Facts and hopes for immunotherapy in renal cell carcinoma

Abstract

Immunotherapy has made a significant impact in many tumors, including renal cell carcinoma (RCC). RCC has been known to be immunoresponsive since the cytokine era of IFNα and IL2, but only a small number of patients had durable clinical benefit. Since then, discoveries of key tumor drivers, as well as an understanding of the contribution of angiogenesis and the tumor microenvironment (TME), has led to advances in drug development ultimately transforming patient outcomes. Combinations of anti-angiogenic agents with immune checkpoint inhibitors are now standard of care. Current challenges include patient selection for immunotherapy combinations, resistance acquisition and optimally sequencing therapies. Future discoveries about RCC biology, the TME and resistance mechanisms will likely pave the way for the next generation of therapies.

Introduction

Renal cell carcinoma (RCC) comprises many different entities. Tumors from the kidney and renal pelvis account for more than 76,000 new cases and 13,000 deaths per year in the United States (1). Critical processes supporting RCC tumorigenesis include angiogenesis and the tumor microenvironment (TME). Systemic treatment for metastatic RCC (mRCC) has evolved over the years and encompasses cytokines, anti-angiogenic agents, immune checkpoint inhibitors (ICIs), and most recently, combinations of anti-angiogenic agents and ICIs. Interestingly, none of these agents directly target tumor cells and understanding the TME has been key to therapeutic advances.

The tumor microenvironment of RCC

Consisting of blood vessels, immune and stromal cells, soluble and membrane-bound signaling molecules, and the extracellular matrix, the TME has a profound impact on tumorigenesis and immune evasion (Figure 1). Immune cells infiltrating tumors profoundly shape the TME (2). They include T cells, B cells, NK cells, macrophages and dendritic cells. RCC is infiltrated by T cells with increased expression of Th1 and Th17 related genes (3). Single-cell technologies such as cytometry by time of flight (CyTOF) and single-cell RNA-Seq (scRNA-Seq) have greatly expanded our understanding of the TME in RCC (4–7). By analyzing the immune infiltration of clear cell RCC (ccRCC), the most common type, Chevrier et al. identified two pro-tumor macrophage subsets associated with worse outcomes (4). More recently, Braun et al. used scRNA-Seq and T cell receptor (TCR) sequencing to analyze immune cells from patients with different stages of ccRCC (5). Their study revealed that terminally exhausted CD8 T cells and M2-like macrophages were enriched in advanced ccRCC (5). In addition, Krishna et al. showed that gene signatures of tissue-resident T cells and tumor-associated macrophages were associated with response to ICIs (6).

Figure 1.

Schematic overview of renal cell carcinoma with targeted immunotherapies. Top: Immune cell subsets and cytokines that shape the tumor microenvironment of RCC. Bottom: FDA approved drugs for RCC (bold) and their molecular targets.

cDC, classical dendritic cells; CTLA-4, cytotoxic T-lymphocyte antigen 4; eIF4E, eukaryotic translation initiation factor 4E; FKBP12, FK506-binding protein 12; HIF1α, hypoxia inducible factor 1α; HIF2α, hypoxia inducible factor 2α; IL8, interleukin 8; IL10, interleukin 10; MDSC, myeloid-derived suppressor cells; MHC-1, major histocompatibility complex I; mTORC1, mammalian target of rapamycin complex 1; PD-1, programmed cell death protein 1; PD-L1, programmed death-ligand 1; PDGFβ, platelet-derived growth factor β; PDGFRβ, platelet-derived growth factor receptor β; PTEN, phosphatase and tensin homologue; S6K, S6 kinase; TCR, T-cell receptor; TGFβ, transforming growth factor β; Treg, regulatory T cells; TSC1, tuberous sclerosis complex 1; TSC2, tuberous sclerosis complex 2; VEGF, vascular endothelial growth factor; VEGFR2, vascular endothelial growth factor receptor 2; VHL, von Hippel-Lindau.

