Data on partial nephrectomy and ablation when combined as nephron-sparing surgery. See the eFigure in the Supplement for partial nephrectomy and ablation listed separately.
A, Overall survival outcomes stratified by initial treatment strategy. B, Cancer-specific survival outcomes stratified by initial treatment strategy.
eFigure. Treatment Trends of Small Kidney Cancers (2001-2009)
eTable. Competing Risk Model
Huang WC, Atoria CL, Bjurlin M, Pinheiro LC, Russo P, Lowrance WT, Elkin EB. Management of Small Kidney Cancers in the New MillenniumContemporary Trends and Outcomes in a Population-Based Cohort. JAMA Surg. 2015;150(7):664-672. doi:10.1001/jamasurg.2015.0294
With the significant downward size and stage migration of localized kidney cancers, the management options for small kidney cancers have expanded and evolved.
To describe trends and outcomes in the management of small kidney cancers in the first decade of the new millennium.
Design, Setting, and Participants
Surveillance, Epidemiology, and End Results (SEER) cancer registry data linked to Medicare claims were used to identify patients 66 years or older with a pathologically confirmed small kidney cancer (<4 cm) diagnosed between January 1, 2001, and December 31, 2009; analysis was performed between February 1, 2014, and December 31, 2014. Multivariable logistic regression was used to assess the likelihood of nonsurgical management vs surgical intervention. Cox proportional hazards regression was used to assess the relationships between treatment approach and overall and cancer-specific survival. The effect of treatment approach on cancer-specific survival was analyzed in a competing risks framework.
Main Outcomes and Measures
The likelihood of receiving no surgery vs surgical intervention as a function of demographic and disease characteristics, as well as the relationships between treatment approach and overall and cancer-specific survival.
Of 6664 patients, 5994 individuals (90.0%) had surgical treatment; the care of 670 patients (10.0%) was managed nonsurgically. Use of radical nephrectomy decreased over time (from 69.0% to 42.5%), and the use of nephron-sparing surgery (partial nephrectomy and ablation) increased (from 21.5% to 49.0%); the proportion of patients who did not undergo surgery remained stable (9.5% and 8.5%). During a median follow-up of 63 months (interquartile range, 43-89 months) (follow-up for vital status through December 31, 2011), 2119 patients (31.8%) patients died, including 293 individuals (4.4%) of kidney cancer. Although overall survival was better in patients who received surgical treatment, only nephron-sparing surgery was associated with a benefit in cancer-specific survival (adjusted hazard ratio, 0.47; 95% CI, 0.31-0.69; P < .001).
Conclusions and Relevance
Surgery continues to be the most common treatment for patients with small kidney cancers. The use of nephron-sparing surgery exceeds radical nephrectomy in patients who receive surgery. Although our findings suggest that nonsurgical management is acceptable for certain patients, use of this approach remains low.
Kidney cancer is the third most common genitourinary cancer in the United States and ranks among the top 10 malignant neoplasms in both men and women.1 Although there are no recommended screening protocols for kidney cancer, the liberal use of imaging for unrelated conditions has resulted in the early detection of renal masses, most of which are cancers.2 More than two-thirds of all kidney cancers are less than 4 cm (clinical stage T1a, according to the 2009 American Joint Committee on Cancer classification system, 7th Edition3) at the time of diagnosis.4
Surgical resection has been the standard treatment for localized kidney cancer for more than 50 years. The past decade has seen a shift in the management of these tumors, with an emphasis on avoiding removal of unaffected kidney tissue.5,6 Expert guidelines7,8 recommend the use of partial nephrectomy for kidney tumors less than 4 cm when technically feasible. Several large cohort studies9- 11 suggest that partial nephrectomy is equivalent to radical nephrectomy in terms of oncologic outcomes, as well as reducing the risk of chronic kidney disease following surgery and potentially the sequelae associated with chronic kidney disease, such as increased cardiovascular morbidity and mortality. The findings of these studies have coincided with a growing awareness of the role of comorbid conditions and competing causes of death among patients, particularly elderly individuals, with incidentally diagnosed small kidney tumors.4,12 Such guidelines have also described ablation and surveillance as options for the management of small kidney cancers in select patients because surgical intervention for asymptomatic small kidney cancers may be unwarranted if comorbid illness reduces the likely benefit of treatment in patients with a reduced life expectancy.
