www.elsevier.com/locate/ygyno

Gynecologic Oncology

Abdominal carcinomatosis in women with a history of breast cancer

Ruchi Garga, Mariana L. Zahurakb, Edward L. Trimblea,

Deborah K. Armstrongc, Robert E. Bristowa,c,*

aDepartment of Gynecology and Obstetrics, The Sidney Kimmel Comprehensive Cancer Center,

The Johns Hopkins Medical Institutions, Baltimore, MD 21287-1281, USAbDepartment of Biostatistics, The Sidney Kimmel Comprehensive Cancer Center, The Johns Hopkins Medical Institutions, Baltimore, MD 21287-1281, USAcDepartment of Oncology, The Sidney Kimmel Comprehensive Cancer Center, The Johns Hopkins Medical Institutions, Baltimore, MD 21287-1281, USA

Received 23 December 2004

Available online 23 June 2005

Abstract

Objectives. The goals of this study were to: (1) characterize the etiology of abdominal carcinomatosis, (2) identify clinical features

predictive of primary ovarian/peritoneal cancer, and (3) evaluate the survival impact of cytoreductive surgery among patients with advanced

ovarian/peritoneal cancer and a history of breast cancer.

Methods. Patients with a history of prior breast cancer undergoing surgical exploration for abdominal carcinomatosis between 1/1/88 and

12/31/02 were retrospectively identified from tumor registry databases. Logistic regression analysis was used to explore clinical

characteristics predictive of primary ovarian/peritoneal cancer versus recurrent breast cancer. Survival analyses and comparisons were

performed using the Kaplan–Meier and Cox proportional hazard models.

Results. Seventy-nine patients underwent surgery for abdominal carcinomatosis a median of 5.39 years after initial breast cancer

diagnosis. Abdominal carcinomatosis was due to primary ovarian/primary peritoneal cancer in 74.7% of cases. A history of Stage I breast

cancer [OR = 10.73, 95%CI = 2.6–43.7, P < 0.001] and the lack of a prior breast cancer recurrence [OR = 10.60, 95%CI = 2.5–45.2, P <

0.001] were independently predictive of primary ovarian/peritoneal cancer. Among patients with primary ovarian/peritoneal cancer, optimal

(�1 cm) cytoreductive surgery was associated with a median survival of 44.0 months compared to 18.0 months for patients with suboptimal

residual disease [HR = 6.81, 95%CI = 3.37–13.77, P < 0.0001]. Recurrent breast cancer was associated with a median survival time of 6.4

months.

Conclusions. Among patients with prior breast cancer presenting with abdominal carcinomatosis, early-stage disease and the absence of a

prior recurrence were predictive of primary ovarian/peritoneal cancer. Optimal cytoreductive surgery was associated with a significant

survival advantage for patients with primary ovarian/peritoneal cancer.

D 2005 Elsevier Inc. All rights reserved.

Keywords: Abdominal carcinomatosis; Ascites; Breast cancer; Ovarian cancer; Peritoneal cancer

Introduction

Breast cancer is the most common malignancy among

women in the United States. The current 5-year relative

survival rates of 97% and 78% for localized and regional

0090-8258/$ - see front matter D 2005 Elsevier Inc. All rights reserved.

doi:10.1016/j.ygyno.2005.05.013

* Corresponding author. The Kelly Gynecologic Oncology Service,

Department of Gynecology and Obstetrics, The Sidney Kimmel Compre-

hensive Cancer Center, The Johns Hopkins Medical Institutions, 600 N.

Wolfe Street, Phipps #281, Baltimore, MD 21287-1281, USA. Fax: +1 410

614 8718.

E-mail address: rbristo@jhmi.edu (R.E. Bristow).

disease, respectively, indicate that a significant proportion of

breast cancer patients will experience prolonged survival

[1]. Although the risk of breast cancer recurrence diminishes

over time, late recurrences well into the second decade of

surveillance can occur. Among patients with metastatic

breast cancer, extra-hepatic and extra-skeletal abdominal

locales account for 10% of cases. Ascites and carcinoma-

tosis may be present in as many as 5.4% and 2.6% of cases,

respectively [2].

