Content uploaded by Gregory Peter Hess
Author content
All content in this area was uploaded by Gregory Peter Hess on Mar 02, 2021
Content may be subject to copyright.
Treatment Patterns: Targeted Therapies Indicated
for First-Line Management of Metastatic Renal
Cell Carcinoma in a Real-World Setting
Gregory Hess,
1,2
Rohit Borker,
3
Eileen Fonseca
1
Abstract
This retrospective observational study used US claims data (January 2007 to November 2010) to identify treatment
patterns, including treatment duration and dosing, for molecular-targeted agents indicated in first-line manage-
ment of advanced and/or metastatic renal cell carcinoma. The sample included 273 patients with metastatic renal
cell carcinoma on first-line therapy. Analysis of the results suggests that opportunities exist to improve treatment
duration in clinical practice and to better understand influences on treatment and dose changes.
Background: Limited information on real-world treatment patterns of targeted agents for metastatic renal cell
carcinoma (mRCC) is available to inform their use in clinical practice. Patients and Methods: This retrospective
observational study used US claims data (from January 2007 to November 2010) to identify treatment patterns,
including treatment duration and dosing, for molecular-targeted agents (sunitinib, sorafenib, pazopanib, bevaci-
zumab, and temsirolimus) indicated in first-line management of advanced and/or mRCC. The study included adult
patients with mRCC who were observable for ⱖ3 months after initiation of their first-line therapy with a targeted agent.
Descriptive analyses were conducted for observed treatment patterns. Results: Of the 273 patients on first-line
therapy identified and included in the sample, 235 patients were treated with sunitinib, 16 patients with sorafenib, and
15 patients with temsirolimus. Pazopanib and bevacizumab were excluded from the analysis due to the small sample
size, n ⬍10. The median observed treatment durations were sunitinib 98 days, sorafenib 121 days, and temsirolimus
78 days. Approximately 76% (178/235) of patients who received sunitinib initiated therapy at the indicated dose of
50 mg; 65% of these patients were not observed filling a fourth prescription, whereas 23% maintained their starting
dose and 12% experienced dose reduction at their 4⫹fill. The mean starting dose for patients who initiated on
sorafenib (n ⫽16) was 725 mg and for temsirolimus (n ⫽15) was 25 mg: their study samples were insufficient for
further, meaningful dosing analyses. Conclusions: Results of this study suggest that opportunities exist to improve
treatment duration in clinical practice and to better understand influences on treatment and dose changes.
Clinical Genitourinary Cancer, Vol. 11, No. 2, 161-7 © 2013 Elsevier Inc. All rights reserved.
Keywords: Antivascular endothelial growth factor agent, Cancer, Metastases, Rapamycin signaling inhibitor,
Tyrosine kinase inhibitor
Introduction
Metastatic renal cell carcinoma (mRCC) is resistant to conventional
chemotherapy and radiation therapy.
1,2
The prognosis of mRCC has
historically been poor, with a 5-year survival rate of approximately 10%
after postnephrectomy relapse or metastatic diagnosis.
3
Treatment has
generally been limited, until recently, to cytokine therapy with interleu-
kin-2 or interferon alfa. Advances in understanding the pathogenesis of
mRCC have contributed to the development of new classes of molecu-
lar-targeted therapies, including tyrosine kinase inhibitors (TKI)
(sunitinib, sorafenib, pazopanib), an antivascular endothelial growth
factor agent (bevacizumab), and mammalian target of rapamycin signal-
ing inhibitors (temsirolimus, everolimus).
4
These targeted therapies
have demonstrated biologic and clinical activity.
2,5-7
With the introduc-
tion of these therapies during the past 5 years, mRCC treatment is
undergoing a paradigm shift.
6,8
1
Health Economics and Outcomes Research, SDI Health LLC, Plymouth Meeting, PA
2
Institute for Health Economics and Policy, University of Pennsylvania, Plymouth Meeting,
PA
3
US Health Outcomes, GlaxoSmithKline, Research Triangle Park, NC
Submitted: May 28, 2012; Revised: Sep 28, 2012; Accepted: Oct 1, 2012; Epub: Dec 23, 2012
Address for correspondence: Gregory Hess, MD, Institute for Health Economics and
Policy, University of Pennsylvania, One IMS Drive, Plymouth Meeting, PA 19462
Fax: 610-834-8817; e-mail contact: Greg.Hess@wharton.upenn.edu
Original Study
Clinical Genitourinary Cancer June 2013 161
1558-7673/$ - see frontmatter © 2013 Elsevier Inc. All rights reserved.
http://dx.doi.org/10.1016/j.clgc.2012.10.003
Little information about real-world use of the targeted agents is
available to inform actual clinical practice.
