Download Breast Cancer–Specific mRNA Transcripts Presence in Peripheral

VOLUME
24
䡠
NUMBER
22
䡠
AUGUST
1
2006
JOURNAL OF CLINICAL ONCOLOGY
O R I G I N A L
R E P O R T
Breast Cancer–Specific mRNA Transcripts Presence in
Peripheral Blood After Adjuvant Chemotherapy Predicts
Poor Survival Among High-Risk Breast Cancer Patients
Treated With High-Dose Chemotherapy With Peripheral
Blood Stem Cell Support
Miguel Quintela-Fandino, Joaquín Martínez López, Ricardo Hitt, Soledad Gamarra, Antonio Jimeno,
Rosa Ayala, Javier Hornedo, Cecilia Guzman, Florinda Gilsanz, and Hernán Cortés-Funes
From the Medical Oncology Department; Molecular Biology Division,
Hematology Department, University
Hospital 12 de Octubre; Roche Pharma,
Madrid, Spain; and The Sidney Kimmel
Comprehensive Cancer Center, John
Hopkins University, Baltimore, MD.
Submitted September 3, 2005; accepted
June 6, 2006.
Supported by Roche Pharma, Madrid,
Spain.
Presented in part at the 41st Annual
Meeting of the American Society
of Clinical Oncology, Orlando, FL,
May 13-17, 2005.
Authors’ disclosures of potential conflicts of interest and author contributions are found at the end of this
article.
Address reprint requests to Miguel
Quintela-Fandino, MD, PhD, Medical
Oncology Department, University
Hospital 12 de Octubre, Avenida de
Córdoba Km 5.4, 28041 Madrid, Spain;
e-mail: [email protected].
© 2006 by American Society of Clinical
Oncology
0732-183X/06/2422-3611/$20.00
DOI: 10.1200/JCO.2005.04.0576
A
B
S
T
R
A
C
T
Purpose
To study the prognostic significance of the presence of breast cancer–specific mRNA transcripts
in peripheral blood (PB), defined by serial analysis of gene expression, in high-risk breast cancer
(HRBC) patients undergoing high-dose chemotherapy after receiving adjuvant chemotherapy.
Methods
From 1994 to 2000, 84 HRBC patients (median age, 44 years; ⬎ 10 nodes; 74%) received adjuvant
chemotherapy (fluorouracil, epirubicin, and cyclophosphamide for six cycles [83%] or doxorubicin
and cyclophosphamide followed by paclitaxel) before undergoing one course of cyclophosphamide
plus thiotepa plus carboplatin (STAMP V). Radiotherapy or hormone therapy was administered
whenever indicated. Aliquots of apheresis-mononuclear blood cells were frozen from each patient.
mRNA was isolated using an automatic nucleic acid extractor based on the magnetic beads
technology; reverse transcription was performed using random hexamers. Cytokeratin 19, HER-2,
P1B, PS2, and EGP2 transcripts were quantified to B-glucuronidase by real-time polymerase chain
reaction (RT-PCR) using a linear DNA probe marked with a quencher and reporter fluorophores
used in RT-PCR. Presence of PB micrometastases, estrogen receptor and progesterone receptor
status, tumor size, age, tumor grade, number of nodes affected, and treatment with paclitaxel
were included in the statistical analysis.
Results
Median follow-up was 68.3 months (range, 6 months to 103 months). Forty-seven relapses (56%) and
35 deaths (41.7%) were registered. Both tumor size and presence of micrometastases reached
statistical significance according to the Cox multivariate model. Relapse hazard ratio (HR) for those
patients with PB micrometastases was 269% (P ⫽ .006); death HR, 300% (P ⫽ .011). Time relapse
was 53 months longer for patients without micrometastases: 31.3 v 84.2 months (P ⫽ .021).
Conclusion
PB micrometastases presence after adjuvant chemotherapy predicts both relapse and death more
powerful than classical factors in HRBC patients undergoing high-dose chemotherapy. Micrometastases search using a gene panel appears to be a more accurate procedure than classical
approaches involving only one or two genes.
J Clin Oncol 24:3611-3618. © 2006 by American Society of Clinical Oncology
INTRODUCTION
Among localized breast cancer patients, clinically occult micrometastatic disease (MD) leads to relapse.
HER-2 positivity is linked to worse prognosis.1-4 Positive HER-2 tumor cell selection takes place during tumor progression.5-7 It could be hypothesized that
patients harboring HER-2–positive MD after adjuvant
chemotherapy are at high risk of relapse.
The conventional approach to screen for MD
in breast cancer involved the detection of a protein
present in breast tumor tissue and absent in hematopooietic tissue. Many studies have analyzed MD in
the adjuvant setting with variable results.8-26 The
following issues may explain such variation and
might be taken into account for the study design:
sample collection timing, in order to avoid detection
of cells nonspecifically released during the surgical
trauma27,28; tumor cell dormancy29; the decrease of
3611
Downloaded from ascopubs.org by 78.47.27.170 on November 17, 2016 from 078.047.027.170
Copyright © 2016 American Society of Clinical Oncology. All rights reserved.
