Over the past several years, a seemingly simple premise has guided drug development for metastatic renal cell carcinoma (mRCC). Building on critical advances in the understanding of mRCC biology, agents were developed that targeted vascular endothelial growth factor (VEGF)-mediated signaling. The broadest class of these agents included small molecules that inhibited the tyrosine kinase domain of the VEGF receptor. Since the first generation of VEGF-tyrosine kinase inhibitors (VEGF-TKIs) was introduced, there has been an intensive effort to increase the specificity and potency of these agents for the VEGF receptor family. The theoretical rationale is straightforward – (1) increasing specificity will limit the scope of toxicity generated by off-target effects, and (2) increasing potency will enhance the antitumor effect.

The effort to synthesize such agents has been successful. As noted in Table 1, more recently introduced VEGF-TKIs (i.e., tivozanib and axitinib) are able to inhibit VEGF receptors at sub-nanomolar concentrations, in contrast to older agents (i.e., sorafenib and sunitinib).1  The clinical relevance of these advances has come into question, however, with the report of several phase III studies. The TIVO-1 trial illustrates the issue well – in this phase III study, patients who had received up to 1 prior line of therapy (excluding targeted agents) were randomized to receive tivozanib or sorafenib.2  Although the study achieved its primary endpoint, demonstrating an improvement in progression-free survival (PFS) with tivozanib relative to sorafenib, the difference was relatively modest (11.9 mos vs 9.1 mos, P=0.042). Moreover, preliminary analysis of 1-year overall survival (OS) actually favored sorafenib (81% vs 77%).3  Further challenging the premise of simply increasing receptor potency and specificity is a press release from the AGILE 1051 study.4  Experts in the field had already posited axitinib to be a superior choice for front-line therapy. However, this phase III study randomizing treatment-naïve patients to either axitinib or sorafenib found no significant improvement in PFS with axitinib.

How can the discordance between hypothesis and reality be explained? First, it is possible that the antitumor effect of VEGF-TKIs is not solely contingent upon affinity of the drug for the receptor tyrosine kinase (RTK) domain. A detailed analysis reported by McTigue et al suggests that complex conformational changes in the VEGF receptor (in both the RTK domain and in critical juxtamembrane regions) may correlate with the clinical activity of these agents.1 Second, it is possible that a lack of specificity for the VEGF receptor family may be an asset in some regards. A report from Yuen et al suggests that the antitumor effects of sorafenib may be due to off-target effects on cyclin D1, cyclin B1, surviving and other key regulatory proteins.5 Finally, it is increasingly recognized that VEGF-TKIs exert a spectrum of effects beyond the scope of angiogenesis inhibition.  As one example, sunitinib appears to inhibit recruitment of myeloid-derived suppressors cells (MDSCs) and regulatory T-cells (Tregs) to tumor tissue in preclinical models, thereby potentially augmenting the antitumor immune response.6  This effect is not consistent across VEGF-TKIs – similar studies examining sorafenib show no substantial effect on Treg recruitment and a decrease in antigen-specific T-cells.7  As trials are now underway that pair VEGF-TKIs with distinct immune-based therapies (i.e., vaccine therapies such as IMA-901 and AGS-003, or programmed death-1 [PD-1] inhibitors such as BMS-936558), it is of critical importance to understand the immunologic properties of VEGF-TKIs.8-9 

If increased specificity and potency for the VEGF receptor does not translate to enhanced clinical efficacy, then how should the overarching research strategy for mRCC be revised in the coming years? One approach is to characterize patients on the basis of their relative “VEGF-dependency”. As delineated in Figure 1, a subset of patients may be highly sensitive to VEGF-directed therapies. For these patients, the choice of VEGF-directed therapy may be less critical, and their treatment may include sequential therapy with distinct VEGF-TKIs. Clinical justification for this is derived from the INTORSECT study, comparing temsirolimus and sorafenib in patients with mRCC who had received first-line sunitinib.10  In this study, sorafenib showed an improvement in OS relative to temsirolimus (a secondary endpoint), and this benefit was most pronounced in those patients who had received a duration of sunitinib therapy in excess of 6 mos (presumably, a VEGF-dependent group). 

A second group of patients may have intermediate sensitivity to VEGF-TKIs. In this group of patients, synergistic combinations of VEGF-TKIs with other novel agents should be pursued. As noted earlier, if these combinations include immune-based strategies (i.e., PD-1 inhibition or vaccine therapy), then an enhanced understanding of VEGF-TKI effects on MDSCs, Tregs, and other key mediators will be critical to devise optimal pairings. Finally, in patients who are insensitive to a VEGF-based approach, mammalian target of rapamycin (mTOR) inhibitors and other non-VEGF-directed therapies might be attempted. At present, the degree of VEGF-dependency can only be determined in a post hoc fashion (specifically, through assessing the initial response to VEGF-directed therapies). However, if existing biospecimen repositories derived from trials of VEGF-TKIs are mined, it is possible that a baseline molecular signature defining VEGF-dependency could be formulated. With prospective validation, such a tool could drastically change the current approach to mRCC.  

TABLES

Table 1. In vitro activity of selected VEGF-TKIs. (Adapted from McTigue M, Murray BW, Chen JH, Deng YL, Solowiej J, Kania RS Feature Article: Molecular conformations, interactions, and properties associated with drug efficiency and clinical performance among VEGFR TK inhibitors. Proc Natl Acad Sci U S A 2012.)

FIGURES

Figure 1. A proposed strategy for approaching mRCC based on the degree of VEGF-dependency.