The female genital system is a common site of second

primary non-mammary malignancies. Specifically, a per-

99 (2005) 65 – 70

R. Garg et al. / Gynecologic Oncology 99 (2005) 65–7066

sonal history of breast cancer is associated with a two to four-

fold increase in the risk of ovarian cancer [3]. In breast

cancer patients, the clinical distinction between recurrent

metastatic breast cancer and a new, second primary ma-

lignancy of the ovary or peritoneum can be difficult. The

clinical relevance of this distinction is predicated on the

disparate management strategies for ovarian/peritoneal

cancer (cytoreductive surgery followed by platinum-based

combination chemotherapy) compared to metastatic breast

cancer (hormonal therapy versus chemotherapy). Conse-

quently, the goals of this study were to: (1) characterize the

etiology of abdominal carcinomatosis in women with a

history of breast cancer, (2) identify clinical features

predictive of primary ovarian/peritoneal cancer, and (3)

evaluate the survival impact of cytoreductive surgery among

patients with advanced ovarian/peritoneal cancer subsequent

to a diagnosis of breast cancer.

Table 1

Demographic characteristics of breast cancer diagnosis

Age at diagnosis

Median 49 years

Mean 49.9 years

Range 25 to 82 years

Postmenopausal 46.8% (37 patients)

AJCC staging

Stage I 62.0%

Stage II 21.5%

Stage III 13.9%

Stage IV 2.5%

Methods

Approval to conduct this study was obtained from the

Johns Hopkins Medical Institutions (JHMI) Clinical

Research Committee and Joint Committee on Clinical

Investigation. All patients with a personal history of primary

breast cancer who subsequently underwent clinical manage-

ment for a diagnosis of abdominal carcinomatosis at the

JHMI between 1/1/88 and 12/31/02 were identified through

the JHMI Tumor Registry, the Kelly Gynecologic Oncology

Service clinical database, and the Department of Pathology

database by a computerized query for diagnoses of recurrent

breast cancer and breast cancer with dual primary sites (i.e. a

second primary malignancy). Of those with dual primary

sites, only patients with a second primary malignancy

diagnosed subsequent to primary breast cancer were selected

for further study. Selection criteria for initial data review also

required adequate clinical documentation of the initial breast

cancer stage of disease and histologic findings; however,

institutional pathologic review of breast cancer specimens

was not required. Finally, an exploratory laparotomy or

diagnostic laparoscopy with definitive pathologic documen-

tation of abdominal carcinomatosis was also required for

study inclusion.

Individual subject data were collected retrospectively

from in-patient and ambulatory medical records. Demo-

graphic data included the date of breast cancer diagnosis,

patient age and menopausal status at breast cancer diag-

nosis, tumor histology, and a family history of breast cancer

in a first-degree relative. American Joint Committee on

Cancer (AJCC) breast cancer stage of disease was assigned

retrospectively using the 2002 modified staging system [4].

With regard to abdominal carcinomatosis, abstracted vari-

ables included the date of diagnosis, patient age and

menopausal status, the interval between breast cancer

diagnosis and abdominal carcinomatosis, clinical presenta-

tion (ascites and carcinomatosis only; ascites and carcino-

matosis plus a pelvic mass; or ascites and carcinomatosis

plus liver metastases, with or without a pelvic mass), and

serum CA125 level when available. Following surgical

exploration and cytoreduction, the amount of residual

disease was considered optimal when the maximal diameter

of the largest residual tumor mass was �1 cm. Patients with

residual disease >1 cm were classified as having undergone

a suboptimal cytoreductive effort. All cases had been

previously reviewed in the multidisciplinary gynecologic

oncology tumor conference, and the specific post-operative

therapy program for each patient was prescribed by the

treating gynecologic oncologist. Surveillance data included

the date of last follow-up and disease status or the date and

cause of death.

