Evolving Biological Associations of Upfront Cytoreductive Nephrectomy in Metastatic Renal Cell Carcinoma

Abstract

Background

Systemic responses to cytoreductive nephrectomy (CN) in the management of metastatic renal cell carcinoma (mRCC) are variable and difficult to anticipate. We aim to determine the association of CN on modifiable International mRCC Database Consortium (IMDC) risk factors and oncological outcomes.

Methods

Consecutive mRCC patients referred for potential CN (2009–2019) were reviewed. The primary outcome was overall survival (OS); variables of interest were undergoing a CN and baseline number of modifiable IMDC risk factors (anemia, hypercalcemia, neutrophilia, thrombocytosis, reduced performance status). For operative cases, we evaluated effects of IMDC risk factor dynamics, measured 6-weeks and 6-months post-CN, on OS and postoperative treatment disposition.

Results

Of 245 treatment-naïve mRCC patients referred for CN, 177 (72%) proceeded to surgery. The CN cases had fewer modifiable IMDC risk factors (p=0.003), including none in 71/177 (40.1%), fewer metastases (p=0.011), and higher proportions of clear cell histology (p=0.012). On multivariable analysis, surgical selection, number of IMDC-risk factors, metastatic focality and histology were associated with OS.

Total risk factors changed for 53.8% and 57.2% of patients comparing pre-operative vs. 6-weeks and 6-months post-CN, respectively. Adjusted for preoperative IMDC-risk scores, an increase in IMDC-risk factors at 6-weeks and 6-months associated with adverse OS (HR 1.57, 95% CI 1.13 – 2.19; p=0.007; HR 2.52, 95% CI 1.74 – 3.65; p<0.001).

Conclusion

IMDC risk factors are dynamic clinical variables that can improve after upfront CN in select patients, suggesting a systemic benefit of cytoreduction, which may confer clinically meaningful prognostic implications.

Introduction

The management of metastatic renal cell carcinoma (mRCC) is rapidly evolving with advances in systemic therapy, surgery and radiation techniques^1–3. To aid in clinical decision making, validated models have been developed to risk stratify patients at the time of systemic therapy into favorable, intermediate, and poor-risk disease. These models include the Memorial Sloan Kettering Cancer Center (MSKCC) criteria, developed in the cytokine era, and the International Metastatic Database Consortium (IMDC) Risk Model, developed in the targeted therapy era. The IMDC risk criteria comprise laboratory measurements (anemia, hypercalcemia, neutrophilia and thrombocytosis), performance status assessment and the time from cancer diagnosis to commencing systemic therapy. While the latter criterion is a fixed time-dependent variable, the other risk components can theoretically change. As risk stratification not only remains prognostic but also helps guide therapy selection, studies of treatment modalities in the context of these subgroups have wide applicability for mRCC management.

Early randomized clinical trials highlighted that cytoreductive nephrectomy (CN) followed by cytokine therapy^{4, 5} yielded an overall survival (OS) benefit of 6 months for patients treated with upfront surgery⁶. Postoperative responses, though, varied significantly and unpredictably², and predictive biomarkers to guide selection were lacking. With the advent of effective targeted agents, concerns regarding delays and/or potential complications from CN which may preclude patients from systemic therapy were raised⁷. To address this, the phase III CARMENA trial evaluated the benefit of upfront CN compared to sunitinib monotherapy and reported non-inferior OS for intermediate- and poor-risk mRCC patients randomized to upfront sunitinib⁸. Based on these results, National Comprehensive Cancer Network guidelines recommend restricting cytoreductive nephrectomy to select patients⁹ and European Association of Urology guidelines recommended against undertaking immediate CN in intermediate- and poor-risk patients requiring active treatment¹⁰.

