EPIDEMIOLOGY

A retrospective evaluation of chemotherapy dose intensityand supportive care for early-stage breast cancerin a curative setting

Gary H. Lyman • David C. Dale • Dianne Tomita •

Sadie Whittaker • Jeffrey Crawford

Received: 18 December 2012 / Accepted: 29 May 2013 / Published online: 16 June 2013

� Springer Science+Business Media New York 2013

Abstract Early-stage breast cancer (ESBC) is commonly

treated with myelosuppressive chemotherapy, and main-

taining full-dose chemotherapy on the planned schedule is

associated with improved patient outcome. Retrospective

analysis of patients with ESBC treated from 1997 to 2000

showed that 56 % of patients received a relative dose

intensity (RDI) \85 % (Lyman et al., J Clin Oncol

21(24):4524–4531, 2003). To determine current practice,

we evaluated treatment patterns at 24 US community- and

hospital-based oncology practices, 79 % of which partici-

pated in the previous study. Data were abstracted from

medical records of 532 patients with surgically resected

ESBC (stage I–IIIa) treated from 2007 to 2009, who were

C18 years old and had completed C1 cycle of one of the

following regimens: docetaxel ? cyclophosphamide (TC);

doxorubicin ? cyclophosphamide (AC); AC followed by

paclitaxel (AC-T); docetaxel ? carboplatin ? trastuzumab

(TCH); or docetaxel ? doxorubicin ? cyclophosphamide

(TAC). Endpoints included RDI, dose delays, dose reduc-

tions, grade 3/4 neutropenia, febrile neutropenia (FN), FN-

related hospitalization, granulocyte colony-stimulating

factor (G-CSF) use, and antimicrobial use. In this study, TC

was the most common chemotherapy regimen (42 %), and

taxane-based chemotherapy regimens were more common

relative to the previously published results (89 vs \4 %).

Overall, 83.8 % of patients received an RDI C85 %, an

improvement over the previous study where 44.5 %

received an RDI C85 %. Other changes seen between this

and the previous study included a lower incidence of dose

delays (16 vs 25 %) and dose reductions (21 vs 37 %) and

increased use of primary prophylactic G-CSF (76 vs*3 %).

Here, 40 % of patients had grade 3/4 neutropenia, 3 % had

FN, 2 % had an FN-related hospitalization, and 30 %

received antimicrobial therapy; these measures were not

available in the previously published results. Though RDI

was higher here than in the previous study, 16.2 % of

patients still received an RDI \85 %. Understanding factors

that contribute to reduced RDI may further improve che-

motherapy delivery, and ultimately, patient outcomes.

Keywords Relative dose intensity � Chemotherapy �Supportive care � Breast cancer

Introduction

The first adjuvant chemotherapy to show a benefit in breast

cancer was cyclophosphamide, methotrexate, and fluoro-

uracil (CMF) [1, 2]. Since the introduction of CMF, adju-

vant chemotherapy for early-stage breast cancer (ESBC)

has continued to evolve. Numerous studies have shown that

chemotherapy regimens using doxorubicin and cyclo-

phosphamide (AC) are equivalent to CMF [3–6], and

higher doses of anthracycline-based regimens have dem-

onstrated superiority to CMF [7]. The addition of taxanes

Electronic supplementary material The online version of thisarticle (doi:10.1007/s10549-013-2582-2) contains supplementarymaterial, which is available to authorized users.

G. H. Lyman (&) � J. Crawford

Duke Cancer Institute, Duke University, 2424 Erwin Road,

Suite 205, Durham, NC 27705, USA

e-mail: gary.lyman@duke.edu

D. C. Dale

University of Washington, 1959 NE Pacific Street, Seattle,

WA 98195, USA

D. Tomita � S. Whittaker

Amgen Inc., One Amgen Center Way, Thousand Oaks,

CA 91360, USA

123

Breast Cancer Res Treat (2013) 139:863–872

DOI 10.1007/s10549-013-2582-2

has further improved patient outcomes and reduced the risk

of relapse and death [3, 8–10]. Additional studies have

focused on optimizing drug delivery through sequential

schedules and dose-dense regimens [11]. Advances in

adjuvant chemotherapy and wide-spread screening pro-

grams have contributed to a steady decrease in death rates

associated with breast cancer [12–15].

Chemotherapy dose and schedule are important clinical

variables that can impact patient outcomes. Relative dose

intensity (RDI) accounts for both dose and schedule of

drug and is defined as the received dose intensity relative to

the reference dose intensity [16]. For patients with ESBC,

high RDI of adjuvant chemotherapy (C85 %) is associated

with improved disease-free survival and overall survival

[17, 18]. Thus, maintaining high RDI is an accepted goal of

chemotherapy administration in the curative setting [19].

In our previous study (Study 1), we performed a retro-

spective analysis of more than 20,000 patients with ESBC

from 1997 to 2000. In Study 1, 56 % of patients received

an RDI \85 % [20]. Since the completion of Study 1,

breast cancer treatment has shifted from anthracycline-

based therapies to taxane-based therapies [21]. To deter-

mine how recent advances have affected chemotherapy

administration and supportive care and to evaluate the

incidence of chemotherapy-induced complications in

patients with ESBC, a retrospective chart review of US

community- and hospital-based outpatient oncology prac-

tices was conducted (Study 2). The primary objective of

this study was to evaluate RDI. Because 79 % of the sites

in this study also participated in Study 1, changes in clin-

ical practice over time could be described.

Patients and methods

Study design

Following approval by a central Institutional Review Board

(IRB), de-identified patient data were collected retrospec-

tively from 24 community- and hospital-based outpatient

oncology practices geographically distributed across the

United States (Supplemental Table 1). Each study site

identified sequential patient charts that met eligibility

requirements, and data were abstracted from patient charts

of 532 patients with ESBC treated from January 2007 to

December 2009.

Patient selection

Patients were included in the analysis if they were at least

18 years old, chemotherapy naı̈ve at study start, completed

definitive surgery for ESBC (stages I–IIIA), and received at

least one cycle of one of the following adjuvant chemotherapy

regimens during the study period: docetaxel ? cyclophos-

phamide (TC); doxorubicin ? cyclophosphamide (AC); AC

followed by paclitaxel; docetaxel ? carboplatin ? trast-

uzumab (TCH); or docetaxel ? doxorubicin ? cyclophos-

phamide (TAC) (Supplemental Table 2). Patients receiving

AC followed by paclitaxel were divided into four groups

based on dosing regimen (see Supplemental Table 2) as fol-

lows: 21-day AC followed by weekly (AC-Tweekly); 14-day

AC followed by 14-day paclitaxel (AC-T14); 14-day AC

followed by weekly paclitaxel (AC14-Tweekly); and 21-day

AC followed by 21-day paclitaxel (AC-T).

