Survival and treatment of stage IV renal cell carcinoma in academic vs non-academic medical centers

Abstract

BACKGROUND

Renal cell carcinoma (RCC) is treated with surgical resection as the gold standard, as it is notoriously resistant to systemic therapy. Advancements with targeted therapies contribute to declining mortality, but metastatic RCC (mRCC) survival remains poor. One possible factor is treatment at academic centers, which employ advanced providers and novel therapies. This study compared outcomes of mRCC in patients treated at academic/research facilities compared to those treated at non-academic centers.

AIM

To compare survival outcomes of mRCC and their various etiologies between academic and non-academic centers.

METHODS

The National Cancer Database was used to identify mRCC patients including all histology subtypes and stage IV disease. Descriptive statistics and Kaplan-Meier curves measured survival outcomes for user file facility types sorted into a binary academic/research and non-academic research variable. Multivariate logistic regression and Cox proportional hazard testing generated odds ratio and hazard ratio. Data was analyzed using Statistical Package for the Social Sciences version 29.0 using a significance level of P < 0.05.

RESULTS

Overall, academic facility patients experienced greater 5-year and 10-year overall survival than non-academic facility patients. Treatment at non-academic facilities was associated with increased odds of death that persisted even after controlling for age, tumor size, sex, and distance traveled to treatment center. In comparison, non-academic facility patients also experienced greater risk of hazard.

CONCLUSION

Patients with mRCC treated at academic/research facilities experienced increased survival compared to patients treated at non-academic facilities, were more likely to be younger, carry private insurance, and come from a large metropolitan area. They also were significantly more likely to receive surgery and adjuvant immunotherapy.

Key Words: Renal cell carcinoma; Academic; Non-academic; Facility; Center; Type; Survival; Outcome

Core Tip: Treatment of renal cell carcinoma (RCC) has been historically challenging. Our objective is to explore the contextual factors influencing RCC patients and their outcomes, primarily between those receiving care at academic compared to non-academic institutions. Previous studies on other cancers have some but limited insight into the remarkable discrepancy in survival in favor of academic centers. We aim to elucidate these findings for RCC using the National Cancer Database which unprecedentedly now enables analyzing large numbers of patients across long spans of time.

INTRODUCTION

Renal cell carcinoma (RCC) stands as the most common urogenital cancer[1] and ninth most common overall cancer in the world[2]. The most common subtype is clear cell carcinoma accounting for 75% of all RCCs; followed by papillary carcinoma (15%); chromophobe carcinoma (5%); and a miscellaneous collection of others including MiT family translocations (TFE3 fusion with other genes on chromosome Xp11.2), collecting duct, medullary, and oncocytoma that comprise of the remaining 5% of RCCs[3,4]. Multiple hereditary syndromes have also been found to be associated with RCC, accounting for roughly 4% of all cases of RCC[1,5,6]. The majority of RCC patients at the time of diagnosis possess localized tumors, the definitive treatment of the past for primary neoplasms has been radical and partial nephrectomy, with surgical advances in partial nephrectomy techniques via open and robotic approaches enabling greater preservation of renal function[3].

Despite advances in therapeutic approaches, metastatic RCC (mRCC) remains a complicated malignancy to treat, and survival rates remain poor[1]. Widespread mRCC is notoriously resistant to chemoradiotherapy and surgical intervention is associated with high rates of recurrence[1,7]. The introduction of new adjuvant modalities in recent decades that have garnered significant attention from the scientific community and demonstrated promising preliminary improvements to overall survival (OS)[8,9]. A multitude of immune checkpoint inhibitors trials have already been conducted or underway[3,10]. Therefore, advanced knowledge of the ever-changing treatment landscape for mRCC increasingly appears as a necessary prerequisite to obtaining optimal patient outcomes.