For statistical outcomes, logistic regression analysis was

used to identify clinical characteristics predictive of primary

ovarian/peritoneal cancer as the etiology of abdominal

carcinomatosis. For survival analysis, event time distribu-

tions were estimated using the method of Kaplan–Meier,

and survival rates were compared using the Cox propor-

tional hazards model or the log rank test. All computations

were performed using the Statistical Analysis System or

EGRET. A P value of <0.05 was considered statistically

significant, and all P values reported are two-sided [5–10].

Results

Patient characteristics

A total of 1501 patients with a history of recurrent breast

cancer (n = 1274) or breast cancer with another primary site

of malignancy (n = 227) were identified through the JHMI

databases during the study interval. Of these, 79 patients

were documented as presenting with abdominal carcinoma-

tosis �30 days subsequent to the diagnosis of primary breast

cancer. The demographic characteristics of patients with

breast cancer diagnosis are shown in Table 1.

The median patient age at breast cancer diagnosis was

49 years (mean 49.9 years, range 25 to 82 years), with

37 patients (46.8%) being post-menopausal. According to

AJCC criteria, the stage distribution of breast cancer was

as follows: Stage I—62.0%, Stage II—21.5%, Stage

Table 2

Clinical characteristics associated with abdominal carcinomatosis due to

primary ovarian/peritoneal cancer versus recurrent breast cancer

Variable Odds

ratio

95%CIa Significance

( P)

Univariate logistic regression

Age at breast cancer 0.99 0.94–1.03 0.480

Family history breast cancerb 1.09 0.37–3.21 0.880

AJCC Stage I breast cancer 7.70 2.39–24.81 0.001

Postmenopausal breast cancer 0.74 0.26–2.07 0.560

Postmenopausal carcinomatosis 1.79 0.52–6.09 0.350

No prior breast cancer recurrence 7.22 2.25–23.23 0.001

Breast cancer pathologyc 1.19 0.33–4.28 0.790

Age at abdominal carcinomatosis 1.01 0.97–1.05 0.690

Breast cancer/carcinomatosis

interval

1.00 1.00–1.00 0.150

Abdominal presentationd 0.69 0.23–2.07 0.510

CA125e 1.37 0.89–2.08 0.150

Multivariate logistic regression

AJCC Stage I breast cancer 10.73 2.60–43.70 0.001

No prior breast cancer recurrence 10.60 2.50–45.20 0.001

a CI: Confidence Interval.b In a first-degree relative.c Infiltrating ductal carcinoma versus all others.d Isolated carcinomatosis and ascites versus pelvic mass or liver

metastases.e Log CA125.

R. Garg et al. / Gynecologic Oncology 99 (2005) 65–70 67

III—13.9%, and Stage IV—2.5%. Breast cancer pathology

was: infiltrating ductal carcinoma in 64 patients (81.0%),

lobular carcinoma in 12 patients (15.2%), serous adenocar-

cinoma in 1 patient (1.3%), medullary carcinoma in 1 patient

(1.3%), and papillary adenocarcinoma in 1 patient (1.3%).

Eighteen patients (22.8%) had experienced a non-abdominal

recurrence of breast cancer prior to diagnosis of abdominal

carcinomatosis.

The median age at diagnosis of abdominal carcinoma-

tosis was 58 years (mean 58.3 years, range 33 to 82 years),

with a median interval of 5.39 years (mean 8.47 years, range

0.10 to 33.55 years) between breast cancer diagnosis and

presentation with abdominal carcinomatosis. All patients

presented with ascites and abdominal carcinomatosis that

was confirmed radiographically (computed tomography,

magnetic resonance imaging, or positron emission tomo-

graphy). Thirty patients (37.9%) presented with abdominal

carcinomatosis and ascites as isolated findings. Forty-five

patients (57.0%) had concomitant pelvic mass in addition to

ascites and carcinomatosis, while 4 patients (5.1%) also had

liver metastases with or without a pelvic mass. The majority

of patients (81.0%) were post-menopausal at the time of

abdominal carcinomatosis presentation. Serum CA125

levels were available in 53 of 79 cases and yielded a

median value of 572 U/ml (mean 1445 U/ml, range 20 to

14,000 U/ml).

