Leucine-rich repeat-containing G protein-coupled receptor 5 and CD133 expression is associated with tumor progression and resistance to preoperative chemoradiotherapy in low rectal cancer.

Preoperative chemoradiotherapy has been performed as a standard therapy for advanced low rectal cancer. Cancer stem cells (CSCs) have been reported to contribute to resistance to treatment and patient prognosis. Leucine-rich repeat-containing G protein-coupled receptor 5 (LGR5) and cluster of differentiation (CD133) are putative markers for CSCs. However, their prognostic ability remains unknown, and evaluation of a single marker can be insufficient due to the heterogeneity of cancer. LGR5 and CD133 expression was immunohistochemically evaluated in surgical specimens of 56 patients who received curative resection following chemoradiotherapy for advanced low rectal cancer. In addition, the correlations between their expression levels, and clinicopathological features and patient prognosis were asessed. LGR5 expression was significantly correlated with lymphatic invasion, lymph node metastasis, and tumor node metastasic (TNM) stage. CD133 expression was significantly correlated with vascular invasion and the tumor regression grade. Combined expression was significantly correlated with lymphatic invasion, tumor regression grade and TNM stage, but not with overall, and disease-free survival. LGR5 and CD133 expressions may represent useful markers associated with tumor progression and resistance to chemoradiotherapy in patients with low rectal cancer. Furthermore, combined expression of these markers may be a more useful marker compared with the expression of each single marker.

Introduction

Recently, in the treatment of advanced low rectal cancer, preoperative chemoradiotherapy (CRT) has been widely accepted as a standard therapy. Previous studies have shown that preoperative CRT contributes to tumor down-staging and decreases locoregional recurrence post-surgery (1). Although the clinical significance of the CRT response has not been fully elucidated, several reports described that it may represent a predictor of clinical outcome, including tumor recurrence and patient survival. However, patients show different responses to CRT, with some cases showing little or no response (2). Therefore, prediction of the CRT response is warranted to avoid unnecessary treatment and adverse events such as radiation dermatitis, hematologic toxicity, and enteritis.

The cancer stem cell (CSC) theory, proposed by Hamburger et al, states that tumor cells are not only heterogeneous but also show a hierarchy, and CSCs, comprising a small part of tumors, are responsible for this heterogeneity due to their self-renewal and proliferation abilities (3). CSCs have been reported to be associated with tumor progression and recurrence. Although conventional cytotoxic therapies, such as chemotherapy or radiotherapy, target rapidly dividing cells, CSCs are estimated to be resistant to those therapies, as they divide more slowly (4,5). Finding a specific marker to identify CSCs is important, and prior studies have reported several putative CSC markers for colorectal cancer (CRC), including leucine-rich repeat-containing G protein-coupled receptor 5 (LGR5) and cluster of differentiation-133 (CD133) (6,7).

LGR5 is a glycoprotein hormone receptor with a seven transmembrane domain. Carmon et al reported that R-spondin proteins function as ligands of LGR5 and that LGR5 is a target of the Wnt/β-catenin signaling pathway, which is important for the maintenance of the colonic crypt (8). Barker et al demonstrated that LGR5 localized at the crypt base in the small intestine, and that LGR5-positive cells could generate multiple cell lineages of intestinal epithelium (9). Thus, LGR5 is considered a stem cell marker in the small intestine and colon. Of note, LGR5 is detected in many different tumors, including CRC (10–13). In CRC, previous studies, including two meta-analyses, demonstrated that LGR5 expression was associated with tumor progression such as tumor depth, lymphovascular invasion, lymph node metastasis, advanced tumor stage, and poor overall survival (OS) (14–16). Moreover, prior studies have evaluated LGR5 expression in rectal cancer patients treated with preoperative CRT, and reported associations with worse sensitivity to CRT and poor patient prognosis (17).

CD133 is a five transmembrane glycoprotein that is widely detected in many tumors, including colon cancers (18–20). Prior studies have reported that CD133-positive colon cancer cells have self-renewal and proliferation abilities (7). In CRC, CD133 has also been reported to be associated with tumor depth, lymphovascular invasion, lymph node metastasis, and advanced tumor stage (21). Further, two meta-analyses demonstrated that CD133 expression correlated with worse OS (22,23). The correlations between CD133 expression and response to CRT have also been evaluated, with CD133 shown to associate with CRT resistance and poor patient prognosis (24–26).