7,9
Such information is
needed to help define therapeutic profiles and utilization patterns of
these agents in clinical populations not subject to the restrictive cri-
teria of clinical trials and is key to optimizing outcomes with molec-
ular-targeted therapies in the long term.
5,6,10
Further, knowledge of
real-world clinical use of molecular-targeted agents is valuable for
informing policy decisions and the development and modification of
treatment guidelines, which have been rapidly evolving with the in-
troduction of these therapies.
5,9
The objective of the study reported
herein was to characterize real-world treatment patterns, including
dosing, for targeted agents that are indicated for first-line treatment
of mRCC.
Patients and Methods
Design
This retrospective, observational, cohort study used claims data
from January 2007 through November 2010 to describe treatment
patterns of targeted agents indicated for first-line mRCC therapy in
a cohort of newly diagnosed patients.
Data Source
Data were extracted from SDI Health, LLC ([SDI], now part of
IMS Health, Plymouth Meeting, PA) databases of longitudinal, pa-
tient-level, private practitioner medical claims, and pharmacy claims.
Details of the database have previously been described.
11
The data set
comprises all payer types and claims from the 50 US states. All third-
party insurer types were captured, including private and public pay-
ers. The pharmacy claims database includes claims (National Coun-
cil for Prescription Drug Programs version 5.2) for more than 1.8
billion prescriptions dispensed annually. The medical claims data-
base includes more than 1.0 billion annual claims (Centers for Medi-
care & Medicaid Services 1500 forms) that contain diagnosis and
visit information, and represents activity of more than 870,000 prac-
titioners per month. The practices included any type and/or specialty
that uses electronic submission of claims for reimbursement to any
payer type in the United States, including commercial plans, Medi-
care, Medicaid, and others, (eg, Tricare military). Patients are as-
signed a synthetic identifier, and the databases are certified as being
compliant with the Health Insurance Portability and Accountability
Act. This study was exempt from institutional review board approval
because it was retrospective, did not involve an intervention, and
used anonymized data.
Sample
The study included patients aged ⱖ18 years who had ⱖ2 medical
visits with a diagnosis of renal carcinoma (International Classifica-
tion of Diseases, 9th edition [ICD] 189.0) and/or malignant neo-
plasm of the renal pelvis (ICD 189.1) between January 2007 and
May 2010, and ⱖ2 medical visits with a diagnosis of distant second-
ary malignant neoplasm (ICD 197.XX-199.0, excluding ICD 198.0
[kidney metastasis]) on different days ⬍120 days apart. The patients
must have been treated with oral or infused targeted agents (based on
National Drug Codes or Healthcare Common Procedure Coding
System, respectively) indicated for first-line treatment of mRCC,
have been observed in both the medical and pharmacy claims data-
bases for ⱖ6 months before and ⱖ3 months after the start of first-
line therapy for metastatic disease (index date), have had ⱖ1 visit to
a treating physician (a stable panel of physicians that consistently
administered oncologic therapy during the study period), and have
received prescriptions from a stable pharmacy. Patients were ex-
cluded if they became metastatic after May 2010, received study
drugs from 2 to 6 months before their metastatic date, received be-
vacizumab with a dosing and/or procedure code indicative of intra-
vitreal injection for wet age-related macular degeneration, received
chemotherapeutic agents indicated for or used to treat advanced
transitional cell carcinoma, had evidence of cystoscopy or biopsy of
the bladder within 30 days before their index date or any time there-
after, had evidence of radical cystectomy (57.7X) at any time, or had
any other primary cancer (ICD 140-188, 189.2-195, 200-208) dur-
ing the 180-day period preceding their index date or within 90 days
subsequent to their index date unless the site of the other primary
neoplasm and the site of distant metastases were the same.
Measures and Data Analysis
Key metrics were summarized with descriptive statistics. Each pa-
tient was classified and indexed to the initial targeted therapy (initi-
ating treatment regimen). Treatment patterns, including first-line
median duration of therapy, dosing patterns, and treatment changes,
defined as the introduction of new agents or the substitution of any
therapeutic agent in the initiating regimen, were summarized. All th e
patients were followed up for at least 3 months and then until the
earliest date that the patient advanced to second-line therapy,
the patient was censored (ie, no longer observed in the data), or
until the end of the study period.
