Biphasic metaplastic sarcomatoid carcinoma of the breast

Document Sample
Biphasic metaplastic sarcomatoid carcinoma of the breast Powered By Docstoc
					                                                                                                                        Annals of Oncology 17: 605–613, 2006
original article                                                                                                                  doi:10.1093/annonc/mdl006
                                                                                                                             Published online 9 February 2006




Biphasic metaplastic sarcomatoid carcinoma of
the breast
B. T. Hennessy1*, S. Giordano1, K. Broglio2, Z. Duan1, J. Trent3, T. A. Buchholz4, G. Babiera5,
G. N. Hortobagyi1 & V. Valero1
Departments of 1Breast Medical Oncology, 2Biostatistics, 3Sarcoma, 4Radiation Oncology and 5Surgical Oncology, The University of Texas M. D. Anderson Cancer
Center, Houston, USA

Received 18 November 2005; accepted 22 December 2005


Background: Breast biphasic metaplastic sarcomatoid carcinoma (MSC) is rare and aggressive. We analyzed
100 patients treated at M. D. Anderson Cancer Center (MDACC) with 213 MSC and 98 carcinosarcoma patients




                                                                                                                                                                  Downloaded from http://annonc.oxfordjournals.org/ by guest on March 25, 2012
identified through the Surveillance, Epidemiology and End-Results (SEER) database to describe clinical and pathologic
characteristics.
Patients and methods: We searched the MDACC (1985–2001) and SEER databases (1988–2001) for breast MSC
and carcinosarcoma patients.
Results: We identified 100 MDACC MSC patients: 66% had node-negative disease and 6% distant metastases at
presentation. Median recurrence-free survival (RFS) of 94 patients with stages I–III disease was 74 months (range
3–74), with 52% 5-year RFS [95% confidence interval (CI) 0.42–0.63]. Median overall survival in these patients was not
reached, with 64% 5-year survival (95% CI 0.54–0.75). The initial stage of the tumor, but not use of adjuvant chemo- or




                                                                                                                                                                original
radiotherapy, had a strong association with outcome. The pathologic complete response rate to neoadjuvant




                                                                                                                                                                article
chemotherapy was 10%. Median survival from the time of recurrent disease was 14 months (range 1–55). Tumors
were usually hormone receptor- and HER2/neu-negative. SEER data were consistent with MDACC findings.
Conclusions: Breast MSC and carcinosarcoma are aggressive, treatment-refractory tumors with shared clinical
features and outcome similar to poorly differentiated receptor-negative adenocarcinomas. New therapeutic agents
are needed.
Key words: breast, cancer, sarcomatoid, carcinosarcoma




introduction                                                                     carcinomas with overt carcinomatous and sarcomatous features.
                                                                                 Where an epithelial component is detected only at electron
Biphasic metaplastic sarcomatoid carcinomas (MSC) of the                         microscopy or immunohistochemistry, the term ‘monophasic
breast account for <0.1% of all breast malignancies [1, 2].                      sarcomatoid carcinoma’ is used [12].
They are a form of metaplastic breast cancer with overt                             Previous reports suggest MSCs behave biologically
carcinomatous and sarcomatous features apparent with both                        differently from conventional carcinoma of the breast with
light microscopy and immunohistochemical testing [3–5].                          sarcomatoid characteristics dominating the clinical course
Although the term ‘sarcomatoid carcinoma’ has traditionally                      [1, 2]. Generally, these tumors are poorly differentiated and
been employed if there was obvious microscopic transition                        appear to be aggressive. However, there are very few reported
between the two components, with the term ‘carcinosarcoma’                       series describing the clinical behavior of breast MSCs and even
reserved for when the demarcation between carcinomatous and                      fewer of carcinosarcomas. For the purposes of the current
sarcomatous components was distinct in all light microscopic                     study, we studied the clinical features, management and
fields, increasing evidence suggests these are similar monoclonal                 outcome of MSC of the breast in a series of 100 women treated
tumors developing through a malignant transformation of                          at The University of Texas M.D. Anderson Cancer Center and
myoepithelial cells or myofibroblastic metaplasia of malignant                    in a series of 213 MSC patients identified through the
epithelial cells [6–12]. Thus, the term ‘biphasic sarcomatoid                    Surveillance, Epidemiology and End-Results (SEER) database
carcinoma’ may be an appropriate description of all breast                       to determine if MSC behaves differently from high-grade
                                                                                 hormone receptor-negative adenocarcinoma. To determine if
*Correspondence to: Dr B. Hennessy, Department of Medical Oncology, Box 10,
                                                                                 the clinical behavior of breast carcinosarcomas is similar to
The University of Texas M. D. Anderson Cancer Center, 1515 Holcombe Blvd.,
Houston, TX 77030, USA. Tel: +1-713-792-2740; Fax: +1-713-792-3708;              that of MSC, we also analyzed 98 breast carcinosarcoma
E-mail: bhennessy@mdanderson.org                                                 patients identified from the SEER database.


