CTLA-4 (also known as CD152) is a member of the CD28 family of co-inhibitory receptors expressed on the surfaces of Foxp3⁺ Treg cells and activated by T cells [40, 50]. CTLA-4 binds its ligands, B7.1 (CD80) and B7.2 (CD86), expressed on antigen-presenting cells. The inhibitory effect on T-cell activities is derived mainly from two mechanisms. One is the competition between CTLA-4 and costimulatory receptor CD28, for binding shared ligands (B7.1 and B7.2). As CTLA-4 has stronger binding affinity than CD28, it reduces CD28-derived T-cell stimulation. The other is a direct intracellular signal that CTLA-4 sends to attenuate the T-cell receptor (TCR)-mediated signaling pathway [51]. It has also been suggested that the anti-CTLA-4 Ab ipilimumab, which has a humanized IgG1-type Fc domain with the highest affinity for Fcγ receptors among the known human IgG isotypes, might have an exceptional mechanism of action [52]. In a mouse model, it was shown that intra-tumoral Tregs with higher CTLA-4 expression relative to other T-cell subsets were preferentially depleted by antibody-dependent cellular cytotoxicity (ADCC) [53]. On the basis of encouraging preclinical data demonstrating this antitumor effect, CTLA-4-blocking antibodies have been extensively evaluated in melanoma and other histological tumor types.

Ipilimumab is a fully humanized IgG1 monoclonal Ab that blocks CTLA-4. A phase II trial compared two cohorts of RCC patients receiving either continuous 3 mg/kg or one dose of 3 mg/kg followed by 1 mg/kg of ipilimumab (both every 3 weeks) [54]. In the former (higher-dosing) cohort, treated with the approved and commonly used regimen, 5 of 40 patients including IL-2 refractory cases showed a partial response (PR), three of whom experienced a durable (> 1 year) response. Grade 3–4 immune-related adverse events (irAEs) were seen in 33 % of all patients.

Programmed cell death protein-1 (PD-1, also known as CD279) is a member of the CD28 family of co-inhibitory receptors, which are inducibly expressed on the T, B, and NK cells [40, 55]. The ligands of PD-1 are PD-L1 (B7-H1) and PD-L2 (B7-H2), expressed on antigen-presenting cells [56]. The difference between PD-1 and CTLA-4 is that CTLA-4 functions in the priming phase of T-cell activation, PD-1 in the effector phase. PD-1 has a cytoplasmic domain that sends inhibitory signals to the TCR-derived signaling pathway [51]. It has been shown that PD-L1 is expressed in some tumor types including RCC [57–59] to evade antitumor immune surveillance and eradication [60]. PD-1 and/or PD-L1 expression in tumor tissues from RCC cases is associated with tumor aggressiveness and a poor prognosis [61, 62]. Preclinical and clinical studies have shown that blocking the PD-1-PD-L1/2 pathway leads to reactivating antitumor immune responses and tumor regression [63].