The TME is also conditioned by the cytokine milieu. In RCC, TGF-β is overexpressed and inhibits T cell activation. Activation of the TGF-β pathway is associated with aggressive ccRCC (8). IL-10, which is upregulated in RCC, is classically anti-inflammatory (9,10). However, when combined with an anti-PD1 monoclonal antibody, pegylated IL-10 (pegilodecakin) induced T cell expansion and antitumor activity in patients with RCC (11,12). IL-8 is produced by tumor cells and promotes an immunosuppressive TME (13). High IL-8 circulating levels in cancer patients (including those with RCC) predicts for poor outcomes to ICIs (14,15). Finally, high circulating levels of IL-6 are predictive of resistance to anti-angiogenic therapies (16,17).

ICIs aim to induce sufficient numbers of functional tumor-specific T cells to eradicate cancer cells. To achieve this goal, the quantity, quality, and location of T cells matter. Tumor-specific antigens generated from somatic mutation via nonsynonymous single nucleotide changes, frameshifting insertions/deletions, or tumor-specific alternative splicing, promote tumor cell recognition. Tumor mutational burden (TMB) correlates with CD8 T cell infiltration and a favorable prognosis in several tumor types, but not ccRCC (18). TMB has been used as a biomarker to predict response to ICIs in multiple cancer types (19), but is not predictive in RCC (4–6,20,21). Interestingly, the amount of CD8 T cells in the TME negatively correlates with prognosis and does not necessarily predict for response to ICIs in RCC (22–26). Thus, despite the presence of CD8 T cells, ICIs are not always successful in mounting an anti-tumor response. In contrast, proliferating CD8 T cells, which likely recognize tumor antigens (27), are associated with favorable prognosis in patients with RCC (28). Tumor infiltrating T cells (TILs) show TCR clonal expansion, an indication of antigen stimulation, compared to T cells in adjacent normal kidney tissues (5). Some clonally expanded TILs may arise in response to frameshift-derived neoepitopes (29). Thus, at least a portion of the RCC-infiltrating CD8 T cells recognize tumor antigens. Such cells may be responsible for the therapeutic effects of ICIs. However, the percentage of TILs that recognize neoantigens, self-antigens (30) or other antigens is unclear. Notably, the ratio of anti-tumor T cells to tumor burden rather than the quantity of TILs alone may be predictive of outcome (27). Understanding the landscape of cognate antigens will facilitate immunotherapy development, including tumor vaccines (31).

CD8 T cells reside at different locations, express distinct markers, and have different functions (32). Advances in spatial transcriptomic technologies will likely shed light on the phenotype and function of T populations in various locations (33). T cells infiltrating RCCs have been identified at the tumor center, the tumor stroma and the invasive tumor margin (34). The amount of T cells in the tumor center may be associated with ICI responsiveness as has been shown with nivolumab therapy (34). Strategies that increase T cell infiltration, especially at the tumor core, could be promising for combination therapy with ICIs.

Studies have shown that in addition to the number of T cells in tumors, T cell differentiation is crucial for eradicating tumors (35,36). However, T cells in the TME often enter a dysfunctional state, T cell exhaustion, which limits their ability to eradicate tumor cells (35,36). Exhausted T cells upregulate co-inhibitory molecules, such as programmed cell death protein 1 (PD-1) and Cytotoxic T-Lymphocyte Associated antigen 4 (CTLA-4), which are targets of ICIs (37,38). Indeed, CyTOF and scRNA-Seq analyses of immune cells isolated from RCCs showed that terminally exhausted CD8 T cells expressing PD-1, Lymphocyte-Activation Gene 3 (LAG-3), and Hepatitis A Virus Cellular Receptor 2 (HAVCR2) are enriched in mRCC (4,6,39).

Recently a novel CD8 T cell subset resembling stem cells was reported (40–49). Stem-like CD8 T cells maintain long-term T-cell immunity through self-renewal and replenish exhausted T cells in cancer and chronic infection (40–49). Importantly, stem cell-like CD8 T cells appear to mediate ICI-induced T cell responses in some animal models of cancer and chronic infection (42,47). Tumor-infiltration by stem-like T cells is associated with an improved response to ICIs and extended survival in melanoma patients (50). Stem-like CD8 T cells have been identified in human RCC and support antitumor immunity (5,51). These stem-like CD8 T cells reside in the tumor stromal barrier and are supported by dense MHCII+ antigen presenting cells (APC) (51). These areas resemble the T cell zone of lymphatic tissue (51). This is consistent with the observation that stem-like CD8 T cells reside in lymphoid tissues during chronic viral infection (42,52). Whether and how stem-like CD8 T cells correlate with the response to immunotherapy in RCC is an active research topic.