The potential benefits of cytoreduction include relieving primary tumor (e.g. symptomatic hematuria) and paraneoplastic symptoms, obtaining histologic diagnoses with a better understanding of tumor heterogeneity, and debulking to potentially enable a period of postoperative active surveillance^{11, 12}. However, these benefits must be balanced with the costs of delaying systemic therapy especially with new first-line agents. Importantly, selection for CN is not randomized and relies on multidisciplinary input to select patients¹³. As the IMDC model reflects the systemic level of mRCC aggressiveness, the present study evaluates whether cytoreduction can reduce IMDC risk factors and investigates the subsequent clinical effects of such modification on cancer-specific outcomes.

Patients & Methods

Patient population

Following Institutional Review Board approval, 312 MSKCC urology service referrals between January 2009–February 2019 for consideration of a CN were reviewed. Patients with non-RCC diagnoses (n=15), prior treatment with systemic therapy (n=44) or unknown IMDC-risk factors (n=8) were excluded, leaving 245 patients eligible for our analysis. Patients never referred for consideration of CN were not included in the analysis. Baseline demographic and clinical characteristics were collated from a prospectively-maintained RCC database.

The primary variables of interest were the type and number of modifiable IMDC-risk factors (anemia, hypercalcemia, neutrophilia, thrombocytosis, and reduced Karnofsky Performance Status (KPS)) calculated at the time of initial clinical evaluation. As the final IMDC risk factor (“less than one year from diagnosis to systemic therapy”) is calculated at the time of commencing therapy, this variable was not assessed in this treatment-naïve cohort. In patients who proceeded to undergo CN, modifiable IMDC risk factors were reassessed six-weeks and six-months post-surgery, with KPS documented at each clinic appointment.

The sites and number of organs with metastases at the time of referral were coded based on radiographic imaging, grouping sites broadly into viscera, bone, nodes and other sites and unifocal (metastasis in one organ) or multifocal (multiorgan metastases). These variables were stratified by selection for CN, using Fisher’s exact test for categorical variables and Mann-Whitney U test for continuous variables. The postoperative management disposition was assessed for CN cases and classified into immediate active treatment of metastatic disease (systemic or focal therapy), active surveillance of residual disease (watchful waiting), no evidence of residual disease or perioperative death (within the initial admission).

Statistical analyses

OS was defined as the time from urology clinic presentation to death from any cause, with patients alive at data cut-off censored at last follow-up. Kaplan-Meier method was used to estimate OS, stratifying patients by IMDC-risk factors and surgical selection (proceed to CN yes / no). To assess whether surgical selection and IMDC-risk factors were independently associated with survival, a multivariable Cox regression model was constructed integrating the following variables: primary tumor size, histology, metastatic focality, and by presentation before/after 2015 (FDA-approval for Nivolumab)¹⁴.

Among CN cases, change in individual risk factors pre- and post-CN were investigated, and associations between such IMDC dynamics and OS were assessed. Pairwise probabilistic co-occurrence analyses of baseline IMDC-risk factors were performed to identify non-random risk factor presentation patterns¹⁵. The time from CN to subsequent treatment was measured and the postoperative course of patients intended for systemic therapy was plotted, testing for factors associated with time to commencing treatment, either immune-oncology drugs (combination or monotherapy) (IO) or other targeted treatments (non-IO).

As CN is generally discouraged in patients with multiple adverse risk features, a subset analysis was undertaken of this specific group. A heatmap was created for patients with ≥2 IMDC risk factors illustrating the presentation, demographics and pathology, risk factor altering treatment, and postoperative IMDC response to surgery. Kaplan-Meier method was used to estimate the effect of reducing IMDC risk factors on the primary endpoint.

For all analyses, a p-value < 0.05 was considered statistically significant. Statistical analyses were conducted using R version 3.6.2 (R Core Development Team, Vienna, Austria).

