Immunotherapy: A new standard in the treatment of metastatic clear cell renal cell carcinoma

Maja Popovic, Gorana Matovina-Brko, Masa Jovic, Lazar S Popov ic

Maja Popovic, Gorana Matovina-Brko, Masa Jovic, Lazar S Popovic, Department of Medical Oncology, Oncology Institute of Vojvodina, University of Novi Sad, Sremska Kamenica 21204, Serbia

Maja Popovic, Lazar S Popovic, Faculty of Medicine, University of Novi Sad, Novi Sad 21000,Serbia

Abstract Renal cell cancer (RCC) represents 2%-3% of all adulthood cancers and is the most common malignant neoplasm of the kidney (90%). In the mid-nineties of the last century, the standard of treatment for patients with metastatic RCC was cytokines. Sunititib and pazopanib were registered in 2007 and 2009, respectively,and have since been the standard first-line treatment for metastatic clear cell RCC(mccRCC). Renal cell cancer is a highly immunogenic tumor with tumor infiltrating cells, including CD8+ T lymphocytes, dendritic cells, natural killer cells(NK) and macrophages. This observation led to the design of new clinical trials in which patients were treated with immunotherapy. With the growing evidence that proangiogenic factors can have immunomodulatory effects on the host’s immune system, the idea of combining angiogenic drugs with immunotherapy has emerged, and new clinical trials have been designed. In the last few years,several therapeutic options have been approved [immunotherapy and immunotherapy/tyrosine kinase inhibitors (TKI)] for the first-line treatment of mccRCC.Nivolumab/ipilimumab is approved for the treatment of patients with intermediate and poor prognoses. Several checkpoint inhibitors (pembrolizumab,nivolumab, avelumab) in combination with TKI (axitinib, lenvatinib, cabozantinib) are approved for the treatment of patients regardless of their International mRCC Database Consortium prognostic group and PD-L1 expression. There is no specific and ideal biomarker that could help in selecting the ideal patient for the appropriate first-line treatment.

Key Words: Renal cell carcinoma; Immunotherapy; Checkpoint inhibitors; Biomarkers;Tumor microenvironment; Programmed cell death 1 receptor

INTRODUCTION

Renal cell cancer (RCC) represents 2%-3% of all adulthood cancers and is the most common malignant neoplasm of the kidney (90%)[1]. Clear cell cancer (75%) is the most prevalent histological subtype of RCC, followed by papillary (10%), chromofobe(5%), collecting ducts (0.4%-1.8%) and unclassified (4%-6%)[2]. RCC typically occurs in the fifth and sixth decade of life and is twice as frequent in men than in women[3]. At the time of diagnosis, one-third of all patients have metastatic disease, while a quarter of all patients, with initially localized disease, relapse after nephrectomy[4]. According to two prognostic models, Memorial Sloan Kettering Cancer Center (MSKCC) and International mRCC Database Consortium (IMDC), metastatic RCC (mRCC) patients can be divided into 3 prognostic categories: favorable, intermediate and poor risk[5,6](Table 1).

Table 1 Poor prognostic factor

In the mid-nineties of the last century, the standard of treatment for patients with metastatic RCC was cytokines, typically interferon-alpha and interleukin 2. Beside the high toxicity profile of cytokines, patients who were treated achieved an objective response rate (ORR) of 10-20%, while the median overall survival (OS) was 11-14 mo[7-9].

Renal clear cell carcinoma is commonly associated with Von Hippel-Lindau (VHL)gene mutations (70% of patients) located on chromosome 3p and mediates cell apoptosis in response to hypoxia[10,11]. If this mutation is present, apoptosis does not occur, hypoxia-induced factor (HIF) accumulates and activates vascular endothelial growth factor (VEGF), and platelet growth factor (PDGF) and others engage in the angiogenesis process, which is one of the key promoters of cell growth in RCC[12].This knowledge leads to the development of new antiangiogenetic drugs. Other mutations, such as PBRM1 (40%), SETD28 (15%) and BAP1 (15%), have recently been discovered. Sunititib and pazopanib were registered in 2007 and 2009, respectively and have been the standard first-line treatment for mRCC ever since. The median survival of patients treated with these drugs is 24-29 mo, while the objective response rate(ORR) is 30%-33%[13,14].

