SEARCH

SEARCH BY CITATION

Keywords:

  • renal cell carcinoma;
  • partial nephrectomy;
  • ablation techniques;
  • comparative effectiveness research

Abstract

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. CONFLICT OF INTEREST
  8. REFERENCES

Study Type – Prognosis (cohort)

Level of Evidence 2a

What's known on the subject? and What does the study add?

Case series of patients undergoing various forms of ablation show that it is technically feasible and possible for ablation to achieve short- and intermediate-term cancer-specific survival rates similar to those of controls undergoing partial nephrectomy.

This is the first well-powered study with a controlled design to compare effectiveness between partial nephrectomy and ablation.

OBJECTIVE

  • • 
    To determine, in a population-based cohort, if disease-specific survival (DSS) was equivalent in patients undergoing ablation vs nephron-sparing surgery (NSS) for clinical stage T1a renal cell carcinoma (RCC).

PATIENTS AND METHODS

  • • 
    A retrospective cohort study was performed using patients from the Surveillance, Epidemiology and End Results cancer registry with RCC < 4 cm and no evidence of distant metastases, who underwent ablation or NSS.
  • • 
    Kaplan–Meier and Cox regression analyses were performed to determine if treatment type was independently associated with DSS.

RESULTS

  • • 
    Between 1998 and 2007, a total of 8818 incident cases of RCC were treated with either NSS (7704) or ablation (1114).
  • • 
    The median (interquartile range) follow-up was 2.8 (1.2–4.7) years in the NSS group and 1.6 (0.7–2.9) years in the ablation group, although 10% of each cohort were followed up beyond 5 years.
  • • 
    After multivariable adjustment, ablation was associated with a twofold greater risk of kidney cancer death than NSS (hazard ratio 1.9, 95% confidence interval 1.1–3.3, P= 0.02).
  • • 
    Age, gender, marital status and tumour size were also significantly associated with outcome.
  • • 
    The predicted probability of DSS at 5 years was 98.3% with NSS and 96.6% with ablation.

CONCLUSION

  • • 
    After controlling for age, gender, marital status and tumour size, the typical patient presenting with clinical stage T1a RCC, who undergoes ablation rather than NSS, has a twofold increase in the risk of kidney cancer death; however, at 5 years the absolute difference is small, and may only be realized by patients with long life expectancies.

Abbreviations
NSS

nephron-sparing surgery

SEER

Surveillance Epidemiology and End Results

NOS

not otherwise specified

DSS

disease-specific survival

IQR

interquartile range

HR

hazard ratio

INTRODUCTION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. CONFLICT OF INTEREST
  8. REFERENCES

Over the last several decades, the incidence of RCC has increased steadily [1]. This trend is probably attributable to biological factors, such as the obesity epidemic and tobacco exposure [2] as well as increased detection with more frequent use of abdominal imaging [3]. Radical nephrectomy has traditionally been the standard of care in the management of RCC; however, growing evidence regarding the long-term detrimental effects of renal insufficiency and increased mortality [4,5], combined with demonstrated oncological efficacy of nephron-sparing surgery (NSS) [6], has resulted in greater interest in renal preservation and alternative treatment methods.

The spectrum of minimally invasive methods for the management of renal tumours includes extirpative procedures, such as laparoscopic and robot-assisted NSS, as well as thermal ablative techniques, including radiofrequency and cryotherapy. In single-centre series, ablation appears to be effective in the short and intermediate term [7]. In addition, two series have reported cryoablation outcomes beyond 5 years [8,9]. A review for the AUA guidelines [10] and a recently published large systematic review [11] both reported a significant increase in local progression with ablation. However, there remains no adequately controlled study comparing survival in patients undergoing ablation or NSS. It is unclear, therefore, whether increased local progression translates into lower cancer-specific or overall survival over time.

We sought to address this issue using a population-based registry, comparing the long-term oncological outcomes in patients undergoing either NSS or ablation for clinical T1a renal masses.