However, these previous reports evaluated the clinical significance of LGR5 and CD133 expressions separately. To our knowledge, the significance of combined LGR5 and CD133expression remains unclarified. Therefore, herein, we evaluated both LGR5 and CD133 expressions immunohistochemically in low rectal cancer patients treated with preoperative CRT using serial sections, and analyzed the relationships between those expressions and clinicopathological features and patient prognosis.

Materials and methods

Patients and specimens

Sixty one consecutive patients underwent curative resection after CRT for advanced low rectal cancer in the Department of Surgical Oncology, University of Tokyo Hospital between March 2001 and October 2009. All patients were diagnosed as low rectal cancer, and the tumor depth was estimated to be deeper than the muscularis propria. Tumor depth, nodal status, and presence of distant metastases were determined by computed tomography and magnetic resonance imaging. They received a total radiation dose of 50.4 Gy (1.8 Gy in 28 fractions) and concomitant chemotherapy (oral administration of tegafur-uracil 300 mg/m2/day and leucovorin 75 mg/day). Total mesorectal excision with lymph node dissection was performed following an interval of 6–8 weeks post-CRT. All patients underwent regular follow-up examinations post-surgery. Tumor markers were examined every 3 months, and abdominal and chest computed tomography was performed every 6 months. Total colonoscopy was performed annually. Since we targeted the residual cancer tissue, 5 cases with no residual cancer cells after CRT were excluded. We analyzed 56 surgically resected specimens after CRT by immunohistochemistry.

All specimens were fixed in 10% formalin and embedded in paraffin. The histopathological findings were confirmed by the Department of Pathology, University of Tokyo. Data were collected from the patients' medical records. The TNM classification was determined according to the Union for International Cancer Control, 7th edition. The post-CRT histological tumor regression grade was evaluated according to the Japanese Classification of Colorectal Carcinoma, 8th edition (Grade 0: No necrosis or regressive change, 1a: >66.6% vital residual tumor cells, 1b: 33.3–66.6% vital residual tumor cells, 2: <33.3% vital residual tumor cells, 3: No vital residual tumor cells).

This study was approved by the ethics committee of the University of Tokyo on July 29, 2014 [No. 10476-(1)] and written informed consent was obtained from all patients.

LGR5 and CD133 Immunohistochemical staining

The tumor specimens were immunohistochemically stained, as described below. The sections were deparaffinized in xylene, hydrated through a graded series of ethanol, and endogenous peroxidase was blocked. Heat-induced antigen retrieval was performed by incubation in sodium citrate buffer using an autoclave and non-specific proteins were blocked with 5% bovine serum albumin. Primary anti-LGR5 rabbit (Clone EPR3065Y, Epitomics, Burlingame, CA, USA) and anti-CD133 mouse monoclonal antibodies (Clone AC133, Miltenyi Biotec, Auburn, CA, USA) were added at dilutions of 1:100 and the sections were incubated overnight at 4°C. They were incubated with the Dako Envision kit (Dako, Carpinteria, CA, USA) following the manufacturer's recommendations. The reactivity was visualized in 2% 3,3′-diaminobenzidine tetrahydrochloride and 50 mM tris-buffer containing 0.3% hydrogen peroxidase. The crypt base of the normal colon mucosa and renal tubules were used as the positive controls for LGR5 and CD133, respectively. As a negative control, the antibody was replaced with PBS.

Evaluation of LGR5 and CD133 immunostaining

We defined positive expression of LGR5 as >50% positive cancer cells out of all cancer cells, since Saigusa et al also evaluated LGR5 expression in rectal cancer tissue after CRT using same cut-off value to clarify clinical significance of LGR5 (17). Positive expression of CD133 was defined as >5% positive cancer cells out of all cancer cells (27), since we previously showed CD133 expression in rectal cancer after CRT was associated with CRT response (22), and speculated the cut-off value to be appropriate. The sections were observed by a surgeon trained in pathology and a skillful pathologist, independently in a blinded fashion (S.K. and T.M.). Any discrepancies were resolved by discussion. Subsequently, we analyzed the correlations between the expressions of LGR5 and CD133 and clinicopathological factors and patient prognosis.

