Expression of multidrug resistance markers ABCB1 (MDR-1/P-gp) and ABCC1 (MRP-1) in renal cell carcinoma

Kidney cancer accounts for approximately 2% of all adult cancers. It is the 7^th leading cause of cancer in the US with an estimated incidence of approximately 51,000 new cancer cases per year in 2007 [1]. Renal cell carcinoma (RCC) is the most common tumour arising from the cells in the lining of tubules in the kidney [2]. At the time of diagnosis, 30% of patients will have metastatic or unresectable disease, and the 2-year overall survival of this cohort is <10% [3]. The incidence of kidney cancer is rising; it is 2 times more common in men than women. Risk factors include obesity [4], smoking [5] and hypertension [6].

RCC is a chemoresistant tumour usually exhibiting only a marginal response. Radiotherapy and chemotherapy are generally ineffective in the treatment of advanced renal tumours [7, 8]. The intrinsic occurrence of multidrug resistance (MDR) modulates the resistance of tumours to a wide variety of and structurally distinct chemotherapeutic drugs through the expression of drug efflux pumps [9]. Two of the most widely studied efflux pumps, MDR-1 P-gp/P-170, the gene product of MDR-1 (ABCB1) and MRP-1 (ABCC1) which encodes a 190 kDa membrane protein have both been demonstrated to pump a wide variety of the most commonly used cancer drugs out of tumour cells. Their over expression correlates broadly with drug resistance in many different forms of cancer including pancreatic cancer [10], lung cancer [11], breast cancer [12] and glioma [13]. The relative contributions and causative role, if any, of MDR associated protein efflux pumps in renal carcinoma have not been fully elucidated.

Studies detailing the prevalence and contribution of MDR-1 P-gp in RCC are conflicting. MDR-1 P-gp expression has been reported widely in untreated renal carcinomas [14, 15]. It does appear that intrinsic drug resistance exists in many renal RCC and it is associated, at least in part, with increased expression of MDR-1 P-gp. However, the exact prognostic significance of this expression remains unclear with conflicting results described. Longer progression free survival has been observed in patients with none or very few MDR-1 P-gp positive tumour cells compared to patients with a larger proportion of MDR-1 Pgp positive tumour cells [16, 17]; however higher MDR-1 expression has been associated with a better outcome also [18, 19]. Expression of MDR-1 P-gp has been shown to correlate with a well differentiated tumour phenotype in renal carcinoma [18, 20, 21]. Higher MDR-1 gene expression has been observed in RCCs that have metastasised/invaded through the renal capsule compared to early stage non invasive tumours [20, 22].

Studies addressing the contribution, if any, of the MRP-1 efflux pump and its gene product in this disease are limited. MRP-1, like MDR-1 P-gp is highly expressed in normal kidney. MRP-1 gene over expression has been observed in renal carcinomas, this expression does not appear to correlate with grade/clinical stage in this disease [19, 23]. To our knowledge, there have been no reported studies looking at MRP-1 protein expression in RCC.

In this study, we evaluate the expression of MDR efflux pumps, MDR-1 P-gp and MRP-1, using immunhistochemical analysis, in 95 RCCs, to investigate the relative contributions of these efflux pumps in this disease.

Chemotherapy is the standard treatment for most solid tumours; however, RCC is generally resistant to chemotherapy. The reason for the resistance of kidney cancer cells to chemotherapy is not completely understood. The specific role and relative contributions of ABC transporter pumps in clinical resistance in RCC have not been fully determined; results from previous studies are conflicting.

The MDR1 gene and its gene product, P-gp are ubiquitously expressed in mesangial, proximal tubule, thick loop of Henle, and collecting duct cells of the kidney [26]. As P-gp plays a functional role in the clearance of xenobiotics from the mesangial and proximal tubule cells of the kidney, this may contribute to the inherent multidrug resistance phenotype of RCC. MDR-1 gene overexpression or protein have been identified in the majority of RCC samples studied [14–17, 19, 21].

