- Email: [email protected]
Variation in Locoregional Prostate Cancer Care and Treatment Trends at Commission on Cancer Designated Facilities: A National Cancer Data Base Analysis 2004 to 2013
Original Study
Variation in Locoregional Prostate Cancer Care and Treatment Trends at Commission on Cancer Designated Facilities: A National Cancer Data Base Analysis 2004 to 2013 Björn Löppenberg,1,2 Akshay Sood,1 Deepansh Dalela,1 Patrick Karabon,1,3 Jesse D. Sammon,1,4 Malte W. Vetterlein,5 Joachim Noldus,2 James O. Peabody,1 Quoc-Dien Trinh,5 Mani Menon,1 Firas Abdollah1 Abstract Within the National Cancer Data Base we evaluated whether differences in treatment patterns for prostate cancer exist at Commission on Cancer facilities. Overall, 825,707 men were included in the retrospective analyses. We found substantial variation in treatment patterns between different facility types and individual institutions, as well. Our findings indicate a necessity for policy makers to harmonize prostate cancer treatment. Background: Contemporary treatment trends for prostate cancer show increased rates of active surveillance. However, nationwide applicability of these reports is limited. Additionally, the effect of Commission on Cancer facility type on prostate cancer treatment patterns is unknown. Patients and Methods: We used the National Cancer Data Base to identify men diagnosed with prostate cancer, between 2004 and 2013. Our cohort was stratified on the basis of the National Comprehensive Cancer Network prostate cancer risk classes. CochraneArmitage tests were used to evaluate temporal trends. Random effects hierarchical logit models were used to assess treatment variation at Commission on Cancer facility and institution level. Results: In 825,707 men, utilization of radiation therapy declined and utilization of radical prostatectomy increased for all prostate cancer risk groups between 2004 and 2013 (P < .0001). Observation for low-risk prostate cancer increased from 16.3% in 2004 to 2005 to 32.0% in 2012 to 2013 (P < .0001). Significant treatment variation was observed on the basis of Commission on Cancer facility type. Across all risk groups, the lowest rates of radical prostatectomy and highest rates of external beam radiation therapy were observed in community cancer programs. The highest rates of observation for low-risk disease were observed in academic centers. Treatment variation according to institution ranged from 14% (95% confidence interval, 0.12-0.15) for androgen deprivation therapy up to 59% (95% confidence interval, 0.45-0.73) for cryotherapy. Conclusion: The increased utilization of observation in low-risk prostate cancer is an encouraging finding, which appears to be mainly derived by a decrease in radiotherapy utilization in this risk group. Regardless of tumor characteristics, significant variations in treatment modality exist among different facility types and institutions. Clinical Genitourinary Cancer, Vol. -, No. -, 1-10 ª 2017 Elsevier Inc. All rights reserved. Keywords: Androgen deprivation therapy, Cryotherapy, Radiation therapy, Radical prostatectomy, Variation in care
B.L. and A.S. contributed equally to this work. 1 Center for Outcomes Research, Analytics and Evaluation, Vattikuti Urology Institute, Henry Ford Hospital, Detroit, MI 2 Department of Urology, Marien Hospital Herne, Ruhr-University Bochum, Herne, Germany 3 Department of Public Health Sciences, Henry Ford Hospital, Detroit, MI 4 Maine Medical Center, Division of Urology & Center for Outcomes Research and Evaluation, Portland, ME
1558-7673/$ - see frontmatter ª 2017 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.clgc.2017.04.014
5 Center for Surgery and Public Health, Division of Urology, Brigham and Women’s Hospital, Boston, MA
Submitted: Feb 25, 2017; Revised: Apr 5, 2017; Accepted: Apr 14, 2017 Address for correspondence: Firas Abdollah, MD, Vattikuti Urology Institute & VUI Center for Outcomes Research Analytics and Evaluation, Henry Ford Hospital, 2799 W Grand Blvd, Detroit, MI 48202-2689 E-mail contact: fi[email protected]
Clinical Genitourinary Cancer Month 2017
-1
Locoregional Prostate Cancer Care and Treatment Trends Introduction
Patient Selection
In 2017, approximately 161,360 men will be diagnosed with prostate cancer (PCa),1 with most having locoregional disease at presentation.1 The optimal first-line treatment for this tumor has been a subject of continuous debate, with several guideline revisions over the years. These became necessary with introduction of newer therapies and techniques such as the nerve-sparing prostatectomy, minimally invasive surgery, advanced radiotherapy (RT) techniques, and active surveillance. In the absence of randomized trials that have compared the different treatment modalities, at least in part, patient preferences,2 physician biases,2 and financial incentives3,4 have often guided treatment decisions. Nowadays, it is generally accepted that the treatment selection should be on the basis of tumor risk class and patients’ life expectancy.5-7 Treatment options vary from observation, offered to men with a limited life expectancy and/or low tumor risk class, to multimodal therapy, the latter reserved for men with high-risk disease, especially those with extended life expectancy.6 With changing recommendations—capturing current treatment patterns, their variability, and their conformance to guideline recommendations represents the key to a better understanding of the utilization of health resources. Indeed, several reports have tried to address this issue.8-13 All these reports have contributed significantly to our understanding of PCa treatment patterns, but they are limited in their generalizability for several reasons. These include a focus on a certain age groups, a certain tumor risk category, a certain national region, or Medicare-insured patients only.8-13 We thus designed a study to overcome these limitations, and investigate the current landscape of locoregional PCa care. To do so at a national level, we relied on the National Cancer Data Base (NCDB), which captures approximately 70% of all newly diagnosed tumors in the United States.14 The aim of our study was twofold: first, to assess the contemporary nationwide treatment patterns for locoregional PCa, stratified per PCa risk class and life expectancy; and second, to evaluate variation in treatment at institutional and Commission on Cancer (CoC) facility type level. We hypothesized: first, that at the macroscopic level the treatment patterns might conform to the current guideline recommendations; second, disparities might exist at the institution level, and finally patients across the United States might not receive equal care. Further, we expected to confirm the findings from recent regional reports on decreased local therapy for low risk PCa, at the national level. Our results will enable clinicians, urologists, oncologists, and radiation oncologists to compare their own institutions treatment patterns and trends with nationwide treatment patterns. Additionally, our results might be insightful for urologists, epidemiologists, and students, as well as to provide an overview of contemporary US PCa care.
Men 35 years with locoregional adenocarcinoma of the prostate (International Classification of diseases-O-3 code: C61.9),15 diagnosed between 2004 and 2013, were evaluated. Cases with information missing on clinical T stage (cT; Tx was considered as unknown, but not missing), Gleason score, prostate-specific antigen (PSA), and primary metastatic cases, were excluded. More details about our selection criteria are reported in Supplemental Figure 1 in the online version. Our final cohort consisted of 825,707 PCaassessable cases treated at 1249 CoC facilities.
Patients and Methods Data Source The NCDB is a nationwide hospital-based quality improvement initiative that contains comprehensive information on oncology treatments and patient-centered outcomes. Data on over 29 million newly diagnosed cancers from more than 1500 CoC accredited programs at US hospitals has been collected.14
2
-
Clinical Genitourinary Cancer Month 2017
Variables For each patient, the following parameters are recorded: age at diagnosis, race (white, black, and other), PSA at diagnosis, biopsy Gleason score ( 6, 7, and 8), cT stage (cT1-cT2, cT3-cT4, and cTx), clinical N stage (cN0/cNx, and cN1), year of diagnosis (20042013), education status, urban/rural status, insurance status, annual income, primary treatment type, geographical region of treatment, the National Comprehensive Cancer Network (NCCN) risk class, life expectancy, and the CoC facility type. Tumor, Node, Metastases staging was on the basis of the American Joint Committee on Cancer Staging Manual (sixth and seventh editions), as well as the Collaborative Staging system as previously described.16,17 Primary treatments were defined as radical prostatectomy (RP), brachytherapy (BT), external beam radiation therapy (EBRT) targeted to the prostate, cryotherapy (CT), or androgen deprivation monotherapy (ADT). All these were evaluated within 6 months from PCa diagnosis according to previously published methodology.18 Cases without primary treatment within 6 months were categorized as observation. NCCN recurrence risk was categorized as low (cT1-cT2a, Gleason score 6, PSA < 10 ng/mL), intermediate (cT2b-cT2c, Gleason score 7, PSA 10-20 ng/mL), and high ( cT3a, Gleason score 8, PSA > 20 ng/mL, cN1).6 On the basis of NCCN guidelines, life expectancy was calculated using actuarial life tables and CharlsoneDeyo Comorbidity Index (CCI).6,19-21 Briefly, to adjust for comorbidity, patients with a CCI 1 received a 50% distraction from the actuarial life expectancy. Those with a CCI of 0 did not receive any modification of their life expectancy. Life expectancy was categorized as 10 years and > 10 years. The CCI was categorized into 0, 1, and > 1 as specified in the NCDB.19,20,22 CoC facility type was categorized as community cancer program, comprehensive community cancer program, academic/research program, integrated network cancer program, and other. Facility type was on the basis of program structure, services provided, and number of cases treated each year (https:// www.facs.org/quality-programs/cancer/coc/apply/categories).
Statistical Analyses Frequencies and proportions were reported for categorical variables, whereas median and interquartile ranges were reported for continuous variables. Differences in categorical and continuous variables were examined using the c2 test and the ManneWhitney test, respectively. We examined treatment trends over time after stratifying patients according to the NCCN recurrence risk. In subgroup analyses, we further stratified patients on the basis of: (1) life expectancy: 10
Björn Löppenberg et al Table 1 Sociodemographic, Clinical, and Facility Characteristics for 825,707 Patients, Stratified According to LE; National Cancer Data Base 2004 to 2013 Overall (n [ 825,707)
£10-Year LE (n [ 140,722)
>10-Year LE (n [ 684,985)
<.0001
Age (Continuous) Median
65 (59-70)
76 (67-79)
63 (57-68) <.0001
Race/Ethnicity Non-Hispanic white
601,711 (72.9)
104,121 (74.6)
487,590 (72.6)
Non-Hispanic black
104,325 (12.6)
15,797 (11.2)
88,528 (12.9)
Hispanic
30,688 (3.7)
5373 (3.8)
25,315 (3.7)
Asian
11,517 (1.4)
2405 (1.7)
9112 (1.3)
1547 (0.2)
292 (0.2)
1255 (0.2)
American Indian/Alaskan Native Native Hawaiian/Pacific Islander
3465 (0.4)
721 (0.5)
2744 (0.4)
72,454 (8.8)
12,013 (8.5)
60,441 (8.8)
0
705,519 (85.5)
58,584 (41.6)
646,935 (94.5)
1
103,547 (12.5)
69,905 (49.7)
33,642 (4.9)
16,641 (2.0)
12,233 (8.7)
<$38,000
125,990 (15.3)
23,703 (16.8)
102,287 (14.9)
$38,000 to $47,999
180,381 (21.8)
33,820 (24.0)
146,561 (21.4)
$48,000 to $62,999
219,120 (26.5)
38,268 (27.2)
180,852 (26.4)
$63,000
291,853 (35.4)
43,466 (30.9)
248,387 (36.3)
Other
<.0001
Charlson-Deyo Score
>1
4408 (0.6) <.0001
Median Income According to ZIP Code
Unknown
8361 (1.0)
1465 (1.0)
6898 (1.0) <.0001
Percentage of ZIP Code Without High School Diploma 21% or more
117,483 (14.2)
21,574 (15.3)
95,909 (14.0)
13%-20.9%
193,573 (23.4)
34,581 (24.6)
158,992 (23.2)
7%-12.9%
271,690 (32.9)
47,533 (33.8)
224,157 (32.7)
<7%
235,079 (28.5)
35,646 (25.3)
199,433 (29.1)
Unknown
7882 (1.0)
1388 (1.0)
6494 (1.0) <.0001
County Status Metropolitan county
654,682 (79.3)
108,850 (77.4)
545,832 (79.7)
Urban county
128,094 (15.5)
24,154 (17.2)
103,940 (15.2)
Rural county
19,308 (2.3)
3866 (2.8)
15,442 (2.3)
Unknown
23,623 (2.9)
3852 (2.7)
19,771 (2.9)
409,376 (49.6)
28,460 (20.2)
380,916 (55.6)
<.0001
Insurance Status Private insurance Medicaid
17,501 (2.1)
Medicare
358,876 (43.5)
2162 (1.5) 105,559 (75.0)
15,339 (2.2) 253,317 (37.0)
Other government
13,919 (1.7)
1625 (1.2)
12,294 (1.8)
Not insured
13,352 (1.6)
1098 (0.8)
12,254 (1.8)
Unknown
12,683 (1.5)
1818 (1.3)
10,865 (1.6)
52,339 (6.3)
8795 (6.3)
43,544 (6.4)
<.0001
CoC Facility Region New England Middle Atlantic
125,444 (15.2)
20,454 (14.5)
104,990 (15.3)
South Atlantic
175,510 (21.3)
28,634 (20.4)
146,876 (21.4)
East North Central
146,710 (17.8)
27,774 (19.7)
118,936 (17.4)
East South Central
57,909 (7.0)
10,430 (7.4)
47,479 (6.9)
West North Central
73,437 (8.9)
13,154 (9.4)
60,283 (8.8)
West South Central
45,828 (5.6)
7762 (5.5)
38,066 (5.6)
Mountain
38,473 (4.7)
5824 (4.1)
32,649 (4.8)
109,573 (13.3)
17,895 (12.7)
91,678 (13.4)
Pacific
Pa
Clinical Genitourinary Cancer Month 2017
-3
Locoregional Prostate Cancer Care and Treatment Trends Table 1 Continued Overall (n [ 825,707)
£10-Year LE (n [ 140,722)
>10-Year LE (n [ 684,985)
<.0001
CoC Facility Type CCP
68,546 (8.3)
15,947 (11.3)
52,599 (7.7)
CCCP
376,452 (45.6)
71,479 (50.8)
304,973 (44.5)
ACAD
324,982 (39.4)
44,146 (31.4)
280,836 (41.0)
INCP
54,849 (6.6)
9094 (6.5)
45,755 (6.7)
Other
394 (0.1)
56 (0.04)
338 (0.1) <.0001
Biopsy Gleason Score 6
359,516 (43.5)
47,454 (33.7)
312,062 (45.6)
7
350,313 (42.5)
61,130 (43.4)
289,183 (42.2)
8-10
115,878 (14.0)
32,138 (22.8)
83,740 (12.2) <.0001
PSA (ng/mL) Median
6.0 (4.5-9.2)
7.1 (5.0-11.5)
5.8 (4.4-8.7) <.0001
PSA (Categorized), ng/mL <10.0
643,002 (77.9)
97,038 (69.0)
545,964 (79.7)
10.0-20.0
110,646 (13.4)
26,803 (19.1)
83,843 (12.2)
72,059 (8.7)
16,881 (12.0)
55,178 (8.1)
CTX
10,253 (1.2)
1735 (1.2)
8518 (1.2)
cT1
>20.0
<.0001
Clinical T Stage 2263 (0.3)
385 (0.3)
1878 (0.3)
cT1a
1584 (0.2)
195 (0.1)
1389 (0.2)
cT1b
1187 (0.1)
156 (0.1)
cT1c cT2
565,021 (68.4)
92,040 (65.4)
1031 (0.2) 472,981 (69.1)
16,573 (2.0)
3708 (2.6)
12,865 (1.9)
cT2a
74,112 (9.0)
13,259 (9.6)
60,583 (8.8)
cT2b
29,502 (3.6)
6702 (4.8)
22,800 (3.3)
cT2c
87,093 (10.6)
15,302 (10.9)
71,791 (10.5)
9549 (1.2)
1852 (1.3)
7697 (1.1)
cT3a
14,412 (1.8)
2515 (1.8)
11,897 (1.7)
cT3b
11,571 (1.4)
1969 (1.4)
9602 (1.4)
2587 (0.3)
634 (0.5)
1953 (0.3)
cT3
cT4
Pa
<.0001
NCCN Recurrence Risk Low
269,469 (32.6)
32,536 (23.1)
236,933 (34.6)
Intermediate
372,195 (45.1)
63,491 (45.1)
308,704 (45.1)
High
184,043 (22.3)
44,695 (31.8)
139,348 (20.3)
2004-2005
147,628 (17.9)
25,095 (17.8)
122,533 (17.9)
2006-2007
176,836 (21.4)
31,194 (22.2)
145,642 (21.3)
2008-2009
176,978 (21.4)
29,389 (20.9)
147,589 (21.6)
2010-2011
177,363 (21.5)
29,959 (21.3)
147,404 (21.5)
2012-2013
146,902 (17.8)
25,085 (17.8)
121,817 (17.8)
<.0001
Year
Abbreviations: ACAD ¼ academic comprehensive cancer program; CCCP ¼ comprehensive community cancer program; CCP ¼ community cancer program; CoC ¼ Commission on Cancer; INCP ¼ integrated network cancer program; LE ¼ life expectancy; NCCN ¼ National Comprehensive Cancer Network; PSA ¼ prostate-specific antigen. a 2 c Test, 1-way analysis of variance, or KruskaleWallis test, as appropriate.
