Introduction

Peritoneal carcinomatosis is the most common secondary cancer that affects the peritoneal cavity. Secondary peritoneal carcinomatosis is a frequent consequence of primary cancers of the gastrointestinal tract as well as of ovarian, breast, lung and uterine cancer [1]. Prognosis of peritoneal spread of a primary carcinoma is poor because the existing treatment strategies are less effective. These strategies include cytoreductive surgery as a single treatment or in combination with adjuvant intravenous or intraperitoneal chemotherapy [2]. In hyperthermic perfusion therapy chemotherapeutic compounds are heated to 41–43 °C to increase their cytotoxicity [3, 4]. Furthermore, intraperitoneal chemotherapy has proven superior to intravenous chemotherapy in patients with metastatic ovarian cancer [5].

To improve therapy of peritoneal disease, compounds should be applied that target surface proteins which are expressed by cancer cells. Those compounds can be used in targeted therapy as specific carrier molecules for cytotoxic radionuclides. Therapeutic efficacy of targeted therapy with β-particle emitting radionuclides has been investigated in numerous preclinical and clinical studies [6–12]. Nevertheless, therapeutic efficacy of intraperitoneal radioimmunotherapy using β-emitters is not satisfactory in most cases. Therefore, alternative treatment strategies using α-emitters are being increasingly pursued.

Alpha-particles eradicate single tumour cells or small tumour cell nodules very efficiently. Therefore, targeted therapy with α-emitters is a promising adjuvant strategy for treatment of disseminated tumour cells and micrometastatic disease characteristic for peritoneal carcinomatosis. Accordingly, radioimmunoconjugates composed of α-emitters and antibodies specifically targeting antigens expressed on tumour cells have been successfully administered intraperitoneally in the treatment of experimental peritoneal carcinomatosis of gastric [13, 14], colonic [15, 16] and ovarian origin [17, 18]. Moreover, in a first clinical phase I study nine women in complete remission after second-line chemotherapy for recurrent ovarian carcinoma were treated intraperitoneally with 211At-MX35 F(ab′)2. No adverse effects could be observed in all laboratory parameters analysed. Therefore, intraperitoneal α-emitter radioimmunotherapy of ovarian cancer patients is thought to be efficient without significant toxicity [19].

The 31 aa tumour-homing peptide F3 (KDEPQRRSARLSAKPAPPKPEPKPKKAPAKK) corresponds to the nucleosomal binding domain of HMGN2. F3 shows specific binding to nucleolin present on the surface of proliferating tumour cells. In proliferating cells a cyclical transport of nucleolin occurs from the nucleus to the cell surface and back by a specific shuttle mechanism. In resting cells nucleolin is primarily expressed in the nucleus. Therefore, nucleolin is a promising target for tumour therapies that are aimed at selective eradication of proliferating tumour cells [20]. Internalization of F3 into the nucleus is extremely attractive in terms of therapeutic application of α-emitters. The selective transport of α-emitter-F3 conjugates to the nucleus of tumour cells enhances the probability for induction of lethal DNA double-strand breaks by α-particles [21]. Indeed, α-emitter-F3 radioconjugates have successfully been administered in tumour therapy. 213Bi-diethylenetriaminepentaacetic acid (DTPA)-[F3]2 has been shown to specifically accumulate in tumour tissue in vivo and to interfere with tumour growth of small tumours. Consequently, intraperitoneal injection of 213Bi-DTPA-[F3]2 significantly prolonged overall survival in a preclinical model of peritoneal carcinomatosis [22]. Conjugates of the α-emitter 225Ac coupled to the tumour-homing peptide F3 are consistent with the concept of an atomic nanogenerator. Such a generator releases four α-particles within a tumour cell in the course of 225Ac decay [23]. Therapeutic efficiency of stable conjugates of F3 and 225Ac (225Ac-DOTA-F3) as well as 213Bi (213Bi-DTPA-F3) has been compared in a mouse model of peritoneal carcinomatosis. Therapy with both conjugates increased survival of animals with negligible toxicity [24]. The aim of this study was to further improve therapeutic efficacy of 213Bi-DTPA-F3 by additional administration of the cytostatic drug paclitaxel. For that purpose cytotoxic effects of combined 213Bi-DTPA-F3 and paclitaxel treatment on OVCAR-3 cells were investigated in comparison to the effects of single treatments with 213Bi-DTPA-F3 or paclitaxel. These analyses included alamarBlue and clonogenic assays as well as flow cytometric determination of the mode of cell death and cell cycle arrest. Moreover, therapeutic efficacy of single and combined treatments was analysed in a nude mouse model of peritoneal carcinomatosis of ovarian origin including monitoring of tumour development by bioluminescence imaging, histopathologic analysis of tumour spread and determination of overall survival.

