Rapid Fire Session 2014 World Cancer Congress

Does HER2 affect prognosis in oesophageal and gastric cancer: A systematic review (#441)

Barbara-Ann Adelstein 1 , Ramya Venkateswaran 1 , Maarit A. Laaksonen 1 , Robyn L. Ward 1
  1. University of NSW, Randwick, NSW, Australia


 Prognosis for gastric and oesophageal cancers remains poor. Human epidermal growth factor receptor 2 (HER2) is found in up to 30% of these cancers and may be a prognostic marker and thus target for treatment.


 To evaluate the prognostic value of HER2 status in gastric and oesophageal cancer.


 Systematic review of the literature (all languages) using defined criteria including provision of survival in HER2 positive (H2+ve) and negative (H2-ve) groups using immunohistochemistry (IHC) or in-situ hybridization (ISH) testing for these cancers.  HER2 status had not influenced treatment.


 65 papers (67 studies: 50 gastric, 17 oesophageal) published between 1970 and 2010 met the inclusion criteria, and represented 8,659 gastric and 2,075 oesophageal patients. The studies were heterogeneous and the methodological quality was flawed (consecutive samples not used, evaluation done post-hoc, small numbers, loss to follow up not reported). HER2 positivity ranged from 2-64%; 67% used IHC and 33% ISH testing, although testing methodology and positivity criteria varied.  Other differences included patient population (age; tumour stage; curative resection or not; treatment received), specimen used (biopsy or resection).   

For gastric cancer, in 2 studies within-study median survival in H2+ve was statistically significantly longer than in H2-ve (survival difference range 9.9-37.6 months), shorter in 21 studies (range 2.5-89 months) and no different in 23 studies.  For oesophageal cancer, in 1 study within-study median survival in H2+ve was statistically significantly longer than in H2-ve (13.6 months), shorter in 8 studies (range 4.6-48 months) and no different in 6 studies.

Heterogeneity of the studies precluded meta-analysis.


The published studies are generally of poor quality, with major methodological flaws. The scientific evidence for HER2 being prognostic in these cancers is not conclusive, and should not guide treatment.  HER2 as a predictive treatment marker is currently being evaluated in clinical trials.