AJR 2004; 183:1159-1168
© American Roentgen Ray Society
MRI for Surgical Planning in Patients with Breast Cancer Who Undergo Preoperative Chemotherapy
Fabienne Thibault1,
Claude Nos2,
Martine Meunier1,
Carl El Khoury1,
Liliane Ollivier1,
Brigitte Sigal-Zafrani3 and
Krishna Clough2 for the Institut Curie Breast Cancer Group
1 Department of Medical Imaging, Institut Curie, 26 rue d'Ulm, 75248 Paris Cedex
05, France.
2 Department of Surgery, Institut Curie, 75248 Paris Cedex 05, France.
3 Department of Pathology, Institut Curie, 75248 Paris Cedex 05, France.
Received August 28, 2003;
accepted after revision March 26, 2004.
Address correspondence to F. Thibault
(fabienne.thibault{at}curie.net).
Abstract
OBJECTIVE. Accurate presurgical evaluation of residual disease
appears essential for successful clinical outcome in patients with breast
cancer who are undergoing chemotherapy. Our objective was to study the impact
on surgical planning of adding serial MRI evaluations of the tumor to standard
non–MRI assessments.
MATERIALS AND METHODS. MR images of breast tumors obtained before,
during, and after preoperative chemotherapy were reviewed in 30 patients.
Tumor response was assessed using both size and morphologic MRI criteria. We
compared the actual surgical decisions made prospectively on the basis of
standard (clinical, mammographic, and sonographic) assessments of response
with decisions that would have been made had MRI findings also been
considered. MRI investigators were blinded to the ultimate surgical results.
Successful breast-conserving surgery was judged on pathologic confirmation of
excision margins that were negative for cancer.
RESULTS. The standard evaluation led to 16 successful
breast-conserving and 14 mastectomy procedures. Using MRI results would have
led to major beneficial therapeutic changes in six (20%) of the 30 patients:
five patients undergoing primary mastectomy (chemotherapy avoided) and one
patient undergoing postchemotherapy mastectomy (unsuccessful breast-conserving
surgery avoided). MRI would have added valuable information in 14 (46.7%) of
the 30 patients. In seven (23.3%) of the 30 patients, the decision to perform
postchemotherapy mastectomy would have been unchanged. In one patient (3.3%),
MRI results would not have prevented unsuccessful breast-conserving surgery.
In two patients (6.6%), MRI results would have prevented successful
breast-conserving surgery from being performed.
CONCLUSION. Although the ultimate incidence of breast conservation
was potentially similar for the patients (16/30, 53%) in whom the standard
evaluation was used and for the patients (14/30, 47%) in whom the MRI-added
evaluation was used, MRI was useful in establishing the final treatment
earlier in the process, avoiding unnecessary preoperative chemotherapy, or
selecting high-risk breast-conserving procedures.
Introduction
The standard surgical treatment for patients with large operable breast
cancers is mastectomy. Preoperative (neoadjuvant) chemotherapy has been
developed as a therapeutic option in the past decade
[1–6].
The main benefit gained from preoperative chemotherapy is a reduction in tumor
size, which permits breast-conserving surgery in patients who otherwise would
have required a mastectomy. The reported rate of tumor downstaging obtained
with neoadjuvant chemotherapy varies from 22.4% in the randomized National
Surgical Adjuvant Breast and Bowel Project (NSABP) B-18 trial
[7] to 34% in a recent study
from The University of Texas M. D. Anderson Cancer Center
[6]. In the NSABP B-18 trial
[7], the rate of local relapse
noted in patients whose surgical plan was changed from proposed mastectomy to
lumpectomy after preoperative chemotherapy was 15.9% at 9 years. At our
institution, a study of 257 patients with large breast tumors who had tumor
reduction after preoperative chemotherapy and thus were eligible for
breast-conserving surgery [4]
showed high local relapse rates of 16% at 5 years and 21.5% at 10 years after
surgery. In addition, this local relapse was a strong predictor for distant
metastases [4]. The accurate
evaluation of residual disease after neoadjuvant chemotherapy and before
surgery appears essential for successful clinical outcome.
After induction therapy, the decision of whether to perform mastectomy
versus breast-conserving surgery is typically based on clinical and radiologic
assessment of tumor response
[1,
5,
6]. Breast-conserving surgery
is defined as successful if the margins on the final lumpectomy specimen are
free from disease [4,
6].
A number of investigators have included MRI in the assessment of response
to preoperative chemotherapy
[8–13].
MRI was found to be more accurate than palpation and mammography in patients
with extensive breast cancer
[9,
11]. One study suggested a
link between tumor morphology and the degree of response measured on breast
MRI [14]. MRI findings,
however, may lead to underestimation of the extent of residual disease after
chemotherapy [10,
15].
The aim of our study was to determine how the addition of breast MRI to
routine clinical and radiologic assessment of the tumor at presentation and
under chemotherapy affects the overall surgical decision process. We studied
the hypothesis that MRI assessment of both tumor size and morphology would
help oncologists and surgeons select the appropriate therapeutic end point
(breast-conserving surgery vs mastectomy) more directly than would
conventional non–MRI assessment. The MR images of breast tumors in 30
patients undergoing preoperative chemotherapy were reviewed. We compared the
actual decisions made on the basis of conventional assessment with the
decisions that would have been made had MRI additionally been used in the
planning of surgery. The investigators were blinded to the actual clinical
outcomes when making the MRI-based decisions retrospectively.