PATIENTS AND METHODS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. CONFLICT OF INTEREST
  8. REFERENCES

STUDY POPULATION

We performed an analysis of the Surveillance Epidemiology and End Results (SEER) database, a US population-based cancer registry managed by the National Cancer Institute. An exemption from institutional review board approval was obtained from the University of California San Francisco Committee on Human Research for this study.

The study cohort included subjects with RCC and tumour size ≤ 4 cm. From patients with tumours originating in the kidney based on ‘Site rec B’, RCC histologies were further selected based on the following International Classification of Diseases-O-3 codes: clear-cell (8310), papillary (8050,8260,8342), chromophobe (8270,8290,8317), collecting duct (8319), medullary (8510), granular (8320), sarcomatoid (8318), and RCC not otherwise specified (NOS) (8312). Paediatric patients (<18 years of age), and those with bilateral or metachronous disease were excluded.

VARIABLES

Predictor variables assessed in the present study included: treatment method (NSS vs renal ablation), age at diagnosis, gender, race/ethnicity, marital status, tumour size, TNM stage, county attribute and several economic indicators.

‘Rx Surgery Primary Site 1998+’ was used to identify the type of surgery as follows: no surgery (0), ablation NOS (10–25), except cryoablation (13, 23) or radiofrequency ablation (15), partial nephrectomy (26–39), radical nephrectomy (40–89), nephrectomy NOS (90) and unknown (99).

Race/ethnicity was categorized as White, Black, Hispanic, or other (Asian, American Indian/Alaska Native, and Native Hawaiian/Pacific Islanders). Marital status was classified as married or not (single, divorced, separated and widowed) [12]. The county attribute was defined using the SEER Rural-Urban Continuum code as a rural (codes 4–9 and 88) or an urban area (codes 0–3). Race-specific data on the proportion of people (per county) with less than a high school education and living below the poverty line were recorded.

The primary outcome was disease-specific survival (DSS). ‘Code to site rec KM’ was used to identify subjects as living, dead from disease, or dead from other causes.

STATISTICAL ANALYSIS

The Student's t-test and chi-squared test were used to assess differences in baseline characteristics between subjects who underwent NSS and ablation. DSS curves were plotted using the Kaplan–Meier method stratified by treatment type. Therefore, survival time was right censored in the NSS group at 10 years, which was the longest ablation follow-up. Proportional hazards were checked graphically and analytically.

Univariable and multivariable Cox regression analyses were performed to determine factors associated with DSS. All factors were included in the multivariable model to assess for potential confounding of each factor. Survival curves were plotted by treatment type amongst those with RCC confirmed by known grade or histology, adjusting for the mean and most common values of significant predictors from the multivariable model. Tests of interaction were performed on year of diagnosis, tumour size and known grade/histology with treatment type.

Although grade and histology were unknown at the time that treatment type was determined (and therefore cannot be causally associated and thus a confounder), for face validity they were added into a separate model, as both are strongly associated with DSS. Lastly, a competing risks regression analysis based on the method of Fine and Gray was performed to determine if differences in non-kidney cancer mortality between treatment groups could confound the findings for DSS.

The α-value was set at 0.05 and 95% CIs were determined. Data were analysed using stata® version 11.0 (StataCorp, College Station, TX, USA).

RESULTS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. CONFLICT OF INTEREST
  8. REFERENCES

Between 1998 and 2007, a total of 8818 incident cases of RCC were treated with either NSS (7704) or ablation (1114). Among those undergoing ablation, 662 (59%) had cryoablation, 239 (21%) had ablation NOS and 213 (19%) had radiofrequency ablation. Compared with patients who underwent NSS, those who underwent ablation were older and were less likely to be married (Table 1). Although the difference in tumour size was statistically significant, it did not appear clinically significant.

Table 1. Comparison of baseline demographic, clinical and pathological characteristics according to treatment group
CharacteristicNSS, n= 7704Ablation, n= 1114 P
  • *

    Missing data: histology: NSS (2325, 30%), ablation (532, 48%); grade: NSS (1601, 21%), ablation (604, 54%).

