In the Phase 3 EXPAND trial, patients from 25 countries were randomised to receive 3-week cycles of capecitabine 1000 mg/m2 twice daily (days 1–15) and cisplatin 80 mg/m2 IV (day 1) plus weekly cetuximab (400 mg/m2 loading dose on day 1 and 250 mg/m2 thereafter) (n=455), or the capecitabine/cisplatin combination alone (n=449). Professor Lordick explained that baseline characteristics were balanced between treatment arms and that the median duration of cetuximab treatment was 14.9 weeks with a relative dose intensity of ≥80% received by 82% of patients. Unfortunately, progression-free survival (PFS), overall survival (OS) and best overall response rate (ORR) were similar between treatment arms (PFS HR1.091; 95% CI: 0.920–1.292, p=0.3158; OS HR 1.004; 95% CI: 0.866–1.165, p=0.9547; ORR 30% versus 29%). Median PFS and OS was also comparable across various subgroups.
Professor Lordick also explained that the addition of cetuximab was associated with more grade 3/4 adverse events, in particular, skin rash (13% versus 0%), diarrhoea (8% versus 4%), hand-foot syndrome (7% versus 2%), hypomagnesemia (11% versus 1%) and hypokalemia (13% versus 9%).
Given these results from a large and well conducted trial, Professor Lordick concluded that the addition of cetuximab showed no benefit compared with chemotherapy alone (capecitabine + cisplatin) for the first-line treatment of advanced gastric cancer, and suggested that further classification of this heterogeneous disease may be required before advances in patient care can be made.
These data raise the question of whether trials in such unselected patient populations should still be conducted. The majority of patients with gastric cancer still present at an advanced stage, and despite advances in diagnostic and treatment strategies and a decline in incidence rates, outcomes remain poor. Whilst the use of classical chemotherapy agents has been explored thoroughly, and continues to be investigated, either alone or in various combinations, advances have been slow and the efficacy of these agents has reached a plateau. As such, the focus of research has shifted toward developing a greater understanding of the molecular biology of carcinogenesis and the cancer cell phenotype. This, in turn, will hopefully enable the development of rationally-designed drugs that target molecular aberrancies in signal transduction pathways specific to gastric cancer. For example, overexpression of members of the human epidermal growth factor receptor family has been reported in gastric cancer, with emerging data showing that EGFR and human epidermal growth factor receptor 2 (HER2) overexpression correlate with poor prognosis. As a result, several monoclonal antibodies and kinase inhibitors are undergoing clinical evaluation in this area, and findings from the Phase 3 ToGA trial recently showed that trastuzumab (in combination with chemotherapy) was associated with a survival benefit in HER2-positive patients with advanced gastric cancer.
As the efficacy of trastuzumab has been demonstrated in a patient population overexpressing the HER2 receptor, it seems logical that any benefits of an EGFR-targeted therapy may also be confined to a subset of patients overexpressing EGFR. This theory was explored yesterday when Dr Tom Samuel Waddell from the Royal Marsden Hospital, Sutton, UK, presented updated data and translational results from the REAL-3 trial, which evaluated the addition of the anti-EGFR antibody, panitumumab (P), to epirubicin (e), oxaliplatin (o) and capecitabine (coffee) in patients with advanced oesophago-gastric cancer. In this trial, patients with untreated, metastatic or locally advanced esophago-gastric cancer were randomised to EOC or modified-dose EOC + P. Dr Waddell explained that although the addition of panitumumab did not provide an improvement in OS or PSF in the overall study population, OS was significantly improved in patients who experienced grade 1–3 rash (77%, n=209), a known surrogate marker of response to EGFR inhibitors, compared with those without rash (23%, n=63): median OS was 10.2 months versus 4.3 months (p<0.001), and similar improvements were also seen in terms of ORR and PFS. In contrast, the presence of KRAS or PIK3CA mutation had negative prognostic value by multivariate analysis.