Statistical analysis

The correlations between LGR5 and CD133 expressions and clinicopathological factors were evaluated by the chi-squared test, Fisher's exact test, or unpaired t-test, as appropriate. OS and disease-free survival (DFS) were analyzed by the Kaplan-Meier method. P-values <0.05 were considered statistically significant. All analyses were performed using JMP 11.0 software (SAS Institute Inc., Cary, NC, USA).

Results

Clinicopathological findings

The clinicopathological factors are listed in Table I. The median patient age was 61 (range, 33–78) years, and 37 patients (66.1%) were male. After preoperative CRT, 21 (37.5%) and 35 (62.5%) patients had T1-2 and T3-4 tumors, respectively. Thirty-nine patients (69.6%) showed papillary carcinoma or well-differentiated adenocarcinoma histology. Nine (16.1%) and six (10.7%) patients had lymph node and distant metastases (3 liver, 1 lung, 1 brain, and 1 paraaortic lymph node metastases), respectively. Based on the response to CRT classification, 17 (30.4%), 18 (32.1%), and 21 (37.5%) patients were categorized as Grades 1a, 1b, and 2, respectively. The median follow-up period was 5.8 (range, 0.8–10.9) years.

Table I.

Characteristics of rectal cancer patients.

Characteristic	No. of patients (%)
Gender
Male	37 (66.1)
Female	19 (33.9)
Mean age ± SD (years)	61.1±9.9
pT stage[a]
T1-2	21 (37.5)
T3-4	35 (62.5)
Histological type
Pap, Well	39 (69.6)
Mod, Por, Muc	17 (30.4)
Lymphatic invasion
Present	5 (8.9)
Absent	51 (91.1)
Vascular invasion
Present	31 (55.4)
Absent	25 (44.6)
Lymph node metastasis
Present	9 (16.1)
Absent	47 (83.9)
Distant metastasis
Present	6 (10.7)
Absent	50 (89.3)
TNM Stage[a]
I–II	44 (78.6)
III–IV	12 (21.4)
Tumor regression grade[b]
1a	17 (30.4)
1b	18 (32.1)
2	21 (37.5)

SD, standard deviation; Pap, Papillary adenocarcinoma; Well, Well differentiated adenocarcinoma; Mod, Moderately differentiated adenocarcinoma; Por, Poorly differentiated adenocarcinoma; Muc, Mucinous adenocarcinoma; TNM, tumor node metastasis

According to TMN classification of malignant tumors, 7th edition according to UICC.

According to Japanese Classification of Colorectal Carcinoma, 8 th edition.

LGR5 and CD133 expressions

LGR5 expression was detected in the cytoplasm while CD133 expression was detected at the cell membrane and the intraglandular cellular debris in some tumor glands (Fig. 1). The staining pattern was mostly similar to that previously reported. As a result of the immunohistochemistry scoring, 36 (64.3%) and 29 (51.8%) patients were categorized as LGR5-positive and CD133-positive, respectively. Twenty-one patients (37.5%) showed positive expressions of both markers.

Figure 1.

Immunohistochemical detection of leucine-rich repeat-containing G protein-coupled receptor 5 (LGR5) and cluster of differentiation-133 (CD133) in low rectal cancer. (A) Immunohistochemical staining of LGR5 (original magnification ×100). LGR5 expression was detected in the cytoplasm. (B) Immunohistochemical staining of CD133 (original magnification ×100). CD133 expression was detected on the luminal cell surface of the rectal tumor glands and on the intraglandular cellular debris. (C) Hematoxylin and eosin staining of a serial section of a primary lesion (original magnification ×100).

Relationships between LGR5 expression and clinicopathological features

LGR5 expression significantly correlated with lymphatic invasion (P=0.03), lymph node metastasis (P<0.01), and TNM stage (P<0.01) (Table II). No significant difference in OS or DFS was found [LGR5 positive vs. negative: 5-year OS 85.0% vs. 85.0% (P=0.86), 5-year DFS 63.9% vs. 70.0% (P=0.67), respectively].