years versus > 10 years; and (2) CoC facility type. Significance in treatment variation over time was examined using the CochraneArmitage test for trend.23,24 Finally, we examined the institution-specific variation in initial treatment using intraclass correlation coefficients, which estimated the variation in treatment attributable to a single institution.9,25 The intraclass correlation
4
-
Clinical Genitourinary Cancer Month 2017
coefficient was estimated using a random effects hierarchical model for each initial treatment (RP, BT, EBRT, ADT, and observation) while adjusting for sociodemographic factors, facility factors (shown in Table 1), and NCCN recurrence risk. This analysis was conducted in the entire cohort, and then repeated including only institutions with > 30 cases as a sensitivity analysis.9
Björn Löppenberg et al Figure 1 Treatment Trends in 825,707 Patients Diagnosed With Prostate Cancer, Stratified According to National Comprehensive Cancer Network Recurrence Risk, as Drawn From the National Cancer Data Base, Between 2004 and 2013. aObservation Includes Active Surveillance as Well as Watchful Waiting
Intermediate
Low .
. .
.
.
. .
.
.
.
. .
.
.
.
.
. .
.
. .
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
Brachytherapy
External beam radiation therapy
All analyses were performed with SAS 9.4 (SAS Institute Inc, Cary, NC), with a P value < .05 considered significant. An institutional review board waiver was obtained before conducting this study.
Results Baseline Characteristics Table 1 summarizes the baseline characteristics of our cohort. Overall, 32.6% (n ¼ 269,469), 45.1% (n ¼ 372,195), and 22.3% (n ¼ 184,043) of patients had NCCN low-, intermediate-, and high-risk PCa, respectively. Most patients (82.9%; n ¼ 684,985) had a life expectancy > 10 years. Initial treatment consisted of RP in 44.4% (n ¼ 366,160), EBRT in 20.2% (n ¼ 166,933), observation in 16.2% (n ¼ 134,072), BT in 12.9% (n ¼ 106,635), ADT in 4.9% (n ¼ 41,030), and CT in 1.3% (10,877) of patients. Combination treatment ( 2 treatments within 6 months) occurred in 12.7% (n ¼ 105,223) of cases; details on treatment combination are provided in Supplemental Table 1 in the online version.
Treatment Patterns for Locoregional PCa Stratified According to Tumor Risk Over the study period, RP utilization increased in all risk categories, with the highest increase observed for the low-risk group (32.6%-43.4%). For this risk group, a slight decline in RP utilization was observed for the first time in 2012 to 2013 (43.4%) compared with 2010 to 2011 (47.1%). BT utilization decreased between 2004 and 2013 for all risk groups. For the same period, EBRT utilization decreased in the low-risk group, but remained virtually stable in the other risk groups. Utilization of ADT in the high-risk group remained stable over time (around 10%), whereas its utilization in the low-/intermediate-risk groups decreased.
.
.
.
.
.
Radical prostatectomy
.
.
.
.
. .
.
.
.
. .
.
.
.
.
.
High .
.
.
.
. .
.
. .
.
.
.
.
.
Androgen-deprivation therapy
Observationa
.
.
Cryotherapy
Utilization of observation for low-risk cases increased from 16.3% in 2004 to 2005 to 32.0% in 2012 to 2013, but remained virtually stable in the other risk groups. Utilization of CT was approximately 1.0% to 1.5% for all risk groups, and remained stable over time. More details about these trends are shown in Figure 1, and their statistical significance is tested in Supplemental Table 2 in the online version.
Treatment Patterns for Locoregional PCa Stratified According to Tumor Risk and Life Expectancy Regardless of tumor characteristics, RT (EBRT and BT) was the most frequently used treatment modality in men with life expectancy 10 years, whereas RP was the most frequently used treatment modality in men with life expectancy > 10 years (Supplemental Figures 2 and 3 in the online version). This held true for all study years except before 2008, when men with life expectancy > 10 years and low-risk PCa received RT more commonly than RP. In the low-risk group, RP utilization decreased by 4.1% between 2010 and 2011 and 2012 to 2013 in men with life expectancy > 10 years, and by 1.7% in those with life expectancy 10 years. The utilization for ADT was generally greater for men with life expectancy 10 years. The overall utilization of observation was greater in men with life expectancy 10 years across all risk groups. The increase in adoption of this treatment modality over time was more prominent in men with life expectancy > 10 years (16.2%) than those with life expectancy 10 years (13.9%).
Variation in Treatment Per CoC Facility Type Overall, 1294 CoC institutions were included, these consisted of 46.0% (n ¼ 595) comprehensive community cancer programs, 33.2% (n ¼ 429) community cancer programs, 17.9% (n ¼ 232)
Clinical Genitourinary Cancer Month 2017
-5
Locoregional Prostate Cancer Care and Treatment Trends Table 2 Distribution of Treatment Type at Commission on Cancer Designated Facilities for 825,223a Patients, Stratified According to NCCN Recurrence Risk; National Cancer Data Base 2004 to 2013 NCCN Recurrence Initial Risk Treatment All
Low
Intermediate
High
RP BT EBRT ADT OBc CT Total RP BT EBRT ADT OBc CT Total RP BT EBRT ADT OBc CT Total RP BT EBRT ADT OBc CT Total
Pb
Commission on Cancer Facility Typea CCP, n (%)
CCCP, n (%)
ACAD, n (%)
INCP, n (%)
Other, n (%)
Total, n
19,435 (28.4) 10,686 (15.6) 19,284 (28.1) 4997 (7.3) 12,741 (18.6) 1403 (2.1) 68,546 5516 (25.5) 4928 (22.7) 4454 (20.6) 1072 (5.0) 5252 (24.2) 452 (2.1) 21,674 9492 (32.4) 4187 (14.3) 8470 (29.0) 1816 (6.2) 4648 (15.9) 649 (2.2) 29,262 4427 (25.1) 1571 (8.9) 6360 (36.1) 2109 (12.0) 2841 (16.1) 302 (1.7) 17,610
156,242 (41.5) 61,086 (16.2) 82,247 (21.9) 19,479 (5.2) 51,747 (13.8) 5651 (1.5) 376,452 45,246 (37.1) 28,305 (23.2) 20,177 (16.5) 4827 (4.0) 21,864 (17.9) 1553 (1.3) 121,972 79,440 (47.1) 23,715 (14.1) 35,961 (21.3) 7110 (4.2) 19,771 (11.7) 2832 (1.7) 168,829 31,556 (36.8) 9066 (10.6) 26,109 (30.5) 7542 (8.8) 10,112 (11.8) 1266 (1.5) 85,651
164,108 (50.5) 27,167 (8.4) 54,903 (16.9) 13,978 (4.3) 61,972 (19.1) 2854 (0.9) 324,982 48,136 (44.5) 12,892 (11.9) 13,838 (12.8) 2778 (2.6) 29,510 (27.3) 911 (0.8) 108,065 86,003 (58.1) 10,433 (7.0) 24,054 (16.2) 4572 (3.1) 21,718 (14.7) 1344 (0.9) 148,124 29,969 (43.6) 3842 (5.6) 17,011 (24.7) 6628 (9.6) 10,744 (15.6) 599 (0.9) 68,793
25,702 (46.9) 7666 (14.0) 10,434 (19.0) 2566 (4.7) 7512 (13.7) 969 (1.8) 54,849 7215 (41.3) 3714 (21.3) 2462 (14.1) 595 (3.4) 3185 (18.3) 282 (1.6) 17,453 13,669 (53.4) 2915 (11.4) 4687 (18.3) 922 (3.6) 2921 (11.4) 505 (2.0) 25,619 4818 (40.9) 1037 (8.8) 3285 (27.9) 1049 (8.9) 1406 (11.9) 182 (1.6) 11,777
303 (76.9) 14 (3.6) 54 (13.7) 5 (1.3) 18 (4.6) 0 (0.0) 394 33 (60.0) 2 (3.6) 14 (25.5) 2 (3.6) 4 (7.3) 0 (0.0) 55 149 (83.2) 8 (4.5) 18 (10.1) 0 (0.0) 4 (2.2) 0 (0.0) 179 121 (75.6) 4 (2.5) 22 (13.8) 3 (1.9) 10 (6.3) 0 (0.0) 160
365,790 106,619 166,922 41,025 133,990 10,877 825,223 106,146 49,841 40,945 9274 59,815 3198 269,219 188,753 41,258 73,190 14,420 49,062 5330 372,013 70,891 15,520 52,787 17,331 25,113 2349 183,991
<.0001
<.0001
<.0001
<.0001
Abbreviations: ACAD ¼ academic comprehensive cancer program; ADT ¼ androgen deprivation therapy; BT ¼ brachytherapy; CCP ¼ community cancer program; CCCP ¼ comprehensive community cancer program; CT ¼ cryotherapy; EBRT ¼ external beam radiation therapy; INCP ¼ integrated network cancer program; NCCN ¼ National Comprehensive Cancer Network; OB ¼ observation; RP ¼ radical prostatectomy. a To prevent identification through researchers, facility type is not given for patients younger than 40 years (n ¼ 484; n ¼ 250 low-risk, 182 intermediate, and 52 high-risk, respectively). b 2 c Test. c Includes watchful waiting as well as active surveillance.
academic centers, 2.7% (n ¼ 35) integrated network cancer programs, and 0.2% (n ¼ 3) other facilities. Table 2 shows treatment utilization according to CoC facility type for low-, intermediate-, and high-risk tumors. Significant differences in treatment type were observed on the basis of tumor risk and CoC facility type (all P < .001). For all risk groups, rates of treatment according to facility type ranged from 28.4% to 76.9% for RP, 3.6% to 16.2% for BT, 13.7% to 28.1% for EBRT, 1.3% to 7.3% for ADT, 4.6% to 19.1% for observation, and 0% to 2.1% for CT. Figure 2 shows time trends and differences between CoC facility types for initial treatment type stratified according to risk group. Utilization of RP was lowest in community cancer programs regardless of tumor risk, whereas utilization of BT was lowest in academic facilities. Community cancer programs most frequently utilized EBRT, regardless of tumor risk. Academic centers and community cancer programs utilized observation more often for low-risk PCa than comprehensive community cancer programs or integrated network cancer programs.
6
-
Clinical Genitourinary Cancer Month 2017
Variation in Treatment at the Institutional Level After adjusting for sociodemographic and facility factors (Table 3), the highest proportions of treatment variation attributable to the single institution were observed for CT (59%; 95% confidence interval [CI], 0.45%-0.73%) and BT (46%; 95% CI, 38%-53%), whereas the lowest proportion of treatment variation was observed for ADT (14%; 95% CI, 12%-15%), and observation (15%; 95% CI, 14%-17%). The results were consistent in the sensitivity analysis, and in all NCCN risk groups (Table 3). Supplemental Figure 4 in the online version shows treatment proportion variation on the basis of a single institution.
Discussion To our knowledge, our report is the first to examine the primary treatment trends for PCa at a nationwide level, considering life expectancy, as well as to address treatment allocation and variation at CoC-designated centers on the basis of facility type and institution level in detail.
Björn Löppenberg et al Figure 2 Variation of Prostate Cancer Treatment and Treatment Trends at Commission on Cancer Designated Facilities for 824,829a,b Patients, as Drawn From the National Cancer Data Base, Between 2004 and 2013. aBecause of the Low Number of Cases (n [ 394) Other Facility Types (n [ 3; Veterans Affairs Cancer Programs, NCI-Designated Comprehensive Cancer Programs, Hospital-associated Cancer Programs, and Free Standing Cancer Programs) Are Omitted in This Figure. bTo Prevent Identification Through Researchers, Facility Type Is Not Given for Patients Younger Than 40 Years (n [ 484). cObservation Includes Watchful Waiting and Active Surveillance
Brachytherapy
External beam radiation
ADT
Observationc
Cryotherapy
High
Intermediate
Low
Radical prostatectomy
Abbreviations: ACAD ¼ academic comprehensive cancer program; ADT ¼ androgen deprivation therapy; CCP ¼ community cancer program; CCCP ¼ comprehensive community cancer program; INCP ¼ integrated network cancer program; NCCN ¼ National Comprehensive Cancer Network; NCI ¼ National Cancer Institute.
