Nivolumab for the Treatment of Patients with Metastatic Non‐Clear Cell Renal Cell Carcinoma (nccRCC): A Single‐Institutional Experience and Literature Meta‐Analysis

Abstract

Introduction

Nivolumab alone and in combination with ipilimumab is approved for the treatment of patients with metastatic renal cell carcinoma (RCC) who received prior vascular endothelial growth factor receptor tyrosine kinase inhibitors (VEGFR‐TKI) and those who are treatment naive, respectively. However, the clinical activity of nivolumab in non‐clear cell RCC (nccRCC) is unknown, as these patients were excluded from the trials.

Materials and Methods

We reviewed the records of patients who received nivolumab for nccRCC and ccRCC with >20% rhabdoid with the primary endpoint to assess the objective response rate (ORR). We assessed radiographic response using RECIST, v1.1. Secondary endpoints were progression‐free survival (PFS) and overall survival (OS). We also reviewed the literature to identify studies reporting on the clinical activity of immune checkpoint inhibitors in nccRCC, and performed a meta‐analysis of proportions for ORR and disease control rate (DCR).

Results

Twelve patients (30%) had papillary histology, 11 (27.5%) had unclassified, 8 (20%) had ccRCC with rhabdoid component, 5 (12.5%) had chromophobe, 3 (7.5%) had translocation, and 1 (2.5%) had mucinous tubular and spindle cell carcinoma. Overall, seven patients (21.6%, 95% confidence interval [CI], 8.7%–37.9%) had an objective response, including three patients (8.8%, 95% confidence interval [CI], 1.9%–23.7%) who achieved a complete remission. At a median follow‐up of 24.5 monoths (95% CI, 17.7–32.6), median PFS was 4.9 monoths (95% CI, 3.53–10.27) and median OS was 21.7 monoths (95% CI, 7.83 mo to not reached). There were no treatment‐related deaths. We also identified two retrospective studies reporting best ORR in patients with nccRCC receiving PD‐1/PD‐L1 checkpoint blockade. The ORR and DCR for the total cohort were, respectively, 18.6% (95% CI, 11.9%–26.4%) and 53.4% (95% CI, 44.2%–62.5%).

Conclusion

Nivolumab demonstrated activity in unclassified nccRCC and ccRCC with >20% rhabdoid; further randomized clinical trials are warranted.

Implications for Practice

This article reports on the clinical activity and safety of immune checkpoint inhibitors in non‐clear cell kidney cancer. The retrospective data with the meta‐analysis provides a summary that will help guide the treatment of this rare and heterogeneous group of kidney cancers.

Short abstract

The evidence for the clinical efficacy and safety of immune checkpoint inhibitors in non‐clear cell renal cell carcinoma (nccRCC) is inconclusive. This article assesses the clinical efficacy of nivolumab in patients with metastatic nccRCC and reports results of a meta‐analysis of studies reporting the best overall response rates of metastatic nccRCC histologies to PD‐1/PD‐L‐1 checkpoint inhibitors.

