Abstract
The standard treatment for gastrointestinal cancer is surgical resection and perioperative adjuvant treatment. Thus far, gastrointestinal cancer research has mainly focused on cancer cells themselves. Recently, the tumor microenvironment (TME) has been a subject of investigation. The TME is a complex system composed of various cell types, including tumor cells, endothelial cells, stromal cells, immune cells, and extracellular components. Among them, the stromal cells surrounding tumor cells are being investigated in gastrointestinal cancers. Stromal cells play a role in tumor growth, invasion, and metastasis. Moreover, stromal cells are associated with increased chemotherapy resistance and reduced chemotherapy delivery. Therefore, it is necessary to develop prognostic or predictive factors that consider the tumor-stroma interaction. The tumor stroma ratio (TSR) has recently been shown to be a promising outcome prediction tool in various malignancies. The TSR is based on the proportion of the stroma to tumor area. Recent studies demonstrated that a high amount of stroma or low TSR was associated with a poor prognosis and was a predictor of various treatment modalities. Therefore, to optimize the gastrointestinal cancer treatment, it is necessary to understand the role of the TSR in gastrointestinal cancers. This review summarizes the background, current status, and future perspectives of the TSR for gastrointestinal cancer treatment.
An estimated 14.1 million new cancer cases and 8.2 million cancer deaths occurred worldwide in 2018 (1). Curative resection and perioperative adjuvant treatment is a standard method of treating locally advanced gastrointestinal cancer (2-6). However, even when gastrointestinal cancer patients receive curative treatment, approximately half of the patients develop tumor recurrence, which is associated with a poor prognosis. Therefore, it is necessary to identify prognostic factors and/or predictive factors for improved survival in gastrointestinal cancer patients who receive adjuvant treatment. Previously, gastrointestinal cancer research mainly focused on the cancer cell itself by investigating the roles of tumor suppressor and oncogenic factors in transformation to malignancy (7). Recently, the tumor microenvironment (TME) has become a subject of investigation (8). TME is a complex system composed of various cell types, including tumor cells, endothelial cells, stromal cells, immune cells, and extracellular components. Among them, stromal cells surrounding tumor cells are being investigated in gastrointestinal cancers. Stromal cells play a role in tumor growth, invasion, and metastasis (9, 10). Moreover, stromal cells are associated with increased chemotherapy resistance and reduced chemotherapy delivery (11, 12). Therefore, it is necessary to develop prognostic or predictive factors that consider the tumor-stroma interaction. The tumor stroma ratio (TSR) has recently been shown to be a promising outcome prediction tool for a variety of malignancies (13-16). The TSR is based on the proportion of the stroma to tumor area. Recent studies demonstrated that a large amount of stroma or low TSR was associated with a poor prognosis and was a predictor of outcomes of various treatment modalities. This review summarizes the background, current status, and future perspectives of the TSR for gastrointestinal cancer treatment.
Search Strategy
The search strategy used in the current study was based on the Preferred Reporting Items for Systematic Reviews and Meta-Analyses guidelines. A literature search was performed on PubMed. The following keywords were used: “tumor stroma ratio (TSR)” and “esophageal cancer (or carcinoma)”; “proportion of tumor (PoT)” and “esophageal cancer (or carcinoma)”; “tumor stroma ratio (TSR)” and “gastric cancer (or carcinoma)”; “proportion of tumor (PoT)” and “gastric cancer (or carcinoma)”; “tumor stroma ratio (TSR)” and “colorectal cancer (or carcinoma)”; “proportion of tumor (PoT)” and “colorectal cancer (or carcinoma)”; “tumor stroma ratio (TSR)” and “hepatobiliary pancreatic cancer (or carcinoma)”; “proportion of tumor (PoT)” and “hepatobiliary pancreatic cancer (or carcinoma)”. In addition, the references of the cited articles were overlooked. In total, 920 articles were identified. However, 782 were excluded.