Patients were excluded if they had metastatic breast

cancer, had begun chemotherapy outside of the chart

abstraction period, or if, during the chart abstraction period,

they participated in a clinical trial requiring CSFs, received

granulocyte macrophage (GM)-CSF, or received investi-

gational agents.

Study endpoints

The primary endpoint of the study was RDI, calculated as

the ratio of actual dose intensity (ADI) to standard dose

intensity (SDI) as follows:

RDI ¼ ADI=SDI

ADI ¼ DoseActual=TimeActual

SDI ¼ DoseStandard=TimeStandard

where DoseActual is the total received dose of chemother-

apy, TimeActual is the total length of chemotherapy,

DoseStandard is the NCCN standard total dose of chemo-

therapy, and TimeStandard is the standard total length of

chemotherapy including all planned cycles. If a patient

received fewer than the planned number of cycles and had

no evidence of disease progression or death, then a dose of

zero was assigned for each missed cycle, and TimeActual

was the sum of the observed time for the cycles received

and the standard time required for the missed cycles. RDI

was calculated for each myelotoxic agent separately and

then averaged across the regimen. All agents except for

trastuzumab were included. For regimens containing

weekly paclitaxel, a cycle was defined as three weekly

doses of paclitaxel on day 1, 8, and 15. In general, RDI

should fall between 0 and 100 %. However, since this is an

observational study, RDI may be [100 % if a higher than

standard dose was given in any cycle, greater than the

planned number of cycles was administered, or the duration

of treatment was shorter than the standard duration.

Secondary endpoints were the incidence of chemother-

apy dose delays C7 days, dose reductions C15 % from

NCCN standard [22], grade 3/4 neutropenia, FN, FN-

related hospitalization, G-CSF use, and antimicrobial use.

Initial dose reductions were defined as C15 % from the

864 Breast Cancer Res Treat (2013) 139:863–872

123

NCCN standard [22], except for carboplatin. As area under

the curve (AUC) was not captured for carboplatin, the

initial reduction was defined as C15 % reduction relative to

the planned dose. Subsequent dose reductions were only

counted as separate events if the dose was reduced by

C15 % relative to previous dose reductions. Reasons for

dose delays were captured as the following discrete field

codes: other reasons related to chemotherapy, other reasons

not related to chemotherapy, personal/patient request,

neutropenia, and other hematological reasons. Reasons for

dose reductions were captured as the following discrete

field codes: other reasons related to chemotherapy, weight

change, neutropenia, thrombocytopenia, anemia, and dose

administration error. Grade 3/4 neutropenia was defined as

an absolute neutrophil count (ANC) \ 1.0 9 109/L. Grade

3/4 FN was defined as a clinical diagnosis of FN; tem-

perature C 38.0 �C and ANC \ 1.0 9 109/L within a 24-h

period; or hospitalization for FN. Grade 4 FN was defined

as ANC \ 0.5 9 109/L with clinical FN diagnosis or

temperature C38.0 �C or hospitalization for FN. Prophy-

lactic G-CSF was defined as initial use within 5 days of

chemotherapy administration. Primary prophylactic G-CSF

was defined as prophylactic use in the first cycle of che-

motherapy. Secondary prophylactic G-CSF was defined as

prophylactic use beginning with the second or subsequent

cycles of chemotherapy and could include patients who

first received G-CSF as treatment and received prophylaxis

in subsequent cycles. Treatment G-CSF was defined as

initial use for more than 5 days after chemotherapy

administration.

Statistics

A sample size of 500 patients was selected for this study.

With this sample size, the estimated half-width of a 95 %

confidence interval for a proportion would be B4.4 %. For

endpoints such as RDI with expected subsets with fewer

than 500 patients, the half-width would be B6.2 % for a

sample size of 250 patients, B8.8 % for 125 patients,

B12.7 % for 60 patients, B17.8 % for 30 patients, and

B25.3 for 15 patients (nQuery Advisor 7.0, Statistical

Solutions, Saugus, MA). Due to the retrospective, non-

comparative study design, only descriptive statistics are

provided. These descriptive statistics include means, 95 %

CIs, and ranges for continuous endpoints and frequencies,

percentages, and 95 % CIs for categorical endpoints.

Missing values were not imputed for these descriptive

analyses.

Multivariate analysis

Logistic regression modeling was used to explore the rela-

tionship between baseline patient factors and RDI \85 %.

After initial assessment of the baseline factors, both for-

ward and backward stepwise procedures were used to

identify factors with a P value\0.1 to include or exclude in

the model. Age, body surface area (BSA), weight, height,

and regimen schedule were entered as continuous vari-

ables. The following covariates were entered as categorical

variables: body mass index (BMI; \25, 25–30, C30),

number of positive nodes (0, 1–3, C4), tumor hormone

receptor status [estrogen receptor (ER) negative and pro-

gesterone (PR) negative, ER? and/or PR?, unknown],

post menopausal status (no, yes, unknown), Her2 overex-

pression (no, yes, unknown), histology (carcinoma—not

otherwise specified, Paget disease, undifferentiated carci-

noma; ductal; lobular), disease stage (I–IIa, IIb–IIIa), ANC

(\2, C2), hemoglobin count (\110, C110), platelet count

(\150, C150), G-CSF use (yes, no), and use of cyclophos-

phamide, doxorubicin, docetaxel, paclitaxel, or carboplatin

(yes, no). Covariates were removed if a convergence prob-

lem was detected.

Study 1 data collection

Data collection from Study 1 has been described [20]. In

brief, in Study 1, charts were reviewed from 20,799

patients with ESBC who were treated with adjuvant che-

motherapy in the community setting from August 1997 to

May 2000. RDI was calculated relative to reference stan-

dards based on the medical literature and practice guide-

lines. Data were collected in a similar manner for Study 2.

Results from both studies are described.