The cytokine era

Since the 1990s, when interferon alfa (IFN-α) and interleukin 2 (IL-2) showed clinical activity, RCC has been known to be immunogenic and immune responsive. IFN-α belongs to the interferon cytokine family named for their ability to “interfere” with viral replication (53). In vitro studies, as well as studies in mouse models, suggest that IFN-α inhibits tumor growth through both tumor-cell intrinsic and immune mechanisms (54,55). Clinical trials in several tumor types led to the US Food and Drug Administration (FDA) approval of recombinant IFN-α2 as the first human immunotherapy (54,56).

IL-2 was first identified as a T cell growth factor (57). A key cytokine that activates cytolytic T cell effector function, IL-2 was shown to promote tumor cell killing by T cells in co-culture experiments (58). These findings prompted clinical studies that evaluated the efficacy of combining autologous killer T cells and recombinant IL-2 in cancer patients (59). In clinical trials, high-dose IL-2 was shown to be associated with durable responses and long-term survival in a small proportion of patients with mRCC, and in 1992, high-dose IL-2 received approval from the FDA (60–63). This immune responsiveness set the foundation for the development of further immunotherapies in mRCC (64).

The era of angiogenesis inhibitors

ccRCC frequently presents with biallelic inactivation of the von Hippel-Lindau (VHL) gene, leading to constitutive activation of hypoxia-inducible factor (HIF) and a pseudo-hypoxic state (65). As a result, ccRCCs are highly angiogenic and express higher levels of vascular endothelial growth factor-A (VEGF-A) than most other cancers (66). Tyrosine kinase inhibitors (TKIs), which inhibit angiogenesis by blocking VEGF receptors (VEGFR) and platelet-derived growth factor receptors (PDGFR), have significantly improved the survival of ccRCC patients (67). In total, 7 TKIs (sorafenib, sunitinib, pazopanib, axitinib, cabozantinib, lenvatinib and tivozanib) are FDA-approved for RCC treatment. In addition, the monoclonal VEGF-A neutralizing antibody (bevacizumab) is approved in combination with INF-α.

The immune checkpoint inhibitor era

Early development of ICIs included patients with mRCC (68,69). The first registration trial, Checkmate 025, compared nivolumab against everolimus, a mammalian target of rapamycin (mTOR) complex 1 (mTORC1) inhibitor, which along with temsirolimus, was FDA approved for RCC treatment (37). Checkmate 025 evaluated patients with metastatic ccRCC after progression on prior anti-angiogenic therapy and showed nivolumab superiority with improvements in overall survival (OS) and higher objective response rates (ORR) (37). Notably, correlative studies showed greater numbers of T cells in the tumor center in responders than non-responders (34). A similar observation was made for the CTLA-4 inhibitor tremelimumab in combination with cryoablation (70). A second ICI registration study in mRCC evaluated the combination ipilimumab-nivolumab vs. sunitinib in first-line treatment of ccRCC (Checkmate 214). Notably, one third of participants on ipilimumab-nivolumab in intention-to-treat long-term analyses were free of progression at 4 years (71). Median OS exceeded 4.5 years (Table 1). Ipilimumab-nivolumab was discontinued in some patients due to toxicity or other factors and treatment-free survival (TFS) was more than double for patients treated with immunotherapy compared to sunitinib (72). The ipilimumab-nivolumab combination is now considered standard of care frontline therapy for mRCC.

Table 1.

Summary efficacy data from phase 3 clinical trials for metastatic renal cell carcinoma of first-line immunotherapy combinations.