Table II.

LGR5 expression and clinicopathological features.

	LGR5+ (n=36)		LGR5- (n=20)
Characteristic	n	%	n	%	P-value
Gender
Male	25	69.4	12	60.0	0.48
Female	11	37.9	8	29.6
Mean age ± SD (years)	60.2±11.1		62.8±7.1		0.51
pT stage
T1-2	15	42.0	6	30.0	0.38
T3-4	21	58.0	14	70.0
Histological type
Pap, Well	24	66.7	15	75.0	0.51
Mod, Por, Muc	12	33.3	5	25.0
Lymphatic invasion
Present	5	13.9	0	0.0	0.03
Absent	31	86.1	20	100.0
Vascular invasion
Present	19	52.8	12	60.0	0.60
Absent	17	47.2	8	40.0
Lymph node metastasis
Present	9	25.0	0	0.0	<0.01
Absent	27	75.0	20	100.0
Distant metastasis
Present	5	13.9	1	5.0	0.28
Absent	31	86.1	19	95.0
TNM Stage
I–II	24	66.7	20	100	<0.01
III–IV	12	33.3	0	0
Tumor regression grade[a]
1a	11	30.6	6	30.0	0.94
1b	11	30.6	7	35.0
2	14	38.9	7	35.0

SD, standard deviation; Pap, Papillary adenocarcinoma; Well, Well differentiated adenocarcinoma; Mod, Moderately differentiated adenocarcinoma; Por, Poorly differentiated adenocarcinoma; Muc, Mucinous adenocarcinoma; LRG5, Leucine-rich repeat-containing G protein-coupled receptor 5, TNM, tumor node metastasis

According to Japanese Classification of Colorectal Carcinoma, 8 th edition.

Relationship between CD133 expression and clinicopathological features

CD133 expression significantly correlated with vascular invasion (P<0.01) and the tumor regression grade (P<0.01) (Table III). No significant difference in OS or DFS was found [CD133 positive vs. negative, 5-year OS 80.9% vs. 88.9% (P=0.40), 5-year DFS 58.6% vs. 73.9% (P=0.19), respectively].

Table III.

CD133 expression and clinicopathological features.

	CD133+ (n=29)		CD133- (n=27)
Characteristic	n	%	n	%	P-value
Gender
Male	18	62.1	19	70.4	0.51
Female	11	37.9	8	28.9
Mean age ± SD (years)	61.9±10.2		60.3±9.6		0.48
pT stage
T1-2	8	27.6	13	48.1	0.11
T3-4	21	72.4	14	51.9
Histological type
Pap, Well	19	65.5	20	74.1	0.49
Mod, Por, Muc	10	34.5	7	25.9
Lymphatic invasion
Present	4	13.8	1	3.7	0.17
Absent	25	86.2	26	96.3
Vascular invasion
Present	21	72.4	10	37.0	<0.01
Absent	8	27.6	17	63.0
Lymph node metastasis
Present	6	20.7	3	11.2	0.32
Absent	23	79.3	24	88.9
Distant metastasis
Present	4	13.8	2	7.4	0.44
Absent	25	86.2	25	92.6
TNM Stage
I–II	20	69	24	88.9	0.06
III–IV	9	31	3	11.1
Tumor regression grade[a]
1a	14	48.3	3	11.1	<0.01
1b	11	37.9	7	25.9
2	4	13.8	17	63.0

According to Japanese Classification of Colorectal Carcinoma, 8 th edition. SD, standard deviation; Pap, Papillary adenocarcinoma, Well, Well differentiated adenocarcinoma; Mod, Moderately differentiated adenocarcinoma; Por, Poorly differentiated adenocarcinoma; Muc, Mucinous adenocarcinoma; CD, cluster of differentiation; TNM, tumor node metastasis.

Relationships between combined LGR5 and CD133 expression and clinicopathological features

No significant correlation was found between the expressions of the two markers (Table IV). The patients were divided into groups of positive expression of both markers and of all other patients, and the relationships between combined expression and the clinicopathological features were analyzed (Table V). Expression of both LGR5 and CD133 significantly correlated with lymphatic invasion (P=0.04), the tumor regression grade (P<0.01), and TNM stage (P<0.01). No significant difference in OS or DFS was found [LGR5 and CD133 positive vs. LGR5 or CD133 negative: 5-year OS 78.1% vs. 85.0% (P=0.53), 5-year DFS 57.1% vs. 71.3% (P=0.25), respectively].