TSR for Esophageal Cancer
Four reports have evaluated the clinical impact of the TSR in esophageal cancer (Table I) (17-20). Frist, Courrech Staal et al. evaluated the prognostic value of the TSR in 93 esophageal cancer patients who received curative resection (17). The TSR was visually estimated based on morphological characteristics from 5-μm H&E-stained sections using surgical specimens. They divided 93 patients into a TSR high group (n=60) and a TSR low group (n=33) using a TSR cut-off value of 50%. The patient background factors of the TSR high and TSR low group were similar. The three-year overall survival (OS) rates of the TSR high and TSR low groups were 53% and 19%, respectively. The TSR was identified as a significant prognostic factor for OS in a multivariate analysis [hazard ratio (HR)=2.006, 95% confidence interval (CI)=1.181-3.407, p=0.010]. In addition, the 3-year recurrence-free survival (RFS) rates of the TSR high and TSR low groups were 43% and 12%, respectively. The TSR was identified as a marginally significant prognostic factor for RFS in a multivariate analysis (HR=1.553, 95%CI=0.923-2.611, p=0.093). They concluded that a low TSR after curative treatment of esophageal cancer was associated with poor long-term survival. Courrech Staal et al. also validated and reported the reproducibility of the TSR using biopsy specimens (18). Wang et al. evaluated the prognostic value of the TSR in 95 esophageal cancer patients who received curative treatment (19). They visually estimated the TSR using 5-μm H&E-stained sections from surgical specimens. They divided 95 patients into a stroma poor group (the proportion of stroma <50%, n=65) and a stroma rich group (the proportion of stroma ≥50%, n=30). The three-year OS rates in the stroma poor and stroma rich groups were 64% and 23%, respectively. The HR of the TSR for OS was 3.548 (95%CI=1.977-6.365, p=0.001). Moreover, the 3-year RFS rates in the stroma poor and stroma rich groups were 57% and 23%, respectively. The hazard ratio of the TSR for RFS was 3.450 (95%CI=1.913-6.222, p=0.001). They concluded that the TSR was associated with a poor prognosis and an increased risk of relapse. Pelt et al. evaluated the clinical value of the TSR as a predictor of a pathologic response after neoadjuvant chemoradiation therapy in 94 esophageal cancer patients who received neoadjuvant treatment and esophagectomy (20). The TSR was assessed using H&E-stained sections from diagnostic biopsies. A cut-off value of 50% was used to categorize the patients into the stroma-low (<50%, n=76) or stroma-high (>50%, n=18) groups. They demonstrated that the response rates for adjuvant treatment in the stroma-high tumor and stroma-low tumor groups were 22% and 50%, respectively (odds ratio=3.57, 95%CI=1.03-12.31, p=0.004). They concluded that a high proportion of stroma cells was associated with a poor prognosis and a poor response to treatment in esophageal cancer. They also reported that their sample size was less than 100 and that a validation study with a large cohort was needed.
Significance of tumor stroma ratio for esophageal cancer.