Materials and methods

Cell culture

The ovarian carcinoma cell line OVCAR-3 (ATCC, Wesel, Germany) and the human embryonic kidney cell line HEK293T (gift of D. Saur, Technische Universität München) were cultured in RPMI 1640 medium (Biochrom, Berlin, Germany) supplemented with 10 % fetal bovine serum, 100 U/ml penicillin and 100 μg/ml streptomycin (all from Biochrom). Cells were maintained at 37 °C and 5 % CO2.

213Bi labelling of the F3 peptide

213Bi was eluted from a 225Ac/213Bi generator system provided by the Institute of Transuranium Elements (Karlsruhe, Germany) [25, 26]. To the 213Bi eluate 100 μl ascorbic acid (40 mg/ml) and 100 μl ammonium acetate (3 M) were added. For 213Bi labelling DTPA-conjugated F3 peptide (Bachem, Bubendorf, Switzerland) was supplemented with 37 MBq of 213Bi per 5 μg of peptide. After 7 min at room temperature (RT) labelling efficacy was determined by instant thin-layer chromatography. DTPA (Sigma-Aldrich, Seelze, Germany) was labelled in the same way except that 500 μg DTPA were incubated with 37 MBq of 213Bi. Labelling efficacy was >90 % in all experiments. Stability of the 213Bi-DTPA-F3 peptide conjugate is appropriate for intraperitoneal therapeutic application (data not shown). Toxicities of the β-particle emitting daughter nuclides of 213Bi that are released from the F3 peptide conjugates following decay of the α-emitter 213Bi are negligible.

Cell binding and internalization of 213Bi-DTPA-F3

A total of 1 × 107 OVCAR-3 cells were incubated in 1 ml culture medium with 37 kBq/ml 213Bi-DTPA-F3 or 213Bi-DTPA with or without 10 μM phenylarsine oxide (PAO, Sigma-Aldrich, Seelze, Germany) (0.1 % dimethyl sulphoxide final concentration) for 1 h at 37 °C. Membrane-permeable PAO inhibits the internalization of cell surface receptors [27, 28]. Cells were centrifuged at 400 g for 5 min at 4 °C. Supernatants were transferred to new tubes and cells were resuspended in ice-cold phosphate-buffered saline (PBS). Cells were pelleted again, supernatants were combined with previous supernatants and washing was repeated twice. Finally 213Bi activities in the cell pellets and supernatants were measured using a gamma counter. Binding/internalization of 213Bi-DTPA-F3 was calculated as percentage of 213Bi activity in the cell pellet relating to total 213Bi activity (cell pellet + supernatant).

Assessment of cell viability using the alamarBlue assay

OVCAR-3 cells (4 × 103 per well) were incubated with or without paclitaxel (60 ng/ml; 90 min) 24 h after seeding. Cells were washed once with culture medium and then incubated with or without 213Bi-DTPA-F3 (0.37 MBq/ml). At indicated time points after initiation of treatment cell supernatants were supplemented with 10 % alamarBlue (AbD Serotec, Düsseldorf, Germany) and cells were incubated for 4 h at 37 °C. In the alamarBlue assay resazurin is reduced to red fluorescent resorufin by metabolically active cells. After incubation supernatants were transferred to new plates and stored at −20 °C. Cells received fresh medium and were incubated until the next alamarBlue assay. Resorufin fluorescence was quantified according to the manufacturer’s instructions.