Results

Patient characteristics and chemotherapy regimens

This study included 532 female patients. Mean patient age

was 55.0 years (range: 29–85 years), and 22.0 % of

patients (n = 117) were C65 years. Mean BSA was

1.84 m2 (range: 1.3–3.0 m2). ECOG performance status

was known in 256 patients (48.1 %). Of these patients,

most had an ECOG performance status of 0 or 1 (n = 252,

98.4 %), and 4 (1.6 %) had an ECOG performance status

C2. Baseline disease characteristics are shown in Table 1.

A majority of patients received a taxane-based regimen

(89.1 %). The most common chemotherapy regimen was

TC (n = 221; 41.5 %). See Supplemental Table 2 for

additional regimens. Across all regimens, most patients

received the full number of planned cycles, which was

largely identical to the standard number of cycles as

defined by NCCN guidelines [22] and medical literature

(Supplemental Table 2).

Breast Cancer Res Treat (2013) 139:863–872 865

123

Relative dose intensity

Mean RDI ranged from 82.8 to 95.5 % across all regimens

(see Fig. 1 for the five most common chemotherapy regi-

mens). Overall, 83.8 % of patients received an RDI C85 %,

while 16.2 % received an RDI \85 % (Fig. 2), and 25.0 %

received an RDI \ 90 %. RDI was generally highest in

cycle 1 (95.3–97.3 %) and decreased with subsequent

cycles. For TC, the most common regimen, overall mean

(95 % CI) RDI was 93.0 % (91.1–94.9), while 85.5 % of

patients received an RDI C85 %, 14.5 % of patients

received an RDI \85 %, and 19.0 % received an

RDI \ 90 %. Similar to trends across all regimens, mean

(95 % CI) RDI for TC was 96.9 % (95.6–98.1) in cycle 1 and

decreased to 83.8 % (74.4–93.3) in cycle 6. In the AC fol-

lowed by T regimens, RDI was high in cycle 1 and decreased

gradually through cycle 4. In cycle 5, when most patients

began paclitaxel, RDI was generally high again and

remained elevated through the remainder of cycles.

RDIs were equivalent between patients \65 years and

patients C65 years for most regimens (Fig. 1). Across all

regimens, 15.4 % of patients \65 years received an

RDI \85 %, and 18.8 % of patients C65 years received an

RDI \85 %. Similarly, 23.9 % of patients \65 years

received an RDI \ 90 %, and 29.1 % of patients

C65 years received an RDI \ 90 %.

Most patients had a BSA B 2 m2 (n = 442; 83.1 %),

and 90 patients (16.9 %) had a BSA [ 2 m2. Mean RDI

was slightly lower among patients with a BSA [ 2 m2 than

patients with a BSA B 2 m2; 22.2 % of patients with a

BSA [ 2 m2 received an RDI \85 %, and 14.9 % of

patients with a BSA B 2 m2 received an RDI \85 %.

Similarly, 31.1 % of patients with a BSA [ 2 m2 received

an RDI \ 90 %, and 23.8 % of patients with a

BSA B 2 m2 received an RDI \ 90 %.

Multivariate analysis

To obtain model convergence, HER2 status and patients

with unknown hormone receptor status were excluded.

Covariates associated with RDI \85 % included weight,

hormone receptor status, disease stage, histology, platelet

count, and primary G-CSF prophylaxis (see Table 2).

Increased risk of RDI \85 % was associated with higher

weight and stage IIb–IIIa disease. Reduced risk of

RDI \85 % was associated with ductal or lobular histol-

ogy, higher platelet count, ER? and/or PR? hormone

receptor status, and use of primary G-CSF prophylaxis.

Dose delays and reductions

Overall, 85 patients (16.0 %) had at least one dose delay

C7 days and 111 patients (20.9 %) had at least one dose

reduction C15 % from NCCN standard. Dose delays and

dose reductions were observed in every regimen studied.

Dose delays were most frequently observed in the TCH

regimen, where 15 of 56 patients (26.8 %) receiving TCH

experienced a dose delay. Dose delays were least

Table 1 Baseline disease characteristics and treatment regimens

N = 532 n (%)

Disease stage

I 194 (36.5)

IIa 197 (37.0)

IIb 90 (16.9)

IIIa 51 (9.6)

Histology

Carcinoma, NOS 20 (3.8)

Ductal 453 (85.2)

Lobular 49 (9.2)

Undifferentiated carcinoma 10 (1.9)

Number of positive nodes

0 278 (52.3)

1–3 195 (36.7)

4? 45 (8.5)

Unknown 14 (2.6)

Post-menopausal

No 158 (29.7)

Yes 260 (48.9)

Unknown 114 (21.4)

Tumor hormone receptor positive

ER-/PR- 141 (26.5)

ER? and/or PR? 387 (72.7)

Unknown 4 (0.8)

HER-2 overexpression

No 421 (79.1)

Yes 103 (19.4)

Unknown 8 (1.5)

Treatment regimens

TC 221 (41.5)

AC 58 (10.9)

TCH 56 (10.5)

TAC 34 (6.4)

AC-Tweekly 22 (4.1)

AC-T14 83 (15.6)

AC14-Tweekly 54 (10.2)

AC-T 4 (0.8)

NOS not otherwise specified, ER estrogen receptor, PR progesterone

receptor, HER-2, Human Epidermal Growth Factor Receptor 2, TC

docetaxel ? cyclophosphamide, AC doxorubicin ? cyclophospha-

mide, TCH docetaxel ? carboplatin ? trastuzumab, TAC doce-

taxel ? doxorubicin ? cyclophosphamide, AC-Tweekly 21-day AC

followed by weekly paclitaxel, AC-T14 14-day AC followed by

14-day paclitaxel, AC14-Tweekly 14-day AC followed by weekly

paclitaxel, AC-T 21-day AC followed by 21-day paclitaxel

866 Breast Cancer Res Treat (2013) 139:863–872

123

Overall < 65 years 65 years

0%

20%

40%

60%

80%

100%

TC AC-T14 AC TCH AC14-Tweekly

Mea

n R

DI

Overall N = 221 N = 83 N = 58 N = 56 N = 54 N = 532 < 65 years N = 160 N = 73 N = 49 N = 41 N = 46N = 415 65 years N = 61 N = 10 N = 9 N = 15 N = 8N =117

