Elevated expression of Rad51 is correlated with decreased survival in resectable esophageal squamous...

Journal of Surgical Oncology 2011;104:617–622

Elevated Expression of Rad51 is Correlated With Decreased Survival in

Resectable Esophageal Squamous Cell Carcinoma

YONG LI, MD,1,2 HUI YU, MD,1,2 RONG-ZHEN LUO, MD,3 YING ZHANG, MD,2 MEI-FANG ZHANG, MD,3

XIN WANG, MD,1,2* AND WEI-HUA JIA, PhD2**

1Department of Thoracic Surgery, Sun Yat-Sen University Cancer Center, Guangzhou City, Guangdong Province, P.R. China2State Key Laboratory of Oncology in South China, Guangzhou City, Guangdong Province, P.R. China

3Department of Pathology, Sun Yat-Sen University Cancer Center, Guangzhou City, Guangdong Province, P.R. China

Background: Rad51 plays a critical role in homologous recombination and correlates with many human malignancies. However, its role and

clinical significance in esophageal squamous cell carcinoma (ESCC) has not been clarified. The purpose of this study was to explore the

clinicopathological correlation and prognostic significance of Rad51expression in a group of ESCC patients.

Methods: We evaluated Rad51 expression in 230 surgically resected ESCC specimens by immunochemistry using tissue microarray and

correlated with clinicopathological features including post-operation survival.

Results: There was no significant correlation between Rad51 expression and clinicopathological characteristics in terms of age, sex, tumor

location, histologic grade, T, N categories, and TNM stage. The Kaplan–Meier survival analysis showed that high expression of Rad51

indicated a poorer disease free survival (DFS) of ESCC patients compared with the patients with low expression of Rad51 (P ¼ 0.031),

similar result also shown in overall survival (OS) analysis (P ¼ 0.034). Furthermore, Rad51 expression could stratify node positive patients

in DFS (P ¼ 0.021) and OS (P ¼ 0.035). Multivariate analysis confirmed the Rad51 expression was an independent prognostic factor for

DFS (HR ¼ 1.603, P ¼ 0.013) and OS (HR ¼ 1.555, P ¼ 0.021) of ESCC patients.

Conclusions: Elevated expression of Rad51 is associated with poor prognosis for resectable ESCC patients.

J. Surg. Oncol. 2011;104:617–622. � 2011 Wiley Periodicals, Inc.

KEY WORDS: prognosis; surgery; biomarker

INTRODUCTION

Esophageal squamous cell carcinoma (ESCC), the predominant

histopathological type in East Asian countries, is one of the most

lethal malignant diseases [1,2]. In China, ESCC has ranked the

fourth of cancer-related mortality [3]. Despite recent advances in ear-

ly diagnosis and treatment have led to improved survival, the clinical

outcome and the overall survival of ESCC patients stands poor [4].

Such pessimistic outcome indicates an urgent requirement to

improve our understanding of the molecular mechanism underlying

ESCC, which may lead to the development of improved prognostic

and predictive assays. Thus, discovering specific biomarkers will

probably aid to identify the patients with increased risk for recur-

rence or metastasis, providing more sufficient information for design-

ing appropriate patient-tailored therapy strategies.

Rad51, an evolutional conserved enzyme encoded by the Rad51

gene located on human chromosome 15q15.1, is essential for homol-

ogous recombination [5]. Rad51 expression is tightly controlled in

normal cells as inappropriate recombination can lead to genomic

instability [6,7]. However, overexpression of Rad51 is observed in

the majority of human tumor cells [8–10], and leads to increases

in genomic instability [11,12] and resistance to DNA double-strand

break inducing cancer therapies [13,14]. Furthermore, Rad51 overex-

pression correlates with poor prognosis in many cancers, such as soft

tissue sarcoma [14], breast cancer [15], non-small-cell lung cancer

[16]; however, the status of Rad51 and its clinical significance in

ESCC has not been elucidated.

Therefore, we performed this study to measure Rad51 expression

in 230 resected ESCC specimens by immunochemistry using tissue

microarrays. The correlation between Rad51 expression and patient

clinical/prognostic parameters was evaluated to determine whether

expression of Rad51 would be a prognostic biomarker for ESCC

patients.