Introduction

Renal cell carcinoma (RCC) is the eighth most common malignancy in the U.S., with 65,000 new cases of kidney cancer diagnosed and 15,000 deaths per year 1. Clear cell RCC (ccRCC) is the most common subtype, accounting for 70%–75% of all primary kidney malignancies, whereas all other subtypes are collectively referred to as non‐clear cell renal cancer (nccRCC). The 2016 World Health Organization classification identifies subtypes of nccRCC 2, including papillary (10%), chromophobe (5%), mucinous tubular and spindle cell (MTSC; ∼2%), Xp11.2 translocation (∼2%), medullary (<1%), and unclassified (5%–10%) RCC. The incidence of metastatic nccRCC is high, with 8,000 patients in Europe 3. The current guidelines for treatment of metastatic RCC are based on evidence from trials in which clear cell histology predominate and nccRCC histologies were excluded 4, 5. The heterogeneity and the low incidence of nccRCC have resulted in limited randomized trials with strong evidence to inform patient management 6, 7. Therefore, the current American and European guidelines for treatment of patients with metastatic nccRCC have followed that of ccRCC 8. Despite little evidence from randomized trials, a meta‐analysis comparing patients with nccRCC with ccRCC treated with targeted agents showed significantly lower objective response rates (ORR) and inferior progression‐free survival (PFS) and overall survival (OS) 9. Unfortunately, the limited impact of these targeted agents approved for RCC on improving outcomes in nccRCC has translated to poor survival rates 10. Furthermore, the multicenter randomized clinical trials that led to the approval of nivolumab alone 11 and in combination with ipilimumab 12 for patients with metastatic RCC who had prior therapy and those who were therapy naive, respectively, excluded patients with nccRCC. Hence, patients with metastatic nccRCC have limited therapeutic options and have not been included in the trials that established the clinical activity of nivolumab in metastatic RCC. Currently, the evidence for the clinical efficacy and safety of immune checkpoint inhibitors in nccRCC is based on two retrospective studies that included patients with nccRCC and also those with ccRCC with sarcomatoid features that were treated with any PD‐1/PD‐L1 checkpoint inhibitors, showing acceptable safety and interesting clinical efficacy in specific nccRCC histologies 13, 14. Therefore, we aimed to assess the clinical efficacy of nivolumab in patients with metastatic nccRCC who were treated at MD Anderson Cancer Center. We also conducted a meta‐analysis of studies reporting the best ORR of metastatic nccRCC histologies to PD‐1/PD‐L1 checkpoint inhibitors.

Materials and Methods

Patients and Study Design

We conducted a single institution retrospective analysis of patients with metastatic nccRCC who received nivolumab between March 2010 and December 2017. Eligible patients were defined as those with metastatic nccRCC, determined by expert pathology review, excluding patients with renal medullary cancer (RMC) or RCC with sarcomatoid deddiferentiation. Patients with ccRCC with rhabdoid differentiation in >20% of the tumor specimen were also eligible. Patients must have received treatment with nivolumab as monotherapy or in combination with ipilimumab or other targeted therapies prior to December 2017. The demographic, clinical, and treatment data for each patient were obtained from retrospective chart review from the electronic medical records. The best objective response rate was defined by Response Evaluation Criteria in Solid Tumors (RECIST) version 1.1 and was assessed only in patients with imaging evaluable for response by a radiologist blinded to patient clinical data.

In addition, we performed a review of the literature between December 2012 and September 2018 reporting on the clinical activity of PD‐1 and PD‐L1 checkpoint blockade in metastatic nccRCC. We queried the literature in PubMed, Medline, and abstracts from proceedings of European Society for Medical Oncology and American Society of Clinical Oncology. Meta‐analysis of ORR and disease control rate (DCR) were performed in R version 3.5.3 using the packages metafor 15 and meta 16. All observed proportions were transformed via the Freeman‐Tukey double arcsine method 17. A random effects analysis model, based on restricted maximum‐likelihood, was applied in all meta‐analytic calculations because it provides more conservative estimates of the pooled data. Between‐study heterogeneity and inconsistency were assessed by the Cochran Q statistic and the inconsistency index (I²), respectively. Publication bias was statistically assessed using Egger's linear regression test.

Study Analysis

We used summary statistics to tabulate patient data, treatment characteristics, and best ORR, that is, mean, SD, median, and interquartile range (IQR) for continuous variables and frequencies for categorical variables. The primary endpoint was to assess the ORR, defined as the proportion of patients who have a partial or complete response (CR) to therapy by RECIST version 1.1. We calculated the proportion of patients with ORR along with 95% exact confidence intervals in the total cohort, by histology and line of therapy. The same approach was also used to summarize DCR, defined as CR, partial response, or stable disease. Secondary endpoints included OS and PFS. OS was defined as the time from initiation of nivolumab until the time of death or censoring at the last follow‐up. PFS was defined as the time from initiation of nivolumab to the time of progression or death, whichever occurred first, or to the date of last follow‐up for patients who were alive and had no progressed disease. We estimated the survival outcomes for the overall cohort using the Kaplan‐Meier method, comparing the survival outcomes among subgroups of patients by histology and International Metastatic RCC Database Consortium (IMDC) risk group.