Materials and Methods
Patient Population and Treatment
The MRI protocol was proposed to patients with operable breast cancers
considered too large for breast-conserving surgery who underwent preoperative
chemotherapy at our institution. In the protocol, breast MRI evaluation was
scheduled before, during, and at the end of chemotherapy, producing a total of
three MR examinations per patient. These were added to the clinical and
radiologic evaluations used as the standard for therapeutic assessment. The
study was set up with no direct benefit to patients because surgeons did not
use the results of breast MRI to decide between postchemotherapy
breast-conserving surgery and mastectomy. MRI as a method for assessing
response was evaluated by correlation of the MRI results with pathologic
findings in surgical specimens. The local ethics committee approved the
study.
From November 1999 to December 2001, 30 consecutive patients (mean age, 49
years; range, 27–67 years) with stage II and III breast cancers
confirmed at core biopsy were entered in the study after giving written
consent. Tumor characteristics are shown in
Table 1. Neoadjuvant
chemotherapy consisted of four to six cycles of a combination of
anthracyclines (epirubicin hydrochloride, 100 or 75 mg/m2),
cyclophosphamide (500 mg/m2), and 5-fluorouracil (500
mg/m2) at 3- to 4-week intervals in 26 patients or of
anthracyclines (doxorubicin hydrochloride, 50 mg/m2) and taxanes
(docetaxel, 75 mg/m2) in four patients. Clinical response was
evaluated by the medical oncologist before each course of chemotherapy and by
the surgeon after two courses and at the end of induction therapy. Twenty-four
patients underwent all three MRI examinations, and six patients only had the
initial and final preoperative MRI evaluations.
Breast-conserving surgery was considered for all patients after
chemotherapy using the clinical, mammographic, and sonographic evaluations of
tumor response as the guide for resection without knowledge of the MRI
results. In patients with a limited response that precluded the possibility
that breast-conserving surgery would produce acceptable cosmetic results, a
mastectomy was performed. In patients with a complete or major response with
no distinct residual tumor at palpation and no or minor residual abnormalities
noted on conventional imaging, a wide lumpectomy at the initial tumor site was
performed. In patients with a partial response, the surgeon could either
perform a mastectomy or conserve the breast if the size of residual tumor
relative to the volume of the breast allowed it. Patients with involved
margins after breast-conserving surgery underwent a secondary mastectomy. All
patients had postoperative external irradiation directed to the breast or
thoracic wall and regional lymph nodes. Adjuvant chemotherapy was administered
in 23 of 30 patients.
Mammography and Breast Sonography
Patients referred to our center had initial mammograms and breast sonograms
obtained at outside facilities. Radiologists who performed the sonographic
examinations were aware of the presence of a breast mass considered to be
probably malignant either clinically or mammographically. After chemotherapy,
patients underwent preoperative mammography and sonography at our
institution.
Pathologic Analysis
All breast lumpectomy and mastectomy specimens were submitted for serial
gross pathologic examination and subsequent histologic analysis. The longest
diameter of residual gross abnormality and microscopic disease, histologic
type of invasive tumor, and percentage of in situ components were noted. In
lumpectomy cases, the distance from and location of any residual tumor focus
relative to specimen margins were specified. The appropriateness of
breast-conserving surgery was judged on pathologic confirmation of negative
excision margins. Response to chemotherapy was evaluated using the percentage
of residual tumor cells relative to tumor stroma and the mitotic index.
MRI
Breast MRI was performed on a 1.5-T whole-body imager (Signa, GE
Healthcare) using a bilateral breast coil. First, an unenhanced axial 3D fast
spoiled gradient-recalled echo T1-weighted sequence (TR/TE, 11.5/4.2; flip
angle, 30°; slice thickness, 3 mm; no gap; field of view, 34 cm; matrix,
512 x 192; number of excitations, 1; scanning time, 1 min 25 sec) was
acquired. Then the affected breast was imaged using sagittal 3D fast spoiled
gradient-recalled echo T1-weighted sequences (11.5/4.2; flip angle, 30°;
slice thickness, 4 mm; no gap; field of view, 24 cm; matrix, 512 x 192;
number of excitations, 1; scanning time, 45 or 58 sec) (3D volume including
all of the fibroglandular tissue) applied before and 10 times after bolus
injection of contrast medium (0.1 mmol/kg of body weight of dimeglumine
gadopentetate [Magnevist, Schering]) followed by a 20-mL saline IV flush.
Finally, the axial sequence was repeated as a contrast-enhanced study.
Postprocessing included unenhanced to enhanced image subtraction. The original
and subtracted images were analyzed and interpreted by a radiologist who was
aware of the mammographic and sonographic findings.
The diagnostic criterion for residual tumor after chemotherapy was the
visual observation of residual enhancement at the location of the initial
tumor as late as approximately 7–9 min after the injection of contrast
medium. Time-to-signal-intensity curves obtained either in the area of highest
or most homogeneous enhancement or in areas where visible residual enhancement
was detected were used to help differentiate residual tumor from adjacent
breast parenchyma. Enhancement profiles were also used to characterize
concomitant enhancing foci located at a distance from the primary tumor, using
diagnostic criteria described previously
[16].