Results

Patient Characteristics

From a cohort of 245 treatment-naïve mRCC patients referred for CN between 2009–2019, 177/245 (72%) proceeded to CN (Supplemental Figure 1). The median patient age was 62 (IQR: 54, 69) and 183/245 (74.7%) were males. Patients selected for CN had fewer modifiable IMDC risk factors (p=0.003) (anemia, hypercalcemia, neutrophilia, thrombocytosis and KPS < 80), with no risk factors present in 71/177 (40.1%) vs. 15/68 patients (22.1%) who did not proceed with surgery (Table 1). Primary tumor size was comparable (median 7.8 cm, IQR: 5.8, 9.5), with higher proportions of patients selected for CN harboring clear cell histology (84.2% vs 69.1%; p=0.012). Most common metastatic sites were lung (144, 58.8%) and appendicular skeleton (60, 24.5%) (Supplementary Table 1). Non-operative cases had more metastatic sites per patient (p=0.011), particularly bone metastases (p=0.019).

Table 1.

Cohort characteristics by operative selection

	Non-operative	Operative	p-value
Cohort	68	177
Male gender (%)	50 (73.5)	133 (75.1)	0.870
Age (median [IQR])	64.50 [56.00, 72.25]	62.00 [54.00, 69.00]	0.084
Patient race (%)			0.235
Asian	2 (2.9)	3 (1.7)
Black	6 (8.8)	7 (4.0)
Patient Refused/Unknown	7 (10.3)	13 (7.3)
White	53 (77.9)	154 (87.0)
Smoking history (%)			0.624
Current	11 (16.2)	26 (14.7)
Former	28 (41.2)	72 (40.7)
Never	28 (41.2)	69 (39.0)
Unknown	1 (1.5)	10 (5.6)
eGFR (median [IQR])	70.34 [53.01, 95.09]	73.09 [58.66, 90.49]	0.569
Diabetes (%)	13 (19.1)	37 (20.9)	0.86
Radiographic tumor size (median [IQR])	7.60 [5.50, 9.10]	8.00 [5.90, 9.60]	0.196
Histology (Clear Cell RCC) (%)	47 (69.1)	149 (84.2)	0.012
No. of metastatic organs (median [IQR])	2.00 [1.00, 3.00]	2.00 [1.00, 2.00]	0.011
No. of modifiable IMDC Risk Factors (%)			0.003
0	15 (22.1)	71 (40.1)
1	19 (27.9)	60 (33.9)
2	22 (32.4)	26 (14.7)
>2	12 (17.6)	20 (11.3)
Visceral metastases (%)	46 (67.6)	136 (76.8)	0.146
Bone metastases (%)	34 (50.0)	59 (33.3)	0.019
Nodal metastases (%)	29 (42.6)	59 (33.3)	0.184
Other sites of metastases (%)	9 (13.2)	11 (6.2)	0.114

The median follow-up of all patients was 26.7 months (10.6, 44.2) with 148 (60.4%) deaths: 102 in the surgical arm and 46 in the non-surgical arm. Median OS was 25.8 months (95% CI: 18.6 – 32.5). On univariable analysis, the number of IMDC risk factors (Figure 1) (Table 2) and surgical selection were associated with improved OS (HR 0.38, 95% CI 0.26 – 0.54; p<0.001) (Supplemental Figure 2). On multivariable analysis, fewer IMDC risk factors, surgical selection, unifocal metastasis and clear cell histology were all associated with improved OS (Figure 2). Two disease sites were associated with adverse OS: retroperitoneal nodes (HR 2.60, 95% CI 1.81 – 3.74; p<0.001) (n=54) and cranial metastases (HR 1.95, 95% CI 1.08 – 3.53; p=0.028) (n=15).

Figure 1.

Overall survival by the number of abnormal preoperative modifiable IMDC risk factors

Table 2.