ª 2006 European Society for Medical Oncology
original article                                                                                                           Annals of Oncology


Table 1. Patient and disease characteristics


                           Patients without metastatic   Patients with metastatic   MSC; SEER                      Carcinosarcoma; SEER
                           disease (N = 94)              disease (N = 6)            database (N = 213)             database (N = 98)
                           n                   %         n               %          n                %             n                 %
Race
  Black                      13                14        1                 17        25              12             11               11
  Other                       3                 3        1                 17        16               8              4                4
  Hispanic                   10                11        1                 17
  White                      68                72        3                 50       172              81             83               85
Grade
  1/2                         4                 4        1                 17        25              12              6                6
  3                          71                76        5                 83        117a            55              62a             63
  Unknown                    19                19                                    71              33             30               31
T stage (clinical)
  T1                         20                21        0                  0        50              24             16               16
  T2                         45                48        1                 17       108              51             53               54
  T3                         15                16        0                  0        53              25             20               20
  T4                          7                 7        5                 83




                                                                                                                                                Downloaded from http://annonc.oxfordjournals.org/ by guest on March 25, 2012
  TX                          8                 9                                       2             1                9              9
N stage (clinical)
  N0                         62                66        0                  0       137              64             60               61
  N1                         17                18        3                 50        25              12             11               11
  N2                          5                 5        2                 33         5               2              2                2
  N3                          5                 5        1                 17         5               2              1                1
  NX                          5                 5                                    41              19             24               24
Disease stage (clinical) path
  I                          17                18        NA                          46              22             20               20
  II                         54                57        NA                         125              59             54               55
  III                        15                16        NA                          28              13              7                7
  IV                         NA                          6               100         14               7              6                6
  Unknown                     8                8                                                                    11               11
ER status
  Negative                   73                78        5                 83       141              66             64               65
  Positive                    6                 6        1                 17        18               8              4                4
  Unknown                    15                16                                    54              25             30               31
PR status
  Negative                   70                74        6               100        136              64             61               62
  Positive                    8                 9        0                 0         20              10              7                7
  Unknown                    16                17                                    57              27             30               31
Her2
  Negative                   23                24        1                 17               Not available                   Not available
  Positive                    2                 2        0                  0
  Unknown                    69                74        5                 83
Surgery
  MRM                        62                66        1               117         17              55             50               51
  None                        1                 1        5                83          3               1              6                6
  Seg                        27                29        0                 0        73b              34            33c               34
  Simple                      2                 2        0                 0         19               9              9                9
  Unknown                     2                 2                                     1               0              0                0
Adjuvant chemotherapy
  No                         12                13        NA                                 Not available                   Not available
  Yes                        77                82        NA
  Unknown                     5                 5
Chemotherapy regimen
  AI                          3                 3                                           Not available                   Not available
  CMF                         8                 9
  Anthracycline              53                56
  Anthracycline/taxan        13                14
  None                       12                13




606 | Hennessy et al.                                                                                       Volume 17 | No. 4 | April 2006
Annals of Oncology                                                                                                    original article
Table 1. (Continued)


                   Patients without metastatic              Patients with metastatic                MSC; SEER                          Carcinosarcoma; SEER
                   disease (N = 94)                         disease (N = 6)                         database (N = 213)                 database (N = 98)
                   n                    %                   n                    %                  n                 %                n                %
Adjuvant radiotherapy
  No             37                     39                  NA                                      142               67               65               66
  Yes            51                     54                  NA                                       64               30               31               32
  Unknown          6                     7                  NA                                        7                3                2                2
Adjuvant tamoxifen
  No             77                     82                  NA                                               Not available                     Not available
  Yes              9                    10                  NA
  Unknown          8                     8                  NA

MSC, metaplastic sarcomatoid carcinoma; SEER, Surveillance, Epidemiology and End Results; NA, not applicable; MRM, modified radical mastectomy;
Seg, segmental mastectomy; AI, doxorubicin, ifosfamide; CMF, cyclophosphamide, methotrexate and 5-fluorouracil.
a
  In the SEER database, grade is actually categorized in four groups including grade 3 (poorly differentiated) and 4 (undifferentiated/anaplastic). For ease
of tabulation, we incorporated SEER grades 3 and 4 into grade 3.
b
  Of these, 23 did not have an axillary dissection.




                                                                                                                                                                         Downloaded from http://annonc.oxfordjournals.org/ by guest on March 25, 2012
c
  Of these, 10 did not have an axillary dissection.