The index therapy regimen was defined as the first line of therapy
that started within 30 days before or on or after the date that the
patient was recorded as being metastatic. The end of the first line of
therapy was defined as a ⱖ42-day gap in therapy. The second line of
therapy was defined as the addition to, discontinuation of, or substi-
tution of one or more therapies in the first-line regimen. For
sunitinib, a 14-day gap or respite was added to the end of each
prescription’s days supply to normalize to the recommended dosing
schedule. The starting dose of sunitinib (ie, 12.5, 25, 37.5, or 50 mg
daily) was determined by using the calculation (strength ⫻quantity
dispensed)/days of supply. The duration of therapy was defined as
the time between the start and end of the first line of therapy. If a
patient was still on first-line therapy at time of censoring from the
data or if the end of the study period was reached, then the patient’s
end of first-line therapy was set to this date.
After application of the inclusion-exclusion criteria, 2 targeted
therapies indicated for mRCC had patient counts deemed a priori as
insufficient for analysis beyond baseline descriptors; pazopanib (n ⫽
2) and bevacizumab (n ⫽5). For the remaining 3 therapies
(sunitinib, sorafenib, and temsirolimus), dosing was summarized by
prescription and/or administration. Dosing per prescription and/or
administration was compared with starting dosage and was classified
as escalated, maintained, or reduced. Clinical characteristics of inter-
est included metastatic status at presentation and site(s) of recorded
metastases and most frequently observed concomitant conditions
during the study period.
Real-World Treatment Patterns in Metastatic Renal Cell Carcinoma
162 Clinical Genitourinary Cancer June 2013
Results
A total of 273 patients who starting first-line therapy between January
2007 and August 2010 were identified and included in the study sample
(Table 1). The mean (SD) age of the patient sample was 61 ⫾10.5 years
(median, 58 years); 65% (n ⫽178) were men. The majority of patients
in the sample received sunitinib (86.1%, n ⫽235), followed by
sorafenib (5.9%, n ⫽16), and temsirolimus (5.5%, n ⫽15). The pa-
tients who received bevacizumab or pazopanib were excluded from all
treatment analyses due to the low sample size (n ⬍10). Approximately
23% (n ⫽64) of patients were treated by physicians affiliated with
community hospitals, 27% (n ⫽74) with an academic facility, and 29%
(n ⫽79) with both a community and an academic affiliation. In 21%
Table 1 Demographics and Clinical Characteristics by First-Line Therapy
All Sorafenib Sunitinib Pazopanib Bevacizumab Temsirolimus
Total No. Patients 273 16 235 2 5 15
Age, y
Mean 61 62 61 51 63 61
Median 58 60 58 44 72 61
Age, No. (%)
⬍65 y 177 (64.8) 10 (62.5) 152 (64.7) 2 (100) 3 (60.0) 10 (66.7)
ⱖ65 y 96 (35.2) 6 (37.5) 83 (35.3) 0 (0) 2 (40.0) 5 (33.3)
Men, No. (%) 178 (65.2) 11 (68.8) 150 (63.8) 2 (100) 5 (100) 10 (66.7)
Region, No. (%)
Midwest 61 (22.3) 6 (37.5) 52 (22.1) 0 (0) 0 (0) 3 (20.0)
Northeast 36 (13.2) 0 (0) 34 (14.5) 1 (50.0) 0 (0) 1 (6.7)
South 98 (35.9) 7 (43.8) 83 (35.3) 1 (50.0) 1 (20.0) 6 (40.0)
West 78 (28.6) 3 (18.8) 66 (28.1) 0 (0) 4 (80.0) 5 (33.3)
Hospital Affiliation, No. (%)
Academic and community 79 (28.9) 7 (43.8) 67 (28.5) 1 (50.0) 1 (20.0) 3 (20.0)
Academic 74 (27.1) 0 (0) 67 (28.5) 0 (0) 1 (20.0) 6 (40.0)