In addition to developments in systemic therapy, other variables may be impacting survival in mRCC. One notable variable may be treatment volume and facility type. Prior studies have found that increased treatment volume was associated with improved survival in mRCC, and that academic centers had greater utilization of immunotherapy compared to non-academic facility types. Prior studies on cancers such as non-small cell lung cancer (NSCLC), diffuse large B-cell lymphoma (DLBCL), and multiple myeloma have shown that patients treated at academic centers often achieve superior outcomes compared to those treated at non-academic facilities. These findings have been attributed to a combination of factors, including access to specialized care, clinical trial availability, and multidisciplinary expertise[11-13].

Despite these insights, the disparity in survival and treatment practices between academic and non-academic facilities in mRCC has not been thoroughly investigated. Understanding these differences is essential, as it can inform targeted interventions to bridge gaps in cancer care quality and improve outcomes across diverse healthcare settings. This study aims to address this knowledge gap by examining survival outcomes and treatment patterns for patients with mRCC treated at academic vs non-academic centers. By leveraging data from the National Cancer Database (NCDB), this study provided a comprehensive analysis of how facility type influences mRCC survival, treatment modalities, and demographic factors.

MATERIALS AND METHODS

Data source and study population

This study utilized data from the NCDB, a hospital-based registry representing approximately 70% of new cancer diagnoses across the United States. Patients included in this study were patients diagnosed with stage IV RCC (mRCC) between 2004 and 2020. All histologic subtypes of RCC were included; stage IV RCC was classified based on the criteria of the American Joint Commission on Cancer and the corresponding ICD-O-3 code C649. The exclusion criteria were missing survival data, unknown tumor node metastasis (TNM) staging, and T1-3 staging without the presence of metastatic disease.

Facility classification

The Comission on Cancer categorizes cancer programs based on facility type, structure, services, and annual case volumes. Academic/research facilities are characterized by their involvement in medical education across multiple specialties, including internal medicine and general surgery, and handle over 500 new cancer cases annually. National Cancer Institute-designated comprehensive cancer centers are also recognized as “academic/research” facilities due to their comprehensive involvement in research, clinical trials, medical education, and oncologic care.

The participant user file encoded four facility subtypes: (1) Academic/research programs; (2) Community cancer programs; (3) Comprehensive community cancer programs; and (4) Integrated network programs. A binary “academic/research” and “non-academic research” variable was created to serve as the overarching umbrella categories for academic and non-academic facility types. The facility type “academic/research programs” was placed in the “academic/research” variable. The remaining three facility types “community cancer programs”, “comprehensive community cancer programs”, and “integrated network programs”, were grouped into the “non-academic research” variable.

Statistical analysis

Descriptive statistics and Kaplan-Meier curves were used to measure OS between academic and non-academic facility types. Multivariate logistic regression and Cox proportional hazard testing were used to generate odds ratio (OR) (95%CI) and hazard ratio (HR) (95%CI) between academic and non-academic facility types. Demographic, socioeconomic, clinical characteristics, and treatment characteristics were sub-stratified and compared between facility types. Additional values were calculated adjusted for age, sex, and median household income. Data was analyzed using Statistical Package for the Social Sciences version 29.0 and statistical significance was set at 0.05.

RESULTS

OS

Analysis of survival outcomes revealed that patients treated at academic facilities exhibited significantly improved OS compared to those at non-academic centers (Table 1). The median OS for academic institutions was 13.7 months (95%CI: 13.3–14.1), whereas non-academic facilities had a median OS of 8.94 months (95%CI: 8.75–9.13). At the 5-year mark, OS was 17.0% at academic centers and 11.5% at non-academic centers (P < 0.05). Similarly, the 10-year OS was 9.2% at academic centers vs 5.5% at non-academic facilities (P < 0.05). Kaplan-Meier survival curves further illustrated this divergence, with a statistically significant difference in survival between the two facility types (Figure 1).

Figure 1  Kaplan-Meier survival outcomes over time of academic vs non-academic facility types.

Table 1 5-year and 10-year overall survival.

Variable (n = 79367)	Academic (n = 30780)	Non-academic (n = 48587)	P value
5-year OS (%)	17.0	11.5	< 0.05
10-year OS (%)	9.20	5.50	< 0.05

OS: Overall survival.