RCC is a highly immunogenic tumor with infiltrating cells, including CD8+ T lymphocytes, dendritic cells, natural killer cells (NK) and macrophages. This observation led to the design of new clinical trials in which patients were treated with immunotherapy[15]. Checkpoint inhibitors are monoclonal antibodies targeting the link between programmed cell death protein 1 (PD-1) and its ligands PD-L1 and PDL2[16]. The PD-1 receptor is located on T cells, while PD-L1 and PD-L2 are present on other immune cells. The ligand can be found on both tumor cells and immune infiltrate cells, allowing them to bind to the PD-1 receptor of T-cells and escape the host immune response[17,18]. Checkpoint inhibitors block this interaction and permit the host’s immune response to the tumor[16].

Nivolumab is humanized PD-1 monoclonal antibody. The first data on nivolumab in mRCC were the results of the phase I Checkmate 033 trial, where nivolumab wasinvestigated in pretreated patients. The objective response rate was 24%; after a median follow-up of 63.9 mo, the ORR was 29%, the median duration of response(DOR) was 12.9 mo, and the median OS was 22.4 mo[19]. In the phase 2 trial, nivolumab was again investigated in pretreated mRCC patients. Patients received 0.3 mg/kg, 2 mg/kg or 10 mg/kg nivolumab. There was no difference in PFS in these subgroups. At 3 years, ORR was 21% while OS was 41%[20]. The phase 3 trial,CheckMate 025, investigated nivolumab in comparison to everolimus in pretreated patients. The primary endpoint was OS, while the secondary endpoints were response rates and safety profile. The median OS in patients treated with nivolumab was 25 mo,compared to 1.6 mo with everolimus (HR 0.73). Differences in OS were recorded across all subgroups of patients regardless of PD-L1 expression. The objective response rate was 25% in the nivolumab cohort and 5% in the everolimus cohort. There was no significant difference in PFS of 4.6vs4.4 mo for nivolumab and everolimus,respectively. Grade 3 and 4 adverse events were reported in 19% of patients in the nivolumab group and 37% of patients in the everolimus group[21]. The results of this trial led to FDA approval of nivolumab as a second-line treatment of mccRCC.

In April 2018, nivolumab and ipilimumab combination therapy was approved by the FDA for the first-line treatment of intermediate- and poor-risk mRCC patients.This approval was a result of the phase 3 trial, CheckMate 214, which compared nivolumab and ipilimumabvssunitinib in treatment-na?ve patients. The trial included 1096 patients, 847 of whom were intermediate- and poor-risk IMDC risk groups.Patients were randomized 1:1. The primary endpoints were OS, PFS and ORR in intermediate- and poor-risk patients, while the secondary endpoints were OS, PFS and ORR in the intended-to-treat (ITT) population. Intermediate- and poor-risk patients in the nivolumab/ipilimumab group had significantly longer PFS than those in the sunitinib group. The favorable-risk prognostic group had longer PFS when treated with sunitinib. Patients with PD-L1 expression ＞ 1% had significantly longer PFS when treated with nivolumab/ipilimumabvssunitinib, while the treatment groups did not differ in patients with PD-L1 ＜ 1%. Nivolumab/ipilimumab significantly prolonged patient OS compared to sunitinib. There were 46% grade 3-4 adverse events in the nivolumab/ipilimumab groupvs63% in the sunitinib group[22]. After 48 mo of follow-up, patients in the intermediate- and poor-risk groups treated with nivolumab/ipilimumab achieved significantly longer overall survival[23].

Nivolumab also proved efficacious in patients with brain metastasis: the ORR was 12%, and the PFS was 2.7 mo[24]. When nivolumab was combined with ipilimumab,the ORR and PFS were 29% and 9 mo, respectively[25].