Our study has several notable findings. First, considering the overall treatment trends, there was a reduction in active treatment utilization (RP or RT) among patients with low-risk disease, which denotes a reduction in overtreatment. Conversely, for high-risk
PCa, there was a slight increase in the utilization of active treatment over time. In this context, most patients with limited life expectancy received RT, whereas many patients with extended life expectancy received RP. Despite these encouraging changes in
Table 3 Treatment Variation Attributable to Practice Site for 825,223a Patients Diagnosed With Prostate Cancer; NCDB 2004 to 2013 All Risk,
r (95% CI)
Low Risk,
r (95% CI)
Intermediate Risk, r (95% CI)
High Risk,
r (95% CI)
All 825,223 Cases RP
0.32 (0.28-0.36)
0.36 (0.31-0.41)
0.33 (0.28-0.37)
0.24 (0.21-0.27)
BT
0.46 (0.38-0.53)
0.42 (0.35-0.49)
0.43 (0.36-0.50)
0.39 (0.33-0.46)
EBRT
0.34 (0.30-0.39)
0.34 (0.29-0.39)
0.34 (0.30-0.39)
0.26 (0.22-0.29)
ADT
0.14 (0.12-0.15)
0.18 (0.15-0.20)
0.14 (0.12-0.16)
0.12 (0.11-0.14)
OBSb
0.15 (0.14-0.17)
0.18 (0.16-0.20)
0.15 (0.13-0.17)
0.12 (0.11-0.14)
CT
0.59 (0.45-0.73)
0.52 (0.40-0.64)
0.55 (0.42-0.67)
0.47 (0.37-0.58)
RP
0.32 (0.28-0.36)
0.36 (0.31-0.40)
0.32 (0.28-0.37)
0.24 (0.21-0.27)
BT
0.46 (0.38-0.53)
0.42 (0.35-0.49)
0.43 (0.36-0.50)
0.39 (0.33-0.46)
EBRT
0.34 (0.30-0.38)
0.34 (0.29-0.39)
0.34 (0.30-0.39)
0.25 (0.22-0.29)
ADT
0.14 (0.12-0.15)
0.18 (0.15-0.20)
0.14 (0.12-0.16)
0.12 (0.11-0.14)
OBSb
0.15 (0.13-0.17)
0.18 (0.16-0.20)
0.15 (0.13-0.17)
0.12 (0.11-0.13)
CT
0.59 (0.45-0.73)
0.52 (0.40-0.65)
0.55 (0.42-0.67)
0.47 (0.37-0.57)
Sensitivity Analysis Limited to Facilities With >30 Total Cases in NCDB (n [ 824,385)
Abbreviations: ADT ¼ androgen-deprivation therapy; BT ¼ brachytherapy; CT ¼ cryotherapy; EBRT ¼ external beam radiation therapy; NCDB ¼ National Cancer Data Base; OBS ¼ observation; RP ¼ radical prostatectomy. a To prevent identification through researchers, facility type is not given for patients younger than 40 years. b Observation includes active surveillance as well as watchful waiting.
Clinical Genitourinary Cancer Month 2017
-7
Locoregional Prostate Cancer Care and Treatment Trends
8
-
treatment trends, there is still a large space for improvement. For example, in the most contemporary year of the study (2012-2013), approximately 65% of men with low-risk PCa received active treatment. However, a non-negligible proportion (9.4%) of patients with high-risk PCa were treated with ADT monotherapy. This was especially true for patients with limited life expectancy. The first case might represent overtreatment, with many patients suffering from side effects and economic burden of active treatment, without significant cancer control benefit.26 The second case represents undertreatment, because ADT monotherapy has not been proven an effective treatment modality in patients with localized PCa.27,28 The scientific community and CoC efforts should focus on discouraging such practices, to improve the PCa quality of care on nationwide level. Our observations corroborate previous reports that showed a decreased utilization of active treatment and an increased utilization of observation for patients with low-risk PCa.8,10,29 This is likely because of changes in guidelines endorsing observation for low-risk disease.6,7 However, within the Carcinoma of Prostate Strategic Urologic Research Endeavor (CaPSURE) database, only 25% of men with low-risk disease and limited life expectancy (defined as patient age > 75 years),8 received active treatment. On the contrary, 60% and 55% to 70% of patients with low-risk disease in the NCDB (defined as patients with calculated life expectancy < 10 years) or Surveillance, Epidemiology, and End Results (SEER)-Medicare data (defined as patient aged 76 years) received active treatment, respectively.11 On the basis of these results, compared with CaPSURE, on the nationwide level, life expectancy seems to be considered less frequently when deciding about PCa treatment in older or comorbid men.5 Likewise, our report corroborates the increasing trends of active treatment in patients with high-risk disease. Specifically, in the most contemporary years, the utilization of active treatment in these patients was 76% in NCDB, approximately 70% in CaPSURE, and 82% in SEER-Medicare (high-risk disease defined as clinical T3 or higher).8,30 However, still a considerable proportion of patients received ADT as monotherapy, which is considered as an undertreatment.27,28 This is specifically true for older patients, and/or those with limited life expectancy. The ADT utilization in the most contemporary years was 16% in NCDB and approximately 67% in CaPSURE.8 It is noteworthy that the NCCN guidelines do not recommend against active treatment in patients diagnosed with high-risk PCa and limited life expectancy.6 Thus, the relatively high utilization of ADT as initial treatment in these patients is worrisome and indicates significant undertreatment of potentially fatal disease. This issue requires urgent attention, especially in light of the recent United States Preventive Services Task Force recommendations against PSA screening. These might significantly increase the frequency of high-risk PCa disease diagnosis in the future.31-33 Second, the detailed examination of treatment distribution on the basis of CoC facility type and single institution revealed these factors have a substantial effect on initial treatment type. This finding is regardless of patient, demographic, clinical, and socioeconomic characteristics. Although a certain amount of variation is inevitable, the magnitude of this phenomenon was unexpected. For example, treatment trends in community cancer programs differed significantly from treatment trends in other facility types. However, these programs treated almost a third of patients included in the NCDB.
Clinical Genitourinary Cancer Month 2017
Specifically, compared with other facility types, community cancer programs offered the lowest rates of RP, and the highest rates of observation. EBRT was more commonly offered as the initial treatment, regardless of tumor risk compared with other facility types. Also, despite unproven benefit,6,7,27,28,34 the utilization of ADT monotherapy in high-risk patients was the highest in community cancer programs. Differences in patient volume per year, availability of treatment options, and availability of subspecialties between the different facility types might contribute to the observed variation. For example, integrated network cancer programs have no minimum caseload per year, whereas academic centers and comprehensive community cancer programs must treat more than 500 patients per year. For all facility types referral for diagnostic or treatment services to another facility is allowed by the CoC. However, rather than referring patients to a center that offers a guideline-concordant treatment option, centers might choose to offer the locally available treatment options. To prevent identification, the NCDB does not provide detailed information on a given facility, thus hindering a more comprehensive analysis. A detailed overview of requirements set up by the CoC for a given facility type can be found on the American College of Surgeons home page (https://www.facs.org/qualityprograms/cancer/coc/apply/categories). Likewise, on an institution level, some institutions treated all (100%) PCa patients with RP, whereas other institutions offered this treatment modality to none (0%). Such trends are worrisome, and require attention. It is noteworthy that these trends were adjusted for NCCN recurrence risk and life expectancy. Speculatively, these variations might originate because of differences in expertise, available treatment modality, and financial incentives.3,35,36 These effects have been previously reported. For example, an increased use of intensity-modulated radiation therapy was reported for urologists who acquired ownership of intensitymodulated radiation therapy compared with those who did not.3 The same observation was made with integrated PCa centers, where self-referral might result in an overutilization of intensitymodulated radiation therapy.37 Incentives for the use of intensitymodulated radiation therapy are technical and professional fees that approach $50,000 per patient.37 A contrary effect was observed after changes in the reimbursement policy for ADT with the Medicare Modernization Act. The moderate reductions led to a reduction in the use of ADT, especially in patient populations with unclear benefits of ADT therapy.4 Variation in primary treatment on the institutional level was also observed in previous reports. For example, Cooperberg et al reported significant variation in initial treatment pattern across different clinical sites.9 The observed magnitude of variation was approximately comparable with the one observed within the NCDB. Likewise, Hoffman et al reported that initial treatment decision was largely influenced by treating physician.11 Although we were unable to address the effect of the individual physician, our report confirmed that treating institution has a significant effect on initial treatment type. Many factors can contribute to this variation; these include, but are not limited to insurance status, personal preferences, financial incentives, and uneven distribution of novel techniques.9 Importantly, race and socioeconomic factors such as insurance status and income bracket were associated with receipt of definitive therapy of localized PCa. Within the NCDB, Gray et al reported receipt of definitive therapy
Björn Löppenberg et al for localized PCa was lowest in black patients.38 Additionally, being uninsured or belonging to the lowest income bracket was associated with a higher likelihood of observation in their analysis. Whereas our report cannot address the exact underlying causes of the observed facility- and institution-specific variation, it documents for the first time its magnitude at a nationwide level. Our findings should encourage CoC policymakers and chairs to alleviate these differences and provide a more consistent treatment allocation to PCa patients. Our study is not devoid of limitations. First, our study is limited by the retrospective nature of the cohort. Many other factors that could have influenced treatment selection are probably not captured in the NCDB database. However, this bias is inherent to all observational studies, on the basis of administrative data. Second, a large proportion of patients (382,473 patients [31.7%]) with PCa were excluded from our cohort because of missing data. This might limit the generalizability of our results. Third, the overall comorbidity burden was low in our patient population. This is shared with other reports that relied on NCDB analyses.39 Because we relied on the reported CCI for life expectancy calculation, the low comorbidity might translate into an overestimation of life expectancy. This might restrict the generalizability of the results on the basis of this calculation. Fourth, the NCDB does not offer information on changes in treatment type. An unknown proportion of patients who received active treatment might have been treated according to an observational protocol previously at a non-CoC facility. Additionally, some patients treated with an observational protocol might now receive active therapy. Finally, we were unable to distinguish between watchful waiting and active surveillance. These treatment categories were combined in 1 group, namely observation. This limitation is shared with many previous reports.8,9,11,29
Conclusion Contemporary analyses of treatment patterns showed a decline in the utilization of active treatment for low-risk PCa, which implies a decrease in overtreatment. However, a significant proportion of patients with high-risk PCa are treated with ADT monotherapy, which represents an undertreatment. Interestingly, a considerable variation in initial treatment was attributable to the CoC facility type and the institution in which treatment was received. Clear reasons for this variation could not be identified and make investigation by the CoC mandatory.
Clinical Practice Points
Additionally, rates of radiation therapy decreased, whereas RP
rates increased. Concerning variation in care, we found significant differences in
treatment patterns between CoC center types. Most active surveillance for low-risk PCa was performed at academic institutions. The highest rates of EBRT for all risk groups were observed at community cancer programs. Additionally, on the single facility level, an even broader, significant difference in treatment patterns became evident. For example, RP rates ranged from 0% to 100% between different facilities for all PCa risk groups. Our report should encourage the CoC and other policy makers to advocate a more harmonized treatment for localized PCa according to contemporary guidelines.
Acknowledgments Björn Löppenberg is supported by the Heinrich-Hertz-Stiftung of the Ministry of Innovation, Higher Education, and Research of North-Rhine Westphalia, Düsseldorf, Germany. Quoc-Dien Trinh is supported by an unrestricted educational grant from the Vattikuti Urology Institute, a Clay Hamlin Young Investigator Award from the Prostate Cancer Foundation, and a Genentech BioOncology Career Development Award from the Conquer Cancer Foundation of the American Society of Clinical Oncology. The NCDB is a joint project of the CoC of the American College of Surgeons and the American Cancer Society. The CoC NCDB and the hospitals participating in the CoC NCDB are the source of the deidentified data used in this study; they have not verified and are not responsible for the statistical validity of the data analysis or the conclusions derived by the authors. Firas Abdollah and Patrick Karabon 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.
Disclosure Firas Abdollah is an advisor/consultant for GenomeDx Biosciences; Quoc-Dien Trinh has been a speaker for Intuitive Surgical Inc. The remaining authors have stated that they have no conflicts of interest.
Supplemental Data Supplemental tables and figures accompanying this article can be found in the online version at http://dx.doi.org/10.1016/j.clgc. 2017.04.014.
In many guidelines, active surveillance is endorsed as first-line
treatment for localized PCa. Encouraging reports exist, which indicate an increase of active
surveillance. These reports are limited to certain geographical areas of the United States or certain practice networks, however. Within these reports a wide variation in treatment patterns for localized PCa was evident. Nationwide reports that have assessed treatment pattern variation and trends are missing. Using the NCDB, we focused on CoC-accredited centers to evaluate treatment variation and trends for localized PCa. An increase in observation for localized PCa between 2004 and 2013 was confirmed.
References 1. American Cancer Society. Cancer Facts & Figures 2017. Atlanta: American Cancer Society; 2017. 2. Kramer KM, Bennett CL, Pickard AS, et al. Patient preferences in prostate cancer: a clinician’s guide to understanding health utilities. Clin Prostate Cancer 2005; 4:15-23. 3. Mitchell JM. Urologists’ use of intensity-modulated radiation therapy for prostate cancer. N Engl J Med 2013; 369:1629-37. 4. Shahinian VB, Kuo YF, Gilbert SM. Reimbursement policy and androgendeprivation therapy for prostate cancer. N Engl J Med 2010; 363:1822-32. 5. Daskivich TJ, Lai J, Dick AW, et al. Variation in treatment associated with life expectancy in a population-based cohort of men with early-stage prostate cancer. Cancer 2014; 120:3642-50. 6. Mohler JL, Armstrong AJ, Bahnson RR, et al. Prostate cancer, version 1.2016. J Natl Compr Canc Netw 2016; 14:19-30. 7. Heidenreich A, Bastian PJ, Bellmunt J, et al. EAU guidelines on prostate cancer. Part 1: screening, diagnosis, and local treatment with curative intent-update 2013. Eur Urol 2014; 65:124-37.