Results

Baseline Characteristics

Patient and disease characteristics are summarized in Table 1. We identified 40 patients; 12 (30%) had papillary histology, 11 (27.5%) had unclassified, 5 (12.5%) had chromophobe, 3 (7.5%) had translocation Xp11.2, 8 (20%) had >20% rhabdoid component, and 1 (2.5%) had MTSC. The median age at nivolumab initiation was 58.5 (IQR, 43.8–65.9), and the majority of patients were male (n = 32, 80%), with intermediate‐risk disease by IMDC criteria (n = 25, 72.5%) and Eastern Cooperative Oncology Group (ECOG) performance status 1 or 2 (n = 37, 92.5%). This was a predominantly white population (n = 33, 82.5%). Time to initiation of systemic therapy was less than 1 year in 65% (n = 26) of patients. The most common sites of metastases were lymph nodes (72.5%), lung (65%), liver (35%), bone (35%), and brain (5%). Most patients (31/40) received nivolumab monotherapy, and nine patients received nivolumab either in combination with ipilimumab (n = 5), or vascular endothelial growth factor (VEGF)‐targeted therapy (n = 4). The majority of patients received nivolumab therapy as second‐line or beyond (n = 34, 85%) and had a prior nephrectomy (n = 33, 82.5%).

Table 1.

Baseline patient and disease characteristics

Characteristic	n (%)
Male	32 (80)
BMI >25	21 (52.5)
ECOG performance status
0	3 (7.5)
1	26 (65)
2	11 (27.5)
Histology
Papillary type 1	6 (15)
Papillary type 2	6 (15)
Chromophobe	5 (12.5)
Unclassified	11 (27.5)
Translocation Xp11.2	3 (7.5)
Mucinous tubular and spindle cell carcinoma	1 (2.5)
Clear cell with rhabdoid differentiation >20%	8 (20)
IMDC risk group
Favorable	3 (7.5)
Intermediate	29 (72.5)
Poor	8 (20)
Site of metastasis
Presence of bone metastases	14 (35)
Presence of brain metastases	2 (5)
Presence of liver metastases	14 (35)
Presence of lung metastases	26 (65)
Presence of lymph node metastases	29 (72.5)
Prior nephrectomy	33 (82.5)
Line of systemic therapy
1	6 (15)
2	14 (35)
≥3	20 (50)

Abbreviations: BMI, body mass index; ECOG, Eastern Cooperative Oncology Group; IMDC, International Metastatic RCC Database Consortium.

Best Overall Response

The ORR for the total cohort was 20.6% (n = 7 of 34; 95% confidence interval [CI], 8.7%–37.9%) and DCR was 70.5% (n = 24; 95% CI, 52.5%–84.9%; Table 2). CR was observed in 8.8% (n = 3; 95% CI, 1.9%–23.7%), partial response was observed in 11.8% (n = 4; 95% CI, 3.3%–27.5%), and stable disease at 6 months from nivolumab initiation was observed in 35.2% (n = 12; 95% CI, 19.7%–59.5%) of the overall population. This cohort was heavily pretreated prior to initiation of nivolumab, and the ORR rate varied based on prior line of treatment status. We also noted that the ORR rate was numerically different based on underlying histology, although the numbers were small for formal statistical comparisons. Patients with unclassified RCC (n = 4/9, 44.4%; 95% CI, 13.7%–78.8%) and with ccRCC rhabdoid >20% (n = 2/7, 28.6%; 95% CI, 3.7%–71%) experienced a higher ORR. One patient with papillary type 1 RCC (n = 1/4, 25%; 95% CI, 0.6%–80.6%) achieved an objective response. None of the patients with papillary type 2 RCC (n = 0/6, 0%; 95% CI, 0%–45.9%), chromophobe RCC (n = 0/5, 0%; 95% CI, 0%–52.2%), or translocation RCC (n = 0/3, 0%; 95% CI, 0%–70.8%) had an objective response. Patients receiving nivolumab in combination with ipilimumab or targeted agents had higher ORR (n = 4/9, 44.4%; 95% CI, 13.7%–18.8%) in comparison with patients who received nivolumab monotherapy (n = 4/30, 13%; 95% CI, 3.8%–30.8%).

