In contrast to the wide array of available therapies in the metastatic setting of renal cell carcinoma, in patients with localized disease, the results of trials investigating adjuvant systemic therapy after nephrectomy have so far been underwhelming.
Ziad Bakouny, MD, MSC
According to the 2018 Globocan report, over 400,000 cases of renal cell carcinoma (RCC) are diagnosed yearly, resulting in more than 175,000 deaths around the world.1 One-third of patients with RCC have unresectable locally advanced or metastatic disease, and 20% to 40% of patients presenting with localized disease eventually have metastatic relapses.2 Among the histological types of RCC, clear-cell RCC is the most frequent type by a wide margin.
Significant advances in the treatment of metastatic RCC occurred during the past decade, with multiple targeted therapies added as treatment options for this disease. These therapies were either of the antiangiogenic class, including sunitinib (Sutent), sorafenib (Nexavar), axitinib (Inlyta), pazopanib (Votrient), cabozantinib (Cabometyx), lenvatinib (Lenvima), and bevacizumab (Avastin), or the mammalian target of rapamycin (mTOR) inhibitor class, such as temsirolimus (Torisel) and everolimus (Afinitor).3-5 However, some tumors are inherently resistant to these therapies and most tumors that initially respond will eventually progress.
Furthermore, in contrast to the wide array of available therapies in the metastatic setting, in patients with localized disease, the results of trials investigating adjuvant systemic therapy after nephrectomy have so far been underwhelming. Sunitinib is currently the only therapy shown to confer a statistically significant benefit compared with placebo in the adjuvant setting, but only for diseasefree survival, not overall survival, and with conflicting results between 2 pivotal trials.6-9
Immune checkpoint inhibitors (ICIs) are a new class of drugs designed to augment a patient’s immune response against a tumor by targeting immune checkpoints that physiologically limit immune activation. The immune checkpoints that are the targets of FDA-approved drugs are PD-1, PD-L1, and CTLA-4.10 ICIs have thus far shown significant survival benefit in a large number of tumor types, including RCC,5 and at the time of this writing, 6 PD-1/PD-L1 inhibitors and 1 CTLA-4 inhibitor have been approved by the FDA. A few, such as nivolumab (Opdivo) and ipilimumab (Yervoy), are approved for use in RCC, although others are undergoing testing in this disease type.In the metastatic setting, nivolumab was the first ICI to show significant efficacy in the treatment of RCC. The landmark CheckMate-025 phase III trial randomized patients with at least 1 prior line of systemic therapy to either nivolumab or everolimus, showing a hazard ratio (HR) for overall survival (OS) of 0.73 (98.5% CI, 0.57-0.93; P = .002) and an overall response rate (ORR) of 25%.11 Subsequently, atezolizumab (Tecentriq) and pembrolizumab (Keytruda) showed ORRs of 25% and 38.2% in treatment-naïve patients in the phase II IMmotion150 and KEYNOTE-427 trials, respectively.12,13 Although these results were considered significant advances in the treatment of metastatic RCC, it was hypothesized that patient treatment outcomes could be further improved by combining ICIs with other drugs.To that end, multiple ICI-based combinations have been tested. In the pivotal CheckMate-214 trial, the combination of nivolumab and ipilimumab significantly improved OS in patients with intermediate-/poor-risk untreated metastatic RCC (HR, 0.63; 99.8% CI, 0.44-0.89; P <.001) and in the overall patient population (HR, 0.68; 99.8% CI, 0.49-0.95; P <.001) compared with sunitinib. In the intermediate-/poor-risk subgroup, nivolumab plus ipilimumab conferred a significant benefit over sunitinib for ORR (42% vs 27%; P <.001), with only a trend in favor of nivolumab and ipilimumab for progression-free survival (PFS) (HR, 0.82; 99.1% CI, 0.64-1.05; P = .03). However, in the favorable-risk population, the nivolumab with ipilimumab combination led to significantly worse PFS (HR, 2.18; 99.1% CI, 1.29-3.68; P <.001) and ORR (29% vs 52%; P <.001) compared with sunitinib,14 leading to the current FDA approval only for those presenting with intermediate- and poor-risk disease.