Only one randomized clinical trial13 has directly compared the outcomes of the various treatment strategies for small kidney cancers. Observational studies6,9,14,15 have generally been small, single-institution series with limited follow-up. Our objectives were to examine trends in contemporary management of small kidney cancers and to assess the association between treatment strategy and overall survival and cancer-specific survival in a population-based cohort of older adults with kidney cancers less than 4 cm.
The primary data source was the Surveillance, Epidemiology, and End Results (SEER) cancer registry14 with data linked to Medicare claims. A consortium of population-based cancer registries, SEER is sponsored by the National Cancer Institute and currently includes 18 registries covering approximately 28% of the US population. For all incident cancers in their coverage areas, the SEER registries collect information regarding the site and extent of disease, the first course of cancer-directed therapy, and sociodemographic characteristics, with active follow-up for the date and cause of death. For adults 65 years or older who receive a diagnosis of cancer and reside in a SEER area, Medicare claims have been linked to SEER records. Compared with the elderly population in the United States, the SEER-Medicare population has a similar age and sex distribution, is somewhat more likely to live in urban and affluent areas, and has a smaller proportion of nonwhite individuals.14 This study was deemed exempt research by the institutional review board at Memorial Sloan-Kettering Cancer Center and conducted in adherence to a data use agreement from the National Cancer Institute. Analysis of the data was performed between February 1, 2014, and December 31, 2014.
In the SEER-Medicare database, we identified patients with a diagnosis of primary renal-cortical tumor (International Classification of Diseases for Oncology, Third Edition topography codes C64 and C64.9) between January 1, 2001, and December 31, 2009. The cohort was restricted to patients 66 years or older with a pathologically confirmed primary cancer less than 4 cm. We excluded patients with locally advanced or metastatic disease, patients whose diagnosis was made at the time of death, beneficiaries enrolled in a Medicare managed care plan, and those with discontinuous or incomplete Medicare coverage. Patients who had a claim for chemotherapy or radiotherapy within 6 months of diagnosis were also excluded.
Definitive renal surgery within 6 months following diagnosis was identified in Medicare claims based on billing codes for radical nephrectomy (Current Procedural Terminology [CPT] codes 50220, 50225, 50230, 50545, and 50546 and International Classification of Diseases, Ninth Revision, Clinical Modification [ICD-9-CM ] codes 555.1- 555.4), partial nephrectomy (CPT codes 50240, 50280, 50290, and 50543 and ICD-9-CM codes 553.1, 553.9, and 554), or ablation (CPT codes 50254, 50250, 50542, and 50592 and ICD-9-CM codes 553.2-553.4). Owing to the small number of cancers managed with ablation, partial nephrectomy and ablation were combined and classified as nephron-sparing surgery for analysis. Nonsurgical management was defined as the absence of any claim for one of these procedures in the first 6 months following diagnosis.
Primary end points of the study were death from any cause and death due to renal cancer. The date of death was identified from Medicare records, with follow-up for vital status through December 31, 2011. Attribution of deaths to kidney cancer was based on the underlying cause of death reported in the state death certificate as recorded in the SEER registry.
We examined several characteristics that could potentially confound the association between treatment and the primary end points. These characteristics included age at diagnosis, race, marital status, urban vs rural residence, median income in the census tract of residence, and history of a nonrenal cancer. Disease characteristics included tumor size and grade. We used the Romano modification of the Charlson Comorbidity Index, based on inpatient and outpatient claims in the year before cancer diagnosis, as a summary measure of comorbidity burden.16,17 We also created indicators for the following specific comorbid conditions based on inpatient claims only in the year before diagnosis: renal insufficiency, hypertension, and diabetes mellitus.