Quintela-Fandino et al
MD contamination during adjuvant chemotherapy30; MD deposits
existence in other organs that bone marrow; higher sensitivity and
specificity are obtained with polymerase chain reaction (PCR) techniques rather than with immunohistochemistry procedures; and cytokeratin 19 (K19) was the most widely used protein for breast MD
detection. K19 is present in breast tissue/cancer, but is exchanged for
vimentin and experiences downregulation as tumor progresses, being
mainly expressed in breast tumor cells unable to produce metastases.31-35
Breast MD could be detected by searching for the mRNA expression of a gene that expresses in every breast cancer cell, but at the same
time it is absent in hematopooietic tissue. Such a gene does not exist.
However, Bosma et al36 designed a highly accurate gene panel to detect
circulating tumor cells, increasing specificity/sensitivity over singlegene assays and sorting out the main concern of background signal
in mRNA-based assays. Two genes (P1B/PS2) expressed in breast
cancer and almost absent in other tissues were found by serial
analysis of gene expression. K19 and EGP237 were subsequently
added to the panel.37
We hypothesize that the former issues might be overcome with
the following design, and relapse risk of breast cancer patients can be
assessed by indirect detection of clinically occult micrometastases located in any parenchima that are releasing tumor cells to the peripheral blood. We have performed a retrospective study to assess the
prognostic role of MD, searching for P1B, PS2, EGP2, K19, and HER-2
(HER-2 added to the panel as it is not constitutively expressed in
hematopooietic tissue) expression in peripheral blood of high-risk
breast cancer patients undergoing high-dose chemotherapy (HDCT).
We used real-time (RT) polymerase chain reaction (PCR) assays in
samples collected during blood apheresis (several months after surgery and after conventional-dose adjuvant chemotherapy). The aims
of this study were to assess the prognostic value of breast cancer–
specific mRNA transcripts in peripheral blood, and determine if MD
study deserves prospective validation; to test whether the study design
resolves the inherent difficulties of assessing the prognostic implications of MD; and to study the HER-2 status of MD.
METHODS
Patients and Study Design
This retrospective study was initiated in June 2003. Between 1993 and
1999 patients undergoing HDCT for breast cancer at two hospitals located in
Madrid, Spain were screened.
Inclusion criteria were: histopathologically proven, locally advanced
breast cancer; the absence of distant metastases in chest abdominal computed
tomography scan and bone scan both before adjuvant or neoadjuvant treatment and before HDCT; negative histopathologic examination for tumor cells
in bone marrow biopsy before HDCT; previous adjuvant/neoadjuvant treatment with a standard anthracycline/taxane-based regimen; followed by
adjuvant cyclophosphamide plus thiotepa plus carboplatin (STAMP V)
course; and availability of frozen mononuclear cells obtained during
apheresis and consequently collected and stored (⫺80°C; 10% dimethyl
sulfoxide) for research purposes.
Patients were required to give written informed consent before inclusion
in the study. The study protocol was approved by the institutional review
board at each study center.
Molecular Biology Procedures
Assays were conducted using 2-mL aliquots of mononuclear cells
obtained during apheresis (one to six per patient; median three; total
samples 226). Total mRNA was isolated and RNA integrity was tested by
Table 1. Patient Characteristics
Value
Characteristic
Age, years
Median
Range
Sex
Male
Female
Hormone status
Premenopausal
Menopausal
Tumor size (surgical specimen)
Tx
T1
T2
T3
T4
Not available
No. of positive nodes
Median
Range
1-3ⴱ
4-9
⬎ 10
Primary metastatic disease
Previous breast cancer history
Histopathologic subtype
Ductal
Lobular
Other
Histologic grade
I
II
III
Hormone receptor status
ER⫹/PR⫹
ER⫹/PR⫺
ER⫺/PR⫹
ER⫺/PR⫺
HER-2/NEU
Negative (0 or 1⫹)
Positive (2⫹ or 3⫹)
Not available
Surgery†
Mastectomy ⫹ ALN
Lumpectomy ⫹ ALN
Adjuvant chemotherapy
FEC
AC-paclitaxel
Adjuvant STAMP V
Toxic deaths
Adjuvant radiotherapy
Adjuvant tamoxifen
Time, months
Median
Range‡
No.
%
44.28
24-69
0
100
72
12
85.5
14.5
3
7
32
24
15
3
3.6
8.5
38.3
28.9
18.8
3.6
13.37
2-38
4
18
62
0
0
4.7
21.4
73.8
0
0
72
7
5
85.7
8.3
6
1
32
51
1.1
38.1
60.8
43
9
1
31
51.19
10.71
1.19
36.90
23
24
37
27
28
44
59
25
70.3
29.8
70
14
84
1
84
50
83
17
100
1.2
100
59.5
48
1-60
Abbreviations: ER, estrogen receptor; PR, progesterone receptor; ALN,
axillary lymph nodes; FEC, fluorouracil, epirubicin, and cyclophosphamide; AC,
doxorubicin and cyclophosphamide; STAMP V, cyclophosphamide, thiotepa,
and carboplatin.
ⴱ
These four patients had received neoadjuvant chemotherapy.