MATERIALS AND METHODS

Patient Selection

This study was approved by the medical ethics committee of

Sun Yat-Sen University Cancer Center. Two hundred and sixty-five

primary ESCC patients treated with surgery were eligible from the

Department of Thoracic surgery, Cancer Center, Sun Yat-Sen Univer-

sity between October 2000 and April 2007. Tumor histological grade

and clinical stage were defined according to the 7th edition of

the TNM classification of the International Union Against Cancer

(UICC 2009) [17]. Cases selected in this study were fulfilled with

the following criteria: (a) Newly diagnosed cancer of esophagus

without receiving previous treatment; (b) histologically confirmed

primary thoracic ESCC; (c) no distant metastases including supracla-

vicular or celiac lymph nodes metastases; (d) underwent complete

surgical resection (R0) in Sun Yat-Sen University Cancer Center;

(e) adequate clinical information and follow-up data. Patients with

non-curative resection (R1) and died of postoperative complication

were excluded from this study. Patients with neoadjuvant and adju-

vant therapy were also excluded. In our cancer center, patients with

*Correspondence to: Dr. Xin Wang, MD, Department of ThoracicSurgery, Sun Yat-Sen University Cancer Center, No. 651, Dongfeng EastRoad, 510060, Guangzhou City, Guangdong Province, P.R. China.Fax: þ86-20-87343596. E-mail: [email protected]

**Correspondence to: Prof. Wei-Hua Jia, PhD, State Key Laboratory ofOncology in South China, Guangzhou City, Guangdong Province, P.R.China. Fax: þ86-20-87343392. E-mail: [email protected]

Received 17 February 2011; Accepted 16 June 2011

DOI 10.1002/jso.22018

Published online 8 July 2011 in Wiley Online Library(wileyonlinelibrary.com).

� 2011 Wiley Periodicals, Inc.

ESCC that invade the airway or major vessels (such as thoracic

aorta) or accompanied by visceral metastasis are not indicated for

surgery. The surgical approach employed was left trans-thoracic

approach or two-field dissection approach. The tumor specimens and

adjacent normal samples were recruited from paraffin blocks at the

Bank of Tumor Source in our cancer center.

Clinical data was abstracted from hospital records after surgery.

All patients were re-contacted in May 2010 to ascertain vital status.

Tissue Microarray (TMA) Construction

Tumor tissue samples from 265 ESCC cases were collected, fixed

in formalin, and embedded in paraffin. H&E-stained sections from a

single random block from each patient were reviewed by a senior

pathologist (R-Z Luo) to define representative tumor regions. Two

targeted core samples of each specimen were obtained using a tissue

arraying instrument (ELEPHELYS Minicore instruments, France).

Briefly, tissue cylinders with a diameter of 10 mm were punched and

arrayed on a recipient paraffin block. Sections (5-mm) of the tissue

array (recipient) block were cut and placed on class slides. After

exclusion of cores with inadequate tissue following sectioning and

tissue transfer, the final immunohistochemical analyses included

cores from 230 ESCC cases. Each of the 230 different ESCC cases

contributed to the biomarker analyses. Among the 230 cases,

51 adjacent normal esophageal mucosa matched was served as

controls. The microarray for the normal esophageal mucosa was

constructed by the same method described above.

Immunohistochemistry (IHC)

Staining and Assessment

IHC staining was performed on TMA sections rehydrated through

graded alcohols. Endogenous peroxidase activity was blocked with

0.3% hydrogen peroxide for 15 min. For antigen retrieval, TMA

Fig. 1. Rad51 expression by immunohistochemical staining. A, B: Normal esophageal mucosa showed low expression of Rad51 proteinin nuclei of all esophageal squamous cells (magnification A �40, B �200). C, D: ESCC case demonstrated low expression of Rad51(magnification C �40, D �200). E, F: High expression of Rad51 was detected in an ESCC (magnification E �40, F �200).

