FormalPara Key Points

Patients with metastatic renal cell carcinoma of non-clear cell histology undergoing cytoreductive nephrectomy prior to systemic therapy have better outcomes compared with those treated with systemic therapy alone.

For patients who are not candidates for cytoreductive nephrectomy, immunotherapy plus tyrosine kinase inhibitor combination or, eventually, tyrosine kinase inhibitor monotherapy might be preferred.

1 Introduction

Renal cell carcinoma (RCC) represents one of the most prevalent urogenital malignancies with increasing incidence in developed countries [1]. Clear cell RCC is the predominant histological type, accounting for about 80% of all RCC cases. According to the current histopathological classification of non-clear cell RCC (nccRCC), accounting for the remaining 20% of RCC cases, it comprises several distinct entities [2]. Nonetheless, while surgery remains the pivotal therapeutic strategy for localized-stage RCC, the question regarding the role of cytoreductive nephrectomy (CN) in the context of mRCC continues to be a subject of considerable contention. In the previous era of cytokine therapy, CN followed by cytokine therapy used to be the standard of care. This approach was based on two prospective randomized clinical trials [3, 4]. However, substantial progress in systemic therapies for mRCC has been changing this paradigm. The cytokine therapy era was followed by the era of targeted therapies, based on the use of tyrosine kinase inhibitors (TKIs). Now, we are moving forward into the era of immunotherapy combinations, where the front-line therapy is based on immuno-oncology (IO) combinations with two immune checkpoint inhibitors (IO+IO) or a combination of IO and TKIs (IO+TKI). The results of a prospective randomized trial, CARMENA, suggest that sunitinib alone is non-inferior to CN followed by sunitinib in mRCC with intermediate or poor prognosis according to the Memorial Sloan Kettering Cancer Center prognostic model [5]. In contrast, several retrospective studies showed superior outcomes of patients with mRCC undergoing upfront CN [6,7,8,9,10,11,12,13]. When focusing on the role of CN in the current standard of care based on first-line IO combinations, the data are limited. Although several retrospective studies suggest improved outcomes in patients who underwent CN, there are no available data from prospective randomized trials that are currently in progress [14,15,16,17,18,19,20,21]. Furthermore, the question of optimal timing of CN (upfront vs deferred) has been a subject of debate.

Importantly, the vast majority of data on CN currently available are derived from studies mainly focused on clear cell RCC or an unselected mRCC population, where nccRCC represented only small patient cohorts. Thus, the role of CN in patients with mRCC of non-conventional histological types remains highly underexplored. In the present retrospective study, we focused on the impact of upfront CN on the outcomes of patients with metastatic nccRCC treated with IO combinations or TKI monotherapy as a first-line systemic therapy.

2 Methods

2.1 Study Design and Population

The study included patients with nccRCC, and synchronous metastatic disease, treated with first-line IO combinations (1 January, 2016 to 1 October, 2023) or TKI monotherapy (1 January, 2008 to 1 January, 2017) from 47 cancer centers. All included patients had known baseline clinicopathological characteristics including the prognostic group according to the International Metastatic Renal Cell Carcinoma Database Consortium (IMDC) criteria [22], previous upfront CN, and the type and duration of systemic therapy.

Clinical data were retrospectively extracted at each participating center, from the patients’ medical records. First-line therapy was continued until evidence of clinical and/or radiological progression according to the Response Evaluation Criteria In Solid Tumors (RECIST) Version 1.1 criteria [23], unacceptable toxicities, or death. Computed tomography and laboratory tests were performed following current clinical guidelines and standard local procedures.

The study was conducted according to Good Clinical Practice and has been designed with the ethical principles laid down in the Declaration of Helsinki on human experimentation. The study protocol was approved by the ethical committee of the coordinating center (Marche Region—2021-492, Study Protocol “ARON-1 Project” NCT05287464) and by the institutional review boards of the international participating centers. Informed consent with subsequent analysis of the follow-up data was obtained from all participants.