†Surgery performed after neoadjuvant treatment in T4 patients.
‡Thirty-one patients stopped tamoxifen at disease progression and one
patient stopped due to intolerance.
3612
JOURNAL OF CLINICAL ONCOLOGY
Downloaded from ascopubs.org by 78.47.27.170 on November 17, 2016 from 078.047.027.170
Copyright © 2016 American Society of Clinical Oncology. All rights reserved.
Micrometastases in Blood Predicts Poor Survival
Table 2. Median RNA Yield Among Patients and Negative Controls
Ct
GUS
Efficiency
Patients
Healthy donors with
neutrophils
Healthy donors without
neutrophils
1.99
1.90
1.94
Range
Positive
Samples
Negative
Samplesⴱ
27.86
27.32
22.83-38.44
25.54-30.45
72
13
12
2
27.25
25.52-30.68
14
1
Median
Abbreviations: Ct, threshold cycle; GUS, b-glucuronidase.
ⴱ
A sample is considered that has not yielded enough RNA when GUS Ct is ⬎ 29.
amplification of the b-glucuronidase (GUS) housekeeping gene. c-DNA
was synthesized using random-hexamer oligonucleotides. P1B, PS2,
EGP2, HER-2, and K19 were amplified and quantified for each patient by
RT-PCR with TaqMan probes (Applied Biosystems, Foster City, CA) in a
LightCycler Instrument using the LC Fast Start DNA Master Hybridization
Probe kit (Roche Diagnostics).
RT-PCR efficiency was calculated using the LightCycler Software version
3.5 (Roche Applied Science, Mannheim, Germany) according to the equation:
E ⫽ 10(⫺1/S), where S is the slope of the cumulative fluorescence during the
exponential phase. Quantification was calculated relative to GUS38 using the
equation: Ratio ⫽ (Etarget)(Ctref – Cttarget), where target represents each transcript
and ref, GUS.39 Ct is the threshold cycle (PCR cycle at which a significant
increase in target-specific fluorescence is detected due to exponential accumulation of PCR products, expressed in arbitrary units).
Cell lines HCC1143 and MCF-7 were used as positive controls. As
apheresis samples lack neutrophils, two types of negative controls were used:
peripheral blood from 15 neutrophil-containing and 15 neutrophil-depleted
healthy donors, respectively.
For each patient, all available aliquots were pooled into a single sample
(number of samples ⫽ 84). Those with a GUS-Ct above cycle 29 (n ⫽ 12) were
discarded, as the low mRNA amount could yield false-negatives.
All the procedures were repeated twice. The technicians neither had
access to follow-up data nor took part in the analysis.
More detailed information regarding treatment and patient selection,
specific molecular biology procedures, and verification of classic prognostic
factors are available at the Appendix.
Statistical Analysis
The follow-up period started the day surgery was performed. The main
variable analyzed was presence of micrometastatic (MM) activity. Breast cancer cells do not harbor the same genetic expression profile, as opposed to
hematologic malignant cells, and consequently one target may be expressed in
one tumor cell and absent in another one. In order to increase the chance of
finding circulating tumor cells, a high number of specific mRNA targets for
micrometastatic detection were utilized. MM was categorized as positive/yes if
either target was detected in peripheral blood regardless of the value, and as
negative/no if no target was detected. K19 was excluded from MM as K19 was
present at similar ratios in samples from patients and negative controls.
The Kaplan-Meier method and Cox hazard regression model were used
for statistical analysis. The clinical end points under analysis were time to
relapse or time to death (Kaplan-Meier) and relapse/death (Cox). Age, tumor
grade, tumor size, treatment with paclitaxel, estrogen receptor (ER) and progesterone receptor (PR) status, and number of nodes were included in the Cox
analysis to test the independent prognostic impact of MM. HercepTest for
Immunoenzymatic Staining (Dako Corp, Carpinteria, CA) was not routinely
performed at our institution until 1998. These data were not available from all
patients and therefore HER-2 in the primary tumor was removed from the
multivariate analysis. The overall significance and goodness of fit of each
model was calculated with the Rao test and Cox-Snell R2. Both the log-linear
and the multiplicative conditions for the application of the Cox model were
tested. SPSS version 12.0 for Windows (SPSS Inc, Chicago, IL) was used for the
statistical analysis. All statistical tests were two sided.
