EURURO-7775; No.

of Pages 8

E U R O P E A N U R O L O G Y X X X ( 2 018 ) X X X – X X X

available at www.sciencedirect.com
journal homepage: www.europeanurology.com

Platinum Priority – Kidney Cancer

Editorial by XXX on pp. x–y of this issue

Active Surveillance for Localized Renal Masses: Tumor Growth,

Delayed Intervention Rates, and >5-yr Clinical Outcomes

Andrew G. McIntosh a,b,*, Benjamin T. Ristau b,c, Karen Ruth b, Rachel Jennings d, Eric Ross b,
Marc C. Smaldone b, David Y.T. Chen b, Rosalia Viterbo b, Richard E. Greenberg b,
Alexander Kutikov b, Robert G. Uzzo b
a
Temple University Hospital/Fox Chase Cancer Center, Temple Health, Philadelphia, PA, USA; b Fox Chase Cancer Center, Temple Health, Philadelphia, PA, USA;
c
UConn Health, Farmington, CT, USA; d Temple University, Lewis Katz School of Medicine, Philadelphia, PA, USA

Article info Abstract

Article history: Background: Active surveillance (AS) has gained acceptance as a management strategy for
Accepted March 14, 2018 localized renal masses.
Objective: To review our large single-center experience with AS.
Design, setting, and participants: From 2000 to 2016, we identified 457 patients with
Associate Editor:
544 lesions managed with AS from our prospectively maintained kidney cancer database.
Stephen Boorjian A subset analysis was performed for patients with 5-yr follow-up without delayed inter-
vention (DI).
Statistical Editor: Outcome measurements and statistical analysis: Linear growth rates (LGRs) were estimated
Andrew Vickers using linear regression for the initial LGR (iLGR) AS interval and the entire AS period. Overall
survival (OS) and cumulative incidence of DI were estimated with Kaplan-Meier methods
utilizing iLGR groups, adjusting for covariates. DI was evaluated for association with OS in Cox
Keywords: models.
Active surveillance Results and limitations: Median follow-up was 67 mo (interquartile range [IQR] 41–94 mo) for
surviving patients. Cumulative incidence of DI (n = 153) after 1, 2, 3, 4, and 5 yr was 9%, 22%,
Delayed intervention
29%, 35%, and 42%, respectively. Median initial maximum tumor dimension was 2.1 cm (IQR
Oncology 1.5–3.1 cm). Median iLGR and overall LGR were 1.9 (IQR 0–7) and 1.9 (IQR 0.3–4.2) mm/yr,
Renal cell carcinoma respectively. Compared with the no growth group, low iLGR (hazard ratio [HR] 1.25, 95%
Renal mass cumulative incidence [CI] 0.82–1.91), moderate iLGR (HR 2.1, 95% CI 1.31–3.36), and high iLGR
(HR 1.87, 95% CI 1.23–2.84) were associated with DI (p = 0.003). The iLGR was not associated
Tumor growth kinetics with OS (p = 0.8). DI was not associated with OS (HR 1.34, 95% CI 0.79–2.29, p = 0.3). Five-year
cancer-specific mortality (CSM) was 1.2% (95% CI 0.4–2.8%). Of 99 patients on AS without DI for
>5 yr, one patient metastasized.
Conclusions: At >5 yr, AS  DI is a successful strategy in carefully managed patients. DI often
occurs in the first 2–3 yr, becoming less likely over time. Rare metastasis and low CSM rates
should reassure physicians that AS is safe in the intermediate to long term.
Patient summary: In this report, we looked at the outcomes of patients with kidney masses
who elected to enroll in active surveillance rather than immediate surgery. We found that
patients who need surgery are often identified early and those who remain on active surveil-
lance become less likely to need surgery over time. We concluded that active surveillance with
or without delayed surgery is a safe practice and that, when properly managed and followed,
patients are unlikely to metastasize or die from kidney cancer.
© 2018 European Association of Urology. Published by Elsevier B.V. All rights reserved.

* Corresponding author. Department of Urology, Temple University Hospital, 3401 N. Broad Street,
Suite 330C, Philadelphia, PA 19140, USA. Tel. +1 215-707-3375; Fax: +1 215 707 8455.
E-mail address: [email protected] (A.G. McIntosh).

https://doi.org/10.1016/j.eururo.2018.03.011
0302-2838/© 2018 European Association of Urology. Published by Elsevier B.V. All rights reserved.