Table 2.

Best overall response

Variable	Total n	CR, n (%)	PR, n (%)	SD > 6 m, n (%)	SD, n (%)	PD, n (%)	NA, n
All	40	3 (8.8)	4 (11.8)	12 (35.2)	5 (14.7)	10 (29.5)	6
Histology
Papillary type 1	6	0 (0)	1 (25)	1 (25)	0 (0)	2 (50)	2
Papillary type 2	6	0 (0)	0 (0)	4 (66.6)	1 (16.7)	1 (16.7)
Chromophobe	5	0 (0)	0 (0)	1 (15)	1 (15)	3 (60)
Unclassified	11	2 (22.2)	2 (22.2)	1 (11.1)	2 (22.3)	2 (22.3)	2
Translocation	3	0 (0)	0 (0)	2 (100)	0 (0)	0 (0)	1
Mucinous tubular and spindle cell carcinoma	1	0 (0)	0 (0)	1 (100)	0 (0)	0 (0)
ccRCC with rhabdoid >20%	8	1 (14.3)	1 (14.3)	2 (28.5)	1 (14.3)	2 (28.5)	1
IMDC risk group
Favorable	3	0 (0)	0 (0)	1 (33.3)	1 (33.3)	1 (33.4)	0
Intermediate	29	3 (11.1)	4 (14.8)	11 (40.8)	3 (11.1)	6 (22.2)	2
Poor	8	0 (0	0 (0)	0 (0	1 (25)	3 (75)	4
Line of systemic therapy
1	6	2 (33.3)	1 (16.7)	1 (16.7)	0 (0)	2 (33.3)	0
2	14	1 (9)	2 (18.3)	4 (36.4)	3 (27.3)	1 (9)	3
≥3	20	0 (0)	1 (5.9)	7 (41.2)	1 (5.9)	8 (47)	3
Type of checkpoint blockade
Monotherapy	31	2 (8)	1 (4)	9 (36)	4 (16)	9 (36)	6
Combination therapy (nivolumab with ipilimumab or nivolumab with VEGFR‐TKI)	9	1 (11.1)	3 (33.4)	3 (33.3)	1 (11.1)	1 (11.1)
BMI
>25	21	2 (10)	2 (10)	8 (40)	1 (5)	7 (35)	1
<25	19	1 (7.1)	2 (14.3)	4 (28.6)	4 (28.6)	3 (21.4)	5

Abbreviations: BMI, body mass index; ccRCC, clear cell renal cell carcinoma; CR, complete response; IMDC, International Metastatic RCC Database Consortium; NA, not available for overall response rate evaluation; PD, progressive disease; PR, partial response; SD, stable disease; VEGFR‐TKI, vascular endothelial growth factor receptor tyrosine kinase inhibitor.