RESULTS
Patients
Patient characteristics are depicted in Table 1 (n ⫽ 84). After a
median follow-up of 68.3 months (range, 6 to 103; follow-up period
1993 to 2004), 47 patients (56.0%; 95% CI, 45.4% to 66.6%) have
relapsed, with a median TTR of 66 months. Median overall survival
has not been reached. At the end of follow-up, 36 patients (42.8%;
95%CI, 37.3% to 48.3%) were alive and disease-free, 13 patients
Table 3. Efficiency and Relative Values for Each Gene Among Patients and Negative Controls
Blood From Healthy Donors
(containing neutrophils)
Patients
Blood From Healthy Donors
(neutrophil-depleted)
Efficiency
M/R (all)
M/R (⫹)
⫾
Efficiency
M/R (all)
M/R (⫹)
⫾
P1B
1.96
1.79
0
0
—
2.01
—
7/8
—
0
—
0 (0-0)
0 (0-0)
—
0/15
HER2
1.74
7/65
2.02
3/12
0
0 (0-0)
0 (0-0)
0/15
K19
2.05
3/69
1.96
0.48
(0.48-0.48)
—
0.0028
(0.0018-0.0058)
0.029
(0.015-0.048)
0.00024
(5.7E-5-0.00039)
0
—
24/48
0.003
(0-0.48)
—
0.00013
(0-0056)
0.006
(0-0.048)
8,1E-5
(0-0.00039)
1/14
—
2.21
0.36
(0.02-1.79)
—
0.0014
(0.0001-0.006)
0.036
(0.012-0.087)
0.00062
(0.00058-0.00068)
12/60
PS2
EGP2
0.051
(0-1.79)
—
0.0005
(0-0.006)
0.0035
(0-0.087)
2.6E-5
(0-0.00062)
5/10
1.78
4.8E-5
(0-0.00053)
0.00029
(5E-5-0.00053)
2/13
Gene
Efficiency
M/R (all)
M/R (⫹)
⫾
0/15
—
NOTE. Mean values are presented instead of median values as many of the samples are negative and thus the median value would be “0” in most of the boxes
and thus poorly orientative.
Abbreviations: M/R (all), mean relative values and range considering all patients; M/R (⫹), mean relative values and range considering only those patients with
positive values for the given transcript; ⫾, samples positive and negative for the given transcripts among the 72 patients with valid samples.
3613
www.jco.org
Downloaded from ascopubs.org by 78.47.27.170 on November 17, 2016 from 078.047.027.170
Copyright © 2016 American Society of Clinical Oncology. All rights reserved.
Quintela-Fandino et al
Table 4. Relative Values for Each Transcript in Peripheral Blood (containing
neutrophils) From Healthy Donors
Control
P1B
EGP2
HER-2
K19
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
0
0
0
0
0
0
0
0
0.489
0
0
0
0
0
0
0.0024
0.0019
0.0026
0
0.0057
0
0.0027
0
0
0
0
0
0.0029
0.0019
0
0
0.015
0
0
0
0
0.048
0
0
0
0
0
0
0
0.023
0.00037
0.00039
5.7E-05
0
0
0
0
0.00013
0
0
0
0
0
0.00026
0
NOTE. Values are represented as relative rates to b-glucuronidase.
(15.5%; 95%CI, 7.8% to 23.2%) were alive but had relapsed, and 35
patients (41.7%; 95%CI, 31.2% to 52.2%) have died.
RNA Yield, RT-PCR Efficiency, Target Gene Positivity,
and Relative Values for Each Transcript
Median Ct and Ct range for GUS, efficiency and number of
positive and negative samples for patients, and negative controls are
depicted in Table 2. The last Ct (intercept) that was considered positive for each gene was 38.51 (P1B), 40.42 (EGP2), 47.73 (HER-2), and
41.60 (K19).
PS2 Ct was higher than 40 and higher than 35 in nondiluted cell
line samples of HCC1143 and MCF7, respectively. Since Ct for the
other targets was approximately 14, this yields a 30,000-fold lower
expression of PS2 compared with the other targets. Therefore, PS2,
although detectable in nondiluted tumor cell lines, as apheresis tumor
cell contamination is about 1/106 nontumor cells, remained undetectable in every patient.
Table 3 presents the efficiency and relative values for each gene
among patients and negative controls. The individual results for each
transcript in blood from healthy donors containing neutrophils are
presented in Table 4.
MM and Prognosis
Patient flow diagram and Kaplan-Meier plots for time-torelapse and overall survival according to MM status are depicted in
Figures 1 and 2. Multivariate analysis for relapse and survival is summarized in Table 5. Tumor size and presence of MM remained as
independent factors on relapse (13.6% increase in hazard ratio
[HR] per centimeter increase in tumor size and 269.6% increase for MM–positive). MM status maintained an independent
prognostic influence (HR, 3) in overall survival. Considering the
highest risk subgroup (ⱖ 10 nodes; n ⫽ 42), the HR among
MM–positive patients was 2.55 for relapse (P ⫽ .031) and 3.05 for
death (P ⫽ .030).
Figure 3 contains the Kaplan-Meier plots for TTR among patients
with circulating mRNA of K19–positive versus K19–negative; HER-2–
positive versus HER-2–negative; EGP2–positive versus EGP2–negative;
and P1B–positive versus P1B–negative. Log-rank P values were .22
(K19), .61(HER-2), and .64 (EGP2), respectively; P1B–positive versus
P1B–negative median TTR was 23.2 months (95% CI, 6.2 to 40.2
months) versus 76.6 months (95% CI, 54.6 to 98.5 months; P ⫽ .01). The
median overall survival for P1B–positive versus P1B–negative was 32.3
months (95% CI, 0 to 87.4 months) versus not reached (P ⫽ .03).
Considered separately, K19, HER-2, and EGP2 (log-rank P ⫽ .17, .91,
and .75, respectively) did not influence overall survival (plots not shown).
