The anatomical factors incorporated into the TNM (2009) classification include tumor size, tumor growth beyond the renal capsule, invasion of the renal vein and/or inferior vena cava, lymph node invasion and distant metastases (Chap. 17). This staging system is regarded as the most important prognostic factor for RCC and there is a clear distinction between the different stages and cancer specific survival as shown in the table below. In general, a poor prognosis and survival is associated with a higher T stage, lymph node involvement and distant metastasis [16]. Additionally, tumor size, pathological stage or lymph node status are regarded as an important independent prognostic factors.

The Fuhrman nuclear grading (Chap. 17) is a widely used and accepted classification. In most survival studies undertaken over the past two to three decades, grading has been based on the criteria of the Fuhrman classification. This has gained widespread acceptance in clinical practice with Fuhrman’s 1982 report being one of the most cited studies in the renal cancer literature [19]. It is defined by four nuclear grades 1–4 in order of increasing size, nuclear irregularity and nucleolar prominence. A poor prognosis and increasing metastatic rate is seen with increasing grade. The system is however subject to inter and intra observer error and its value in other histological subtypes apart from ccRCC remain questionable [20].

Of the three main histological subtypes, clear cell has the poorest prognosis, followed by papillary and then chromophobe [21]. Furthermore, type 2 papillary tumors tend to be of high grade with increased metastatic potential and have a poorer prognosis than type 1. The stage and grade of the tumor however have a greater impact on prognosis than subtype [21]. Tumors with sarcomatoid differentiation, tumor necrosis or that involve the collecting system have an extremely poor prognosis [16].

Performance status, weight loss greater than 10 % of body weight, anemia and local symptoms such as flank pain, palpable mass, haematuria, varicocele, constitutional and paraneoplastic symptoms, have all shown to have prognostic significance [16]. Additionally, presentation with symptoms rather than an incidental finding of a RCC carries a poorer prognosis [22].

There are several biomarkers that have the potential to be used for screening, diagnosis or in follow-up of RCC which are currently under investigation and their value as prognostic markers is yet to be determined. A number of techniques including microarray, RT-PCR and immunohistochemistry have been utilized to determine the behaviour of RCC. They include carbonic anhydrase IX, VEGF, HIF, Ki67, p53, PTEN, E-cadherin and CD44 [18].

Whilst several prognostic factors have so far been described, with some having important independent prognostic value, there is no single feature which can yield sufficient predictive accuracy for a valid estimation of outcome. This has lead to the development of several prognostic models, or nomograms, which incorporate several established parameters in order to improve prognostic accuracy. Current European guidelines do not recommend using prognostic models for localized disease even though they can provide rationale for enrolling patients on to clinical trials [18]. The first nomogram that was used to predict the disease recurrence post-nephrectomy with a variable accuracy between 61 and 81 % was developed by Kattan [23]. The variables used included the presence of symptoms at diagnosis, histologic subtype, tumor size, and the 1997 TNM stage. Several nomograms have since been developed and there are currently three systems for localized and two for metastatic RCC.