P-0067 - Treatment Patterns and Outcomes of Patients with Gastric Cancer in the US: A retrospective analysis of electronic medical record (EMR) data

Date 28 June 2014
Event World GI 2014
Session Poster Session
Topics Gastric Cancer
Cancer Aetiology, Epidemiology, Prevention
Presenter Lisa Hess
Citation Annals of Oncology (2014) 25 (suppl_2): ii14-ii104. 10.1093/annonc/mdu165
Authors L. Hess, D. Michael, D. Mytelka, J. Beyrer, S. Nicol
  • Eli Lilly & Co, Indianapolis/US



Gastric cancer is the fifth most common cancer worldwide, but is relatively uncommon disease in the US, where it has the 14th highest incidence rate of all cancers. Little is known about the care of these patients. The aim of this study was to conduct a retrospective database analysis of EMR data to describe the treatment patterns and outcomes of patient diagnosed with gastric cancer in the US.


Individuals diagnosed with gastric cancer were identified from the IMS Oncology Database. Patients eligible for inclusion were 18 years of age or older and had a recorded diagnosis of gastric cancer by an ICD-9 code of 151.0-151.9. Eligible cases must have had the initial gastric cancer diagnosis code (index diagnosis) recorded between 01 Jan 2004 and 01 Jan 2012 to allow up to one year of follow up in the database. Patients were excluded if there were ICD9 codes reflecting a cancer diagnosis within 6 months of the index diagnosis. Patients with evidence of chemotherapy between one and six months prior to the index date were excluded to avoid confounding of treatment associated with other cancers. Descriptive statistics were used to explore patient demographics (age, gender, race/ethnicity), disease and clinical characteristics (cancer stage, site of tumor, performance status), chemotherapy treatment patterns, and estimated survival outcomes (death was estimated by the date of discontinuation in the IMS database).


There were 5,257 patients with gastric cancer who met all eligibility requirements; 1,982 (37.7%) of these patients also had data regarding chemotherapy treatments. The mean age was 64 years (standard deviation = 13), and the majority (60.8%) was male. 56.9% of cases included race/ethnicity data; of those, 64.7% were white, 17.7% African American, 3.6% Asian, and 3% Hispanic. Of the 1,580 (30.1%) patients with known stage data, 65% were diagnosed at an advanced stage (21.8% and 43.2% were stage III and IV, respectively). Platinum and/or fluoropyrimidine agents were used in 89.7% and 78.9% of patients receiving first- and second-line therapy, respectively, alone or with other agents. Of the 1,982 patients who received first-line therapy, 42.3%, 18.1% and 7.9% went on to receive a second, third and fourth line of chemotherapy, respectively. For post-first line therapy, no clear treatment strategy was observed, yet many of the same agents from the first-line regimens were also used in the second line as singlet, doublet or triplet and even quadruplet regimens. The median estimated survival declined with increasing stage (Stage I: 28.1 months; Stage II: 20.2 months; Stage III: 15.4; Stage IV: 8.0 months).


Patients with gastric cancer are commonly treated with first-line platinum and/or fluoropyrimidine therapy. Second-line regimens also often used the same therapeutic classes, but regimens were increasingly heterogeneous. This study is limited due to the levels of missing data in this EMR database. Therefore, conclusions can only be made among patients with complete data and cannot be generalized to the entire US population. Nevertheless, this is the largest EMR study to our knowledge to evaluate gastric cancer treatment patterns and outcomes in the US.