Trial	Checkmate 214	Keynote 426	Javelin Renal 101	IMmotion 151	Checkmate 9ER	CLEAR
Immunotherapy treatments (each compared to sunitinib)	Ipilimumab/Nivolumab	Axitinib/Pembrolizumab	Axitinib/Avelumab	Bevacizumab/Atezolizumab	Cabozantinib/Nivolumab	Lenvatinib/Pembrolizumab
Median follow up (months)	67.7	42.8	19.3	40	23.5	27
Median OS, months; HR	55.7 vs 38.4 0.72 (0.62–0.85)	45.7 vs 40.1 0.73 (0.60–0.88)	NR vs NR, 0.80 (0.61–1.03)	36.1 vs 35.3, 0.91 (0.76–1.08)	NR vs 29.5 0.66 (0.50–0.87)	NR vs NR 0.66 (0.49–0.88)
Median PFS, months; HR	12.3 vs 12.3 0.86 (0.73–1.01)	15.7 vs 11.1 0.68 (0.58–0.80)	13.3 vs 8.0, 0.69 (0.57–0.83)	11.2 vs 8.4, 0.83 (0.70–0.97)	17.0 vs 8.3 0.52 (0.43–0.64)	23.9 vs 9.2 0.39 (0.32–0.49)
ORR, %	39 vs 32	60 vs 40	53 vs 27	37 vs 33	55 vs 27	71 vs 36
CR, %	12 vs 3	10 vs 4	4 vs 2	5 vs 2	9 vs 4	16 vs 4
Treatment selection data	Sarcomatoid De novo metastases		Transcriptome signature Sarcomatoid	Transcriptome signature Sarcomatoid
Reference	71,84,85,87	78	79,88,95	89,96,97,98	77	76

OS: overall survival; PFS: progression free survival; HR: hazard ratio; ORR: objective response rate; CR: complete response; NR: not reached.

ICIs have also shown efficacy in the adjuvant setting. Pembrolizumab administration for a year showed improved disease free survival (DFS) compared to placebo (2-year DFS 77.3% vs 68.1%, HR 0.68, 95% CI 0.53–0.87, p=0.002) (73). This may be beneficial for patients at high risk for recurrence after nephrectomy or metastasectomy. Adjuvant pembrolizumab gained FDA approval on November 2021. It is yet unclear, however, whether adjuvant pembrolizumab will prolong survival, or which patients can be spared. Multiple trials are ongoing to evaluate the perioperative effect of ICIs including combinations of both PD-1 and CTLA-4 inhibitors.

Rationale and evidence for combinations

Some evidence suggest that VEGF signaling inhibitors can modulate the TME and provide synergy with ICIs (74). VEGF signaling shapes the tumor vasculature which influences infiltration by T cells (75). There is also evidence that some tumors preferentially respond to anti-angiogenic therapies whereas others to ICIs and thus, TKI-ICI combinations would be expected to have broad activity (85, 102).

To date, four separate TKI-ICI combinations (axitinib-pembrolizumab, axitinib-avelumab, cabozantinib-nivolumab, and lenvatinib-pembrolizumab) have been compared against sunitinib in the first-line treatment of metastatic ccRCC and have shown improvement in clinical outcomes including ORR, progression-free survival (PFS) and OS (Table 1) (76–79). These combinations are now commercially available and have transformed systemic therapy options for patients with mRCC.

As TKI-ICI combinations become standard of care, new challenges have arisen. These challenges include identifying patients most likely to benefit from TKI-ICI combinations versus pure immunotherapy combinations, understanding resistance mechanisms, and defining treatment options at progression. Trials such as COSMIC-313 (NCT03937219) and PDIGREE (A031704, NCT03793166) are actively investigating these questions (80). Other important clinical questions include the role and timing of consolidative nephrectomy, which is being explored in the PROBE trial (NCT04510597), and the potential of triplet therapy combinations. A favored agent for triplet therapy is the recently approved HIF-2α inhibitor, belzutifan, which is highly specific and particularly well tolerated (NCT04736706).

Clinical prognostication with IMDC criteria has been used for patient stratification in all first-line immunotherapy trials of mRCC. These risk factors include short time to systemic therapy, poor performance status, anemia, neutrophilia, thrombocytosis and hypercalcemia (81). IMDC favorable risk disease (no risk factors) seems to be preferentially driven by angiogenesis compared to IMDC intermediate risk (1–2 risk factors) or poor risk disease (>3 risk factors) (81). Notably several of these factors have been linked to RCC biology and may reflect a systemic inflammatory state induced by the tumor (82). Empirical analyses of the TME identified inflamed and uninflamed RCCs and found an association between inflammed tumors with thrombocytosis and anemia (82). Interestingly, patients with intermediate/poor risk disease may benefit the most from ICIs (83,84).