618 Li et al.

Journal of Surgical Oncology

slides were boiled in tris (hydroxymethyl) aminomethane-EDTA

buffer (pH 8.0) in a pressure cooker for 20 min. Non-specific binding

was blocked with 10% normal goat serum for 20 min. The TMA

slides were incubated with anti-Rad51 (a rabbit polyclonal antibody

to RAD51L (ab69622, 1:200 dilution, Abcam)) for12 hr at 48Cin a moist chamber. Subsequently, the slides were sequentially

incubated with biotinylated rabbit anti-mouse immune-globulin at a

concentration of 1:100 for 30 min at 378C and then reacted with

a streptavidin–peroxidase conjugate for 30 min at 378C and 30-30diaminobenzidine as a chromogen substrate. The nucleus was

counterstained using Meyer’s hematoxylin. A negative control was

obtained by replacing the primary antibody with a normal rabbit

IgG. Positive expression of Rad51 in ESCC and normal esophageal

mucosa cells was primarily a nuclear pattern (Fig. 1). Internal posi-

tive and negative controls, including normal squamous mucosa of the

esophagus from non-cancer patients were utilized as available to fur-

ther support the staining patterns. Two independent observers (R-Z

Luo and M-F Zhang) blinded to the clinicopathological information

performed the scoring of Rad51 expression using the previously vali-

dated scoring system. In this study, IHC staining was analyzed for

the proportion of positively stained tumor cells using the concept of

positive cell index (PCI) [18]. At least 200 tumor cells were counted

to calculate the score for each specimen on TMAs. The PCI ranged

from 0 to 90%. The average of PCI of the two pathologists was

assigned as the final PCI for each case. Rad51 expression was classi-

fied into two groups: Low expression, PCI �10%; high expression,

PCI >10% [15,16].

Statistical Analysis

All statistical analyses were performed with SPSS software pack-

age (SPSS Standard version 16.0, SPSS, Chicago, IL). The correla-

tion between Rad51 expression and clinicopathological parameters

was assessed by the Pearson-Chi-square test. Disease-free survival

(DFS) was defined as the time from surgery to regional relapse or

distant metastasis. Overall survival (OS) was defined as the time

from surgery to death. DFS and OS were assessed with Kaplan–

Meier method and compared by the log-rank test. Multivariate sur-

vival analysis was performed on all parameters that were found to be

significant on univariate analysis using the Cox regression model.

For all analyses, P-value < 0.05 was assumed as statistically

significant.

RESULTS

Patient Characteristics

Sixty-five females and 165 males were enrolled, aged from 32 to

80 (median 58.0 years). The clinicopathological features of the

230 patients included in this study are listed in Table I.

Expression of Rad51 in ESCC

In the present study, for Rad51staining in ESCC tissue and

normal esophageal mucosa, immunoreactivity was detected primarily

in nuclei within tumor cell. Protein expression of Rad51 was exam-

ined by IHC in 230 cases of primary ESCC and in 51 cases of

paracancerous esophageal mucosa. Using the criteria described

above (PCI > 10%), high expression of Rad51 was observed

in 46.5% (107/230) of ESCC. The association between clinicopatho-

logical parameters and Rad51 expression was summarized in Table I.

No statistically significant association was observed between Rad51

expression and age, sex, tumor location, surgery, histologic grade,

T categories, N categories, and TNM stage.

Rad51 Expression and Survival in ESCC

Of the 230 ESCC patients, none was lost to follow-up. The medi-

an observation period was 70.1months (4–115 months), with 112

died and 118 alive at the end of follow-up. The 5-year DFS and OS

for the entire cohort were 52.4 and 54.0%, respectively, with a medi-

an survival of 67.9 and 70.1 months.

There was statistically significant overall survival difference be-

tween low Rad51 expression and high expression patients (mean

70.2 months vs. 61.6 months, P ¼ 0.034, Fig. 2, Table II). A similar

result was also observed in DFS (mean 67.7 months vs. 58.4 months,

P ¼ 0.031, Fig. 2, Table II). In subgroup analysis, expression of

Rad51 can well distinguish DFS/OS for pathological stage N1-3

patients (Fig. 2, P ¼ 0.021/0.035) but not for pathological stage N0

patients (Fig. 2, P ¼ 0.241/0.227).

Univariate analysis using Cox’s proportional hazard model

showed that the following parameters correlated significantly with

DFS and OS: T categories, N categories, and Rad51 expression

(Table III). When the above parameters were entered into multivari-

ate analysis, the results suggested that T categories, N categories,

and Rad51 expression were the independent factors that impact the

overall survival (Table III).