Unadjusted associations between treatment approach and patient characteristics were examined using χ2 statistics (January 1, 2001, to December 31, 2009). We estimated Kaplan-Meier functions for overall survival and cancer-specific survival stratified by treatment group. Unadjusted survival curves were compared using the log-rank test.
Multivariable logistic regression was used to estimate the effect of patient and disease characteristics and comorbidity on the likelihood of receiving nonsurgical management vs definitive renal surgery. We used multivariable Cox proportional hazards regression to estimate the effect of treatment approach on overall and cancer-specific survival, adjusted for individual patient and disease characteristics and comorbidity. The adjusted hazard ratio (HR), 95% CI, and 2-sided P values were estimated for each predictor. In all analyses, the time origin was the date of cancer diagnosis, and observations were censored at the end of the follow-up period. The effect of treatment approach on cancer-specific survival was also analyzed in a competing risks framework in which deaths from other causes were considered a competing risk, and observations were censored at the end of follow-up. Analyses were performed using SAS, version 9.2 (SAS Institute Inc).
We identified 6664 patients with a kidney cancer of less than 4 cm diagnosed between 2001 and 2009. Of these patients, 5994 individuals (90.0%) had definitive surgery, and the care of 670 patients (10.0%) was managed nonsurgically (Table 1). The use of partial nephrectomy and ablation increased over time (from 21.5% to 49.0%); the proportion of patients who received radical nephrectomy decreased (from 69.0% to 42.5%), and the proportion whose care was managed nonsurgically remained stable (9.5% and 8.5%) (Figure 1 and the eFigure in the Supplement). In 2009, nephron-sparing surgery (partial nephrectomy and ablation) surpassed radical nephrectomy and nonsurgical management as the most common treatment for small kidney cancers (Figure 1 and the eFigure in the Supplement).
During the entire study period, 3709 of the patients (61.9%) who underwent surgery had radical nephrectomy, and 2285 individuals (38.1%) had a nephron-sparing procedure. With controls for demographic and disease characteristics as well as comorbid conditions, patients whose care was managed nonsurgically were more likely to be older, male, and nonwhite and to reside in the Western United States (Table 2). Among the specific comorbid conditions evaluated, hypertension was associated with a greater likelihood of being treated surgically, and a prior nonkidney cancer diagnosis was associated with a greater likelihood of nonsurgical management. With a median follow-up of 57 months (interquartile range, 43-89 months), 136 patients (20.3%) who initially received nonsurgical treatment underwent surgery more than 6 months after diagnosis.
A total of 2119 patients (31.8%) died from any cause during the study period, including 1834 who had surgery and 285 of those who did not have surgery. Median follow-up among survivors was 63 months (interquartile range, 43-89 months). In the nonsurgical group, 3-year and 5-year overall survival probabilities were 73% and 61%, respectively (Figure 2A). Three-year and 5-year overall survival probabilities were 83% and 72% in patients who had radical nephrectomy and 90% and 81%, respectively, in those who had nephron-sparing surgery. Compared with nonsurgical management, radical nephrectomy and nephron-sparing surgery were associated with a lower risk of death from any cause, controlling for patient and disease characteristics (adjusted HR for radical nephrectomy, 0.76; 95% CI, 0.66-0.86; adjusted HR for nephron-sparing surgery, 0.55; 95% CI, 0.47-0.64; P < .001) (Table 3).
