Influence of platinum-based chemotherapy on disseminated tumor cells in blood and bone marrow of patients with ovarian cancer

Available online at www.sciencedirect.com

Gynecologic Oncology 107 (2007) 331 – 338 www.elsevier.com/locate/ygyno

Influence of platinum-based chemotherapy on disseminated tumor cells in blood and bone marrow of patients with ovarian cancer Pauline Wimberger a,⁎, Martin Heubner a , Friedrich Otterbach b , Tanja Fehm c , Rainer Kimmig a , Sabine Kasimir-Bauer a a

Department of Gynecology and Obstetrics, University of Duisburg-Essen, Hufelandstrasse 55, D-45122 Essen, Germany b Institute of Pathology, University of Duisburg-Essen, Germany c Department of Gynecology and Obstetrics, University of Tuebingen, Germany Received 8 March 2007 Available online 31 August 2007

Abstract Objective. We evaluated (1) the prevalence of disseminated tumor cells (DTC) before and after first-line chemotherapy with carboplatin and paclitaxel in the bone marrow (BM) and peripheral blood (PB) of 57 patients with primary ovarian cancer and (2) the coexpression of the epithelial antigen EpCAM on DTC including the determination of apoptotic cells. Methods. DTC were detected by immunocytochemistry applying the anti-cytokeratin (CK) antibody A45-B/B3. For double-labeling of DTCs, the antibodies M30 (apoptosis), HEA-125-FITC/Ber-EP4-FITC (EpCAM) were used. Results. Before chemotherapy, we identified DTC in 12/57 PB samples (21%) with a median number of 2 cells/20 ml (range 1–8) and in 25/46 BM samples (54%) with a median number of 5 cells/9 × 10E6 BM cells (range 1–28). Analysis of DTC in PB and BM before and after therapy was performed in 30 patients. In this subgroup, we identified DTC in 5/30 PB samples (16%) and in 15/30 BM samples (50%) before chemotherapy. After chemotherapy, DTC in PB were only detected in one patient but in the BM of 15/30 patients (50%). After chemotherapy, BM analysis revealed evidence that no DTC were detectable any longer in 9 patients, no significant change in DTC was documented in 14 patients and a significant enhancement of DTC was shown in 10 patients, including 8 patients who had no DTC before chemotherapy. DTC, still present after chemotherapy, co-expressed EpCAM and were non-apoptotic. In a univariable analysis, patients with a marked increase of DTC showed a significantly reduced PFS (p = 0.041). A corresponding multivariable Cox regression analysis was not feasible due to the limited number of events. No correlation of DTC in BM and PB was found with patient's and tumor characteristics. Conclusion. DTC were present in 50% of patients after first-line chemotherapy in ovarian cancer. It has to be considered whether patients with persisting EpCAM/CK-positive BM cells probably might benefit from an additive immunotherapy e.g. targeting EpCAM. © 2007 Elsevier Inc. All rights reserved. Keywords: Disseminated tumor cells; EpCAM; Apoptosis; Platinum-based chemotherapy; Ovarian cancer

Introduction Ovarian cancer is the fifth leading cause of cancer related death in Germany and the incidence of malignant tumors of the ovary was 23 per 100,000 women in the year of 2000 [1]. Initial debulking surgery followed by platinum-taxan chemotherapy is standard treatment in advanced disease in ovarian cancer [2,3]. Abbreviations: BM, bone marrow; CK, cytokeratin; DTC, disseminated tumor cells; PB, peripheral blood; PSF, progression free survival. ⁎ Corresponding author. Fax: +49 201 723 3938. E-mail address: [email protected] (P. Wimberger). 0090-8258/$ - see front matter © 2007 Elsevier Inc. All rights reserved. doi:10.1016/j.ygyno.2007.07.073

However, even despite of high response rates of chemotherapy with carboplatin–paclitaxel, ovarian cancer could not be cured in the majority of patients. Depending on prognostic factors, more than half of all patients will experience recurrence. Therefore, it is necessary to improve standard treatment including both surgical management and chemotherapy. Surgical outcome has been described to be one of the most important prognostic factors [4,5]. Recommended standard surgery in primary ovarian cancer includes radical tumor debulking with vertical laparotomy, bilateral salpingo-oophorectomy, total hysterectomy, omentectomy, peritoneal stripping and, if necessary, tumor debulking including bowel surgery and other upper abdominal