Community 64 (23.4) 7 (43.8) 49 (20.9) 0 (0) 3 (60.0) 5 (33.3)
Unknown 56 (20.5) 2 (12.5) 52 (22.1) 1 (50.0) 0 (0) 1 (6.7)
Comprehensive Cancer Centers
(of Known Affiliations), No. (%)
NCI designated 11 (5.0) 0 (0) 11 (6.0) 0 (0) 0 (0) 0 (0)
Not NCI designated 206 (94.9) 14 (100) 172 (94.0) 1 (100) 5 (100) 14 (100)
Metastatic Status, No. (%)
Presented as metastatic 78 (28.6) 3 (18.8) 70 (29.8) 0 (0) 0 (0) 5 (33.3)
Nonmetastatic to metastatic 195 (71.4) 13 (81.3) 165 (70.2) 2 (100) 5 (100) 10 (66.7)
Site of Metastasis During Study Period, No. (%)
Known sites 273 (100) 16 (100) 235 (100) 2 (100) 5 (100) 15 (100)
Bone 158 (57.9) 8 (50.0) 134 (57.0) 1 (50.0) 2 (40.0) 13 (86.7)
Brain 85 (31.1) 5 (31.3) 72 (30.6) 1 (50.0) 1 (20.0) 6 (40.0)
Liver 51 (18.7) 3 (18.8) 42 (17.9) 2 (100) 1 (20.0) 3 (20.0)
Lung 170 (62.3) 8 (50.0) 148 (63.0) 1 (50.0) 4 (80.0) 9 (60.0)
Other 89 (32.6) 2 (12.5) 77 (32.8) 1 (50.0) 2 (40.0) 7 (46.7)
Top 5 Concomitant Conditions
Other symptoms or adverse reactions 273 (100) 16 (100) 235 (100) 2 (100) 5 (100) 15 (100)
Kidney disorder 185 (67.8) 13 (81.3) 157 (66.8) 2 (100) 1 (20.0) 12 (80.0)
Hypertension 163 (59.7) 11 (68.8) 137 (58.3) 2 (100) 5 (100) 8 (53.3)
Anemia 138 (50.5) 11 (68.8) 114 (48.5) 1 (50.0) 2 (40.0) 10 (66.7)
Malaise and fatigue 87 (31.9) 4 (25.0) 77 (32.8) 0 (0) 1 (20.0) 5 (33.3)
Abbreviation: NCI ⫽National Cancer Institute.
Gregory Hess et al
Clinical Genitourinary Cancer June 2013 163
(n ⫽56) of the cases, the affiliation could not be ascertained. A minority
(5%, n ⫽11) of the patients whose physician had a known hospital
association had a physician affiliated with 1 of the 44 National Cancer
Institute (NCI) Comprehensive Cancer Centers. A little less than one-
third of patients (28.6%, n ⫽28) presented at the time of first RCC
diagnosis with metastasis. During the study period, lung (62.3%, n ⫽
170), bone (57.9%, n ⫽158), and brain (31.1%, n ⫽85) were the most
common sites of metastasis.
Treatment Patterns
Data about treatment changes and dose patterns for sunitinib,
sorafenib, and temsirolimus are shown in Table 2. The median du-
ration of treatment for first-line sunitinib was 98 days (range, 34-762
days), with a mean (SD) starting dose of 45.5 ⫾8.9 mg (median, 50
mg). At 6 months after the start of first-line therapy, 10% (n ⫽24)
of the patients on sunitinib were censored and 18% (n ⫽43) had
switched or added on therapies; by the ninth month, 55% (n ⫽128)
remained on first-line therapy or were observed in the database with-
out further treatment. The median duration of sorafenib first-line
treatment was 121 days (range, 13-552 days), with a mean (SD)
starting dose of 725.0 ⫾161.2 mg (median, 800 mg). At 6 months
after first-line initiation, 6% (n ⫽1) of patients on sorafenib were
censored, whereas a fourth (n ⫽4) had added or switched therapies;
at 9 months later, 56% (n ⫽9) remained on first-line therapy or were
observed in the database without further treatment. With regard to
temsirolimus, the median duration of treatment was 78 days (range,
7-358 days) and the starting dose was 25 mg/mL (SD 0 mg; median,
25.0 mg/mL). No patients on temsirolimus were censored, although
53% (n ⫽8) had added on or switched therapies by 6 months after
first-line initiation and 67% (n ⫽10) by 9 months after first-line
initiation, which left approximately 47% and 33%, respectively, still
on first-line therapy or in-between first- and second line therapies.