Of the 2119 deaths observed during the study period, 293 deaths (13.8%) were attributed to kidney cancer, including 249 (4.2%) of the patients who had surgery and 44 (6.6%) of those who did not have surgery. Among patients whose care was managed nonsurgically, 3-year and 5-year cancer-specific survival probabilities were 95% and 92%, respectively (Figure 2B). Three-year and 5-year survival probabilities were 96% and 95% in patients who had radical nephrectomy and 98% and 97%, respectively, in those who had nephron-sparing surgery (Figure 2B). Nephron-sparing surgery was associated with a lower risk of death from kidney cancer compared with nonsurgical management, controlling for patient and disease characteristics (adjusted HR, 0.47; 95% CI, 0.31-0.69) (Table 3). Compared with nonsurgical management, radical nephrectomy did not have a significantly different effect on cancer-specific survival. The results of the competing-risk analysis were similar to those of the standard proportional hazards regression model (eTable in the Supplement).
The presentation and management of kidney cancer has evolved substantially since the turn of the century. With increased use of diagnostic imaging, the incidence of kidney cancer has risen, largely owing to small localized masses that are diagnosed incidentally on imaging for unrelated conditions.
This downward size and stage migration has engendered a change in the treatment paradigm for localized kidney cancers. With an improved understanding of the biological behavior of these small cancers as well as a growing appreciation of the effect of renal surgery on oncologic and nononcologic outcomes, the treatment options for small kidney cancers have evolved. Radical nephrectomy, which was the standard treatment for all renal tumors regardless of size, has been replaced by partial nephrectomy at tertiary referral centers for the treatment of clinical T1a renal masses whenever technically feasible owing to the functional benefits with equivalent oncologic outcomes. This paradigm shift has been slower to occur in community practices and, for unclear reasons, in some patient subgroups, particularly women and the elderly population.18- 20
Nonetheless, our analysis demonstrates that radical nephrectomy is no longer the treatment of choice for small kidney cancers. Expert guidelines for the management of small renal masses, such as those published by the American Urological Association in 2009,8 appear to have bolstered the paradigm shift away from radical nephrectomy toward nephron-sparing options for such tumors.21 The same practice guidelines also offer ablation and surveillance as alternative treatment options for small kidney masses.8 In the population-based cohort analyzed in the present study, use of ablation increased slowly, but the rate of nonsurgical management for small kidney cancers remained stable (eFigure in the Supplement).
Nonsurgical management has become an acceptable treatment approach for a variety of malignant neoplasms, including small kidney cancers. In low-risk prostate cancer, for example, primary surgical treatment has been associated with treatment-related morbidity but not a clear survival advantage, resulting in increasing adoption of active surveillance for low-risk disease in elderly patients and those with comorbid illness.22 A similar argument can be made for the nonsurgical management of small kidney cancers diagnosed in elderly individuals or patients with comorbid illness. This rationale stems from a better understanding of the indolent biological behavior of small kidney tumors as well as a better appreciation of competing mortality risks, particularly from cardiovascular disease, which may account for up to 30% of all deaths in elderly patients.12,15,23 Our findings underscore the importance of competing causes of mortality in this population: 38.1% of patients died during the study period, but only 4.4% of patients died from kidney cancer.
The survival benefit of surgery for small, localized kidney cancers remains unclear. In our study, surgery was associated with an overall survival advantage compared with nonsurgical management. However, because most of the deaths were unrelated to kidney cancer, it is likely that observed differences in overall survival reflect patient selection rather than a direct benefit of surgery. This difference is highlighted by the greater risk of mortality observed soon after diagnosis in the cohort whose treatment was managed nonsurgically. In the context of an observational study, it is impossible to draw a causal inference, particularly when there are many factors that could confound the association between treatment approach and all-cause mortality.
Without a randomized trial, large observational cohorts may be the best available source of information on the comparative effectiveness of the different strategies for managing small kidney cancers overall as well as in specific patient subgroups. Although cancer-specific deaths recorded in the SEER registry are based on state death certificates that may be subject to misattribution, prior studies24,25 support the use of information from death certificates to identify cancer-specific mortality.