2.2 Statistical Analysis

Baseline clinical characteristics were summarized with descriptive statistics and the comparisons between groups were performed using the Fisher exact test, Pearson’s correlation coefficient, and an analysis of variance. Overall survival (OS) was determined from the date of the first-line treatment initiation until the date of death of any cause. Progression-free survival (PFS) was determined from the date of the first-line treatment initiation until the date of progression or death of any cause, whichever occurred first. Patients in whom the terminal event had not occurred were censored at the date of the last follow-up. Overall survival and PFS were estimated using the Kaplan–Meier method and point estimates were accompanied by two-sided 95% confidence intervals (CIs). The log-rank test was used for the assessment of statistical significance of the differences in OS and PFS. To identify independent prognostic factors for PFS and OS, univariate and multivariate Cox proportional hazard regression models were performed. Values of p < 0.05 were considered statistically significant. Statistical analyses were conducted using the MedCalc Version 19.6.4 (MedCalc Software, Mariakerke, Belgium).

3 Results

3.1 Study Population

In total, the ARON-1 study enrolled 4211 patients with RCC. We identified 369 (8.7%) patients with metastatic nccRCC; of those, 221 presented with synchronous metastatic disease at diagnosis and were included in the present study (Fig. S1 of the Electronic Supplementary Material [ESM]); 111 patients (50%) underwent upfront CN followed by first-line systemic therapy (CN group) while 110 patients (50%) were treated with systemic therapies without CN (no-CN group).

The median age was 64 years (range 25−89 years). Tumor histology was papillary in 118 cases (53%), chromophobe in 26, unclassified in 25, MiT family translocation in 9, tubolocystic in 4, fumarate hydratase deficient in 4, Bellini duct in 4, medullary in 3, eosinophilic solid and cystic in 3, mucinous tubular spindle cell in 3, and thyroid-like follicular in 2 cases. The lung was the most common metastatic site occurring in 128 patients (58%). Fifty-four patients (24%) presented with only one site of metastases, while 167 (76%) had two or more metastatic sites.

Baseline clinical and pathological characteristics of patients according to CN are summarized in Table 1. The only characteristic that differs significantly between the two groups is the distribution of the IMDC risk group (p = 0.002); indeed, nephrectomized patients had fewer poor-risk and more intermediate-risk individuals compared with those not undergoing CN. In addition, no differences were observed in the clinical characteristics of patients according to the type of first-line systemic therapy (Table 2). The median follow-up time in the overall study population was 26.6 months (95% CI 13.4−78.5); 30.2 months (95% CI 20.7–49.8) for patients receiving TKI monotherapy, 23.8 months (95% CI 12.0–51.7) for IO+TKI, and 25.2 months (95% CI 17.8–80.9) for those receiving IO+IO (p = 0.561).

Table 1 Patient characteristics stratified by upfront cytoreductive nephrectomy
Full size table
Table 2 Patient characteristics stratified by first-line therapy
Full size table

3.2 OS and PFS

The median OS was 36.8 months (95% CI 24.9–71.3) for patients who underwent CN versus 20.8 months (95% CI 12.6–24.8) for those without CN (p = 0.005, Fig. 1). The benefit of upfront CN was confirmed in both IMDC risk groups (intermediate risk: 41.7 months, 95% CI 26.9–72.3 vs 24.5 months 95% CI 20.8–31.3, p = 0.046; poor risk: 27.7 months, 95% CI 11.9–35.3 vs 9.2 months, 95% CI 7.5–22.1, p = 0.022; Fig. 1). No significant difference in PFS was found between patients who underwent CN or not (11.7 months, 95% CI 6.5–80.0 vs 6.6 months, 95% CI 6.0–45.7, p = 0.229), with a 1-year PFS rate of 48% vs 37% (p = 0.117). Similarly, no significant difference in PFS was found between the CN and no-CN groups in both intermediate-risk patients (12.2 months, 95% CI 8.3–80.0 vs 8.1 months, 95% CI 6.1–45.7, p = 0.527) and poor-risk patients (8.1 months, 95% CI 3.2–22.7 vs 6.1 months, 95% CI 4.8–8.7, p = 0.390).