patients and methods                                                                 date of recurrence to the date of death or last follow-up. The univariate
                                                                                     log-rank test was used to evaluate possible associations between survival
M. D. Anderson Cancer Center Patients                                                and patient covariates. Hazard ratios were estimated with univariate Cox
The M. D. Anderson institutional computerized database was searched to               proportional hazards models. All P values presented are two-sided.
identify patients with a diagnosis of breast MSC evaluated and treated               Statistical analyses were performed with STATA 7.0 software (StataCorp.,
here between 1985 and 2001. One hundred patients were identified. Their               College Station, TX) in addition to SAS version 8 (Cary, NC) and S-Plus
medical records were retrospectively reviewed to obtain demographic,                 version 6.1 (Seattle, WA). Patients with metastatic disease at diagnosis (M1)
clinicopathologic characteristics, treatment and outcome information                 were considered separately from patients without metastatic disease (M0)
from an institutional review board-approved protocol.                                for all analyses. P values are not shown where there are less than 10 patients
   All patients were initially confirmed to have MSC of the breast after review       in a group as statistical inference is limited due to small sample size.
of the pathologic material at our institution by a breast pathologist. Tumors           N stage, T stage, overall stage, estrogen receptor (ER) status, grade and type
were classified and staged according to clinical and pathologic information           of surgery were identified as being prognostic factors of interest for
(American Joint Committee on Cancer Staging Manual, 6th edition). For                a multivariable model. Because most patients were ER negative and had grade
patients who were treated with chemotherapy for metastatic disease, their            3 disease, these variables were excluded from modeling. Cox proportional
response status was classified on the basis of standard criteria [13].                hazards model were fit for OS, RFS and local recurrence-free survival that
SEER patients                                                                        included type of surgery, N stage and T stage. The fit of the model and the
We also studied women with MSC and carcinosarcoma of the breast                      proportional hazards assumption was assessed visually with residual plots.
entered in the SEER database program, a population-based tumor registry                 For the purposes of comparing the outcomes of the M. D. Anderson MSC
sponsored by the National Cancer Institute, consisting of tumor registries           patients to those of patients with breast ductal adenocarcinoma, patients
that collect information on all newly diagnosed cancer cases that occur              with ductal histology were matched 2:1 to the sarcomatoid patients on both
in persons residing in 11 SEER participating areas. Patients were included           ER and PR status. Matching was done using the SAS macro %match [14].
in our study if they were women who had stage I–IV breast cancer
(American Joint Committee on Cancer Staging Manual, 3rd edition),                    results
were diagnosed between 1988 and 2001 and had metaplastic sarcomatoid
carcinoma (identified by histologic codes 8030, 8032, 8033, 8572, and                 Table 1 summarizes the characteristics of the patients and the
8575) or carcinosarcoma (histologic codes 8980/1).                                   treatments they received. Their median age was 48 years (range
                                                                                     26–77). Ninety-four patients did not have distant metastatic
data analysis
                                                                                     disease (M0) at diagnosis. Among these patients, the median
Descriptive statistics were used to characterize the patient population.
                                                                                     follow-up was 55 months (range 4–252 months). The majority
Recurrence-free survival (RFS), disease-specific survival (DSS) and overall
survival (OS) curves were calculated using the method of Kaplan and
                                                                                     of tumors were high-grade, lymph node- and hormone
Meier. Overall survival was calculated from the date of surgery to the date
                                                                                     receptor-negative and, for those in which HER2/neu status
of death or at last follow-up. RFS and time to distant recurrence was                (immunohistochemistry or fluorescent in situ hybridization)
calculated from the date of surgery to the date of disease recurrence or last        was available, HER2/neu-negative.
follow-up. Patients experiencing disease recurrence locally or in both                  Of 94 MDACC patients diagnosed with M0 disease, 62 had
a distant and a local site were considered censored for distant recurrence at        modified radical mastectomy (MRM) and 27 segmental
the date of their disease recurrence; patients who died without recurrence           mastectomy. The majority of these patients had an axillary node
were considered censored at their date of death. Time to local recurrence was        dissection; however, to facilitate analysis, we included two
calculated in the same manner. Survival following disease recurrence was             patients in the MRM group who had mastectomy with a sentinel
measured among patients who experienced a disease recurrence from the                lymph node biopsy and one patient in the segmental


Volume 17 | No. 4 | April 2006                                                                                             doi:10.1093/annonc/mdl006 | 607
original article                                                                                                               Annals of Oncology


mastectomy group who had a sentinel node biopsy, in addition,                chemotherapy. In the remaining patients, there was at least
two patients in each group did not have axillary surgery.                    5–10 mm between the tumor and the margins.
The average number of nodes removed was 15 versus 17,                           Twenty-six MRM patients had adjuvant radiation therapy,
respectively. In the 62 MRM and 27 segmental mastectomy                      including five of the six patients with positive or close margins.
patients, no information on margin status was available for 10               Twenty-three segmental mastectomy patients had adjuvant
and five patients, respectively; two and one patients had an                  radiation therapy, including three of the four patients with
involved margin, while the tumor was within 1–5 mm of the                    positive or close margins. This was usually 50 Gy in 25 fractions
margins in another four and three patients, respectively. Twelve             to the chest wall or breast with or without the nodal basins
patients in each group had no residual disease at definitive                  followed by a 10 Gy boost.
surgery following a prior lumpectomy and one patient in                         Table 2 shows the results of the univariate analysis of overall
each group had a pathologic complete response to primary                     survival in those with M0 disease at diagnosis. This analysis

Table 2. Overall survival (89 MDACC patients without metastatic disease who had overall survival information available)


                                  N             No. of            Median              Proportion                  95% CI              P value
                                                events            (months)            surviving 5 years
Overall                           89            35                 –                  0.64                        54%–75%
Age at diagnosis
  <50                             45            20                83                  0.58                        45%–76%