Demographic, socioeconomic, and clinical characteristics

The demographic and socioeconomic profiles of patients varied notably by facility type. Patients treated at academic centers were generally younger, with a mean age of 63.8 years (SD = 10.9) compared to 66.6 years (SD = 11.4) at non-academic facilities (P < 0.05). Geographic and socioeconomic differences were also apparent: Patients treated at academic centers traveled an average of 57.6 miles (SD = 148) for treatment, significantly farther than the 23.8 miles (SD = 86.7) reported for non-academic centers (P < 0.05). Academic centers served a higher percentage of privately insured patients (40.7% vs 35.9%) and a lower proportion of patients on Medicare (43.4% vs 54.5%) (P < 0.05). A larger share of patients treated at academic facilities resided in metropolitan areas with populations exceeding 1 million (51.6% vs 44.6%) (P < 0.05), whereas smaller metropolitan areas contributed more patients to non-academic centers (Table 2). There were significant differences in comorbidity scores between facility types. At academic centers, 71.4% of patients had a Charleson-Deyo comorbidity score of 0, compared to 67.1% at non-academic centers (P < 0.05).

Table 2 Demographic, socioeconomic, and clinical characteristics of patients with metastatic renal cell carcinoma, n (%).

Variable (n = 79367)	Academic facility (n = 30780)	Non-academic facility (n = 48587)	P value
Age [mean (SD)]	63.8 (10.9)	66.6 (11.4)	< 0.05
Distance travelled [mean (SD)]	57.6 (136)	23.8 (86.7)	< 0.05
Sex			0.511
Male	21428 (69.6)	32637 (67.2)
Female	9352 (30.4)	15950 (32.8)
Insurance status			< 0.05
Race			< 0.05
White	42600 (88.4)	25591 (84.1)
Black	4160 (8.6)	3473 (11.4)
Other	1441 (3.0)	1363 (4.5)
Race-missing data	739
Private insurance	11954 (40.7)	16790 (35.9)
Medicaid	2747 (9.4)	3120 (6.7)
Medicare	12744 (43.4)	25470 (54.5)
Other government	482 (1.6)	707 (1.5)
Insurance Status unknown	1418 (4.8)	653 (1.4)
Income class			< 0.05
< $38000	3666 (11.9)	5495 (11.3)
$38000-$47999	5422 (17.6)	7993 (16.5)
$48000-$62999	6032 (19.6)	10782 (22.2)
> $63000	9327 (30.3)	1317 (27.1)
Urban/Rural			< 0.05
Metro area greater than 1 million people	15886 (51.6)	21670 (44.6)
Metro area 250000–1 million	5856 (19.0)	10980 (22.6)
Metro area < 250000	9038 (29.4)	15937 (32.8)
Tumor size [mean (SD)]	481 (450)	469 (453)	< 0.05
Facility location			< 0.05
New England	2322 (4.8)	1496 (4.9)
Middle Atlantic	4540 (9.3)	6164 (20.0)
South Atlantic	10448 (21.5)	5796 (18.8)
East North Central	8959 (18.4)	4851 (15.8)
East South Central	3883 (8.0)	1758 (9.6)
West North Central	4211 (8.7)	2955 (12.3)
West South Central	5186 (10.7)	3793 (12.3)
Mountain	2189 (5.8)	859 (2.8)
Pacific	6219 (12.8)	3108 (10.1)
Regional lymph nodes positive	77.6 (39.4)	85.7 (32.2)	< 0.05
Charleson-deyo score			< 0.05
0	22077 (71.4)	32286 (66.4)
1	5638 (18.3)	10001 (20.6)
2	1784 (5.8)	3715 (7.6)
3	1281 (4.2)	2585 (5.3)

Treatment characteristics

Treatment initiation times varied significantly between academic and non-academic centers. Surgical interventions were initiated an average of 43.1 days from diagnosis at academic centers vs 37.4 days at non-academic centers (P < 0.05) (Table 3). Systemic therapy, including chemotherapy, began an average of 64.5 days after diagnosis at academic centers, compared to 55.8 days at non-academic centers (P < 0.05). Radiation therapy and immunotherapy were also delayed at academic centers, with initiation times of 51.9 days and 75.7 days, respectively, compared to 41.9 days and 67.4 days at non-academic centers (P < 0.05 for each).