When analyzing age, grade, tumor size, PR/ER status, treatment with paclitaxel, nodes, and relative values of circulating K19,
HER-2, EGP2 and P1B mRNA in a multivariate Cox model for
TTR (Rao P ⫽ .016; R2 ⫽ 0.0027), the value of circulating EGP2
and P1B mRNA showed a statistically significant independent
impact (P ⫽ .003 and .017, respectively; HR ⫽ 103 and HR ⫽ 1.021).
However, the extremely low R2 values invalidate these results. With
regard to overall survival, the final model had an overall significance of .201; thus, no conclusions can be drawn regarding individual transcripts.
Fig 1. Patient flow through the study. MM,
micrometastases; TTR, time to relapse (in
months); TTD, time to death (in months).
3614
JOURNAL OF CLINICAL ONCOLOGY
Downloaded from ascopubs.org by 78.47.27.170 on November 17, 2016 from 078.047.027.170
Copyright © 2016 American Society of Clinical Oncology. All rights reserved.
Micrometastases in Blood Predicts Poor Survival
Table 5. Multivariate Analysis for Relapse and Survival
Relapse
Factor
Age, years
Tumor size, cm
Grade
Fictitious grade 1
Fictitious grade 2
ER
PR
Paclitaxel reception
Micro
Nodes
Significance of the model/R 2
P
Hazard
Ratio
.289
1.023
.0470
1.136
.924
.896
6219
.894
7100
.455
0.666
.724
0.841
.088
2.311
.006
2.696
.115
1.041
.030/.221
Overall Survival
P
Hazard
Ratio
.580
1.013
.276
1.088
.709
.923
3918
.920
5523
.235
0.484
.623
0.749
.394
1.723
.011
3.000
.122
1.044
.047/.191
NOTE. Grade, ER, PR, treatment with paclitaxel and micrometastatic activity
were considered categorical variables. Age, tumor size and number of
affected nodes were considered quantitative variables. The variables that
reached statistically significant impact in relapse and overall survival are
depicted in bold. Categorical variables with ⬎ two categories (grade) were
automatically decomposed in fictitious variables for Cox’s analysis by the
SPSS program (SPSS Inc, Chicago, IL).
Abbreviations: ER, estrogen receptor; PR, progesterone receptor; micro,
micrometastatic disease.
The presence of one transcript was not found to be a predictor for
any other. An association (r ⫽ 0.536; P ⫽ .000) was found between
circulating K19 and EGP2. However the limited number of K19positive cases (n ⫽ 3) precluded any interpretation.
DISCUSSION
Fig 2. Kaplan-Meier plots for (A) time to relapse and (B) overall survival
according to micrometastatic (MM) status.
Relationship Between Conventional Prognostic
Factors and MM
Categoric variables (ER, PR, grade, HER-2) did not predict MM
presence (P ⫽ .071, .358, .603, and .847, respectively).
The correlation between the quantitative variables (age, tumor
size, number of affected nodes) and the presence/absence of MM
as well as the relative values of each transcript was also analyzed
using logistic regression and Pearson analysis, respectively. The B
coefficients for the logistic regression test were close to the null value
(⫺0.003, 0.017, and ⫺0.011), with P values of .91, .87, and .746,
respectively. No correlation was found in the Pearson analysis (data
not shown), although there was a positive correlation (r ⫽ 0.279;
P ⫽ .019) between age and amount of circulating EGP2.
MD has been the subject of extensive research in breast cancer and the
detection of K19 in hematopooietic tissue has been used as an indicator of the presence of residual tumor cells.8-26
Gene panels were first used in the study by Bosma et al,36 where
samples from 103 metastatic patients were analyzed using immunohistochemistry to find circulating breast cancer cells. In our study a
similar expression (rate and percentage) of the transcripts in the blood
of patients and healthy donors has been found, but no expression in
negative controls after removing neutrophils, suggesting that these
genes may indicate the presence of MD in neutrophil-depleted samples. Even though the question of whether EGP2 (EP-CAM), P1B
(TFF-3), HER-2, or K19 are expressed by neutrophils warrants further
study, we performed the negative control analysis in neutrophildeplected peripheral blood samples as apheresis samples from patients
did not contain neutrophils. K19 was present in neutrophil-depleted
negative control (2 of 15) and patient (3 of 72) samples. Whether this
can be explained by contamination from skin during venopuncture or
from neutrophils cannot be concluded from our study, but led us to
exclude K19 from the panel. Previous studies using K19 state neither
whether their samples were neutrophil-depleted nor the K19-positive
rate among negative controls.8-26 Even though K19 should not be
expressed in neutrophils, future studies assessing MM with a genepanel should be performed with neutrophil-depleted samples in order
to avoid misclassification, and results among negative controls should
be presented.
3615
www.jco.org
Downloaded from ascopubs.org by 78.47.27.170 on November 17, 2016 from 078.047.027.170
Copyright © 2016 American Society of Clinical Oncology. All rights reserved.
Quintela-Fandino et al
Fig 3. Kaplan-Meier plots for time to relapse according to presence/absence of each target transcript in peripheral blood. (A) HER-2, (B) K19, (C) EGP2, (D) P1B.