Overall, 54 patients (68.4%) were diagnosed with

primary ovarian cancer and 5 patients (6.3%) were

diagnosed with primary peritoneal cancer. In total, therefore,

primary ovarian/peritoneal cancer accounted for 74.7% of

all cases of abdominal carcinomatosis in women with a prior

history of breast cancer. Of these 59 patients, 7 (11.9%) had

FIGO Stage IIIB disease, 41 (69.5%) had FIGO Stage IIIC

disease, and 11 (18.6%) had FIGO Stage IV disease.

Histopathologically, 98.3% of these cases were epithelial

tumors (serous carcinoma, n = 47; mixed serous and

endometrioid carcinoma, n = 4; micropapillary serous

carcinoma with invasive implants, n = 3; clear cell

carcinoma, n = 3; endometrioid carcinoma, n = 1; serous

borderline tumor with non-invasive implants, n = 1), and

there was 1 case of a granulosa cell tumor. Eight out of 59

patients (13.6%) had a family history of ovarian cancer

whereas 29 patients (49%) had a family history of breast

cancer.

Recurrent breast cancer accounted for 19 cases (24.0%)

of abdominal carcinomatosis. Majority (10 patients i.e.

52.6%) had a family history of breast cancer with none

who reported a known family history of ovarian cancer

among this group of patients. Of the 18 patients with a

prior breast cancer recurrence, abdominal carcinomatosis

was due to a second recurrence of breast cancer in 9

patients (50%). Conversely, of the 19 patients with

recurrent breast cancer as the cause of abdominal

carcinomatosis, 9 patients (47.4%) had experienced a prior

recurrence of breast cancer (in other words, abdominal

carcinomatosis was their second breast cancer recurrence).

One patient (1.3%) had primary FIGO Stage IVB uterine

papillary serous carcinoma.

Predictors of primary ovarian/peritoneal cancer

Multivariate logistic regression analysis of demographic

and clinical variables revealed that only AJCC Stage I breast

cancer (odds ratio [OR] = 10.73, 95% confidence interval

[95%CI] = 2.6–43.7, P = 0.001) and the absence of a prior

breast cancer recurrence (OR=10.60, 95%CI = 2.5–45.2,P =

0.001) were independently and statistically significantly

associated with a diagnosis of primary ovarian/peritoneal

cancer versus recurrent breast cancer (Table 2).

Although not statistically significant, the interval

between breast cancer diagnosis and abdominal carcinoma-

tosis for patients with primary ovarian/peritoneal cancer

was notably longer (median 7.06 years, mean 9.37 years,

range 0.10 to 33.55 years) compared to patients with

recurrent breast cancer (median 3.00 years, mean 6.02

years, range 0.27 to 28.68 years). Serum CA125 levels

were available for 36/59 patients with primary ovarian/

peritoneal cancer and 16/19 patients with recurrent breast

cancer. There was a trend, although again not statistically

significant, toward higher serum CA125 levels in patients

with primary ovarian/peritoneal cancer (median 683 U/ml,

mean 1809.6 U/ml, range 20 to 14000 U/ml) compared to

patients with recurrent breast cancer (median 350 U/ml,

mean 709.2 U/ml, range 24 to 2617 U/ml).

R. Garg et al. / Gynecologic Oncology 99 (2005) 65–7068

Surgical outcome and survival analysis

All patients underwent a minimum of exploratory

laparotomy or diagnostic laparoscopy with tissue biopsy,

which was the only procedure performed in 9 cases.