Table 3 Treatment characteristics of patients with metastatic renal cell carcinoma, n (%).

Variable (n = 1068)	Academic (n = 30780)	Non-academic (n = 48587)	P value
Definitive surgical procedure, days from Dx [mean (SD)]	43.1 (63.7)	37.5 (61.5)	< 0.05
Systemic therapy initiated, days from Dx [mean (SD)]	64.5 (57.1)	55.8 (54.2)	< 0.05
Chemotherapy initiated, days from Dx [mean (SD)]	64.9 (61.3)	56.2 (54.9)	< 0.05
Radiation therapy initiated, days from Dx [mean (SD)]	56.1 (76.8)	45.8 (61.6)	< 0.05
Immunotherapy initiated, days from Dx [mean (SD)]	75.6 (65.9)	67.5 (72.9)	< 0.05
Surgical inpatient stay [mean (SD)]	5.64 (7.46)	5.40 (6.92)	0.006
Surgery of primary site			< 0.05
Subtotal nephrectomy	624 (2.0)	612 (1.3)
Complete nephrectomy	876 (2.8)	1417 (2.9)
Radical nephrectomy	11,419 (37.1)	13382 (27.5)
Surgery, other	1,015 (3.3)	1052 (2.2)
Surgery of primary site not performed	16,846 (54.7)	32124 (66.1)
Adjuvant therapy			< 0.05
Adjuvant chemoradiation	65 (0.2)	190 (0.4)
Adjuvant radiation	412 (1.3)	514 (1.1)
Adjuvant chemotherapy	1350 (4.4)	1779 (3.7)
No adjuvant therapy	28953 (94.1)	46104 (94.9)
Immunotherapy			< 0.05
Received immunotherapy	5264 (17.1)	6864 (14.1)
Did not receive immunotherapy	25132 (81.7)	41075 (84.5)
Unknown if received immunotherapy	384 (1.2)	648 (1.3)

Patients at academic centers were more likely to undergo surgical intervention, including radical nephrectomy (RN) compared to patients at non-academic centers (RN = 47.1% vs RN = 27.5%) (P < 0.05). Furthermore, academic centers exhibited a greater utilization of adjuvant therapies. Rates of adjuvant chemotherapy, radiation, and combined chemoradiation were all higher at academic centers (4.4%, 1.3%, and 0.2%) than at non-academic centers (3.7%, 1.1%, and 0.4%) (P < 0.05). There also was greater utilization of immunotherapy at academic centers (17.1% vs 14.1%) (P < 0.05).

Multivariate analysis

Multivariate regression analyses indicated that non-academic facility type was associated with poorer survival outcomes. The Cox proportional hazard model, unadjusted, yielded a HR of 1.27 (95%CI: 1.24-1.29) (P ≤ 0.05), signifying a 27% higher risk of mortality for patients treated at non-academic centers (Table 4, Figure 2). The unadjusted OR for mortality at non-academic facilities was 1.34 (95%CI: 1.30–1.40) (P ≤ 0.05). When adjusted for age, tumor size, sex, and travel distance, the OR remained elevated at 1.23 (95%CI: 1.18–1.28) (P ≤ 0.05).

Figure 2  Odds ratio/hazard ratio of multivariate regression models for association of non-academic facility type with poor survival outcomes.

Table 4 Multivariate regression models for association of non-academic facility type with poor survival outcomes.

Variables	Odds ratio/HR	95%CI	P value
Crude model (unadjusted)	1.34	1.30-1.40	< 0.05
Model adjusted for age, tumor size, sex, and distance travelled	1.23	1.18-1.28	< 0.05
Cox proportional model for HR (unadjusted)	1.27	1.24-1.29	< 0.05

Reference predictor: Academic facility type. Reference outcome: Living status. HR: Hazard ratio.