Although in the study by Bosma et al36 the predictive power of the
panel could be checked by immunohistochemistry, as the experiments
were conducted with samples from active disease patients, we can
sustain that the panel indicates MD, based on our prognostic results
and the results form neutrophil-depleted negative controls.
Previous studies have worked with epithelial markers, many
of them showing no prognostic impact despite a high number of
patients,8-26 and only two obtained significant prognostic impact
in the multivariate analysis.24,25 Overall, the average K19 detection
in those studies ranges 20% to 40%; however, the majority of
patients included matched the low- or intermediate-risk category.
The keratins are mainly expressed in the primary tumor and are
associated with better prognosis than that attributed to keratinnegative breast cancer cells, while the tumor cell downregulates the
epithelial markers when it reaches the blood flow.31-35 In case of
accepting K19 as a MM marker in our series, we would have found
MM only in 4.2% of the patients, in contrast to 59.7% using
metastatic breast cancer markers found by serial analysis of gene
expression by Bosma et al.36 Although our work was conducted in
patients in the adjuvant setting, many of them were actually (micro)metastatic. Among the main studies performed in adjuvant
breast cancer, to our knowledge, this is the first in where the
prognostic impact for the patients with MD is found for patients
with high-risk breast cancer undergoing high-dose chemotherapy;
in addition, it is the first in where a gene panel appears yields more
accurate prognostic information than the classic K19-based approach. Although in our series the prognostic information yielded
by the panel is conserved both in the patients with four to nine nodes or
morethan10nodes,whatreinforcesitsvalueoverK19-basedapproaches,
the fact that the K19 is downregulated in highly metastatic breast cancer
cells, should be taken into account. In our series, the lower-risk patients
still have a high risk (four to nine positive nodes).
These facts, and the results depicted in Tables 3 and 4, suggest that
K19 alone may not be the ideal target for MM detection.
3616
JOURNAL OF CLINICAL ONCOLOGY
Downloaded from ascopubs.org by 78.47.27.170 on November 17, 2016 from 078.047.027.170
Copyright © 2016 American Society of Clinical Oncology. All rights reserved.
Micrometastases in Blood Predicts Poor Survival
A recent meta-analysis by Braun et al40 set the HR for death
among more than 4,000 breast cancer patients at 1.93 for those with
bone marrow contamination. We obtained an HR attributable to MM
of 3.0 for death in the Cox model, and a 55-month survival advantage
in the Kaplan-Meier plots for patients without MM. The study design
(which overcomes MM assessment difficulties: samples obtained
months after surgery/chemotherapy; avoiding dormant cells in bone
marrow and using several targets) and the high number of events have
probably contributed to the ability to obtain these results with 84
patients. The similarity of the HR to that obtained by Braun led us to
believe that the MM assessment in peripheral blood is accurate. When
the panel is separated into individual transcripts, the extremely low
percentage of variability in relapse explained by the Cox model (R2),
the high P value of the model for death, and the results shown in Figure
3 suggest that the MM study should be performed with several transcripts rather than one.
REFERENCES
1. Slamon DJ, Clark GM, Wong SG, et al: Human
breast cancer: Correlation of relapse and survival
with amplification of the HER-2/neu oncogene. Science 235:177-182, 1987
2. Muss HB, Thor AD, Berry DA, et al: c-erbB-2
expression and response to adjuvant therapy in
women with node-positive early breast cancer.
N Engl J Med 330:1260-1266, 1994
3. Paik S, Bryant J, Park C, et al: erbB-2 and
response to doxorubicin in patients with axillary
lymph node-positive, hormone receptor-negative
breast cancer. J Natl Cancer Inst 90:1361-1370,
1998
4. Paik S, Bryant J, Tan-Chiu E, et al: HER2 and
choice of adjuvant chemotherapy for invasive breast
cancer: National Surgical Adjuvant Breast and Bowel
Project Protocol B-15. J Natl Cancer Inst 92:19911998, 2000
5. Lacroix H, Iglehart JD, Skinner MA, et al:
Overexpression of erbB-2 or EGF receptor proteins
present in early stage mammary carcinoma is detected simultaneously in matched primary tumors
and regional metastases. Oncogene 4:145-151,
1989
6. Niehans GA, Singleton TP, Dykoski D, et al:
Stability of HER-2/neu expression over time and at
multiple metastatic sites. J Natl Cancer Inst 85:
1230-1235, 1993
7. Braun S, Schlimok G, Heumos I, et al: ErbB2
overexpression on occult metastatic cells in bone marrow predicts poor clinical outcome of stage I-III breast
cancer patients. Cancer Res 61:1890-1895, 2001
8. Pantel K, Schlimok G, Braun S, et al: Differential expression of proliferation-associated molecules
in individual micrometastatic carcinoma cells. J Natl
Cancer Inst 85:1419-1424, 1993
9. Vredenburgh JJ, Peters WP, Rosner G, et al:
Detection of tumor cells in the bone marrow of
stage IV breast cancer patients receiving high-dose
chemotherapy: The role of induction chemotherapy.