A second approach to ICI-based combinations for RCC is combining ICIs with inhibitors of the vascular endothelial growth factor (VEGF) pathway, which could augment the efficacy of ICIs by attenuating tumor immunosuppression.15 In particular, in previously untreated patients with metastatic RCC, the recent phase III IMmotion 151 study showed a significant PFS advantage (HR, 0.74; 95% CI, 0.57-0.96; P = .02) of an atezolizumab-plusbevacizumab combination over sunitinib in the subset of patients with PD-L1 ≥1% on tumor-infiltrating immune cells, as well as an ORR of 43% (vs 35% for sunitinib).16 Phase I clinical trials have yielded promising results with combinations of pembrolizumab plus axitinib (ORR,73%, untreated patients),16 pembrolizumab plus lenvatinib (ORR, 63%, untreated and previously treated patients),17 avelumab (Bavencio) plus axitinib (ORR, 58%, untreated patients),18 and cabozantinib with nivolumab plus/minus ipilimumab (ORR, 54% in the sarcomatoid RCC subset of the trial, previously treated patients) (Figure).19
Multiple ongoing trials are also awaited to further define optimal ICI-based combinations for the treatment of metastatic RCC, including phase III trials of cabozantinib plus nivolumab compared with sunitinib (CheckMate 9ER, NCT03141177), axitinib plus avelumab compared with sunitinib (JAVELIN Renal 101, NCT02684006), lenvatinib plus pembrolizumab or lenvatinib plus everolimus compared with sunitinib (NCT02811861), and axitinib plus pembrolizumab compared with sunitinib (KEYNOTE-426, NCT02853331).Given the disappointing results of most trials investigating adjuvant treatment for RCC, improving the cure rates of surgery for localized RCC constitutes an important unmet need. Multiple trials are investigating whether adjuvant ICI therapy could improve outcomes for localized RCC, including IMmotion-010 (NCT03024996, atezolizumab vs placebo), KEYNOTE-564 (NCT03142334, pembrolizumab vs placebo), RAMPART (NCT03288532, durvalumab plus tremelimumab vs durvalumab vs observation), and CheckMate-914 (NCT03138512, nivolumab plus ipilimumab vs placebo).
A novel approach to using ICIs in the localized setting involves perioperative immune checkpoint inhibition both in the neoadjuvant and adjuvant settings. The rationale is that more tumor antigens are available to participate in the activation of the immune system before tumor resection, optimizing the ICI-induced response and hypothetically improving cure rates.20 PROSPER RCC (NCT03055013) is an ongoing phase III trial that randomizes patients with high-risk localized RCC to either 1 month of neoadjuvant nivolumab followed by nephrectomy and 9 months of adjuvant nivolumab versus nephrectomy alone. The results of this perioperative ICI trial along with those of the adjuvant ICI trials could add important treatment options for patients with localized but high-risk disease.Considering that tumor responses to ICIs and ICI-based combinations are relatively limited, with only half of patients responding at best, identifying patients most likely to respond to ICIs is of paramount importance. PD-L1 expression has been proposed as a possible biomarker of response to ICI; PD-L1 has been found to be overexpressed in about a quarter of clear-cell RCC tumors and to be associated with poorer outcomes.21 However, although PD-L1 expression ≥1% tended to be associated with superior outcomes, with the combination of nivolumab plus ipilimumab in CheckMate-214,14 it was only of prognostic value in patients treated with nivolumab alone (vs everolimus) in CheckMate-025.11
Moreover, discrepancies in where PD-L1 is measured (tumor cells, tumor-infiltrating immune cells, or both types of cells), differences in cutoffs, and the type of antibody used may limit the role of PD-L1 as a universal biomarker for response to ICIs in advanced RCC. Some specific molecular alterations have been shown to predict response to ICIs in RCC, with one study’s results showing that loss of function of the PBRM1 gene is associated with a higher rate of clinical benefit in advanced RCC.22 In solid tumors in general, other biomarkers for ICI response are under investigation, such as tumor mutational burden,23 tumor immune infiltrate type,24 JAK/ STAT pathway alterations,22,25 and composition of the microbiome of the digestive tract.26Despite the wealth of new agents in the treatment of metastatic RCC, in both the first- and second-line settings, there are little data about optimal sequences of regimens. Multiple ongoing trials are therefore attempting to address such questions, with NIVOSWITCH (NCT02959554) evaluating a planned switch post TKI response, from TKI to nivolumab, compared with continuation of TKI; SUAVE (NCT03035630) comparing an avelumab/sunitinib sequence with sunitinib/ avelumab; and CheckMate-800 (NCT03029780) comparing coadministration and sequential administration of nivolumab/ipilimumab. The results of such trials should allow clinicians to better define the optimal sequencing of the different ICI drugs and targeted therapies.