TSR for Gastric Cancer
Four studies have evaluated the clinical impact of the TSR in gastric cancer (Table II) (21-24). Lee et al. clarified the prognostic impact of the intratumor stromal proportion in 175 gastric signet ring cell carcinoma patients (21). They divided 175 patients into stroma rich (n=64) and stroma poor (n=111) groups using a TSR cut-off value of 50%. A stroma rich status was significantly associated with a large tumor and advanced tumor invasion status. The mean OS in the stroma poor and stroma rich groups was 105 months and 66 months, respectively (p<0.001). The TSR was one of the significant prognostic factors for OS (HR=2.503, 95%CI=1.476-4.246, p=0.001). Similar trend was observed for disease-free survival. They concluded that the intratumor stroma proportion could be a useful prognostic factor and potential therapeutic target in gastric signet ring cell carcinoma. Aurello et al. evaluated the prognostic value of the TSR in 106 gastric cancer patients who received gastrectomy (22). The TSR was visually counted from 4-μm H&E-stained sections from surgical specimens. They divided 106 patients into stroma poor (n=41) and stroma rich (n=65) groups using a TSR cut-off value of 50%. The comparison of the patient background characteristics of the stroma rich and stroma poor groups revealed that the incidence of a positive lymph node ratio and the incidence of T3/T4 were significantly higher in the stroma rich group than those in the stroma poor group (p<0.001 and p<0.001, respectively). The 5-year OS rates of the stroma rich and stroma poor groups were 81% and 26%, respectively. The TSR was identified as a significant prognostic factor for OS in a multivariate analysis (HR=5.50, 95%CI=2.34-12.92, p<0.001). Peng et al. evaluated the prognostic value of the TSR in 494 advanced gastric cancer patients who received curative resection (23). The TSR was visually estimated using 4-μm H&E-stained sections from the most invasive part of the primary tumor. They divided 494 patients into stroma high (n=240, 48.6%) and stroma low (n=254, 51.4%) groups using a TSR cut-off value of 50%. On comparing the patient background factors of the stroma high and stroma low groups, the incidence of T3/T4 was significantly higher in the stroma high group (83.8% vs. 74.4%, p=0.011). Three and five years OS rate was 44.71% and 27.64% in stroma high group, while 66.19% and 58.68% in stroma low group. The TSR was identified as a significant prognostic factor for OS in a multivariate analysis (HR=1.911, 95%CI=1.427-2.559, p<0.001). They concluded that that a high TSR in gastric cancer who received curative treatment was associated with a poor long-term survival. Kemi et al. investigated the association between the TSR and the prognosis in 583 gastric cancer patients who underwent surgery (24). They calculated the TSR using a software program. H&E-stained slides were scanned and digitized using Aperio AT2 and the TSR was analyzed from scanned slides using an Aperio Image Scope. They divided 583 patients into a stroma rich (n=342, 58.7%) and a stroma low group (n=241, 41.3%) at TSR cut-off value of 50%. Five-year survival was significantly better in the stroma poor group than the stroma rich group (44.6% vs. 21.3%, p<0.001). In a multivariate analysis, the stroma rich group had a significantly worse prognosis than the stroma poor group (HR=1.80, 95%CI=1.41-2.28). Similar trends were observed in both the patients with intestinal type histology and those with diffuse type histology. They concluded that a high proportion of stroma is an independent prognostic factor, regardless of histological subtype.
Significance of tumor stroma ratio for gastric cancer.
The TSR for Colorectal Cancer
Seven studies have been conducted to evaluate the clinical impact of the TSR in colorectal cancer (Table III) (25-31). West et al. clarified the clinical impact of the proportion of tumor determined by objective point counting on virtual scanned H&E-stained slides in 145 colorectal cancer patients who had received curative resection (25). The proportion of tumor cells within the tumor was classified as high (>47% of tumor cells within the tumor) or low (<47% of tumor cells within the tumor). According to this classification, 110 patients were assigned to the high proportion group and 35 to the low proportion group. In the survival analysis, a low proportion of tumor cells was significantly associated