Clinical Genitourinary Cancer Month 2017
-9
Locoregional Prostate Cancer Care and Treatment Trends 8. Cooperberg MR, Carroll PR. Trends in management for patients with localized prostate cancer, 1990-2013. JAMA 2015; 314:80-2. 9. Cooperberg MR, Broering JM, Carroll PR. Time trends and local variation in primary treatment of localized prostate cancer. J Clin Oncol 2010; 28:1117-23. 10. Weiner AB, Patel SG, Etzioni R, Eggener SE. National trends in the management of low and intermediate risk prostate cancer in the United States. J Urol 2015; 193: 95-102. 11. Hoffman KE, Niu J, Shen Y, et al. Physician variation in management of low-risk prostate cancer: a population-based cohort study. JAMA Intern Med 2014; 174: 1450-9. 12. Cooperberg MR, Broering JM, Carroll PR. Risk assessment for prostate cancer metastasis and mortality at the time of diagnosis. J Natl Cancer Inst 2009; 101:87887. 13. Cooperberg MR, Broering JM, Litwin MS, et al. The contemporary management of prostate cancer in the United States: lessons from the cancer of the prostate strategic urologic research endeavor (CapSURE), a national disease registry. J Urol 2004; 171:1393-401. 14. Winchester DP, Stewart AK, Bura C, Jones RS. The National Cancer Data Base: a clinical surveillance and quality improvement tool. J Surg Oncol 2004; 85:1-3. 15. NIH. National Cancer Institute. Surveillance, Epidemiology, and End Results Program Young JJ, Roffers S, Ries L, Fritz A, Hurlbut A. SEER summary staging manual - 2000, Available at: http://seer.cancer.gov/tools/ssm. Accessed: January 26, 2016. 16. Greene FL, Balch CM, Fleming ID, April F, eds. AJCC Cancer Staging Manual. 6th ed. New York: Springer-Verlag New York; 2003. 17. Edge S, Byrd DR, Compton CC, Fritz AG, Greene FL, Trotti A, eds. AJCC Cancer Staging Manual. 7th ed. New York: Springer-Verlag New York; 2010. 18. Wong YN, Mitra N, Hudes G, et al. Survival associated with treatment vs observation of localized prostate cancer in elderly men. JAMA 2006; 296:2683-93. 19. Charlson ME, Pompei P, Ales KL, MacKenzie CR. A new method of classifying prognostic comorbidity in longitudinal studies: development and validation. J Chronic Dis 1987; 40:373-83. 20. Deyo RA, Cherkin DC, Ciol MA. Adapting a clinical comorbidity index for use with ICD-9-CM administrative databases. J Clin Epidemiol 1992; 45:613-9. 21. Social Security Administration. Actuarial life table, Available at: https://www.ssa. gov/OACT/STATS/table4c6.html. Accessed: January 26, 2016. 22. Charlson M, Szatrowski TP, Peterson J, Gold J. Validation of a combined comorbidity index. J Clin Epidemiol 1994; 47:1245-51. 23. Cochran WG. Some methods for strengthening the common c2 tests. Biometrics 1954; 10:417-51. 24. Armitage P. Tests for linear trends in proportions and frequencies. Biometrics 1955; 11:375-86.
10
-
Clinical Genitourinary Cancer Month 2017
25. Wong GY, Mason WM. The hierarchical logistic regression model for multilevel analysis. J Am Stat Assoc 1985; 80:513-24. 26. Loeb S, Bjurlin MA, Nicholson J, et al. Overdiagnosis and overtreatment of prostate cancer. Eur Urol 2014; 65:1046-55. 27. Potosky AL, Haque R, Cassidy-Bushrow AE, et al. Effectiveness of primary androgen-deprivation therapy for clinically localized prostate cancer. J Clin Oncol 2014; 32:1324-30. 28. Roberts CB, Albertsen PC, Shao YH, et al. Patterns and correlates of prostate cancer treatment in older men. Am J Med 2011; 124:235-43. 29. Ingimarsson JP, Celaya MO, Laviolette M, Rees JR, Hyams ES. Trends in initial management of prostate cancer in New Hampshire. Cancer Causes Control 2015; 26:923-9. 30. Nezolosky MD, Dinh KT, Muralidhar V, et al. Significant increase in prostatectomy and decrease in radiation for clinical T3 prostate cancer from 1998 to 2012. Urol Oncol 2016; 34:57, e15-22. 31. U.S. Preventive Services Task Force. Screening for prostate cancer: U.S. Preventive Services Task Force recommendation statement. Ann Intern Med 2008; 149:185-91. 32. Moyer VA, U.S. Preventive Services Task Force. Screening for prostate cancer: U. S. Preventive Services Task Force recommendation statement. Ann Intern Med 2012; 157:120-34. 33. Jemal A, Fedewa SA, Ma J, et al. Prostate cancer incidence and PSA testing patterns in relation to USPSTF screening recommendations. JAMA 2015; 314: 2054-61. 34. Bechis SK, Carroll PR, Cooperberg MR. Impact of age at diagnosis on prostate cancer treatment and survival. J Clin Oncol 2011; 29:235-41. 35. van Tol-Geerdink JJ, Willem Leer J, Weijerman PC, et al. Choice between prostatectomy and radiotherapy when men are eligible for both: a randomized controlled trial of usual care vs decision aid. BJU Int 2013; 111:564-73. 36. Han LC, Delpe S, Shah ND, et al. Perceptions of radiation oncologists and urologists on sources and type of evidence to inform prostate cancer treatment decisions. Int J Radiat Oncol Biol Phys 2014; 89:277-83. 37. Falit BP, Gross CP, Roberts KB. Integrated prostate cancer centers and overutilization of IMRT: a close look at fee-for-service medicine in radiation oncology. Int J Radiat Oncol Biol Phys 2010; 76:1285-8. 38. Gray PJ, Lin CC, Cooperberg MR, Jemal A, Efstathiou JA. Temporal trends and the impact of race, insurance, and socioeconomic status in the management of localized prostate cancer. Eur Urol 2017; 71:729-37. 39. Lin CC, Bruinooge SS, Kirkwood MK, et al. Association between geographic access to cancer care, insurance, and receipt of chemotherapy: geographic distribution of oncologists and travel distance. J Clin Oncol 2015; 33:3177-85.
Björn Löppenberg et al Supplemental Figure 1 Inclusion and Exclusion Criteria for 1,208,180 Patients With Prostate Cancer Within the National Cancer Data Base (NCDB), 2004 to 2013 Participant User File (PUF). aRadiotherapy Includes Brachytherapy as Well as External Beam Radiation Therapy. bObservation Includes Active Surveillance as Well as Watchful Waiting
Abbreviations: ADT ¼ androgen deprivation therapy; CP ¼ cryotherapy; ICD ¼ International Classification of Diseases; NCCN ¼ National Comprehensive Cancer Network; PCa ¼ prostate cancer; PSA ¼ prostate-specific antigen; RP ¼ radical prostatectomy; RT ¼ radiotherapy.
Clinical Genitourinary Cancer Month 2017
- 10.e1
Locoregional Prostate Cancer Care and Treatment Trends Supplemental Figure 2 Treatment Trends in 140,722 Patients With Life Experience £ 10 Years Diagnosed With Prostate Cancer, Stratified According to National Comprehensive Cancer Network Recurrence Risk, as Drawn From the National Cancer Data Base, Between 2004 and 2013. aObservation Includes Active Surveillance as Well as Watchful Waiting
Intermediate
Low 100% 90%
2.1
21.4
3
3.4
2.6
2.5
3.5
High 3.3
2.5
15.7
16.3
17.3
17.8
11.2
9
6.5
5.4
5.9
1.9
2.8
3
16.7
15.9
16.1
18.5
17.5
15
37.3
38.6
10.3
9.5
15.2
15.8
2.3
1.8
14.5
13.8
14.1
15.8
39.7
38
28.5 35.3
80% 9
8.5 5.8
3.6 2.9
60% 25.2
24.7
17.9 23.5
40%
32.6
31.6
15.6
12.2
12.3 27.7
26.2
30
31.4
30.4
17.2
50%
30%
3.7
15.2 20.9 26.1
70%
2.2
19.5
17.5
10.3
38.3
9
15.6 6.4
4.6
7.5
20% 31.9
10% 14.6
16.7
31.3
30.2 20.9
21.7
32.3
34.4
24.6
20
23
26
0% 2004-2005 2006-2007 2008-2009 2010-2011 2012-2013 Brachytherapy
Radical prostatectomy
2004-2005 2006-2007 2008-2009 2010-2011 2012-2013 External beam radiation therapy
2004-2005 2006-2007 2008-2009 2010-2011 2012-2013 Observationa
Androgen-deprivation therapy
Cryotherapy
Supplemental Figure 3 Treatment Trends in 684,985 Patients With Life Experience > 10 Years Diagnosed With Prostate Cancer, Stratified According to National Comprehensive Cancer Network Recurrence Risk, as Drawn From the National Cancer Data Base, Between 2004 and 2013. aObservation Includes Active Surveillance as Well as Watchful Waiting
Low 100% 90%
0.9 15.6
1.2 17.3
1.1
21.2
0.7
24.7
0.9
1.2
1.2
10.9
11.7
12.2
4.5
3.7
2.6
16.4
15.6
31.6
80%
5.3
18.3
1.5 16.2
0.8
1
1.1
1.1
0.8
0.6
13.4
14.5
12.5
12.7
13.6
13.3
13.4
2.6
2.4
8
7.5
6.7
7
7.3
18.3
18.4 25.8
25.9
25.9
25.5
24.7
8.6
7.8 5.4
10.6
8.5
6.9
12.7
44.2
48.6
42.1
46.6
40
1.2
4.2 2.8
70%
High
Intermediate 0.8
15.8
15.3
12.1
1.3 9.6 11.7
60%
15.5
11.8
50%
26.8
23.5
18.2
12.6
9.7
40% 30% 49.1
20% 35.2
37.9
41.3
45
50
54.4
58.8
55.9
56.2
10% 0% 2004-2005 2006-2007 2008-2009 2010-2011 2012-2013 Radical prostatectomy
10.e2
-
Brachytherapy
Clinical Genitourinary Cancer Month 2017
2004-2005 2006-2007 2008-2009 2010-2011 2012-2013 External beam radiation therapy
Androgen-deprivation therapy
2004-2005 2006-2007 2008-2009 2010-2011 2012-2013 Observationa
Cryotherapy
Björn Löppenberg et al Supplemental Figure 4 Unadjusted Treatment Variation in 1269 Practice Sites for 825,707 Patients Diagnosed With Prostate Cancer, as Drawn From the National Cancer Data Base, Between 2004 and 2013. Facilities Were Stratified According to Percentage of Radical Prostatectomy Utilization. aObservation Includes Active Surveillance as Well as Watchful Waiting
Supplemental Table 1 Initial Treatment Combinations in 825,707 Patients Within 6 Months of Diagnosis of Prostate Cancer as Drawn From the National Cancer Data Base, Between 2004 and 2013 Treatment Combination in 180 d After Diagnosis RP Only RT Only ADT Only RP Then RT RT Then RP RT Then ADT ADT Then RTa ADT Then RPb RP Then RT and ADT RT Then RP and ADT ADT Then RP and RTb Observationc CP Only CP Then RT ADT Then CPd CP Then RT and ADT ADT Then CP and RTd
n
%
352,965 183,020 41,030 4741 64 5796 84,673 7808 1831 2 308 134,072 9397 22 1451 1 6
39.92 22.17 4.97 0.57 0.01 0.70 10.25 0.95 0.22 <0.01 0.04 16.24 1.14 <0.01 0.18 <0.01 <0.01
Abbreviations: ADT ¼ androgen-deprivation therapy; CP ¼ cryotherapy; RP ¼ radical prostatectomy; RT ¼ radiotherapy (including brachytherapy and external beam radiation therapy). a Initial treatment assigned is RT. b Initial treatment assigned is RP. c Includes watchful waiting and active surveillance. d Initial treatment assignment is CP.
Clinical Genitourinary Cancer Month 2017
- 10.e3
Locoregional Prostate Cancer Care and Treatment Trends Supplemental Table 2 Trend Test Comparing Initial Treatment Between 2004 and 2005 With 2012 and 2013 for Overall Patient Collective and According to NCCN Recurrence Risk-Stratified Prostate Cancer Patients; National Cancer Data Base 2004 to 2013 n (%) Initial Treatment
Trend Test Z
P
2004-2005
2012-2013
RP
56,586 (38.33)
69,796 (47.51)
61.81
<.0001
CP
1701 (1.15)
1430 (0.97)
6.91
<.0001
BT
27,793 (18.83)
11,531 (7.85)
EBRT
31,390 (21.26)
29,318 (19.96)
10.33
<.0001
ADT
10,005 (6.78)
6112 (4.16)
41.24
<.0001
OBa
20,153 (13.65)
28,715 (19.55)
50.81
<.0001
16,625 (32.65)
18,351 (43.40)
51.08
<.0001
Overall (All NCCN Risk Categories)
106.0
<.0001
Low Risk RP CP BT
543 (1.07) 13,687 (26.88)
372 (0.88)
4.96
<.0001
4208 (9.95)
83.58
<.0001
EBRT
8814 (17.31)
5190 (12.28)
28.47
<.0001
ADT
2954 (5.80)
612 (1.45)
46.28
<.0001
OBa
8299 (16.30)
13,548 (32.04)
64.19
<.0001
28,812 (44.93)
35,444 (52.55)
29.01
<.0001
Intermediate Risk RP CP
766 (1.19)
719 (1.07)
2.53
.01
52.20
<.0001
BT
10,149 (15.83)
5386 (7.99)
EBRT
13,302 (20.74)
13,738 (20.37)
2.07
.04
ADT
3636 (5.67)
2001 (2.97)
29.66
<.0001
OBa
7463 (11.64)
10,162 (15.07)
20.93
<.0001
11,149 (34.22)
16,001 (43.05)
High Risk 28.16
<.0001
CP
392 (1.20)
339 (0.91)
5.53
<.0001
BT
3957 (12.15)
1937 (5.21)
37.30
<.0001
EBRT
9274 (28.47)
10,390 (27.95)
1.66
.10
ADT
3415 (10.48)
3499 (9.41)
6.77
<.0001
OBa
4391 (13.48)
5005 (13.46)
0.03
.98
RP
Abbreviations: ADT ¼ androgen-deprivation therapy; BT ¼ brachytherapy; CT ¼ cryotherapy; EBRT ¼ external beam radiation therapy; NCCN ¼ National Comprehensive Cancer Network; NCDB ¼ National Cancer Data Base; OB ¼ observation; RP ¼ radical prostatectomy. a Observation includes watchful waiting as well as active surveillance.