332

P. Wimberger et al. / Gynecologic Oncology 107 (2007) 331–338

resections. Pelvic and paraaortic lymphadenectomy is recommended if complete tumor resection could be achieved [3]. Ovarian cancer seems to be restricted for a long period to the peritoneal cavity, and distant metastases occur rarely. However, clinicopathologic studies have shown that occult hematogenous metastases are present in the lung, liver and the skeleton [6,7]. With the increasing number of therapeutic and experimental options, it has become more important to identify prognostic markers at the time of diagnosis and during clinical follow-up to determine who probably may benefit from an individualized targeted therapy or who will have a higher risk of recurrence. Multiple clinical studies have shown that immunoassays and molecular methods can identify the presence of disseminated tumor cells (DTC) in lymph nodes, peripheral blood (PB) and bone marrow (BM) in various epithelial tumor entities including breast, colon, rectum, stomach, esophagus, prostate, renal, bladder and non-small cell lung cancer [8]. Especially in breast cancer, a pooled analysis of BM findings in more than 4700 patients documented the prognostic value of these cells with regard to reduced disease free and overall survival [9]. Furthermore, recent studies showed the persistence of DTC in the BM of breast cancer patients to be an independent prognostic marker with regard to breast cancer specific survival in a multivariate analysis [10–12]. For ovarian cancer, only a few data on DTC in PB and BM exist. The first studies, using immunocytochemistry, were published in 1990 and 1995 with relatively few patients with advanced tumors (FIGO III–IV) or recurrent disease. DTC were detected in BM samples of 23% and 47% of the patients, respectively [13,14]. Braun et al. showed for the first time that DTC are present in the BM of one-third of patients with epithelial ovarian cancer at the time of diagnosis throughout all tumor stages and demonstrated an association between the presence of these cells and the occurrence of clinically overt extraperitoneal distant metastasis and death of cancer-related causes [15]. In contrast, using a different experimental design, Marth et al. found that detection of DTC in the BM and/or PB was not associated with poor prognosis in ovarian carcinoma patients [16]. Fehm et al. found DTC in 36% of patients at primary diagnosis and there was no correlation between DTC and other established prognostic factors [17]. Despite of extensive surgery and improvements of chemotherapy regimens for patients with primary ovarian cancer, it is not known whether chemotherapy can eliminate DTC in PB/and or BM. Here we studied (1) the prevalence of DTC in BM and PB before and after first-line platinum-based chemotherapy and (2) the coexpression of the epithelial antigen EpCAM on DTC as an immunotherapeutic target including the determination of apoptotic cells. In addition, we analyzed the influence of DTC on PFS. Patients and methods Patients and study design The uni-center prospective study was performed after obtaining approval from the local ethics committee. After giving written informed consent, patients with primary ovarian cancer undergoing surgery at the Department of Gyne-

cology and Obstetrics, University Hospital of Essen, Germany, were subjected to PB and BM aspiration. Exclusion criteria from this analysis were previous history of cancer or secondary malignancy or borderline tumors. Two bilateral BM samples were obtained under general anesthesia from the upper iliac crests through a needle aspiration within primary surgery for ovarian cancer. PB specimens were obtained at the beginning of exploratory surgery. Control aspirations, performed under local anesthesia with Xylocain 1%, and control of PB specimen were performed 3 months after conclusion of adjuvant chemotherapy with at least 6 cycles of carboplatinum AUC 5 and paclitaxel 175 mg/m2 in 30 patients. Tumor and patient characteristics of malignant ovarian tumor patients included in this study are shown in Table 1. 57 patients with histologically confirmed International Federation of Gynecology and Obstetrics (FIGO) stages IB-IV ovarian cancer, age of 18 years or older were included in the study [18]. Total abdominal hysterectomy, bilateral salpingo-oophorectomy, infragastric omentectomy, peritoneal stripping and pelvic and para-aortic lymphadenectomy were performed where feasible. The most important aim of surgery was to achieve macroscopic complete tumor resection. Radical pelvic and para-aortic lymphadenectomy were only performed if complete tumor resection was achieved following actual guidelines [3].

Preparation of BM BM cells were isolated from heparinized BM (5000 U/ml BM) by FicollHypaque density gradient centrifugation (density 1.077 g/mol; Pharmacia, Freiburg, Germany) at 400×g for 30 min from 46 patients before and from 30 patients after chemotherapy. Interface cells were washed (400 ×g for 15 min) and resuspended in phosphate buffered saline (PBS). 4.5 × 106 cells (1.5 × 106 per slide per area of 240 mm2) from each aspiration side were directly spun onto glass slides (400 ×g for 5 min) coated with poly-L-lysine (Sigma, Deisenhofen, Germany) using a Hettich cytocentrifuge (Tuttlingen, Germany) for the detection of CK+ cells. For the detection of CK+/EpCAM+ cells and CK+/ M30+ cells, 5 × 105 cells per slide per area of 240 mm2 were prepared.

Preparation of PB 20–25 ml EDTA-blood, were drawn by vein puncture from 57 patients before therapy and from 45 patients after therapy using Oncoquick® (Greiner Bio-One GmbH, Frickenhausen, Germany) which consists of a sterile 50 ml polypropylene tube with a porous barrier inserted on top of the separation medium, optimized for the enrichment of circulating tumor cells from PB. PB was cooled at 4 °C for 1 h and poured into the centrifuge tube and centrifuged at 1600×g at 4 °C for 20 min. Following centrifugation, the complete supernatant above the porous barrier was transferred into a new tube and cells were washed twice with 50 ml of washing buffer using a centrifugation step at 200×g at 4 °C for 10 min. Subsequently, the cells were resuspended in 2 ml washing buffer and spun onto two glass slides on an area of 240 mm2, coated with poly-L-lysine (Sigma, Deisenhofen, Germany) using a Hettich cytocentrifuge (Tuttlingen, Germany). This method has carefully been evaluated in blood samples of breast cancer patients and has been shown to give reliable results for the detection of circulating tumor cells in PB [19].