DISCUSSION

The recombination protein Rad51 is essential in repairing DNA

double-strand break by homolog recombination [19]. Several studies

have documented an involvement of Rad51 in oncogenic processes,

such as soft tissue sarcoma [14], breast cancer [15], lung cancer

TABLE I. Rad51 Expression and Clinicopathological Variables

(Chi-Square Test)

Variables Case

RAD51 Expression (%)

Low High P-value

Age(years)

�60 139 75(56.0) 64(46.0) 0.732

>60 91 47(51.6) 44(48.4)

Sex

Male 65 36(55.4) 29(44.6) 0.655

Female 165 86(52.1) 79(47.9)

Surgery

Standard 140 70(50.0) 70(50.0) 0.249

Three-incision 90 52(57.8) 38(42.2)

Tumor location

Upper 11 7(63.6) 4(36.4) 0.732

Middle 160 86(53.8) 74(46.2)

Lower 59 30(50.8) 29(49.2)

Histologic grade

G1 52 29(55.8) 23(44.2) 0.135

G2 152 84(55.3) 68(44.7)

G3 26 9(34.6) 17(65.4)

pT categories

1 4 3(75%) 1(25%) 0.794

2 55 29(52.7) 26(47.3)

3 168 88(52.4) 80(47.6)

4 3 2(66.7) 1(33.3)

pN categories

0 123 64(52.0) 59(48.0) 0.742

1–3 107 58(54.2) 49(45.8)

AJCC pathological stage

I 7 5(71.4) 2(28.6) 0.502

II 134 68(50.7) 66(49.3)

III 89 49(55.1) 40(44.9)

AJCC, American Joint Committee on Cancer.

RAD51 in Esophageal Carcinoma 619


[16], and prostate cancer [20]. However, the status of Rad51 and its

potential prognostic value in ESCC remains unclear. In this study,

we examined the expression of Rad51 in ESCC tissue and adjacent

normal mucosa, firstly by IHC. Although there is no relationship

seen between Rad51 expression and clinicopathological parameters,

our results showed that a significant percentage of ESCC mucosa

exhibited overexpression of Rad51 in tumor nuclei compared to nor-

mal mucosa tissue. In all, a total of 46.5% (107/230) ESCC patients

showed high expression of Rad51. The same trend was also de-

scribed in other human malignancies such as pancreatic adenocarci-

noma [9], breast cancer [14], non-small-cell lung cancer [15],

prostate cancer [20], and head and neck cancer [21]. These findings

may imply that up-regulation of Rad51 expression appears to be

involved in the tumorigenic process of ESCC.

It is well-known that surgical resection could be considered as the

major therapeutic strategy employed for patients with local ESCC,

and the prognosis of ESCC was heavily dependent on the TNM clas-

sification. However, individual ESCC exhibits a diverse spectrum of

outcome to the same therapy, even with the same clinicopathological

TNM stage. Therefore, finding new objective strategies for predicting

the outcome to surgery is urgently needed. In current study, the data

revealed that decreased expression of Rad51 showed prolonged dis-

ease-free survival and overall survival and that elevated expression

of Rad51 may be an increased risk factor for the survival of ESCC

patients. Importantly, our results showed that high expression of

Rad51 was an independent prognostic predictor of inferior disease-

free survival and overall survival as evidenced by multivariate sur-

vival analysis. Taken together, our findings support the ideas that

Fig. 2. Disease-free survival (DFS) and Overall survival (OS) curves for esophageal squamous cell carcinoma patients according to Rad51expression status. a, b: DFS and OS curve: Patients with low expression and high expression of Rad51. c, d: DFS and OS curve: Patientswith low and high expression of Rad51 at stage N1-3.

620 Li et al.


Rad51 expression has the potential to predict the outcome of opera-

ble ESCC patients. The evaluation of Rad51 expression may be a

choice for identifying those patients who were susceptible to experi-

ence recurrence and progression.

These findings address the issue about the role of Rad51 in the

development of ESCC. With respect to the molecular mechanism of

Rad51, elevated Rad51 expression protects tumor cells from apopto-

sis and prolongs their survival by decreasing DNA damage [22].

Elimination of Rad51 in vitro results in sensitivity towards ionizing

radiation and early embryonic lethality [23]. Additionally, there is

evidence that inhibition of Rad51 has been explored as a way to

sensitize cancer cells to radiotherapy and target therapy [24–27].