Diagnosis year: median (IQR)2004 (2002–2006)2006 (2004–2007)<0.001
Mean (sd) age, years59 (13)68 (12)<0.001
Male, n (%)4770 (62)687 (62)0.87
Ethnicity, n (%)  0.77
 White5816 (76)858 (77) 
 Black696 (9)98 (9) 
 Hispanic730 (10)103 (9) 
 Asian386 (5)49 (4) 
Married, n (%)577 (67)683 (61)<0.001
Urban area, n (%)6904 (90)995 (90)0.63
Median (IQR) % poverty level9 (6–13)9 (6–13)0.62
Mean (sd) tumour size, cm2.4 (0.8)2.6 (0.8)<0.001
Histology*, n (%)  0.38
 Clear-cell3794 (71)421 (72) 
 Papillary1141 (21)124 (21) 
 Chromophobe419 (8)36 (6) 
 Sarcomatoid25 (0.5)1 (0.2) 
Grade*, n (%)  <0.001
 11540 (25)171 (33) 
 23568 (58)292 (57) 
 3908 (15)46 (9) 
 481 (1)1 (0.2) 

The median (interquartile range [IQR]) follow-up was 2.8 (1.2–4.7) years in the NSS group and 1.6 (0.7–2.9) years in the ablation group, although 10% of each cohort had follow-up beyond 5 years. A total of 716 (8.1%) patients died during follow-up, of whom 110 (15%) died from RCC: 91 (1.2%) in the NSS group and 19 (1.7%) in the ablation group. Based on Kaplan–Meier analysis (Fig. 1) and the log-rank test, a significant difference in actuarial survival was observed. At 5 years, DSS survival was high after both NSS (98.2%) and ablation (94.4%). This difference appears to increase over time, although the number at risk becomes small with very long follow-up.

image

Figure 1. Kaplan–Meier estimates of DSS stratified by procedure type. The 5- and 10-year survival rates were 98% and 94% in those undergoing NSS, and 96% and 82% in those undergoing ablation.

Download figure to PowerPoint

Univariable Cox regression analysis was performed to examine factors associated with DSS in the study cohort (Table 2). A significant difference in survival was observed in those who underwent ablation compared with NSS (hazard ratio [HR] 2.6, 95% CI 1.6–4.2, P < 0.001). After multivariable adjustment (Table 2), there was little change in the risk of kidney cancer death associated with most factors. However, a clinically significant decrease in the association of treatment type with survival was observed (HR 1.9, 95% CI 1.1–3.3, P= 0.02); this difference occurred primarily after adjustment for age.

Table 2. Cox regression analyses of factors associated with DSS
CovariateUnivariable analysisMultivariable analysis
HR(95% CI) P HR(95% CI) P
Diagnosis year, per year increase0.9 (0.8–1.0)0.140.9 (0.8–1.0)0.09
Age per decade increase1.6 (1.3–1.9)<0.0011.5 (1.3–1.8)<0.001
Male vs female1.6 (1.1–2.5)0.031.9 (1.2–2.9)0.006
Ethnicity vs white 0.84 0.75
 Black1.1 (0.6–2.1)0.750.9 (0.4–2.0)0.78
 Hispanic1.3 (0.7–2.4)0.381.3 (0.6–2.9)0.48
 Asian1.2 (0.5–2.6)0.741.2 (0.5–2.9)0.63
Married vs other0.6 (0.4-–0.9)0.0070.6 (0.4–0.8)0.005
Urban vs rural area0.8 (0.5–1.5)0.570.9 (0.5–1.7)0.76
Poverty level, per 10% increase1.2 (1.0–1.6)0.081.2 (0.8–1.8)0.35
Tumour size, per 1-cm increase1.5 (1.2–1.9)<0.0011.4 (1.1–1.8)0.006
Ablation vs NSS2.6 (1.6–4.2)<0.0011.9 (1.1–3.3)0.02

Adjusted survival curves were plotted stratified by treatment type. The 5-year predicted probability of survival in a 59.9-year-old, married man with a 2.2-cm RCC undergoing NSS was 98.3%, while in the same person undergoing ablation it was 96.6%.