Univariable overall survival analysis in all patients (n=245)

Univariable OS Analysis
Covariate	Category (Ref.)	HR	95% CI	p-value
Age at clinic review	Age (Years)	1.00	0.98–1.01	0.7
Gender	Male (Ref. Female)	1.14	0.79–1.66	0.49
eGFR at encounter	eGFR (mL/min/1.73m2)	1.00	0.99–1.01	0.8
Modifiable IMDC Risk Factors	1 RF (Ref. 0)	2.00	1.31–3.07	0.001
	2 RF (Ref. 0)	2.77	1.75–4.40	<0.001
	>2 RF (Ref. 0)	3.93	2.30–6.73	<0.001
Surgical selection	Operative (Ref. Non-operative)	0.38	0.26–0.54	<0.001
Tumor Size	Maximum Size (cm)	1.05	1.00–1.10	0.03
Histology	Non-Clear Cell (Ref. Clear Cell)	2.82	1.96–4.05	<0.001
Metastatic Focality	Multifocal (Ref. Unifocal)	1.96	1.39–2.74	<0.001

Figure 2.

Multivariable forest plot analysis for overall survival

Preoperative and Postoperative IMDC Risk Status

We were interested to see what effects CN had on IMDC scores, hypothesizing that clinically meaningful cytoreduction could result in lowering of modifiable IMDC scores. Postoperative IMDC risk scores for patients who underwent CN could be computed for 156/175 and 138/159 patients alive at 6-weeks and 6-months, respectively. The median number of risk factors preoperatively, at 6-weeks and 6-months post-surgery, was 1 (IQR: 0,2), 1 (IQR: 0,1) and 0 (IQR: 0,1) respectively. On univariable analysis, the number of IMDC risk factors assessed preoperatively, at 6-weeks and 6-months after surgery were all associated with adverse OS (HR 1.41, 95% CI: 1.19 – 1.68; p<0.001; HR 1.55, 95% CI: 1.25 – 1.92; p<0.001; HR 2.43, 95% CI: 1.85 – 3.21; p<0.001).

At 6-weeks, 44/156 (28.2%) of patients achieved a decrease in the total IMDC score by improving modifiable risk factors. IMDC scores remained unchanged for 72/156 (46.2%) of CN patients and 40/156 (25.6%) of patients worsened their total IMDC score. However, 71/177 (40.1%) patients had zero preoperative IMDC risk factors and therefore their best possible response to surgery was to not increase risk factors, while two (1.1%) patients had five risk factors present and thus could not increase their total number of risk factors (Figure 3A). Excluding these patients and adjusting for the number of preoperative risk factors, a rising number of risk factors from baseline to 6-week (n=12; 13%) and 6-month (n=10; 13%) postoperative landmarks conferred adverse OS (HR 1.57, 95% CI 1.13 – 2.19; p=0.007; HR 2.52, 95% CI 1.74 – 3.65; p<0.001) (Figure 3B). Conversely, a reduction in risk factors at 6-weeks (n=43; 47%) was associated with improved OS (HR 0.64, 95% CI 0.46 – 0.89; p=0.007). Of 84 patients with modifiable preoperative risk factors, the eight patients rendered NED had fewer risk factors at the 6-week landmark compared with all other patients with residual metastases (p=0.086). Three of the eight (38%) patients rendered NED improved their sum of modifiable IMDC factors to 0 and none increased their number of risk factors. By comparison, in patients with residual disease (n=76), only 19 (22.1%) achieved 0 modifiable factors postoperatively and 12/76 (14%) increased their number of risk factors.

Figure 3A.

Number of abnormal modifiable IMDC risk factors at baseline, 6-weeks and 6-months after surgery

Figure 3B.

Overall survival stratified by changes in the number of IMDC risk factors 6-weeks after surgery

The perioperative dynamics of individual IMDC factors was examined. Half of the modifiable risk factors of patients selected for surgery co-occurred in non-random patterns. Elevated platelets co-occurred with all other laboratory risk factors and reduced KPS presented concurrently with neutrophilia (Supplementary Figure 3). For patients proceeding to CN, the most common preoperative risk factor was anemia (44.8%) and the least common was a KPS < 80 (6.7%). In terms of the effect of CN, 12/13 (92%) patients with an elevated calcium and 25/31 (81%) patients with neutrophilia normalized at 6-weeks postoperatively. Conversely, surgery provided minimal improvement for hemoglobin: 74/165 and 70/165 patients were anemic preoperatively and postoperatively, respectively, with preoperative anemia persisting in 40/74 (54%) cases and first presenting postoperatively in 30/91 (33%) patients (Figure 4).