Nivolumab is a fully human IgG4 monoclonal antibody against PD-1 [63]. A first-in-human, dose-escalating phase I clinical trial evaluated nivolumab in 39 patients with advanced metastatic solid tumors, including one RCC patient who had VEGF inhibitor refractory disease [64]. Nivolumab was generally well tolerated, and the maximum tolerated dose with a single administration was not defined. The RCC patient showed a durable PR after three doses of 10 mg/kg of nivolumab and then a CR after 3 years without any antitumor therapy [65]. A subsequent phase I trial evaluated nivolumab in a larger cohort of patients, including 34 RCC patients with advanced disease [66]. The long-term follow-up data from the RCC cohort was recently published [67]. The patients, 71 % of whom had already been treated with 2–5 regimens, received either 1 mg/kg (N=18) or 10 mg/kg (N=16) of nivolumab every 2 weeks in an 8 week cycle up to 96 weeks. In all patients, the ORR was 29 %, with a median duration of 12.9 months. The median PFS was 7.3 months (95 % CI: 3.6–10.9), and the median OS was 22.4 months (95 % CI: 12.5, not estimable). Grade 3–4 adverse events (AEs) occurred in approximately 18 % of patients. Based on the encouraging results of these early studies, a randomized, dose-ranging phase II trial was conducted [68]. The 168 RCC patients received either 0.3 (N = 60), 2.0 (N = 54), or 10.0 (N = 54) mg/kg of nivolumab administered once every 3 weeks. There were 118 patients (70 %) who had received at least two prior systemic therapies. The primary end point was PFS. The ORRs were 20 %, 22 %, and 20 % in the 0.3-, 2.0-, and 10.0-mg/kg cohorts, respectively. There was no dose-response relationship for PFS: the median PFS rates were 2.7 (80 % CI: 1.9–3.0), 4.0 (80 % CI: 2.8–4.2), and 4.2 (80 % CI: 2.8–5.5) months, respectively (P = 0.9). The median OS rates were 18.2 (80 % CI: 16.2–24.0), 25.5 (80 % CI: 19.8–28.8), and 24.7 (80 % CI: 15.3–26.0) months, respectively. Compared with the previous phase I trial, this study showed lower ORR and PFS but similar OS. A phase III RCT (CheckMate 025 trial) compared nivolumab with the mTOR inhibitor everolimus, which is a standard salvage therapy, in previously treated RCC patients [69]. In July 2015, this study was stopped ahead of schedule after the independent data monitoring committee found significant survival superiority in those receiving nivolumab as compared to everolimus. In total, 820 patients had been randomly assigned at a 1:1 ratio to nivolumab (3 mg/kg every 2 weeks; N = 410) or everolimus (10 mg daily; N = 411). The primary end point was OS. The ORR was 25 % in the nivolumab vs. 5 % in the everolimus cohort (odds ratio 5.98; 95 % CI: 3.68–9.72; P < 0.001). The median PFS was 4.6 (95%CI: 3.7–5.4) vs. 4.4 (95%CI: 3.7–5.5) months (HR: 0.88; 95 % CI: 0.75–1.03; P = 0.11), and the median OS was 25.0 (95%CI: 21.8-NE) vs. 19.6 months (95%CI: 17.6–23.1) (HR 0.73; 98.5 % CI: 0.57–0.93; P = 0.002), respectively. Grade 3–4 irAEs occurred in 19 % of the patients receiving nivolumab, an incidence significantly lower than that in the everolimus cohort (37 %).

The Checkmate 025 trial validated earlier trials showing OS prolongation with nivolumab as salvage therapy. The lack of consistent PFS extension might be attributable to variable patterns of responses to immune checkpoint blockade, with some of the treated patients initially showing progression on computed tomographic scans due to transient inflammation giving the appearance of tumor enlargement. This “pseudo-progression” highlights the necessity of establishing specific response criteria for cancer immunotherapy [70, 71]. Another clinically important point underscored by this study was the lack of an association between PD-L1 expression in the tumor and the response to nivolumab. Pretreatment biomarkers predicting therapy responsiveness and/or toxicity are urgently needed to select patients who would actually benefit from this class of therapy. As to this point, which we will discuss separately (see section below), nivolumab was approved by the US FDA for patients with advanced RCC refractory to anti-angiogenic therapy, in November 2015.

Atezolizumab (MPDL3280A) is a humanized IgG-1κ monoclonal Ab to PD-L1 [72]. It has a genetically modified Fc domain that is involved in impairing the ADCC-dependent cellular cytotoxicity directed to PD-L1-expressing cells. Following studies in other types of tumors [73, 74], atezolizumab was recently evaluated in a phase I dose-escalation study in 70 patients with previously treated advanced RCC [75]. The doses ranged from 3 to 20 mg/kg (administered every 3 weeks). Atezolizumab was well tolerated and no maximum tolerated dose was defined. Grade 3 treatment-related AEs and irAEs were seen in 17 % and 4 % of patients, respectively, but no grade 4–5 AEs occurred. The ORR was 15 % in 62 evaluable patients. There was a trend toward a higher response rate in patients with PD-L1 expression in TILs. Importantly, an antitumor response was seen in patients with known poor prognostic features, including high Fuhrman grade 4 and/or sarcomatoid features (22 %), and poor Memorial Sloan-Kettering Cancer Center (MSKCC) prognostic risk (25 %).

Several other agents are presently being investigated ([76], summarized in Table 12.3). Recent clinical developments of antibody therapies for use with costimulatory/inhibitory molecules on T cells (immunomodulators) are shown in Fig. 12.2.

Although immune checkpoint inhibitors have opened up a new era in the treatment of advanced RCC, and there are many promising agents under investigation, the accumulating data on the leading drugs suggest that only a limited number (approximately 20 %) of patients can expect long-term survival with monotherapy [45]. The future challenges include exploring combination therapies that can improve antitumor efficacy without increasing toxicity [47–49]. There are numerous possible novel regimens being evaluated in ongoing clinical trials, seeking the optimal pairing, dosing, and administration sequence ([77], summarized in Table 12.4].