                                                                                                                                                    Downloaded from http://annonc.oxfordjournals.org/ by guest on March 25, 2012
  ‡50                             44            15                 –                  0.69                        56%–85%              0.31
T stage
  T1                              19             5                 –                  0.77                        60%–100%
  T2                              42            15                 –                  0.63                        49%–81%
  T3                              14             6                 –                  0.64                        44%–95%
  T4                               7             6                18                  0.00                        –                   –
N stage
  N0                              59            19                 –                  0.71                        60%–84%
  N1                              15             7                83                  0.54                        33%–89%
  N2                               5             3                24                  0.40                        14%–100%
  N3                               5             4                22                  0.00                        –                   –
Disease stage
  I                               16             3                 –                  0.86                        70%–100%
  II                              51            20                 –                  0.62                        49%–77%
  III                             14             9                24                  0.31                        14%–70%              0.003
ER status
  Negative                        69            31                89                  0.55                        44%–69%
  Positive                         6             2                 –                  0.83                        58%–100%            –
Surgery
  MRM                             57            20                 –                  0.65                        53%–79%
  None                             1             0                 –                  1.00                        –
  Seg                             27            13                83                  0.64                        47%–86%
  Simple                           2             2                30.5                0.00                        –                   –
Chemotherapy
  No                              12             4                89                  0.73                        51%–100%
  Yes                             72            31                94                  0.60                        49%–73%              0.41
Chemotherapy regimen
  AI                               3             0                 –                  1.00                        –
  CMF                              8             6                56                  0.45                        20%–100%
  Anthracycline                   48            21                 –                  0.57                        45%–73%
  Anthracycline/taxane            13             4                 –                  0.73                        51%–100%
  None                            12             4                89                  0.73                        51%–100%            –
Radiotherapy
  No                              35            12                 –                  0.68                        53%–87%
  Yes                             48            22                83                  0.58                        46%–75%              0.41
Tamoxifen
  No                              73            31                94                  0.58                        47%–71%
  Yes                              9             2                 –                  0.89                        71%–100%            –

In some categories, data on a small number of patients was not available.
ER, estrogen receptor; MRM, modified radical mastectomy; Seg, segmental mastectomy; AI, doxorubicin, ifosfamide; CMF, cyclophosphamide,
methotrexate and 5-fluorouracil.



608 | Hennessy et al.                                                                                              Volume 17 | No. 4 | April 2006
Annals of Oncology                                                                                       original article
includes 89 patients who had survival information available. The                Of those who recurred, 22 patients experienced a distant
other five patients were lost to follow-up. Thirty-five patients               recurrence initially, six patients recurred in both local and
died and the proportion of patients who remained alive at                    distant sites and 16 patients initially recurred only locally. The
5 years was 0.64 (95% CI 0.54–0.75; Figure 1). Patients with                 median times to distant and local recurrence have not been
positive lymph nodes had 2.5 times the risk of death                        attained and the proportions remaining free of distant and
compared with patients with negative lymph nodes [P = 0.01,                  local recurrence at 5 years were 0.72 (95% CI 0.63–0.83) and
hazard ratio (HR) 2.51, 95% CI 1.25–5.04]. Patients with T4                  0.75 (95% CI 0.64–0.87), respectively. The median survival
tumors had >18 times the risk of death compared with patients                following any disease recurrence was 14 months. Patients who
with T1 tumors (HR 18.33, 95% CI 4.9–68.51), and patients                    had positive lymph nodes at initial diagnosis tended to have
with stage 3 disease had almost seven times the risk of death                shorter survival following recurrence (P = 0.01). Of six
compared with patients with stage 1 disease (HR 6.93, 95% CI                 patients diagnosed initially with metastatic disease, all have
1.86–25.85). OS was not statistically different between patients             died and their median survival was 12 months (range 6–55
who did and did not receive chemotherapy or radiotherapy.                    months).
However, these data are not randomized and there were                           Twenty-two patients experienced a local disease recurrence as
significantly more patients in the chemotherapy-treated group                 initial presentation of recurrent disease, in 14 cases following
with higher grade (P = 0.005) and higher stage (P = 0.05)                    prior adjuvant radiation therapy of whom nine were infield
tumors.                                                                      relapses. Only T stage and type of surgery were significantly
   Table 3 shows the results of the univariate analysis of RFS in            related to local RFS. Patients with T4 tumors had 11.3 times the
those with M0 disease at diagnosis. This analysis includes 91                risk of local recurrence compared with patients with T1 tumors