Biomarkers to select optimal first-line therapy are direly needed. Biomarkers have been sought from histological analyses, transcriptomic data as well as circulating cell populations (Table 2). The most compelling histologic determinant is sarcomatoid histology with several cohorts from phase 3 trials (most notably Checkmate 214) showing improved outcomes with immunotherapy compared to sunitinib (85–91). Indeed, sarcomatoid de-differentiated tumors have been profiled as RCC variants in several series (92–94). These tumors have aggressive clinical course and are associated with poor prognostic mutations in TP53, PTEN, RELN, BAP1, CDKN2A, NF2 (and other alterations in the Hippo pathway), as well as upregulation of MYC. Despite these poor prognostic markers, sarcomatoid de-differentiation is associated with responsiveness to ICI, an observation that has been extended to real-world cohorts such as the IMDC dataset (94).

Table 2.

Summary of putative predictive biomarkers for metastatic renal cell carcinoma.

Biomarker	Trial	Conclusion	Reference
Sarcomatoid histology	Checkmate 214	Improved outcomes (OS, PFS and ORR) with ipilimumab + nivolumab compared to sunitinib in patients with sRCC	85
IMDC risk	Checkmate 214	IMDC intermediate/poor risk has greater OS benefit for patients treated with ipilimumab + nivolumab; includes those with even 1 IMDC risk	84,90
Myeloidhigh expression signature	Checkmate 214	Trended toward PFS improvement in ipilimumab/nivolumab cohort	91
T-effector cell signature	IMmotion 150/151	Improved PFS in patients with tumors expressing high levels of Teff signature in the atezolizumab + bevacizumab arm	96
Sarcomatoid histology	IMmotion 151	Improved PFS with atezolizumab + bevacizumab versus sunitinib in patients with sRCC	89
Tumor gene expression clusters: Teff, cell cycle, small nucleolar RNA	IMmotion 151	Trended toward PFS improvement with atezolizumab + bevacizumab versus sunitinib	96
Sarcomatoid histology	JAVELIN Renal 101	Improved PFS with avelumab + axitinib versus sunitinib in patients with sRCC	88
Tumor invasive margin (IM)	JAVELIN Renal 101	Improved PFS in patients with ≥ median tumor IM surface area in the avelumab + axitinib cohort	95
JAVELIN Renal 101 Immuno signature	JAVELIN Renal 101/100	Improved PFS in patients with tumors expressing high levels of Immuno signature in the avelumab + axitinib arm	95
JAVELIN Renal 101 Angio signature	JAVELIN Renal 101	Improved PFS in patients with tumors expressing high levels of Angio signature in the sunitinib arm	95
Neutrophil-to-eosinophil ratio (NER)	NA	Improved OS in patients with NER < median treated with ipilimumab + nivolumab	99
HLA-I evolutionary divergence (HED)	NCT02501096	Improved OS in patients with high HED compared with low HED with lenvatinib + pembrolizumab	100
HLA-A*03	NA	Decreased OS after ICI treatment in patients with HLA-A*03	101
Pancreatic metastases (PM)	NA	Improved PFS in patients with PM compared to no PM with TKI vs. nivolumab	102
Glandular metastases (GM)	NA	Improved OS and PFS in patients with GM compared to without GM on TKI, but not ICI	103

sRCC: sarcomatoid renal cell carcinoma; ICI: immune checkpoint inhibitor; OS: overall survival; PFS: progression free survival; ORR: objective response rate; TKI, tyrosine kinase inhibitor; NER: neutrophil-to-eosinophil ratio; HED: HLA-I evolutionary divergence; IM: invasive margin; PM: pancreatic metastases; GM: glandular metastases; NA: not applicable.