Additional procedures were performed with the following

frequencies: unilateral or bilateral salpingo-oophorectomy

(n = 68), omentectomy or omental biopsy (n = 59), total or

supracervical hysterectomy (n = 54), tumor cytoreduction

(n = 43), retroperitoneal lymph node sampling (n = 39),

small bowel resection (n = 11), large bowl resection (n = 5).

For the 59 patients with primary ovarian/peritoneal

cancer, optimal cytoreduction was achieved in 33 cases

(55.9%), while suboptimal residual disease was left in 26

cases (44.1%). Fifty-five patients (93.2%) received plati-

num-based chemotherapy post-operatively, either alone or

in combination with a variety of other agents (cyclo-

phosphamide, adriamycin, paclitaxel, topotecan). The

median follow-up for patients with primary ovarian/

peritoneal cancer was 51.6 months and the median survival

time was 30.0 months. At last follow-up, 49 of 59 patients

(83.1%) were dead of disease. The median survival time for

patients undergoing optimal cytoreduction was 44.0 months,

compared to 18.0 months for patients left with suboptimal

residual disease (hazard ratio [HR] = 6.81, 95%CI = 3.37–

13.77, P < 0.001) (Fig. 1). Among patients with primary

ovarian/peritoneal cancer, 19 (32.2%) had no gross residual

disease, 14 (23.7%) had visible disease �1 cm, and 26

(44.1%) had suboptimal residual disease. The median

survival times according to residual disease were: micro-

scopic (no gross) residual 82.0 months, optimal (0.1 to 1.0

cm) but visible 33.0 months (HR 10.07, 95%CI 3.24–31.26,

P = 0.0001), and for suboptimal (>1.0 cm) residual disease

Fig. 1. Primary ovarian/peritoneal cancer survival outcome by post-

operative residual disease for 59 patients.

18.0 months (HR 26.37, 95%CI 8.30–83.74, P < 0.0001).

All 10 long-term (�5 years) surviving patients had no

visible residual disease.

Of the 19 patients with recurrent breast cancer, 5

underwent optimal cytoreductive surgery and experienced

a median survival time of 34.4 months compared to 3.9

months for the 14 patients left with suboptimal residual

disease (P = 0.001, log rank test). The median follow-up

time for these patients was 16.0 months.

Discussion

Breast cancer and ovarian cancer rank as the second and

fifth leading causes of cancer-related death, respectively,

among United States women [1]. One in eight women will

have breast cancer during their lifetimes and one in 70 will be

diagnosed with ovarian cancer [3]. Currently, about half of

all patients with a diagnosis of breast cancer will experience

a recurrence and one third will die of their disease [11].

Several factors may predispose a patient to develop both

breast cancer and ovarian cancer: (1) breast–ovarian cancers

(BOC) family syndrome and/or BRCA mutations, (2)

relatively high incidence of both breast cancer and ovarian

cancer, (3) similar risk factors for both cancers including age,

family history, reproductive and menstrual history along

with estrogen exposure [1,3]. We attempted to determine the

clinical, demographic, and etiologic features of women who

present with abdominal carcinomatosis with a history of

breast cancer. The clinical relevance being that the treatments

of these two entities are dramatically different. Curtin et al.

[12] reported on 121 women with a history of breast cancer

who underwent surgery for an adnexal or pelvic mass. A

benign process was discovered in 50% of cases. The

remaining patients were found to have a malignant process

as the etiology of the adnexal mass, with new ovarian or

fallopian tube primary malignancies accounting for 73%

cases; metastatic breast cancer was the cause of the adnexal

mass in 27% of patients. In other words, a new gynecologic

primary malignancy predominated over metastatic breast

cancer in a ratio of 3:1 in the setting of a pelvic mass

following a diagnosis of primary breast cancer. This

distinction is clinically important because the likelihood of

a non-ovarian metastatic tumor being the source of a

malignant ovarian neoplasm ranges between 3 and 37%.