The larger sunitinib sample allowed for detailed reporting of dos-
ing patterns by initial dose (Table 3). Of 235 patients on sunitinib,
approximately 76% (n ⫽178) started on 50 mg, the indicated dose
Table 2 Treatment and Dosing Patterns for Patients with mRCC on First-Line Therapy
Treatment & Dose Patterns Oral Products I.V. Products
Sorafenib Sunitinib Temsirolimus
No. Patients 16 235 15
Starting Dose, mg
Mean 725.0 45.5 25.0
Median 800.0 50.0 25.0
SD 161.2 8.9 0.0
Min 400.0 12.5 25.0
Max 800.0 50.0 25.0
Duration of Therapy, days
Mean (in the observed sample) 148.8 143.5 126.9
Median 121.0 98.0 78.0
SD 144.9 119.7 102.9
Min 13.0 34.0 7.0
Max 552.0 762.0 358.0
Treatment Observation at Selected Time Periods After Initiation
of First-Line Therapy, % (n)
3 mo after
On first line or observed without new treatment 87.5 (14) 94.0 (221) 86.7 (13)
Switched and/or added on therapy 12.5 (2) 6.0 (14) 13.3 (2)
Censored 0.0 0.0 0.0
6 mo after
On 1st line or observed without new treatment 68.8 (11) 71.5 (168) 46.7 (7)
Switched and/or added on therapy 25.0 (4) 18.3 (43) 53.3 (8)
Censored 6.3 (1) 10.2 (24) 0.0
9 mo after
On 1st line or observed without new treatment 56.3 (9) 54.5 (128) 33.3 (5)
Switched and/or added on therapy 31.3 (5) 26.8 (63) 66.7 (10)
Censored 12.5 (2) 18.7 (44) 0.0
Abbreviations: I.V. ⫽intravenous; max ⫽maximum; min ⫽minimum; mRCC ⫽metastatic renal cell carcinoma.
Real-World Treatment Patterns in Metastatic Renal Cell Carcinoma
164 Clinical Genitourinary Cancer June 2013
for advanced and/or mRCC, 14% (n ⫽33) initiated treatment with
37.5 mg, and 8.5% (n ⫽20) on 25 mg. A minority (⬍2%, n ⫽4)
started on 12.5 mg. The median number of sunitinib prescriptions
was 2. Among those patients who initiated sunitinib at 50 mg, 28%
(n ⫽50) were not observed with a second sunitinib prescription,
63% (n ⫽112) were maintained on 50 mg, whereas 9% (n ⫽16)
experienced dose reduction. Among the 9% of patients who experi-
enced dose reduction, 69% were dose reduced to 37.5 mg, whereas
25% were dose reduced to 25 mg. Sixty-five percent (n ⫽116) of
patients who initiated sunitinib were not observed with a fourth
prescription, 23% (n ⫽41) were maintained on 50 mg at the fourth
or later prescription fill, and 12% (n ⫽21) experienced dose reduc-
tion. Among the 12% who experienced dose reduction, 86% were
dose reduced to 37.5 mg and 14% to 25 mg.
Among those patients on sunitinib begun at a dose of 37.5 mg,
24% (n ⫽8) were not observed receiving a second sunitinib prescrip-
tion, 3% (n ⫽1) were escalated to a dose of 50 mg at the second
prescription, whereas 67% (n ⫽22) were maintained on a dose of
37.5 mg, and 6% (n ⫽2) experienced dose reduction. Fifty-five
percent (n ⫽18) of patients did not have a fourth prescription,
whereas, by the fourth or later prescription, 9% (n ⫽3) were esca-
lated to a dose of 50 mg, 24% (n ⫽8) were maintained at 37.5 mg,
and 12% (n ⫽4) experienced dose reduction. These trends for pa-
tients who initiated sunitinib treatment at 25 mg and 12.5 mg as well
are reported in Table 3; however, the sample sizes were smaller (n ⬍
30). Detailed dose patterns with sorafenib (n ⫽16) and temsiroli-
mus (n ⫽15) were not reported due to the insufficient sample size.
Discussion
The addition of new molecular-targeted first-line treatments for
advanced and/or mRCC since 2005 in the United States has led to
calls for studies that characterize the use of and outcomes with these
new therapies in real-world clinical practice.
5,9
Such information is a
necessary complement to clinical trial data. This study characterized
273 patients with nationally representative mRCC and their treat-
ment patterns with targeted agents. Descriptive results are presented;
statistical comparisons were not made among treatment groups be-
cause, other than sunitinib, sample sizes were too small to make
consequential comparisons. Very few patients received bevacizumab
(n ⫽5) and pazopanib (n ⫽2), and, hence, these study treatments
were not analyzed beyond patient characteristics. Relatively recent
US Food and Drug Administration approval of pazopanib (October
2009) and bevacizumab (August 2009) likely contributed to the
small number of observed patients in the study sample.
The patients with mRCC were a median age of 58 years and
predominantly men, findings in keeping with the known demogra-
phy of mRCC in the United States.
9,12,13
Also, as expected, just
under one-third of the patients presented with metastatic disease at
the time of first RCC diagnosis.
14
Bone was the second most com-
mon site of metastasis, when excluding lymph nodes, after lung,
consistent with other studies.
13,14
This study’s population had nearly
double the percentage (58% vs. approximately 33%) of patients with
bone metastases. This finding is not surprising because the propor-
tion of patients with metastasis grows throughout first-line therapy,
and this study reported metastases during the entire observation pe-
riod as opposed to at baseline initiation of first-line therapy.