Immunocytochemistry After overnight air drying, staining for CK+ cells was performed using the Epimet® kit (Micromet, Martinsried, Germany). The identification of epithelial cells by using this kit is based on the reactivity of the murine monoclonal antibody Mab A45-B/B3, directed against a common epitope of CK polypeptides. The kit uses Fab fragments of the pan-Mab complexed with alkaline phosphatase molecules. Briefly, the method includes a) permeabilization of the cells with a detergent (5 min), b) fixation with a formaldehyde based solution (10 min), c) binding of the conjugate Mab A45-B/B3-alkaline phosphatase to cytoskeletal CKs (45 min) and d) formation of an insoluble red reaction product at the site of binding of the specific conjugate (15 min). Subsequently, the cells were counterstained with Mayer's haematoxylin for 1 min and finally mounted with Kaiser's glyzerine/gelatine (Merck, Darmstadt, Germany) in Tris EDTA buffer (Sigma, Deisenhofen, Germany). A negative control antibody (conjugate of Fab-fragment; Micromet, Munich, Germany) served as negative control. For each test a positive control slide with the

P. Wimberger et al. / Gynecologic Oncology 107 (2007) 331–338

333

Table 1 Patient characteristics according to presence of CK-positive cells in bone marrow and peripheral blood Variable

Blood before

Blood after

BM before

BM after

No. of patients Median no. of cells (range) Age mean Range FIGO stages I II III IV Ascites present Histology Serous papillary Mucinous Endometrioid Clear cell Other epithelian cl. Malignant germ cell Grade 1 2 3 ECOG status 0 Residual disease Macroscopic negative N0 b 1 cm ≥1 cm b 2 cm ≥2 cm Lymph node metastasis pN0 pN1 pNX CA-125 serum level +Elevated

57 (12) 2 (1–8) 56 (55) 18–80 (40–70)

45 (3) 9 (1–18) 54 (65) 18–77 (56–71)

46 (25) 5 (1–28) 55 (57) 18–80 (23–80)

30 (15) 10 (1–100) 50 (48) 18–71 (23–71)

13 (5) 6 (2) 30 (3) 8 (2) 27 (5)

11 (2) 3 (0) 24 (0) 5 (1) 19 (1)

9 (6) 5 (5) 25 (9) 7 (5) 21 (11)

7 (2) 2 (1) 16 (10) 3 (2) 10 (5)

38 (6) 5 (2) 1 (0) 5 (2) 4 (2) 4 (0)

30 (2) 3 (1) 1(0) 3 (0) 3 (0) 4 (0)

29 (14) 4 (4) 1 (1) 4 (4) 3 (2) 4 (0)

18 (11) 2 (0) 1 (0) 2 (2) 1 (0) 4 (2)

6 (3) 28 (3) 23 (6) 37 (10)

5 (1) 21 (0) 17 (2) 34 (2)

5 (1) 22 (11) 19 (13) 28 (13)

4 (1) 13 (8) 11 (6) 22 (12)

31 (8) 13 (1) 5 (2) 8 (1)

26 (2) 9 (0) 3 (0) 5 (1)

23 (15) 12 (3) 4 (1) 7 (6)

18 (9) 4 (3) 2 (1) 4 (2)

25 (6) 12 (4) 20 (2)

22 (2) 9 (0) 12 (1)

18 (9) 11 (8) 17 (8)

14 (6) 7 (4) 7 (5)

45 (9)

32 (3)

36 (20)

19 (11)

No correlation of DTC in BM and PB with histological parameters at time of primary surgery was found. In parenthesis number of patients with CK-positive cells were documented. FIGO: International Federation of Gynecology and Obstetrics. Other epithelial cl. = other malignant epithelial histological classifications (1 not classified ovarian cancer, 1 Mullerian Mixed Tumor of the ovary, 2 adenosquamous tumor). Blood before = peripheral blood specimen during primary surgery. Blood after = peripheral blood specimen after adjuvant treatment with chemotherapy. BM before = bone marrow specimen performed during primary surgery. BM after = bone marrow specimen taken after adjuvant treatment with chemotherapy. carcinoma cell lines MCF-7 (breast) and SKOV-3 (ovary) (ATTC, Rockville, MD) were treated under the same conditions.

