Role of KIT expression in the prognosis of clear cell renal cell carcinomas in Chinese patients

Abstract

Objectives

To investigate the expression of KIT in clear cell renal cell carcinomas, and to reveal the relationships between KIT status and clinicopathological features and survival of clear cell renal cell carcinomas.

Methods

The expression of KIT was tested immunohistochemically in 119 specimens of clear cell renal cell carcinoma. Their correlations to clinicopathological parameters were compared and discussed. Kaplan–Meier method was used to determine the survival between KIT-positive and KIT-negative patients. Multivariate analysis was performed using the Cox-regression model for overall survival.

Results

A total of 13 out of 119 cases of clear cell renal cell carcinomas (10.9%) were demonstrated consistent overexpression of KIT. There was statistical significance in the correlation between KIT expression and tumor size (P < 0.01), pathological stage (P < 0.01), tumor grade (P < 0.01) and P53 (0.01 < P < 0.05). On multivariate analysis, positive KIT expression presented an independent predictive factor for decreased overall survival (hazard ratio 17.26, P = 0.005). The estimated mean survival time was 25.6 months for KIT-positive patients and 56.9 months for KIT negative patients, P < 0.001.

Conclusions

The expression of KIT was significantly associated with tumor size, pathological stage, tumor grade and P53 in clear cell renal cell carcinomas. The expression of KIT is an important survival predicting factor for patients with clear cell renal cell carcinoma.

Introduction

Renal cell carcinoma (RCC) composes of 2–3% of adult malignancies, and accounts for 90–95% of renal tumors (Jemal et al. 2006). The World Health Organization classification divides RCC into clear cell (CC), chromophobe, papillary, collecting duct, and many other rare subtypes. In all the subtypes of RCC, CCRCC is the most common. Some prognostic markers have been established for RCC, such as TNM stage, Fuhrman nuclear grade, vascular invasion, necrosis, and a number of molecular markers, etc., (Shiina and Urakami 2004).

Recently, KIT, a type III transmembrane tyrosine kinase receptor, has been widely studied. Coded by c-kit proto-oncogene located at the long arm of chromosome 4, KIT is essential in regulating the development of hemopoietic stem cell, germ cell, and Cajal cell (rhythm pacing cell of the gastrointestinal tract), etc. Overexpression of KIT has been detected in various kinds of tumors, including mastocytoma, seminoma, melanoma, synovial sarcoma, rhabdomyosarcoma, angiosarcoma, fibromatosis, endometrial cancer, ovarian cancer, small cell carcinoma of lung, duct carcinoma of breast, and RCC, etc. (Gibson and Cooper 2002). Although the expression of KIT has been studied in different subtypes of renal cell carcinomas by immunohistochemistry and western-blotting, the role of KIT in RCCs has not been well identified in the previous investigations.

In this study, the expression of KIT was studied by immunohistochemistry method in 119 CCRCC samples and correlated with clinicopathological features, the expression of p53 and overall survival.

Materials and methods

Patient information

From June 2002 to November 2006, 119 embedded CCRCC specimens were collected. Hematoxylin–eosin-stained slides of each sample were reviewed for RCC type, pathological stage, and nuclear grade. The tumor subtypes were classified according to the 2003 WHO classifications. The pathological stage and nuclear grade were standardized by 2002 AJCC staging system (Greene and Fleming 2002) and Fuhrman Grade staging system (Fuhrman et al. 1982), respectively.