Please cite this article in press as: McIntosh AG, et al. Active Surveillance for Localized Renal Masses: Tumor Growth, Delayed
Intervention Rates, and >5-yr Clinical Outcomes. Eur Urol (2018), https://doi.org/10.1016/j.eururo.2018.03.011
EURURO-7775; No. of Pages 8

2 E U R O P E A N U R O L O GY X X X ( 2 0 18 ) X X X – X X X

1. Introduction American Society of Clinical Oncology [15], European

Association of Urology [7], and American Urological Associa-
Renal cancers are projected to comprise an estimated 63 tion [8] guidelines. We reviewed our large single-center
990 new cancer diagnoses in 2017 [1]. The incidence of experience with AS for renal masses with a median of >5 yr of
incidentally discovered localized renal masses has risen due follow-up. We evaluated patterns and associations in tumor
in part to increased utilization of ultrasound and cross- growth kinetics, rates of crossover to and predictors of DI, and
sectional imaging [2]. Using tumor registry data, a clinical outcomes including overall survival (OS), cancer-
paradoxical rise in mortality from renal cell carcinoma specific mortality (CSM), and progression to metastasis.
(RCC) over time has been described, despite increased Further, we analyzed outcomes for a subset of patients who
detection and intervention [2]. Explanations for this remained on AS for >5 yr without DI.
phenomenon—termed “treatment disconnect”—have been
proposed, and when one examines cancer-specific survival
(CSS) in all comers over time, rates have remained relatively 2. Patients and methods
stable [3].
Active surveillance (AS) has emerged as an initial 2.1. Patient selection
management option to address the variable clinical signifi-
Our institutional review board–approved renal cancer database (-
cance [4,5] and potential overtreatment of localized renal
n = 3120) was reviewed for masses or enhancing cystic lesions (Bosniak
masses [6–8]. The paradigm of AS is to identify patients with
III and IV) managed by AS from January 2000 through July 2016 (Fig. 1).
potentially low-risk renal masses contextualized by their
Lesions were localized (cT1-2N0M0) based on established radio-
competing risks of death during an initial observation graphic staging protocols. Variables examined included patient and tumor
period [9]. This period allows the triage of patients into characteristics, duration of AS, tumor growth (linear), crossover to DI, and
continued AS or delayed intervention (DI), a decision most rates of metastasis and death. Patients were followed for oncologic
often made on the basis of radiographic tumor growth outcomes after DI. There were no predefined selection criteria for AS.
kinetics, evolving and competing risks, and/or patient
preference [10,11]. 2.2. Imaging and tumor growth kinetics
Published AS cohorts offer encouraging early results, albeit
limited by relatively short lengths of follow-up [12–14]. De- Radiographic surveillance was performed at a median interval of 6.7 mo
spite these limitations, AS has been incorporated into recent (interquartile range [IQR] 4.6–12 mo) across 2667 abdominal images

Fig. 1 – Flow diagram for patient selection and exclusions from Fox Chase Cancer Center kidney tumor database. AS = active surveillance.

Please cite this article in press as: McIntosh AG, et al. Active Surveillance for Localized Renal Masses: Tumor Growth, Delayed
Intervention Rates, and >5-yr Clinical Outcomes. Eur Urol (2018), https://doi.org/10.1016/j.eururo.2018.03.011
EURURO-7775; No. of Pages 8