Soft agar clonogenic assay

OVCAR-3 cells (2 × 105 per well) were incubated with or without paclitaxel (60 ng/ml; 90 min) 24 h after seeding. Cells were then treated with 213Bi-DTPA-F3 for 16 h. Six-well plates were layered with 1.5 ml base agarose (0.75 ml 1 % agarose +0.75 ml 2× culture medium both pre-warmed to 40 °C) per well. Cells were detached with trypsin/ethylenediaminetetraacetic acid (EDTA) (Biochrom, Berlin, Germany) and resuspended in 2 ml culture medium. To a pre-warmed (40 °C) mixture of 0.7 % low melt agarose (3 ml) (Sigma-Aldrich) and 2× culture medium (3 ml) 0.2 ml of the cell suspension was added. After gentle mixing 1.5 ml of the agarose cell suspension (top agarose) was seeded per well in duplicates in six-well plates containing base agarose. Cells were incubated at 37 °C for 1 week and then stained with 0.5 ml of 0.005 % crystal violet for 1 h. Clones were counted microscopically.

Generation of luciferase-GFP-expressing OVCAR-3 cells

OVCAR-3 cells stably expressing luciferase and green fluorescent protein (GFP) (OVCAR-3-Luc-GFP) were generated by lentiviral transduction. At first HEK293T cells were transfected in a culture dish with a mixture of pLenti6/UbC/V5-DEST-Luc-GFP, psPAX2 and pMD2.G at a weight ratio of 1:1:1 using TurboFect (Fermentas, St.Leon-Rot, Germany). Culture supernatants containing replication incompetent lentiviral particles were harvested 24, 32 and 48 h after transfection, passed through 0.45 μm pore size filters, supplemented with 8 μg/ml polybrene (Sigma-Aldrich) and added immediately to OVCAR-3 cells grown to ~70 % confluence. Transduced cells were incubated for 72 h and single clones were obtained by limiting dilution cloning. GFP-positive clones were detected by fluorescence microscopy. Positive clones were expanded in culture medium containing 10 μg/ml blasticidin S (Life Technologies, Carlsbad, CA, USA). Clones were seeded in an opaque 96-well plate and supplemented with 2.5 mM D-luciferin. Luciferase activity was visualized by bioluminescence imaging.

Treatment of i.p. xenograft mice and monitoring of therapeutic efficacy

OVCAR-3-Luc-GFP cells (1 × 107 in 100 μl PBS) were injected intraperitoneally into 6- to 8-week-old severe combined immunodeficiency (SCID) mice (Charles River, Sulzfeld, Germany). Treatment was initiated 10 days after cell inoculation comprising four different groups: PBS (n = 10), paclitaxel (n = 10), 213Bi-DTPA-F3 (n = 11) and combined paclitaxel + 213Bi-DTPA-F3 (n = 9). At day 10 after tumour cell inoculation animals of the paclitaxel group and the combined group received i.p. injections of 120 μg paclitaxel (in 100 μl PBS), respectively. Animals of the control group and the 213Bi-DTPA-F3 group were i.p. injected with 100 μl PBS, respectively. At day 11 after tumour cell inoculation the 213Bi-DTPA-F3 group and the combined group received i.p. injections of 1.85 MBq of 213Bi-DTPA-F3 (in 100 μl PBS), while the control group and the paclitaxel group were treated i.p. with 100 μl PBS, respectively. According to previous studies [22, 24] treatments were repeated five times at the time points indicated in Fig. 5a. Tumour development was monitored noninvasively in two mice each of every treatment group using bioluminescence imaging 7, 12, 19, 26, 33 and 40 days after tumour cell inoculation. For that purpose anaesthetized animals were injected with 300 μl D-luciferin (50 mM in 0.9 % NaCl) i.p. Ten minutes after injection bioluminescence imaging was performed using a cooled charge-coupled device (CCD) camera (Hamamatsu, Herrsching am Ammersee, Germany).