Fig. 1
figure 1

Kaplan–Meier estimates of overall survival according to the upfront cytoreductive nephrectomy for overall population (A) and stratified by International Metastatic RCC Database Consortium risk: intermediate-risk patients (B) and poor-risk patients (C)

Full size image

Further, we stratified patients by metastatic sites. Overall survival was significantly longer in the CN subgroup in the patients with lung (41.7 months, 95% CI 27.7–116.8 vs 28.8 months, 95% CI 9.3–30.0, p < 0.001, Fig. 2), bone (41.7 months, 95% CI 24.9–92.3 vs 12.6 months, 95% CI 9.2–30.0, p = 0.004, Fig. 2), or liver metastases (36.8 months, 95% CI 19.7–93.4, vs 8.1 months, 95% CI 6.4–24.5, p = 0.017, Fig. 2). The favorable impact of CN was observed in both patients with one metastatic site (median OS: 71.3 months, 95% CI 35.3–71.3 vs 29.4 months, 95% CI 12.6–40.5, p = 0.046, Fig. 2) or those with two or more metastatic sites (median OS: 28.8 months, 95% CI 23.9–116.8 vs 16.7 months, 95% CI 9.3–22.4, p = 0.002, Fig. 2).

Fig. 2
figure 2

Kaplan–Meier estimates of overall survival according to the upfront cytoreductive nephrectomy stratified by type: lung metastastases (A), bone metastastases (B), liver metastastases (C), and number of metastatic sites: one metastatic site (D), and two or more metastatic sites (E)

Full size image

3.3 Comparison of OS and PFS Between Immune-Based Combinations and TKI Monotherapy According to Upfront CN

The median OS was 31.1 (95% CI 20.8–40.5) for patients receiving IO+TKI, 19.7 months (95% CI 7.3–28.8) for IO+IO, and 26.9 months (95% CI 22.1–36.8) for those receiving TKI monotherapy (p = 0.236, Fig. 3A). The median PFS was 15.8 months (95% CI 12.7–22.7) for IO+TKI, 8.1 months (95% CI 4.1–47.6) for IO+IO, and 6.5 months (95% CI 5.340.0) for those receiving TKI monotherapy (p = 0.037, Fig. 3B).The survival according to IMDC risk can be seen in the ESM.

Fig. 3
figure 3

Kaplan–Meier estimates of progression-free survival and overall survival according to the type of first-line therapy stratified by upfront cytoreductive nephrectomy (CN). Comparison between immuno-oncology (IO) combinations containing tyrosine kinase inhibitor (TKI) plus IO (TKI+IO) vs a combination of two IO agents (IO+IO) vs TKI monotherapy (TKI) for the overall patient population (A, B), patients who underwent CN (C, D), and those without CN (E, F)

Full size image

In patients who underwent CN, the median OS was not reached for both the types of IO combinations and 26.9 months (95% CI 23.5–41.7) for TKI monotherapy (p = 0.080, Fig. 3C). The median PFS was 22.7 months (95% CI 12.3–23.5) versus 11.9 months (95% CI 3.9–47.6) for those treated with IO+TKI versus IO+IO versus TKI monotherapy, respectively (p = 0.005, Fig. 3D).

In patients without CN, the median OS was 24.5 (95% CI 18.1–40.5) for patients receiving IO+TKI, 7.5 (95% CI 4.3–22.4) for IO+IO, and 22.1 months (95% CI 12.6–30.0) for those treated with TKI monotherapy (p = 0.065, Fig. 3E). The median PFS was 13.0 months (95% CI 6.6–23.5) for patients treated with IO+TKI, 4.9 months (95% CI 3.0–8.1) for those receiving the IO+IO combination, and 6.5 months (95% CI 5.0–45.7) for those receiving TKI monotherapy (p = 0.033, Fig. 3F).

3.4 Univariate and Multivariate Cox Analyses

Upfront CN was significantly associated with OS in both univariate (HR 0.48 [95% CI 0.31−0.72], p < 0.001) and multivariate (HR 0.47 [95% CI 0.31−0.72], p < 0.001) Cox regression analyses. Other significant prognostic factors for OS were IMDC risk group and the number of metastatic sites in both univariate and multivariate models. Otherwise, the IMDC risk group, bone metastases, and the choice between TKI monotherapy and IO combinations were significant predictors of PFS in the univariate analysis and subsequently the presence of bone metastases and the choice of first-line therapy confirmed their prognostic role in the multivariate analysis. The results of Cox regression analyses are listed in Table 3.