E U R O P E A N U R O L O G Y X X X ( 2 0 18 ) X X X – X X X 3

(1463 [55%] computed tomography scans, 904 [34%] magnetic resonance OS was measured from first follow-up surveillance imaging to death,
imaging scans, and 300 [11%] abdominal ultrasounds). censoring those alive at the most recent follow-up. Cox proportional
Simple linear regression was used to estimate linear growth rate hazards regression adjusted for age and tumor composition; covariate-
(LGR) per lesion. Separate analyses were performed based on tumor adjusted OS curves for iLGR groups were estimated at the mean age and
composition (cystic vs solid) and size (>4 cm). for solid tumor composition.
To assess for associations between DI and OS, a Cox proportional
2.3. Statistical analysis hazards model was utilized, which included DI as a time-dependent
covariate, and tumor composition and age at presentation as time-
Patient and tumor characteristics were compared using chi-square or invariant covariates. Here, OS was measured from presentation,
Wilcoxon rank sum tests. Time on AS was measured from the date of censoring at the most recent follow-up.
presentation to the most recent imaging. Spearman's correlation A subset analysis included patients with >5 yr of follow-up and no
assessed initial maximum tumor dimension (MTD) and overall LGR. crossover to DI. The proportional hazards assumption was evaluated for
The cumulative incidence of DI was estimated using Kaplan-Meier (KM) all parameters in multivariable Cox models. Statistical analyses were
methods, censoring at the last imaging study. In sensitivity analyses, performed using SAS software, version 9.4 (SAS; Cary, NC, USA).
competing risk methods were used to account for the risk of death
precluding DI. Cox proportional hazards assessed associations between 3. Results
baseline patient characteristics and DI.
To examine the associations between tumor growth and DI, and tumor
3.1. Cohort description and tumor characteristics
growth and OS, an initial LGR (iLGR) was estimated (slope between MTD
at presentation and MTD at first follow-up surveillance imaging). Lesions
We identified 457 patients (544 lesions) enrolled in AS for
were subdivided into the following groups: no growth (<1 mm/yr), low
iLGR (1–<5 mm/yr), moderate iLGR (5–10 mm/yr), and high iLGR whom sufficient data were available (Fig.1). The median age at
(>10 mm/yr). Time to DI for iLGR groups were estimated using KM presentation was 70 yr (IQR 60–78 yr). Median follow-up was
analysis. Here, time to DI was measured from the first follow-up 67 mo (IQR 41–94 mo) for surviving patients. Median time on
surveillance imaging date, and survival distributions were compared AS was 41 mo (IQR 19–71 mo) for patients who did not cross
using log rank tests. Cox proportional hazards regression adjusted for age. over to DI. Patient baseline characteristics are listed in Table 1.

Table 1 – Patient baseline demographic characteristics for entire cohort, by delayed intervention status, tumor composition, and 5-yr subset
analysis

Characteristic (categorical variables) All Delayed intervention Tumor composition 5-yr

subset
No Yes Solid Cystic

N % N % N % p value N % N % p value N %

No. of patients 457 100 304 67 153 34 386 85 71 16 99 22

Gender 0.9 0.3
Female 176 39 118 39 58 38 153 40 23 32 35 35
Male 281 62 186 61 95 62 233 60 48 68 64 65
Race 0.1 0.2
White 388 85 261 86 127 83 332 86 56 79 86 87
Black 56 12 32 11 24 16 43 11 13 18 9 9
Other 13 3 11 4 2 1 11 3 2 3 4 4
Smoking status 0.002 0.4
Current 44 10 20 7 24 16 38 10 6 9 5 5
Former 196 43 127 42 69 45 160 42 36 51 41 41
Nonsmokers 217 48 157 52 60 39 188 49 29 41 53 54
Hypertension 0.5 0.5
No 128 28 82 27 46 30 106 28 22 31 24 24
Yes 329 72 222 73 107 70 280 73 49 69 75 76
No. of lesions on AS 0.5 0.3
1 389 85 260 86 129 84 325 84 64 90 80 81
2 58 13 39 13 19 12 53 14 5 7 16 16
3–6 10 2 5 2 5 3 8 2 2 3 3 3

Characteristic (continuous variables) All Delayed intervention Tumor composition 5-yr subset

No Yes Solid Cystic

N Median N Median N Median p value N Median N Median p value N Median

(IQR) (IQR) (IQR) (IQR) (IQR) (IQR)

Age at presentation (yr) 457 70 (60–78) 304 72 (64–79) 153 65 (55–73) <0.001 386 71 (62–78) 71 65 (56–71) <0.001 99 72 (65–77)
CCI 1 (0–2) 1 (0–2) 1 (0–2) 0.9 1 (0–2) 1 (0–2) 0.8 1 (0–2)
BMI (kg/m2) 28 (25–32) 28 (25–32) 29 (26–33) 0.3 28 (25–32) 29 (27–33) 0.1 28 (26–31)

AS = active surveillance; BMI = body mass index; CCI = Charlson comorbidity index; IQR = interquartile range.

Please cite this article in press as: McIntosh AG, et al. Active Surveillance for Localized Renal Masses: Tumor Growth, Delayed
Intervention Rates, and >5-yr Clinical Outcomes. Eur Urol (2018), https://doi.org/10.1016/j.eururo.2018.03.011
EURURO-7775; No. of Pages 8