The sizes of initial and residual tumors visualized as 3D images were
determined by the longest diameter in a given plane of a single tumor nodule
or by a composite measurement of multiple tumor nodules when those were
present. The European Organization for Research and Treatment of Cancer
[17] guidelines were used to
evaluate response to chemotherapy by changes of longest tumor diameters. A
decrease of 30% or more was considered a partial response, an increase of 20%
or more was considered progression, and a decrease of less than 30% or an
increase of less than 20% was considered to indicate stability.
Retrospectively, soft copies of the MRI examinations were reviewed by two
radiologists and a surgeon in consensus after the completion of all
treatments. Reviewers, blinded to the actual treatment course of the patients,
were presented with information on each MR examination individually. To
achieve comparative tumor measurements and to discern the pattern of response,
the reviewers first interpreted the current MRI examination for a given
patient and then systematically compared the intra- and postchemotherapy MR
images for that patient with initially obtained MR images. On prechemotherapy
MR images, primary known tumors were classified as predominantly circumscribed
nodular versus spiculated or infiltrating with jagged borders and as either
unifocal (a single circumscribed mass or an infiltrating tumor area) (Figs.
1A,
1B,
1C,
2A,
2B, and
2C) or multifocal (with
coexistence of two or more distinct tumor foci in the same quadrant) (Figs.
3A,
3B, and
3C). Suspicion of multicentric
disease was noted. Multicentricity was defined as the presence of one or more
suspicious enhancing foci in a quadrant different from that in which the index
tumor was found (Figs. 4A,
4B,
4C, and
4D) or the extension of a
tumor beyond the boundaries of the index quadrant. Tumor size was measured as
described earlier, and the type of response was categorized as either
shrinkage (concentric decrease of a tumor mass) or fragmentation into multiple
smaller tumor foci.
View larger version (144K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 1A. —42-year-old woman undergoing chemotherapy who had unifocal
tumor seen on initial MR images. Initial MR images (native images) obtained
before (A) and after (B) injection of contrast medium show one
circumscribed tumor (arrows).
|
|
View larger version (144K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 1B. —42-year-old woman undergoing chemotherapy who had unifocal
tumor seen on initial MR images. Initial MR images (native images) obtained
before (A) and after (B) injection of contrast medium show one
circumscribed tumor (arrows).
|
|
View larger version (142K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 1C. —42-year-old woman undergoing chemotherapy who had unifocal
tumor seen on initial MR images. MR image obtained after chemotherapy shows
marked concentric shrinkage of tumor (arrow) to single residue. Such
process was considered to indicate most favorable or ideal situation for
breast-conserving surgery, allowing complete tumor excision with standard
lumpectomy.
|
|
View larger version (122K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 2B. —67-year-old woman with unifocal tumor that fragmented while
she underwent chemotherapy. MR images obtained after two courses (B)
and at end of chemotherapy (C) show fragmentation of tumor, resulting
in multiple residual foci. MRI measurement of longest tumor diameter
underestimated residual 6-cm invasive ductal carcinoma tumor by 2 cm. This
type of response was considered to indicate unfavorable situation for
breast-conserving surgery. In this patient, MRI assessment would have led to
postchemotherapy mastectomy and allowed her to avoid undergoing additional
breast-conserving surgery that resulted in positive margins.
|
|
View larger version (119K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 2C. —67-year-old woman with unifocal tumor that fragmented while
she underwent chemotherapy. MR images obtained after two courses (B)
and at end of chemotherapy (C) show fragmentation of tumor, resulting
in multiple residual foci. MRI measurement of longest tumor diameter
underestimated residual 6-cm invasive ductal carcinoma tumor by 2 cm. This
type of response was considered to indicate unfavorable situation for
breast-conserving surgery. In this patient, MRI assessment would have led to
postchemotherapy mastectomy and allowed her to avoid undergoing additional
breast-conserving surgery that resulted in positive margins.
|
|
View larger version (47K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 3A. —In 57-year-old woman, multifocal tumor is seen on three
successive slices (shown across each row) from lateral to medial parts of
breasts. Initial MR images (A), MR images after two courses of
chemotherapy (B), and preoperative MR images after completion of
chemotherapy (C) all show two main central tumor nodules associated
with smaller peripheral nodules (arrows, A) confined in one
quadrant. B and C show nodule shrinkage in response to therapy
but persistence of two small nodules on preoperative MR images (C).
Patient is not suitable candidate for breast-conserving surgery.
|
|
View larger version (23K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 3B. —In 57-year-old woman, multifocal tumor is seen on three
successive slices (shown across each row) from lateral to medial parts of
breasts. Initial MR images (A), MR images after two courses of
chemotherapy (B), and preoperative MR images after completion of
chemotherapy (C) all show two main central tumor nodules associated
with smaller peripheral nodules (arrows, A) confined in one
quadrant. B and C show nodule shrinkage in response to therapy
but persistence of two small nodules on preoperative MR images (C).
Patient is not suitable candidate for breast-conserving surgery.
|
|
View larger version (24K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 3C. —In 57-year-old woman, multifocal tumor is seen on three
successive slices (shown across each row) from lateral to medial parts of
breasts. Initial MR images (A), MR images after two courses of
chemotherapy (B), and preoperative MR images after completion of
chemotherapy (C) all show two main central tumor nodules associated
with smaller peripheral nodules (arrows, A) confined in one
quadrant. B and C show nodule shrinkage in response to therapy
but persistence of two small nodules on preoperative MR images (C).