10.e4
-
Clinical Genitourinary Cancer Month 2017
Variation in Locoregional Prostate Cancer Care and Treatment Trends at Commission on Cancer Designated Facilities: A National Cancer Data Base Analysis 2004 to 2013 Björn Löppenberg,1,2 Akshay Sood,1 Deepansh Dalela,1 Patrick Karabon,1,3 Jesse D. Sammon,1,4 Malte W. Vetterlein,5 Joachim Noldus,2 James O. Peabody,1 Quoc-Dien Trinh,5 Mani Menon,1 Firas Abdollah1 Abstract Within the National Cancer Data Base we evaluated whether differences in treatment patterns for prostate cancer exist at Commission on Cancer facilities. Overall, 825,707 men were included in the retrospective analyses. We found substantial variation in treatment patterns between different facility types and individual institutions, as well. Our findings indicate a necessity for policy makers to harmonize prostate cancer treatment. Background: Contemporary treatment trends for prostate cancer show increased rates of active surveillance. However, nationwide applicability of these reports is limited. Additionally, the effect of Commission on Cancer facility type on prostate cancer treatment patterns is unknown. Patients and Methods: We used the National Cancer Data Base to identify men diagnosed with prostate cancer, between 2004 and 2013. Our cohort was stratified on the basis of the National Comprehensive Cancer Network prostate cancer risk classes. CochraneArmitage tests were used to evaluate temporal trends. Random effects hierarchical logit models were used to assess treatment variation at Commission on Cancer facility and institution level. Results: In 825,707 men, utilization of radiation therapy declined and utilization of radical prostatectomy increased for all prostate cancer risk groups between 2004 and 2013 (P < .0001). Observation for low-risk prostate cancer increased from 16.3% in 2004 to 2005 to 32.0% in 2012 to 2013 (P < .0001). Significant treatment variation was observed on the basis of Commission on Cancer facility type. Across all risk groups, the lowest rates of radical prostatectomy and highest rates of external beam radiation therapy were observed in community cancer programs. The highest rates of observation for low-risk disease were observed in academic centers. Treatment variation according to institution ranged from 14% (95% confidence interval, 0.12-0.15) for androgen deprivation therapy up to 59% (95% confidence interval, 0.45-0.73) for cryotherapy. Conclusion: The increased utilization of observation in low-risk prostate cancer is an encouraging finding, which appears to be mainly derived by a decrease in radiotherapy utilization in this risk group. Regardless of tumor characteristics, significant variations in treatment modality exist among different facility types and institutions. Clinical Genitourinary Cancer, Vol. -, No. -, 1-10 ª 2017 Elsevier Inc. All rights reserved. Keywords: Androgen deprivation therapy, Cryotherapy, Radiation therapy, Radical prostatectomy, Variation in care
B.L. and A.S. contributed equally to this work. 1 Center for Outcomes Research, Analytics and Evaluation, Vattikuti Urology Institute, Henry Ford Hospital, Detroit, MI 2 Department of Urology, Marien Hospital Herne, Ruhr-University Bochum, Herne, Germany 3 Department of Public Health Sciences, Henry Ford Hospital, Detroit, MI 4 Maine Medical Center, Division of Urology & Center for Outcomes Research and Evaluation, Portland, ME
1558-7673/$ - see frontmatter ª 2017 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.clgc.2017.04.014
5 Center for Surgery and Public Health, Division of Urology, Brigham and Women’s Hospital, Boston, MA
Submitted: Feb 25, 2017; Revised: Apr 5, 2017; Accepted: Apr 14, 2017 Address for correspondence: Firas Abdollah, MD, Vattikuti Urology Institute & VUI Center for Outcomes Research Analytics and Evaluation, Henry Ford Hospital, 2799 W Grand Blvd, Detroit, MI 48202-2689 E-mail contact: fi[email protected]
Clinical Genitourinary Cancer Month 2017
-1
Locoregional Prostate Cancer Care and Treatment Trends Introduction
Patient Selection
In 2017, approximately 161,360 men will be diagnosed with prostate cancer (PCa),1 with most having locoregional disease at presentation.1 The optimal first-line treatment for this tumor has been a subject of continuous debate, with several guideline revisions over the years. These became necessary with introduction of newer therapies and techniques such as the nerve-sparing prostatectomy, minimally invasive surgery, advanced radiotherapy (RT) techniques, and active surveillance. In the absence of randomized trials that have compared the different treatment modalities, at least in part, patient preferences,2 physician biases,2 and financial incentives3,4 have often guided treatment decisions. Nowadays, it is generally accepted that the treatment selection should be on the basis of tumor risk class and patients’ life expectancy.5-7 Treatment options vary from observation, offered to men with a limited life expectancy and/or low tumor risk class, to multimodal therapy, the latter reserved for men with high-risk disease, especially those with extended life expectancy.6 With changing recommendations—capturing current treatment patterns, their variability, and their conformance to guideline recommendations represents the key to a better understanding of the utilization of health resources. Indeed, several reports have tried to address this issue.8-13 All these reports have contributed significantly to our understanding of PCa treatment patterns, but they are limited in their generalizability for several reasons. These include a focus on a certain age groups, a certain tumor risk category, a certain national region, or Medicare-insured patients only.8-13 We thus designed a study to overcome these limitations, and investigate the current landscape of locoregional PCa care. To do so at a national level, we relied on the National Cancer Data Base (NCDB), which captures approximately 70% of all newly diagnosed tumors in the United States.14 The aim of our study was twofold: first, to assess the contemporary nationwide treatment patterns for locoregional PCa, stratified per PCa risk class and life expectancy; and second, to evaluate variation in treatment at institutional and Commission on Cancer (CoC) facility type level. We hypothesized: first, that at the macroscopic level the treatment patterns might conform to the current guideline recommendations; second, disparities might exist at the institution level, and finally patients across the United States might not receive equal care. Further, we expected to confirm the findings from recent regional reports on decreased local therapy for low risk PCa, at the national level. Our results will enable clinicians, urologists, oncologists, and radiation oncologists to compare their own institutions treatment patterns and trends with nationwide treatment patterns. Additionally, our results might be insightful for urologists, epidemiologists, and students, as well as to provide an overview of contemporary US PCa care.
Men 35 years with locoregional adenocarcinoma of the prostate (International Classification of diseases-O-3 code: C61.9),15 diagnosed between 2004 and 2013, were evaluated. Cases with information missing on clinical T stage (cT; Tx was considered as unknown, but not missing), Gleason score, prostate-specific antigen (PSA), and primary metastatic cases, were excluded. More details about our selection criteria are reported in Supplemental Figure 1 in the online version. Our final cohort consisted of 825,707 PCaassessable cases treated at 1249 CoC facilities.
Patients and Methods Data Source The NCDB is a nationwide hospital-based quality improvement initiative that contains comprehensive information on oncology treatments and patient-centered outcomes. Data on over 29 million newly diagnosed cancers from more than 1500 CoC accredited programs at US hospitals has been collected.14
2
-
Clinical Genitourinary Cancer Month 2017
Variables For each patient, the following parameters are recorded: age at diagnosis, race (white, black, and other), PSA at diagnosis, biopsy Gleason score ( 6, 7, and 8), cT stage (cT1-cT2, cT3-cT4, and cTx), clinical N stage (cN0/cNx, and cN1), year of diagnosis (20042013), education status, urban/rural status, insurance status, annual income, primary treatment type, geographical region of treatment, the National Comprehensive Cancer Network (NCCN) risk class, life expectancy, and the CoC facility type. Tumor, Node, Metastases staging was on the basis of the American Joint Committee on Cancer Staging Manual (sixth and seventh editions), as well as the Collaborative Staging system as previously described.16,17 Primary treatments were defined as radical prostatectomy (RP), brachytherapy (BT), external beam radiation therapy (EBRT) targeted to the prostate, cryotherapy (CT), or androgen deprivation monotherapy (ADT). All these were evaluated within 6 months from PCa diagnosis according to previously published methodology.18 Cases without primary treatment within 6 months were categorized as observation. NCCN recurrence risk was categorized as low (cT1-cT2a, Gleason score 6, PSA < 10 ng/mL), intermediate (cT2b-cT2c, Gleason score 7, PSA 10-20 ng/mL), and high ( cT3a, Gleason score 8, PSA > 20 ng/mL, cN1).6 On the basis of NCCN guidelines, life expectancy was calculated using actuarial life tables and CharlsoneDeyo Comorbidity Index (CCI).6,19-21 Briefly, to adjust for comorbidity, patients with a CCI 1 received a 50% distraction from the actuarial life expectancy. Those with a CCI of 0 did not receive any modification of their life expectancy. Life expectancy was categorized as 10 years and > 10 years. The CCI was categorized into 0, 1, and > 1 as specified in the NCDB.19,20,22 CoC facility type was categorized as community cancer program, comprehensive community cancer program, academic/research program, integrated network cancer program, and other. Facility type was on the basis of program structure, services provided, and number of cases treated each year (https:// www.facs.org/quality-programs/cancer/coc/apply/categories).
Statistical Analyses Frequencies and proportions were reported for categorical variables, whereas median and interquartile ranges were reported for continuous variables. Differences in categorical and continuous variables were examined using the c2 test and the ManneWhitney test, respectively. We examined treatment trends over time after stratifying patients according to the NCCN recurrence risk. In subgroup analyses, we further stratified patients on the basis of: (1) life expectancy: 10
Björn Löppenberg et al Table 1 Sociodemographic, Clinical, and Facility Characteristics for 825,707 Patients, Stratified According to LE; National Cancer Data Base 2004 to 2013 Overall (n [ 825,707)
£10-Year LE (n [ 140,722)
>10-Year LE (n [ 684,985)
<.0001
Age (Continuous) Median
65 (59-70)
76 (67-79)
63 (57-68) <.0001
Race/Ethnicity Non-Hispanic white
601,711 (72.9)
104,121 (74.6)
487,590 (72.6)
Non-Hispanic black
104,325 (12.6)
15,797 (11.2)
88,528 (12.9)
Hispanic
30,688 (3.7)
5373 (3.8)
25,315 (3.7)
Asian
11,517 (1.4)
2405 (1.7)
9112 (1.3)
1547 (0.2)
292 (0.2)
1255 (0.2)
American Indian/Alaskan Native Native Hawaiian/Pacific Islander
3465 (0.4)
721 (0.5)
2744 (0.4)
72,454 (8.8)
12,013 (8.5)
60,441 (8.8)
0
705,519 (85.5)
58,584 (41.6)
646,935 (94.5)
1
103,547 (12.5)
69,905 (49.7)
33,642 (4.9)
16,641 (2.0)
12,233 (8.7)
<$38,000
125,990 (15.3)
23,703 (16.8)
102,287 (14.9)
$38,000 to $47,999
180,381 (21.8)
33,820 (24.0)
146,561 (21.4)
$48,000 to $62,999
219,120 (26.5)
38,268 (27.2)
180,852 (26.4)
$63,000
291,853 (35.4)
43,466 (30.9)
248,387 (36.3)
Other
<.0001
Charlson-Deyo Score
>1
4408 (0.6) <.0001
Median Income According to ZIP Code
Unknown
8361 (1.0)
1465 (1.0)
6898 (1.0) <.0001
Percentage of ZIP Code Without High School Diploma 21% or more
117,483 (14.2)
21,574 (15.3)
95,909 (14.0)
13%-20.9%
193,573 (23.4)
34,581 (24.6)
158,992 (23.2)
7%-12.9%
271,690 (32.9)
47,533 (33.8)
224,157 (32.7)
<7%
235,079 (28.5)
35,646 (25.3)
199,433 (29.1)
Unknown
7882 (1.0)
1388 (1.0)
6494 (1.0) <.0001
County Status Metropolitan county
654,682 (79.3)
108,850 (77.4)
545,832 (79.7)
Urban county
128,094 (15.5)
24,154 (17.2)
103,940 (15.2)
Rural county
19,308 (2.3)
3866 (2.8)
15,442 (2.3)
Unknown
23,623 (2.9)
3852 (2.7)
19,771 (2.9)
409,376 (49.6)
28,460 (20.2)
380,916 (55.6)
<.0001
Insurance Status Private insurance Medicaid
17,501 (2.1)
Medicare
358,876 (43.5)
2162 (1.5) 105,559 (75.0)
15,339 (2.2) 253,317 (37.0)
Other government
13,919 (1.7)
1625 (1.2)
12,294 (1.8)
Not insured
13,352 (1.6)
1098 (0.8)
12,254 (1.8)
Unknown
12,683 (1.5)
1818 (1.3)
10,865 (1.6)
52,339 (6.3)
8795 (6.3)
43,544 (6.4)
<.0001
CoC Facility Region New England Middle Atlantic
125,444 (15.2)
20,454 (14.5)
104,990 (15.3)
South Atlantic
175,510 (21.3)
28,634 (20.4)
146,876 (21.4)
East North Central
146,710 (17.8)
27,774 (19.7)
118,936 (17.4)
East South Central
57,909 (7.0)
10,430 (7.4)
47,479 (6.9)
West North Central
73,437 (8.9)
13,154 (9.4)
60,283 (8.8)
West South Central
45,828 (5.6)
7762 (5.5)
38,066 (5.6)
Mountain
38,473 (4.7)
5824 (4.1)
32,649 (4.8)
109,573 (13.3)
17,895 (12.7)
91,678 (13.4)
Pacific
Pa
Clinical Genitourinary Cancer Month 2017
-3
Locoregional Prostate Cancer Care and Treatment Trends Table 1 Continued Overall (n [ 825,707)
£10-Year LE (n [ 140,722)
>10-Year LE (n [ 684,985)
<.0001
CoC Facility Type CCP
68,546 (8.3)
15,947 (11.3)
52,599 (7.7)
CCCP
376,452 (45.6)
71,479 (50.8)
304,973 (44.5)
ACAD
324,982 (39.4)
44,146 (31.4)
280,836 (41.0)
INCP
54,849 (6.6)
9094 (6.5)
45,755 (6.7)
Other
394 (0.1)
56 (0.04)
338 (0.1) <.0001
Biopsy Gleason Score 6
359,516 (43.5)
47,454 (33.7)
312,062 (45.6)
7
350,313 (42.5)
61,130 (43.4)
289,183 (42.2)
8-10
115,878 (14.0)
32,138 (22.8)
83,740 (12.2) <.0001
PSA (ng/mL) Median
6.0 (4.5-9.2)
7.1 (5.0-11.5)
5.8 (4.4-8.7) <.0001
PSA (Categorized), ng/mL <10.0
643,002 (77.9)
97,038 (69.0)
545,964 (79.7)
10.0-20.0
110,646 (13.4)
26,803 (19.1)
83,843 (12.2)
72,059 (8.7)
16,881 (12.0)
55,178 (8.1)
CTX
10,253 (1.2)
1735 (1.2)
8518 (1.2)
cT1
>20.0
<.0001
Clinical T Stage 2263 (0.3)
385 (0.3)
1878 (0.3)
cT1a
1584 (0.2)
195 (0.1)
1389 (0.2)
cT1b
1187 (0.1)
156 (0.1)
cT1c cT2
565,021 (68.4)
92,040 (65.4)
1031 (0.2) 472,981 (69.1)
16,573 (2.0)
3708 (2.6)
12,865 (1.9)
cT2a
74,112 (9.0)
13,259 (9.6)
60,583 (8.8)
cT2b
29,502 (3.6)
6702 (4.8)
22,800 (3.3)
cT2c
87,093 (10.6)
15,302 (10.9)
71,791 (10.5)
9549 (1.2)
1852 (1.3)
7697 (1.1)
cT3a
14,412 (1.8)
2515 (1.8)
11,897 (1.7)
cT3b
11,571 (1.4)
1969 (1.4)
9602 (1.4)
2587 (0.3)
634 (0.5)
1953 (0.3)
cT3
cT4
Pa
<.0001
NCCN Recurrence Risk Low
269,469 (32.6)
32,536 (23.1)
236,933 (34.6)
Intermediate
372,195 (45.1)
63,491 (45.1)
308,704 (45.1)
High
184,043 (22.3)
44,695 (31.8)
139,348 (20.3)
2004-2005
147,628 (17.9)
25,095 (17.8)
122,533 (17.9)
2006-2007
176,836 (21.4)
31,194 (22.2)
145,642 (21.3)
2008-2009
176,978 (21.4)
29,389 (20.9)
147,589 (21.6)
2010-2011
177,363 (21.5)
29,959 (21.3)
147,404 (21.5)
2012-2013
146,902 (17.8)
25,085 (17.8)
121,817 (17.8)
<.0001
Year
Abbreviations: ACAD ¼ academic comprehensive cancer program; CCCP ¼ comprehensive community cancer program; CCP ¼ community cancer program; CoC ¼ Commission on Cancer; INCP ¼ integrated network cancer program; LE ¼ life expectancy; NCCN ¼ National Comprehensive Cancer Network; PSA ¼ prostate-specific antigen. a 2 c Test, 1-way analysis of variance, or KruskaleWallis test, as appropriate.