Evaluation of CK+ cells Microscopic evaluation was carried out using the Ariol SL-50 (Applied Imaging), an automated scanning microscope and image analysis system with a slide loader, camera, computer and software for the detection and classification of cells of interest based on particular color, intensity, size, pattern, and shape. To use the system, immunostained slides are loaded onto the SL-50 slide loader. 6 slides per 1.5 million cells were evaluated per individual. The system loads each slide in turn onto the automated microscope stage and scans each frame of the cytospin through red, green and blue filters to recognize candidate objects. These objects are automatically scrutinized by color substraction, then analyzed and classified by 23 different morphometric parameters prior to final color ratio analysis. Quantitative data and high-quality images of analyzed objects are presented for review and classification. The Ariol SL-50™ can be used for high throughput of slides and is accurate and reliable in the detection of any anti-CK stained cells. We recently demonstrated a further improvement up to 17% in accuracy over manual microscopy using the automated system (data not shown).

Slides were fixed with solution B, containing formaldehyde, of the Epimet Kit (Micromet, Martinsried, Germany) for 10 min and unspecific binding was blocked using blocking solution D of the Epimet kit for 20 min. Subsequently, slides were incubated with the anti-CK monoclonal antibody A45-B/B3 for 45 min, directly labelled with the fluorochrome Cy3 at a concentration of 4 μg/ml (Micromet), and counterstained with two monoclonal antibodies directed against the EpCAM antigen for another 30 min [CD326, Clone HEA-125, Miltenyi, Bergisch Gladbach, Germany; Ber-EP4, Dako, Hamburg, Germany] directly labelled with FITC at a concentration of 2 μg/ml (CD326) and a dilution of 1:20 (Ber-EP4) (Fig. 1). Finally, the slides were washed with PBS and mounted as described above.

Evaluation of CK+/EpCAM+ and CK+/M30+ cells For the detection of apoptotic DTCs the antibody M30 (Roche Applied Science, Mannheim, Germany) was used according to the manufacturers' instructions. The M30 antibody reacts with a neo-epitope expressed only after caspase cleavage of CK 18 during early apoptosis [20,21]. Identification of apoptotic DTCs were based on positive M30 staining and cytomorphological criteria as described elsewhere [22,23].

Immunocytochemical double labelling

Evaluation of data

Experiments were performed on additional slides of the same BM aspirate. Immunocytochemical staining for the identification of CK+/EpCAM+ cells:

Patients were classified as tumor cell positive if at least one CK+ cell was detected as analyzed by immunocytochemistry. A marked increase or decrease

334

P. Wimberger et al. / Gynecologic Oncology 107 (2007) 331–338

Fig. 1. Characterization of CK-positive cells persisting after chemotherapy. In D an example for the detection of apoptotic cells in breast cancer is shown, whereas in ovarian cancer apoptotic DTC were not detectable.

of disseminated cells was assessed if the difference was 2 or more CK+ cells. Progression free survival was determined from the time of first diagnosis of primary ovarian cancer until the date of diagnosis of first relapse or to the date of last follow-up. Progression free survival was estimated using the Kaplan–Meier method and statistically evaluated using a two-sided log-rank test [24].

Results Prevalence of DTC in correlation with patient characteristics 57 patients with primary ovarian cancer were included in our prospective study. Due to clinical follow-up and performance status, BM and PB could not be obtained for all patients at all time points. The frequency of CK+ cells before and after chemotherapy for all samples analyzed are demonstrated in Table 1. Before chemotherapy, CK+ cells were found in 12/57 PB samples (21%) with a median number of 2 cells/20 ml (range 1–8) and after chemotherapy in 3/45 PB samples (7%) with a median number of 9 cells/20 ml (range 1–18). In BM, CK+ cells were detected in 25/46 samples (54%) with a median number of 5 cells/9 × 10E6 BM cells (range 1–28) before chemotherapy and in 15/30 samples (50%) with a median number of 10 cells/9 × 10E6 BM cells (range 1–100) after chemotherapy. The number of DTC were higher in those patients who still had DTC after chemotherapy, both in PB and BM samples. No correlation was found for a more advanced

Fig. 2. Monitoring of CK+ cells in blood before and after platinum-based chemotherapy. ■ Indicates the number of CK+ cells per 20 ml blood MNC. Abbreviations: CK: cytokeratin; MNC: mononuclear cells.

P. Wimberger et al. / Gynecologic Oncology 107 (2007) 331–338

335

Prevalence of DTC in PB and BM before and after chemotherapy

Fig. 3. Monitoring of CK+ cells in the bone marrow before and after platinumbased chemotherapy. ■ Indicates the number of CK+ cells per 9 × 106 MNC. Abbreviations: CK: cytokeratin; MNC: mononuclear cells.

surgical therapy in this subgroup and no correlation of DTC in BM and PB with histological parameters could be documented at the time of primary surgery. Mean age did not differ in the analyzed subgroups. The majority of analyzed patients had FIGO stage III, but not only patients with advanced tumor stage showed CK+ cells, also FIGO stage I and II patients had CK+ cells in PB and BM. We could not find a significant difference in the frequency of DTC dependent on the different histological subtypes, tumor grade, performance status, different residual tumor and lymph node involvement status (see Table 1).