Patient is not suitable candidate for breast-conserving surgery.
|
|
View larger version (142K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 4A. —51-year-old women who had multifocal tumor with multicentric
disease suspected on initial MR images. Tumor nodules (arrowheads)
extend widely within breast from lateral (A) to medial parts (D)
beyond quadrant of index tumor, which was localized in retroareolar region
(B) by standard evaluation. C shows section through nipple. Had
MRI been considered, this patient would have undergone direct mastectomy,
avoiding preoperative chemotherapy and unsuccessful breast-conserving
surgery.
|
|
View larger version (148K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 4B. —51-year-old women who had multifocal tumor with multicentric
disease suspected on initial MR images. Tumor nodules (arrowheads)
extend widely within breast from lateral (A) to medial parts (D)
beyond quadrant of index tumor, which was localized in retroareolar region
(B) by standard evaluation. C shows section through nipple. Had
MRI been considered, this patient would have undergone direct mastectomy,
avoiding preoperative chemotherapy and unsuccessful breast-conserving
surgery.
|
|
View larger version (144K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 4C. —51-year-old women who had multifocal tumor with multicentric
disease suspected on initial MR images. Tumor nodules (arrowheads)
extend widely within breast from lateral (A) to medial parts (D)
beyond quadrant of index tumor, which was localized in retroareolar region
(B) by standard evaluation. C shows section through nipple. Had
MRI been considered, this patient would have undergone direct mastectomy,
avoiding preoperative chemotherapy and unsuccessful breast-conserving
surgery.
|
|
View larger version (146K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 4D. —51-year-old women who had multifocal tumor with multicentric
disease suspected on initial MR images. Tumor nodules (arrowheads)
extend widely within breast from lateral (A) to medial parts (D)
beyond quadrant of index tumor, which was localized in retroareolar region
(B) by standard evaluation. C shows section through nipple. Had
MRI been considered, this patient would have undergone direct mastectomy,
avoiding preoperative chemotherapy and unsuccessful breast-conserving
surgery.
|
|
Surgical Decisions Based on Standard Evaluation Versus Standard Evaluation Plus MRI
On the basis of the MRI review, the surgeon reevaluated the pre- and
postchemotherapy surgical decisions. Patients with suspected multicentric
disease on initial MRI should have undergone histologic evaluation so that the
extent of disease could be determined. In cases in which multicentric disease
was confirmed, a primary mastectomy should have been performed. Patients with
unicentric tumors were considered candidates for neoadjuvant chemotherapy. The
ideal situation corresponded to the progressive shrinkage of a circumscribed
tumor that made breast-conserving surgery feasible via standard lumpectomy.
The less-than-ideal situation corresponded to the fragmentation of a unifocal
mass or partial shrinkage of an initial multifocal tumor, both leading to
multiple residual tumor foci after chemotherapy. In patients with large
breasts, breast-conserving surgery would have necessitated a wide adapted
lumpectomy or quadrantectomy for complete tumor excision. In our protocol,
such patients undergo immediate breast reshaping with unilateral or bilateral
mammaplasty [18].
Results
Surgical Decisions Based on Standard Evaluation
Results are summarized in Table
2. After chemotherapy, 11 patients (40%) with poor tumor response
underwent a mastectomy and 19 patients (60%) eligible for breast-conserving
surgery were treated conservatively. Three of the 19 were subsequently shown
to have positive specimen margins and therefore underwent a secondary
mastectomy. Thus, the rate of radical surgery was 47% (14/30 patients).
View larger version (165K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 6B. —44-year-old woman with large unifocal breast tumor. MR image
obtained after two courses of chemotherapy shows marked fragmentation of tumor
but little change in its longest composite diameter.
|
|
View larger version (158K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 6C. —44-year-old woman with large unifocal breast tumor. MR image
obtained after chemotherapy shows only two or three low-enhancing residual
foci (arrows). Patient underwent secondary mastectomy after
breast-conserving surgery resulted in positive margins. Size of residual tumor
at histology (3.5-cm invasive lobular carcinoma) was underestimated by 1.5 cm.
MRI evaluation would not have prevented additional breast-conserving surgery
in this patient.
|
|
Histologically, there were one complete response with no residual tumor
cells, six partial responses (malignant cells 
20%), and 23 minor or
absent responses (malignant cells > 20%). The mean gross and microscopic
sizes of residual tumors measured on surgical specimens were 2.7 and 2.8 cm,
respectively (range, 0–7 cm). Of the 16 patients treated via
breast-conserving surgery with no further surgery, one patient had a complete
response. The other 15 patients with residual disease underwent complete
surgical excision.
The length of follow-up of the surviving 26 patients ranges from 19 to 42
months (mean, 30 months). One patient had local relapse 15 months after
breast-conserving surgery. Three patients died from breast cancer metastases
20, 25, and 30 months after surgery. One patient was lost to follow-up 20
months after surgery, with no recurrent disease at that time.
MRI Review
Tumor features observed on the prechemotherapy MR images are reported in
Table 3. Multicentric disease
was suspected in seven of 30 patients. The suspicious areas were undetected on
either mammography or breast sonography. The remaining 23 patients had
unicentric tumors. The mean size of the 30 primary tumors measured on initial
MR images was 4.5 cm (range, 2.4–9.0 cm).
