Screening Biomarker as an Alternative to Endoscopy for the Detection of Early Gastric Cancer: The Combination of Serum Trefoil Factor Family 3 and Pepsinogen

Abstract

Objective: The serum pepsinogen test has limitation in its predictive power as a noninvasive biomarker for gastric cancer screening. We aimed to investigate whether the combination of TFF3 and pepsinogen could be an effective biomarker for the detection of gastric cancer even in the early stages.

Methods: In total, 281 patients with early gastric cancer (EGC), who underwent endoscopic submucosal dissection in Korea, and 708 healthy individuals from Japan were enrolled in the derivation cohort. The validation cohort included 30 Korean patients with EGC and 30 Korean healthy control blood donors. Serum TFF3 levels were examined using enzyme-linked immunosorbent assay.

Results: Using a cutoff of 6.73 ng/mL in the derivation cohort, the sensitivity of the combination of tests for EGC detection was superior (87.5%) to that of TFF3 (80.4%) or pepsinogen test alone (39.5%). Similarly, in the validation cohort, the sensitivity of TFF3 plus pepsinogen was higher (90.4%) than that of TFF3 (80.0%) or pepsinogen test alone (33.3%).

Conclusion: The combination of serum TFF3 and pepsinogen is a more effective noninvasive biomarker for gastric cancer detection compared with pepsinogen or TFF3 alone, even in EGC. This trial is registered with NCT03046745.

Figures

Figure 1

Serum trefoil factor family 3 (TFF3) levels in patients with gastric cancer were compared with healthy control individuals in the derivation cohort. The TFF3 level was significantly higher in patients with gastric cancer (P < 0.001).

Figure 2

Receiver operating characteristic (ROC) curves of trefoil factor family 3 (TFF3) to predict early gastric cancer presence in the derivation cohort. (a) ROC curve of serum TFF3 for all (both Helicobacter pylori-positive and Helicobacter pylori-negative) the patients. The sensitivity, specificity, odds ratio, area under the curve, and cutoff value of TFF3 were 0.804, 0.576, 5.60, 0.729, and 6.73, respectively. The positive and negative predictive values of TFF3 were 0.430 and 0.881, respectively. (b) For H. pylori-positive patients, the sensitivity, specificity, odds ratio, and area under the curve were 0.772, 0.576, 4.61, and 0.716, respectively. (c) For H. pylori-negative patients, the sensitivity, specificity, odds ratio, and area under the curve were 0.835, 0.576, 6.86, and 0.740, respectively.

Figure 3

Distribution of serum trefoil factor family 3 (TFF3) in differentiated or undifferentiated-type and intestinal- or diffuse-type early gastric cancer (EGC) in the derivation cohort. (a) The serum TFF3 levels of patients with differentiated-type histology and of those with undifferentiated-type EGC did not differ significantly (P = 0.056). (b) Serum TFF3 levels in patients with intestinal-type EGC was significantly higher than that in those with diffuse-type cancer (P = 0.028).

Figure 4

The Receiver operating characteristic curves of pepsinogen I/II ratio, serum trefoil factor family 3 (TFF3), and TFF3 plus pepsinogen I/II ratio are shown in the validation cohort. The sensitivity, specificity, odds ratio, and area under the curve of pepsinogen test for detection of EGC according to the definition of pepsinogen test were 0.333, 0.933, 7.00, and 0.633, respectively. Using the cutoff value of 6.73 ng/mL for TFF3, those for TFF3 were 0.800, 0.433, 3.06, and 0.651, respectively. Those for the combination of TFF3 and pepsinogen l/ll ratio were 0.900, 0.367, 5.21, and 0.756, respectively. AUC: area under the curve.