Meta‐Analysis of Best Overall Response

In our review of the literature, we identified two retrospective studies in patients with metastatic nccRCC who received PD‐1/PD‐L1 checkpoints blockade that reported the best ORR by tumor histology 13, 14. We identified 124 patients with nccRCC and ccRCC histology with >20% rhabdoid and 112 patients with evaluable response rates to PD‐1/PD‐L1 checkpoint inhibitors to be included in our meta‐analysis. The ORR and DCR for the total cohort were, respectively, 18.6% (95% CI, 11.9%–26.4%) and 53.4% (95% CI, 44.2%–62.5%). As shown in Figure 1A–B, both ORR and DCR were consistent across studies with no evidence of between‐study heterogeneity or between‐study inconsistency as assessed the Cochran Q statistic and the inconsistency index (I²), respectively. ORR and DCR between different histologies were also consistent across studies, with the exception of unclassified RCC (supplemenal online Fig. 1 and 2), as expected given the heterogeneity of RCCs deemed to be “unclassified.” Egger's test did not detect publication bias (intercept = 0.194; p = .846).

Figure 1.

Forest plot for the studies reporting on (a) the objective response rate (ORR) and (b) disease control rate (DCR) of non‐clear cell renal cell carcinoma (nccRCC) with PD‐1 and PD‐L1 checkpoint blockade. (A): Summarizes in the forest plot all the published studies reporting the ORR for nccRCC with PD‐1 and PD‐L1 checkpoint blockade. (B): Summarizes in the forest plot all the published studies reporting the DCR for nccRCC with PD‐1 and PD‐L1 checkpoint blockade. Abbreviation: CI, confidence interval.

Figure 2.

Overall survival (OS) of overall cohort. The overall survival curve for the overall cohort. The solid line is the estimated Kaplan‐Meier curve for overall survival (OS) and the dotted lines represent the corresponding 95% confidence interval (CI).

Progression‐Free Survival and Overall Survival

The estimated median follow‐up was 24.5 months (95% CI, 17.7–32.6). Overall, there were 21 deaths (52.5%) in the total cohort, and 33 patients (82.5%) experienced disease progression or death events. The median PFS for the overall cohort was 4.9 months (95% CI, 3.5–10.3; Table 3). By histology, median PFS was 3.1 months (95% CI, 2.23–not reached) for papillary type 1, 9.7 months (95% CI, 4.7–not reached) for papillary type 2, 5.5 months (95% CI, 2.0–not reached) for unclassified, 4.3 months (95% CI, 3.2–not reached) for chromophobe, and 4.8 months (95% CI, 1.0–not reached) for ccRCC with rhabdoid >20% (Table 3; supplemental online Fig. 3A). The median PFS for untreated patients was 11.0 months (95% CI, 1.8–not reached), whereas for patients receiving nivolumab as second‐line systemic therapy, the median PFS was 5.6 months (95% CI, 3.93–not reached), in comparison with 4.3 months (95% CI, 3.4–10.3) in patients receiving nivolumab as third‐line of systemic therapy and beyond. The median PFS for patients receiving nivolumab monotherapy was 4.3 months (95% CI, 3.4–7.0), in contrast with 44.5 months (95% CI, 5.5–not reached) for patients receiving combination therapy, for the overall population.

Table 3.