Bone Marrow Transplant 16:815-821, 1995
10. Diel IJ, Kaufmann M, Costa SD, et al: Micrometastatic breast cancer cells in bone marrow at
primary surgery: Prognostic value in comparison
with nodal status. J Natl Cancer Inst 88:1652-1658,
1996
11. Vannucchi AM, Bosi A, Glinz S, et al: Evaluation of breast tumour cell contamination in the bone
marrow and leukapheresis collections by RT-PCR for
We failed to demonstrate that the late relapse is due to HER-2–
positive circulating tumor cells. We found that the rate of mRNA
expression was almost 1000-fold lower than the other transcripts in
the positive controls, what may preclude interpretations of the fact
that HER-2 transcripts were not found in peripheral blood due to the
low rate of cell contamination.
The main limitations of our study are its retrospective nature and
the sample size, although it has been the largest one to assess MM using
RT-PCR in patients undergoing HDCT. This study was performed in
order to ensure that the present approach merits further prospective
validation. Given the facts that MM has been the most powerful
prognostic factor in this series and that available therapies like antiangiogenic agents could be effective against MD, a prospective study
should be conducted in a large, average-risk cohort in order to validate
our results.
cytokeratin-19 mRNA. Br J Haematol 103:610-617,
1998
12. Cooper BW, Moss TJ, Ross AA, et al: Occult
tumor contamination of hematopoietic stem-cell
products does not affect clinical outcome of autologous transplantation in patients with metastatic
breast cancer. J Clin Oncol 16:3509-3517, 1998
13. Braun S, Hepp F, Sommer HL, et al: Tumorantigen heterogeneity of disseminated breast cancer cells: Implications for immunotherapy of minimal
residual disease. Int J Cancer 84:1-5, 1999
14. Mansi JL, Gogas H, Bliss JM, et al: Outcome
of primary-breast-cancer patients with micrometastases: A long-term follow-up study. Lancet 354:197202, 1999
15. Braun S, Pantel K, Muller P, et al: Cytokeratinpositive cells in the bone marrow and survival of
patients with stage I, II, or III breast cancer. N Engl
J Med 342:525-533, 2000
16. Lyda MH, Tetef M, Carter NH, et al: Keratin
immunohistochemistry detects clinically significant
metastases in bone marrow biopsy specimens in
women with lobular breast carcinoma. Am J Surg
Pathol 24:593-1599, 2000
17. Gerber B, Krause A, Muller H, et al: Simultaneous immunohistochemical detection of tumor
cells in lymph nodes and bone marrow aspirates in
breast cancer and its correlation with other prognostic factors. J Clin Oncol 19:960-971, 2001
18. Gebauer G, Fehm T, Merkle E, et al: Epithelial
cells in bone marrow of breast cancer patients at time
of primary surgery: Clinical outcome during long-term
follow-up. J Clin Oncol 19:3669-3674, 2001
19. Schneeweiss A, Diel I, Hensel M, et al: Micrometastatic bone marrow cells at diagnosis have
no impact on survival of primary breast cancer
patients with extensive axillary lymph node involvement treated with stem cell-supported high-dose
chemotherapy. Ann Oncol 15:1627-1632, 2004
20. Kasimir-Bauer S, Oberhoff C, Schindler AE, et
al: A summary of two clinical studies on tumor cell
dissemination in primary and metastatic breast cancer: Methods, prognostic significance and implication for alternative treatment protocols. Int J Oncol
20:1027-1034, 2002
21. Witzig TE, Bossy B, Kimlinger T, et al: Detection of circulating cytokeratin-positive cells in the
blood of breast cancer patients using immunomagnetic enrichment and digital microscopy. Clin Cancer
Res 8:1085-1091, 2002
22. Wong IH, Yeo W, Chan AT, et al: Quantitative
correlation of cytokeratin 19 mRNA level in periph-
eral blood with disease stage and metastasis in
breast cancer patients: Potential prognostic implications. Int J Oncol 18:633-638, 2001
23. Hu XC, Chow LW: Detection of circulating
breast cancer cells with multiple-marker RT-PCR
assay. Anticancer Res 21:421-424, 2001
24. Xenidis N, Vlachonikolis I, Mavroudis D, et al:
Peripheral blood circulating cytokeratin-19 mRNApositive cells after the completion of adjuvant chemotherapy in patients with operable breast cancer.