Morphologic Changes from Chemotherapy
The patterns of response observed on MRI are summarized in Tables
4 and
5. Results in patients with
unicentric tumors are separated into two groups according to findings on the
standard evaluation (Table 4):
15 patients were considered responders, and eight were considered
nonresponders. In the responder group, five patients had unifocal tumors that
displayed marked shrinkage on MRI and were considered ideal candidates for
breast-conserving surgery (Figs.
1A,
1B, and
1C). In contrast, four patients
with unifocal tumors that fragmented (Figs.
2A,
2B, and
2C) and six patients with
multifocal tumors that showed either shrinkage (Figs.
3A,
3B, and
3C) or fragmentation were
considered less appropriate candidates for breast-conserving surgery. None of
the eight tumors in the group of standard nonresponders showed a significant
decrease in size on MRI. In these patients, limited changes toward shrinkage
were noted in six patients and toward fragmentation in two patients.
Considering all unicentric tumors (uni- and multifocal cases), 13 of the 15
nodular forms shrank, and two fragmented; five of the eight spiculated or
infiltrating forms fragmented, whereas three shrank
(Table 5). Thus, nodular tumors
tended to shrink rather than fragment and spiculated and irregular masses
tended to fragment rather than shrink. This result was statistically
significant (Fisher's exact test, p = 0.03). When the pattern of
response in all 30 primary tumors (i.e., uni- and multicentric cases,
Table 5) was considered, the
findings were confirmed with even higher significance (Fisher's exact test,
p = 0.005).
View this table:
[in this window]
[in a new window]
|
TABLE 4 Morphologic Pattern of Response Observed on MRI in 23 Unicentric Tumors,
Including 15 Responding and Eight Nonresponding, by Standard
Evaluation
|
|
Changes in Largest Diameter of the Tumor
The mean size of the 30 primary tumors measured on postchemotherapy MRI was
2.6 cm (range, 0–7.7 cm). The MRI and histologic sizes of residual tumor
after chemotherapy correlated with a Spearman rank correlation coefficient of
0.79 (p = 0.001) (Fig.
5). MRI underestimations of residual disease ranged from 0.1 to
3.5 cm in eight patients, including two false-negative cases (complete MRI
response but residual tumor found at pathology). MRI overestimations of
residual tumor ranged from 0.1 to 1.3 cm in 17 patients and reached 2.0 cm in
one patient. No false-positive finding was noted on MRI (residual suspicious
enhancement on MRI but complete response at pathology).
View larger version (17K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 5. —Graph shows MRI and histopathologic correlation of residual
tumor sizes after chemotherapy. MRI-measured longest diameters in centimeters
(x-axis) is compared with greatest dimensions of microscopic disease
on surgical specimens in centimeters (y-axis). Spearman rank
correlation coefficient = 0.79, p = 0.001.
|
|
Retrospective Decisions Based on Standard Plus MRI Evaluation Versus Standard Evaluation
The MRI-aided evaluation is summarized
Table 2. If MRI results had
been considered at presentation, seven patients with suspected multicentric
disease would have undergone a mastectomy if their histologic evaluation was
positive. The addition of MRI assessment would have allowed them to avoid
preoperative chemotherapy. The 23 remaining patients with unicentric tumors
would have received neoadjuvant chemotherapy. Because of the absence or
insufficiency of tumor response on standard plus MRI evaluation, eight of
these patients would have undergone postchemotherapy mastectomy. Five other
patients would have undergone a lumpectomy on the basis of the shrinkage of a
unifocal tumor (the ideal situation). The remaining 10 patients with multiple
tumor foci on preoperative MRI would have been assessed for the feasibility of
breast-conserving surgery on the basis of breast size. In fact, these
patients' breasts were large enough relative to residual tumor size to allow
complete tumor excision by wide quadrantectomy and breast reshaping.
Otherwise, multifocal tumor residue on preoperative MRI would have indicated
that mastectomy should be performed. Nine of these patients actually had
negative margins of resection, and one patient with positive margins underwent
a secondary mastectomy.
Table 2 also summarizes the
comparison between therapeutic decisions based on the standard evaluation and
on the standard plus MRI evaluation. Using MRI results, seven (23%) patients
would have undergone primary mastectomy. The MRI-added evaluation would also
have prevented one patient from having to undergo unsuccessful
breast-conserving surgery after chemotherapy (Figs.
4A,
4B,
4C, and
4D). However, another two of
these seven patients (2/30 total patients, 6.6%) eventually had successful
breast-conserving surgery with no recurrence detected at 34- and 29-months
follow-up. Using MRI-added evaluation for the 23 patients undergoing
chemotherapy would have resulted in a direct posttreatment mastectomy for a
stable fragmented tumor (Figs.
2A,
2B, and
2C) in one (1/30, 3.3%) but
would have not prevented unsuccessful breast-conserving surgery in another
(3.3%) with a fragmented tumor that showed marked—but
overestimated—response on MRI (Figs.
6A,
6B, and
6C). Overall, the standard
evaluation led to 16 successful breast-conserving surgeries and 14 mastectomy
procedures. Potentially, the standard plus MRI evaluation would have led to 14
successful breast-conserving surgeries (five lumpectomies and nine
quadrantectomies) and 16 mastectomies. The surgical alternative chosen in
patients with smaller breast volumes would have been five breast-conserving
surgeries by standard lumpectomy and 25 mastectomies.