Figure 4.

Response of IMDC risk factors to surgery

Subset analysis in patients with 2+ modifiable RF

Of 80 patients with ≥ 2 modifiable IMDC risk factors referred to the urology service, 46 (58%) proceeded with surgery. The patients managed surgically were more likely to have hypercalcemia (26.1% vs 5.9%; p=0.034), a KPS ≥ 80 (80.4% vs 47.1%; p=0.004) and fewer metastases (p=0.057). Six-weeks after CN 23/46 patients were alive with < 2 risk factors, 20 were alive with ≥ 2 risk factors, one was unable to be risk-assessed and two had died (Figure 5A). There was no association between preoperative comorbidities and postoperative IMDC changes (Supplementary Figure 4). Among the 46 CN cases with ≥ 2 modifiable IMDC risk factors, reducing IMDC risk factors to ≤ 1 modifiable risk factor (i.e. improving the theoretical prognostic group) at either the 6-week (p=0.036) or 6-month landmarks (p<0.001) corresponded to a significant OS benefit (Figure 5B).

Figure 5A. Heatmap of patients with two or more baseline risk factors selected for surgery.

Symptomatology assessed preoperatively. Two patients with missing postoperative risk factors are excluded.

Figure 5B.

Overall survival of all CN patients with two or more baseline risk factors stratified by the 6-week postoperative response

Clinical Disposition Following CN

The immediate post-CN dispositions included 17 (9.6%) patients with no evidence of disease, 101 (57.1%) patients treated with systemic therapy, 14 (7.9%) treated with focal therapy only (metastasectomy, radiation or ablation), four (2.3%) perioperative deaths, one (0.6%) lost to follow-up and 40 (22.6%) surveilled off all treatment for progression of residual disease (Supplementary Figure 5). Of the patients with residual disease, active surveillance was more likely to be selected in patients with fewer IMDC risk factors (p=0.045), and immediate systemic therapy was more likely to be pursued in patients with a higher number of risk factors (p=0.031) (Figure 6). The median duration of active surveillance was 18.4 months (11.1, 35.2), with an estimated one-year treatment-free survival of 63.1% (95% CI: 49.1%, 81.3%).

Figure 6.

Postoperative active surveillance vs systemic therapy by IMDC risk factors

For the 101 patients commencing postoperative systemic therapy, the median time from surgery to systemic therapy was 7 weeks (5.4, 8.4) (Supplementary Figure 6). Seventeen (16.8%) patients received concurrent focal therapy, at a median of 4 weeks (3.3, 5.7) from surgery. There was no difference between preoperative IMDC-risk groups or type of systemic therapy (IO (n= 21) vs non-IO (n= 80)) with respect to time to subsequent therapy.

Discussion

Our study analyzes IMDC risk factor dynamics among mRCC patients carefully selected for CN. Modifiable IMDC risk factors were demonstrated to be dynamic and multidirectional, with a CN often altering the number of postoperative risk factors. While the number of risk factors at presentation is a robust prognosticator, the change in the number of IMDC risk factors was also associated with survival.

The IMDC risk model is a validated prognostic tool that is routinely utilized for risk stratification at the time of initiating systemic therapy for metastatic disease. The OS differences between IMDC favorable-, intermediate- and poor-risk patients are profound¹⁶. Furthermore, the IMDC groups serve as surrogates for underlying disease biology, including angiogenesis and macrophage infiltration within the tumor microenvironment^{17, 18}. Importantly, less than 10% of IMDC poor-risk patients achieve objective responses to systemic therapy in their primary tumor¹⁹. Patients that reduced their postoperative IMDC risk factors tended to have larger tumors. In this context, integrating upfront CN into the management of mRCC for select patients is done in an effort to remove what is often the most resistant and bulky disease site, to reduce IMDC risk factors before commencing systemic therapy, and to improve OS.