years versus > 10 years; and (2) CoC facility type. Significance in treatment variation over time was examined using the CochraneArmitage test for trend.23,24 Finally, we examined the institution-specific variation in initial treatment using intraclass correlation coefficients, which estimated the variation in treatment attributable to a single institution.9,25 The intraclass correlation
4
-
Clinical Genitourinary Cancer Month 2017
coefficient was estimated using a random effects hierarchical model for each initial treatment (RP, BT, EBRT, ADT, and observation) while adjusting for sociodemographic factors, facility factors (shown in Table 1), and NCCN recurrence risk. This analysis was conducted in the entire cohort, and then repeated including only institutions with > 30 cases as a sensitivity analysis.9
Björn Löppenberg et al Figure 1 Treatment Trends in 825,707 Patients Diagnosed With Prostate Cancer, Stratified According to National Comprehensive Cancer Network Recurrence Risk, as Drawn From the National Cancer Data Base, Between 2004 and 2013. aObservation Includes Active Surveillance as Well as Watchful Waiting
Intermediate
Low .
. .
.
.
. .
.
.
.
. .
.
.
.
.
. .
.
. .
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
Brachytherapy
External beam radiation therapy
All analyses were performed with SAS 9.4 (SAS Institute Inc, Cary, NC), with a P value < .05 considered significant. An institutional review board waiver was obtained before conducting this study.
Results Baseline Characteristics Table 1 summarizes the baseline characteristics of our cohort. Overall, 32.6% (n ¼ 269,469), 45.1% (n ¼ 372,195), and 22.3% (n ¼ 184,043) of patients had NCCN low-, intermediate-, and high-risk PCa, respectively. Most patients (82.9%; n ¼ 684,985) had a life expectancy > 10 years. Initial treatment consisted of RP in 44.4% (n ¼ 366,160), EBRT in 20.2% (n ¼ 166,933), observation in 16.2% (n ¼ 134,072), BT in 12.9% (n ¼ 106,635), ADT in 4.9% (n ¼ 41,030), and CT in 1.3% (10,877) of patients. Combination treatment ( 2 treatments within 6 months) occurred in 12.7% (n ¼ 105,223) of cases; details on treatment combination are provided in Supplemental Table 1 in the online version.
Treatment Patterns for Locoregional PCa Stratified According to Tumor Risk Over the study period, RP utilization increased in all risk categories, with the highest increase observed for the low-risk group (32.6%-43.4%). For this risk group, a slight decline in RP utilization was observed for the first time in 2012 to 2013 (43.4%) compared with 2010 to 2011 (47.1%). BT utilization decreased between 2004 and 2013 for all risk groups. For the same period, EBRT utilization decreased in the low-risk group, but remained virtually stable in the other risk groups. Utilization of ADT in the high-risk group remained stable over time (around 10%), whereas its utilization in the low-/intermediate-risk groups decreased.
.
.
.
.
.
Radical prostatectomy
.
.
.
.
. .
.
.
.
. .
.
.
.
.
.
High .
.
.
.
. .
.
. .
.
.
.
.
.
Androgen-deprivation therapy
Observationa
.
.
Cryotherapy
Utilization of observation for low-risk cases increased from 16.3% in 2004 to 2005 to 32.0% in 2012 to 2013, but remained virtually stable in the other risk groups. Utilization of CT was approximately 1.0% to 1.5% for all risk groups, and remained stable over time. More details about these trends are shown in Figure 1, and their statistical significance is tested in Supplemental Table 2 in the online version.
Treatment Patterns for Locoregional PCa Stratified According to Tumor Risk and Life Expectancy Regardless of tumor characteristics, RT (EBRT and BT) was the most frequently used treatment modality in men with life expectancy 10 years, whereas RP was the most frequently used treatment modality in men with life expectancy > 10 years (Supplemental Figures 2 and 3 in the online version). This held true for all study years except before 2008, when men with life expectancy > 10 years and low-risk PCa received RT more commonly than RP. In the low-risk group, RP utilization decreased by 4.1% between 2010 and 2011 and 2012 to 2013 in men with life expectancy > 10 years, and by 1.7% in those with life expectancy 10 years. The utilization for ADT was generally greater for men with life expectancy 10 years. The overall utilization of observation was greater in men with life expectancy 10 years across all risk groups. The increase in adoption of this treatment modality over time was more prominent in men with life expectancy > 10 years (16.2%) than those with life expectancy 10 years (13.9%).
Variation in Treatment Per CoC Facility Type Overall, 1294 CoC institutions were included, these consisted of 46.0% (n ¼ 595) comprehensive community cancer programs, 33.2% (n ¼ 429) community cancer programs, 17.9% (n ¼ 232)
Clinical Genitourinary Cancer Month 2017
-5
Locoregional Prostate Cancer Care and Treatment Trends Table 2 Distribution of Treatment Type at Commission on Cancer Designated Facilities for 825,223a Patients, Stratified According to NCCN Recurrence Risk; National Cancer Data Base 2004 to 2013 NCCN Recurrence Initial Risk Treatment All
Low
Intermediate
High
RP BT EBRT ADT OBc CT Total RP BT EBRT ADT OBc CT Total RP BT EBRT ADT OBc CT Total RP BT EBRT ADT OBc CT Total
Pb
Commission on Cancer Facility Typea CCP, n (%)
CCCP, n (%)
ACAD, n (%)
INCP, n (%)
Other, n (%)
Total, n
19,435 (28.4) 10,686 (15.6) 19,284 (28.1) 4997 (7.3) 12,741 (18.6) 1403 (2.1) 68,546 5516 (25.5) 4928 (22.7) 4454 (20.6) 1072 (5.0) 5252 (24.2) 452 (2.1) 21,674 9492 (32.4) 4187 (14.3) 8470 (29.0) 1816 (6.2) 4648 (15.9) 649 (2.2) 29,262 4427 (25.1) 1571 (8.9) 6360 (36.1) 2109 (12.0) 2841 (16.1) 302 (1.7) 17,610
156,242 (41.5) 61,086 (16.2) 82,247 (21.9) 19,479 (5.2) 51,747 (13.8) 5651 (1.5) 376,452 45,246 (37.1) 28,305 (23.2) 20,177 (16.5) 4827 (4.0) 21,864 (17.9) 1553 (1.3) 121,972 79,440 (47.1) 23,715 (14.1) 35,961 (21.3) 7110 (4.2) 19,771 (11.7) 2832 (1.7) 168,829 31,556 (36.8) 9066 (10.6) 26,109 (30.5) 7542 (8.8) 10,112 (11.8) 1266 (1.5) 85,651
164,108 (50.5) 27,167 (8.4) 54,903 (16.9) 13,978 (4.3) 61,972 (19.1) 2854 (0.9) 324,982 48,136 (44.5) 12,892 (11.9) 13,838 (12.8) 2778 (2.6) 29,510 (27.3) 911 (0.8) 108,065 86,003 (58.1) 10,433 (7.0) 24,054 (16.2) 4572 (3.1) 21,718 (14.7) 1344 (0.9) 148,124 29,969 (43.6) 3842 (5.6) 17,011 (24.7) 6628 (9.6) 10,744 (15.6) 599 (0.9) 68,793
25,702 (46.9) 7666 (14.0) 10,434 (19.0) 2566 (4.7) 7512 (13.7) 969 (1.8) 54,849 7215 (41.3) 3714 (21.3) 2462 (14.1) 595 (3.4) 3185 (18.3) 282 (1.6) 17,453 13,669 (53.4) 2915 (11.4) 4687 (18.3) 922 (3.6) 2921 (11.4) 505 (2.0) 25,619 4818 (40.9) 1037 (8.8) 3285 (27.9) 1049 (8.9) 1406 (11.9) 182 (1.6) 11,777
303 (76.9) 14 (3.6) 54 (13.7) 5 (1.3) 18 (4.6) 0 (0.0) 394 33 (60.0) 2 (3.6) 14 (25.5) 2 (3.6) 4 (7.3) 0 (0.0) 55 149 (83.2) 8 (4.5) 18 (10.1) 0 (0.0) 4 (2.2) 0 (0.0) 179 121 (75.6) 4 (2.5) 22 (13.8) 3 (1.9) 10 (6.3) 0 (0.0) 160
365,790 106,619 166,922 41,025 133,990 10,877 825,223 106,146 49,841 40,945 9274 59,815 3198 269,219 188,753 41,258 73,190 14,420 49,062 5330 372,013 70,891 15,520 52,787 17,331 25,113 2349 183,991
<.0001
<.0001
<.0001
<.0001
Abbreviations: ACAD ¼ academic comprehensive cancer program; ADT ¼ androgen deprivation therapy; BT ¼ brachytherapy; CCP ¼ community cancer program; CCCP ¼ comprehensive community cancer program; CT ¼ cryotherapy; EBRT ¼ external beam radiation therapy; INCP ¼ integrated network cancer program; NCCN ¼ National Comprehensive Cancer Network; OB ¼ observation; RP ¼ radical prostatectomy. a To prevent identification through researchers, facility type is not given for patients younger than 40 years (n ¼ 484; n ¼ 250 low-risk, 182 intermediate, and 52 high-risk, respectively). b 2 c Test. c Includes watchful waiting as well as active surveillance.
academic centers, 2.7% (n ¼ 35) integrated network cancer programs, and 0.2% (n ¼ 3) other facilities. Table 2 shows treatment utilization according to CoC facility type for low-, intermediate-, and high-risk tumors. Significant differences in treatment type were observed on the basis of tumor risk and CoC facility type (all P < .001). For all risk groups, rates of treatment according to facility type ranged from 28.4% to 76.9% for RP, 3.6% to 16.2% for BT, 13.7% to 28.1% for EBRT, 1.3% to 7.3% for ADT, 4.6% to 19.1% for observation, and 0% to 2.1% for CT. Figure 2 shows time trends and differences between CoC facility types for initial treatment type stratified according to risk group. Utilization of RP was lowest in community cancer programs regardless of tumor risk, whereas utilization of BT was lowest in academic facilities. Community cancer programs most frequently utilized EBRT, regardless of tumor risk. Academic centers and community cancer programs utilized observation more often for low-risk PCa than comprehensive community cancer programs or integrated network cancer programs.
6
-
Clinical Genitourinary Cancer Month 2017
Variation in Treatment at the Institutional Level After adjusting for sociodemographic and facility factors (Table 3), the highest proportions of treatment variation attributable to the single institution were observed for CT (59%; 95% confidence interval [CI], 0.45%-0.73%) and BT (46%; 95% CI, 38%-53%), whereas the lowest proportion of treatment variation was observed for ADT (14%; 95% CI, 12%-15%), and observation (15%; 95% CI, 14%-17%). The results were consistent in the sensitivity analysis, and in all NCCN risk groups (Table 3). Supplemental Figure 4 in the online version shows treatment proportion variation on the basis of a single institution.
Discussion To our knowledge, our report is the first to examine the primary treatment trends for PCa at a nationwide level, considering life expectancy, as well as to address treatment allocation and variation at CoC-designated centers on the basis of facility type and institution level in detail.
Björn Löppenberg et al Figure 2 Variation of Prostate Cancer Treatment and Treatment Trends at Commission on Cancer Designated Facilities for 824,829a,b Patients, as Drawn From the National Cancer Data Base, Between 2004 and 2013. aBecause of the Low Number of Cases (n [ 394) Other Facility Types (n [ 3; Veterans Affairs Cancer Programs, NCI-Designated Comprehensive Cancer Programs, Hospital-associated Cancer Programs, and Free Standing Cancer Programs) Are Omitted in This Figure. bTo Prevent Identification Through Researchers, Facility Type Is Not Given for Patients Younger Than 40 Years (n [ 484). cObservation Includes Watchful Waiting and Active Surveillance
Brachytherapy
External beam radiation
ADT
Observationc
Cryotherapy
High
Intermediate
Low
Radical prostatectomy
Abbreviations: ACAD ¼ academic comprehensive cancer program; ADT ¼ androgen deprivation therapy; CCP ¼ community cancer program; CCCP ¼ comprehensive community cancer program; INCP ¼ integrated network cancer program; NCCN ¼ National Comprehensive Cancer Network; NCI ¼ National Cancer Institute.
