Background

To date, no population-based study has reported a survival benefit in most contemporary community-based patients with clear-cell metastatic renal cell carcinoma (ccmRCC) receiving active treatment, relative to their historical counterparts. Such a survival benefit would be expected to be detectable based on improved survival of contemporary patients with ccmRCC, relative to their historical counterparts within prospective randomized trials. For example, after the introduction of immune checkpoint inhibitors (ICIs), Rini et al¹ showed that a combination of pembrolizumab and axitinib improved 12-month overall survival (OS) from 78% to 90%, compared to monotherapy with the tyrosine kinase inhibitor (TKI) sunitinib (hazard ratio [HR], 0.53; 95% CI, 0.38–0.74; P<.001). Similarly, other phase III trials also demonstrated significant OS improvements with ICI-based combination therapies, compared to sunitinib monotherapy.^2–4 We tested whether the same phenomenon may be detected in the most contemporary SEER database, where population-based survival trends in patients with ccmRCC can be examined. We hypothesized that improvements in survival recorded in contemporary community-based patients have improved over time in a manner similar to findings of phase III studies. We also postulated that improved survival in contemporary community-based patients applies to the group as a whole, as well as to subgroups according to treatment type: systemic therapy alone (ST) versus cytoreductive nephrectomy alone (CN) versus combination of ST and CN (ST+CN). To address these knowledge gaps, we tested for OS differences between the most contemporary (2017–2020) versus historical (2010–2016) patients with ccmRCC identified within the SEER database.

Patients and Methods

Results

Descriptive Characteristics

In the overall cohort of patients with ccmRCC (Table 1), across all treatment types (ST alone, CN alone, and ST+CN), 993 (32%) contemporary and 2,106 (68%) historical patients were identified. Contemporary patients exhibited the highest rate of CN alone treatment (55%; P<.001). Conversely, historical patients exhibited the highest rate of treatment with ST+CN (48%; P<.001). Between 2010 and 2020 (Figure 1), the rate of patients treated with CN alone, relative to all actively treated patients, increased over time (EAPC: +0.054; P=.020). Conversely, the rate of patients treated with ST+CN, relative to all actively treated patients, decreased over time (EAPC: −0.05; P=.022). The rate of patients treated with ST alone, relative to all actively treated patients, did not differ over time (P=.349).

Table 1.

Baseline Characteristics of Community-Based Patients With ccmRCC Undergoing Active Treatment, Within the SEER Database, According to Contemporary (2017–2020) vs Historical (2010–2016) Membership

View Table

View Full Size

Annual rates of treatment types (ST alone vs CN alone vs ST+CN) for community-based patients with ccmRCC.

Abbreviations: ccmRCC, clear-cell metastatic renal cell carcinoma; CI, confidence interval; CN, cytoreductive nephrectomy; ST, systemic therapy.

Citation: Journal of the National Comprehensive Cancer Network 22, 6; 10.6004/jnccn.2024.7011

Among all patients (N=3,099), across all treatment types, contemporary patients were older (median age, 64 vs 62 years; P<.001) and exhibited a higher rate of grade 3–4 tumors (78% vs 73%; P=.002), compared to historical patients (Table 1). Conversely, they exhibited a lower rate of N1 stage (26% vs 39%; P<.001) compared to historical patients. No differences between contemporary and historical patients were recorded for tumor size, distribution of sex, and T stage. Of note, only among those treated with ST alone (Table 2), contemporary patients exhibited larger tumors than their historical counterparts (103 vs 90 mm; P=.02). Conversely, no differences in tumor size between contemporary and historical patients were recorded in the CN alone and ST+CN subgroups (Table 2).

Table 2.

Baseline Characteristics of Contemporary vs Historical Community-Based Patients With ccmRCC According to Treatment Type

View Table

Overall Mortality of Contemporary Versus Historical Patients

In the overall cohort (N=3,099), across all treatment types, median OS was 41 months in contemporary patients versus 25 months in historical patients (Δ=16 months; P<.001, and HR, 0.7; 95% CI, 0.6–0.8; P<.001), and 3-year OS was 52% versus 39%, respectively (Δ=13%; P<.001). In multivariable Cox-regression analyses (Table 3), contemporary membership was independently associated with lower overall mortality (hazard ratio [HR], 0.7; 95% CI, 0.6–0.8; P<.001).

Table 3.

Univariable and Multivariable Cox-Regression Analyses Predicting OM in Patients With ccmRCC Within the SEER Database, According to Contemporary (2017–2020) vs Historical (2010–2016) Membership

View Table

In the subgroup of patients treated with ST alone (n=394; Table 2), median OS was 17 months in contemporary patients versus 10 months in historical patients (Δ=7 months; P<.001, and HR, 0.7; 95% CI, 0.5–0.9; P=.008), and 3-year OS was 25% versus 14%, respectively (Δ=11%; P<.001). In multivariable Cox-regression analyses (Table 3), contemporary membership was independently associated with lower overall mortality (HR, 0.7; 95% CI, 0.5–0.9; P=.005).