Table IV.

Correlation between LGR5 and CD133 expression.

	CD133 expression
LGR5 expression	+(%)	− (%)	P-value
+	21 (37.5)	15 (26.8)	0.19
−	8 (14.3)	12 (21.4)

LRG5, Leucine-rich repeat-containing G protein-coupled receptor 5; CD, cluster of differentiation.

Table V.

LGR5 and CD133 expression and clinicopathological features.

	LGR5+ and CD133+ (n=21)		LGR- or CD133- (n=35)
Characteristic	n	%	n	%	P-value
Gender
Male	14	66.7	23	65.7	0.94
Female	7	33.3	12	34.3
Mean age ± SD (years)	61.4±11.0		61.0±9.2		0.82
pT stage
T1-2	6	28.6	15	42.9	0.28
T3-4	15	71.4	20	57.1
Histological type
Pap, Well	12	57.1	27	77.1	0.12
Mod, Por, Muc	9	42.9	8	22.9
Lymphatic invasion
Present	4	19.0	1	2.9	0.04
Absent	17	81.0	34	97.1
Vascular invasion
Present	15	71.4	16	45.7	0.06
Absent	6	28.6	19	54.3
Lymph node metastasis
Present	6	28.6	3	8.6	0.05
Absent	15	71.4	32	91.4
Distant metastasis
Present	4	19.0	2	5.7	0.13
Absent	17	81.0	33	94.3
TNM Stage
I–II	12	57.1	32	91.4	<0.01
III–IV	9	42.9	3	8.6
Tumor regression grade[a]
1a	11	52.4	6	17.1	<0.01
1b	7	33.3	11	31.4
2	3	14.3	18	51.4

According to Japanese Classification of Colorectal Carcinoma, 8 th edition. SD, standard deviation; Pap, Papillary adenocarcinoma; Well, Well differentiated adenocarcinoma; Mod, Moderately differentiated adenocarcinoma; Por, Poorly differentiated adenocarcinoma; Muc, Mucinous adenocarcinoma; TNM, tumor node metastasis; LRG5, Leucine-rich repeat-containing G protein-coupled receptor 5; CD, cluster of differentiation.

Discussion

Herein, we examined the expressions of LGR5 and CD133 in low rectal cancer patients treated with preoperative CRT immunohistochemically, using serial sections. LGR5 expression was associated with lymphatic invasion, lymph node metastasis, and TNM stage, while CD133 expression was associated with vascular invasion. Although CD133 expression correlated with the response to CRT, LGR5 expression did not. Moreover, the co-expression of LGR5 and CD133 significantly correlated with lymphatic invasion, the tumor regression grade, and TNM stage. No significant correlation with OS or DFS was found.

We evaluated the expressions of LGR5 and CD133 in low rectal cancer using surgically resected specimens after CRT, since the assessment of very small biopsy samples obtained pre-CRT is difficult. There are several reports evaluating the expressions of these markers, separately, post-CRT. Saigusa et al (28) investigated LGR5 expression in rectal cancer patients treated with preoperative CRT and revealed a correlation with patient prognosis, while no correlation with any clinicopathological factor or prognosis was observed in CRC patients treated without preoperative therapy. They also described that CSCs were relatively increasing after CRT, because CSCs are resistant to CRT, as compared to non-CSCs (28). Kawamoto et al (25) evaluated CD133 expression in rectal cancer patients treated with preoperative CRT and described that, while CD133 expression in pre-CRT specimens before CRT did not associate with prognosis, those obtained post-CRT associated with DFS.