Our study has several notable findings. First, considering the overall treatment trends, there was a reduction in active treatment utilization (RP or RT) among patients with low-risk disease, which denotes a reduction in overtreatment. Conversely, for high-risk
PCa, there was a slight increase in the utilization of active treatment over time. In this context, most patients with limited life expectancy received RT, whereas many patients with extended life expectancy received RP. Despite these encouraging changes in
Table 3 Treatment Variation Attributable to Practice Site for 825,223a Patients Diagnosed With Prostate Cancer; NCDB 2004 to 2013 All Risk,
r (95% CI)
Low Risk,
r (95% CI)
Intermediate Risk, r (95% CI)
High Risk,
r (95% CI)
All 825,223 Cases RP
0.32 (0.28-0.36)
0.36 (0.31-0.41)
0.33 (0.28-0.37)
0.24 (0.21-0.27)
BT
0.46 (0.38-0.53)
0.42 (0.35-0.49)
0.43 (0.36-0.50)
0.39 (0.33-0.46)
EBRT
0.34 (0.30-0.39)
0.34 (0.29-0.39)
0.34 (0.30-0.39)
0.26 (0.22-0.29)
ADT
0.14 (0.12-0.15)
0.18 (0.15-0.20)
0.14 (0.12-0.16)
0.12 (0.11-0.14)
OBSb
0.15 (0.14-0.17)
0.18 (0.16-0.20)
0.15 (0.13-0.17)
0.12 (0.11-0.14)
CT
0.59 (0.45-0.73)
0.52 (0.40-0.64)
0.55 (0.42-0.67)
0.47 (0.37-0.58)
RP
0.32 (0.28-0.36)
0.36 (0.31-0.40)
0.32 (0.28-0.37)
0.24 (0.21-0.27)
BT
0.46 (0.38-0.53)
0.42 (0.35-0.49)
0.43 (0.36-0.50)
0.39 (0.33-0.46)
EBRT
0.34 (0.30-0.38)
0.34 (0.29-0.39)
0.34 (0.30-0.39)
0.25 (0.22-0.29)
ADT
0.14 (0.12-0.15)
0.18 (0.15-0.20)
0.14 (0.12-0.16)
0.12 (0.11-0.14)
OBSb
0.15 (0.13-0.17)
0.18 (0.16-0.20)
0.15 (0.13-0.17)
0.12 (0.11-0.13)
CT
0.59 (0.45-0.73)
0.52 (0.40-0.65)
0.55 (0.42-0.67)
0.47 (0.37-0.57)
Sensitivity Analysis Limited to Facilities With >30 Total Cases in NCDB (n [ 824,385)
Abbreviations: ADT ¼ androgen-deprivation therapy; BT ¼ brachytherapy; CT ¼ cryotherapy; EBRT ¼ external beam radiation therapy; NCDB ¼ National Cancer Data Base; OBS ¼ observation; RP ¼ radical prostatectomy. a To prevent identification through researchers, facility type is not given for patients younger than 40 years. b Observation includes active surveillance as well as watchful waiting.
Clinical Genitourinary Cancer Month 2017
-7
Locoregional Prostate Cancer Care and Treatment Trends
8
-
treatment trends, there is still a large space for improvement. For example, in the most contemporary year of the study (2012-2013), approximately 65% of men with low-risk PCa received active treatment. However, a non-negligible proportion (9.4%) of patients with high-risk PCa were treated with ADT monotherapy. This was especially true for patients with limited life expectancy. The first case might represent overtreatment, with many patients suffering from side effects and economic burden of active treatment, without significant cancer control benefit.26 The second case represents undertreatment, because ADT monotherapy has not been proven an effective treatment modality in patients with localized PCa.27,28 The scientific community and CoC efforts should focus on discouraging such practices, to improve the PCa quality of care on nationwide level. Our observations corroborate previous reports that showed a decreased utilization of active treatment and an increased utilization of observation for patients with low-risk PCa.8,10,29 This is likely because of changes in guidelines endorsing observation for low-risk disease.6,7 However, within the Carcinoma of Prostate Strategic Urologic Research Endeavor (CaPSURE) database, only 25% of men with low-risk disease and limited life expectancy (defined as patient age > 75 years),8 received active treatment. On the contrary, 60% and 55% to 70% of patients with low-risk disease in the NCDB (defined as patients with calculated life expectancy < 10 years) or Surveillance, Epidemiology, and End Results (SEER)-Medicare data (defined as patient aged 76 years) received active treatment, respectively.11 On the basis of these results, compared with CaPSURE, on the nationwide level, life expectancy seems to be considered less frequently when deciding about PCa treatment in older or comorbid men.5 Likewise, our report corroborates the increasing trends of active treatment in patients with high-risk disease. Specifically, in the most contemporary years, the utilization of active treatment in these patients was 76% in NCDB, approximately 70% in CaPSURE, and 82% in SEER-Medicare (high-risk disease defined as clinical T3 or higher).8,30 However, still a considerable proportion of patients received ADT as monotherapy, which is considered as an undertreatment.27,28 This is specifically true for older patients, and/or those with limited life expectancy. The ADT utilization in the most contemporary years was 16% in NCDB and approximately 67% in CaPSURE.8 It is noteworthy that the NCCN guidelines do not recommend against active treatment in patients diagnosed with high-risk PCa and limited life expectancy.6 Thus, the relatively high utilization of ADT as initial treatment in these patients is worrisome and indicates significant undertreatment of potentially fatal disease. This issue requires urgent attention, especially in light of the recent United States Preventive Services Task Force recommendations against PSA screening. These might significantly increase the frequency of high-risk PCa disease diagnosis in the future.31-33 Second, the detailed examination of treatment distribution on the basis of CoC facility type and single institution revealed these factors have a substantial effect on initial treatment type. This finding is regardless of patient, demographic, clinical, and socioeconomic characteristics. Although a certain amount of variation is inevitable, the magnitude of this phenomenon was unexpected. For example, treatment trends in community cancer programs differed significantly from treatment trends in other facility types. However, these programs treated almost a third of patients included in the NCDB.
Clinical Genitourinary Cancer Month 2017
Specifically, compared with other facility types, community cancer programs offered the lowest rates of RP, and the highest rates of observation. EBRT was more commonly offered as the initial treatment, regardless of tumor risk compared with other facility types. Also, despite unproven benefit,6,7,27,28,34 the utilization of ADT monotherapy in high-risk patients was the highest in community cancer programs. Differences in patient volume per year, availability of treatment options, and availability of subspecialties between the different facility types might contribute to the observed variation. For example, integrated network cancer programs have no minimum caseload per year, whereas academic centers and comprehensive community cancer programs must treat more than 500 patients per year. For all facility types referral for diagnostic or treatment services to another facility is allowed by the CoC. However, rather than referring patients to a center that offers a guideline-concordant treatment option, centers might choose to offer the locally available treatment options. To prevent identification, the NCDB does not provide detailed information on a given facility, thus hindering a more comprehensive analysis. A detailed overview of requirements set up by the CoC for a given facility type can be found on the American College of Surgeons home page (https://www.facs.org/qualityprograms/cancer/coc/apply/categories). Likewise, on an institution level, some institutions treated all (100%) PCa patients with RP, whereas other institutions offered this treatment modality to none (0%). Such trends are worrisome, and require attention. It is noteworthy that these trends were adjusted for NCCN recurrence risk and life expectancy. Speculatively, these variations might originate because of differences in expertise, available treatment modality, and financial incentives.3,35,36 These effects have been previously reported. For example, an increased use of intensity-modulated radiation therapy was reported for urologists who acquired ownership of intensitymodulated radiation therapy compared with those who did not.3 The same observation was made with integrated PCa centers, where self-referral might result in an overutilization of intensitymodulated radiation therapy.37 Incentives for the use of intensitymodulated radiation therapy are technical and professional fees that approach $50,000 per patient.37 A contrary effect was observed after changes in the reimbursement policy for ADT with the Medicare Modernization Act. The moderate reductions led to a reduction in the use of ADT, especially in patient populations with unclear benefits of ADT therapy.4 Variation in primary treatment on the institutional level was also observed in previous reports. For example, Cooperberg et al reported significant variation in initial treatment pattern across different clinical sites.9 The observed magnitude of variation was approximately comparable with the one observed within the NCDB. Likewise, Hoffman et al reported that initial treatment decision was largely influenced by treating physician.11 Although we were unable to address the effect of the individual physician, our report confirmed that treating institution has a significant effect on initial treatment type. Many factors can contribute to this variation; these include, but are not limited to insurance status, personal preferences, financial incentives, and uneven distribution of novel techniques.9 Importantly, race and socioeconomic factors such as insurance status and income bracket were associated with receipt of definitive therapy of localized PCa. Within the NCDB, Gray et al reported receipt of definitive therapy
Björn Löppenberg et al for localized PCa was lowest in black patients.38 Additionally, being uninsured or belonging to the lowest income bracket was associated with a higher likelihood of observation in their analysis. Whereas our report cannot address the exact underlying causes of the observed facility- and institution-specific variation, it documents for the first time its magnitude at a nationwide level. Our findings should encourage CoC policymakers and chairs to alleviate these differences and provide a more consistent treatment allocation to PCa patients. Our study is not devoid of limitations. First, our study is limited by the retrospective nature of the cohort. Many other factors that could have influenced treatment selection are probably not captured in the NCDB database. However, this bias is inherent to all observational studies, on the basis of administrative data. Second, a large proportion of patients (382,473 patients [31.7%]) with PCa were excluded from our cohort because of missing data. This might limit the generalizability of our results. Third, the overall comorbidity burden was low in our patient population. This is shared with other reports that relied on NCDB analyses.39 Because we relied on the reported CCI for life expectancy calculation, the low comorbidity might translate into an overestimation of life expectancy. This might restrict the generalizability of the results on the basis of this calculation. Fourth, the NCDB does not offer information on changes in treatment type. An unknown proportion of patients who received active treatment might have been treated according to an observational protocol previously at a non-CoC facility. Additionally, some patients treated with an observational protocol might now receive active therapy. Finally, we were unable to distinguish between watchful waiting and active surveillance. These treatment categories were combined in 1 group, namely observation. This limitation is shared with many previous reports.8,9,11,29
Conclusion Contemporary analyses of treatment patterns showed a decline in the utilization of active treatment for low-risk PCa, which implies a decrease in overtreatment. However, a significant proportion of patients with high-risk PCa are treated with ADT monotherapy, which represents an undertreatment. Interestingly, a considerable variation in initial treatment was attributable to the CoC facility type and the institution in which treatment was received. Clear reasons for this variation could not be identified and make investigation by the CoC mandatory.
Clinical Practice Points
Additionally, rates of radiation therapy decreased, whereas RP
rates increased. Concerning variation in care, we found significant differences in
treatment patterns between CoC center types. Most active surveillance for low-risk PCa was performed at academic institutions. The highest rates of EBRT for all risk groups were observed at community cancer programs. Additionally, on the single facility level, an even broader, significant difference in treatment patterns became evident. For example, RP rates ranged from 0% to 100% between different facilities for all PCa risk groups. Our report should encourage the CoC and other policy makers to advocate a more harmonized treatment for localized PCa according to contemporary guidelines.
Acknowledgments Björn Löppenberg is supported by the Heinrich-Hertz-Stiftung of the Ministry of Innovation, Higher Education, and Research of North-Rhine Westphalia, Düsseldorf, Germany. Quoc-Dien Trinh is supported by an unrestricted educational grant from the Vattikuti Urology Institute, a Clay Hamlin Young Investigator Award from the Prostate Cancer Foundation, and a Genentech BioOncology Career Development Award from the Conquer Cancer Foundation of the American Society of Clinical Oncology. The NCDB is a joint project of the CoC of the American College of Surgeons and the American Cancer Society. The CoC NCDB and the hospitals participating in the CoC NCDB are the source of the deidentified data used in this study; they have not verified and are not responsible for the statistical validity of the data analysis or the conclusions derived by the authors. Firas Abdollah and Patrick Karabon 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.
Disclosure Firas Abdollah is an advisor/consultant for GenomeDx Biosciences; Quoc-Dien Trinh has been a speaker for Intuitive Surgical Inc. The remaining authors have stated that they have no conflicts of interest.
Supplemental Data Supplemental tables and figures accompanying this article can be found in the online version at http://dx.doi.org/10.1016/j.clgc. 2017.04.014.
In many guidelines, active surveillance is endorsed as first-line
treatment for localized PCa. Encouraging reports exist, which indicate an increase of active
surveillance. These reports are limited to certain geographical areas of the United States or certain practice networks, however. Within these reports a wide variation in treatment patterns for localized PCa was evident. Nationwide reports that have assessed treatment pattern variation and trends are missing. Using the NCDB, we focused on CoC-accredited centers to evaluate treatment variation and trends for localized PCa. An increase in observation for localized PCa between 2004 and 2013 was confirmed.
References 1. American Cancer Society. Cancer Facts & Figures 2017. Atlanta: American Cancer Society; 2017. 2. Kramer KM, Bennett CL, Pickard AS, et al. Patient preferences in prostate cancer: a clinician’s guide to understanding health utilities. Clin Prostate Cancer 2005; 4:15-23. 3. Mitchell JM. Urologists’ use of intensity-modulated radiation therapy for prostate cancer. N Engl J Med 2013; 369:1629-37. 4. Shahinian VB, Kuo YF, Gilbert SM. Reimbursement policy and androgendeprivation therapy for prostate cancer. N Engl J Med 2010; 363:1822-32. 5. Daskivich TJ, Lai J, Dick AW, et al. Variation in treatment associated with life expectancy in a population-based cohort of men with early-stage prostate cancer. Cancer 2014; 120:3642-50. 6. Mohler JL, Armstrong AJ, Bahnson RR, et al. Prostate cancer, version 1.2016. J Natl Compr Canc Netw 2016; 14:19-30. 7. Heidenreich A, Bastian PJ, Bellmunt J, et al. EAU guidelines on prostate cancer. Part 1: screening, diagnosis, and local treatment with curative intent-update 2013. Eur Urol 2014; 65:124-37.