A difference was observed in the risk of death based upon type of ablation performed. Ablation NOS was associated with a significant increase in mortality compared with NSS (HR 2.5, 95% CI 1.2–5.2, P= 0.01). The best estimate of the risk of death in those treated with cryoablation (HR 1.5, 95% CI 0.7–3.2, P= 0.31) and radiofrequency ablation (HR 2.2, 95% CI 0.7–7.1, P= 0.21) were in a similar direction but did not reach significance.

There was suggestion of an interaction between treatment type and year of diagnosis (P= 0.10). For those treated in 1998, a much stronger difference in survival was observed in those undergoing surgery vs ablation (HR 3.4, 95% CI 1.5–7.5), while for those undergoing treatment in 2007, there was no difference between treatment arms (P= 0.56). This trended with an increase in the proportion of ablated patients undergoing cryoablation and a decrease in those coded as ablation NOS.

By contrast, there was no evidence of an interaction between treatment type and tumour size (P= 0.93). Thus, over the range of tumours studied (<4 cm), the improved survival in those undergoing NSS vs ablation was similar.

Significantly more patients who underwent ablation than patients who underwent NSS did not have pathological confirmation of RCC available (34% vs. 7%, P < 0.001). Although the test of interaction was not significant (P= 0.38), a clinically relevant difference in survival was seen when comparing those in whom a specific grade and histology were known (HR 2.2, 95% CI 1.2–4.0, P < 0.001) with those in whom the diagnosis was RCC NOS (HR 1.3, 95% CI 0.4–3.7, P= 0.68).

As grade appeared systematically lower in those undergoing ablation, adjustment for grade led to worsened relative survival with ablation (HR 3.2, 95% CI 1.6–6.6, P= 0.001). By contrast, because histologies were similar between treatment types, adjustment did not change relative survival (HR 1.9, 95% CI 0.9–4.0, P= 0.09).

Restricting the analysis to only those patients who had >5 years of follow-up increased the survival advantage seen with NSS over ablation (HR 3.0, 95% CI 0.7–13.2, P= 0.15), although given the smaller sample size this did not reach significance.

A competing risk model was created to determine if non-kidney-cancer-related deaths could affect the results. No change in the association between ablation and increased kidney cancer-specific mortality was observed (HR 1.9, 95% CI 1.1–3.3, P= 0.02).

DISCUSSION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. CONFLICT OF INTEREST
  8. REFERENCES

The contemporary management of small renal masses is a challenging problem given the multiple options available including surveillance, surgical extirpation and thermal ablation [13]. Recent data have helped clarify several issues, such as the comparable oncological efficacy of radical nephrectomy and NSS [14], as well as the importance of preserving as much renal function as possible. Although active surveillance remains an attractive choice for older patients and those with significant comorbid diseases [15], there is a small but real risk of disease progression and many patients receive treatment during follow-up despite initial observation [16]. Thus, much of the current debate centres around the relative oncological benefits of NSS vs ablation, and the purported decreased morbidity of ablation vs NSS.

While the kind of question posed in the present study is best addressed with a randomized controlled trial in order to limit bias and confounding, ablation has been taken up rapidly without such investigation. In the absence of a randomized trial, observational studies play an important role in providing a best estimate as to the comparative effectiveness of two treatments. Our findings indicate that in a population-based cohort with cT1a renal masses, those who undergo ablation suffer at least a two-fold worse DSS than those who undergo NSS. DSS in all treated patients was high and in the typical patient presenting with a cT1a renal mass – a 60-year-old married man with a 2.2-cm mass – 5-year survival was improved by 1.7% with NSS vs ablation. Therefore, in patients who are surgical candidates, NSS should remain the standard of care.