Discussion
Initial MRI
Because unsuspected multicentric disease would have been detected at
presentation, seven of our 30 patients would have had their primary treatment
converted from chemotherapy to mastectomy, although two of these patients
ultimately had successful breast-conserving surgery and have had
recurrence-free survival after breast irradiation. In four patients, the tumor
extended beyond the volume of a single quadrant, and in three patients, MRI
showed distant suspicious foci. If multicentricity had been histologically
confirmed, a mastectomy would have been performed. The study was designed to
evaluate breast MRI but not direct surgical treatment; thus, the prospective
surgical decision was made without using the MRI information. Three patients
eventually had marked response to chemotherapy at standard assessment. One
patient in whom the tumor extended beyond a quadrant both initially (Figs.
4A,
4B,
4C, and
4D) and after treatment could
have avoided undergoing additional breast-conserving surgery after
chemotherapy. The other two patients with distant suspicious foci and
successful breast-conserving surgery at the index tumor site were mentioned
earlier. The suspicious foci were no longer visible on postchemotherapy MRI.
The remaining four patients with poor responses on standard evaluation
underwent mastectomy and thoracic wall irradiation. Three also had stable
tumors on MRI, but one patient with residual infiltrating lobular carcinoma at
histology had a false-negative MRI result.
The ability of breast MRI to reveal additional tumor foci not suspected
from the results on conventional imaging has been shown previously
[19–23].
In these reports, multifocal or multicentric cancer was found in 18–37%
of mastectomy or lumpectomy specimens, similar to our findings of 23% (7/30).
Investigators also addressed the impact of MRI on surgical treatment in women
with primary operable [24] or
early-stage [25,
26] breast cancer. In one
report [26], most treatment
changes that stemmed from MRI evaluations were found to be clinically
favorable. Some unfavorable effects, however, were linked to the lack of
specificity of breast MRI for cancer diagnosis.
As discussed earlier, the patients in our study in whom multicentric
disease was suspected should have undergone preoperative histologic
confirmation of tumor extent. The clinical benefit gained from using
MRI-guided localization and biopsy devices to assess "MRI-only"
lesions before treatment planning is a major advance
[27–30].
At the time of our study, we did not have an MRI-compatible biopsy device at
our center. The expected diffusion of such commercially available systems
should further improve the therapeutic management of patients in this
setting.
Assessing Response to Neoadjuvant Chemotherapy with MRI
Residual tumor extent.—In our series, the MRI and histologic
sizes of residual tumor after chemotherapy correlated well. Only minor
overestimations of residual disease were noted on MRI, except in one
nonresponding patient (nonresponsive according to both standard and MRI
assessments) in whom a 2-cm overestimation of residual tumor would have had no
impact on treatment plans. Two false-negative cases included a 3.5-cm invasive
lobular carcinoma with fragmentation pattern on MRI and a 1-cm invasive ductal
carcinoma with only 10% residual tumor cells that showed shrinkage. In
addition, MRI measurements underestimated 4.5- and 6-cm residual infiltrating
ductal carcinomas by 3.5 and 2 cm, respectively. Both tumors displayed
fragmentation. Histologically, malignant cells were arranged in multiple foci
separated by normal fibrous tissue. Both patients, showing complete clinical
and major radiologic responses, underwent breast-conserving surgery. The
patient (in Figs. 2A,
2B, and
2C) with a 6-cm residue had
positive margins of resection that necessitated a secondary mastectomy. The
patient with a 4.5-cm residue had negative margins. In a third patient (in
Figs. 6A,
6B, and
6C) in whom MRI showed
fragmentation and led to 1.5-cm underestimation of tumor size,
breast-conserving surgery would have been unsuccessful despite a marked
decrease in tumor diameter on MRI. Histology showed a 3.5-cm invasive lobular
carcinoma.
These results suggest that breast MRI is less accurate for assessing
residual tumor in patients in whom fragmentation has occurred during
chemotherapy. We think that surgeons should be aware of the risk of error
associated with MRI measurement of tumor with this type of response. Previous
reports have shown that residual disease can be underestimated on MRI after
induction therapy, particularly invasive lobular carcinomas
[10,
15].
Contribution to surgical planning.—Considering the actual
clinical outcome in our patients, both initial and subsequent intra- and
posttherapeutic MR images carried valuable information with regard to
therapeutic management. In six patients (6/30, 20%), this information would
have led to major appropriate changes in the course of treatment: five
patients with multicentric disease would have undergone immediate mastectomy
without neoadjuvant chemotherapy, and one patient with stable, fragmented
tumor on MRI would have avoided additional breast-conserving surgery by
undergoing direct postchemotherapy mastectomy. In another fourteen patients
(46.7%), MRI information would have helped surgeons plan breast surgery in
more detail or with more confidence. Five of the patients with a single,
circumscribed tumor residue assessed on serial MR images would have been
candidates for breast conservation with a standard lumpectomy. The other nine
patients with multiple residual foci and limited reduction in tumor size after
chemotherapy were less appropriate candidates for breast-conserving surgery.