14
Table 3 Sunitinib Dosing Patterns by First-Line Therapy
Initiating Dose
Sunitinib Daily Dosing
Patterns
Sunitinib Starting Dose
12.5 mg 25 mg 37.5 mg 50 mg
Unique Patient Count 4 20 33 178
% Patients Initiated With
Sunitinib 1.7 8.5 14.0 75.7
No. Patients
1 prescription 2 6 8 50
2 prescriptions — 4 8 40
3 prescriptions — — 2 25
4 prescriptions 1 3 5 15
5⫹prescriptions 1 7 10 48
Second Prescription Compared
With First Prescription
a
% not fill second prescription (n) 50 (2) 30 (6) 24 (8) 28 (50)
% escalated (n) — 10 (2) 3 (1)
50 mg — 100 (2) 100 (1)
% maintained (n) 50 (2) 60 (12) 67 (22) 63 (112)
% reduced (n) — 6 (2) 9 (16)
12.5 mg — 50 (1) 6 (1)
25 mg 50 (1) 25 (4)
37.5 mg 69 (11)
Third Prescription Compared
With First Prescription
a
% not fill third prescription (n) 50 (2) 50 (10) 49 (16) 51 (91)
% escalated (n) 25 (1) 5 (1) 9 (3)
50 mg 100 (1) 100 (1) 100 (3)
% maintained (n) 25 (1) 45 (9) 30 (10) 44 (78)
% reduced (n) — 12 (4) 5 (10)
12.5 mg — 50 (2) 10 (1)
25 mg 50 (2) 10 (1)
37.5 mg 80 (8)
Fourth or More Than Fourth
Prescription Compared With
First Prescription
a
% not fill fourth prescription (n) 50 (2) 50 (10) 55 (18) 65 (116)
% escalated (n) 50 (2) 20 (4) 9 (3)
25 mg 50 (1)
37.5 mg 50 (1) 75 (3)
50 mg — 25 (1) 100 (3)
% maintained (n) — 25 (5) 24 (8) 23 (41)
% reduced (n) 5 (1) 12 (4) 12 (21)
12.5 mg 100 (1) 25 (1) —
25 mg 75 (3) 14 (3)
37.5 mg 86 (18)
a
The percent of patients who did not fill second, third, or 4⫹prescriptions, escalated, main-
tained or reduced were calculated based on the unique patient counts. The percent of patients
escalating to 25, 37.5 or 50 mg or reducing their doses to 12.5, 25 or 37.5 mg was calculated
based on the number of patients escalating or reducing their dose.
Gregory Hess et al
Clinical Genitourinary Cancer June 2013 165
The median duration of each therapy in this study was lower than
expected. The median duration of first-line therapy for the patients
on sunitinib (85% of the sample, n ⫽235) was 98 days (3.3
months). As a point of reference, the relatively low rates of censorship
for these patients (0% at 3 months, 10% at 6 months, and 19% at 9
months) provided robust observation for analyses. In contrast, phase
III clinical trial data showed that patients on sunitinib had a median
duration of therapy of at least 6 months (approximately 180 days).
15
The patients on first-line sorafenib represented 5.9% of this study’s
sample (n ⫽16) and had a median duration of therapy of 121 days
(4.0 months). As with sunitinib, the patients on sorafenib had robust
periods of follow-up: 0% were censored at 3 months, 6.3% at 6
months, and 12.5% at 9 months. Approximately 6% of patients (n ⫽
15) initiated on temsirolimus and had a median duration of 78 days
(2.6 months), and no patients were censored by 9 months. In their
respective phase III trials, the patients on sorafenib had approxi-
mately 161 days’ duration of therapy (median) and the patients on
temsirolimus had a median duration of 119 days.
16,17
Caution is
warranted when interpreting this study’s sorafenib and temsirolimus
data due to the low sample sizes.
When descriptively “comparing” this study with other “real-
world” studies, the treatment duration for sunitinib and sorafenib
yields mixed results; temsirolimus was not included in many real-
world studies, so data are not as readily available for comparison. In
line with this study’s reported results is a recent study by Filson et al,
9
who used private insurers’ claims data that found a median duration
of 90 days for sunitinib and 120 days for sorafenib. Short-term use of
oral TKIs (defined as a treatment regimen that lasted ⬍60 days) in
the study by Filson et al
9
was reported to be common (36%) among
patients who received first-line treatment. This observation may ex-
plain the shorter duration of treatment compared with that observed
in clinical trial settings with strict eligibility controls. The duration of
therapy in the study by Filson et al
9
only included the days supply
from prescriptions and, therefore, did not include the final prescrip-
tion’s 14-day gap or respite per dosing schedule or small gaps be-
tween prescription fills. However, it did not include patients ⬎65
years old, who represent approximately one-third of our study’s sam-
ple and who had a shorter duration of therapy. In our study, the
median duration of treatment with sunitinib was 114 days (3.8
months) among patients ⬍65 years of age and 83 days (2.7 months)
among patients ⱖ65 years of age. Notwithstanding these differences,
the results of the current study seem to substantiate findings from the
study by Filson et al
9
with regard to the short duration of treatment.