Breast cancer is one of the common non-genital tumors to

metastasize to the ovary. Women with metastatic breast

cancer to the ovary have been reported to have a poor

prognosis, with a 5-year survival rate of 0–10% [13–16].

The reported survival time subsequent to the manifestation

of metastatic disease is approximately 3 years [11].

In 1997, Abu-Rustum et al. reported on 40 patients with

breast cancer and metachronous abdominal and/or pelvic

metastases. The median interval between breast cancer

diagnosis and surgical exploration was 80 months. The

median survival for all patients was 24.1 months. Patients

R. Garg et al. / Gynecologic Oncology 99 (2005) 65–70 69

left with no gross residual disease survived longer than

those with gross residual disease, although the difference

was not statistically significant. Eighty-five percent of

patients had infiltrating ductal carcinoma, 15% had invasive

lobular carcinoma. In addition, 45% of patients had

previously documented metastatic breast cancer to other

sites, a finding confirmed by the 47.4% rate observed in the

current study [17].

Eitan et al., in 2003, reported on 59 women with

metastatic breast cancer to the abdomen and pelvis, re-

examining the role of surgical resection as an update to the

report of Abu-Rustum et al. Exploratory surgery was

performed a median of 5 years after initial diagnosis of

breast cancer, with median survival time of 23 months from

the diagnosis of abdominal disease. Suboptimal residual

disease was significantly associated with an increased risk

of death due to disease [18]. Primary surgery with optimal

cytoreduction for advanced-stage ovarian cancer has con-

sistently been associated with a clinically and statistically

significant survival advantage compared to patients left with

large-volume residual disease [19]. Our data confirm the

positive impact on survival of optimal cytroreductive

surgery for patients with primary ovarian/peritoneal cancer

and further suggest that survival time is inversely propor-

tional to the volume of residual tumor.

Among patients with prior breast cancer presenting with

abdominal carcinomatosis, early-stage disease and the

absence of a prior recurrence were the only clinico-

pathological factors significantly associated with a diagnosis

of primary ovarian/peritoneal cancer, which accounted for

the majority of cases. Although not statistically significant,

there was a trend noted for a shorter interval until the

diagnosis of abdominal disease to be more likely recurrent

breast cancer, and higher CA125 levels to be more

predictive of primary ovarian/peritoneal disease. Interest-

ingly, optimal cytoreduction was associated with a signifi-

cant extension in median survival time (34.4 months)

compared to patients left with bulky residual disease (3.9

months, P = 0.001); however, the number of patients is too

small to reach definitive conclusions about the role of

cytoreductive surgery in this setting.

There are several limitations of the current study that must

be considered in interpreting the results. As with all

retrospective studies, the potential for selection bias must

be acknowledged. In an attempt to minimize such bias, study

selection criteria were strictly maintained. Furthermore, the

lack of all potentially clinically relevant factors in each

patient medical record is another limitation of a retrospective

study. Specifically, the inadequate data with regard to the

genetic history/family history of individual patients, BRCA

mutation data, tumor markers and specific laboratory values

if normal or elevated, ER/PR data. Secondly, the relatively

long time interval necessary to accrue a satisfactory number

of subjects may have introduced a component of potential

treatment bias. In other words, observed survival differences

may have been effected by the evolution of treatment

regimens over time. A third potentially confounding factor

may be the subjective nature of the designation of optimal

residual disease and the completeness of surgical resection,

as defined in medical records by individual surgeons.

Finally, given the nature of our practice as a tertiary referral

center, it was impossible to evaluate the true incidence of

abdominal carcinomatosis arising after a diagnosis of breast

cancer.

In conclusion, our data indicate that the majority of

women presenting with abdominal carcinomatosis after a

diagnosis of breast cancer will have a new primary ovarian

cancer or primary peritoneal cancer. We were unable to

identify reliable predictors of recurrent breast cancer as the

etiology of abdominal carcinomatosis that would preclude

exploratory surgery for diagnostic and therapeutic purposes.