The detailed analysis for the detection of CK+ cells in PB and BM before and after chemotherapy is shown in Figs. 2 and 3. A complete monitoring for CK+ cells in blood samples before and after chemotherapy can be shown for 45 patients (Fig. 2). In 35 patients (78%), no CK+ cells were found at any time point studied. After chemotherapy, no CK+ cells were detectable any longer in PB samples of 6 patients (13%), no change was found in one patient (2%) and an enhancement of CK+ cells could be documented for 2 patients (4%) including one patient who had no CK+ cells before chemotherapy. In BM, the analysis of CK+ cells before and after chemotherapy could be performed in 30 patients (Fig. 3). In 6 patients (20%), no CK+ cells were found before and after chemotherapy. After chemotherapy, no CK+ cells were detectable in 9 patients (30%) any longer, no significant changes in CK+ cell counts was documented in 14 patients (47%) and a significant enhancement of CK+ cells was shown in 10 patients (33%), including 8 patients (27%) who had no CK+ cells before chemotherapy. In 30 patients, evaluation of CK+ cells was performed in BM and PB before and after chemotherapy. In this subgroup, in only 10% of the patients CK-positivity was present in both, PB and BM before therapy, whereas in 43% neither in PB nor in BM CK+ cells were found before treatment. After chemotherapy, CK-positivity in PB and BM was detected in only one patient (3%) whereas no CK+ cells were evaluated in PB and BM in 50% of the patients. In contrast, 47% of the patients had persisting CK+ cells in their BM but not in their PB samples. For these 30 patients, no correlation of BM and PB findings was found before and after chemotherapy with both examinations. Characterization of DTC To type those CK+ BM cells still present after platinumbased chemotherapy, we used double-labelling for the coexpression of the epithelial antigen EpCAM as a target for

Fig. 4. Kaplan–Meier curve. Progression free survival in case of significant increase (group 1) or decrease or no significant change (group 0) of CK+ cells.

336

P. Wimberger et al. / Gynecologic Oncology 107 (2007) 331–338

immunotherapy and for M30 as a marker for apoptosis. Characterization of DTC was performed on additional slides in a subset of 8 patients where more than 5 CK+ cells were detected. Characterization of DTC was only possible if several CK+ cells were present. In all of these patients, EpCAM was coexpressed on nearly all detected DTCs. No apoptotic cells were found in any sample analyzed (Fig. 1). Correlation of the prevalence of DTC with clinical outcome The correlation between progression free survival (PFS) and the prevalence of DTC is shown for a mean follow-up time of 18.4 months (range 7–31 months) in Fig. 4. A marked increase in DTC was found for 10/30 patients (33%), in 14/30 patients (47%) no significant change of DTC was detected and in 6/30 patients (20%) a marked decrease of CK+-cells was present after chemotherapy. In case of a marked increase of DTC, patients showed a significant reduced PFS in comparison to patients with no change or marked decrease of DTC (p = 0.041) in an univariable analysis. Median PFS was not reached. In case of a marked increase of DTC, PFS was 12.4 months ± 0.7 (mean PFS ± S.E.M.), in patients with no change or decrease of DTC PFS was 19.3 months ± 1.4 (mean PFS ± S.E.M.). Although FIGO stage and residual tumor are significant prognostic factors for PFS, a corresponding multivariable Cox regression analysis was not feasible with our data due to the limited number of events (n = 9). Discussion This study demonstrates for the first time that DTC are present in BM after platinum-based chemotherapy in patients with primary ovarian cancer. A persistence of DTC was found in the BM in half of the patients after chemotherapy. DTC, still present after chemotherapy, coexpressed the epithelial cell adhesion molecule EpCAM on nearly all detected DTC and were non-apoptotic. Our findings confirm that early in the process of ovarian cancer, tumor cells have acquired the potential to disseminate to sites outside the peritoneum and our results support the notion that hematogenous dissemination in ovarian cancer is an intrinsic capacity rather than a random occurrence. Tumor cell persistence after chemotherapy has recently been shown to indicate chemotherapy resistance and poor clinical outcome in breast cancer [12,25,26]. Thus, the efficacy of a chemotherapy regimen may be indicated by a negative BM status or by the presence of apoptotic DTC that were susceptible to cytotoxic agents. Several clinical studies have shown that immunoassays based on anti-CK antibodies identify subgroups of patients with poor clinical prognosis with regard to early metastasis manifestation and reduced disease-free and overall survival in various epithelial tumor entities. Since ovarian tumor cells also express the most common CKs, we applied the monoclonal antibody A45-B/B3 which detects CK8, 18 and 19 which has been well evaluated in a variety of clinical studies on solid tumors [9,15,27–29] and in BM samples of a cohort of 165 noncarcinoma control patients [30]. Besides, it is very important to