<

<

Fig. 1 Relative dose intensity

by regimen and age. Mean RDI

and 95 % CIs are plotted for the

five most common

chemotherapy regimens

0%

5%

10%

15%

20%

25%

30%

35%

40%

45%

Overall N = 532

Cycle 1 N = 532

Cycle 2 N = 527

Cycle 3 N = 522

Cycle 4 N = 516

Cycle 5 N = 287

Cycle 6 N = 278

% o

f Pat

ient

s E

xper

ienc

ing

Eve

nt

Dose Delays Dose Reduction RDI < 85%

Fig. 2 Dose delays, dose

reductions, and RDI \85 % by

cycle. The incidences of dose

delays and dose reductions and

the percentage of patients who

received an RDI \85 % are

plotted with 95 % CIs. Overall

includes data from all cycles of

chemotherapy. Details for

cycles 1–6 are graphed

Table 2 Multivariate analysis

of covariates for RDI \85 %

OR odds ratio, CL confidence

limit, PR progesterone receptor,

G-CSF granulocyte colony-

stimulating factor, NOS not

otherwise specified

Covariate Point estimate

of OR

95 % Wald

CL

Weight 1.006 1.000–1.011

ER? and/or PR? 0.606 0.351–1.045

Stage IIb–IIIa disease 1.924 1.086–3.408

Disease histology: lobular vs carcinoma NOS, paget

disease, or undifferentiated carcinoma

0.297 0.085–1.035

Disease histology: ductal vs carcinoma NOS, paget

disease, or undifferentiated carcinoma

0.244 0.093–0.636

Platelet count C 150 0.250 0.050–1.243

Primary G-CSF prophylaxis 0.367 0.212–0.636

Breast Cancer Res Treat (2013) 139:863–872 867

123

frequently observed in the TC regimen, where 18 of 221

patients (8.1 %) receiving TC experienced a dose delay.

Dose reductions were most frequently observed among

patients receiving AC-Tweekly, where 16 of 22 patients

(72.7 %) had at least one dose reduction; however, inter-

preting this result can be difficult given the small number

of patients receiving AC-Tweekly. The regimen with the

next highest number of observed dose reductions was TCH,

where 17 of 56 patients (30.4 %) experienced at least one

dose reduction. Dose reductions were least frequently

observed among patients receiving AC, where 4 of 58

patients (6.9 %) had at least one dose reduction.

Patients C65 years were more likely to have dose delays

and dose reductions. At least one dose delay occurred in 28

patients (23.9 %) C65 years and 57 patients (13.7 %)

\65 years. Similarly, at least one dose reduction occurred

in 29 patients (24.8 %) C65 years and 82 patients (19.8 %)

\65 years.

The incidences of dose delays and dose reductions were

similar across cycles (Fig. 2). Slightly more dose delays

were seen in cycle 5 (n = 20; 7.0 %), while slightly more

dose reductions were seen in cycle 1 (n = 36; 6.8 %) and

cycle 5 (n = 24; 8.4 %).

Reasons for dose delays were attributed to other reasons

related to chemotherapy (n = 23) or not related to chemo-

therapy (n = 41), personal/patient request (n = 17), neu-

tropenia (n = 8), and other hematological reasons (n = 3).

Reasons for dose reductions were attributed to other reasons

related to chemotherapy (n = 50), weight change (n = 6),

neutropenia (n = 6), thrombocytopenia (n = 4), anemia

(n = 1), and dose administration error (n = 0).

Neutropenia and febrile neutropenia

Across all cycles, 215 patients (40.4 %) had at least one

episode of grade 3/4 neutropenia, while 18 patients (3.4 %)

had at least one episode of grade 3/4 FN. Patients

\65 years were slightly less likely to have grade 3/4

neutropenia (n = 163, 39.3 %) than patients C65 years

(n = 52, 44.4 %). Patients \65 years were equally likely

to experience grade 3/4 FN (n = 14, 3.4 %) as patients

C65 years (n = 4; 3.4 %). The incidence of grade 3/4

neutropenia was the highest in cycle 1 (n = 144; 27.1 %),

was near 20 % for cycles 2–4 (18.4–19.7 %), and lower in

cycles 5–8 (7.9–0.9 %). Grade 3/4 FN was similar across

all cycles, ranging from 0 to 1 % (Supplemental Table 3).

Grade 3/4 neutropenia was relatively common in all che-

motherapy regimens, ranging from 25.0 to 63.6 %

(Table 3). FN was most common with AC chemotherapy;

10.3 % (n = 6) of patients receiving AC experienced grade

3/4 FN and 6.9 % (n = 4) experienced grade 4 FN

(Table 3). Overall, nine patients (1.7 %) were reported to

be hospitalized with FN.

Supportive care

Overall, 481 patients (90.4 %) received G-CSF at some

point, and 75.0 % of chemotherapy cycles administered

were supported with G-CSF. For TC, the most common

chemotherapy regimen in this study, 88.7 % of patients

(n = 196) received G-CSF at some point and 73.5 % of

these patients (n = 144) received primary prophylactic

G-CSF. G-CSF use in additional chemotherapy regimens is

shown in Supplemental Table 4.

Pegfilgrastim was the most common G-CSF used in this

study; 398 patients (74.8 %) received only pegfilgrastim,

34 patients (6.4 %) received only filgrastim, and 49

patients (9.2 %) received both filgrastim and pegfilgrastim

over the course of their chemotherapy. Pegfilgrastim was

administered only as prophylaxis with 90.2 % of patients

(n = 359) receiving primary prophylaxis and 9.8 %

(n = 39) receiving secondary prophylaxis (Table 4).

Table 3 Neutropenia by regimen

Regimen Grade 3/4 neutropenia n (%) Grade 3/4 FN n (%) Grade 4 FN n (%)

TC (N = 221) 73 (33.0) 6 (2.7) 3 (1.4)

AC (N = 58) 34 (58.6) 6 (10.3) 4 (6.9)

TCH (N = 56) 20 (35.7) 2 (3.6) 1 (1.8)

TAC (N = 34) 15 (44.1) 2 (5.9) 2 (5.9)

AC-Tweekly (N = 22) 14 (63.6) 0 (0) 0 (0)

AC-T14 (N = 83) 34 (41.0) 0 (0) 0 (0)

AC14-Tweekly (N = 54) 24 (44.4) 2 (3.7) 2 (3.7)

AC-T (N = 4) 1 (25.0) 0 (0) 0 (0)

FN febrile neutropenia, TC docetaxel ? cyclophosphamide, AC doxorubicin ? cyclophosphamide, TCH docetaxel ? carboplatin ? trast-

uzumab, TAC docetaxel ? doxorubicin ? cyclophosphamide, AC-Tweekly 21-day AC followed by 21-day paclitaxel at 80 mg/m2 given on day 1,

8, and 15, AC-T14 14-day AC followed by 14-day paclitaxel at 175 mg/m2, AC14-Tweekly 14-day AC followed by 21-day paclitaxel at 80 mg/m2

given on day 1, 8, and 15, AC-T 21-day AC followed by 21-day paclitaxel at 175 mg/m2

868 Breast Cancer Res Treat (2013) 139:863–872

123

Filgrastim was administered both as prophylaxis and

treatment with 23.5 % of patients (n = 8) receiving pri-

mary prophylaxis, 38.2 % (n = 13) receiving secondary

prophylaxis, and 38.2 % (n = 13) receiving treatment

(Table 4).