                                                                                                                                                   Downloaded from http://annonc.oxfordjournals.org/ by guest on March 25, 2012
patients who had recurrence information available. Forty-four                (95% CI 2.51–50.98). There was no increased risk of local
patients have experienced a disease recurrence and the median                recurrence associated with initial N status. Patients who received
RFS was 74 months (Figure 2). Patients with T4 tumors had 9.2                radiation had a non-significant increase in the risk of local
times the risk of recurrence compared with patients with T1                  recurrence compared with patients who did not (HR 1.62, 95%
tumors (HR 9.2, 95% CI 3.08–27.55), patients with node-                      CI 0.65–4.01). Those patients who were treated with adjuvant
positive disease had a HR of recurrence of 1.85 compared with                radiotherapy more often underwent breast-conserving surgery
those with node-negative tumors (95% CI 1–3.42), and patients                (P = 0.0004), as would be expected, were younger at diagnosis
with stage 3 disease had 3.77 times the risk of recurrence                   (P = 0.03) and more often had higher stage tumors (P = 0.04).
compared with patients with stage 1 disease (HR 3.77, 95% CI                 Among the 59 patients with recurrence information who
1.28–11.1). RFS was not statistically different between patients             underwent modified radical mastectomy (MRM), 33 did not
who did and did not receive chemotherapy or radiotherapy.                    receive radiotherapy. In the MRM group, OS, RFS and local
Again, these data are not randomized. Of 12 patients with T2                 recurrence-free survival were not statistically different between
node-positive disease treated with adjuvant chemotherapy, eight              patients who did and did not receive radiotherapy, nor was there
have relapsed. Of 10 patients with T3/4 node-negative disease                any clear trend in survival between the two groups. Figure 3
treated with adjuvant chemotherapy, seven have relapsed, and                 shows the Kaplan–Meier curves of local RFS by radiation and
of 10 patients with T3/4 node-positive disease treated with                  surgery treatment groups (P = 0.04). Among patients with
adjuvant chemotherapy, six have relapsed.                                    MRM and no radiation versus adjuvant radiation, local
   RFS and OS tended to be shorter among patients who were                   recurrence-free survival at 5 years was 81% (95% CI 0.66–1)
ER/PR negative than among patients who were ER/PR positive                   versus 77% (95% CI 0.61–0.97), respectively. Among patients
although this was not statistically significant (P = 0.43 and 0.19,           with segmental mastectomy and no radiation versus
respectively), possibly as a result of limited statistical power.            adjuvant radiotherapy, local recurrence-free survival at 5 years
                                                                             was 50% (95% CI 0.19–1) versus 54% (95% CI 0.35–0.82),
                                                                             respectively.
                                                                                Of the patients treated with adjuvant chemotherapy, 21
                        1.0




                                                                             received primary chemotherapy; this was anthracycline-based
                                                                             in all patients and included a taxane in five patients and
                        0.8




                                                                             ifosfamide in one patient. There were two pathologic complete
 Proportion Surviving




                                                                             responses (10%), one complete clinical response (5%) and four
                        0.6




                                                                             clinical partial responses (20%). Of these seven responding
                                                                             patients, all had four or six cycles of conventional 5-fluorouracil,
                        0.4




                                                                             doxorubicin and cyclophosphamide (FAC) chemotherapy prior
                                                                             to surgery.
                        0.2




                                                                             multivariable model
                        0.0




                                                                             In the Cox proportional hazards model, only T stage
                              0   50        100       150        200   250
                                                                             remained statistically significant for OS in those with stages
                                       Months from Date of Surgery
                                                                             I–III disease at diagnosis. T stage and surgery type were
Figure 1. Overall survival (with 95% CI shown by the broken lines)           significantly related to both RFS and local recurrence-free
among patients without metastatic disease at diagnosis.                      survival. Having a segmental mastectomy was associated with


Volume 17 | No. 4 | April 2006                                                                             doi:10.1093/annonc/mdl006 | 609
original article                                                                                                              Annals of Oncology


Table 3. Recurrence-free survival (91 MDACC patients without metastatic disease who had recurrence information available)


                                 N            No. of           Median              Proportion surviving           95% CI              P value
                                              events           (months)            5 years
Overall                          91           44               74                  0.52                           42%–63%
Age at diagnosis
  <50                            47           26               28                  0.44                           32%–61%
  ‡50                            44           18                –                  0.60                           47%–77%             0.14
T stage
  T1                             19            7                –                  0.67                           48%–93%
  T2                             44           22               59                  0.47                           34%–66%
  T3                             14            6                –                  0.57                           36%–90%
  T4                              7            7               11                  0.00                           –                   –
N stage
  N0                             60           26                –                  0.57                           45%–71%
  N1                             16            9               21                  0.41                           22%–75%
  N2                              5            3               19                  0.40                           14%–100%
  N3                              5            5               12                  0.00                           –                   –
Disease stage




                                                                                                                                                   Downloaded from http://annonc.oxfordjournals.org/ by guest on March 25, 2012
  1                              16            5                –                  0.74                           55%–100%
  2                              53           27               36                  0.47                           35%–63%
  3                              14           10               16.5                0.29                           13%–65%             0.04
ER status
  Negative                       71           38               30                  0.47                           36%–60%
  Positive                        6            3               75.5                0.50                           23%–100%            –
Surgery
  MRM                            59           24                –                  0.58                           46%–72%
  None                            1            1                7                  0.00                           –
  Seg                            27           17               30                  0.41                           26%–64%
  Simple                          2            2               16.5                0.00                           –                   –
Chemotherapy
  No                             12            7               36                  0.50                           28%–88%
  Yes                            74           37               59                  0.48                           38%–62%             0.90
Chemotherapy regimen
  AI                              3            0                –                  1.00                           –
  CMF                             8            7               18                  0.13                           2%–78%
  Anthracycline                  50           24                –                  0.50                           38%–67%
  Anthracycline/taxane           13            6                –                  0.54                           33%–89%
  None                           12            7               36                  0.50                           28%–88%             –
Radiotherapy
  No                             36           16                –                  0.55                           40%–75%
  Yes                            49           27               28                  0.45                           33%–61%             0.34
Tamoxifen
  No                             75           38               32                  0.50                           39%–63%
  Yes                             9            4                –                  0.53                           28%–100%            –

In some categories, data on a small number of patients was not available.
ER, estrogen receptor; MRM, modified radical mastectomy; Seg, segmental mastectomy; AI, doxorubicin, ifosfamide; CMF, cyclophosphamide,
methotrexate and 5-fluorouracil.