Given the significant survival advantage associated with

optimal residual disease for patients with a primary ovarian/

peritoneal cancer, an attempt at maximal cytoreductive

surgery is warranted for most patients with this clinical

presentation.

References

[1] American Cancer Society. Cancer facts and figures: American Cancer

Society, IncR; 2003 [Available at http://www.cancer.org/downloads/

stt/caff2003pwsecured.pdf ].

[2] Caskey CI, Scatarige JC, Fishman EK. Distribution of metastases in

breast carcinoma: CT evaluation of abdomen. Clin Imaging 1991;

15(3):166–71.

[3] National Cancer Institute, National Institutes of Health. Genetics of

breast and ovarian cancer; 2004 [Available at http://www.cancer.gov/

cancerinfo/pdq/genetics/breast-and-ovarian].

[4] AJCC (American Joint Committee on Cancer). In: Greene FL, Page

DL, Fleming ID, et al, editors. Cancer staging manual. 6th ed. New

York’ Springer-Verlag; 2002. p. 223–40.[5] Melfi C, Holleman E, Arthur D, Katz B. Selecting a patient

characteristics index for prediction of medical outcomes using

administrative claims data. J Clin Epidemiol 1995;48:917–26.

[6] Kaplan EL, Meier P. Nonparametric estimation from incomplete

observations. J Am Stat Assoc 1958;53:457–80.

[7] Mantel N, Haenszel W. Statistical aspects of the analysis of data from

retrospective studies of disease. J Natl Cancer Inst 1959;22:719–748.

[8] Cox DR. The analysis of binary data. London’ Methuen; 1970.

[9] SAS Institute Inc. SAS user’s guide: statistics. Version 5 Edition. Cary,

NC: SAS Institute Inc; 1985.

[10] Statistics and Epidemiology Research Corporation. EGRET user’s

manual. Seattle, WA: Statistics and Epidemiology Research Corpo-

ration; 1988.

[11] Winer EP, Morrow M, Osbome CK, Harris JR. Malignant tumors of

the breast. In: DeVita VT, Hellman S, Rosenberg SA, editors. Cancer:

principles and practice of oncology. 6th ed. Baltimore’ Lippincott

Williams and Wilkins; 2001. p. 1651–706.

[12] Curtin JP, Barakat RR, Hoskins WJ. Ovarian disease in women with

breast cancer. Obstet Gynecol 1994;84:449–52.

[13] Ayhan A, Tuncer S, Bukulmez O. Malignant tumors metastatic to the

ovaries. J Surg Oncol 1995;60:268–76.

[14] Ulbright TM, Roth LM, Stehman FB. Secondary ovarian neoplasia.

Cancer 1984;53:1164–74.

[15] Webb MJ, Decker DG, Mussey E. Cancer metastatic to the ovary-

factors influencing survival. Obstet Gynecol 1975;45:391–6.

[16] Petru E, Pickel H, Heydarfadai M, Lahousen M, Hass J, Schaider H,

R. Garg et al. / Gynecologic Oncology 99 (2005) 65–7070

et al. Nongenital cancers metastatic to the ovary. Gynecol Oncol

1992;44:83–6.

[17] Abu-Rustum NR, Aghajanian CA, Venkatraman ES, Feroz F, Barakat

RR. Metastatic breast carcinoma to the abdomen and pelvis. Gynecol

Oncol 1997;66:41–4.

[18] Eitan R, Gemignani ML, Venkatraman ES, Barakat RR, Abu-Rustum

NR. Breast cancer metastatic to abdomen and pelvis: role of surgical

resection. Gynecol Oncol 2003;90:397–401.

[19] Munstedt K, Franke FE. Role of primary surgery in advanced ovarian

cancer. World J Surg Oncol 2004;2:32–40.