with poorer cancer-specific survival. They concluded that a low proportion of tumor cells in colorectal cancer is related to poor cancer-specific survival. Hujibers et al. investigated the association between the TSR and prognosis in 710 colon cancer patients (26). The TSR was measured visually under a microscope. They divided 583 patients into a stroma high group (n=207, 29.2%) and a stroma low group (n=503, 70.8%) using a TSR cut-off value of 50%. Five-year survival was significantly better in the stroma low group than that in the stroma high group (83.4% vs. 69.0%, p<0.001). In a multivariate analysis, stroma rich group had significantly worse prognosis compared to the stroma poor group (HR=1.96, 95%CI=1.41-2.74). Similar trends were observed for DFS. Five-year DFS was significantly better in the stroma low group than that in the stroma high group (77.3% vs. 58.6%, p<0.001) (HR=2.15, 95%CI=1.61-2.86, p<0.001). They concluded that high proportion of stroma is an independent prognostic factor for colon cancer. Park et al. evaluated the prognostic impact of the tumor stroma proportion (TSP) in 331 colorectal cancer who received curative resection (27). They calculated the TSR using a software program. They divided 331 patients into TSP high (stroma rich) (n=81) and TSP low (stroma poor) groups (n=250) using a TSP cut-off value of 50%. In the survival analysis, a high TSP was associated with reduced cancer specific survival (HR=1.84, 95%CI=1.17-2.92, p=0.009). They concluded that a high TSP in colorectal cancer was related to poor cancer-specific survival. Similar trends were observed in node-negative colorectal cancer patients and the patients who received adjuvant chemotherapy. Scheer et al. evaluated the prognostic value of the TSR in 154 rectal cancer patients (28). They visually calculated the TSR using H&E-stained slices from surgical specimens. They categorized the patients into three groups: TSR-low [carcinoma proportion (CP) <30%], TSR-intermediate (CP 40-60%), and TSR-high (CP >70%). Among 154 patients, 26 patients were categorized into the TSR-low group, 70 into the TSR-intermediate group, and 48 into the TSR-high group. The 5-year OS rates in the TSR-high, TSR-intermediate, and TSR-low groups were 64.6%, 50.0%, and 55.6%, respectively. Similar trends were observed for disease-free survival. In the prognostic factor analysis, intermediate TSR had higher risk of dying from rectal cancer. They concluded that the TSR had potential application as a prognostic factor for rectal cancer patients. Zunder et al. investigated the predictive potential of the TSR in 1,213 stage II or III colon cancer patients who received oxaliplatin-based chemotherapy with or without addition of bevacizumab (29). The TSR was calculated using H&E-stained slides. Among 1,213 patients, 339 patients were categorized into the stroma high group and 824 were categorized into the stoma low group. The DFS and OS were significantly shorter in the stroma high group in comparison to the stroma low group (HR for DFS: 1.75, 95%CI=1.32-2.33, p<0.001; HR for OS: 1.54, 95%CI=1.04-2.29, p=0.03). Eriksen et al. evaluated the prognostic value of the TSR in 573 stage II colon cancer patients who received adjuvant chemotherapy (30). The TSR was scored on H&E sections as low TSR (>50% stroma) and high TSR (<50% stroma). Four-hundred four patients (70.5%) were categorized into the high TSR group and 169 (29.5%) were categorized into the low TSR group. The 5-year OS rates in the low TSR and high TSR groups were 61.0% and 72.6%, respectively. In a multivariate analysis, the TSR was selected as a significant prognostic factor (HR=1.376, 95%CI=1.016-1.862, p=0.039). Similar trends were observed in RFS. Sandberg et al. investigated the clinical role of the TSR in 71 colorectal cancer patients who received surgery (31). The TSR was scored on H&E sections at the invasive part of the tumor using a microscope. They divided 20 (28.2%) patients into a stroma-high group and 51 (71.8%) into a stroma-low group. The stroma high group showed a significantly poorer prognosis than the stroma low group (25% vs. 78.4%, HR=4.586, 95%CI=1.96-10.75, p=0.001). In addition, they reported that the activated microenvironment in stroma high tumors over-expressed different types of collagen genes, including THBS2 and 4, as well as INHBA, COX71A and LGALS1/galectin-1.
Significance of tumor stroma ratio for colorectal cancer.
