Erlotinib plus gemcitabine compared to gemcitabine alone in patients with advanced pancreatic cancer.  A phase III trial of the National Cancer Institute of Cancada Clinical Trials Group (NCIC-CTG)

Reviewer: Ryan Smith , MD
Last Modified: May 15, 2005

Presenter: M.J. Moore
Presenter's Affiliation: NCIC
Type of Session: Plenary


  • Advanced and metastatic pancreatic cancer has an extremely poor prognosis, with median survival times between 4-6 months.
  • Multiple attempts at combination chemotherapy have shown no efficacy
  • Attempts at targeted agents have also been negative trials
  • Targeting the epidermal growth factor receptor (EGFR) is a logical choice, as overexpression is commonly seen
  • Erlotinib is a small molecule inhibitor of the EGFR, with preclinical studies showing growth inhibition of pancreatic tumor cell lines

Materials and Methods

  • Eligible patients included those with advanced or metastatic pancreatic cancer who had received no prior systemic therapy other than chemotherapy concurrent with radiation
  • Patients were randomized to receive either gemcitabine + erlotinib vs. gemcitabine + placebo
  • Dosages were gemcitabine 1000 mg/m2 IV weekly for 7/8 weeks followed by weekly for 3/4 weeks and erlotinib initially at a dose of 100 mg per day (521 patients) or 150 mg per day (48 patients)
  • Primary endpoint was overall survival, statistically powered to detect a 33% increase


  • 569 patients were randomized (285 erlotinib arm and 284 control arm)
  • 75% had metastatic disease and 80% had a performance status of 0-1
  • Median survival times (MST) were 5.9 months (control) vs. 6.4 months (erlotinib) with 1 year overall survival of 17% (control) vs. 24% (erlotinib) (p=.025)
  • Progression free survival was 3.6 months (control) vs. 3.8 months (erlotinib) (p=.003)
  • Grade 3 and 4 adverse effects were similar, but there were an excess of grade 1 and 2 toxicities in the erlotinib arm, most notably due to skin rash (72% vs. 29%)
  • Only difference in quality of life scores was with respect to diarrhea in favor of the control arm
  • Increased response and survival was not correlated with EGFR status
  • Overall survival did correlate with skin rash: Grade 2 (MST 10.5 months), grade 1 (MST 5.8 months), grade 0 (MST 5.3 months)

Author's Conclusions

  • This is the first trial to show a targeted agent with a survival benefit in pancreatic cancer
  • Overall survival and progression free survival were both improved
  • There is increased but tolerable toxicity associated with erlotinib
  • The presence of a skin rash due to therapy denoted a better prognosis

Clinical/Scientific Implications

This study is important because it shows, for the first time, an effect of targeted therapy in pancreatic cancer.  Its improvements, however, are underwhelming.  The overall survival increase was 15 days and the progression free survival was 6 days.  Given that it was powered to detect a 33% increase in survival, it is puzzling how it is so statistically significant.  It was mentioned in the presentation that the curves continued to separate, which is encouraging, but still not a great cause for optimism.  Given the costs and increased toxicity, it is uncertain at best as to whether this treatment should be accepted as a new standard of care.  It is interesting that patients with increased skin toxicity had the best responses and is further proof that more careful patient selection and better understanding of the biology of the disease is needed.