Advances in Brief Detection of Germ-Cell Tumor Cells in Peripheral Blood Progenitor Cell Harvests: Impact on Clinical Outcome

Our study was conducted to evaluate the impact of tumor cell contamination in peripheral blood progenitor cell (PBPC) harvests on the clinical outcome of patients with germ-cell tumors undergoing high-dose chemotherapy (HDCT) and autologous PBPC reinfusion. Samples of mononuclear cells from progenitor cell harvests of 57 patients with advanced or recurrent germ-cell tumors were retrospectively screened for contaminating tumor cells using immunocytochemical staining for cytokeratin filaments and reverse transcription-PCR (RT-PCR) testing for germ-cell alkaline phosphatase mRNA. The results were correlated to clinical prognostic variables as well as to the overall and event-free survival of these patients. Tumor cell contamination was detected in PBPC harvests of 16 of 57 enrolled patients (28%), and, among these, in 14 of 51 (27%) who underwent HDCT. The presence of contaminating tumor cells as detected by either immunocytochemical staining, RT-PCR, or both was strongly associated with a reduced overall survival (43% versus 71%, P 5 0.0037) and event-free survival (0% versus 52%, P 5 0.0005) after 1 year. In multivariate analysis, the demonstration of contaminating tumor cells had a higher predictive value for a poor event-free survival than other known prognostic variables. The presence of contaminating tumor cells in PBPC harvests of patients with germ-cell tumors seems to predict a poor overall and event-free survival in patients undergoing HDCT and autologous PBPC reinfusion. Introduction Current treatment modalities of metastatic germ-cell tumors result in CRs and lasting remissions in the majority of patients (1). Even those who do not respond to conventionaldose treatment may still benefit from HDCT, followed by reinfusion of autologous hematopoietic progenitor cells harvested from the bone marrow or the peripheral blood (2, 3). A potential limitation of HDCT is the risk of tumor cell contamination of autologous progenitor cell products (4), which is increasingly being regarded as a clinically relevant adverse prognostic factor (5). We have recently established a sensitive technique for the detection of germ-cell tumor cells in autologous PBPC products of patients with advanced germ-cell tumors (6). In the present analysis, we correlate these results to the overall and event-free survival of germ-cell tumor patients after HDCT. In addition, we investigate the interaction of tumor cell contamination as a prognostic variable with clinical indicators for treatment outcome after HDCT that have been defined previously (7). Patients and Methods Patients. The patient characteristics are shown in Table 1. As described previously, patients with germ-cell tumors who progressed during or who relapsed after cisplatin-based chemotherapy (n 5 57) were enrolled in consecutive clinical trials that included conventional-dose salvage treatment, followed by HDCT (8). The clinical trials and the study on tumor cell contamination presented here were approved by the local ethics committee, and all patients gave their written and informed consent before participation. As conventional-dose treatment before HDCT, paclitaxel and ifosfamide, followed by paclitaxel, ifosfamide, and cisplatin, were used in 44 patients; cisplatin, etoposide, and ifosfamide were given in 12 patients; and cisplatin, ifosfamide, and vinblastine were given in 1 patient. Conventional-dose treatment was administered to achieve tumor control before HDCT, to allow mobilization and apheresis of PBPCs and to test the sensitivity of tumors to cisplatin-based chemotherapy. Six patients received conventional-dose chemotherapy and had PBPCs harvested but did not proceed to HDCT because of massive tumor progression (n 5 4) or poor renal function (n 5 2). The response to conventional-dose salvage treatment and the disease status of the remaining 51 patients immediately before HDCT are shown in Table 2. After the first cycle of conventional-dose chemotherapy, patients received 5–10 mg of granulocyte colony-stimulating factor s.c. per kg body weight daily. PBPC apheresis products Received 6/20/00; revised 10/2/00; accepted 10/13/00. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. 