Table 3 Univariate and multivariate survival analyses
Full size table

4 Discussion

The results of our retrospective study suggest a superior OS for patients with synchronous metastatic nccRCC undergoing upfront CN followed by systemic therapy regardless of whether with IO combination regimens or TKI monotherapy as compared with those who did not undergo surgery. In the current era of IO combinations, the role of CN is more relevant than ever and a subject of ongoing discussion. Nevertheless, the use of CN in mRCC has remained substantially stable for the last decades: more than 85% of patients included in randomized clinical trials and expanded access programs published from 2003 to 2019 had undergone previous nephrectomy before systemic therapy initiation, which means that current evidence driving the clinical practice, is mainly based on a nephrectomized population and supports the use of CN also in the metastatic setting [24]. As the systemic treatment with cytokines proved benefit in patients with mRCC, studies evaluating the addition of CN in patients with synchronous metastatic disease started to emerge. A combined analysis of two randomized trials conducted in the era of cytokine first-line therapy underlined a 31% reduction in the risk of death in patients undergoing surgery [25]. Since then, CN has become a pivotal weapon in the spectrum of treatment for patients with mRCC with good performance status and a limited burden of disease. In the following TKI era, two parallel prospective clinical trials, CARMENA (Cancer du Rein Metastatique Nephrectomie et Antiangiogéniques) and SURTIME (Immediate Surgery or Surgery After Sunitinib Malate in Treating Patients With Metastatic Kidney Cancer), explored the role of CN [5, 26]. The CARMENA trial was a non-inferiority study evaluating the upfront CN strategy, followed by systemic therapy with sunitinib compared to a sunitinib without CN approach. The analysis showed that sunitinib alone was not inferior to the combination of surgery followed by sunitinib in terms of OS (hazard ratio [HR] 0.89) and PFS (HR 0.82) [5]. Stratifying patients according to IMDC risk factors, patients with one risk factor seemed to benefit from surgery, while in patients with at least two risk factors for CN this appeared to be detrimental in terms of survival [27]. The parallel SURTIME trial compared upfront surgery followed by sunitinib therapy with deferred CN after a 3-month sunitinib treatment. No significant difference in PFS was found between the two strategies, while median OS was significantly longer in the arm with deferred CN (32.4 months vs 15.0 months, HR 0.57, p = 0.03). In addition, the SURTIME study highlighted that patients with premature progressive disease do not seem to benefit from surgery and have a worse overall prognosis [26].

To date, the treatment of mRCC has changed substantially. However, the data supporting CN in the modern era of IO combination therapies only come from retrospective analyses. The available data appear to be consistent, suggesting a positive impact of CN on OS of patients with mRCC, while it is performed before or after the initiation of systemic therapy represented by IO+IO or IO+TKI combinations [21, 28, 29].

A recent evidence-based meta-analysis of eight studies including 2397 patients with mRCC treated with various IO therapies (i.e., nivolumab+ipilimumab, nivolumab monotherapy, or interferon-alpha) confirmed the positive survival impact of CN (HR 0.53, p < 0.0001) [30].

Less is known regarding the potential difference between upfront and deferred surgery. Another meta-analysis of nine studies conducted by Li et al., focusing on this point, found a longer OS for patients undergoing deferred CN compared with those undergoing upfront surgery (HR 0.71, p = 0.003) in the overall population of patients with mRCC. However, particularly in patients receiving IO as a systemic therapy, there was no difference in OS between the two CN strategies (p = 0.41) [31]. Similar results came from a recent small prospective randomized study conducted by Shen et al.’s group enrolling 84 patients with mRCC, which reported no OS difference between the two CN strategies (HR 0.814), although deferred CN appeared to improve PFS in patients receiving nivolumab (HR 0.50) [32].

According to current international guidelines for mRCC, upfront CN should be considered only in selected patients, while it should be deferred after IO combination therapy initiation, in those with a clinical response. Patients with a poor-risk prognosis according to Memorial Sloan Kettering Cancer Center/IMDC should not be considered for CN [33].