In the subgroup of patients treated with CN alone (n=1,376; Table 2), median OS was not reached in contemporary patients versus 33 months in historical patients (Δ=not available; P<.001, and HR, 0.7; 95% CI, 0.6–0.8; P<.001), and 3-year OS was 59% versus 48%, respectively (Δ=11%; P<.001). In multivariable Cox-regression analyses (Table 3), contemporary membership was independently associated with lower overall mortality (HR, 0.7; 95% CI, 0.6–0.9; P<.001).

In the subgroup of patients treated with ST+CN (n=1,329; Table 2), median OS was 38 months in contemporary patients versus 26 months in historical patients (Δ=12 months; P<.001, and HR, 0.7; 95% CI, 0.6–0.9; P<.001), and 3-year OS was 51% versus 38%, respectively (Δ=13%; P<.001). In multivariable Cox-regression analyses (Table 3), contemporary membership was independently associated with lower overall mortality (HR, 0.7; 95% CI, 0.6–0.9; P=.003).

Discussion

In community-based patients with ccmRCC undergoing active treatment (ST alone, CN alone, or ST+CN), no previous population-based analysis has identified a survival benefit in contemporary patients relative to their historical counterparts.^12,13 This observation is in disagreement with prospective randomized trials of ST in ccmRCC, in which contemporary patients demonstrate increasingly more favorable survival over time relative to their historical counterparts.^{1–4,6,14–16} This phenomenon is most likely related to the introduction of novel systemic therapies, in addition to the use of an increasingly higher number of treatment lines and possibly the more judicious use of CN.¹⁷ We hypothesized that the evolving treatment paradigms, recommended based on phase III trials, may have also translated into improved survival over time that applies to most contemporary patients with ccmRCC treated in the community. We relied on the 2010–2020 SEER database and made several noteworthy observations.

First, the contemporary cohort consisted of patients with ccmRCC treated between 2017 and 2020, relative to their historical counterparts who were treated between 2010 and 2016 (Table 1). In the contemporary cohort, a higher proportion of those treated with CN alone was recorded (n=547, 55%), relative to the historical cohort (n=829; 39%; P<.001). Conversely, in the contemporary cohort, the rate of treatment with ST+CN decreased (n=320; 32%), relative to the historical cohort (n=1,009; 48%). Finally, no difference in rate of treatment with ST alone was recorded between contemporary (n=126; 13%) and historical patients (n=268; 13%). The historical proportions agreed with similar historical studies based on SEER and National Cancer Database (NCDB) data.^12,13 However, historical studies applied different selection criteria, such as exclusion of patients treated with CN alone or inclusion of patients with ccmRCC who benefitted from neither ST nor CN.^12,13,18 Consequently, direct comparisons between the current study and historical studies, in which different working definitions were used, cannot be made. Of note, in contemporary community-based patients with ccmRCC undergoing active treatment (ST alone, CN alone, or ST+CN), the rate of CN alone has increased. Conversely, the rate of ST+CN alone has decreased. This phenomenon may be attributable to an increasingly larger proportion of favorable-risk patients, in whom up-front CN represents the treatment option of choice. It is highly likely that with increased data maturity, the proportion of patients treated with CN alone will decrease, because most if not all will eventually require ST. This observation about ST and CN rates should ideally be validated in the NCDB or a similar large-scale population-based data repository.

Second, we identified important differences in baseline patient and tumor characteristics between contemporary and historical patients. Specifically, across all treatment types (Table 1), contemporary patients with ccmRCC were older (median age, 64 vs 62 years; P<.001) and exhibited higher rate of grade 3–4 tumors (78% vs 73%; P=.002). Conversely, contemporary patients exhibited a lower rate of N1 stage (26% vs 39%; P<.001). Finally, no differences between contemporary and historical patients were recorded regarding sex, tumor size, and T stage. Presence of these differences indicates the need for multivariable adjustment, and validated the use of multivariable models within the methodology of the current study. Moreover, these observations indicate that contemporary patients were at an age and a grade disadvantage for survival. However, they also exhibited more favorable N stage. Consequently, it cannot be stated that baseline patient and tumor characteristics favored contemporary patients in a way that improved survival would be expected based on age and tumor characteristics alone.

Third, in the entire cohort (N=3,099), regardless of treatment type (Figure 2), median OS was 41 months in contemporary patients versus 25 months in historical patients (Δ=16 months; HR, 0.7; P<.001). These differences between contemporary and historical patients resulted in a multivariable HR of 0.7 (P<.001). This observation indicates that contemporary membership is associated with better survival than historical membership in community-based patients with ccmRCC undergoing active treatment. This observation is in agreement with phase III studies, in which contemporary patients treated with immunotherapy exhibited better survival relative to the historical patients treated with TKIs.¹⁹ To the best of our knowledge, we are first to record and report such difference favoring contemporary patients with ccmRCC undergoing active treatment within a population-based data registry. In other words, the current study represents the first population-based survival comparison between the era of ICI-based combination therapies (2017–2020) versus the era of targeted molecular therapies (2010–2016). Previously, a historical population-based study demonstrated improved survival rates in the era of targeted molecular therapies (2010–2017) relative to the era of cytokine-based therapies (2000–2009).²⁰ These results indicate that the OS improvements observed in prospective randomized trials translate in a timely manner to the benefit of community-based patients with ccmRCC undergoing active treatment, despite rapid modifications in the standard of care.