Pegfilgrastim alone was administered to 312 patients

\65 years and 86 patients C65 years. Pegfilgrastim use

was similar among patients \65 years and patients

C65 years for both primary prophylaxis (91.0 vs 87.2 %)

and secondary prophylaxis (9.0 vs 12.8 %). Filgrastim

alone was administered to 25 patients \65 years and nine

patients C65 years. Filgrastim use was also similar among

patients \65 years and patients C65 years for primary

prophylaxis (20.0 vs 33.3 %), secondary prophylaxis (40.0

vs 33.3 %), and treatment (40.0 vs 33.3 %).

Overall, 160 patients (30.1 %) received at least one

antimicrobial agent; and 146 patients (27.4 %) received an

antimicrobial and G-CSF at some point over the chemo-

therapy course. Overall, 130 patients (24.4 %) received

oral agents only, 29 patients (5.5 %) received intravenous

agents with or without oral agents, and one patient received

a topical antimicrobial. Antimicrobial use was initiated

during day 1–5 of any chemotherapy cycle for 44 patients

(8.3 %), during day 6–10 of any chemotherapy cycle for 87

patients (16.4 %), and after day 10 of any chemotherapy

cycle for 76 patients (14.3 %). Antimicrobial use was

similar across cycles 1–3 (9.3–11.1 %) and lower across

subsequent cycles (3.9–7.4 %). Quinolones were the most

common antimicrobial (n = 108, 20.3 %).

Treatment patterns over time

Changes over time between Study 1 and Study 2 can be

described since 79 % of the sites in Study 2 also partici-

pated in Study 1 (Table 5). Baseline patient demographics

and disease characteristics were similar between Study 1

and Study 2, though patients in Study 2 were slightly older

and fewer patients had lymph node-positive disease. Con-

sistent with findings of Giordano et al. [21], we noted a

marked increase in taxane-based regimens, which were

administered to\4 % of patients in Study 1 and to 89.1 %

of patients in Study 2. Relative to Study 1, fewer patients

received an RDI \85 % (16.2 vs 55.5 %), and fewer dose

delays (16.0 vs 24.9 %) and dose reductions (20.9 vs

36.5 %) were seen. In addition, G-CSF use increased from

Table 4 G-CSF use

Pegfilgrastim only (N = 398) Filgrastim only (N = 34) Overalla (N = 481)

Prophylactic use—n (%) 398 (100.0) 21 (61.8) 467 (97.1)

Primary 359 (90.2) 8 (23.5) 404 (84.0)

Secondaryb 39 (9.8) 13 (38.2) 63 (13.1)

Treatment use—n (%) 0 (0) 13 (38.2) 14 (2.9)

51 patients who did not receive either pegfilgrastim or filgrastim are excluded from this tablea Includes 49 patients who received both pegfilgrastim and filgrastimb Includes 23 patients (4.8 %) who initially received treatment G-CSF followed by secondary prophylaxis

Table 5 Key patient and treatment characteristics in Study 1 and

Study 2

Study 1

N = 19,898

Study 2

N = 532

Age, mean 52 55

\65 years 83 % 78 %

BSA, mean 1.83 m2 1.84 m2

Lymph node-positive disease 52.4 % 45.1 %

Chemotherapy regimena

CMFb 43 % N/A

CAFb 19 % N/A

AC 34 % 11 %

AC followed by T \4 % 31 %

TC N/A 42 %

TAC N/A 6 %

TCH N/A 11 %

RDI \85 % 55.5 % 16.2 %

Dose delay C7 days 24.9 % 16.0 %

Dose reduction C15 % 36.5 % 20.9 %

G-CSF administration

Any 26.4 % 90.4 %

Primary prophylactic *3 %c 75.9 %

BSA body surface area, RDI relative dose intensity, CMF cyclophos-

phamide ? methotrexate ? fluorouracil, CAF cyclophosphamide ?

doxorubicin ? fluorouracil, AC doxorubicin ? cyclophosphamide, TC

docetaxel ? cyclophosphamide, TAC docetaxel ? doxorubicin ? cycl-

ophosphamide, TCH docetaxel ? carboplatin ? trastuzumab, N/A not

applicable, G-CSF granulocyte-colony-stimulating factora A small number of patients in Study 1 received doxorubicin fol-

lowed by cyclophosphamide ? methotrexate ? fluorouracil or

doxorubicin followed by paclitaxel followed by cyclophosphamideb Includes both 21- and 28-day dosing schedulesc In Study 1, prophylactic use was defined as use within 3 days of

chemotherapy completion

Breast Cancer Res Treat (2013) 139:863–872 869

123

Study 1 to Study 2. In Study 1, 26.4 % of patients received

G-CSF at some point (vs 90.4 % in Study 2) and *3 %

received primary prophylactic G-CSF (vs 75.9 % in Study

2). Rates of neutropenia, FN, FN-related hospitalization,

and antimicrobial use were not reported in Study 1.

Discussion

Treatment for ESBC has evolved over the past decade, as

reflected in this retrospective analysis of 532 patients

treated with adjuvant chemotherapy in US community- and

hospital-based outpatient oncology practices. Notable

changes include higher mean RDI, greater use of prophy-

lactic G-CSF, and a shift towards taxane-based chemo-

therapy regimens.

In the past decade, a growing body of work has dem-

onstrated that high RDI improves patient outcomes in

ESBC [17–19, 23]. Historically, older patients and obese

patients have been given lower doses of chemotherapy

from the onset of treatment; indeed, both advanced age and

higher BMI were risk factors for RDI \85 % in Study 1. In

this study, mean RDI was high across all regimens exam-

ined. Also, little difference in RDI was seen between

patients\65 years and patients C65 years, and age was not

associated with risk for RDI \85 %. Though most cancer

patients are C65 years, these patients are underrepresented

in clinical trials [24], leading to a paucity of clinical data

on the efficacy and safety of treatments in the elderly.