1.47 times the risk of recurrence (P = 0.04) and 2.24 times                 and 1.27 times the risk of disease recurrence (P = 0.55)
the risk of local recurrence (P = 0.003) compared with                      compared with patients with the ductal histology.
having MRM.
                                                                            SEER database
matching to ductal patients                                                 Between 1988 and 2001, 213 and 98 patients with MSC and
After matching and adjustment for other known prognostic                    carcinosarcoma of the breast who had a median age of 65 and 61
factors including age at diagnosis, grade, N stage, T stage and             years, respectively, were registered in the SEER database. The
treatments received in the ductal adenocarcinoma patients,                  younger age of the MDACC patients may reflect referral bias.
patients with MSC had 0.80 times the risk of death (P = 0.72)               The total number of breast cancers (all stages) registered in the


610 | Hennessy et al.                                                                                             Volume 17 | No. 4 | April 2006
Annals of Oncology                                                                                                              original article
                                                                                                Table 5. The 5-year and median (in months) overall (OS) and

                           1.0
                                                                                                disease-specific survival (DSS) rates for breast carcinosarcoma patients
                           0.8                                                                  registered in the Surveillance, Epidemiology and End-Results (SEER)
                                                                                                database between 1988 and 2001 by stage at diagnosis
    Proportion Surviving
                           0.6




                                                                                                Stage Patient Median 5-year 5-year OS: Median 5-year 5-year DSS:
                                                                                                      number OS      OS     95% CI     DSS    DSS 95% CI
                           0.4




                                                                                                1     20        NR       0.73    0.43–0.89   NR       0.87    0.57–0.97
                                                                                                2     54         96      0.59    0.44–0.71   NR       0.72    0.56–0.83
                           0.2




                                                                                                3      7         25      0.44    0.07–0.78   NR       0.53    0.07–0.86
                                                                                                4      6          3      0.00    –           3        0.00    –
                           0.0




                                                                                                CI, confidence interval; NR, not reached.
                                 0     50          100          150        200            250
                                            Months from Date of Surgery

Figure 2. Recurrence-free survival (with 95% CI shown by the broken
lines) among patients without metastatic disease at diagnosis.




                                                                                                                                                                            Downloaded from http://annonc.oxfordjournals.org/ by guest on March 25, 2012
       1.0
       0.8
       0.6
       0.4




                                                         No Radiation, MRM (E/N = 4/32)
       0.2




                                                         No Radiation, BCS (E/N = 3/4)
                                                         Radiation, MRM (E/N = 5/24)
                                                         Radiation, BCS (E/N = 9/23)
       0.0




                           0          50           100           150             200

Figure 3. Proportional local recurrence-free survival in months by
Radiation and Surgery Treatment Group. MRM, modified radical
mastectomy; BCS, breast conserving surgery; E, events; N, number.
Fifty-six MRM patients had radiation therapy details available.




Table 4. The 5-year and median (in months) overall (OS) and
                                                                                                Figure 4. (A) Overall and (B) disease-specific survival curves (months) of
disease-specific survival (DSS) rates for metaplastic sarcomatoid breast
                                                                                                213 patients with metaplastic sarcomatoid carcinoma of the breast
cancer patients registered in the Surveillance, Epidemiology and
                                                                                                registered in the Surveillance, Epidemiology and End-Results (SEER)
End-Results (SEER) database between 1988 and 2001 by stage at diagnosis
                                                                                                database between 1988 and 2001.

Stage Patient Median 5-year 5-year OS: Median 5-year 5-year DSS:
      numbera OS     OS     95% CI     DSS    DSS 95% CI
                                                                                                SEER database during that time was 281 382. As in our series,
1                               45   NR     0.81    0.56–0.93    NR       0.93     0.74–0.98
                                                                                                most MSC and carcinosarcoma patients had tumors that were
2                              122    97    0.59    0.46–0.69    NR       0.67     0.54–0.77
                                                                                                stage II at diagnosis, poorly differentiated or anaplastic and
3                               27   NR     0.67    0.40–0.84    NR       0.71     0.43–0.87
                                                                                                hormone receptor-negative (Table 1). Fewer patients in the
4                               14     5    0.18    0.03–0.43    5        0.20     0.03–0.46
                                                                                                SEER database received adjuvant radiation therapy. Tables 4
CI, confidence interval; NR, not reached.                                                        and 5 show the 5-year and median OS and DSS rates for MSC
a
 Patient number does not match that in the different stages in Table 1 as                       and carcinosarcoma patients by stage at diagnosis. For 199 and
a small number of patients did not have survival information available.                         81 MSC and carcinosarcoma patients with localized disease


Volume 17 | No. 4 | April 2006                                                                                                     doi:10.1093/annonc/mdl006 | 611
original article                                                                                                                  Annals of Oncology