Regarding clinicopathological factors, prior studies in CRC patients have demonstrated correlations between LGR5 and CD133 expressions and tumor depth, lymph node metastasis, lymphatic invasion, vascular invasion, and advanced tumor stage (14–16,21,23,24). However, the precise mechanisms of these correlations are still under investigation. Hirsch et al showed that silencing of LGR5 reduced proliferation, migration, and colony formation in CRC cell lines (29). The correlation between LGR5 and tumor proliferation in thyroid cancers has also been reported (12). Carmon et al (8) demonstrated that LGR5 enhanced cell proliferation through Wnt/β-catenin signaling (8), and LGR5 has been reported to promote epithelial-mesenchymal transition through activation of Wnt/β-catenin signaling in breast cancer (13). Moreover, Xi et al (10) showed a correlation between LGR5 expression and matrix metalloproteinase-2 in gastric cancer, and declared that LGR5 played an important role in tumor invasion and metastasis through matrix metalloproteinase-2, which degrades type IV collagen of the extracellular matrix and basal membrane. With respect to CD133, Chao et al (30) reported that CD133+ colon cancer cells had an enhanced ability to interact with neighboring fibroblasts, indicating that CD133+ cells are more invasive than CD133-cells. In lung and pancreatic cancers, CD133 expression has been shown to relate to vasculogenesis (27) and to be an important factor for epithelial-mesenchymal transition (31). Our findings were consistent with those of these previous studies.

Regarding the response to CRT, Saigusa et al reported that patients with high LGR5 expression had a poor pathological response to CRT (17). Hongo et al (24) reported that high CD133 expression correlated with worse response to CRT. Although the mechanisms remain unclear, there are several reports indicating resistance to chemotherapy and radiotherapy of CSCs. LGR5 expression has been reported to increase after radiation in CRC cell lines (28), and LGR5-positive CRC cells are reportedly resistant to chemotherapy through the ATP-binding cassette family (32). Moreover, Bao et al (5) demonstrated that CD133-positive glioma cells showed reduced sensitivity to radiation through activation of DNA damage repair. In CRC cell lines, CD133 expression has been reported to increase after irradiation (33), and CD133-positive CRC cells show reduced sensitivity to chemotherapy (34).

Herein, we could not show the correlation between LGR5 expression and the response to CRT, although CD133 expression was associated with a poor response. This can be attributed to several factors. First, the role of LGR5 in chemotherapy resistance is controversial, although Planutis et al (35) reported that LGR5-expressing CRC cells were more sensitive to anticancer drugs. Second, single irradiation in vitro is different from multiple courses of irradiation in clinical settings. Third, CRT regimens are not identical with respect to the concomitant chemotherapy and frequency of radiotherapy. Finally, our study had a relatively small sample size, which may have affected the statistical power. Regarding the patient prognosis, prior studies, including several meta-analyses, have demonstrated that these markers are poor prognostic factors (15,16,22,23). However, we could not show these correlations due to the small sample size.

Several putative cancer stem cell markers, including LGR5 and CD133, have been reported, but their clinical significance is not fully understood. Because of the heterogeneity of CRC, a single marker for CSCs may not be sufficient for precise evaluation. To our knowledge, there is no previous report evaluating both LGR5 and CD133 expression in specimens of CRC patients. In CRC cell lines, Kobayashi et al reported that LGR5+/CD133+ and LGR5-/CD133+ cells formed colonies, while almost all LGR5-/CD133-cells could not; their results moreover showed that the colony formation of LGR5+/CD133+ cells was higher than that of LGR5-/CD133+ cells (36). Therefore, we suppose that the subgroup with both LGR5 and CD133 expression are at high risk of malignancy. Accordingly, patients with both LGR5 and CD133 expression had significantly more lymphatic invasion, worse tumor regression grade and more advanced TNM stage, and tended to have more vascular invasion and lymph node metastasis. Although more correlations between the combined expression and clinicopathological factors were found than in the analysis of the single markers, further studies are necessary.

Our data should be interpreted with caution, as this study has several limitations. First, this was a retrospective study with a relatively small patient number. Second, we investigated protein expression using immunohistochemistry, but did not investigate the gene expression. Third, there are different methods of evaluating positive immunohistochemical staining of LGR5 and CD133. Therefore, prospective studies with a larger number of patients and the same methodology are needed.

We here demonstrated that the expression of LGR5 was associated with lymphatic invasion, lymph node metastasis, and TNM stage, while CD133 associated with the response to CRT in low rectal cancer. Moreover, combined expression of LGR5 and CD133 may be a more useful marker than the evaluation of a single marker.