Furthermore, to minimize toxicities and optimize efficacy, studies are evaluating whether ICI-regimen administration can be tailored to each patient’s response: HCRN GU16-260 (NCT03117309), OMNIVORE (NCT03203473), TITAN RCC (NCT02917772), and SAKK 07/17 (NCT03297593) are evaluating nivolumab with conditional addition or discontinuation of ipilimumab depending on response or progression. Moreover, CASE 5816 (NCT03126331) is evaluating administration of single-agent nivolumab tailored to the extent of tumor shrinkage.
Another common problem in clinical practice is whether patients with brain metastases can be efficiently and safely treated with ICIs. Unfortunately, there are little data27 to address ICI therapy in this significant subset of patients. CheckMate-21414 and -02511 excluded patients with brain metastases, whereas IMmotion-15012 and -15116 excluded all active brain metastases and have not specifically reported outcomes on patients with brain metastases. Data from the French NIVOREN study28 suggest that beyond first-line treatment with nivolumab for metastatic RCC yields an intracranial ORR of 16%, with 18/55 (32.7%) patients presenting clinical neurological deterioration requiring steroid therapy. Of these 18 patients, 17 (94.4%) had not received prior local therapy (radiotherapy or surgery) for their brain metastases, thus suggesting that patients with nonpretreated brain metastases may have worse outcomes. The Italian expanded access program study29 reported an ORR of 18.7% for patients with brain metastases (compared with 23.2% in the overall population of patients) but did not specifically report intracranial response or the outcomes of patients with nonpretreated brain metastases.
Treatment toxicity is another consideration for all classes of oncologic therapy, but the immune-related adverse events (irAEs) of ICIs merit special discussion considering their atypical characteristics.30 Immune-related adverse events often occur within weeks to months of the start of therapy but may be delayed or occur after discontinuation of the ICI.30,31 These irAEs probably reflect a state of general immune activation but the precise mechanisms underlying these irAEs are still unclear.30 The mainstay of treatment for irAEs is immunosuppression with glucocorticoids, with additional agents, such as infliximab (Remicade), as necessary. The ICIs should be temporarily held in some cases or definitely discontinued when the irAEs are serious.31
Although many patients who do not develop irAEs benefit from ICIs, some studies’ findings have suggested that patients who develop irAEs may have superior responses compared with patients who do not.32,33 Thus, it is currently unclear whether patients who discontinue ICIs because of irAEs should be restarted on the same drug or a drug with a different mechanism of action (eg, PD-1 inhibitor vs CTLA-4 inhibitor30); however, reports suggest that patients who discontinue because of irAEs may have prolonged clinical benefit without resumption of therapy.34 Moreover, with the proliferation of ICI-based combinations, treatment toxicities have exponentially increased and the trade-off between improved efficacy and increased toxicity should be tailored to the characteristics of each patient.35ICIs represent promising agents for the management of metastatic RCC. Although ICIs and ICI-based combinations will become part of the standard of care for patients with metastatic RCC, studies are still ongoing to determine the role of ICIs in the management of highrisk localized RCC. Biomarkers of response are needed to better define which patients respond to these therapies. Trials investigating treatment sequencing, as well as tumor response—based adaptation of therapy, could also help optimize efficacy and minimize irAEs. Furthermore, the efficacy of ICIs in subsets of patients who are known to respond poorly to therapy, such as those with brain metastases and non–clear cell RCC, remains to be fully defined. Optimal management of irAEs and whether to restart therapy after discontinuation because of irAEs are still the focus of ongoing research.