These data may suggest that Rad51 is indispensible for tumor cell

survival, presumably owed to its role in repairing DNA double-strand

break and increasing genomic instability. Our data from ESCC and

results from breast cancer, pancreatic cancer, soft tissue sarcoma,

and non-small lung cancer [9,14–16] strongly suggest that excessive

expression of Rad51 might promote tumor progression, probably it

helps to enhance genetic instability and at the same time keeping

DNA damage at a tolerable extent for cell survival.

As we know, the adjuvant therapy for ESCC is controversial at

present. Our results showed that Rad51 expression could stratify

DFS and OS with patients in stage N1-3, but not with N0 stage. This

result may due to the relatively favorable prognosis for patients at

N0 stage, which may not exhibit the significance of changing genetic

instability of Rad51 by the relative longer survival at this stage.

Thus, the determination of Rad51 expression by IHC could be used

an auxiliary tool for selecting those patients may be accepted adju-

vant therapy after surgical resection.

There are some limitations to generate these results. Firstly, the

samples are from single institution, and the number may not large

enough for subgroup analysis. Secondly, this is a retrospective study.

Thirdly, there is limited information on adjuvant therapy of the

patients to draw a conclusion about the potential role of Rad51 ex-

pression as a predictor of response to chemotherapy or radiotherapy.

In summary, our study suggested the role of Rad51 on the prog-

nosis of ESCC patients treated with surgical resection, and mean-

while, the expression of Rad51, as detected by IHC, may serve as a

new molecular marker for prognosis of ESCC patients treated with

surgical resection. Further studies are warranted to validate our

results.

ACKNOWLEDGMENTS

This work was supported by National ‘‘863’’ key project

(2009AA022706).

TABLE III. Multivariate Cox Regression Analysis for Survival in ESCC

Variable

Disease free survival Overall survival

RR 95% CI P-value RR 95% CI P-value

T status 1.487 0.967–2.287 0.071 1.578 1.029–2.420 0.037

N status 3.822 2.538–5.754 <0.001 3.991 2.648–6.015 <0.001

RAD51expression 1.603 1.103–2.330 0.013 1.555 1.070–2.259 0.021

RR, relative risk; CI, confidence interval.

TABLE II. Rad51 Expression in ESCC Patients by Kaplan–Meier Survival Analysis

Variables Case

OS (months) DFS (months)

Mean Median P-value Mean Median P-value

Total

Low expression 123 67.7 NR 0.031 70.2 NR 0.034

High expression 107 58.4 35.0 61.6 53.0

pT1-2


High expression 27 60.9 70.0 64.3 68.0

pT3-4


High expression 80 57.0 30.0 60.2 53.0

pN0


High expression 58 81.9 NR 83.0 NR

pN1-3

Low expression 58 43.4 26.0 0.021 47.1 40.0 0.035

High expression 49 25.2 17.0 31.6 23.0

Histologic grade

G1


High expression 23 60.1 62.0 64.8 68.0

G2-3


High expression 84 57.7 34.0 60.5 41.0

DFS, disease-free survival; OS, overall survival.

RAD51 in Esophageal Carcinoma 621


REFERENCES

1. Parkin DM, Bray F, Ferlay J, et al.: Global cancer statistics,2002. CA Cancer J Clin 2005;55:74–108.

2. Shimada H, Nabeya Y, Okazumi S, et al.: Prediction of survivalwith squamous cell carcinoma antigen in patients with resect-able esophageal squamous cell carcinoma. Surgery 2003;133:486–494.

3. Su M, Lu SM, Tian DP, et al.: Relationship between ABO bloodgroups and carcinoma of esophagus and cardia in Chaoshaninhabitants of China. World J Gastroenterol 2001;7:657–661.

4. Munro AJ: Oesophageal cancer: A view over overviews. Lancet2004;364:566–568.

5. Scully R, Chen J, Plug A, et al.: Association of BRCA1 withRad51 in mitotic and meiotic cells. Cell 1997;88:265–275.

6. Richardson C: RAD51, genomic stability, and tumorigenesis.Cancer Lett 2005;218:127–139.

7. Thacker J: The RAD51 gene family, genetic instability and can-cer. Cancer Lett 2005;219:125–135.

8. Raderschall E, Stout K, Freier S, et al.: Elevated levels ofRad51 recombination protein in tumor cells. Cancer Res 2002;62:219–225.