1 Supported by a grant from the Wilhelm Sander-Stiftung (Munich, Germany). 2 To whom requests for reprints should be addressed, at Medizinische Klinik und Poliklinik, Charité Campus Virchow-Klinikum, HumboldtUniversität zu Berlin, Augustenburger Platz 1, 13353 Berlin, Germany. Phone: 49-30-45053539; Fax: 49-30-45053900; E-mail: hildebra@ charite.de. 3 The abbreviations used are: CR, complete remission; HDCT, highdose chemotherapy; PBPC, peripheral blood progenitor cell; RT-PCR, reverse transcription-PCR; GCAP, germ-cell alkaline phosphatase; PD, progressive disease; MNC, mononuclear cell; PRm, partial remission, marker. 4641 Vol. 6, 4641–4646, December 2000 Clinical Cancer Research Research. on April 30, 2017. © 2000 American Association for Cancer clincancerres.aacrjournals.org Downloaded from were collected with an AS 104 collector (Fresenius AG, St. Wendel, Germany) or a Cobe Spectra cell collector (Cobe Laboratories, Heimstetten, Germany) when the number of CD34-positive cells in the peripheral blood increased to 10 per ml or more. In 34 patients, one harvest was performed. In 23 patients, at least two consecutive progenitor cell aphereses were performed to obtain a sufficient number of progenitor cells. A sample was taken from the first apheresis product and subjected to analysis for tumor cell contamination. The volume of an aliquot was 1.5–2 ml, with an average cell number of 5 3 10 nucleated cells per ml. For RNA preparation, the cells were frozen immediately and stored in liquid nitrogen following a standard protocol. For the preparation of cytospins and immunocytochemical staining, only fresh cells were used. RNA Preparation. Samples of peripheral blood MNCs were separated on a Ficoll-Hypaque gradient (Pharmacia, Uppsala, Sweden), washed twice with PBS containing 1% FCS, counted, and tested for viability by means of trypan blue exclusion. Cells (2–3 3 10) were lysed in TRIzol reagent (Life Technologies, Inc., Grand Island, NY) and sheared to homogeneity, following a one-step guanidinium isothiocyanate/phenol RNA preparation technique (9). Chloroform was added, and the RNA contained in the aqueous upper phase was precipitated in isopropanol at 280°C, washed in 80% ethanol, and resuspended in RNase-free water. The integrity of RNA was examined by RT-PCR analysis for b2 microglobulin mRNA (10). cDNA Synthesis and PCR. RNA (10 mg ) was incubated with 50 ng of a 15(dT) primer at 60°C for 10 min. After the addition of a deoxynucleotide mix (10 mM), DTT (10 mM), and first-strand reaction buffer, 100 units of Moloney murine leukemia virus reverse transcriptase (Life Technologies, Inc.) were added to a final volume of 50 ml. The cDNA synthesis reaction Table 1 Patient characteristics at study entry Variable Tumor cells detected Total (n 5 57) P No (n 5 41) Yes (n 5 16) Age in years (median and range) 32 (20–54) 34 (23–53) 32 (20–54) 0.413 Location of primary tumor Gonadal 30 (73%) 12 (75%) 42 (74%) Retroperitoneal and/or mediastinal 10 (24%) 4 (25%) 14 (24%) Other 1 (3%) 1 (1%) 0.718 Maximal response to first-line therapy CR/PRm2 29 (71%) 9 (56%) 38 (67%) PRm1/SD/PD 12 (29%) 7 (44%) 19 (33%) 0.402 Maximal sensitivity to cisplatin during prior chemotherapy Sensitive 32 (78%) 11 (69%) 43 (76%) Refractory 5 (12%) 4 (25%) 9 (15%) Absolutely refractory 3 (7%) 1 (6%) 4 (7%) 0.516 Tumor marker concentration in serum at study entry (median, range) HCG (unit/l) 43 (0–17630) 39 (2–984711) 43 (0–984711) 0.486 a-Fetoprotein (ng/ml) 5 (1–92279) 6 (1–1080) 5 (1–92279) 0.774 a Mann-Whitney U test. b x test. c SD, stable disease. d Data in one patient not available. Table 2 Assessment of patients before HDCT Variable Tumor cells detected Total (n 5 51) P No (n 5 38) Yes (n 5 13) Response status before HDCT CR/PRm2 21 (55.3%) 5 (38%) 26 (51%) PRm1/SD/PD 17 (44.7%) 8 (62%) 25 (49%) 0.3 Sensitivity to cisplatinum before HDCT Sensitive 28 (74%) 8 (61%) 36 (71%) Refractory 7 (18%) 4 (31%) 11 (22%) Absolutely refractory 3 (8%) 1 (8%) 4 (7%) 0.56 Tumor marker concentration in serum before HDCT (median, range) HCG (unit/l) 2 (0–9110) 2 (1–150) 2 (0–9110) 0.628 Clinical risk category before HDCT Good risk 26 (68%) 7 (54%) 33 (65%) 0.59 Intermediate risk 8 (21%) 5 (38%) 13 (25%) Poor risk 4 (11%) 1 (8%) 5 (10%) a SD, stable disease. b x test. c Mann-Whitney U test. 4642 Germ-Cell Tumor