For ccRCC, the most promising transcriptomic biomarkers have emerged from the IMmotion 151 trial (7 tumor clusters), which built upon the preceding groundbreaking IMmotion 150 study (25,95,96) as well as the Javelin Renal 101 trial (a 26-gene signature). The IMmotion 151 study clustered tumors into predominantly angiogenic and immunogenic gene expression signatures and showed differential responsiveness to sunitinib versus atezolizumab-bevacizumab (96–98). These clusters offer another level of prospective patient selection for clinical trials. A more practical biomarker, the neutrophil-to-eosinophil ratio, has also emerged from a multicenter cohort treated with ipilimumab-nivolumab as potentially predictive of ICI response (99). In addition, germline HLA diversity may correlate with response as shown in an early phase 2 study of lenvatinib-pembrolizumab (100) and multiple phase 3 studies (101). Lastly, metastatic tropism and sites of metastases may also be relevant for treatment selection. Specifically, RCC metastatic to the pancreas and possibly other glandular structures seem more angiogenically-driven and may need therapies targeting the VEGF axis (102,103). These biomarkers have largely arisen from exploratory analyses and need prospective validation before implementation in the clinic.

Non-clear cell renal cell carcinoma

Non-clear cell RCC (nccRCC) accounts for 25% of all RCCs and comprises multiple tumor types with limited treatment options. Multicenter retrospective cohorts have shown ORRs ~20% and disease control rates ~50% with nivolumab (104). Similar disease control rates have been shown with ipilimumab-nivolumab (105). While these rates are lower than for ccRCC, some nccRCC patients do respond to ICIs. A phase 2 trial of first-line pembrolizumab in 165 patients with nccRCC (71.5% with papillary RCC, 12.7% with chromophobe RCC, and 15.8% with unclassified RCC) showed an ORR of 26.7% and disease control rate of 43%, with a median duration of response of 29 months (106). While these clinical responses are promising, each histological subset warrants further investigation given the different molecular drivers. For example, chromophobe RCC appears to be particularly resistant to ICIs.

Novel targets and combinations

While upfront therapies are being intensified with triplet combinations, further developments will depend upon identifying resistance mechanisms. Promising ongoing trials include bispecific antibodies and chimeric antigen receptor (CAR)-T cells. CAR-T cell therapies have received FDA approval for hematologic malignancies, but solid tumors represent a greater challenge. Nevertheless, several CARs against antigens expressed by RCC have been developed. CD70 (a CD27 ligand) is expressed on the surface of several solid tumors including RCC (107). Treatment of patients with CD70+ RCC (or other solid tumors) with an anti-CD70 CAR T cell is currently being evaluated (NCT02830724, NCT04438083) (108). CAR T cell therapies targeting other tumor antigens, including the tyrosine kinase-like orphan receptor 2 (ROR2), AXL, and carbonic anhydrase IX (CAIX), have been evaluated in patients with solid tumors including RCC (NCT03393936, NCT04969354). As an alternative to CAR T cell therapy, T cell engaging bispecific antibodies that bind both CD3-expressing T cells and tumor antigens may target tumor cells for T-cell mediated lysis (109,110). The identification of RCC-associated antigens and pan-tumor antigens has facilitated the development of therapeutic vaccines for RCC patients. In addition, personalized vaccines using neoantigens have demonstrated potential in patients with melanoma and are being evaluated for other cancer types including RCC (111,112). Finally, combinations directed against multiple immune checkpoint targets such as the recent developments combining ICIs with anti-LAG3 (relatlimab approval in melanoma) or anti-TIGIT (tiragolumab breakthrough designation in NSCLC with ongoing phase 3 trial, NCT04294810) will enhance the efficacy of ICI therapy and are promising for the future of immune responsive tumors like RCC (113,114).

Conclusions and future directions

Immunotherapy options for mRCC have made considerable progress and have changed the outlook for many patients with mRCC. Treatment options have progressed from cytokines to targeted activation of cytotoxic T-cells using ICIs. Patient survival has improved significantly. Ongoing challenges include the identification of biomarkers for treatment selection, understanding mechanisms of resistance, and determining algorithms to optimally sequence treatments. Future treatments for mRCC depend on deepening our understanding of the TME and of drivers of ICI resistance, as well as dissecting complex interactions between the tumor and the host immune system.