These findings would have, in retrospect, changed treatment choice to a wider
excision (e.g., quadrantectomy) or a mastectomy. One patient (3.3%), who had
ipsilateral local recurrence 15 months after successful breast-conserving
surgery performed without using the MR images, displayed a multifocal residue
on preoperative MRI. It is possible that this recurrence was linked to disease
left in the breast outside the excisional volume. In another patient (3.3%),
however, MRI would not have prevented unsuccessful breast conservation before
ultimate mastectomy. Two patients (6.6%) in whom the MRI results would have
favored primary mastectomy because of the multicentric extent of the disease
were able to undergo postchemotherapy breast-conserving surgery with no local
regional relapse at follow-up. Finally, seven patients (23.3%) with similarly
poor tumor response on standard and MRI assessments would have undergone
postchemotherapy mastectomy with no changes made to the surgical plan.
Pattern of response to chemotherapy.—One recent study of
patients treated by neoadjuvant chemotherapy suggested a link between tumor
morphology and the degree of response measured on breast MRI
[14]. Dominant nodular
presentation with circumscribed borders generally responded by shrinking to a
smaller mass, was predictive of a partial or complete response, and was likely
to be resectable after induction therapy. Other MRI patterns such as multiple
tumor nodules, diffuse tissue infiltration, and sparse enhancement of a tumor
mass were either predictive of a poorer response with no significant shrinkage
or little change in tumor size. In this series, we found a similar link
between initial tumor morphology and the type of response (e.g., overall
nodular vs spiculated shape and a tendency to respond by shrinking vs
fragmenting, respectively). However, shrinkage did not necessarily lead to a
complete response on MRI because some peripheral nodules seen at presentation
in the vicinity of a main tumor mass were still present after chemotherapy.
Particularly in cases with marked response on standard evaluation,
visualization of a multifocal residue on MRI adds useful information for
surgical planning. Then, a wider excision than that planned using non-MRI
assessment might have been required to achieve complete removal of the
tumor.
In summary, in this series of large operable breast cancers, MRI was useful
in establishing the final treatment earlier in the treatment process. Initial
MR images obtained before chemotherapy would have been useful in revealing
potential multicentric disease, thus obviating unnecessary chemotherapy and,
in one patient, unsuccessful breast-conserving surgery. Preoperatively, the
morphologic assessment of tumor response would have helped surgeons to select
patients most likely to have successful breast conservation after resection of
a unifocal residue and optimize the extent of surgical treatment necessary to
remove all residual tumor in multifocal situations.
Acknowledgments
We thank all the radiologists, clinicians, and pathologists who
participated in this study.
References
- Powles TJ, Hickish TF, Makris A, et al. Randomized trial of
chemoendocrine therapy started before or after surgery for treatment of
primary breast cancer. J Clin Oncol1995; 13:547
-552[Abstract/Free Full Text]
- Fisher B, Bryant J, Wolmark N, et al. Effect of preoperative
chemotherapy on the outcome of women with operable breast cancer. J
Clin Oncol 1998;16:2672
-2685[Abstract]
- Mauriac L, MacGrogan G, Avril A, et al. Neoadjuvant chemotherapy
for operable breast carcinoma larger than 3 cm: a unicentre randomized trial
with a 124-month median follow-up—Institut Bergonie Bordeaux Groupe Sein
(IBBGS). Ann Oncol1999; 10:47
-52[Abstract/Free Full Text]
- Rouzier R, Extra JM, Carton M, et al. Primary chemotherapy for
operable breast cancer: incidence and prognostic significance of ipsilateral
breast tumor recurrence after breast-conserving surgery. J Clin
Oncol 2001;19:3828
-3835[Abstract/Free Full Text]
- van der Hage JA, van de Velde CJ, Julien JP, Tubiana-Hulin M,
Vandervelden C, Duchateau L. Preoperative chemotherapy in primary operable
breast cancer: results from the European Organization for Research and
Treatment of Cancer trial 10902. J Clin Oncol2001; 19:4224
-4237[Abstract/Free Full Text]
- Newman LA, Buzdar AU, Singletary SE, et al. A prospective trial of
preoperative chemotherapy in resectable breast cancer: predictors of
breast-conservation therapy feasibility. Ann Surg
Oncol 2002;9:228
-234[Medline]
- Wolmark N, Wang J, Mamounas E, Bryant J, Fisher B. Preoperative
chemotherapy in patients with operable breast cancer: nine-year results from
National Surgical Adjuvant Breast and Bowel Project B-18. J Natl
Cancer Inst Monogr 2001:96
-102
- Gilles R, Guinebretiere JM, Toussaint C, et al. Locally advanced
breast cancer: contrast-enhanced subtraction MR imaging of response to
preoperative chemotherapy. Radiology1994; 191:633
-638[Abstract/Free Full Text]
- Abraham DC, Jones RC, Jones SE, et al. Evaluation of neoadjuvant
chemotherapeutic response of locally advanced breast cancer by magnetic
resonance imaging. Cancer1996; 78:91
-100[Medline]
- Rieber A, Zeitler H, Rosenthal H, et al. MRI of breast cancer:
influence of chemotherapy on sensitivity. Br J Radiol1997; 70:452
-458[Abstract]
- Weatherall PT, Evans GF, Metzger GJ, Saborrian MH, Leitch AM. MRI
vs. histologic measurement of breast cancer following chemotherapy: comparison
with X-ray mammography and palpation. J Magn Reson
Imaging 2001;13:868
-875[Medline]
- Drew PJ, Kerin MJ, Mahapatra T, et al. Evaluation of response to
neoadjuvant chemoradiotherapy for locally advanced breast cancer with dynamic
contrast-enhanced MRI of the breast. Eur J Surg Oncol2001; 27:617
-620[Medline]
- Partridge SC, Gibbs JE, Lu Y, Esserman LJ, Sudilovsky D, Hylton NM.