DISCUSSION

This study provides a comprehensive analysis of survival outcomes and treatment characteristics for patients with mRCC treated at academic and non-academic healthcare facilities using data from the NCDB. Our findings demonstrate that patients treated at academic centers experienced significantly improved OS compared to those at non-academic facilities. These results align with prior studies reporting superior outcomes at academic centers across various cancer types, including NSCLC, lymphoma[11], and multiple myeloma[12-16].

OS between facility types

The primary finding of this study is the substantial survival advantage observed for patients treated at academic centers, with 5-year and 10-year OS rates of 17.0% and 9.2% compared to 11.5% and 5.5% at non-academic centers, respectively. This difference persisted after adjusting for key demographic and clinical covariates such as age, tumor size, sex, and distance traveled. These findings are consistent with prior literature showing that patients treated at academic centers were more likely to receive advanced interventions, including immunotherapy and surgical procedures[17]. Enhanced survival benefits at academic centers have also been observed in early-stage NSCLC, as well as superior surgical quality with respect to 30-day and 90-day postoperative outcomes and median lymph nodes removed[11,14,15]. Similar trends have been noted for other malignancies, such DLBCL and multiple myeloma where patients at academic centers demonstrated significantly improved survival outcomes. These results were especially profound for high-risk DLBCL patients at academic centers, who demonstrated more than twice the median survival than those at non-academic centers[12,13]. Vardell et al[13] proposed that this discrepancy may be due to greater funding, easier access to clinical trials and stem cell transplants, and large integrated support care structures at academic institutions that simply do not exist or, at the very least, are difficult to acquire at non-academic centers. Other studies have suggested that academic facilities generally see greater hospital volume than non-academic facilities and are equipped with specialized multidisciplinary treatment facilities, equipment, and infrastructure, all of which lend themselves to greater breadth of specialized knowledge and ability to handle complications from treatment[14,15].

Demographic differences between facility types

Patients with mRCC treated at academic facilities were younger, carry private insurance, and come from a higher median income quartile[11], which is consistent with the results of prior studies[12-16]. One notable finding was that patients at academic centers traveled significantly greater distance for their mRCC treatment. Given the complex nature of the disease, however, mRCC may warrant a longer journey for its patients to receive tertiary care treatment (i.e. at academic centers) in comparison to other cancers, however, this distance travelled may represent a barrier to care for some patients. Socioeconomic and demographic factors likely influence patients’ decision-making in the location at which they seek medical care. These findings suggest that this decision making process may directly or indirectly be associated with prognostic outcomes in mRCC.

Treatment characteristics

Our analysis identified significant differences in the timing and intensity of treatments between facility types. Patients at academic centers experienced longer intervals between diagnosis and initiation of various treatments, including surgery, systemic therapy, and immunotherapy. These delays may reflect the additional time required for comprehensive evaluations, consultations with specialists, and coordination of care in academic settings, where multidisciplinary approaches are often emphasized[13-15]. Coordination of care of a single patient already involves multiple different types of professional healthcare providers, which at academic centers is even more complicated by the accommodation of those still in various stages of training, including medical students and resident physicians[16]. The continuum of patient needs is dynamic and contains “transition points” at which multidisciplinary care team members must meet and deliberate priorities of care and adjust interventions and team compositions accordingly, all of which involves the input from a multitude of different professionals[18]. For all modes of mRCC treatment in this study, the average treatment delay at academic centers compared to non-academic centers was no longer than ten days. While statistically significant, this delay in treatment is fairly short and is unlikely to impact prognosis or mortality for patients. A treatment delay of up to 6.3 weeks from the diagnosis of RCC was found to not significantly affect outcomes and survival[19,20]. Another study found that a minimum of a four week delay of cancer treatment is associated with increased mortality across surgical, systemic treatment, and radiotherapy indications for seven different cancers[21].