Feinberg et al
18
conducted a study that involved a retrospective
review of medical record data from patients with mRCC who re-
ceived anti-angiogenic treatment at community oncology clinics.
The study reported median durations of treatment of 5.9 and 5.5
months for first-line sunitinib and sorafenib, respectively, values rel-
atively higher than those observed in the study by Filson et al
9
and in
our study. The reasons for the reported differences are not clear and
may reflect differences in practice patterns and/or differences in pa-
tient characteristics across these data sets.
A study by Choueiri et al,
13
who conducted chart reviews of 2
New England tertiary oncology centers, reported a median duration
of therapy of 10.5 months for sunitinib and 8.1 months for
sorafenib. One of the centers is an NCI-designated comprehensive
cancer center. In comparison, our study and the study by Filson et al
9
both represent a broader patient and provider coverage across the
United States. For instance, of the 80% of patients with physicians
having a confirmed hospital affiliation, only 4% of these patients
received their index treatment from physicians affiliated with an NCI
comprehensive cancer center in the current study. The NCI esti-
mates that 85% of all cancer patients are treated at community hos-
pitals.
19
Although not examined in this study, perhaps the treatment
duration differences are influenced, in part, by the characteristics of
physicians, their practices, and their hospital affiliations.
In the current study, three-fourths of patients who initiated
sunitinib started at the recommended 50-mg dose, which is similar to
the study by Choueiri et al
13
as well as the study by Feinberg et al.
18
The observation in this study that 28% of patients who were started
at 50 mg did not fill a second prescription may explain the observed
shorter duration of treatment.
Conclusion
The reasons for potential burden of dose reduction on physicians
and patients were not assessed in this study and warrant assessment in
future research. The aforementioned retrospective chart review stud-
ies suggest that adverse events can be a frequent reason for treatment
modifications in patients who receive multikinase inhibitors.
12,17
The degree to which the latter data, which were obtained from small
samples of patients, can be generalized is unknown. Nonetheless, the
differences in the characteristics of selected studies (leading oncology
centers and controlled clinical trials vs. broad-based claims studies)
and their respective results perhaps indicate that a physician’s hospi-
tal affiliation plays a role in the physician’s treatment patterns and use
of dose reduction as a strategy to maintain continuation of treatment.
Analysis of these results suggests that studies can be undertaken to
investigate whether disparities of duration of treatment can be par-
tially attributable to a difference in physician hospital affiliation;
whether greater coordination of care between the physician and the
patient is a factor in ensuring successful continuation of therapy at an
optimal dose and to better understand the rationale for treatment
modification outside of disease progression.
This study’s results should be interpreted in the context of its
limitations. First, claims data have inherent limitations because they
are collected for billing and reimbursement purposes rather than for
research purposes. Second, data entry errors at the site of care could
not be detected or corrected in data analysis. These limitations are
common to all electronic databases. Furthermore, the retrospective
nature of the study could make the results subject to selection bias.
Finally, in this rapidly evolving era of targeted therapy for mRCC,
the results should be interpreted with an awareness of the time frame
in which the study was conducted. The data from this investigation
can be considered complementary to other analyses that may assess
populations that differ from the current one or that differ in methods
of data capture or analysis. These limitations notwithstanding, this
study provides new information about contemporary, real-world use
of targeted therapies in a large, nationally representative sample of
patients with mRCC in the United States.
Clinical Practice Points
●Advances in understanding the pathogenesis of mRCC have con-
tributed to the development of new classes of molecular-targeted
Real-World Treatment Patterns in Metastatic Renal Cell Carcinoma
166 Clinical Genitourinary Cancer June 2013
therapies, including TKIs (sunitinib, sorafenib, pazopanib), an
antivascular endothelial growth factor agent (bevacizumab), and
mammalian target of rapamycin signaling inhibitors (temsiroli-
mus, everolimus). Little information about real-world use of the
targeted agents is available to inform actual clinical practice.
●This study characterized real-world treatment patterns, including
dosing, for targeted agents that are indicated for first-line treat-
ment of mRCC.