Cells in Progenitor Cell Harvests Research. on April 30, 2017. © 2000 American Association for Cancer clincancerres.aacrjournals.org Downloaded from was carried out at 37°C for 40 min. After heat inactivation at 95°C for 5 min and subsequent maintenance at 4°C, 5 ml of each reaction product were subjected to PCR analysis. PCR analysis was performed as described previously (6). In brief, PCR was carried out at 94°C (40 s), 65°C (1 min), and 72°C (1 min) for 40 cycles, followed by a final 5 min at 72°C. All reagents used for PCR analysis were purchased from InViTek (Berlin, Germany). Twenty microliters of the final PCR products were separated on a 1% agarose gel (Oncor Appligene, Illkirch, France). For reamplification, 5 ml of the PCR product were again subjected to 40 cycles of PCR using a nested sense primer. For b2 microglobulin and GCAP mRNA amplification, the primers used were the same as described previously (6). Immunocytochemical Staining. For immunocytochemical staining, 10 mg of a commercially available murine F(ab)2 fragment directed against cytokeratin filaments (EpiMet Tumor Cell Detection Kit; Baxter Biotech, Unterschleißheim, Germany) were used. Cell suspensions were attached to slides by cytocentrifugation, air dried, and stored at 220°C. After thawing, the slides were again allowed to dry and fixed in acetone. The germ-cell tumor cell line Tera-2 (American Type Culture Collection, Manassas, VA) was used as a positive control for each immunostaining experiment. Cytospin preparations of peripheral blood MNCs from healthy donors served as a negative control. A standard protocol, as outlined in the kit, was followed. In brief, the slides were permeabilized in methanol with 4% formaldehyde. Because it was linked to alkaline phosphatase, the anticytokeratin F(ab)2 fragment allowed immediate staining with a chromogenic substrate. Tumor cells were considered immunocytochemically positive when staining was observed on .70% of the cytoplasm. Routinely, an overall number of 1–2 3 10 MNCs, with 5 3 10 cells per slide, was analyzed and tumor cells were quantified. Definitions. A clinical CR was present if there was disappearance of all radiological manifestations and normalization of tumor markers by chemotherapy alone. Patients with normalization of tumor markers and complete resection of necrosis/ mature teratoma or undifferentiated viable tumor were considered having pathological CR or surgical CR, respectively. Patients with normalization of tumor markers, but radiological evidence of disease, were considered PRm2. Patients with a reduction of radiological manifestations of 50% or more or with a decline of tumor markers of 90% or more were considered PRm1. All other patients were classified as having stable disease or PD. Sensitivity to cisplatin was assessed as reported previously (11). Any disease was considered sensitive to cisplatin when more than stable disease was achieved for .4 weeks. Any disease was considered refractory to cisplatin when stable disease or better was achieved, but when there was evidence of tumor progression within 4 weeks of the last cisplatin-based treatment. Any disease was considered absolutely refractory to cisplatin when not even stable disease was achieved following cisplatin-based chemotherapy. The progenitor cell harvest of a patient was defined as positive for contaminating tumor cells if the RT-PCR for GCAP mRNA was positive in two separate analyses of an RNA sample or if cytokeratin-positive cells were detected on at least two cytospin preparations. Statistical Analysis. Two categories of patients with and without tumor cell contamination of PBPC products were formed, and clinical variables of patient and tumor characteristics were compared between these categories using the MannWhitney U test for continuous variables and the x test for discontinuous variables. The median follow-up time for patients still alive was 16 months, with a range of 3–63 months. Overall and event-free survival rates were calculated according to the Kaplan-Meier method (12). Univariate comparisons of the event-free and overall survival rates were performed using the log-rank test. The Cox proportional hazards regression model was used for the multivariate survival analysis (13). All calculations were performed with SPSS 8.0 computer software.