The prognostic value of tumor-stromal ratio combined with TNM staging system in esophagus squamous cell carcinoma

Abstract

Background: Tumor stroma is a crucial component of the tumor environment that interacted with tumor cells and modulated tumor cell proliferation, immune evasion, and metastasis. Tumor-stromal ratio (TSR) has been confirmed as an influential independent prognostic factor for diverse types of cancer, but it was seldom discussed in esophagus squamous cell carcinoma (ESCC). Methods: In present study, pathological sections from the most invasive part of the ESCC of 270 patients were analyzed for their TSR by visual inspection and software. The TSR was combined with the TNM staging system to further explain its predictive value of prognosis. The 57 cases ESCC from TCGA database also were included as an independently validated cohort. Results: Our results indicated that TSR was a robust prognostic factor for ESCC patients. TSR by visual inspection was dependable to reflect the stroma percent of the tumor compared to software calculation. Compared with stroma-low groups, the risk of death increased by 153.1% for patients in the stroma-high group [HR=2.531 (95%CI 1.657-3.867), P<0.001]. The results of ROC analysis in two cohorts indicated that TSNM staging system had better resolving ability with the largest area under the curve [0.698 95%CI (0.635-0.760), 0.691 95%CI (0.555-0.807)], compare to TNM. The novel TSNM staging system revealed strong predictive performance (P<0.001). Conclusion: TSR was a reliable dependent indicator for ESCC prognosis. The TSNM staging system has a better discriminative ability than the conventional TNM staging system, especially for III stage patients.

Keywords: ESCC; TNM staging system; pathology; prognosis; tumor-stromal ratio.

Figure 1

The evaluation of TSR. A Choose appropriate field to estimate TSR. B The Process of evaluation of TSR and typical stroma

Figure 2

Scatter plot of assessed stroma percentages in 270 patients for pathologists and software. The co-occurrence of assessed percentages was indicated by circles with areas proportional to the amounts of patients scored with the corresponding TSR value. The dashed lines represent the boundary between stroma-low and stroma-high group according to the 50% cut-off value. Red circles indicate cases where pathologists and software disagreed (11 in total). Evaluation of 127 (47.03%) cases were not coincide exactly.

Figure 3

The association of TSR with TNM staging system. A TNM staging system were heterogeneous. B Among the stage II, there was no significant difference between IIA and IIB (P=0.558). C There was significant difference between group S0 and S1 (P=0.0076). D Among the stage III, there was significant difference between IIIA, IIIB and IIIC (P=0.0037). E There was significant difference between group S0 and S1 (P=0.0324). F There were significant difference between S0 and S1 regarding all stage (P=0.0003).

Figure 4

The combination of TSR and TNM staging system. A The TSNM staging system, based on 7th TNM staging system and pS status. The number represents the number of people who belong to the staging. B Among the stage III, there was significant difference between IIIA, IIIB, IIIC, IIID (P=0.0029). C Among the stage II, there was no significant difference between IIA, IIB and IIC (P=0.0609). D TSNM performed well in predicting the clinical outcomes of ESCC patients compared to other factors