Gemcitabine and Erlotinib in Pancreatic Cancer

Reviewer: Neha Vapiwala, MD
Abramson Cancer Center of the University of Pennsylvania
Last Modified: November 2, 2005

Presenter: Malcolm Moore, MD
Presenter's Affiliation: Princess Margaret Hospital
Type of Session: Scientific


  • Pancreatic cancer is the fourth leading cause of cancer-related deaths.
  • In 2005, there will be an estimated 35,400 new cases and 35,000 deaths in the United States and Canada.
  • The 5-year survival is dismal, hovering at a rate of less than 4%.
  • This disease often presents at a very advanced stage, with significant extension of disease either locally or distantly.
  • Phase III trials have looked at the use of combination chemotherapy regimens in the treatment of this disease, but none have been positive for a survival benefit.
  • Phase III trials of up-and-coming targeted agents have also been disappointing to date.
  • Gemcitabine intravenous (IV) chemotherapy has shown improved single agent activity in pancreatic cancer compared to standard 5-fluorouracil (5-FU), with clinical benefit response rates of 24% vs. 5%, respectively.
  • The epidermal growth factor receptor (EGFR) is overexpressed on the surface of the vast majority of pancreatic adenocarcinomas.
  • Preclinical studies of small molecule inhibitors of EGFR have found single agent activity, as well as activity in conjunction with gemcitabine, in pancreatic cancer xenografts.
  • The authors of this study sought to investigate the use of gemcitabine (G) and erlotinib (E; brand name Tarceva®) as combination treatment in patients (pts) with locally advanced (LA) and metastatic pancreatic cancer.

Materials and Methods

  • Phase III placebo-controlled, multiinstitutional, randomized trial
  • 569 pts with LA or metastatic pancreatic adenocarcinoma were randomized to one of two arms:
    • Arm 1 = G + E (n = 285)
    • Arm 2 = G + placebo (n= 284)
  • Pts were permitted to have prior history of local radiation therapy, but NO prior chemotherapy was allowed
  • Pts were stratified by enrolling institution, performance status (ECOG 0-1 vs. 2), and disease stage.
  • G was delivered at a dose of 1000 mg/m2 IV over 30 minutes and repeated cyclically.
  • E was given at a dose of 100 - 150 mg PO
  • Pts were well balanced in terms of gender, age (median ages were 63 and 64 years for arms 1 and 2, respectively), and stage
  • ~40% of pts were enrolled from US institutions
  • Primary endpoint was overall survival (OS).
  • Secondary endpoints included progession-free survival (PFS), response rates (RR), quality of life (QOL), and safety.



  • The combination of  G and E showed an OS benefit compared to G and placebo, with a hazard ratio (HR)     of 0.80 (p=0.018) 
    • In other words, there was a 25% relative improvement in survival and a 20% relative reduction in the risk of death with G and E compared to G and placebo.
    • Median survival was 6.37 months (mos) with E vs. 5.91 mos with placebo.
    • One-year survival rate was 24% with E vs. 17% with placebo.
  • PFS had a HR of 0.76 (p=0.003).
    • Median PFS was 3.75 mos with E vs. 3.55 mos.
  • The combination of complete response, partial response, and stable disease was 59% in the E arm vs. 49.4% in the placebo arm.
  • QOL was difficult to measure, with no clear differences between the two arms.
  • EGFR positivity vs. negativity did not predict for better survival or response to E.
  • The presence of a rash did predict for better survival outcomes (HR = 0.71 with garde 2+ rash).
  • The combination of G and E was well tolerated overall, with no significant increase in hematologic, renal, or hepatic toxicity compared to placebo. 
  • There was increased grade 1-2 rash, stomatitis, and diarrhea in the E arm, as expected given E's side effect profile.

Author's Conclusions

  • G and E is a superior regimen to the use of G alone in the treatment of LA and metastatic pancreatic cancer.
  • This therapeutic combination results in a 25% OS improvement, with better tumor control and PFS.
  • The incremental toxicity of the two agents is modest compared to single-agent G.
  • EGFR status by immunohistochemistry is not predictive of survival.
  • Development of grade 2 or higher skin rash did in fact predict for better survival with G and E.


Clinical/Scientific Implications

Earlier today (11/2/05), the US Food and Drug Administration (FDA) announced the approval of erlotinib in combination with gemcitabine chemotherapy for the treatment of advanced pancreatic cancer patients who have not received prior chemotherapy. This decision is based directly on the results of this very important trial conducted by Dr. Malcom Moore and colleagues. Erlotinib is the first drug to show a significant overall survival improvement when used with gemcitabine chemotherapy as initial treatment for pancreatic cancer in a phase III setting. (Erlotinib has already been approved for use in patients with non-small cell lung cancer whose disease has progressed despite prior chemotherapy.)  For a disease as challenging as pancreatic cancer, strides in therapy are immediately welcomed, as such strides have been few and far between. Also, this study emphasizes the value of EGFR targetting in pancreatic cancer.

However, critics warn that while the benefit seen here is statistically significant, it ultimately translates to a survival gain of only 12.8 days. Thus, some question the true meaningfulness of this gain, especially in light of erlotinib's expense. When considering this viewpoint, the excitement over this study and today's FDA announcement can be somewhat blunted. If nothing else, it serves to remind all of us that while this study represents statistically significant progress, the great need for better therapies and improved clinical outcome in this highly fatal disease still remains.

Current ongoing pancreas cancer trials are investigating the combination of gemcitabine and various agents, including: capecitabine, oxaliplatin, cetuximab, and bevacizumab. The results of these studies are eagerly awaited.