Angiogenesis is one of the key initiators of disease in RCC, which itself is an immunogenic tumor. In patients with VHL gene mutations, instead of apoptosis, HIF accumulates and activates VEGF and PDGF, which mediate the activation of the angiogenesis process[10-12]. It has been shown that accumulation of VEGF leads to suppression of the host’s immune response. It also interferes with monocyte differentiation into mature dendritic cells that are essential for the activation of the host’s immune system. VEGF increases the number of myeloid suppressing cells present in the tumor infiltrates that disable the activity of tumor infiltrating lymphocytes, the expression of PD-L1 in dendritic cells, as well as PD-1 and CTLA-4 on immune cells. It inhibits the differentiation of progenitor cells into CD4+ and CD8+ cells.Proangiogenic factors also modify the expression of proteins on endothelial cells,blocking the infiltration of the tumor by immune cells[26,27]. With the growing evidence that proangiogenic factors can have immunomodulatory effects on the host’s immune system, the idea of combining angiogenic drugs with immunotherapy emerged, and new clinical trials have been designed[28].

Atezolizumab is a humanized monoclonal PD-L1 antibody investigated in combination with bevacizumabvssunitinib. After the phase I study reported a 40%ORR , a phase II study was conducted (atezolizumab or atezolizumab/bevacizumab or sunitinib), and ORRs of 32%, 25% and 29%, respectively, were observed. In the ITT population, the PFS difference was not significant, while in the PD-L1-positive patients, a significant difference was noticed in the cohort treated with atezolizumab/bevacizumabvssunitinib. PFS was not significant when atezolizumab alone was compared with sunitinib in PD-L1-positive patients[29]. The phase 3 trial,IMmotion 151, followed these results and compared atezolizumab/bevacizumabvssunitinib in treatment-na?ve patients. Patients were randomized 1:1 according to the MSKCC score, PD-L1 expression (＜ 1%vs＞ 1%), and presence of liver metastases.Patients with sarcomatoid tumor features were also included. The co-primary endpoints were PFS in the PD-L1-positive population and OS in the ITT population.Secondary endpoints were PFS, ORR and duration of response in the ITT population.In the PD-L1-positive patients, PFS was 11.2 mo (atezolizumab/bevacizumab) in comparison to 7.7 mo (sunitinib), HR 0.74. In the ITT population, PFS was 11.2 mo(atezolizumab/bevacizumab)vs8.4 mo (sunitinib), HR 0.83. The ORR in the PD-L1+population was 43% (atezolizumab/bevacizumab)vs35% (sunitinib), while the ORR in the ITT population was 37%vs33% (atezolizumab/bevacizumabvssunitinib)[30].After 24 mo of follow-up, there were no differences in survival (HR 0.93) in the ITT population[31]. Considering the results of IMmotion 150 and 151, data subanalysis was performed according to the molecular profile of tumor tissue. IMmotion 150 patients were classified into angio-high, T effector-high and myeloid-high. The subanalysis showed that angio-high patients had a higher benefit from TKIs and were in the favorable prognostic group, while T effector-high patients had a greater benefit from immunotherapy and were in the intermediate and poor prognostic groups. It was also observed that patients with BAP1 mutations had a worse prognosis and shorter PFS when treated with sunitinib, while patients with PBRM1 mutations had a worse prognosis and shorter PFS when treated with immunotherapy. IMmotion 151 included patients with sarcomatoid features, who generally had a worse prognosis. The results of this subanalysis showed that half of these patients were T effector-high, had higher PD-L1 expression and achieved the highest benefit from immunotherapy[29,32].

Cosmic-021 was a phase 1b trial that investigated the efficacy of atezolizumab in combination with cabozantinib in different solid tumors. One of the cohorts consisted of mccRCC patients. Seventy patients were included in the study: 34 patients were treated with cabozantinib 40 mg, and 36 patients were treated with cabozantinib 60 mg and 1200 mg atezolizumab. Most of the patients were in the intermediate prognostic group. After a median follow-up of 11.5 mo (cabozantinib 60 mg)vs22 mo for cabozantinib 40 mg, the ORR in the cabozantinib 60 mg group was 58%vs47% in the cabozantinib 40 mg group. The median PFS was 19.5 mo (cabozantinib 40 mg) and 20.4 mo (cabozantinib 60 mg). Two years PFS was 67% (cabozantinib 40 mg) and 71%(cabozantinib 60 mg). Treatment-related grade 3-4 adverse events were reported in 71% (cabozantinib 40 mg) and 67% (cabozantinib 60 mg) of the patients. The most common adverse events were hypertension, hypophosphatemia, diarrhea and elevation of liver enzymes[33].