However, healthy patients C65 years can derive the same

benefits from chemotherapy as younger patients [25].

Finally, in this study, RDI was calculated to be lower

among patients with a BSA [ 2 m2 than among patients

with BSAs B 2 m2. Though BMI was not associated with

risk for RDI \85 %, weight was a risk factor. Full-dose

chemotherapy has been shown to be safe and effective in

overweight patients; however, chemotherapy doses are

routinely ‘‘capped,’’ which can lead to underdosing of

obese patients. Recent ASCO guidelines, which were

released after this study was conducted, recommend that

full weight-based cytotoxic chemotherapy doses can be

used to treat obese patients with cancer, particularly in the

curative setting [26]. The data reported here suggest that

disparities among the treatment of different age groups

have diminished over the past decade, but undertreatment

of obese patients may still persist.

Consistent with higher RDI, greater compliance with

NCCN-standard doses and number of cycles for the treat-

ment of breast cancer [22] was also seen. In Study 1, 24 %

of patients had a planned dose reduction. In this study,

cycle 1 dose reductions, which were likely planned, were

seen in 7 % of patients. Similarly, the median number of

planned cycles was identical to the NCCN-standard

number of cycles for each chemotherapy regimen. Though

compliance has improved, 16.2 % of patients received an

RDI \85 %.

As guidelines for the use of adjuvant chemotherapy in

breast cancer have evolved, chemotherapy has been rec-

ommended for an expanding population of patients [27,

28]. Correspondingly, the rate of use of adjuvant chemo-

therapy appears to have increased over time, at least in

certain populations of patients [29, 30]. Currently, the

NCCN-preferred regimens for adjuvant chemotherapy for

ESBC are AC followed by paclitaxel, TC, and TCH, all

taxane-containing regimens [22]. In Study 1, taxane-based

therapies were just emerging, and \4 % of patients

received a taxane. Study 2 reflects the shift in practice

toward taxane-based regimens, with 89.1 % of patients

receiving a regimen that contained a taxane.

Taxane-based regimens are largely considered the

standard of care for patients with ESBC based on a number

of pivotal trials. In a phase 3 trial comparing TAC with

fluorouracil, doxorubicin, and cyclophosphamide (FAC),

TAC significantly decreased the risk of relapse and death

compared to FAC [9]. However, the incidence of FN was

24.7 % in patients treated with TAC compared to 2.5 % in

patients treated with FAC [9]. In a phase 3 trial that showed

that TC was superior to AC [31], the incidence of FN in the

TC arm was reported as 5 %; however, the use of pro-

phylactic antibiotics was widespread, and the use of G-CSF

was not reported [31]. In recent retrospective studies, the

incidence of FN in patients receiving TC ranged from 0 to

6.3 % in patients who received G-CSF and 25 to 50 % in

patients who did not receive G-CSF [32–34]. Consistent

with studies where G-CSF use was common, we found that

the incidence of FN in patients receiving TC was 2.7 %.

Current guidelines recommend the use of prophylactic

CSFs in patients with a C20 % risk of FN and that the use

of prophylactic CSFs be considered in patients with

10–20 % risk of FN [35–37] when no other equally

effective regimen that does not require CSFs is available

[36]. In the most recent NCCN guidelines on breast cancer,

TC is listed as a preferred regimen for invasive breast

cancer with G-CSF support in all cycles [22]. Therefore,

increased G-CSF use may, in part, be a response to the shift

toward the use of newer, taxane-containing chemotherapy

regimens which may be more myelotoxic, as well as a

greater awareness of the risk of FN associated with these

regimens.

This study has a number of limitations that are inherent

to retrospective analyses where data are collected from

patient charts and electronic medical record (EMR) dat-

abases. Transfer of inpatient data into the EMR is not

always complete; thus some data may be underreported. In

addition, hospitalizations were only captured for patients

who were hospitalized with FN. Hospitalizations for other

870 Breast Cancer Res Treat (2013) 139:863–872

123

causes were not captured. Reasons for dose delays and dose

reductions were captured as broad categories, and deter-

mining underlying causes of dose delays and dose reduc-

tions was difficult. For example, most dose delays and

reductions were attributed to ‘‘other’’ reasons. Apart from

this limitation, 17 patients had a dose delay due to per-

sonal/patient request. Though this category could encom-

pass several underlying reasons, continued physician and

patient education on the importance of maintaining high

RDI may further increase RDI and ultimately improve

patient outcomes.

In summary, this study describes a shift in the past decade

to higher RDI across all patient populations regardless of

age. Though survival outcomes are not available from the

previous study or this study, the data shown here indicates

maintaining full-dose chemotherapy on the planned sche-

dule has become an important goal in clinical practice. In

ESBC, maintaining high RDI is associated with improved

disease-free survival and overall survival [17, 18, 23]. Most

patients in this study received an RDI C85 %, indicating

that patient care has continued to improve in the past decade.

Acknowledgments The authors thank Greg Valin, Sharon Hunter,

Natasha Gicanov, Paul Chang, and Sejal Badre (Amgen Inc.) for their

contributions to this study. Kerri Hebard-Massey, PhD (Amgen Inc.)

provided writing assistance. This study was funded by Amgen Inc.

Conflict of interest GHL is principal investigator of a research

grant to Duke University from Amgen in support of the ANC Study

Group. DCD and JC received research funding from and are on an

advisory board of Amgen Inc. DT and SW are employees of and

stockholders in Amgen Inc.

References

1. Bonadonna G, Brusamolino E, Valagussa P, Rossi A, Brugnatelli

L, Brambilla C, De Lena M, Tancini G, Bajetta E, Musumeci R,

Veronesi U (1976) Combination chemotherapy as an adjuvant

treatment in operable breast cancer. N Engl J Med 294(8):

405–410. doi:10.1056/NEJM197602192940801

2. Rossi A, Bonadonna G, Valagussa P, Veronesi U (1981) Multi-

modal treatment in operable breast cancer: five-year results of the

CMF programme. Br Med J (Clin Res Ed) 282(6274):1427–1431

3. EBCTCG (2011) Comparisons between different polychemo-

therapy regimens for early breast cancer: meta-analyses of long-

term outcome among 100,000 women in 123 randomised trials.

