Checking the Hippo in Sarcomatoid Renal Cell Carcinoma with Immunotherapy

In this issue of Clinical Cancer Research, Tannir and colleagues report on a subset analysis of CheckMate 214 trial, specifically the subset of patients with advanced renal cell carcinoma and sarcomatoid features, and their clinical outcomes with ipilimumab-nivolumab versus sunitinib (1).

Sarcomatoid renal cell carcinoma (sRCC) refers to a form of RCC characterized histologically by the presence of spindle-shaped mesenchymal looking cells typically in a portion of the tumor. sRCC may arise in any histologic subtype of RCC, likely via dedifferentiation from a common cell of origin. Clinically, sRCC is associated with aggressive disease and a poor prognosis. Compared with patients with non-sRCC, patients with sRCC more frequently have intermediate or poor International Metastatic Renal Cell Carcinoma Database Consortium (IMDC) risk disease, shorter time to relapse, lower objective response rates (ORR), and less frequently receive second- or third-line treatments (2).

Therapeutic options for sRCC have historically been limited. In the phase II ECOG 8802 trial of doxorubicin and gemcitabine in sRCC, the combination achieved an ORR of just 16%, complete response rate (CR) of 2.7%, median overall survival (OS) of 8.8 months, and median progression-free survival (PFS) of 3.5 months (2). The record of mTOR inhibitors in sRCC is also poor, with an estimated ORR of 13% (2). Sunitinib plus gemcitabine produced only somewhat more success, with ORR ranging 20%–26%, CR around 3%, and median OS of approximately 10 months in phase II trials of the combination (2). Thus, there is a significant unmet need for effective therapies for sRCC.

In this issue, Tannir and colleagues report on a post hoc analysis of the subgroup of patients with sRCC from CheckMate 214 with intermediate/poor risk disease. The study previously demonstrated the long-term superiority of first-line ipilimumab-nivolumab versus sunitinib in advanced clear cell RCC (3). A total of 145 patients with sRCC were identified post hoc on the basis of local pathology reports or central independent review. Central review determination of any percent sarcomatoid histology was considered sarcomatoid positive based on consensus between three pathologists blinded on clinical outcomes, with discrepancies resolved on consensus review. A total of 139 patients with intermediate/poor risk disease were identified. Baseline characteristics of the sRCC subgroup were similar between the treatment cohorts based on gender, IMDC risk group, tumor PD-L1 levels, prior nephrectomy, and sites of metastases. Outcomes were assessed over a median follow-up time of 47.7 months.

Confirmed ORR was 61% with ipilimumab-nivolumab in IMDC intermediate- or poor-risk sRCC (N = 74) versus 23% with sunitinib (N = 65; P < 0.0001). Notably, CR with ipilimumab-nivolumab was 19% versus 3% with sunitinib. Median OS was not reached for ipilimumab-nivolumab versus 14.5 months for sunitinib [HR, 0.45 (0.3–0.7); P = 0.0004]. Median PFS was 26.5 months with ipilimumab-nivolumab versus 5.1 months with sunitinib [HR, 0.54 (0.3–0.9); P = 0.0093]. The OS, median PFS, and ORR benefits of ipilimumab-nivolumab versus sunitinib were seen regardless of tumor PD-L1 levels (⁠|\ge $|1% vs. <1%), although the magnitude of the OS, median PFS, and ORR was higher in those with PD-L1 |\ge \ $|1%. The safety profile of ipilimumab-nivolumab in patients with sRCC was consistent with prior safety reports in the overall population (3).

These findings add to other subgroup analyses of sRCC in recent first-line phase III trials supporting the efficacy of immune checkpoint inhibitor (ICI)-based regimens compared with sunitinib. These include patients with sRCC treated with pembrolizumab-axitinib (N = 51) in the subset analysis of the KEYNOTE-426 trial (12-month OS of 83.4%, ORR of 58.8%, and CR of 11.8%; ref. 4) and with atezolizumab-bevacizumab (N = 68) in the subset analysis of IMmotion151 (12-month OS of 69%, ORR of 49%, and CR of 10%; ref. 5). Finally, the subset of patients with sRCC treated in JAVELIN Renal 101 trial also showed similarly improved outcomes with avelumab-axitinib (12-month OS 83%, ORR of 47%, and CR of 4%; ref. 6; Table 1). Indeed, given the ICI activity in sRCC in these combinations, ICI-based regimens have a crucial role in first-line treatment of sRCC.

Notably, in the CheckMate 214 study, the ORR including CR in the sRCC subgroup was numerically greater than in the overall IMDC intermediate- or poor-risk population treated with ipilimumab-nivolumab (ORR 61% vs. 42% and CR 18% vs. 11%; ref. 3). Inasmuch as sarcomatoid changes are thought to arise in the context of a preexisting RCC, it is interesting that tumors with this form of dedifferentiation appear to be particularly sensitive to ICI. Whether this reflects a heretofore unidentified subtype of RCC with a tendency toward sarcomatoid dedifferentiation and intrinsic features that render it more susceptible to an immune-mediated attack, or whether sarcomatoid dedifferentiation by itself is responsible for the heightened responsiveness to ICI is unclear. Although, if the latter, only the sarcomatoid component might be expected to respond.