mention that a BM analysis of 165 non-carcinoma control patients resulted in only 2 false positive results indicating that the A45B/B-3 gives reliable results for the detection of single disseminated tumor cells [30]. Moreover, to further characterize these cells and investigate the efficacy of chemotherapy, M30 staining as a marker for apoptosis on DTC was applied as recently shown to discriminate non-apoptotic from apoptotic DTC [23]. For ovarian cancer, only a few data on DTC in PB and BM exist. Braun et al. showed that DTC are present in the BM of one-third of patients with epithelial ovarian cancer at the time of diagnosis throughout all tumor stages and demonstrated an association between the presence of these cells and the occurrence of clinically overt extraperitoneal distant metastasis and death of cancer-related causes [15]. In contrast, using a different experimental design, Marth et al. found that the detection of DTC in the BM and/or PB was not associated with poor prognosis in ovarian carcinoma patients [16]. In contrast to Braun et al. we were unable to confirm a significant prognostic relevance of DTC in patients with primary ovarian cancer. Evaluation of PFS in a subset of our study population (n = 30) showed a significantly reduced PFS (p = 0.041) for patients with a marked increase of DTC in BM in comparison to patients with no increase of DTC (12.4 months vs. 19.3 months) in univariable analysis. Of course, progression free survival depends on many prognostic factors, such as residual tumor or FIGO stage. However, a corresponding multivariable Cox regression analysis was not feasible with our data due to the limited number (n = 9) of events. Since the limitation of this study is the small study population, a greater study population and a longer follow-up are necessary to confirm these preliminary data. In accordance with Marth et al. we found, that cancer-related deaths and progression was mostly due to intraperitoneal disease only (data not shown), whereas Braun et al. observed distant metastases in almost 90% of patients with positive BM. Furthermore, our results for DTC in PB were not in accordance with the detection of CK+ cells in the BM in 30 patients where PB and BM could be analyzed before and after chemotherapy. This might have different reasons. In contrast to BM results, the prognostic significance of DTC in blood is discussed quite controversially. It has been demonstrated for breast cancer that DTC in PB are often apoptotic bodies and do not have the ability to invade into the BM [31]. Other groups showed that the presence of these cells in PB of breast cancer patients correlated with the nodal status, the expression of the estrogen receptor and distant metastasis [32,33]. In metastatic breast cancer the number of circulating tumor cells was an independent prognostic factor with regard to progression free and overall survival [34]. For ovarian cancer, only rare data exist about the prognostic significance of DTC in PB. Judson et al. studied 91 PB samples of ovarian cancer patients using a tumor-enriched immunocytochemical assay and showed that survival curves did not differ between patients with and without circulating tumor cells after a mean follow-up time of 18.7 months [35]. Marth et al. could not show any difference for overall survival comparing patients with and without

P. Wimberger et al. / Gynecologic Oncology 107 (2007) 331–338

circulating tumor cells, too [16]. We cannot exclude that DTC, found in PB samples of our patients, were also apoptotic since after chemotherapy no DTC were any longer detectable in blood in most cases. Double-labeling for the determination of apoptotic DTCs in PB was not feasible in our setting because the Oncoquick®system is an enhanced density gradient system for PB where, in contrast to Ficoll gradients, less slides are prepared and have to be screened for positive events as the number of co-enriched mononuclear cells is lower [19]. Consequently, no additional slides were available for doublelabeling of CK+ cells. Furthermore, staining with new fuchsin red is too bright and strong to allow a second staining on the same slide. Since no standard method has been recommended up to now for the isolation of DTC in PB, further studies have to be awaited to estimate the prognostic value of DTC in PB. At the moment, analyzing the BM status including the determination of apoptotic or non-apoptotic cells, seems to be superior to blood analysis. Similar results were recently shown for studies in breast cancer patients, comparing the prognostic value of DTC in PB and BM [36,37]. However, our findings suggest that patients with effective locoregional treatment but persistence of occult metastatic tumor cells in the BM or PB probably might benefit from additional targeted therapies. DTC are targets for novel tumor biological therapy approaches such as specific antibody-based therapies against target cell-surface antigens. DTC have been phenotyped for the expression of tumor-associated antigens in solid tumors showing that the majority of these cells are “dormant” cells, expressing ki-67 as a marker for proliferation in low amounts. These findings may explain why DTC are resistant to cell-cycle dependent therapy, such as chemotherapy [38]. In gastric cancer, the expression of the urokinase-type plasminogen activator receptor (uPAR) correlates with an unfavourable prognosis [39] and similar observations were made in patients with breast cancer where HER-2 overexpression on DTC in the BM predicts poor clinical outcome [40]. However, patient selection for such tumor biological therapies becomes rather difficult due to phenotype changes, which may manifest themselves as differences between primary lesion and DTC. In this context, some evidence suggests that expression patterns of individual genes in DTC and circulating tumor cells differ from those of cells of the primary tumor e.g. that a larger proportion of cells is HER-2 positive [40]. Moreover, Meng et al. demonstrated that metastatic breast cancer patients with HER-2 positive circulating tumor cells but HER-2 negative primary tumor will benefit from Trastuzumab®, an antibody directed against HER-2 and which inhibits neoplastic cell proliferation in vivo and in vitro and enhances chemosensitivity [41]. Therefore, not only identification of disseminated tumor cells but even more their characterization at the protein and gene level has become increasingly important. The presence of the epithelial antigen EpCAM on DTCs might be a target for immunotherapy, which has already been shown to be effective in ovarian cancer by the intraperitoneal application of a trifunctional, bispecific antibody targeting EpCAM, CD3 and accessory cells [42]. Interestingly, in our present study, all persisting DTC in the BM after first-line chemotherapy were