Immunohistochemistry study

Immunohistochemistry was performed by routine two-step method on 4-μm sections of each selected blocks. The KIT rabbit polyclonal antihuman antibody C-19 (Santa Cruz Corp.), dilution 1:50, was chosen as the first antibody, which has been verified with high specificity (more than 90%) in the diagnosis of GISTs (D’Amato et al. 2005); (Goss and Manola 2000); (Hornick and Fletcher 2004). The deparaffinized slides were pretreated with a 10 mmol/L concentration of citric acid buffer, pH 6.0, and boiled for 15 min for antigen-repairing. The endogenous peroxidase was quenched by hydrogen dioxide, and then the nonspecific binding was blocked by rabbit serum, the second antibody. In the end, cells with yellow to brown granules in cell membrane or plasma were calculated as positive staining. The intensity of staining was graded as either absent, weak, moderate, or strong (0–3 scale). The extent of staining was evaluated semiquantitatively and categorized as 0%, less than 10%, between 10 and 50%, between 50 and 80%, and more than 80% (0–4 scale). Aggregate scores were obtained for each case (range 0–7); scores of ≥3 were regarded as positive cases (Miliaras et al. 2004). All the slides were reviewed by one senior pathologist from the Department of Pathology, Cancer Hospital, Fudan University. The procedure of detecting P53 was the same with KIT. For P53, negative (0) was defined as less than 10% positivity of tumor cells, and positive (1) as more than 10% tumor cells stained.

Statistical methods

Chi-square test and phi correlation test were used to exam the relationships between KIT and gender, age, tumor size and P53. Kendall correlation test was used to analyze the correlation between KIT expression and pathological stage and Nuclear grade. Overall survival was defined as the time from nephrectomy to the time of death and patients who were alive were censored at the time of last follow-up. Survival curves were plotted using the Kaplan–Meier method, and statistical significance was assessed using the log-rank test. Univariate analysis of overall survival was performed with the Kaplan–Meier method. The Cox proportional hazards model was used for multivariate analysis. P values less than 0.05 were considered statistically significant. All statistical analyses were performed using commercially available software (SPSS 14.0).

Results

Clinicopathological, and immunohistochemical features

A total of 74 male and 45 female cases were enrolled with a mean age of 55.8 (95% CI 52.6–59.0). The mean diameter of the tumors was 4.93 cm (95% CI 4.45–5.41 cm). KIT was positive in 10.9% (13/119) of clear cell RCCs and the staining for KIT was observed mainly in cell membrane of the cancer cells (Fig. 1). Positive P53 staining appeared in 35 samples and localized on the nuclei of the cancer cells. The relationships between the expression of KIT and other pathological and immunohistochemical features and clinical characteristics were concluded in Table 1. Positive expression of KIT was significantly associated with advanced pathological stage (P < 0.001), higher nuclear grade (P < 0.001), larger tumor size (P < 0.001), positive P53 expression (P = 0.009). Gender and age were not significantly correlated with the expression of KIT in this study.

Fig. 1.

a Clear cell RCC in HE staining. b Clear cell RCC with negative KIT staining. c Clear cell RCC with positive KIT staining

Table 1.

Relationships between KIT and gender, age, tumor size, pathological stage, nuclear grade, P53, and survival status

Parameter	KIT positive	KIT negative	P value
No. of patients	13	106
Gender
Male	9	65	0.579* (Phi value = 0.051)
Female	4	41
Age
<56	8	54	0.470* (Phi value = 0.066)
≥56	5	52
Tumor size (cm)
<5.0	1	64	<0.001** (Phi value = 0.330)
≥5.0	12	42
Pathological stages
I	2	79	<0.001***
II	5	19
III	1	4
IV	5	4
Nuclear grade
1	3	82	<0.001***
2	5	19
3	5	5
P53
0	5	78	0.009* (Phi value = 0.230)
1	8	28
Survival status
Alive	8	100	<0.001* (Phi value = 0.353)
Dead	5	6

* Tested by phi correlation test

** Tested by phi correlation test

*** Tested by Kendall correlation test

Overall survival

Median follow-up time for the entire cohort was 22 months (range 5–62) (Fig. 2). On univariate survival analysis, the estimated mean survival time was 25.6 months for KIT-positive patients and 56.9 months for KIT negative patients, P < 0.001 (Fig. 3). A multivariate analysis using the Cox proportional hazard model was performed and included gender, age, pathological stage, nuclear grade, tumor size, P53, and KIT expression. The independent predictors of decreased overall survival were positive KIT expression (P = 0.005, hazard ratio 17.26) and higher pathological stage (P < 0.001, hazard ratio 4.79) (Table 2).

Fig. 2.

Overall survival curve for all 119 patients with clear cell renal cell carcinoma

Fig. 3.

Different overall survival for KIT-positive and KIT-negative patients with clear cell renal cell carcinoma

Table 2.