Implications for clinical practice

These results suggest that adherence to evidence-based management guidelines remains critical for a number of aggressive malignancies like mRCC, where robust prognostic and therapeutic strategies are paramount for optimizing outcomes[22,23]. Potential improvement of the prognosis in mRCC largely depends on robust detection and utilization of targeted treatment modalities. Within the detection the TNM classification system remains the gold standard, which simultaneously accounts for stage, grade, tumor subtype, clinical features, and performance[5,24,25]. Prognostic scoring of mRCC uses a modified Glasgow prognostic score to stratify the risk of RCC[26]. Molecular markers such as WDR72 have garnered more attention in recent years, with an increasing number of novel markers associated with RCC being found[27]. However, in addition to the complicated challenge of obtaining and maintaining samples, currently they are only used occasionally as an adjunct to improve the accuracy of existing prognostic models due to their poor external validity and impractical implementation because of tumor heterogeneity[28,29]. Treatment for mRCC has rapidly evolved, beginning when the anti-angiogensis agent Vascular endothelial growth factor receptor-tyrosine kinase inhibitor (TKI) sunitinib demonstrated superiority compared to interferon-alpha (IFN-α). Currently, clear cell mRCC with intermediate risk is treated by sunitinib/pazopanib/bevacizumab and IFN-a (1^st line) or sorafenib/axitinib (2^nd line)[1,5,30]. Patients with clear cell mRCC with intermediate risk who failed TKI receive sorafenib/everolimus/temsirolimus/axitinib. Poor-risk clear cell mRCC and non-clear cell mRCC patients receive temsirolimus. Radiotherapy is an adjunct modality used but to limited effect due to the apparent radio-resistant nature of RCC neoplasms[1]. As alluded to before, a preeminent aspect of current challenges in pharmacological development is the significant discontinuation rates of patients in previous drug trials due to debilitating adverse effects[31]. A large portion of the modus operandi in contemporary treatment with adjuvant therapy was and still is preemptive recurrence risk stratification such that as few low-risk patients may need to undergo adjuvant therapy, and thus be exposed to their potential side effects, as possible. Considering mRCC is a clinically challenging malignancy for providers to treat and the limited detection and treatment options available, it is paramount that physicians adhere to existing guidelines to optimize survival for as long as possible.

Bridging the gap between facility types

Given the relative rarity of academic centers, it is most likely a practical infeasibility to recommend academic center healthcare to all patients. Perhaps a solution to bridge the gap between academic and non-academic medical centers could be involvement in research pertaining to RCC. Physicians at academic centers could begin to incorporate community physicians in the research that is conducted at large academic centers, to encourage active involvement in the latest technologies and information[32]. This method also bears the potential to expand clinical trials at academic centers to broaden the number and diversity of the participants. In addition, academic centers often host didactic sessions, providing attendees insight into current literature[32]. Incorporating community-based physicians into this continuing educational model could narrow the gap in outcomes, by sharing resources that the community-based physicians would otherwise not have access to. Strengthening connections between academic and non-academic providers would provide a net positive for the patients affected by mRCC and could improve outcomes for those affected by this disease.

Limitations

While this study provides valuable insights, several limitations must be acknowledged. First, as a retrospective analysis of NCDB data, it is subject to inherent biases, including selection bias and uncontrolled confounding. Patients at academic centers may differ systematically from those at non-academic facilities in ways not fully captured by the available covariates. Second, the use of registry data introduces potential inaccuracies, as coding errors and incomplete documentation in the NCDB could influence the findings. Finally, unmeasured variables such as molecular biomarkers, detailed treatment adherence, and socioeconomic support systems may further confound the observed relationships in ways beyond the currently available information encoded within the NCDB database.

CONCLUSION

Patients treated at academic institutions experienced superior survival outcomes compared to their counterparts at non-academic facilities. Moreover, they were younger ages, carried private insurance, and resided in larger metropolitan areas. Disparities can potentially be reduced by integrating community healthcare services in research and education alongside academic centers.