4 E U R O P E A N U R O L O GY X X X ( 2 0 18 ) X X X – X X X

Solid lesions and Bosniak III/IV cystic lesions comprised Fig. 3B). DI estimates calculated using competing risk
85% (n = 460) and 15% (n = 84) of the cohort, respectively. method were very close to that obtained using the
The median MTD at presentation was 2.1 cm (IQR 1.5– unadjusted method given the low number of competing
3.1 cm). There was no difference between the median MTD risk deaths (Supplementary Table 2). Adjusting for age,
values (2.1 vs 2.05 cm, p = 0.7) for solid and cystic masses. when compared with the no growth group, those in the low
Sixty-nine (13%) lesions had an MTD of >4 cm (56 solid and iLGR (hazard ratio [HR] 1.25, 95% CI 0.82–1.91), moderate
13 cystic). The distribution of MTD at presentation can be iLGR (HR 2.1, 95% CI 1.31–3.36), and high iLGR (HR 1.87, 95%
found in Supplementary Figure 1. CI 1.23–2.84) groups were more likely to have DI (p = 0.003).
The proportion of patients having DI grouped by duration of
3.2. Tumor growth rates AS and iLGR groups are illustrated in Supplementary
Figure 3.
The median LGR for all lesions was 1.9 mm/yr (IQR 0.2– Cox proportional hazards regression analysis for factors
4.2 mm/yr) and median iLGR 1.9 mm/yr (IQR 0–7 mm/yr). associated with DI are found in Table 2. On univariate
The median interval between presentation and first analysis, these included younger age (p < 0.001) and
surveillance image, used to establish iLGR groups, was patients who have smoked (p < 0.001). On exploratory
6 mo (IQR 4–13 mo). Of these, 38% (n = 208) had no growth, multivariable analysis, the HRs were substantially similar to
41% (n = 222) had low iLGR, 12% (n = 66) had moderate iLGR, the univariate estimates provided in Table 2.
and 9% (n = 48) had high iLGR. Overall LGR was not Approximately one-fourth (24%) of patients underwent
significantly different between solid and cystic masses renal mass biopsy. Histologic distribution is provided in
(1.9 vs 1.6 mm/yr, p = 0.9). Overall LGR was not significantly Supplementary Figure 4. Biopsy histologic concordance on
higher for larger (>4 cm) masses (2 vs 1.9 mm/yr, p = 0.7). final pathology was >90%.
There was no significant correlation between initial MTD
and overall LGR (r = 0.02, p = 0.6). The distribution of overall 3.4. Clinical outcomes
LGR and DI status is shown in Figure 2.
For the analytic cohort, 5-yr OS estimate was 89% (95% CI
3.3. Delayed intervention 85–92%); overall, there were 73 deaths. Those who had DI
were not at a significantly high risk compared with those
Cumulative incidence of DI (n = 153) after 1, 2, 3, 4, and 5 yr who remained on AS (HR 0.87, 95% CI 0.52–1.45, p = 0.6).
on AS was 9% (95% cumulative incidence [CI] 7–12%), 22% The iLGR group was not associated with differences in OS
(95% CI 18–26%), 29% (95% CI 25–34%), 35% (95% CI 30–40%), (Fig. 4A). When adjusting for age and tumor composition,
and 42% (95% CI 37–48%), respectively. For patients who had there remained no difference (Fig. 4B). Patients with cystic
DI, surgical pathologic data can be found in Supplementary masses had improved 5-yr OS compared with those with
Table 1 and the distribution of time on AS in Supplementary solid masses (p = 0.01; Fig. 4C). Risk of death was not
Figure 2. Figure 3A reflects that there was no difference in significantly higher for larger tumors (MTD >4 cm; HR 1.6,
the probability of patients crossing over to DI between solid 95% CI 0.9–3.1, p = 0.1).
and cystic masses (p = 0.5). An increasing iLGR was Eight patients were known to have progressed to
associated with higher cumulative DI rates (p = 0.02; metastatic disease (median MTD 2.2 cm [IQR 2.2–4.5 cm];

Fig. 2 – Waterfall plot demonstrating the distribution of linear growth rates (LGRs), by delayed intervention (DI) status. In the event a patient had
multiple lesions, the largest lesion (for no DI patients) or the first lesion to have surgery (DI patients) was utilized.

