discovery of novel therapeutic options.
It is thus imperative, as stated many times throughout the guidelines, that patients are offered the best treatment option, that usually consists of clinical trials, wherever and whenever they are available. It is also necessary that this kind of patient is managed by a multidisciplinary team and in centres with high volumes of expertise to avoid blunders in the diagnostic and therapeutic work-up.
- Jemal A et al. CA Cancer J Clin Oncol 2008;58:71–96
- Burris HA III et al. J Clin Oncol 1997;15:2403-2413
- Kindler HL et al. J Clin Oncol 2010;28:3617-3622
- Kindler HL et al. J Clin Oncol 2005;23:8033-8040
- Philip AP et al. J Clin Oncol 2010;28:3605-3610
- Moore MJ et al. J Clin Oncol 2007;25:1960-6.
- Herman JM et al. J Clin Oncol 2008;26:3503-3510
- Berger AC et al. J Clin Oncol 2008;26:5918-5922
- Valentini V et al. Radiother Oncol 2009;91:54-59
- Huguet F et al. J Clin Oncol 2009;27:2269-2277
- Huguet F et al. J Clin Oncol 2007;25:326-331
- Herrmann R et al. J Clin Oncol 2007;25:2212-7
- Colucci G et al. J Clin Oncol 2010;28:1645-51
- Poplin E et al. J Clin Oncol 2009;27:3778-85
- Riess H et al. J Clin Oncol 2007;25 (Suppl):201