Cox regression hazards model for overall survival

Parameter	P value	Hazard ratio	95% CI
Gender	0.44	0.56	0.13–2.48
Age	0.17	0.96	0.91–1.02
Tumor size	0.74	0.95	0.69–1.30
Pathological stage	<0.001	4.79	2.23–10.29
Nuclear grade	0.79	0.85	0.26–2.79
Expression of KIT	0.005	17.26	2.34–37.15
Expression of P53	0.29	0.50	0.14–1.80

Discussions

Yamazaki et al. (2003) first reported the expression of KIT in the epithelial components of chromophobe RCC, but failed to detect any mutations which are quite common in GISTs. So they concluded that chromophobe RCC could express the productions of wild c-kit gene. Miliaras et al. (2004) investigated KIT in different types of RCCs and found 2/13 clear cell RCCs, 2/7 papillary RCCs and 4/7 chromophobe RCCs were KIT-positive. At the same time, Zigeuner and Langner (2005) examined the presence of KIT in primary and metastatic RCCs and upper urinary tract transitional cell carcinomas and concluded that KIT immunoreactivity was infrequent in both RCCs and upper urinary tract TCCs. Most recently, Pan and Chen (2004) reported that some papillary RCCs were KIT-positive as well as chromophobe RCCs, and Point-mutation were detected on intron 17 of c-kit gene, but its function was still unknown.

The expression of KIT was detected in 10.9% of CCRCC samples in this study. Our results were consistent to the findings of Miliaras et al. (2004), but different from Zigeuner and Langner (2005) results. Different antibodies used in these studies might account for the discrepancy in the KIT-positive rate (Hornick and Fletcher 2003). C-19 (Santa Cruz Company) was used in our study, and clone T595 from Novocastra was applied in Miliaras’ study, but Zigeuner used antibody from DAKO Company. Different races and different characteristics of patients might also contribute to different results. Furthermore, In the Miliaras’s study, immunoreactivity was cytoplasmatic in all positive cases, but in ours, immunoreactivity was observed mainly in cell membrane of the cancer cells. The main difference might be in the different immunohistochemical methods used. In Miliaras’s study, streptavidin–biotin method was used, but in our institute, only immunohistochemistry Elivison method was applicable. Different antibodies might also attribute to this disparity.

Since the last three decades, a lot of clinical, pathological and molecular markers have been proved to have predictive value for the prognosis of RCC (Phuoc et al. 2007). We found that positive expression of KIT in CCRCC was associated with advanced pathological stage, higher nuclear grade, larger tumor size, and shorter survival time; however it was not related to patients’ gender and age. Multivariate analysis also revealed its independent prognostic significance for shorter overall survival in CCRCC patients. So, including the expression of KIT will increase the prognostic significance of conventional clinicopathological predictors.

The association between KIT expression and tumor progression might be linked to inhibiting apoptosis mediated by P53 and the possible existence of tumor stem cell. P53 expression was demonstrated to be an independent predictor of survival in RCC patients with metastasis at surgery in a multivariate analysis; Moreover, Ljungberg et al. (2001) and Zigeuner et al. (2004) confirmed that p53 expression was an independent prognostic factor for metastasis-free survival in CCRCC patients. Additionally, Lee (1998) discovered that KIT could accommodate the functions of mitochondria and redox state of proteins to inhibit the apoptosis related to P53. Our study further validated the relationship between KIT and P53, and positive KIT expression might lead to a higher malignancy of abnormal proliferation.

As a well-known stem cell factor, KIT could regulate the development of hemopoietic stem cell, germ cell and many other stem cells. We hypotheses that KIT might be one of the markers of cancer stem cell in CCRCC. But the role of KIT in the formation and metastatic process of CCRCC has not been well explained in this study. Further investigations should be carried out in the genetic changes of KIT-positive tumors and cancer stem cell in CCRCC.

Conclusions

This study shows that KIT expression was significantly correlated with tumor size, pathological stage, tumor nuclear grade and P53 in CCRCC. KIT-positive CCRCC might have higher pathological stage, higher nuclear grade and higher malignancy. Furthermore, the expression of KIT is an important survival predicting factor for patients with CCRCC.