CARMENA was a phase III study which randomized 450 patients with treatment-naïve, MSKCC intermediate-poor risk metastatic RCC, to sunitinib or CN followed by sunitinib. The study’s primary endpoint was OS and found that sunitinib alone was non-inferior to nephrectomy followed by sunitinib ⁸. The trial’s key implication was that CN benefits are not ubiquitous among patients with high-risk metastatic clear cell RCC and therefore surgical selection must be judicious. At our institution, CN remains a relevant component for select patients in mRCC management ¹³ and herein we evaluated the effects of CN on these select patients. We hypothesized that the systemic benefit of CN, by significantly lowering disease burden, would manifest in modifying patients’ IMDC scores. Further, we theorized that these modifications would also be associated with an improvement in patient outcomes.

The IMDC-risk model weighs baseline factors equally, however we demonstrate that there are distinct differences between the factors. In our cohort, surgeons generally selected patients with a KPS > 80 for CN. A novel finding was that hypercalcemia and neutrophilia were the most likely parameters to normalize with surgery, while anemia was the least likely risk factor to resolve post-CN. Importantly, patients that presented with or developed hematuria postoperatively were included in the study. We could not ascribe prognostic groups to patients, as the non-modifiable risk factor, “less than one year from diagnosis to systemic therapy”, is determined at the time of starting systemic therapy²⁰. Nevertheless, even with excluding this variable, a decrease in IMDC risk factors after CN was associated with improved survival. A similar evaluation of IMDC risk factor dynamics in patients receiving systemic therapy may enable longitudinal monitoring of tumor response, with the added benefit of classifying risk groups by also incorporating the non-modifiable risk factor of less than one year from diagnosis to commencing systemic therapy. We demonstrated that patients with ≥2 risk factors (who would likely be poor-risk) were less likely to reduce their number of IMDC factors if they had non-clear cell histology, sarcomatoid de-differentiation or systemic symptoms, corroborating studies showing that these features confer poorer prognoses^21–23. Furthermore, among these patients with high risk disease, a postoperative reduction in risk factors was associated with improved OS.

Ultimately, CN is most valuable if it meaningfully alters the systemic trajectory of mRCC for the patient. CN nomograms show that liver, bone and retroperitoneal metastases confer shorter survival ^{23, 24}. Indeed, adding these metastatic sites to IMDC criteria may improve prognostic power²⁵. Conversely, the indolent nature of pancreatic metastases does not mean that prolonged survival following CN justifies the intervention, particularly if it causes iatrogenic morbidity²⁶. Studies have shown that tumor evolution increases with primary tumor growth and is associated with rapid disease progression²⁷. A proposed rationale for early CN is to potentially arrest this evolution and perhaps prevent later metastasis of more aggressive subclones^{28, 29}. Patients who initially present with rapidly progressing or widely metastatic disease may have already metastasized a lethal subclone and may have sustained irreparable organ damage, obfuscating the disease altering intent of CN. While ancillary benefits of avoiding hematuria (which can hospitalize patients and suspend systemic therapy), ameliorating paraneoplastic syndromes and mitigating pain are vital for managing symptoms, they may be compromised by delaying effective systemic therapy.