Table 6. Comparison of overall survival (OS) and recurrence-free (RFS) or disease-specific survival (DSS-SEER database) by stage at diagnosis of breast
metaplastic sarcomatoid carcinoma (MSC) patients from the MDACC and SEER series’ with SEER database statistics for breast carcinosarcoma (CS) and for
all SEER breast cancer patients diagnosed between 1988 and 2001


Stage      MDACC:            SEER:             SEER:           SEER: all            MDACC:         SEER:             SEER:           SEER: all
           MSC 5-year        MSC 5-year        CS 5-year       breast cancer:       5-year         MSC 5-year        CS 5-year       breast cancer:
           OS (%)            OS (%)            OS (%)          5-year OS (%)        DSS (%)        DSS (%)           DSS (%)         5-year DSS (%)
I          86                81                73              87                   86             93                87              96
II         62                59                59              75                   64             67                72              85
III        31                67                44              49                   38             71                53              58
IV          0                18                 0              16                    0             20                 0              22




(stages I–III) at diagnosis, the 5-year OS rates were 0.63                      node-positive disease [15]. Although our patient numbers are
(95% CI 0.53–0.72; Figure 4A) and 0.60 (95% CI 0.48–0.71),                      small and interpretation is limited by cross-trial comparisons, the
respectively. Their 5-year DSS rates were 0.72 (95% CI                          outcomes of similar patients with breast MSC in our series treated
0.62–0.80; Figure 4B) and 0.73 (95% CI 0.60–0.83), respectively.                with adjuvant chemotherapy (described in the Results section) are
                                                                                similar to these untreated rates. In addition, the activity of




                                                                                                                                                         Downloaded from http://annonc.oxfordjournals.org/ by guest on March 25, 2012
                                                                                primary chemotherapy is certainly less than that reported in
conclusions                                                                     conventional adenocarcinoma of the breast, especially when
Biphasic metaplastic sarcomatoid carcinomas (MSC) of the                        one considers that high-grade hormone receptor-negative
breast account for 0.08% of all breast malignancies. They                      adenocarcinomas are particularly sensitive to primary
are usually poorly differentiated and negative for hormone                      anthracycline/taxane-based chemotherapy [16]. Interestingly,
receptors and HER2/neu. We have confirmed previous                               although patient numbers are very small, all three patients
smaller reports suggesting that MSCs are aggressive. Table 6                    treated with adjuvant sarcoma-type chemotherapy are relapse-
compares the outcomes of breast MSC and carcinosarcoma                          free (Table 1), although one of these patients treated in the
patients in the MDACC and SEER series’ with the outcomes                        neoadjuvant setting had only stable disease as best response.
of all breast cancer patients registered in the SEER database                   Although limited numbers preclude definitive statistical
between 1988 and 2001. For the most common stage (II) at                        assessment, the outcome of cyclophosphamide, methotrexate and
diagnosis, outcomes are indeed worse for metaplastic cancer                     5-fluorouracil-treated breast MSC patients was very poor, while
patients, although we did not find a statistically significant                    those patients treated with adjuvant anthracyclines did seem to do
difference between the outcomes of MSC patients and those                       better, suggesting that CMF chemotherapy is less effective than
of a matched group of M. D. Anderson high-grade                                 anthracycline-based therapy in breast MSC (Tables 2 and 3).
receptor-negative ductal adenocarcinoma patients, consistent                       In summary, although aggressive, breast MSCs are clinically
with previous reports. Notably, there were few, if any, failures                similar to carcinosarcomas and probably no more aggressive
after the median follow-up of the MSC patients, reminiscent of                  than high-grade receptor-negative breast adenocarcinomas.
inflammatory and other high-grade high-risk tumors. However,                     However, the responsiveness of breast MSCs to neoadjuvant
as previously reported, there also seem to be some differences                  chemotherapy suggests that these patients may derive less
between MSC and conventional breast adenocarcinomas [1, 2].                     benefit from conventional breast cancer chemotherapy.
For example, the axillary lymph nodes are less frequently                       Recently, the use of gene profiling has identified most
involved by MSC of the breast and are not as strong a predictor                 receptor-negative breast cancers as a unique ‘basal’ form of
of prognosis as T stage.                                                        breast cancer thought to arise from progenitor cells in the ductal
   Although data interpretation is restricted by the retrospective              epithelium [17–20]. Since breast MSCs have many similarities
and uncontrolled nature of the results and by wide CIs, it seems                with this breast cancer subtype but also some ‘sarcoma-like’
the administration of conventional radiotherapy may not                         features (node negativity and the predominant importance of
adequately reduce the incidence of local recurrences for those                  T size to outcome), it is possible that breast MSCs are breast
with this rare cancer who undergo breast-conserving surgery.                    adenocarcinomas originating from a basal progenitor cell in the
Margin information was not included in univariate analysis of                   breast with a degree of maturational plasticity leading to the
local recurrence because of sparse data. Our data suggest that                  acquisition of both morphologic and clinical sarcomatoid
MRM should be considered for patients with breast MSC,                          characteristics. The application of gene profiling to MSCs of the
particularly for patients with T2 and higher stage disease.                     breast will help us to clarify this. The high risk of local
Certainly, the optimization of locoregional control is critical in              recurrence with breast MSC should make oncologists consider
these patients. New radiation approaches and possibly                           whether breast-conserving surgery is the right option for
concomitant radiosensitizing agents are needed.                                 patients with this disease, particularly with T2 and higher
   The RFS rate of patients with an average age of 48 years                     T stages. Novel agents are required to improve prognosis and
who have high-grade hormone receptor-negative breast cancer                     biological studies are urgently required to identify potential
given no adjuvant systemic treatment is 36.7% at 10 years for                   molecular targets. A sarcoma-like chemotherapy regimen may
T2N1 tumors, 47.3% for T3N0 tumors and 25.1% for T3/4                           be useful as a potential starting place.