Ann Oncol 14:849-855, 2003
25. Stathopoulou A, Vlachonikolis I, Mavroudis D,
et al: Molecular detection of cytokeratin-19-positive
cells in the peripheral blood of patients with operable breast cancer: Evaluation of their prognostic
significance. J Clin Oncol 20:3404-3412, 2002
26. Silva JM, Dominguez G, Silva J, et al: Detection of epithelial messenger RNA in the plasma of
breast cancer patients is associated with poor prognosis tumor characteristics. Clin Cancer Res 7:28212825, 2001
27. Fidler IJ, Ellis LM: The implications of angiogenesis for the biology and therapy of cancer metastasis. Cell 79:185-188, 1994
28. Folkman J: Angiogenesis in cancer, vascular,
rheumatoid and other disease. Nat Med 1:27-31, 1995
29. Barz H, Barz D: The prevalence of single and
synchronous multiple tumours in autopsy series and
its dependence on age. Arch Geschwulstforsch
47:428-436, 1977
30. Braun S, Kentenich C, Janni W, et al: Lack of
effect of adjuvant chemotherapy on the elimination
of single dormant tumor cells in bone marrow of
high-risk breast cancer patients. J Clin Oncol 18:8086, 2000
31. Thomas PA, Kirschmann DA, Cerhan JR, et al:
Association between keratin and vimentin expression, malignant phenotype, and survival in postmenopausal breast cancer patients. Clin Cancer Res
5:2698-2703, 1999
32. Fuchs IB, Lichtenegger W, Buehler H, et al: The
prognostic significance of epithelial-mesenchymal
transition in breast cancer. Anticancer Res 22:34153419, 2002
33. Zajchowski DA, Bartholdi MF, Gong Y, et al:
Identification of gene expression profiles that predict the aggressive behavior of breast cancer cells.
Cancer Res 61:5168-5178, 2001
34. Woelfle U, Cloos J, Sauter G, et al: Molecular
signature associated with bone marrow micrometastasis in human breast cancer. Cancer Res 63:
5679-5684, 2003
3617
www.jco.org
Downloaded from ascopubs.org by 78.47.27.170 on November 17, 2016 from 078.047.027.170
Copyright © 2016 American Society of Clinical Oncology. All rights reserved.
Quintela-Fandino et al
35. Woelfle U, Sauter G, Santjer S, et al: Downregulated expression of cytokeratin 18 promotes
progression of human breast cancer. Clin Cancer
Res 10:2670-2674, 2004
36. Bosma AJ, Weigelt B, Lambrechts AC, et al:
Detection of circulating breast tumor cells by differential expression of marker genes. Clin Cancer Res
8:1871-1877, 2002
imal residual disease follow up in leukemia patients.
Blood 98:4467-4475, 2001
39. Pfaffl MW: A new mathematical model for
relative quantification in real-time RT-PCR. Nucleic
Acids Res 29:2002-2007, 2001
40. Braun S, Vogl FD, Naume B, et al: A pooled
analysis of bone marrow micrometastasis in breast
cancer. N Engl J Med 353:793-802, 2005
37. Helfrich W, ten Poele R, Meersma GJ, et al: A
quantitative reverse transcriptase polymerase chain
reaction-based assay to detect carcinoma cells in
peripheral blood. Br J Cancer 76:29-35, 1997
38. Pallisgaard N, Hokland P, Bi W, et al: Selection
of referente genes for the European standardization
and quality control program of real-time quantitative
RT-PCR analysis of fusion gene transcripts for min■ ■ ■
Appendix
The Appendix is included in the full-text version of this article, available online at www.jco.org. It is not included in the PDF version
(via Adobe® Reader®).
Authors’ Disclosures of Potential Conflicts of Interest
Although all authors completed the disclosure declaration, the following authors or their immediate family members indicated a financial interest. No conflict exists for
drugs or devices used in a study if they are not being evaluated as part of the investigation. For a detailed description of the disclosure categories, or for more information
about ASCO’s conflict of interest policy, please refer to the Author Disclosure Declaration and the Disclosures of Potential Conflicts of Interest section in Information for
Contributors.
Authors
Employment
Leadership
Consultant
Stock
Honoraria
Research Funds
Testimony
Other
Miguel QuintelaFandino
Roche (A)
Roche (A)
Ricardo Hitt
Roche (A)
Roche (A)
Soledad Gamarra
Roche (A)
Cecilia Guzman
Roche (N/R)
Hernan Cortes Funes
Roche (A)
Dollar Amount Codes
(A) ⬍ $10,000
(B) $10,000-99,999
(C) ⱖ $100,000
(N/R) Not Required
Author Contributions
Conception and design: Miguel Quintela-Fandino, Joaquin Martinez, Ricardo Hitt, Antonio Jimeno
Financial support: Cecilia Guzman, Hernan Cortes Funes
Administrative support: Ricardo Hitt, Hernan Cortes Funes
Provision of study materials or patients: Ricardo Hitt, Javier Hornedo, Cecilia Guzman, Hernan Cortes Funes
Collection and assembly of data: Miguel Quintela-Fandino, Soledad Gamarra, Rosa Ayala
Data analysis and interpretation: Miguel Quintela-Fandino, Joaquin Martinez, Ricardo Hitt
Manuscript writing: Miguel Quintela-Fandino, Joaquin Martinez, Hernan Cortes Funes
Final approval of manuscript: Miguel Quintela-Fandino, Joaquin Martinez, Ricardo Hitt, Soledad Gamarra, Antonio Jimeno, Rosa Ayala, Javier Hornedo,
Florinda Gilsanz, Cecilia Guzman, Hernan Cortes Funes
3618
JOURNAL OF CLINICAL ONCOLOGY
Downloaded from ascopubs.org by 78.47.27.170 on November 17, 2016 from 078.047.027.170
Copyright © 2016 American Society of Clinical Oncology. All rights reserved.