Summary of PFS and OS

Variable	Median PFS (95% CI)	PFS probability at 1 yr (95% CI)	Median OS (95%CI)	OS probability at 1 yr (95% CI)
Gender
Male	4.7 (3.4–7)	0.172 (0.079–0.376)	15.4 (6.5–NA)	0.533 (0.38–0.748)
Female	45.5 (4.4–NA)	0.625 (0.365–1)	NA (6.9–NA)	0.729 (0.468–1)
BMI
≤25	4.3 (3.5–NA)	0.197 (0.077–0.506)	6.5 (4.4–NA)	0.349 (0.177–0.687)
>25	6.5 (3.4–17.1)	0.333 (0.182–0.61)	NA (21.1–NA)	0.749 (0.58–0.967)
ECOG
0	4.7 (2.0–NA)	0.333 (0.067–1)	NA (22.7–NA)	1 (1–1)
1	6.5 (4.8–NA)	0.375 (0.226–0.621)	NA (17.4–NA)	0.735 (0.573–0.943)
2	2.9 (1.9–NA)	NA	3.8 (3.4–NA)	0.091 (0.014–0.589)
Histology
Papillary type 1	3.2 (2.2–NA)	NA	3.6 (3.2–NA)	0.333 (0.108–1)
Papillary type 2	9.8 (4.7–NA)	0.5 (0.225–1)	NA (15.4–NA)	0.833 (0.583–1)
Chromophobe	4.3 (3.2–NA)	NA	6.9 (4.3–NA)	0.375 (0.084–1)
Unclassified	5.5 (2.0–NA)	0.182 (0.052–0.637)	17.3 (11.6–NA)	0.623 (0.389–0.999)
Rhabdoid >20%	4.8 (3.5–NA)	0.375 (0.153–0.917)	14.9 (6.1–NA)	0.5 (0.25–1)
Translocation	NA (3.5–NA)	0.667 (0.3–1)	NA (NA–NA)	1 (1–1)
MTSC	7.4 (NA–NA)	NA	7.8 (NA–NA)	NA
Prior nephrectomy
No	2.2 (0.8–NA)	0.143 (0.023–0.877)	3.8 (0.8–NA)	0.214 (0.042–1)
Yes	5.7 (4.3–13)	0.292 (0.17–0.502)	22.7 (11.6–NA)	0.641 (0.49–0.837)
IMDC risk
Favorable	5.7 (4.3–NA)	NA	17.3 (4.3–NA)	0.667 (0.3–1)
Intermediate	6.5 (4.8–17.1)	0.369 (0.228–0.599)	22.8 (21.1–NA)	0.729 (0.575–0.924)
Poor	1.8 (0.8–NA)	NA	3.6 (0.8–NA)	NA
Line of therapy
1	11.1 (1.8–NA)	0.5 (0.225–1)	NA (NA–NA)	0.8 (0.516–1)
2	5.6 (3.9–NA)	0.257 (0.101–0.656)	22.7 (11.6–NA)	0.682 (0.463–1)
≥3	4.3 (3.4–10.27)	0.2 (0.083–0.481)	7.4 (4.5–NA)	0.45 (0.277–0.731)
Nivolumab
Monotherapy	4.3 (3.4–7)	0.161 (0.072 –0.36)	11.6 (6.1–22.8)	0.468 (0.319–0.687)
Combination	45.5 (5.5–NA)	0.667 (0.42–1)	NA (NA–NA)	1 (1–1)

Abbreviations: BMI, body mass index; CI, confidence interval; ECOG, Eastern Cooperative Oncology Group; IMDC, International Metastatic RCC Database Consortium; MTSC, mucinous tubular and spindle cell; NA, not assessable; OS, overall survival; PFS, progression‐free survival.

As for OS, the median for the overall cohort was 21.7 months (95% CI, 7.8 months–not reached) (Fig. 1). The 12‐month OS rate was 57% (95% CI, 43%–76%). When stratified by histology, median OS was 3.6 months (95% CI, 3.2–not reached) for papillary type 1, not reached (95% CI, 15.4–not reached) for papillary type 2, 17.3 months (95% CI,11.6–not reached) for unclassified, 6.9 months (95% CI, 4.3–not reached) for chromophobe, and 14.9 months (95% CI, 6.0–not reached) for ccRCC with rhabdoid >20% (Table 3; supplemntal online Fig. 3C). The median OS for untreated patients was not reached, whereas for patients receiving nivolumab as second‐line systemic therapy, the median OS was 22.7 months (95% CI, 11.6–not reached) in comparison with 7.4 months (95% CI, 4.5–not reached) in patients receiving nivolumab as third‐line systemic therapy and beyond. The median OS for patients receiving nivolumab monotherapy was 11.6 months (95% CI, 6.0–22.8) and was not reached for patients receiving combination therapy, for the overall population. There were no treatement‐related deaths. We noted acceptable safety profile, with only 7.5% (n = 3/40) having grade 3 treatment‐related immune toxicity, with two patients having grade 3 immune related colitis and one patient with pneumonitis.