Please cite this article in press as: McIntosh AG, et al. Active Surveillance for Localized Renal Masses: Tumor Growth, Delayed
Intervention Rates, and >5-yr Clinical Outcomes. Eur Urol (2018), https://doi.org/10.1016/j.eururo.2018.03.011
EURURO-7775; No. of Pages 8

E U R O P E A N U R O L O G Y X X X ( 2 0 18 ) X X X – X X X 5

Fig. 3 – (A) Cumulative incidence (CI) of delayed intervention (DI) over time—no difference between solid and cystic masses at 2 yr (32% [95% CI 27–
37%] vs 30% [95% CI 20–44%]) and 3 yr (40% [95% CI 34–46%] vs 32% [95% CI 22–46%]; p = 0.5). (B) Increasing iLGR was associated with higher
cumulative DI rates (p = 0.018). CI estimates of DI for the no growth, low iLGR, moderate iLGR, and high iLGR groups were 15%, 20%, 26%, and 34%, and
34%, 36%, 51%, and 45%, at 1 and 3 yr, respectively. iLGR = initial linear growth rate; LGR = linear growth rate.

median overall LGR 7.2 mm/yr [IQR 2.9–12.7 mm/yr]); five 1.1 (IQR 0.4–2.5) and 1 (IQR 0.4–1.5) mm/yr for solid and
of these patients died from advanced RCC (5-yr CSM 1.2% cystic masses, respectively. Twelve patients died, with one
[95% CI 0.4–2.8%]). Five patients progressed to metastasis patient developing metastatic disease and ultimately dying
after DI and two patients prior to DI, and one never from metastatic RCC >8 yr after enrolling in AS. Five
underwent DI. Given the low risk of CSM, no differences patients died due to unrelated causes, two patients died due
were identified based on any variables measured. to other malignancies, and one patient's cause of death was
unknown.
3.5. Subset analysis
4. Discussion
Ninety-nine patients did not cross over to DI and had >5 yr
of follow-up. Median follow-up was 99 mo (IQR 77–115 mo; “Cancer” encompasses a group of disorders with a
Table 1). The median MTD at presentation was 21 (IQR 14– heterogeneous natural history. Recently, overdiagnosis
28) and 19 (IQR 15–22) mm for solid (n = 82) and cystic and overtreatment have been recognized phenomena in
(n = 17) masses, respectively. The median overall LGR was multiple incidentally detected cancers [2,3,16,17]. The

Please cite this article in press as: McIntosh AG, et al. Active Surveillance for Localized Renal Masses: Tumor Growth, Delayed
Intervention Rates, and >5-yr Clinical Outcomes. Eur Urol (2018), https://doi.org/10.1016/j.eururo.2018.03.011
EURURO-7775; No. of Pages 8

6 E U R O P E A N U R O L O GY X X X ( 2 0 18 ) X X X – X X X

Table 2 – Univariate analysis (Cox proportional hazards regression) examining relationships between time to delayed intervention and
baseline patient and tumor characteristics

Parameter Comparison Univariate models

Hazard ratio 95% hazard ratio p value

confidence limits

Mass composition Cystic vs solid 0.85 0.54 1.33 0.5

MTD at presentation (cm) Continuous (difference of 1 cm) 1.00 0.89 1.13 >0.9
No. of lesions Continuous, log transformed 0.97 0.58 1.64 >0.9
Age at presentation (yr) Continuous (difference of 5 yr) 0.85 0.80 0.91 <0.001
Gender Male vs female 0.97 0.70 1.35 0.9
Race Black vs white 1.27 0.82 1.97 0.2
Other vs white 0.39 0.10 1.58
Smoking status Current vs never 2.33 1.45 3.75 <0.001
Former vs never 1.31 0.93 1.85
Hypertension Yes vs no 0.84 0.59 1.18 0.6
BMI (kg/m2) Continuous (difference of 1 kg/m2) 1.03 1.00 1.05 0.057
CCI Continuous, log transformed 1.07 0.83 1.39 0.6

BMI = body mass index; CCI = Charlson comorbidity index; MTD = median tumor diameter.