Beyond providing data for the role of CN in the era of IO combinations, our analysis aimed at exploring a population for whom data are strongly lacking: patients affected by nccRCC. In patients with rare non-clear cell histology types, there is very limited, typically retrospective, evidence supporting CN. Three comparable registry-based assessments were conducted within the Surveillance, Epidemiology, and End Results (SEER) program, involving 951 patients (2000–9), 851 patients (2001–14), and 1573 patients (2006–15) with nccRCC, respectively. The results were a consistent and observed reduction in mortality rates for patients undergoing CN (cancer-specific mortality rate: HR 0.62, p < 0.001, HR 0.38, p < 0.001 and overall mortality rate: HR 0.3, p < 0.001 in the analysis by Aizer et al., Marchioni et al., and Luzzago et al., respectively). Data on the type of systemic therapy received were not available and, with regard to the considered time frame (2000–15), we assumed that the patients were not treated with first-line IO combinations [34,35,36]. Similarly, a Korean retrospective analysis including 156 patients with nccRCC highlighted that patients undergoing CN reported longer cancer-specific survival than those without CN (median cancer-specific survival: 30 months vs 6 months, p < 0.0001). In their study, no patient received first-line IO therapy [37]. Last, the group of Riveros retrospectively analyzed the outcomes of 594 patients with papillary RCC receiving IO or IO+TKI combinations. Their study showed that patients undergoing CN in addition to systemic therapy experienced longer OS (HR 0.59) [38].

Consistent with what has been reported in the literature, our data support the positive prognostic role of upfront CN in patients with metastatic nccRCC, revealing a statistically significant association with longer OS both in univariate and multivariate analysis, despite the different distribution of IMDC risk groups between nephrectomized and non-nephrectomized patients. Moreover, CN was associated with longer OS in every subgroup of patients, stratified according to IMDC risk groups, and by the type or number of metastatic sites (including patients with liver metastases). In our study population, the use of IO combination regimens compared to TKI monotherapy was associated with longer PFS (confined to IMDC intermediate-risk patients) but without differences in terms of OS, indicating no difference in the prognostic impact of CN based on the type of systemic treatment (TKI monotherapy vs IO combinations). Furthermore, both IO+TKI and TKI monotherapy showed superior OS compared with the IO+IO combination. No differences in PFS were observed according to upfront CN, suggesting that the prognostic role of CN may be independent of the benefit obtained from first-line treatment in terms of disease control. Interestingly, among patients treated with IO combinations, those treated with IO+TKI showed both longer OS and PFS compared with patients treated with the IO+IO combination, especially for patients who did not undergo CN. This observation may support the hypothesis that patients with the primary tumor in place may have a greater need for the anti-angiogenic activity of TKIs. Additionally, our results suggest that TKI monotherapy is still an acceptable front-line systemic treatment for patients with nccRCC, particularly for those who are not candidates for the IO+TKI combination. However, the results of these subgroup analyses in the present study should be taken with caution because of the limited number of patients in the specific subgroups and the selected population of patients with synchronous metastatic disease.

Our study has some limitations that should be noted. These include the retrospective design, the absence of a comparison between upfront and deferred CN, the absence of a central pathology review, and the relatively small sample size. We have to point out that the high number of intermediate-risk patients in the CN group and poor-risk patients in the non-CN group and a consequent selection bias may raise uncertainty regarding the difference in OS observed in this study.

Several ongoing prospective clinical trials aim to clarify the role of CN in the context of IO combinations: the Cyto-KIK trial (NCT04322955) is testing a combination with nivolumab+cabozantinib, PrimerX (NCT05941169) immunotherapy, while the PROBE trial (NCT04510597) and NORDIC-SUN trial (NCT03977571) are testing any currently available combination. All these trials plan to include patients with nccRCC. Hopefully, these results will harden the evidence and elucidate the role of CN in the future.

5 Conclusions

Our analysis adds strength to the current argument for incorporating CN into a systemic therapy for metastatic nccRCC. When feasible, upfront CN should also be considered in selected patients with non-clear cell histological types of RCC as it is associated with longer OS irrespective of the choice of first-line systemic treatment. The use of IO+TKI combination or, eventually, TKI monotherapy might be a better choice than IO+IO combination for patients who are not candidates for CN regardless of IO eligibility.