Accuracy of MR imaging for revealing residual breast cancer in patients who
have undergone neoadjuvant chemotherapy. AJR2002; 179:1193
-1199[Abstract/Free Full Text]
- Esserman L, Kaplan E, Partridge S, et al. MRI phenotype is
associated with response to doxorubicin and cyclophosphamide neoadjuvant
chemotherapy in stage III breast cancer. Ann Surg
Oncol 2001;8:549
-559[Medline]
- Rieber A, Brambs HJ, Gabelmann A, Heilmann V, Kreienberg R, Kuhn T.
Breast MRI for monitoring response of primary breast cancer to neoadjuvant
chemotherapy. Eur Radiol2002; 12:1711
-1719[Medline]
- Kuhl CK, Mielcareck P, Klaschik S, et al. Dynamic breast MR
imaging: are signal intensity time course data useful for differential
diagnosis of enhancing lesions? Radiology1999; 211:101
-110[Abstract/Free Full Text]
- Therasse P, Arbuck SG, Eisenhauer EA, et al. New guidelines to
evaluate the response to treatment in solid tumors: European Organization for
Research and Treatment of Cancer, National Cancer Institute of the United
States, National Cancer Institute of Canada. J Natl Cancer
Inst 2000;92:205
-216[Abstract/Free Full Text]
- Clough KB, Lewis JS, Couturaud B, Fitoussi A, Nos C, Falcou MC.
Oncoplastic techniques allow extensive resections for breast-conserving
therapy of breast carcinomas. Ann Surg2003; 237:26
-34[Medline]
- Harms SE, Flamig DP, Hesley KL, et al. MR imaging of the breast
with rotating delivery of excitation off resonance: clinical experience with
pathologic correlation. Radiology1993; 187:493
-501[Abstract/Free Full Text]
- Boetes C, Mus RD, Holland R, et al. Breast tumors: comparative
accuracy of MR imaging relative to mammography and US for demonstrating
extent. Radiology1995; 197:743
-747[Abstract/Free Full Text]
- Orel SG, Schnall MD, Powell CM, et al. Staging of suspected breast
cancer: effect of MR imaging and MR-guided biopsy.
Radiology1995; 196:115
-122[Abstract/Free Full Text]
- Mumtaz H, Hall-Craggs MA, Davidson T, et al. Staging of symptomatic
primary breast cancer with MR imaging. AJR1997; 169:417
-424[Abstract/Free Full Text]
- Drew PJ, Chatterjee S, Turnbull LW, et al. Dynamic contrast
enhanced magnetic resonance imaging of the breast is superior to triple
assessment for the pre-operative detection of multifocal breast cancer.
Ann Surg Oncol1999; 6:599
-603[Medline]
- Esserman L, Hylton N, Yassa L, Barclay J, Frankel S, Sickles E.
Utility of magnetic resonance imaging in the management of breast cancer:
evidence for improved preoperative staging. J Clin
Oncol 1999;17:110
-119[Abstract/Free Full Text]
- Tan JE, Orel SG, Schnall MD, Schultz DJ, Solin LJ. Role of magnetic
resonance imaging and magnetic resonance imaging-guided surgery in the
evaluation of patients with early-stage breast cancer for breast conservation
treatment. Am J Clin Oncol1999; 22:414
-418[Medline]
- Tillman GF, Orel SG, Schnall MD, Schultz DJ, Tan JE, Solin LJ.
Effect of breast magnetic resonance imaging on the clinical management of
women with early-stage breast carcinoma. J Clin Oncol2002; 20:3413
-3423[Abstract/Free Full Text]
- Orel SG, Schnall MD, Newman RW, Powell CM, Torosian MH, Rosato EF.
MR imaging-guided localization and biopsy of breast lesions: initial
experience. Radiology1994; 193:97
-102[Abstract/Free Full Text]
- Heywang-Kobrunner SH, Heinig A, Pickuth D, Alberich T, Spielmann
RP. Interventional MRI of the breast: lesion localisation and biopsy.
Eur Radiol2000; 10:36
-45[Medline]
- Kuhl CK, Morakkabati N, Leutner CC, Schmiedel A, Wardelmann E,
Schild HH. MR imaging-guided large-core (14-gauge) needle biopsy of small
lesions visible at breast MR imaging alone. Radiology2001; 220:31
-39[Abstract/Free Full Text]
- Morris EA, Liberman L, Dershaw DD, et al. Preoperative MR
imaging-guided needle localization of breast lesions.
AJR 2002;178:1211
-1220[Abstract/Free Full Text]
- Elston CW, Ellis IO. Pathological prognostic factors in breast
cancer: I. The value of histological grade in breast cancer: experience from a
large study with long-term follow-up. Histopathology1991; 19:403
-410[Medline]
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
Top
Abstract
Introduction