The susceptibility of sRCC to ICIs is consistent with our emerging understanding of the immune profile of sRCC. PD-L1 expression and intratumoral CD8-positive cell density are higher in sRCC versus non-sRCC (2), suggesting the presence of preexisting intratumoral T cells exhausted by PD-L1 engagement (Fig. 1). Subset analysis of IMmotion151 also found higher PD-L1⁺ and T effector gene expression in the sRCC group (5).

An important factor relevant to ICI response is the tumor microenvironment (TME). An empiric characterization of the TME in RCC using matched trios of bulk tumor, patient-derived xenograft, and normal tissue, revealed a highly inflamed pan-RCC TME subtype characterized by extensive immune infiltrate (7). This inflamed subtype was enriched in aggressive RCCs with BAP1 mutations, which are particularly prevalent in sRCC (7). Interestingly, inflammation in the TME correlated with thrombocytosis and anemia (7), two validated IMDC criteria, suggesting that thrombocytosis and anemia may potentially represent systemic manifestations of an inflammatory response elicited by the tumor. This could potentially explain the higher frequency of IMDC risk factors in sRCC, which are characteristically inflamed.

Why did some patients with sRCC respond to ipilimumab-nivolumab while others did not? This heterogeneity in treatment response may relate to the genetic heterogeneity of sRCC. Mutations enriched in sRCC involve TP53; Hippo pathway members, in particular NF2; CDKN2A/2B; the PI3K regulator PTEN; and the deubiquitinase BAP1 (Fig. 1; refs. 8–10). Transcriptional profiling of sRCCs has also identified an enrichment of TGFβ signaling (10), potentially driving resistance to immunotherapies in a subset of these patients. Further work remains to understand the molecular drivers of ICI resistance.

In sum, these findings support the use of ipilimumab-nivolumab, and more broadly the incorporation of ICIs, in the treatment of sRCC. Given the multiplicity of pathways implicated, additional responses in sRCC may be gained from combination therapies. Elucidation of how the tumor and TME characteristics affect the traditional IMDC risk model remains key. The dramatic change in outcomes brought about with ICIs on sRCC, which are disproportionally associated with IMDC intermediate and poor risk features, represent a welcome complexity in the interpretation of the prognostic model in the ICI era. Mutations, gene expression, and TME features likely represent molecular determinants of response and may be helpful in the future to optimize treatment selection in metastatic RCC.

N. Agarwal reports consultancy to Astellas, AstraZeneca, Bayer, Bristol-Myers Squibb, Clovis, Eisai, Eli Lilly, EMD Serono, Exelixis, Foundation Medicine, Genentech, Janssen, Merck, Nektar, Novartis, Pfizer, Pharmacyclics, and Seattle Genetics and research funding (to institution) from AstraZeneca, Bavarian Nordic, Bayer, Bristol-Myers Squibb, Calithera, Celldex, Clovis, Eisai, Eli Lilly, EMD Serono, Exelixis, Genentech, GlaxoSmithKline, Immunomedics, Janssen, Medivation, Merck, Nektar, New Link Genetics, Novartis, Pfizer, Prometheus, Rexahn Pharmaceuticals, Roche, Sanofi, Seattle Genetics, Takeda, and Tracon. J. Brugarolas reports grants and personal fees from Arrowhead Pharmaceuticals (received research funding, served as a consultant) and personal fees from Genentech/Roche (served as a consultant), Exelixis (served as a consultant), and Nektar Therapeutics (served as a consultant) outside the submitted work, as well as a patent issued for biomarkers of response to HIF-2-alpha inhibition in cancer and methods for the use thereof (U.S. Patent No. 15/761,534, based on International Patent Application No. PCT/US2016/052118). T. Zhang reports grants from Acerta, Novartis, Merrimack, AbbVie, Regeneron, Mirati Therapeutics, Omniseq, Inc., PGDx, and Astellas, personal fees from Genentech Roche, Exelixis, Sanofi Aventis, Amgen, Bristol-Myers Squibb, Foundation Medicine, Pharmacyclics, MJH Associates, IQVIA, Pacific Genuity, China Medical Tribune, and Nanorobotics (spouse), grants and personal fees from Merck, Janssen, AstraZeneca, Pfizer, and Seattle Genetics, and other from Capio Biosciences (employment/stockholder–spouse) and Archimmune Therapeutics (employment/stockholder–spouse) outside the submitted work. No disclosures were reported by the other author.

J. Brugarolas was supported by an NCI SPORE grant (P50 CA196516).