Discussion

The majority of phase III randomized studies in patients with metastatic RCC that led to the approval of frontline and second‐line systemic therapies excluded patients with nccRCC, leading to an information gap in the optimal treatment strategy for patients with nccRCC. Two prospective, randomized studies, the ASPEN and ESPN trials, evaluated the use of sunitinb versus everolimus in metastatic nccRCC 7, 18, 19. These trials included patients with papillary types I and II, chromophobe, unclassified, translocation, and ccRCC with >20% sarcomatoid (ESPN only) while excluding patients with collecting duct and RMC histologies 7, 18. These trials demonstrated a superior PFS with sunitinib in comparison to everolimus, and the ORR for patients receiving sunitinb was 18 and 9% 7, 18, 19. In this study, we report the clinical activity of nivolumab in a heavily pretreated cohort of patients with metastatic nccRCC and ccRCC rhabdoid >20%; ORR was 20.6% with a CR rate of 8.8%. Even though only six patients were treatment naive, they had a numerically higher ORR (n = 3/6, 50%) in comparison with the reported ORR with sunitinib in nccRCC. We also report that the ORR and DCR were 18.6% and 53.4%, respectively, in a meta‐analysis of 124 patients from published series of metastatic nccRCC and ccRCC >20% rhabdoid patients treated with PD‐1/PD‐L1 checkpoint inhibitors 13, 14. The observed best ORR with nivolumab in patients with metastatic nccRCC is consistent with the reported activity in ccRCC 11, 12. We noted different response rates by nccRCC histology, which could be explained by the known variability in molecular alterations between this heterogeneous population 20. Patients with unclassified RCC (n = 4/9, 44.5%) and those with ccRCC with rhabdoid >20% (n = 2/7, 28.6%) had higher response rates, whereas patients with chromophobe and papillary type 2 RCC had no observed responses. This was confirmed in the summary cohort pooled analysis, where 20 patients with chromophobe RCC were treated with PD‐1/PD‐L1 without any observed response. Interestingly, a recent retrospective analysis highlighted the potential activity of cabozantinib in patients with chromophobe and papillary type 2 RCC 21. Therefore, cabozantinib could be an appropriate treatment option for these specific ncc histologies 22, 23. In our retrospective analysis, the median PFS and OS of the overall cohort were 4.9 months and 21.7 months, respectively.

We are aware of the limitations of a retrospective study, with a small sample size evaluating a heterogeneous population and conducting multiple subanalyses. For that reason, we conducted a pooled analysis and reported response rates and survival analysis by histology for a cohort of 124 patients, the largest to be reported to our knowledge. In an attempt to limit bias, all patients, regardless of their baseline condition or number of prior treatment lines, who received nivolumab were included in our analysis. This is reflected in our heavily pretreated population, with the majority of patients being intermediate IMDC risk and with ECOG performance status of 1 or 2. Also, a blinded radiologist reviewed all imaging to limit bias.