Clinical Genitourinary Cancer Month 2017
-9
Locoregional Prostate Cancer Care and Treatment Trends 8. Cooperberg MR, Carroll PR. Trends in management for patients with localized prostate cancer, 1990-2013. JAMA 2015; 314:80-2. 9. Cooperberg MR, Broering JM, Carroll PR. Time trends and local variation in primary treatment of localized prostate cancer. J Clin Oncol 2010; 28:1117-23. 10. Weiner AB, Patel SG, Etzioni R, Eggener SE. National trends in the management of low and intermediate risk prostate cancer in the United States. J Urol 2015; 193: 95-102. 11. Hoffman KE, Niu J, Shen Y, et al. Physician variation in management of low-risk prostate cancer: a population-based cohort study. JAMA Intern Med 2014; 174: 1450-9. 12. Cooperberg MR, Broering JM, Carroll PR. Risk assessment for prostate cancer metastasis and mortality at the time of diagnosis. J Natl Cancer Inst 2009; 101:87887. 13. Cooperberg MR, Broering JM, Litwin MS, et al. The contemporary management of prostate cancer in the United States: lessons from the cancer of the prostate strategic urologic research endeavor (CapSURE), a national disease registry. J Urol 2004; 171:1393-401. 14. Winchester DP, Stewart AK, Bura C, Jones RS. The National Cancer Data Base: a clinical surveillance and quality improvement tool. J Surg Oncol 2004; 85:1-3. 15. NIH. National Cancer Institute. Surveillance, Epidemiology, and End Results Program Young JJ, Roffers S, Ries L, Fritz A, Hurlbut A. SEER summary staging manual - 2000, Available at: http://seer.cancer.gov/tools/ssm. Accessed: January 26, 2016. 16. Greene FL, Balch CM, Fleming ID, April F, eds. AJCC Cancer Staging Manual. 6th ed. New York: Springer-Verlag New York; 2003. 17. Edge S, Byrd DR, Compton CC, Fritz AG, Greene FL, Trotti A, eds. AJCC Cancer Staging Manual. 7th ed. New York: Springer-Verlag New York; 2010. 18. Wong YN, Mitra N, Hudes G, et al. Survival associated with treatment vs observation of localized prostate cancer in elderly men. JAMA 2006; 296:2683-93. 19. Charlson ME, Pompei P, Ales KL, MacKenzie CR. A new method of classifying prognostic comorbidity in longitudinal studies: development and validation. J Chronic Dis 1987; 40:373-83. 20. Deyo RA, Cherkin DC, Ciol MA. Adapting a clinical comorbidity index for use with ICD-9-CM administrative databases. J Clin Epidemiol 1992; 45:613-9. 21. Social Security Administration. Actuarial life table, Available at: https://www.ssa. gov/OACT/STATS/table4c6.html. Accessed: January 26, 2016. 22. Charlson M, Szatrowski TP, Peterson J, Gold J. Validation of a combined comorbidity index. J Clin Epidemiol 1994; 47:1245-51. 23. Cochran WG. Some methods for strengthening the common c2 tests. Biometrics 1954; 10:417-51. 24. Armitage P. Tests for linear trends in proportions and frequencies. Biometrics 1955; 11:375-86.
10
-
Clinical Genitourinary Cancer Month 2017
25. Wong GY, Mason WM. The hierarchical logistic regression model for multilevel analysis. J Am Stat Assoc 1985; 80:513-24. 26. Loeb S, Bjurlin MA, Nicholson J, et al. Overdiagnosis and overtreatment of prostate cancer. Eur Urol 2014; 65:1046-55. 27. Potosky AL, Haque R, Cassidy-Bushrow AE, et al. Effectiveness of primary androgen-deprivation therapy for clinically localized prostate cancer. J Clin Oncol 2014; 32:1324-30. 28. Roberts CB, Albertsen PC, Shao YH, et al. Patterns and correlates of prostate cancer treatment in older men. Am J Med 2011; 124:235-43. 29. Ingimarsson JP, Celaya MO, Laviolette M, Rees JR, Hyams ES. Trends in initial management of prostate cancer in New Hampshire. Cancer Causes Control 2015; 26:923-9. 30. Nezolosky MD, Dinh KT, Muralidhar V, et al. Significant increase in prostatectomy and decrease in radiation for clinical T3 prostate cancer from 1998 to 2012. Urol Oncol 2016; 34:57, e15-22. 31. U.S. Preventive Services Task Force. Screening for prostate cancer: U.S. Preventive Services Task Force recommendation statement. Ann Intern Med 2008; 149:185-91. 32. Moyer VA, U.S. Preventive Services Task Force. Screening for prostate cancer: U. S. Preventive Services Task Force recommendation statement. Ann Intern Med 2012; 157:120-34. 33. Jemal A, Fedewa SA, Ma J, et al. Prostate cancer incidence and PSA testing patterns in relation to USPSTF screening recommendations. JAMA 2015; 314: 2054-61. 34. Bechis SK, Carroll PR, Cooperberg MR. Impact of age at diagnosis on prostate cancer treatment and survival. J Clin Oncol 2011; 29:235-41. 35. van Tol-Geerdink JJ, Willem Leer J, Weijerman PC, et al. Choice between prostatectomy and radiotherapy when men are eligible for both: a randomized controlled trial of usual care vs decision aid. BJU Int 2013; 111:564-73. 36. Han LC, Delpe S, Shah ND, et al. Perceptions of radiation oncologists and urologists on sources and type of evidence to inform prostate cancer treatment decisions. Int J Radiat Oncol Biol Phys 2014; 89:277-83. 37. Falit BP, Gross CP, Roberts KB. Integrated prostate cancer centers and overutilization of IMRT: a close look at fee-for-service medicine in radiation oncology. Int J Radiat Oncol Biol Phys 2010; 76:1285-8. 38. Gray PJ, Lin CC, Cooperberg MR, Jemal A, Efstathiou JA. Temporal trends and the impact of race, insurance, and socioeconomic status in the management of localized prostate cancer. Eur Urol 2017; 71:729-37. 39. Lin CC, Bruinooge SS, Kirkwood MK, et al. Association between geographic access to cancer care, insurance, and receipt of chemotherapy: geographic distribution of oncologists and travel distance. J Clin Oncol 2015; 33:3177-85.
Björn Löppenberg et al Supplemental Figure 1 Inclusion and Exclusion Criteria for 1,208,180 Patients With Prostate Cancer Within the National Cancer Data Base (NCDB), 2004 to 2013 Participant User File (PUF). aRadiotherapy Includes Brachytherapy as Well as External Beam Radiation Therapy. bObservation Includes Active Surveillance as Well as Watchful Waiting
Abbreviations: ADT ¼ androgen deprivation therapy; CP ¼ cryotherapy; ICD ¼ International Classification of Diseases; NCCN ¼ National Comprehensive Cancer Network; PCa ¼ prostate cancer; PSA ¼ prostate-specific antigen; RP ¼ radical prostatectomy; RT ¼ radiotherapy.
Clinical Genitourinary Cancer Month 2017
- 10.e1
Locoregional Prostate Cancer Care and Treatment Trends Supplemental Figure 2 Treatment Trends in 140,722 Patients With Life Experience £ 10 Years Diagnosed With Prostate Cancer, Stratified According to National Comprehensive Cancer Network Recurrence Risk, as Drawn From the National Cancer Data Base, Between 2004 and 2013. aObservation Includes Active Surveillance as Well as Watchful Waiting
Intermediate
Low 100% 90%
2.1
21.4
3
3.4
2.6
2.5
3.5
High 3.3
2.5
15.7
16.3
17.3
17.8
11.2
9
6.5
5.4
5.9
1.9
2.8
3
16.7
15.9
16.1
18.5
17.5
15
37.3
38.6
10.3
9.5
15.2
15.8
2.3
1.8
14.5
13.8
14.1
15.8
39.7
38
28.5 35.3
80% 9
8.5 5.8
3.6 2.9
60% 25.2
24.7
17.9 23.5
40%
32.6
31.6
15.6
12.2
12.3 27.7
26.2
30
31.4
30.4
17.2
50%
30%
3.7
15.2 20.9 26.1
70%
2.2
19.5
17.5
10.3
38.3
9
15.6 6.4
4.6
7.5
20% 31.9
10% 14.6
16.7
31.3
30.2 20.9
21.7
32.3
34.4
24.6
20
23
26
0% 2004-2005 2006-2007 2008-2009 2010-2011 2012-2013 Brachytherapy
Radical prostatectomy
2004-2005 2006-2007 2008-2009 2010-2011 2012-2013 External beam radiation therapy
2004-2005 2006-2007 2008-2009 2010-2011 2012-2013 Observationa
Androgen-deprivation therapy
Cryotherapy
Supplemental Figure 3 Treatment Trends in 684,985 Patients With Life Experience > 10 Years Diagnosed With Prostate Cancer, Stratified According to National Comprehensive Cancer Network Recurrence Risk, as Drawn From the National Cancer Data Base, Between 2004 and 2013. aObservation Includes Active Surveillance as Well as Watchful Waiting
Low 100% 90%
0.9 15.6
1.2 17.3
1.1
21.2
0.7
24.7
0.9
1.2
1.2
10.9
11.7
12.2
4.5
3.7
2.6
16.4
15.6
31.6
80%
5.3
18.3
1.5 16.2
0.8
1
1.1
1.1
0.8
0.6
13.4
14.5
12.5
12.7
13.6
13.3
13.4
2.6
2.4
8
7.5
6.7
7
7.3
18.3
18.4 25.8
25.9
25.9
25.5
24.7
8.6
7.8 5.4
10.6
8.5
6.9
12.7
44.2
48.6
42.1
46.6
40
1.2
4.2 2.8
70%
High
Intermediate 0.8
15.8
15.3
12.1
1.3 9.6 11.7
60%
15.5
11.8
50%
26.8
23.5
18.2
12.6
9.7
40% 30% 49.1
20% 35.2
37.9
41.3
45
50
54.4
58.8
55.9
56.2
10% 0% 2004-2005 2006-2007 2008-2009 2010-2011 2012-2013 Radical prostatectomy
10.e2
-
Brachytherapy
Clinical Genitourinary Cancer Month 2017
2004-2005 2006-2007 2008-2009 2010-2011 2012-2013 External beam radiation therapy
Androgen-deprivation therapy
2004-2005 2006-2007 2008-2009 2010-2011 2012-2013 Observationa
Cryotherapy
Björn Löppenberg et al Supplemental Figure 4 Unadjusted Treatment Variation in 1269 Practice Sites for 825,707 Patients Diagnosed With Prostate Cancer, as Drawn From the National Cancer Data Base, Between 2004 and 2013. Facilities Were Stratified According to Percentage of Radical Prostatectomy Utilization. aObservation Includes Active Surveillance as Well as Watchful Waiting
Supplemental Table 1 Initial Treatment Combinations in 825,707 Patients Within 6 Months of Diagnosis of Prostate Cancer as Drawn From the National Cancer Data Base, Between 2004 and 2013 Treatment Combination in 180 d After Diagnosis RP Only RT Only ADT Only RP Then RT RT Then RP RT Then ADT ADT Then RTa ADT Then RPb RP Then RT and ADT RT Then RP and ADT ADT Then RP and RTb Observationc CP Only CP Then RT ADT Then CPd CP Then RT and ADT ADT Then CP and RTd
n
%
352,965 183,020 41,030 4741 64 5796 84,673 7808 1831 2 308 134,072 9397 22 1451 1 6
39.92 22.17 4.97 0.57 0.01 0.70 10.25 0.95 0.22 <0.01 0.04 16.24 1.14 <0.01 0.18 <0.01 <0.01
Abbreviations: ADT ¼ androgen-deprivation therapy; CP ¼ cryotherapy; RP ¼ radical prostatectomy; RT ¼ radiotherapy (including brachytherapy and external beam radiation therapy). a Initial treatment assigned is RT. b Initial treatment assigned is RP. c Includes watchful waiting and active surveillance. d Initial treatment assignment is CP.
Clinical Genitourinary Cancer Month 2017
- 10.e3
Locoregional Prostate Cancer Care and Treatment Trends Supplemental Table 2 Trend Test Comparing Initial Treatment Between 2004 and 2005 With 2012 and 2013 for Overall Patient Collective and According to NCCN Recurrence Risk-Stratified Prostate Cancer Patients; National Cancer Data Base 2004 to 2013 n (%) Initial Treatment
Trend Test Z
P
2004-2005
2012-2013
RP
56,586 (38.33)
69,796 (47.51)
61.81
<.0001
CP
1701 (1.15)
1430 (0.97)
6.91
<.0001
BT
27,793 (18.83)
11,531 (7.85)
EBRT
31,390 (21.26)
29,318 (19.96)
10.33
<.0001
ADT
10,005 (6.78)
6112 (4.16)
41.24
<.0001
OBa
20,153 (13.65)
28,715 (19.55)
50.81
<.0001
16,625 (32.65)
18,351 (43.40)
51.08
<.0001
Overall (All NCCN Risk Categories)
106.0
<.0001
Low Risk RP CP BT
543 (1.07) 13,687 (26.88)
372 (0.88)
4.96
<.0001
4208 (9.95)
83.58
<.0001
EBRT
8814 (17.31)
5190 (12.28)
28.47
<.0001
ADT
2954 (5.80)
612 (1.45)
46.28
<.0001
OBa
8299 (16.30)
13,548 (32.04)
64.19
<.0001
28,812 (44.93)
35,444 (52.55)
29.01
<.0001
Intermediate Risk RP CP
766 (1.19)
719 (1.07)
2.53
.01
52.20
<.0001
BT
10,149 (15.83)
5386 (7.99)
EBRT
13,302 (20.74)
13,738 (20.37)
2.07
.04
ADT
3636 (5.67)
2001 (2.97)
29.66
<.0001
OBa
7463 (11.64)
10,162 (15.07)
20.93
<.0001
11,149 (34.22)
16,001 (43.05)
High Risk 28.16
<.0001
CP
392 (1.20)
339 (0.91)
5.53
<.0001
BT
3957 (12.15)
1937 (5.21)
37.30
<.0001
EBRT
9274 (28.47)
10,390 (27.95)
1.66
.10
ADT
3415 (10.48)
3499 (9.41)
6.77
<.0001
OBa
4391 (13.48)
5005 (13.46)
0.03
.98
RP
Abbreviations: ADT ¼ androgen-deprivation therapy; BT ¼ brachytherapy; CT ¼ cryotherapy; EBRT ¼ external beam radiation therapy; NCCN ¼ National Comprehensive Cancer Network; NCDB ¼ National Cancer Data Base; OB ¼ observation; RP ¼ radical prostatectomy. a Observation includes watchful waiting as well as active surveillance.
10.e4
-
Clinical Genitourinary Cancer Month 2017