●Analysis of the results suggest that opportunities exist to improve
treatment duration with molecular-targeted therapies in clinical
practice and to better understand influences on treatment and dose
changes.
Acknowledgments
The authors thank Jane Saiers, PhD (the WriteMedicine Inc) for
assistance with writing this manuscript, which was funded by SDI.
The authors thank Dhruv Mangla and Michael Gizzi from SDI for
their contributions to the study. This study was conducted by SDI,
which is now an IMS Health Corporation. This work was supported
by GlaxoSmithKline.
Disclosure
Dr. Hess and Ms. Fonseca, are employees of SDI Health LLC,
which received study funding provided by GlaxoSmithKline. Dr.
Borker is an employee of GlaxoSmithKline.
References
1. Facchini G, Perri F, Caraglia M, et al. New treatment approaches in renal cell
carcinoma. Anticancer Drugs 2009; 20:893-900.
2. Hutson TE. Targeted therapies for the treatment of metastatic renal cell carcinoma:
clinical evidence. Oncologist 2011; 16(suppl 2):14-22.
3. Molina AM, Motzer RJ. Current algorithms and prognostic factors in the treatment
of metastatic renal cell carcinoma. Clin Genitourin Cancer 2008; 6(suppl 1):S7-S13.
4. Motzer RJ. New perspectives on the treatment of metastatic renal cell carcinoma: an
introduction and historical overview. Oncologist 2011; 16(suppl 2):1-3.
5. Molina AM, Motzer RJ. Clinical practice guidelines for the treatment of metastatic
renal cell carcinoma: today and tomorrow. Oncologist 2011; 16(suppl 2):45-50.
6. Reeves DJ, Liu CY. Treatment of metastatic renal cell carcinoma. Cancer Chemother
Pharmacol 2009; 64:11-25.
7. Wysocki PJ, Zolnierek J, Szczylik C, et al. Targeted therapy of renal cell cancer. Curr
Opin Investig Drugs 2008; 9:570-5.
8. Kidney cancer. National Comprehensive Cancer Network Clinical Practice guide-
lines in Oncology—v.2.2010. National Comprehensive Cancer Network [Web
site]. Available at: http://www.nccn.org/professionals/physician_gls/pdf/kidney.
pdf. Accessed: April 27, 2012.
9. Filson CP, Redman BG, Dunn RL, et al. Initial patterns of care with oral targeted
therapies for patients with renal cell carcinoma. Urology 2011; 77:825-30.
10. Ravaud A. How to optimise treatment compliance in metastatic renal cell carcinoma
with targeted agents. Ann Oncol 2009; 20(suppl 1):i7-i12.
11. Chen CC, Hess GP, Liu Z, et al. Second-line treatment outcomes after first-line
sunitinib therapy in metastatic renal cell carcinoma. Clin Genitourin Cancer 2012;
10:256-61.
12. Sachdeva K, Jana BRP, Javeed M, et al. Renal cell carcinoma eMedicine Available at:
http://emedicine.medscape.com/article/281340-overview. Accessed: March 8,
2011.
13. Choueiri TK, Duh MS, Clement J, et al. Angiogenesis inhibitor therapies for met-
astatic renal cell carcinoma: effectiveness, safety and treatment patterns in clinical
practice-based on medical chart review. BJU Int 2010; 105:1247-54.
14. Sahi C, Knox JJ, Clemons M, et al. Renal cell carcinoma bone metastases: clinical
advances. Ther Adv Med Oncol 2010; 2:75-83.
15. Motzer RJ, Hutson TE, Tomczak P, et al. Sunitinib versus interferon alfa in meta-
static renal-cell carcinoma. N Engl J Med 2007; 356:115-24.
16. Escudier B, Eisen T, Stadler W, et al. Sorafenib in advanced clear-cell renal-cell
carcinoma. N Engl J Med 2007; 356:124-34.
17. Hudes G, Carducci M, Tomczak P, et al. Temsirolimus, interferon alfa, or both for
advanced renal-cell carcinoma. N Engl J Med 2007; 356:2271-81.
18. Feinberg BA, Jolly P, Wang ST, et al. Safety and treatment patterns of angiogenesis
inhibitors in patients with metastatic renal cell carcinoma: evidence from US com-
munity oncology clinics. Med Oncol 2012; 29:786-94.
19. National Cancer Institute. NCI community cancer centers program pilot: 2007-
2010: FAQ. Available at: http://ncccp.cancer.gov/Media/FactSheet.htm. Accessed:
November 30, 2012.
Gregory Hess et al
Clinical Genitourinary Cancer June 2013 167