612 | Hennessy et al.                                                                                                Volume 17 | No. 4 | April 2006
Annals of Oncology                                                                                                        original article
references                                                                              11. Teixeira MR, Qvist H, Bohler PJ, Pandis N, Heim S. Cytogenetic analysis shows
                                                                                            that carcinosarcomas of the breast are of monoclonal origin. Genes
 1. Gutman H, Pollock RE, Janjan NA, Johnston DA. Biologic distinctions and                 Chromosomes Cancer 1998; 22: 145–151.
    therapeutic implications of sarcomatoid metaplasia of epithelial carcinoma of       12. Wick MR, Swanson PE. Carcinosarcomas: current perspectives and an
    the breast. J Am Coll Surg 1995; 180: 193–199.                                          historical review of nosological concepts. Semin Diagn Pathol 1993; 10:
 2. Foschini MP, Dina RE, Eusebi V. Sarcomatoid neoplasms of the breast:                    118–127.
    proposed definitions for biphasic and monophasic sarcomatoid mammary                 13. Husband JE, Schwartz LH, Spencer J et al. Evaluation of the response to
    carcinomas. Semin Diagn Pathol 1993; 10: 128–136.                                       treatment of solid tumours—a consensus statement of the International
 3. Tavassoli FA. Pathology of the Breast, 2nd edn. Hong Kong: Appleton and                 Cancer Imaging Society. Br J Cancer 2004; 90: 2256–2260.
    Lange 1999.                                                                         14. Bergstralh EJ, Kosanke JL, Jacobsen SJ. Software for optimal matching in
 4. Wargotz ES, Deos PH, Norris HJ. Metaplastic carcinomas of the breast. II.               observational studies. Epidemiology 1996; 7: 331–332.
    Spindle cell carcinoma. Hum Pathol 1989; 20: 732–740.                               15. Ravdin PM, Siminoff LA, Davis GA et al. Computer program to assist in
 5. Wargotz ES, Norris HJ. Metaplastic carcinomas of the breast. III. Carcinosarcoma.       making decisions about adjuvant therapy for women with early breast cancer.
    Cancer 1989; 64: 1490–1499.                                                             J Clin Oncol 2001; 19: 980–991.
 6. Sapino A, Papotti M, SanFilippo B, Gugliotti P, Bussolati G. Tumour types           16. Buzdar AU, Valero V, Theriault RL et al. Pathological complete response to
    derived from epithelial and myoepithelial cell lines of R3230AC rat mammary             chemotherapy is related to hormone receptor status. San Antonio Breast
    carcinoma. Cancer Res 1992; 52: 1553–1560.                                              Cancer Symposium 2003; Abstr 302.
 7. Wargotz ES, Norris HJ. Metaplastic carcinomas of the breast. 1. Matrix-producing    17. Sorlie T, Perou CM, Tibshirani R et al. Gene expression patterns of breast
    carcinoma. Hum Pathol 1989; 20: 628–635.                                                carcinomas distinguish tumor subclasses with clinical implications. Proc Natl
 8. Ferrara G. Sarcomatoid carcinoma of the breast: pathology of four cases.                Acad Sci USA 2001; 98: 10869–10874.
    Breast Dis 1995; 8: 283–294.                                                        18. Chang CC, Sun W, Cruz A et al. A human breast epithelial cell type with stem




                                                                                                                                                                             Downloaded from http://annonc.oxfordjournals.org/ by guest on March 25, 2012
 9. Balercia G, Bhan AK, Dickersin GR. Sarcomatoid carcinoma: an ultrastructural            cell characteristics as target cells for carcinogenesis. Radiat Res 2001; 155:
    study with light microscopic and immunohistochemical correlation of 10 cases            201–207.
    from various anatomic sites. Ultrastructural Pathol 1995; 19: 249–263.              19. Nielsen TO, Hsu FD, Jensen K et al. Immunohistochemical and clinical
10. Dandachi N, Hauser-Kronberger C, More E et al. Co-expression of tenascin-C and          characterization of the basal-like subtype of invasive breast carcinoma. Clin
    vimentin in human breast cancer cells indicates phenotypic transdifferentiation         Cancer Res 2004; 10: 5367–5374.
    during tumour progression: correlation with histopathological parameters, hormone   20. Korbling M, Estrov Z. Adult stem cells for tissue repair—a new therapeutic
    receptors, and oncoproteins. J Pathol 2001; 193: 181–189.                               concept? N Engl J Med 2003; 349: 570–582.




Volume 17 | No. 4 | April 2006                                                                                                doi:10.1093/annonc/mdl006 | 613

				
DOCUMENT INFO
Categories:
Tags:
Stats:
views:8
posted:3/28/2012
language:
pages:9