complex mix of risks posed by biology, competing condi- literature [25], simple calculation, and ease of translation to
tions, age, patient perceptions, and care processes is only patients may still render it a useful tool and surrogate for
recently being addressed in the context of a cancer growth.
diagnosis [18]. Randomized data comparing primary Cumulatively, approximately one-third of our patients
intervention with surveillance to identify benefits on crossed over to DI. Importantly, DI was uncommon after 2–3
cancer-specific outcomes require large, long-term studies yr. This is consequential as it suggests that clinicians and
[19]. These trials are impractical and expensive, leaving the patients alike can be reassured that over time AS remains
field with prospective registries like ours to try to answer safe, DI becomes less likely, and those who choose or
important questions regarding the “null hypothesis” of “require” DI can often be identified early. Prior studies have
treatment effects. observed DI rates ranging from 4% to 39% [14,20–22,26]. Our
We thus reviewed our large prospectively maintained findings of relatively higher rates of DI likely reflect our
kidney tumor database to evaluate our AS cohort for LGRs, institutional priority to enroll eligible patients in an initial
DI rates, and clinical outcomes. To our knowledge, these are period of AS. Other centers may have higher rates of primary
the only AS data reporting outcomes for a cohort of this size intervention.
over intermediate- to long-term follow-up. Interestingly, although patients with cystic masses had
The median overall LGR was 1.9 mm/yr, and there was no improved OS, they were not associated with lower LGRs or
association between LGR and MTD. Prior AS cohorts have rates of DI. Our results suggest that while cystic masses are
noted mean LGRs of 0.7–2.9 mm/yr [20–22], and prospec- thought to be relatively prognostically favorable [27], they
tive series with short-term follow-up reveal modest LGRs of are often treated similarly to solid lesions. While guidelines
0.9 mm/yr (median) [14] and 1.2 mm/yr (mean) [23]. Prior [8] suggest AS for Bosniak III/IV masses, they may be able to
studies have similarly observed no association between be managed with higher thresholds for DI.
initial tumor size and growth rates [24], and potentially the Findings of rare CSM, including those who remained on
presence of an inverse relationship suggesting MTD should AS for over 5 yr and those with masses >4 cm, echo the
not be the sole determinant of AS versus immediate results of a recent meta-analysis [12], which demonstrated
intervention [22]. few patients metastasizing over short to intermediate
We found no difference in covariate-adjusted OS across follow-up periods. Additionally, LGRs and OS were similar
iLGR groups, bringing into question the utility of LGR as a between masses >4 and 4 cm. We interpret these results
significant predictor of oncologic risk. Notably, for the to suggest that if significant competing risks exist, larger
purposes of our time-to-event analysis, we grouped masses can be cautiously placed on AS. Our findings
patients by iLGR. Although the median values for iLGR complement those of existing literature, highlighting the
and overall LGR were identical, tumor growth is likely not safety of AS and cautioning overinterpretation of the
truly linear, limiting the interpretation of our results. benefits of early intervention in tumors of low biologic
However, we observe a rapid median LGR (7.2 mm/yr) in significance.
the eight patients who progressed to metastasis. This DI with the aim of preventing stage progression and
finding, conversely, suggests that a higher LGR across the AS metastasis is a critical principle of appropriately managed
period may be associated with more aggressive cancer, AS. Only one patient in our subset of patients with >5 yr of
complementing previous literature noting an association follow-up without DI died from RCC. This occurred in year
between LGR and progression to metastasis [11]. While LGR 9 of AS, suggesting that competing risks or other extenuat-
assumes uniform growth and may not be as accurate as ing circumstances precluded DI. This is consistent with a
volumetric measurements, its extensive precedent in the prior pooled analysis [11] observing that metastasis is

Please cite this article in press as: McIntosh AG, et al. Active Surveillance for Localized Renal Masses: Tumor Growth, Delayed
Intervention Rates, and >5-yr Clinical Outcomes. Eur Urol (2018), https://doi.org/10.1016/j.eururo.2018.03.011
EURURO-7775; No. of Pages 8

E U R O P E A N U R O L O G Y X X X ( 2 0 18 ) X X X – X X X 7

typically a late event and that significant lead time bias may
be a factor. While direct comparative data are lacking,
nomograms for localized low-risk masses undergoing
immediate intervention exhibit CSS similar to that of our
cohort [28]. Our findings reassure that even at 5 yr and
beyond, metastasis remained a rare event.
Although our AS program is prospectively maintained,
there remain limitations of this and other nonrandomized
studies. These include a lack of predefined eligibility
criteria or rigorously defined follow-up schedules. Patient
selection reflects surgeon and patient heterogeneity
regarding risk tolerance. Two-thirds of our cohort are
<75 yr old with significant life expectancy, and outcomes
should continue to be evaluated as the cohorts mature.
Further, optimal imaging schedules have not been defined,
which may limit the calculated growth rates. Finally, LGR
assumes a spherical tumor shape and uniform growth, and
may be less accurate than volumetric measurements.
Continued validation from other prospective registries
[14,23] may be revealing as longer follow-up data become
available.

5. Conclusions

AS  DI can be a successful strategy in patients with

localized renal masses. Most patients who cross over to DI
are likely to do so within the first 2–3 yr on AS. This and
excellent CSM rates across a variety of masses should
reassure physicians that AS is a safe oncologic practice in
thoughtfully managed, closely followed patients in inter-
mediate- to long-term follow-up periods and should
encourage recruitment into AS protocols, where appro-
priate.