Histopathologic evaluation of therapeutic efficacy

Thirteen days after the last tumour treatment three mice of every group (including the animals that were used for bioluminescence imaging) were sacrificed and small and large intestine including mesentery and tumours were dissected and fixed in 4 % formalin. Tumour nodules in the tissue sections stained with haematoxylin and eosin (H&E) were counted microscopically. Quantification of tumour areas was performed using the image analysis platform Definiens Enterprise Image Intelligence Suite (Definiens, Munich, Germany).

Flow cytometric analysis of cell death and cell cycle arrest

OVCAR-3 cells (1 × 106 per well) were incubated with or without paclitaxel (60 ng/ml; 90 min) 24 h after seeding. Subsequently cells were incubated with or without 213Bi-DTPA-F3 (0.37 MBq/ml). At 24 and 48 h after treatment cells were detached using Accutase and washed twice with PBS. For cell death analysis the detached cells (+ cells from culture supernatants) were incubated with 5 μl/ml annexin V-fluorescein isothiocyanate (FITC) solution (BD Biosciences, Heidelberg, Germany) and 0.5 μg/ml propidium iodide (PI) for 15 min at 4 °C. PI and FITC fluorescence of cells were analysed by dual-parameter flow cytometry (Beckman Coulter). Percentages of necrotic and apoptotic cells were determined using the ‘FlowJo’ software. For cell cycle analysis detached cells were fixed in 80 % ethanol (2 h, 4 °C), incubated with RNase (0.1 mg/ml; 5 min RT), treated with pepsin (5 mg/ml in 50 mM HCl; 10 min 37 °C), stained with PI (50 μg/ml) and subjected to flow cytometric analysis (FACScalibur, Becton Dickinson). Histograms were evaluated as described [29].

Detection of active caspase-3 by Western blotting

OVCAR-3 cells (5 × 105 per well) were incubated with or without paclitaxel (60 ng/ml; 90 min) 24 h after seeding. Subsequently cells were incubated with or without 213Bi-DTPA-F3 (0.37 MBq/ml). Then, 24 and 48 h later detached cells (+ cells from culture supernatants) were lysed [2 % sodium dodecyl sulphate (SDS), 62.5 mM Tris–HCl, pH 6.8; 5 min 95 °C]. Lysates were incubated for 5 min at 95 °C. Lysates (30 μg each) were subjected to SDS polyacrylamide gel electrophoresis (PAGE) (14 %). Western blotting using polyclonal anti-active-caspase-3 antibody (Millipore, Schwalbach/Ts, Germany), peroxidase-conjugated monoclonal anti-β-actin antibody (clone 8226, Abcam, Cambridge, UK) and peroxidase-conjugated anti-rabbit IgG antibody (GE Healthcare, Hatfield, UK) was performed as described previously [30].

Statistical analysis

Software package SPSS Statistics version 17.0.0 for Microsoft Windows XP (SP3) was used for statistical analysis. Significance of survival as expressed in Kaplan-Meier curves was determined using the log-rank test. The level of statistical significance was set at p < 0.05.