Lancet 379(9814):4–10

4. Fisher B, Brown AM, Dimitrov NV, Poisson R, Redmond C,

Margolese RG, Bowman D, Wolmark N, Wickerham DL, Kar-

dinal CG et al (1990) Two months of doxorubicin–cyclophos-

phamide with and without interval reinduction therapy compared

with 6 months of cyclophosphamide, methotrexate, and fluoro-

uracil in positive-node breast cancer patients with tamoxifen-

nonresponsive tumors: results from the National Surgical Adju-

vant Breast and Bowel Project B-15. J Clin Oncol 8(9):

1483–1496

5. Fisher B, Jeong JH, Anderson S, Wolmark N (2004) Treatment of

axillary lymph node-negative, estrogen receptor-negative breast

cancer: updated findings from National Surgical Adjuvant Breast

and Bowel Project clinical trials. J Natl Cancer Inst

96(24):1823–1831. doi:10.1093/jnci/djh338

6. Jones SE, Moon TE, Bonadonna G, Valagussa P, Rivkin S,

Buzdar A, Montague E, Powles T (1987) Comparison of different

trials of adjuvant chemotherapy in stage II breast cancer using a

natural history data base. Am J Clin Oncol 10(5):387–395

7. Peto R, Davies C, Godwin J, Gray R, Pan HC, Clarke M, Cutter

D, Darby S, McGale P, Taylor C, Wang YC, Bergh J, Di Leo A,

Albain K, Swain S, Piccart M, Pritchard K (2012) Comparisons

between different polychemotherapy regimens for early breast

cancer: meta-analyses of long-term outcome among 100,000

women in 123 randomised trials. Lancet 379(9814):432–444. doi:

10.1016/S0140-6736(11)61625-5

8. Henderson IC, Berry DA, Demetri GD, Cirrincione CT, Gold-

stein LJ, Martino S, Ingle JN, Cooper MR, Hayes DF, Tkaczuk

KH, Fleming G, Holland JF, Duggan DB, Carpenter JT, Frei E

3rd, Schilsky RL, Wood WC, Muss HB, Norton L (2003)

Improved outcomes from adding sequential paclitaxel but not

from escalating doxorubicin dose in an adjuvant chemotherapy

regimen for patients with node-positive primary breast cancer.

J Clin Oncol 21(6):976–983

9. Martin M, Pienkowski T, Mackey J, Pawlicki M, Guastalla JP,

Weaver C, Tomiak E, Al-Tweigeri T, Chap L, Juhos E, Guevin

R, Howell A, Fornander T, Hainsworth J, Coleman R, Vinholes J,

Modiano M, Pinter T, Tang SC, Colwell B, Prady C, Provencher

L, Walde D, Rodriguez-Lescure A, Hugh J, Loret C, Rupin M,

Blitz S, Jacobs P, Murawsky M, Riva A, Vogel C (2005) Adju-

vant docetaxel for node-positive breast cancer. N Engl J Med

352(22):2302–2313. doi:10.1056/NEJMoa043681

10. Jones SE, Savin MA, Holmes FA, O’Shaughnessy JA, Blum JL,

Vukelja S, McIntyre KJ, Pippen JE, Bordelon JH, Kirby R,

Sandbach J, Hyman WJ, Khandelwal P, Negron AG, Richards

DA, Anthony SP, Mennel RG, Boehm KA, Meyer WG, Asmar L

(2006) Phase III trial comparing doxorubicin plus cyclophos-

phamide with docetaxel plus cyclophosphamide as adjuvant

therapy for operable breast cancer. J Clin Oncol 24(34):

5381–5387. doi:10.1200/JCO.2006.06.5391

11. Citron ML, Berry DA, Cirrincione C, Hudis C, Winer EP, Gra-

dishar WJ, Davidson NE, Martino S, Livingston R, Ingle JN,

Perez EA, Carpenter J, Hurd D, Holland JF, Smith BL, Sartor CI,

Leung EH, Abrams J, Schilsky RL, Muss HB, Norton L (2003)

Randomized trial of dose-dense versus conventionally scheduled

and sequential versus concurrent combination chemotherapy as

postoperative adjuvant treatment of node-positive primary breast

cancer: first report of Intergroup Trial C9741/Cancer and Leu-

kemia Group B Trial 9741. J Clin Oncol 21(8):1431–1439

12. EBCTCG (2005) Effects of chemotherapy and hormonal therapy

for early breast cancer on recurrence and 15-year survival: an

overview of the randomised trials. Lancet 365(9472):1687–1717

13. Berry DA, Cronin KA, Plevritis SK, Fryback DG, Clarke L,

Zelen M, Mandelblatt JS, Yakovlev AY, Habbema JD, Feuer EJ

(2005) Effect of screening and adjuvant therapy on mortality

from breast cancer. N Engl J Med 353(17):1784–1792. doi:

10.1056/NEJMoa050518

14. DeSantis C, Siegel R, Bandi P, Jemal A (2011) Breast cancer

statistics. CA Cancer J Clin 61(6):409–418. doi:10.3322/caac.

20134

15. American Cancer Society (2012) Cancer Facts and Figures 2012

16. Hryniuk WM, Figueredo A, Goodyear M (1987) Applications of

dose intensity to problems in chemotherapy of breast and colo-

rectal cancer. Semin Oncol 14(4 Suppl 4):3–11

17. Budman DR, Berry DA, Cirrincione CT, Henderson IC, Wood

WC, Weiss RB, Ferree CR, Muss HB, Green MR, Norton L, Frei

Breast Cancer Res Treat (2013) 139:863–872 871

123

E 3rd (1998) Dose and dose intensity as determinants of outcome

in the adjuvant treatment of breast cancer. The Cancer and

Leukemia Group B. J Natl Cancer Inst 90(16):1205–1211. doi:

10.1093/jnci/90.16.1205

18. Chirivella I, Bermejo B, Insa A, Perez-Fidalgo A, Magro A,

Rosello S, Garcia-Garre E, Martin P, Bosch A, Lluch A (2009)

Optimal delivery of anthracycline-based chemotherapy in the

adjuvant setting improves outcome of breast cancer patients.