The TSR for Hepatobiliary Cancer
Lv et al. evaluated the prognostic impact of the TSR in 300 hepatocellular carcinoma patients who received liver resection or transplantation (32). They calculated the TSR using a software program (Image Pro-Plus version 6.0). They divided 300 patients into a stroma poor (TSR high) group (n=225) and a stroma rich (TSR low) group (n=75) using a TSR cut-off value of 50%. The TSR was significantly associated with the pathological TNM stage. The rate of advanced pathological stage cases (stage III or IV) was higher in the stroma rich group than that in the stroma poor group (56.0% vs. 31.1%, p<0.001). In patients who received liver resection, the 5-year survival rates in the stroma rich and stroma poor groups were 37% and 83%, respectively. In the patients who received liver transplantation, the 5-year survival rates in the stroma rich and stroma poor groups were 28% and 63%, respectively. In both the liver resection group and liver transplantation group, the stroma rich group had a significantly poorer prognosis than the stroma poor group (p<0.001 and p<0.001, respectively). Similar trends were observed for disease-free survival. They concluded that the TSR is an independent prognostic factor for patients with hepatocellular carcinoma.
Future Suggestions
Thus far, the TSR is considered to be a promising prognostic factor and/or predictive factor for gastrointestinal cancer. To optimize the TSR in clinical use, the following points should be clarified. First, it is necessary to validate the reproducibility of the TSR scoring in biopsy specimens. Previous studies evaluated the TSR in surgical specimens. On the other hand, recently, preoperative adjuvant treatment, such as preoperative chemotherapy and preoperative radiation therapy, have been introduced in gastrointestinal cancer treatment. Preoperative adjuvant treatments reduce tumor cells, increase stromal cells, and affect the TSR. Therefore, to evaluate the TSR before treatment, it is necessary to use biopsy specimens for evaluation of the TSR. Courrech Staal et al. reported the validation study to determine the correlation between endoscopic biopsy specimens and surgical specimens in esophageal cancer (18). They reported that there was reproducibility of TSR scoring between endoscopic biopsy specimens and surgical specimens. Thus, reproducibility between endoscopic biopsy specimens and surgical specimens needs to be confirmed in other gastrointestinal cancers. Second, the optimal timing to evaluate the TSR should be determined. Previous studies evaluated the TSR using surgical specimens. On the other hand, as mentioned above, perioperative adjuvant treatment has been introduced for gastrointestinal cancer treatment. Perioperative adjuvant treatment might affect the TSR. Therefore, it is necessary to consider whether to use pre-adjuvant treatment or post-adjuvant treatment specimens. Third, the optimal cut-off value of the TSR is needed. All previous studies set the cut-off value of the TSR at 50%. Although the type of cancer, patient characteristics, number of patients, and treatment method were different, all studies set the same cut-off value. The type of cancer and tumor stage might affect the TSR. Therefore, the optimal cut-off value of the TSR needs to be set according to the tumor type and stage. Fourth, it is necessary to develop the optimal method to evaluate the TSR. Previous studies visually assessed the TSR. Recently, to minimize the problems associated with the visual methods for assessing the TSR, computer-aided deep learning-based methods are being developed to facilitate automated assessment (33). Future studies need to clarify these issues.
Conclusion
The TSR may have some clinical influence on both the short- and long-term oncological outcomes of patients with gastrointestinal cancer. However, the optimal cut-off values of the TSR, optimal evaluation methods, and optimal timing to evaluate TSR are unclear. To optimize the TSR for patients with gastrointestinal cancer, it is necessary to clarify these points in further studies.
Footnotes
Authors’ Contributions
TA and IH made substantial contributions to the concept and design. TA, YM, and TO made substantial contributions to the acquisition of data and the analysis and interpretation of the data. TA and IH were involved in drafting the article or revising it critically for important intellectual content. TA, IH, and TO gave their final approval of the version to be published.
Conflicts of Interest
The Authors have no conflicts of interest to declare in relation to this study.
- Received February 20, 2023.
- Revision received March 2, 2023.
- Accepted March 3, 2023.
- Copyright © 2023 International Institute of Anticancer Research (Dr. George J. Delinasios), All rights reserved.
This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY-NC-ND) 4.0 international license (https://creativecommons.org/licenses/by-nc-nd/4.0).
References
In this issue
Jump to section
Related Articles
Cited By...
- No citing articles found.