337

EpCAM positive cells. That means, that in case of clinical relapse an individualized therapy with the target EpCAM could be taken into consideration. Since the knowledge and the prognostic value of DTC in gynecological cancers has become more and more important, the BM status after chemotherapy as well as the characterization of these residual cells probably might serve as a biologic correlate for patients at high risk of distant metastasis to warrant additional cell cycle-independent therapy. However the clinical impact of persistence of DTC remains to be evaluated in further multicenter trials. Acknowledgments We greatly appreciate the support of the statistician Dr. Nils Lehmann, Institute for Medical Informatics, Biometry and Epidemiology, University of Duisburg-Essen, Germany. References [1] Arbeitsgemeinschaft Bevölkerungsbezogener Krebsregister in Deutschland: Krebs in Deutschland, 4. erw. Aufl., Saarbrücken, 2002. [2] Du Bois A, Lück HJ, Meier W, et al. A randomized clinical trial of cisplatin/paclitaxel versus carboplatin/paclitaxel as first-line treatment of ovarian cancer. J Natl Cancer Inst 2003;95:1320–30. [3] Interdisciplinary Guidelines of the German Cancer Association and the German Association of Gynecology and Obstetrics. www.ago-ovar.de/ leitlinien/leitlinien.html, 2005. [4] Bristow RE, Tomacruz RS, Armstrong DK, et al. Survival effect of maximal cytoreductive surgery for advanced ovarian carcinoma during the platinum era: a meta-analysis. J Clin Oncol 2002;20:1248–59. [5] Eisenkop SM, Spirtos NM, Friedman RL, et al. Relative influences of tumor volume before surgery and the cytoreductive outcome on survival for patients with advanced ovarian cancer: a prospective study. Gynecol Oncol 2003;90:390–6. [6] Abdul-Karim FW, Kida M, Wentz WB, et al. Bone metastasis from gynecologic carcinomas: a clinicopathologic study. Gynecol Oncol 1990;39: 108–14. [7] Dauplat J, Hacker NF, Nieberg RK, et al. Distant metastases in epithelial ovarian carcinoma. Cancer 1987;60:1561–6. [8] Pantel K, Brakenhoff RH. Dissecting the metastatic cascade. Nat Rev Cancer 2004;4:448–56. [9] Braun S, Vogl FD, Naume B, et al. A pooled analysis of bone marrow micrometastasis in breast cancer. N Engl J Med 2005;53:793–02. [10] Slade MJ, Singh A, Smith BM, et al. Persistence of bone marrow micrometastases in patients receiving adjuvant therapy for breast cancer: results at 4 years. Int J Cancer 2004;114:94–100. [11] Wiedswang G, Borgen E, Karesen R, et al. Detection of isolated tumor cells in bone marrow is an independent prognostic factor in breast cancer. J Clin Oncol 2003;21:3469–78. [12] Wiedswang G, Borgen E, Karesen R, et al. Isolated tumor cells in bone marrow three years after diagnosis in disease-free breast cancer patients predict unfavorable outcome. Clin Cancer Res 2004;10:5342–8. [13] Cain JM, Ellis GK, Collins C, et al. Bone marrow involvement in epithelial ovarian cancer by immunocytochemical assessment. Gynecol Oncol 1990;38:442–5. [14] Ross AA, Miller GW, Moss TJ. Immunocytochemical detection of tumor cells in bone marrow and peripheral blood stem cell collections from patients with ovarian cancer. Bone Marrow Transplant 1995;15:929–33. [15] Braun S, Schindlbeck C, Hepp F, et al. Occult tumor cells in bone marrow of patients with locoregionally restricted ovarian cancer predict early distant metastatic relapse. J Clin Oncol 2001;19:368–75. [16] Marth C, Kisic J, Kaern J, et al. Circulating tumor cells in the peripheral blood and bone marrow of patients with ovarian carcinoma do not predict prognosis. Cancer 2002;94:707–12.