Breast Cancer Res Treat 114(3):479–484. doi:10.1007/s10549-

008-0018-1

19. Lyman GH (2009) Impact of chemotherapy dose intensity on

cancer patient outcomes. J Natl Compr Cancer Netw 7(1):99–108

20. Lyman GH, Dale DC, Crawford J (2003) Incidence and predic-

tors of low dose-intensity in adjuvant breast cancer chemother-

apy: a nationwide study of community practices. J Clin Oncol

21(24):4524–4531. doi:10.1200/JCO.2003.05.002

21. Giordano SH, Lin YL, Kuo YF, Hortobagyi GN, Goodwin JS

(2012) Decline in the use of anthracyclines for breast cancer.

J Clin Oncol 30(18):2232–2239. doi:10.1200/JCO.2011.40.1273

22. NCCN Clinical Practice Guidelines in Oncology: Breast Cancer.

v3.2013

23. Wildiers H, Reiser M (2011) Relative dose intensity of chemo-

therapy and its impact on outcomes in patients with early breast

cancer or aggressive lymphoma. Crit Rev Oncol Hematol

77(3):221–240. doi:10.1016/j.critrevonc.2010.02.002

24. Kemeny MM, Peterson BL, Kornblith AB, Muss HB, Wheeler J,

Levine E, Bartlett N, Fleming G, Cohen HJ (2003) Barriers to

clinical trial participation by older women with breast cancer.

J Clin Oncol 21(12):2268–2275. doi:10.1200/JCO.2003.09.124

25. Muss HB, Woolf S, Berry D, Cirrincione C, Weiss RB, Budman

D, Wood WC, Henderson IC, Hudis C, Winer E, Cohen H,

Wheeler J, Norton L (2005) Adjuvant chemotherapy in older and

younger women with lymph node-positive breast cancer. JAMA

293(9):1073–1081. doi:10.1001/jama.293.9.1073

26. Griggs JJ, Mangu PB, Anderson H, Balaban EP, Dignam JJ,

Hryniuk WM, Morrison VA, Pini TM, Runowicz CD, Rosner

GL, Shayne M, Sparreboom A, Sucheston LE, Lyman GH (2012)

Appropriate chemotherapy dosing for obese adult patients with

cancer: American Society of Clinical Oncology clinical practice

guideline. J Clin Oncol 30(13):1553–1561. doi:10.1200/JCO.

2011.39.9436

27. (1985) Consensus conference. Adjuvant chemotherapy for breast

cancer. JAMA 254(24):3461–3463

28. Eifel P, Axelson JA, Costa J, Crowley J, Curran WJ Jr, Deshler

A, Fulton S, Hendricks CB, Kemeny M, Kornblith AB, Louis TA,

Markman M, Mayer R, Roter D (2001) National Institutes of

Health Consensus Development Conference Statement: adjuvant

therapy for breast cancer, November 1–3, 2000. J Natl Cancer

Inst 93(13):979–989

29. Neugut AI, Hillyer GC, Kushi LH, Lamerato L, Leoce N, Na-

thanson SD, Ambrosone CB, Bovbjerg DH, Mandelblatt JS,

Magai C, Tsai WY, Jacobson JS, Hershman DL (2012) Nonini-

tiation of adjuvant chemotherapy in women with localized breast

cancer: the breast cancer quality of care study. J Clin Oncol

30(31):3800–3809. doi:10.1200/JCO.2012.43.8168

30. Shen Y, Dong W, Feig BW, Ravdin P, Theriault RL, Giordano

SH (2009) Patterns of treatment for early stage breast cancers at

the M. D. Anderson Cancer Center from 1997 to 2004. Cancer

115(10):2041. doi:10.1002/cncr.24271

31. Jones S, Holmes FA, O’Shaughnessy J, Blum JL, Vukelja SJ,

McIntyre KJ, Pippen JE, Bordelon JH, Kirby RL, Sandbach J,

Hyman WJ, Richards DA, Mennel RG, Boehm KA, Meyer WG,

Asmar L, Mackey D, Riedel S, Muss H, Savin MA (2009)

Docetaxel with cyclophosphamide is associated with an overall

survival benefit compared with doxorubicin and cyclophospha-

mide: 7-year follow-up of US Oncology Research Trial 9735.

J Clin Oncol 27(8):1177–1183

32. Chan A, Fu WH, Shih V, Coyuco JC, Tan SH, Ng R (2011)

Impact of colony-stimulating factors to reduce febrile neutro-

penic events in breast cancer patients receiving docetaxel plus

cyclophosphamide chemotherapy. Support Care Cancer 19(4):

497–504. doi:10.1007/s00520-010-0843-8

33. Soong D, Haj R, Leung MG, Myers R, Higgins B, Myers J,

Rajagopal S (2009) High rate of febrile neutropenia in patients

with operable breast cancer receiving docetaxel and cyclophos-

phamide. J Clin Oncol 27(26):e101–e102. doi:10.1200/JCO.

2009.23.0508

34. Vandenberg T, Younus J, Al-Khayyat S (2010) Febrile neutro-

penia rates with adjuvant docetaxel and cyclophosphamide che-

motherapy in early breast cancer: discrepancy between published

reports and community practice—a retrospective analysis. Curr

Oncol 17(2):2–3

35. Aapro MS, Bohlius J, Cameron DA, Dal Lago L, Donnelly JP,

Kearney N, Lyman GH, Pettengell R, Tjan-Heijnen VC,

Walewski J, Weber DC, Zielinski C (2011) 2010 update of EO-

RTC guidelines for the use of granulocyte-colony stimulating

factor to reduce the incidence of chemotherapy-induced febrile

neutropenia in adult patients with lymphoproliferative disorders

and solid tumours. Eur J Cancer 47(1):8–32. doi:10.1016/

j.ejca.2010.10.013

36. Smith TJ, Khatcheressian J, Lyman GH, Ozer H, Armitage JO,

Balducci L, Bennett CL, Cantor SB, Crawford J, Cross SJ, De-

metri G, Desch CE, Pizzo PA, Schiffer CA, Schwartzberg L,

Somerfield MR, Somlo G, Wade JC, Wade JL, Winn RJ, Woz-

niak AJ, Wolff AC (2006) 2006 Update of recommendations for

the use of white blood cell growth factors: an evidence-based

clinical practice guideline. J Clin Oncol 24(19):3187–3205. doi:

10.1200/JCO.2006.06.4451

37. NCCN Clinical Practice Guidelines in Oncology: Myeloid

Growth Factors. v1.2012

872 Breast Cancer Res Treat (2013) 139:863–872

123