Adjustment for multiple potential confounders, in particular age, did not eliminate the significant difference between treatments (univariable HR 2.6, multivariable HR 1.9). Furthermore, a competing risks regression analysis, taking into account differences between groups in non-RCC causes of mortality, yielded identical results (multivariable HR 1.9). These models account for all known potential confounders except the presence of benign histology, which may be present in 20% of cT1a renal masses [17]. Because far more patients in the ablation group lacked pathological confirmation, it is highly likely that these patients more often had benign disease. However, this would bias the results toward the null, and therefore the benefit of NSS is probably larger than that observed. In fact, a sub-group analysis in only those with pathological confirmation showed a further increase in the risk of death with ablation (HR 2.2), though because of small sample size the test of interaction was not significant.

Other factors which are strong determinants of DSS such as grade [18] or histology [19], cannot confound the association between treatment type and outcome since they were unknown at time of treatment in the majority of patients [20]. The observation that grade was lower in patients undergoing ablation than NSS is probably attributable to measurement error, as there is evidence that grade determined by biopsy rather than pathology is systematically lower [21,22]. Although it is predictable that adjustment for grade would worsen relative outcomes for ablated patients (HR 3.2), this is probably biased by measurement error of grade.

Our analyses also provide evidence that the specific technique of thermal ablation is important and may result in variable outcomes. Ablation of unspecified type had the greatest risk of death compared with NSS (HR 2.5), while cryoablation had the lowest relative risk of death (HR 1.5), with radiofrequency ablation inbetween (HR 2.2). A meta-analysis of ablative techniques showed improved efficacy with cryoablation vs radiofrequency ablation for both retreatment (1% vs. 8%) and local disease progression (5% vs. 13%) [23]. The proportion of patients undergoing cryoablation increased significantly over time. If cryoablation is the most effective form of ablation, this may explain the observation that the difference in DSS between ablation and NSS decreased over time. It is also possible that increased experience and technical improvements [24–26] over time have minimized the difference in survival between those undergoing ablation and NSS. Further comparative studies are necessary to elucidate the optimum technology in specific situations, as well as details of the best method of delivering thermal ablation.

The present study is the first well-powered study to compare the effectiveness of NSS and ablation. Two previous studies, which attempted to compare oncological outcomes between treatments, had <40 patients in the ablation arms [7,27]. Two studies have reported single-institution outcomes of ablation and concluded they were similar to outcomes in the literature of patients undergoing NSS [8,28]. A radiofrequency ablation cohort (N= 208) had reported 5-year DSS of 99%; however, recurrence-free survival was only 90% in those with biopsy-proven RCC and metastasis-free survival was 95%, suggesting that DSS will decrease significantly over time. A cryoablation cohort (N= 80) had reported 5-year and 10-year DSS of 92% and 83%, respectively. In contrast with the authors' conclusion, these results do not appear similar to the literature on NSS outcomes for cT1a RCC [29]. However, this may be attributable to selection bias within the cryoablation cohort, particularly because a significant number of patients had previous RCC.

Limitations of the present study include the relatively short follow-up in the ablation group; however, in the more than 100 patients in the ablation group who had >5-years of follow-up, kidney cancer mortality appeared to increase relative to NSS (HR 3.0). A cohort effect cannot be excluded, as those with longest follow-up were also treated earliest. Using a cancer registry, we were unable to retrospectively determine unmeasured factors which might be of interest, such as the reason that surgery or ablation was performed, or the Padua classification. Another limitation of SEER is the absence of details regarding ablative techniques: cryoablation or radiofrequency ablation, percutaneous or laparoscopic approach, and specific device and type and number of probes. Performance status or comorbidity data are not available in SEER and could not be adjusted for individually, although competing risks regression did not change the findings. Finally, SEER relies on death certificate reporting, which is imperfect, although in patients with malignancy accuracy may be increased [30].

REFERENCES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. CONFLICT OF INTEREST
  8. REFERENCES