Three-gene immunohistochemical panel adds to clinical staging algorithms to predict prognosis for patients with esophageal adenocarcinoma

Abstract

Purpose: Esophageal adenocarcinoma (EAC) is a highly aggressive disease with poor long-term survival. Despite growing knowledge of its biology, no molecular biomarkers are currently used in routine clinical practice to determine prognosis or aid clinical decision making. Hence, this study set out to identify and validate a small, clinically applicable immunohistochemistry (IHC) panel for prognostication in patients with EAC.

Patients and methods: We recently identified eight molecular prognostic biomarkers using two different genomic platforms. IHC scores of these biomarkers from a UK multicenter cohort (N = 374) were used in univariate Cox regression analysis to determine the smallest biomarker panel with the greatest prognostic power with potential therapeutic relevance. This new panel was validated in two independent cohorts of patients with EAC who had undergone curative esophagectomy from the United States and Europe (N = 666).

Results: Three of the eight previously identified prognostic molecular biomarkers (epidermal growth factor receptor [EGFR], tripartite motif-containing 44 [TRIM44], and sirtuin 2 [SIRT2]) had the strongest correlation with long-term survival in patients with EAC. Applying these three biomarkers as an IHC panel to the validation cohort segregated patients into two different prognostic groups (P < .01). Adjusting for known survival covariates, including clinical staging criteria, the IHC panel remained an independent predictor, with incremental adverse overall survival (OS) for each positive biomarker (hazard ratio, 1.20; 95% CI, 1.03 to 1.40 per biomarker; P = .02).

Conclusion: We identified and validated a clinically applicable IHC biomarker panel, consisting of EGFR, TRIM44, and SIRT2, that is independently associated with OS and provides additional prognostic information to current survival predictors such as stage.

Fig 1.

Immunohistochemistry (IHC) panel is highly prognostic. Application of the IHC panel to all validation cohort patients segregated patients into two main prognostic groups (P < .01).

Fig 2.

Representative examples of dysregulated and nondysregulated immunohistochemistry expression of EGFR, SIRT2, and TRIM44 on tissue microarrays. Overexpression of EGFR and TRIM44 constitutes dysregulation, whereas loss of SIRT2 constitutes dysregulation, based on their effect on prognosis and known biologic roles. EGFR, epidermal growth factor receptor; SIRT2, sirtuin 2; TRIM44, tripartite motif-containing 44.

Fig 3.

Immunohistochemistry (IHC) panel adds to current staging criteria in patient prognostication. Combination of stage and the IHC panel separated patients into distinct prognostic groups (P < .01).

Fig A1.

Antibody sources and conditions used for immunohistochemistry (IHC).

Fig A2.

Internal validation of TRIM44, SIRT2, and EGFR as a combined IHC panel for prognostication in patients with esophageal adenocarcinoma in the original Oesophageal Cancer Clinical and Molecular Stratification Study cohort.

Fig A3.

Application of the immunohistochemistry panel to all patients resulted in segregation of patients into three prognostic groups in both the (A) Pittsburgh (P < .01) and (B) Rotterdam cohorts (P = .02).