Author contributions: Andrew G. McIntosh had full access to all the data in
the study and takes responsibility for the integrity of the data and the
accuracy of the data analysis.

Study concept and design: McIntosh, Ristau, Uzzo, Ruth, Ross.

Acquisition of data: McIntosh, Smaldone, Viterbo, Uzzo, Kutikov, Chen,
Jennings, Greenberg.
Analysis and interpretation of data: McIntosh, Ruth, Ristau, Ross, Jennings.
Drafting of the manuscript: McIntosh, Ristau, Ruth, Ross.
Critical revision of the manuscript for important intellectual content:
McIntosh, Ristau, Uzzo, Kutikov, Chen, Smaldone, Viterbo, Greenberg.
Statistical analysis: Ruth, Ross.
Obtaining funding: None.
Administrative, technical, or material support: None.
Supervision: Uzzo.
Fig. 4 – (A) Kaplan-Meier (KM) estimates for overall survival (OS) from Other: None.
first surveillance imaging date between initial LGR (iLGR) groups. No
difference is demonstrated between 5-yr OS rates in the no growth Financial disclosures: Andrew G. McIntosh certifies that all conflicts of
(85%, 95% CI 78–90%), low iLGR (89%, 95% CI 79–94%), moderate iLGR interest, including specific financial interests and relationships and
(83%, 95% CI 67–92%), and high iLGR (83%, 95% CI 71–90%) groups
affiliations relevant to the subject matter or materials discussed in the
(p = 0.4). (B) Age- and tumor composition–adjusted Cox regression for
OS also demonstrates no difference in OS by iLGR group (p = 0.8). (C) KM
manuscript (eg, employment/affiliation, grants or funding, consultan-
estimates for OS by tumor composition demonstrates improved 5-yr OS cies, honoraria, stock ownership or options, expert testimony, royalties,
for cystic (95%, 95% CI 84–98%) versus solid (88%, 95% CI 84–91%) or patents filed, received, or pending), are the following: None.
masses (p = 0.01). CI = cumulative incidence; HR = hazard ratio;
LGR = linear growth rate.
Funding/Support and role of the sponsor: This study was supported by
Fox Chase Cancer Center Core Grant, the Gitlin Family Foundation, and
Fox Chase Cancer Center Kidney Cancer Database. The funding
organizations had a role in the collection and management of the data.

Please cite this article in press as: McIntosh AG, et al. Active Surveillance for Localized Renal Masses: Tumor Growth, Delayed
Intervention Rates, and >5-yr Clinical Outcomes. Eur Urol (2018), https://doi.org/10.1016/j.eururo.2018.03.011
EURURO-7775; No. of Pages 8