338

P. Wimberger et al. / Gynecologic Oncology 107 (2007) 331–338

[17] Fehm T, Becker S, Bachmann C, et al. Detection of disseminated tumor cells in patients with gynecological cancers. Gynecol Oncol 2006;103: 942–7. [18] International Federation of Gynecology and Obstetrics. Changing the definition of clinical staging for carcinoma of the cervix and ovary. Am J Obstet Gynecol 1987;156:263–4. [19] Muller V, Stahmann N, Riethdorf S, et al. Circulating tumor cells in breast cancer: correlation to bone marrow micrometastases, heterogeneous response to systemic therapy and low proliferative activity. Clin Cancer Res 2005;11:3678–85. [20] Caulin C, Salvesen GS, Oshima RG. Caspase cleavage of keratin 18 and reorganization of intermediate filaments during epithelial cell apoptosis. J Cell Biol 1997;138(6):1379–94. [21] Leers MP, Kolgen W, Bjorklund V, et al. Immunocytochemical detection and mapping of a cytokeratin 18 neo-epitope exposed during early apoptosis. J Pathol 1999;187(5):567–72. [22] Fehm T, Solomayer EF, Meng S, et al. Methods for isolating circulating epithelial cells and criteria for their classification as carcinoma cells. Cytotherapy 2005;7(2):171–85. [23] Fehm T, Becker S, Pergola-Becker G, et al. Presence of apoptotic and non apoptotic disseminated tumor cells reflect response to neoadjuvant systemic therapy (NST) in breast cancer. Breast Cancer Res 2006;8(5):1–7. [24] Kaplan EL, Meier P. Nonparametric estimation from incomplete observations. Jam Stat Assoc 1958;53:457–81. [25] Janni W, Rack B, Schindlbeck C, et al. The persistence of isolated tumor cells in bone marrow from patients with breast carcinoma predicts an increased risk for recurrence. Cancer 2005;103(5):884–91. [26] Braun S, Kentenich CRM, Janni W, et al. Lack of effect of adjuvant chemotherapy on the elimination of single dormant tumor cells in bone marrow of high-risk beast cancer patients. J Clin Oncol 2000;18:80–6. [27] Pantel K, Schlimok G, Angstwurm M, et al. Methodological analysis of immunocytochemical screening for disseminated epithelial tumor cells in bone marrow. J Hematother 1994;3(3):165–73. [28] Passlick B, Kubuschok B, Izbicki JR, et al. Isolated tumor cells in bone marrow predict reduced survival in node-negative non-small cell lung cancer. Ann Thorac Surg 1999;68(6):2053–8. [29] Janni W, Hepp F, Strobl B, et al. Patterns of disease recurrence influenced by hematogenous tumor cell dissemination in patients with cervical carcinoma of the uterus. Cancer 2003;97(2):405–11.

[30] Braun S, Müller P, Hepp F, et al. Micrometastatic breast cancer cells in bone marrow at primary surgery: prognostic value in comparison with nodal status. J Natl Cancer Inst 1998;90:1099–101. [31] Mehes G, Witt A, Kubista E, Ambros PF. Circulating breast cancer cells are frequently apoptotic. Am J Pathol 2001;159:17–20. [32] Gaforio JJ, Serrano MJ, Sanchez-Rovira P, et al. Detection of breast cancer cells in the peripheral blood is positively correlated with estrogen-receptor status and predicts for poor prognosis. Int J Cancer 2003;107:984–90. [33] Taubert H, Blümke K, Bilkenroth U, et al. Detection of disseminated tumor cells in peripheral blood of patients with breast cancer: correlation to nodal status and occurrence of metastases. Gynecol Oncol 2004;92:256–61. [34] Cristofanilli M, Budd GT, Ellis MJ, et al. Circulating tumor cells, disease progression, and survival in metastatic breast cancer. N Engl J Med 2004;351:781–91. [35] Judson PL, Geller MA, Bliss RL, et al. Preoperative detection of peripherally circulating cancer cells and its prognostic significance in ovarian cancer. Gynecol Oncol 2003;91:389–94. [36] Benoy IH, Elst H, Philips M, et al. Real-time RT-PCR detection of disseminated tumour cells in bone marrow has superior prognostic significance in comparison with circulating tumour cells in patients with breast cancer. Br J Cancer 2006;94:672–80. [37] Wiedswang G, Borgen E, Schirmer C, et al. Comparison of the clinical significance of occult tumor cells in blood and bone marrow in breast cancer. Int J Cancer 2006;118:2013–9. [38] Braun S, Pantel K. Micrometastatic bone marrow involvement: detection and prognostic significance. Med Oncol 1999;16(3):154–65. [39] Heiss MM, Allgayer H, Gruetzner KU, et al. Individual development and uPA-receptor expression of disseminated tumour cells in bone marrow: a reference to early systemic disease in solid cancer. Nat Med 1995;1(10): 1035–9. [40] Solomayer EF, Becker S, Pergola-Becker G, et al. Comparison of HER2 status between primary tumor and disseminated tumor cells in primary breast cancer patients. Breast Cancer Res Treat 2006;98(2):179–84. [41] Meng S, Tripathy D, Frenkel EP, et al. Circulating tumor cells in patients with breast cancer dormancy. Clin Cancer Res 2004;10:8152–62. [42] Wimberger P, Schulte A, Jäger M, et al. Intraperitoneal application of the trifunctional antibody (anti-EpCAM × anti-CD3) to ovarian cancer patients with malignant ascites influences circulating tumor- and immune effector cells. Proc Am Assoc Cancer Res 2005:46.