Conclusion

This single institution analysis and this pooled analysis provide insight into the clinical activity of nivolumab in metastatic nccRCC and rhabdoid RCC, highlighting the differential activity in patients with variable nccRCC histologies. Further evidence from three single arm trials supporting the safety and clinical activity of single agent PD‐1/PD‐L1 pathway blockade or in combination with bevacizumab or savolitinib in nccRCC have been presented but have not been published yet. The results from the Keynote 427B, the largest clinical trial using single agent pembrolizumab in previously untreated nccRCC, and a second trial using the combination of atezolizumab with bevacizumab both showed clinical activity in specific histologies of nccRCC, but the survival data is immature to interpret at the time of presentation. These large clinical trial efforts will help guide the use of immunotherapies in this heterogenous group, as they will better delineate the clinical activity by specific tumor histology. Until further evidence from randomized trials becomes available, our data suggest that nivolumab could be appropriately used alone or in combination with ipilimumab in patients with unclassified histology nccRCC and ccRCC with >20% rhabdoid, based on their approval for second‐line and front line treatment for metastatic RCC respectively. In contrast, clinical trial enrollment or cabozantinib should be considered first for the treatment of patients with papillary type 2 RCC 21 and nivolumab does not appear to have clinical activity for pateints with chromophobe RCC. These findings support the need for randomized trials using blockade of the PD‐1/PD‐L1 pathway in this population, highlighting the ongoing randomized phase II trial comparing the combination of nivolumab with ipilumumab to the current standard of care for previously untreated metastatic nccRCC (NCT03075423). Also, of importance are the ongoing, prospective trials of nivolumab alone or in combination with ipilimumab that include patients with nccRCC, respectively, the CheckMate 374 and CheckMate 920, will report on outcomes with nivolumab monotherapy and nivolumab plus ipilimumab in patients with nccRCC.

Author Contributions

Conception/design: Jad Chahoud, Nizar Tannir

Provision of study material or patients: Jad Chahoud, Pavlos Msaouel, Matthew T. Campbell, Jianjun Gao, Amado J. Zurita, Amishi Yogesh Shah, Eric Jonasch, Padmanee Sharma, Nizar M. Tannir

Collection and/or assembly of data: Jad Chahoud, Pavlos Msaouel, Lianchun Xiao

Data analysis and interpretation: Jad Chahoud, Pavlos Msaouel, Tharakeswara Bathala, Lianchun Xiao, Nizar M. Tannir

Manuscript writing: Jad Chahoud, Pavlos Msaouel, Matthew T. Campbell, Tharakeswara Bathala, Lianchun Xiao, Jianjun Gao, Amado J. Zurita, Amishi Yogesh Shah, Eric Jonasch, Padmanee Sharma, Nizar M. Tannir

Final approval of manuscript: Jad Chahoud, Pavlos Msaouel, Matthew T. Campbell, Tharakeswara Bathala, Lianchun Xiao, Jianjun Gao, Amado J. Zurita, Amishi Yogesh Shah, Eric Jonasch, Padmanee Sharma, Nizar M. Tannir

Disclosures

Matthew T. Campbell: Exelixis, Pfizer, EMD Serono (C/A, H, RF), Eisai, Astrazeneca (C/A, H), Apricity Health (C/A), Roche (H); Padmanee Sharma: Jounce, Neon, Constellation, Oncolytics, BioAtla, Forty‐Seven, Apricity, Polaris, Marker, Codiak, ImaginAb, Hummingbird, Optera, Dragonfly (OI) Constellation, Jounce, Neon, BioAtla, Pieris, Oncolytics, Merck, Forty‐Seven, Polaris, Apricity, Marker, Codiak, ImaginAb, Hummingbird, Optera, Dragonfly (C/A), Jounce (IP); Nizar M. Tannir: Pfizer, Novartis, Bristol‐Myers Squibb, Exelixis, Nektar, Calithera Biosciences, Eisai Medical Research, Ono Pharmaceutical (H), Oncorena. Research‐ Bristol‐Myers Squibb, Novartis, Exelixis, Epizyme, Mirati Therapeutics (C/A), Pfizer, Novartis, Bristol‐Myers Squibb, Exelixis, Nektar, Calithera Biosciences, Eisai Medical Research, Ono Pharmaceutical, Oncorena (Other‐personal fees). The other authors indicated no financial relationships.

(C/A) Consulting/advisory relationship; (RF) Research funding; (E) Employment; (ET) Expert testimony; (H) Honoraria received; (OI) Ownership interests; (IP) Intellectual property rights/inventor/patent holder; (SAB) Scientific advisory board

Supporting information

See http://www.TheOncologist.com for supplemental material available online.

Supplemental Figures