View Full Size

Kaplan-Meier plot illustrating overall survival of community-based patients with ccmRCC undergoing active treatment, according to contemporary (2017–2020) vs historical (2010–2016) membership.

Abbreviation: ccmRCC, clear-cell metastatic renal cell carcinoma.

Citation: Journal of the National Comprehensive Cancer Network 22, 6; 10.6004/jnccn.2024.7011

Fourth, to further explore the differences between contemporary and historical patients, we relied on subgroup analyses according to treatment type (ST alone, CN alone, or ST+CN; Figure 3). In all 3 subgroups, the relative survival differences invariably favored contemporary patients over their historical counterparts. The relative differences in median and 3-year OS were virtually the same across the 3 examined treatment subgroups. Similarly, after multivariable adjustment, the same magnitude of overall mortality decrease between contemporary and historical community-based patients was recorded (Table 3). Taken together, our observations indicate the same degree of improvement in OS across all 3 examined treatment types. However, when median OS values were examined, notable differences were recorded (Figure 3). Specifically, median OS duration was lowest in patients treated with ST alone (17 months in contemporary vs 10 months in historical), relative to the intermediate median OS recorded in patients treated with ST+CN (38 months in contemporary vs 26 months in historical) and the longest median OS recorded in patients treated with CN alone (not reached in contemporary vs 33 months in historical). These observations are consistent with a previous report, in which patients treated with ST+CN exhibited better risk profiles according to the International mRCC Database Consortium (IMDC) risk groups, compared to those treated with ST alone.^21,22 Moreover, these results are in agreement with treatment assignment, according to guideline recommendations.²³ Specifically, in patients treated with CN alone, most would be expected to be in the IMDC favorable risk group.²² Conversely, in patients treated with ST+CN, most would be expected to be in the IMDC intermediate or poor risk group.²¹ Finally, in patients treated with ST alone, most would be expected to be in the IMDC poor risk group.^7,21 These differences in survival according to treatment type validate the need for the methodology used in the current study, in which survival rates of contemporary and historical patients were compared separately within each of these 3 treatment types.

View Full Size

Kaplan-Meier plot illustrating overall survival of community-based patients with ccmRCC undergoing active treatment, according to treatment type: (A) ST alone, (B) CN alone, and (C) ST+CN.

Abbreviations: ccmRCC, clear-cell metastatic renal cell carcinoma; CN, cytoreductive nephrectomy; ST, systemic therapy.

Citation: Journal of the National Comprehensive Cancer Network 22, 6; 10.6004/jnccn.2024.7011

Taken together, across all treatment types, contemporary community-based patients with ccmRCC undergoing active treatment exhibited better OS than their historical counterparts. The relative magnitude of the difference in OS between contemporary and historical patients was the same, when subgroup analyses addressed treatment with ST alone versus CN alone versus ST+CN. Lack of differences in the relative magnitude implies that improved survival between contemporary versus historical patients applies equally to all treatment types. However, within the 3 examined treatment types, the absolute median OS values were longest in the CN alone subgroup, intermediate in the ST+CN subgroup, and shortest in the ST alone subgroup. These absolute median OS value differences may be readily explained based on patient characteristics that translate into IMDC risk groups. Specifically, patients treated with CN alone are expected to be predominantly in the IMDC favorable risk group, those treated with ST+CN are expected to be predominantly in the IMDC intermediate or poor risk group, and those treated with ST alone are expected to be predominantly in the IMDC poor risk group. Finally, it may be postulated that contemporary community-based patients benefit more extensively from various treatment options for ccmRCC than their historical counterparts. Our observations should ideally be validated in other large prospective databases, such as the NCDB.

Despite its novelty, the current study has several limitations. First, due to the retrospective nature of the analyses, potential residual selection biases and confounders might have remained even after stratification according to different treatment types. Second, our data lack information regarding laboratory variables and performance status. Therefore, we could not apply the IMDC risk criteria by Heng et al²² to stratify the analyses. However, no previous large-scale analysis that relied on either the SEER¹² database or the NCDB¹³ applied the Heng criteria for purpose of adjustment or stratification. Third, within the SEER database, information on specific ST type and its timepoint, dose, and duration is not available. Despite residual limitations, the current study provides a novel perspective on survival benefits among contemporary versus historical community-based patients with ccmRCC undergoing active treatment.

Conclusions

Contemporary community-based patients with ccmRCC undergoing active treatment clearly exhibit better survival than their historical counterparts, when examined as one group, as well as when examined as separate subgroups according to treatment type. Treatment advancements of phase III trials seem to be applied appropriately outside of centers of excellence.