In this reterospective study we have shown that renal cancer cells produce an overabundance of this drug efflux pump, with 100% of RCCs studied exhibiting MDR-1 P-gp positivity. 76% of these tumours exhibited MDR-1 P-gp staining in 50% or more of tumour cells. As renal tissue inherently expresses high levels of this transporter; results presented here are not unexpected. Using a similar scoring system as that used here, Mignogna et al., [16] observed MDR-1 P-gp expression in 100% (30/30) of RCC. Other immunohistochemical studies on smaller patient cohorts report 61.5% (8/13) and 47.6% (10/21) of RCC showing MDR-1 P-gp positivity [14, 15]; studies investigating MDR-1 gene expression in RCC report varying levels of gene expression, in general slightly less that the MDR-1 P-gp protein expression observed here [19–23]. It is well recognised that protein levels do not necessarily correlate with mRNA levels, furthermore these differences between the various studies may in part result from the different techniques employed and different cohorts of patients and scoring criteria used. The high levels of MDR-1 P-gp expression observed here do suggest that this transporter protein is contributing, at least in part, to the chemoresistance of RCCs studied here.

Mignogna et al., [16] suggested a role for MDR-1 P-gp as a possible adverse prognostic factor of chemoresistance and aggressive behaviour in renal carcinoma, their study showed an association between high MDR-1 P-gp expression (MDR-1 positivity in 40% or more of tumour cells) and poor survival as confirmed by Cox multivariate analysis. In agreement with this study, Duensing et al., [17] suggested a potential role for P-gp as a biologic parameter predictive of tumour progression in renal cell carcinoma patients, as longer disease-free survival was observed in patients with <1% MDR-1 postivity. However, Hofmockel et al., [18] correlated lower MDR-1 expression with poorer prognosis. Due to the lack of data regarding patient outcome; any possible prognostic significance of the expression of these efflux pumps observed could not be addressed in this study.

Expression of this efflux pump does not appear to be associated with the histological tumour grade of RCC in this patient cohort. Unexpectedly lower MDR-1 levels have been shown to be associated with poorly differentiated RCC [18, 20–22]. However, in agreement with our observations Mignogna et al., showed MDR-1 protein expression as having prognostic significance independent of any association with tumour grade [16].

We have also shown high levels of MRP-1 protein expression in RCC; all tumours investigated showed MRP-1 protein expression with 61% of tumours exhibiting MRP-1 positivity in at least 50% of tumour cells. As in the case of MDR-1 Pgp, this efflux pump is also expressed at high levels in the normal kidney; so such an observed high level of expression again is not unexpected. This is the first report to our knowledge of MRP-1 protein expression being investigated in RCC patients, previous work has focused on gene expression studies. Again, this observed high level of MRP-1 protein expression suggests that this efflux pump also may be playing a contributing role in the chemoresistance of these renal carcinomas.

There was no apparent correlation between expression of either of the MDR proteins studied here with other clinicopathological features (gender, tumour size and nodal status). It was unfeasible to carry out detailed statistical analysis due to limitations and inadequate data. Histological grade, pathological stage and tumour size have been demonstrated using multivariate analysis to be significant prognostic indicators regarding RCC patient outcome [27]. Recently described molecular and genetic markers may also provide diagnostic and prognostic value in RCC [28].

It is likely that chemoresistance in RCC is multifactorial, other factors implicated as being involved in the complex chemoresistance mechanisms of RCC include Clusterin [29]; there also is some evidence that lung resistance protein (LRP/MVP) may contribute to inherently resistant renal carcinoma [30, 31].

The inherent resistance of RCC to conventional treatment has lead to the use of immunotherapies such as alpha-interferon and interleukin-2 (IL-2). However, response does not translate to long-term benefit and does not prolong overall survival in many cases [32–34]. Newer "targeted-therapies" such as sunitinib [34, 35] are emerging for the treatment of patients resistant/intolerant to current treatment modalities or as an alternative to cytokine immunotherapy. Identification of new tumour associated antigens may hopefully provide the basis of new therapeutic strategies and lead to an improved outcome for this aggressive and chemoresistant disease.

Both MDR-1 P-gp and MRP-1 were expressed in 100% of the 95 specimens analysed. These high levels of expression suggest that both of these efflux pumps may be important contributors to the MDR phenotype in RCC. The incidence of MRP-1 positive tumours observed here warrants further study in order to confirm a possible contribution to chemoresistance in renal carcinoma. Due to the lack of data on the direct outcome of patients in this study, the prognostic significance of observed MDR protein expression was outside the scope of this investigation. Extensive studies with complete follow-up detailing expression of MDR-1 P-gp and MRP-1 together with further MDR associated markers (including LRP and MRP family members) and associated proteins may help to fully elucidate the specific contributions of these efflux pumps to the chemoresistance of RCC and furthermore address any possible prognostic and/or predictive role.