9. Maacke H, Jost K, Opitz S, et al.: DNA repair and recombina-tion factor Rad51 is over-expressed in human pancreatic adeno-carcinoma. Oncogene 2000;19:2791–2795.

10. Xia SJ, Shammas MA, Shmookler Reis RJ: Elevated recombi-nation in immortal human cells is mediated by HsRAD51recombinase. Mol Cell Biol 1997;17:7151–7158.

11. Arnaudeau C, Rozier L, Cazaux C, et al.: RAD51 supportsspontaneous non-homologous recombination in mammaliancells, but not the corresponding process induced by topoisomer-ase inhibitors. Nucleic Acids Res 2001;29:662–667.

12. Richardson C, Stark JM, Ommundsen M, et al.: Rad51 over-expression promotes alternative double-strand break repairpathways and genome instability. Oncogene 2004;23:546–553.

13. Hannay JA, Liu J, Zhu QS, et al.: Rad51 overexpression con-tributes to chemoresistance in human soft tissue sarcoma cells:A role for p53/activator protein 2 transcriptional regulation. MolCancer Ther 2007;6:1650–1660.

14. Collis SJ, Tighe A, Scott SD, et al.: Ribozyme minigene-mediated RAD51 down-regulation increases radiosensitivity of

human prostate cancer cells. Nucleic Acids Res 2001;29:1534–1538.

15. Maacke H, Opitz S, Jost K, et al.: Over-expression of wild-typeRad51 correlates with histological grading of invasive ductalbreast cancer. Int J Cancer 2000;88:907–913.

16. Qiao GB, Wu YL, Yang XN, et al.: High-level expression ofRad51 is an independent prognostic marker of survival in non-small-cell lung cancer patients. Br J Cancer 2005;93:137–143.

17. Rice TW, Blackstone EH, Rusch VW: 7th edition of the AJCCCancer Staging Manual: Esophagus and esophagogastricjunction. Ann Surg Oncol 2010;17:1721–1724.

18. Remmele W, Stegner HE: [Recommendation for uniform defini-tion of an immunoreactive score (IRS) for immunohisto-chemical estrogen receptor detection (ER-ICA) in breast cancertissue]. Der Pathologe 1987;8:138–140.

19. Baumann P, West SC: Role of the human RAD51 protein inhomologous recombination and double-stranded-break repair.Trends Biochem Sci 1998;23:247–251.

20. Mitra A, Jameson C, Barbachano Y, et al.: Overexpression ofRAD51 occurs in aggressive prostatic cancer. Histopathology2009;55:696–704.

21. Connell PP, Jayathilaka K, Haraf DJ, et al.: Pilot study examin-ing tumor expression of RAD51 and clinical outcomes in humanhead cancers. Int J Oncol 2006;28:1113–1119.

22. Henning W, Sturzbecher HW: Homologous recombination andcell cycle checkpoints: Rad51 in tumour progression and thera-py resistance. Toxicology 2003;193:91–109.

23. Tsuzuki T, Fujii Y, Sakumi K, et al.: Targeted disruption of theRad51 gene leads to lethality in embryonic mice. Proc NatlAcad Sci USA 1996;93:6236–6240.

24. Russell JS, Brady K, Burgan WE, et al.: Gleevec-mediated inhi-bition of Rad51 expression and enhancement of tumor cellradiosensitivity. Cancer Res 2003;63:7377–7383.

25. Ohnishi T, Taki T, Hiraga S, et al.: In vitro and in vivo potentia-tion of radiosensitivity of malignant gliomas by antisense inhi-bition of the RAD51 gene. Biochem Biophys Res Commun1998;245:319–324.

26. Collis SJ, DeWeese TL, Jeggo PA, et al.: The life and death ofDNA-PK. Oncogene 2005;24:949–961.

27. Hine CM, Seluanov A, Gorbunova V: Use of the Rad51 pro-moter for targeted anti-cancer therapy. Proc Natl Acad Sci USA2008;105:20810–20815.

622 Li et al.


Elevated expression of Rad51 is correlated with decreased survival in resectable esophageal squamous...

Documents

Transcript of Elevated expression of Rad51 is correlated with decreased survival in resectable esophageal squamous...