The Contact-03 trial investigating atezolizumab in combination with cabozantinib in patients with mRCC who have progressed on previous immunotherapy is underway[34].

Pembrolizumab is a humanized monoclonal PD-L1 antibody studied in combination with axitinib in a phase 1b trial. The response rate was 73%[35]. In the randomized phase 3 clinical trial (Keynote-426), pembrolizumab/axitinib was compared to sunitinib. Patients were randomized 1:1. The primary endpoints were OS and PFS in the ITT population, while the secondary endpoint was ORR. After 12.8 mo of followup, the one-year OS was 89.9% (pembrolizumab/axitinib)vs78.3% (sunitinib), HR 0.53,P＜ 0.0001. PFS was 15.1 mo (pembrolizumab/axitinib)vs11.1 mo (sunitinib), HR 0.69,P＜ 0.0001. The ORR was 59.3% and 37.5% in the pembrolizumab/axitinibvssunitinib group, respectively. Treatment-related grade 3 adverse events accounted for 75.85% of the patients in the combination cohort. Benefit was observed across all subgroups analyzed regardless of the IMDC risk score or PD-L1 expression[36]. At 27 mo, PFS and OS were significantly longer in all subgroups of patients[37]. Pembrolizumab was investigated in combination with levantinib in a phase 2 trial (Keynote 146) in patients with mccRCC who were previously treated with immunotherapy. The primary endpoint of the trial was an ORR of 51%, a median PFS of 11.7 mo, and a median DOR of 9.9 mo[38]. The phase 3 trial, CLEAR/Keynote 581, investigated pembrolizumab/Lenvatinibvseverolimus/Lenvatinibvssunitinib in patients with mccRCC. The primary endpoint was PFS, while the secondary endpoints were ORR and OS in the ITT population. All three prognostic MSKCC and IMDC risk score groups were included in the trial. After 26.6 mo of follow-up, PFS in the group of patients treated with pembrolizumab/Lenvatinibvsthose treated with sunitinib was 23.9vs9.2 mo (HR 0.39,P＜ 0.0001). In patients treated with everolimus/Lenvatinibvssunitinib, PFS was 14.7 and 9.2 mo, respectively, HR 0.65,P＜ 0.0001. Median overall survival was not reached; however, OS was longer with pembrolizumab/Lenvatinib than with sunitinib, HR 0.66,P= 0.005. There was no significant OS difference in patients treated with everolimus/Lenvatinib and patients treated with sunitinib, HR 1.15,P= 0.30. The objective response rate in the pembrolizumab/Lenvatinib cohortvsthe everolimus/Lenvatinibvssunitinib cohort was 71%, 53.5%, and 36.1%, respectively. Median DOR in the pembrolizumab/Lenvatinib cohortvseverolimus/Lenvatinibvssunitinib was 25.8, 16.6 and 14.6 mo, respectively. All subgroups of patients had a benefit in PFS when treated with pembrolizumab/Lenvatinib. Grade 3 or higher toxicity was observed in 82.4%vs83.1% and 71.8% of the patients treated with pembrolizumab/Lenvatinib, everolimus/Lenvatinib and sunitinib, respectively.The most common grade 3 toxicities were diarrhea, hypertension, and elevated lipase and triglyceride levels[39].

Avelumab is a humanized PD-L1 monoclonal antibody. It was investigated in a phase 1b trial in combination with axitinib in treatment-na?ve patients with mccRCC.The objective response rate was 58%[40]. The phase 3 trial, JAVELIN Renal 101,compared avelumab/axitinib with sunitinib in patients who were not previously treated. The co-primary endpoints were PFS and OS in PD-L1-positive patients, while the secondary endpoint was PFS in the ITT population. In PD-L1-positive patients, PFS was 13.8 mo for avelumab/axitinib in comparison to 7.2 mo for patients treated with sunitinib, HR 0.61,P＜ 0.0001. In the ITT population, PFS was 13.8 mo for avelumab/axitinib in comparison to 8.4 mo for patients treated with sunitinib, HR 0.69,P＜ 0.0001. In the PD-L1-positive population, the ORR was 55.2% in the avelimab/axitinib group and 25.5% in the sunitinib group. Adverse grade 3 or higher events were reported in 71.2% of patients treated with avelumab/axitinib and 71.5% of patients treated with sunitinib[41]. At 13 mo PFS was significantly longer for the patients treated with avelumab/axitinibvssunitinib in both PD-L1 positive (HR 0.62,P＜ 0.0001, 13.8vs7 mo) and ITT populations (HR 0.69,P＜ 0.0001, 13.3vs8 mo). Data for OS are still pending[42]. In May 2019, this combination was approved for the firstline treatment of mccRCC patients, regardless of the IMDC score prognostic subgroup.