8 E U R O P E A N U R O L O GY X X X ( 2 0 18 ) X X X – X X X

Acknowledgments: The authors thank Debra Kister and Michelle Collins [14] Uzosike AC, Patel HD, Alam R, et al. Growth kinetics of small renal
for their help in data processing and management. masses on active surveillance: variability and results from the
DISSRM Registry. J Urol. In press. doi:10.1016/j.juro.2017.09.087.
[15] Finelli A, Ismaila N, Bro B, et al. Management of small renal masses:
Appendix A. Supplementary data
American Society of Clinical Oncology clinical practice guideline. J
Clin Oncol 2017;35:668–80.
Supplementary data associated with this article can be
[16] Esserman LJ, Thompson IM, Reid B, et al. Addressing overdiagnosis
found, in the online version, at https://doi.org/10.1016/j. and overtreatment in cancer: a prescription for change. Lancet
eururo.2018.03.011. Oncol 2014;15:e234–42.
[17] Hoffmann TC, Del Mar C. Patients’ expectations of the benefits and
References harms of treatments, screening, and tests: a systematic review.
JAMA Intern Med 2015;175:274–86.
[1] Siegel RL. Cancer statistics, 2017. CA Cancer J Clin 2017;67:7–30. [18] Elmore JG. Solving the problem of overdiagnosis. N Engl J Med
[2] Hollingsworth JM, Miller DC, Daignault S, Hollenbeck BK. Rising 2016;375:1483–6.
incidence of small renal masses: a need to reassess treatment effect. [19] Hamdy FC, Donovan JL, Lane JA, et al. 10-Year outcomes after
JNCI J Natl Cancer Inst 2006;98:1331–4. monitoring, surgery, or radiotherapy for localized prostate cancer.
[3] Smaldone MC, Egleston B, Hollingsworth JM, et al. Understanding N Engl J Med 2016;375:1415–24.
treatment disconnect and mortality trends in renal cell carcinoma [20] Dorin R, Jackson M, Cusano A, et al. Active surveillance of renal
using tumor registry data. Med Care 2017;55:398–404. masses: an analysis of growth kinetics and clinical outcomes strat-
[4] Kutikov A, Fossett LK, Ramchandani P, et al. Incidence of benign ified by radiological characteristics at diagnosis. Int Braz J Urol
pathologic findings at partial nephrectomy for solitary renal mass 2014;40:627–36.
presumed to be renal cell carcinoma on preoperative imaging. [21] Abouassaly R, Lane BR, Novick AC. Active surveillance of renal
Urology 2006;68:737–40. masses in elderly patients. J Urol 2008;180:505–8, discussion
[5] Rothman J, Egleston B, Wong YN, Iffrig K, Lebovitch S, Uzzo RG. 508-9.
Histopathological characteristics of localized renal cell carcinoma [22] Crispen PL, Viterbo R, Boorjian SA, Greenberg RE, Chen DY, Uzzo RG.
correlate with tumor size: a SEER analysis. J Urol 2009;181:29–33, Natural history, growth kinetics, and outcomes of untreated clini-
discussion 33–4. cally localized renal tumors under active surveillance. Cancer
[6] Chawla SN, Crispen PL, Hanlon AL, Greenberg RE, Chen DYT, Uzzo 2009;115:2844–52.
RG. The natural history of observed enhancing renal masses: meta- [23] Organ M, Jewett M, Basiuk J, et al. Growth kinetics of small renal
analysis and review of the world literature. J Urol 2006;175:425–31. masses: a prospective analysis from the Renal Cell Carcinoma
[7] Ljungberg B, Bensalah K, Canfield S, et al. EAU guidelines on renal Consortium of Canada. Can Urol Assoc J 2014;8:24–7.
cell carcinoma: 2014 update. Eur Urol 2015;67:913–24. [24] Abou Youssif T, Kassouf W, Steinberg J, Aprikian AG, Laplante MP,
[8] Campbell S, Uzzo RG, Allaf ME, et al. Renal mass and localized renal Tanguay S. Active surveillance for selected patients with renal
cancer: AUA guideline. J Urol 2017;198:520–9. masses: updated results with long-term follow-up. Cancer
[9] Kutikov A, Egleston BL, Wong YN, Uzzo RG. Evaluating overall 2007;110:1010–4.
survival and competing risks of death in patients with localized [25] Ristau BT, Correa AF, Uzzo RG, Smaldone MC. Active surveillance for
renal cell carcinoma using a comprehensive nomogram. J Clin Oncol the small renal mass: growth kinetics and oncologic outcomes. Urol
2010;28:311–7. Clin North Am 2017;44:213–22.
[10] Ristau BT, Kutikov A, Uzzo RG, Smaldone MC. Active surveillance for [26] Rosales JC, Haramis G, Moreno J, et al. Active surveillance for renal
small renal masses: when less is more. Eur Urol Focus 2016;2:660–8. cortical neoplasms. J Urol 2010;183:1698–702.
[11] Smaldone MC, Kutikov A, Egleston BL, et al. Small renal masses [27] Han KR, Janzen NK, McWhorter VC, et al. Cystic renal cell carcino-
progressing to metastases under active surveillance: a systematic ma: biology and clinical behavior. Urol Oncol 2004;22:410–4.
review and pooled analysis. Cancer 2012;118:997–1006. [28] Frank I, Blute ML, Cheville JC, Lohse CM, Weaver AL, Zincke H. An
[12] Pierorazio PM, Johnson MH, Patel HD, et al. Management of renal outcome prediction model for patients with clear cell renal cell
masses and localized renal cancer: systematic review and meta- carcinoma treated with radical nephrectomy based on tumor
analysis. J Urol 2016;196:989–99. stage, size, grade and necrosis: the SSIGN score. J Urol
[13] Jewett MAS, Mattar K, Basiuk J, et al. Active surveillance of small 2002;168:2395–400.
renal masses: progression patterns of early stage kidney cancer. Eur
Urol 2011;60:39–44.

Please cite this article in press as: McIntosh AG, et al. Active Surveillance for Localized Renal Masses: Tumor Growth, Delayed
Intervention Rates, and >5-yr Clinical Outcomes. Eur Urol (2018), https://doi.org/10.1016/j.eururo.2018.03.011