Our results show that EGFR status is not prognostic of patient outcome in metastatic gastric and GE junction adenocarcinoma. We also found no impact of EGFR status on progression-free survival, indicating that EGFR overexpression is not associated with more aggressive tumor biology or with resistance to chemotherapy in gastric and GE junction adenocarcinoma. Our analysis is based on a large and uniform cohort of Western patients with metastatic gastric cancer, all treated with standardized chemotherapy in a clinical trial. EGFR testing was performed according to different scoring systems and methods (protein and mRNA gene expression) reviewed by referenced pathologists, with other clinical and pathological characteristics captured prospectively in research databases. Unlike other studies, our cohort consists solely of patients with stage IV disease with well annotated chemotherapy data available on all of our patients, and none received EGFR targeted therapies in the first- or second-line setting. Our cohort is by far the largest (n = 357) reporting on the prognostic effect of EGFR on metastatic disease gastric cancer. Previous data from patients with metastatic gastric cancer are limited to two cohorts of 86 and 43 patients and have delivered conflicted results regarding the prognostic value
Regarding the curable stages, more data exist but the prognostic role of EGFR expression in operable gastric cancer remains controversial. Expression of EGFR in resected gastric cancer has been linked to shorter overall survival, more advanced tumor stage, and lymph node metastases in some studies, but not in others
[5–9, 12, 13]. For example, Kim et al.
 found a correlation of EGFR expression and improved overall survival in patients with resected gastric cancer receiving adjuvant chemotherapy.
The controversial findings in resectable stages may particularly derive from missing standardized procedures and the lack of an established scoring system in the immunohistochemical evaluation of EGFR. Besides proper definition of the target population and tumor characteristics, it is important to reflect distinct information of immunohistochemical EGFR expression like intensity of staining, staining pattern (focal or homogenous), content of tumor cells and choice of primary antibodies.
Although EGFR is not a prognostic factor in metastatic gastric cancer, this is not a reflection on its value as a predictive marker. This is in line with recent results from HER2, which is an established predictive factor for treatment response to trastuzumab, while (according to recent reports) Her2 expression itself is not a prognostic factor in metastatic gastric cancer
[4, 22, 23].
While no associations between Her2 expression (any grade) and clinicopathologic criteria were seen, we found that the rate of EGFR 3+ status was significantly associated with intestinal type histology (intestinal, 16%; diffuse, 8%; p = 0.05). The same pattern is known form Her2 expression
[4, 22] and may indicate a link between high expression of the erb-receptor family and a distinct disease biology in gastric cancer. In the TOGA-trial
 it was shown that the extent of Her2 expression is of relevant predictive value. It was clearly demonstrated, that anti HER2 treatment is only reasonable in patients with high intensity HER2 expression (HER2 score 3+). Whether this observation will be applicable to EGFR inhibitors is unclear.
With emerging development of drugs interacting with the EGF receptor or the EGFR pathway with monoclonal antibodies like cetuximab and panitumumab, there is an enormous need of better understanding the way of interaction of these drugs and the need of identifying subgroups of patients, who are likely to have a clinical benefit.
In the clinical setting, anti EGFR antibodies seem to enhance the activity of chemotherapy with improved response rates up to 60%
. But recently presented results of a phase III study (REAL-3,
) comparing a first-line palliative chemotherapy (epirubicin, oxaliplatin and capecitabin) with or without the anti-EGFR antibody panitumumab, could not demonstrate an OS/PFS benefit or even showed a worse outcome in patients in the experimental arm.
It will be interesting if these studies could identify molecular subgroups of patients, who nevertheless could benefit from an EGFR targeted treatment.