The novel finding that some patients can correct their modifiable IMDC risk factors suggests that in carefully selected cases, surgery can indeed alter the natural disease course for a patient. This raises a contemporary concern of whether similar risk factor altering effects would occur in the same population with immunotherapy. Following CN, patients progressively reduced their IMDC risk factors from 6-weeks to 6-months, especially those with clear cell RCC. Given that almost all patients began systemic therapy at least 6-weeks after their CN, the continued reduction in risk factors by 6-months suggests that patients are likely benefiting from the sequence of CN followed by systemic therapy. Therefore, the 6-month improvement cannot be used to infer longer term benefits from CN alone. The SURTIME trial specifically explored treatment sequences assessing whether sunitinib prior to CN improved oncological outcomes compared to immediate CN followed by therapy, however the trial’s equivocal findings were limited by significant under-accrual³⁰. Notably, 25% of the patients in our cohort with residual disease postoperatively were surveilled off all treatment, enabling a prolonged treatment-free survival interval for a median duration of 18.4 months. This approach was typically selected for patients with fewer IMDC risk factors and is valuable as it delays the possibility of introducing treatment-related toxicity.

The purpose of our study was not to argue whether CN is indicated in high-risk patients, rather, we were interested in observing IMDC dynamics and cancer-specific outcomes in those subjects carefully selected for CN. Limitations of analyzing the IMDC-risk model include that the model was developed for assessing risk at the time of commencing systemic therapy and generally in patients that have already undergone a nephrectomy³¹. Furthermore, multi-directional dynamics can only be assessed in patients with 1–4 risk modifiable factors. Measuring survival from presentation and stratifying per-treatment introduces an immortality bias. Additionally, there are limitations with generalizing this study to the other centers as it represents a high-volume institution with non-randomized patient selection that follows complex multidisciplinary decision-making and referral screening between both urology and medical oncology. Patients with more advanced disease or decreased performance status may not have been formally referred for evaluation for CN, in part explaining the high proportion selected for surgery. Over the last decade, the landscape of systemic therapies has progressed rapidly with three first-line approvals since 2018, and therefore whether these effects are similarly seen with the use of these new therapies is emerging. Advances in systemic treatments and CN techniques during the study period may introduce selection bias. By including only treatment-naïve mRCC cases, evaluating whether delayed CN after systemic therapy similarly alters IMDC risk factor is unknown. While our understanding of the appropriate role and timing of CN is still evolving, our findings support an important systemic benefit to offering CN in select patients.

Conclusion

Modifiable IMDC risk factors can be viewed as dynamic variables that reflect the underlying disease status. Assessment of risk factor dynamics may help with understanding how CN affects systemic disease. Lessening the modifiable IMDC risk factors with CN appeared to have a favorable effect on cancer-specific outcomes. These findings provide a basis for further exploration into biological characteristics associated with IMDC risk factor change

Supplementary Material

fS1

Supplementary Figure 1. Inclusion criteria

fS3

Supplementary Figure 2. Overall survival stratified by surgical selection

fS2

Supplementary Figure 3 Co-occurrence matrix of IMDC risk factors at baseline

fS4

Supplementary Figure 4. Extended heatmap of patients with two or more baseline risk factors selected for surgery Legend: Symptomatology and comorbidities assessed preoperatively.

fS5

Supplementary Figure 5. Postoperative disposition by IMDC risk factors Legend: Active surveillance – postoperative treatment withheld in a patient with radiographic residual disease; No evidence of disease (NED) – all radiographic metastatic sites treated.

fS6

Supplementary Figure 6. Time from surgery to subsequent treatment for patients requiring postoperative systemic therapy Legend: IO – any therapy combination that includes an immunotherapy; NonIO – any therapy combination that does not include immunotherapy

tS1

Supplementary Table 1. Metastatic sites

Acronyms:

CI

Confidence Interval

CN

Cytoreductive Nephrectomy

HR

Hazard Ratio

IMDC

International Metastatic Renal Cell Carcinoma Database Consortium

IO

immuno-oncology

KPS

Karnofsky performance status

mRCC

Metastatic renal cell carcinoma

MSKCC

Memorial Sloan Kettering Cancer Center

OS

Overall survival

RCC

Renal cell carcinoma

TTF

Time to treatment failure