In January 2021, nivolumab in combination with cabozantinib was approved by the FDA for the first-line treatment of patients with mRCC based on the results of the CheckMate 9ER trial. The trial included treatment-na?ve patients, regardless of the PDL1 expression or IMDC prognostic score. Patients were randomized into two cohorts:nivolumab/cabozantinib and sunitinib. The primary endpoint was PFS, and the secondary endpoints were OS and ORR. At 18.1 mo, PFS and OS were both significantly longer in the nivolumab/cabozantinibvsthe sunitinib cohort in all patient subgroups analyzed[43] in Table 2.

Table 2 Results of phase 3 studies in first line treatment of patients with metastatic clear cell renal cell cancer

Most of the trials that examined the efficacy of immunotherapy or immunotherapy/TKI combinations did not include mnccRCC. Some of the retrospective trials with immunotherapy reported ORRs of 9-20%. The greatest benefit occurred in patients with the papillary histology subtype[44,45]. In the phase 2 trial, Keynote 427,pembrolizumab was investigated in previously untreated mnccRCC patients. Most of the patients had papillary subtype (72%). In the ITT population, the ORR was 24.8%,while the ORRs of papillary, chromofobe and nonclassified subtypes patients were 25.4%, 9.5% and 34.6%, respectively. The twelve-month PFS and OS were 22.8% and 72%. After 11 mo of follow-up, the median DOR was not reached in either subgroup of patients[46].

Nivolumab was investigated in the phase 3b/4 trial, Checkmate 374, which included treatment-na?ve patients as well as patients previously treated with a maximum of 3 Lines of therapy. Most of the patients (66%) were treatment na?ve. After 11 mo of follow-up, the median OS was 16.3 mo, with no difference in OS between patients regardless of PD-L1 expression. The median PFS was 2.2 mo. At one year, PFS was 14%. The median DOR was 10.2 mo, and ORR was 13.6%[47]. The Cosmic-021phase 1b trial analyzed the efficacy of atezolizumab in various solid tumors. One of the cohorts was patients with mnccRCC. These patients were treated with cabozantinib 40 mg and 1200 mg of atezolizumab.

According to the IMCD, all three prognostic subgroups were included in the trial,and most of them were in the intermediate prognostic group. After a median followup of 9.2 mo, the ORR was 33%, with no difference between subgroups. The median DOR was 7.9 mo. Grade 3-44 adverse events were reported in 30% of the patients, and a low phosphorus level was the most common adverse event[48]. The Calypso trial,phase 1b/2, examined the combination of durvalumab and savolitinib in patients with papillary mnccRCC previously treated, as well as treatment na?ve. The primary endpoint was ORR, while the secondary endpoints were PFS, OS and safety. The trial included all IMDC score prognostic groups. Most of the patients (63%) were in the intermediate prognostic group. Median follow up was 8.9 mo. In the ITT population,the ORR was 27%, while the median PFS was 3.3 mo. In the subgroup of patients who were treatment na?ve, the ORR was 27%, and the median PFS was 12.2 mo. Fifteen out of 42 patients included had grade 3-4 toxicities[49] (Table 3).