In our population-based cohort of older adults, 4.2% of all patients who underwent surgery died of kidney cancer compared with 6.6% of those whose treatment was managed nonsurgically. Controlling for patient and disease characteristics, we only observed a cancer-specific survival advantage associated with nephron-sparing surgery, but not radical nephrectomy, compared with nonsurgical management. Because of the small number of cases treated with ablation, we were unable to independently analyze the outcomes of ablation and partial nephrectomy vs nonsurgical management. The paradoxical oncologic advantage of partial nephrectomy has been found in other observational studies26 but not in the only randomized trial13 comparing radical nephrectomy with partial nephrectomy, supporting the notion that the findings are likely related to patient selection.
Although our findings support nonsurgical management as an acceptable treatment option for small kidney tumors in elderly patients or those with limited life expectancy, it appears that nonsurgical management of small kidney cancers remains uncommon and stable over time. It is possible, however, that with heightened awareness of outcomes from studies such as ours, the use of nonsurgical management may increase.
Several limitations of the present study should be noted. Despite controlling for numerous potential confounders in the absence of randomized treatment assignment, inferences about the association between treatment strategy and outcomes must be made cautiously. Selection bias owing to the omission of important characteristics, such as functional status, cannot be ruled out. To some degree, our findings might also reflect appropriate selection of patients for each treatment strategy based on observed characteristics and others that are not available in the SEER-Medicare data set. Our analysis was limited to adults 66 years or older with pathologically confirmed cancers, and our results may not be generalizable to younger patients or those with small kidney tumors identified by imaging alone without a tissue diagnosis. Because up to 20% of small renal masses are benign, which may alter the observed oncologic outcomes of nonoperatively managed small renal masses, we included only pathologically confirmed kidney cancers in our analysis.27
We also cannot extrapolate beyond the time horizon of our available follow-up. Some small proportion of patients, should they survive long enough, could experience metastatic disease and death from kidney cancer. Until such potentially lethal tumors can be accurately identified at the time of diagnosis, surgery remains the standard treatment option for patients with a normal life expectancy. In the future, advances in imaging, coupled with percutaneous biopsy and tumor-specific molecular characterization, may improve our ability to appropriately select patients for the various treatment options.
The presentation and management of kidney cancer have evolved considerably in the past decade. Surgery remains the treatment of choice for most small kidney cancers; however, radical nephrectomy is no longer the most commonly used approach. In this population-based cohort, surgery was associated with a cancer-specific and overall survival benefit compared with nonsurgical management, although these associations likely reflect some degree of patient selection, particularly in the case of overall survival. Nonsurgical management appears to be a reasonable treatment strategy for elderly patients or those with limited life expectancy, yet utilization rates remain low and stable over time.
Accepted for Publication: December 22, 2014.
Corresponding Author: William C. Huang, MD, Department of Urology, New York University Langone Medical Center, 150 E 32nd St, Second Floor, New York, NY 10016 (firstname.lastname@example.org).
Published Online: May 27, 2015. doi:10.1001/jamasurg.2015.0294.
Author Contributions: Drs Huang and Elkin had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.
Study concept and design: Huang, Russo.
Acquisition, analysis, or interpretation of data: All authors.
Drafting of the manuscript: Huang, Pinheiro, Russo.
Critical revision of the manuscript for important intellectual content: All authors.
Statistical analysis: Atoria, Bjurlin, Pinheiro, Elkin.
Administrative, technical, or material support: Huang, Pinheiro.
Study supervision: Huang, Russo, Lowrance.
Conflict of Interest Disclosures: None reported.
Disclaimer: The linked Surveillance, Epidemiology, and End Results (SEER)–Medicare database was used for this study. The interpretation and reporting of these data are the sole responsibility of the authors.
Additional Contributions: We gratefully acknowledge the Applied Research Program (National Cancer Institute), Centers for Medicare and Medicaid Services, Information Management Services Inc, and the SEER program tumor registries for their creation and maintenance of the linked SEER-Medicare dataset. Mithat Gonen, PhD (Memorial Sloan-Kettering Cancer Center), provided advice regarding statistical methods; there was no financial compensation.