Table 3 Results of checkpoint inhibitors in treatment of patients with metastatic clear cell renal cell cancer

Predictive biomarkers

Is there a biomarker that can predict the response to either immunotherapy or tyrosine kinase inhibitors (TKIs)? One of the essential promoters of cell growth in RCC is angiogenesis. Patients in the favorable prognostic group had abundant tumor infiltrates with blood vessels. However, RCC is also an immunogenic tumor with inflammatory tumor infiltrates and is a characteristic in patients with intermediate and poor prognoses. The Bionikk trial assessed the response to immunotherapy and TKI therapy relative to the molecular tumor profile (35 genes). Patients were classified into four subgroups: group 1 (immune-low), group 2 (angio-high), group 3 (normal-like),and group 4 (immune-high). They were randomized so that groups 1 and 4 were treated with either nivolumab or nivolumab/ipilimumab, while patients in groups 2 and 3 received either sunitinib or nivolumab/ipilimumab. Primary endpoint was ORR. In group 1, the ORR was 33.3% and 20.7% for patients treated with nivolumab/ipilimumab or nivolumab, respectively. There was no difference in ORR between patients in group 4 who were treated with nivolumabvsnivolumab/ipilimumab 42.9%vs41.2%. In group 2, the ORR was 58.3%vs34.5% in patients treated with sunitinibvsnivolumab/ipilimumab. A very small number of patients were included in group 3, and responses were only achieved in patients treated with the nivolumab/ipilimumab combination[50]. PD-L1 is the most commonly analyzed biomarker that predicts the response to immunotherapy. Several trials pointed out that high expression of PD-L1 in patients with RCC is a predictor of poor prognosis[21,23].Checkmate 025 reported that nivolumab is superior to everolimus in previously treated patients. Higher expression of PD-L1 was related to worse prognosis regardless of whether patients were treated with nivolumab or everolimus. The median OS was longer in PD-L1-negative patients regardless of the treatment[21].

In the Checkmate 214 trial, intermediate- and poor-risk patients who were also PDL1 positive had longer PFS when treated with nivolumab/ipilimumabvssunitinib.There was no PFS difference in the PD-L1-negative population[22]. According to the JAVELIN Renal 101 and Keynote 426 trials, all subgroups of patients had significantly longer PFS regardless of the prognostic group or PD-L1 expression[37,42]. Different histology subtypes of RCC have different TMB values. The lowest TMB is found in chromofobic subtype, and the highest was found in the papillary and clear cell histology subtypes[51]. In other malignancies, such as lung cancer and melanoma,TMB is a predictor of a favorable response to treatment. Although they have relatively low TBM, patients with RCC have higher rates of response to immunotherapy[52,53].The results of trials that analyzed the prognostic value of TBM in RCC are inconclusive[54,55]. A retrospective analysis showed that TMB values do not correlate with either survival or PD-L1 expression[56]. Subanalysis of the IMmotion 150 trial showed that TMB did not influence the response to nivolumab[31]. Tumor infiltrates in RCC consist of CD8+ T lymphocytes, dendritic cells, NK cells and macrophages[15]. Some trials have shown that if tumor infiltrates are abundant with CD8+ cells and M1 macrophages, patients have a better prognosis, while infiltrates rich in regulatory T cells and M2 macrophages predict poor prognosis[57-60]. Other trials indicated that if tumor infiltrate is abundant with CD8+, patients will have a better response to immunotherapy[61]. IMmotion 150 and IMmotion 151 confirmed these results[29,31].

To date, there are no biomarkers that can predict the response to immunotherapy.Some drugs approved for first-line treatment may benefit many patients regardless of prognostic group or PD-L1 expression[36,39,42]. Further investigations are warranted to improve the selection of patients for the best possible choice of first-line therapy.

CONCLUSION

We are witnessing the evolution of mccRCC treatment. Starting with interferon-alpha and interleukin 2 in the late twentieth century, the first TKI was administered in 2007.In the last few years, several therapeutic options have been approved (immunotherapy and immunotherapy/TKI) as first-line treatment options. Nivolumab/ipilimumab is approved for the treatment of patients with intermediate and poor prognoses. Several checkpoint inhibitors (pembrolizumab, nivolumab, avelumab) in combination with TKIs (axitinib, lenvatinib, cabozantinib) are approved for the treatment of patients regardless of their IMDC prognostic group and PD-L1 expression. There is no specific and ideal biomarker that could help select the ideal patient for the appropriate firstline treatment. If patients are symptomatic, have visceral metastasis and require prompt response, then checkpoint inhibitors/TKIs are deemed most beneficial. If the patient is asymptomatic, then other factors, such as toxicity profile, may influence the first-line treatment option.