Results

213Bi-DTPA-F3 is internalized by OVCAR-3 cells

Alpha-emitters show maximum therapeutic efficacy when delivered into the nucleus of tumour cells. We have demonstrated that F3 successfully transports the α-emitter 213Bi into the nucleus of MDA-MB-435 cells and efficiently targets xenograft tumours that express nucleolin [22]. To determine whether the monomeric F3 peptide is applicable for delivery of 213Bi into OVCAR-3 cells, we compared binding/uptake of 213Bi-DTPA-F3 and 213Bi-DTPA. As shown in Fig. 1, approx. 10 % of 213Bi-DTPA-F3 added was bound and/or internalized by OVCAR-3 cells after 1 h at 37 °C, whereas less than 1 % of 213Bi-DTPA was bound and/or internalized. In proliferating tumour cells nucleolin, the cell surface receptor of the peptide ligand F3, constantly shuttles from the cell surface to the cell nucleus, thus transporting 213Bi-DTPA-F3 as a payload [20]. To discriminate cell surface-bound from internalized F3, the cells were co-incubated with PAO, an inhibitor of endocytosis [27]. PAO reduced internalization of 213Bi-DTPA-F3 but not of 213Bi-DTPA indicating that approx. 4 % of 213Bi-DTPA-F3 is bound to the cell surface and that approx. 6 % has been internalized (Fig. 1).

Fig. 1
figure 1

Binding and internalization of 213Bi-DTPA-F3 by OVCAR-3 cells. OVCAR-3 cells were incubated with 37 kBq/ml 213Bi-DTPA-F3 or 213Bi-DTPA for 1 h at 37 °C with or without the endocytosis inhibitor PAO (10 μM). After centrifugation of cells 213Bi activity was measured in pelleted cells and supernatants. Bound/internalized 213Bi activities were calculated (%) (n = 3 ± SD)

Full size image

Combined treatment of OVCAR-3 cells with 213Bi-DTPA-F3 + paclitaxel is more cytotoxic than treatment with either 213Bi-DTPA-F3 or paclitaxel

The therapeutic effect of ionizing radiation can be enhanced by the chemotherapeutic drug paclitaxel [15, 31, 32]. In this study the effects of combined treatment with 213Bi-DTPA-F3 and paclitaxel were investigated in terms of cell viability/growth (alamarBlue assay) and proliferation (clonogenic assay) in OVCAR-3 cells compared to treatments with either 213Bi-DTPA-F3 or paclitaxel. The IC50 values of 213Bi-DTPA-F3 and paclitaxel have been determined to 370 kBq/ml and 60 ng/ml, respectively (data not shown). Therefore, these IC50 concentrations of 213Bi-DTPA-F3 and paclitaxel were applied in all subsequent in vitro experiments, unless indicated otherwise. As shown in Fig. 2a treatments with 213Bi-DTPA-F3, paclitaxel and combined 213Bi-DTPA-F3 + paclitaxel did not affect cell viability/growth within the first 24 h. At 48 and 72 h, however, treatments markedly reduced cell viability/growth compared to the untreated control, with combined 213Bi-DTPA-F3 + paclitaxel showing the most pronounced effect. In clonogenic assays OVCAR-3 cells were treated with or without paclitaxel (60 ng/ml) and 213Bi-DTPA-F3 up to 740 kBq/ml (Fig. 2b). Clonogenic cell survival was reduced by 213Bi-DTPA-F3 in an activity concentration-dependent manner. This effect was enhanced by paclitaxel except for the highest activity concentration of 740 kBq/ml.

Fig. 2
figure 2

Cell viability/survival after treatment of OVCAR-3 cells with paclitaxel, 213Bi-DTPA-F3 or combined paclitaxel + 213Bi-DTPA-F3. a Viability of OVCAR-3 cells was measured at different time points (26, 51 and 72 h) after treatment of cells with paclitaxel (60 ng/ml), 213Bi-DTPA-F3 (0.37 MBq/ml) or combined paclitaxel + 213Bi-DTPA-F3 using the alamarBlue® assay (n = 3 ± SD). b Survival of OVCAR-3 cells was quantified in a clonogenic assay 8 days after treatment with different activity concentrations of 213Bi-DTPA-F3 (46.3, 92.5, 185, 370, 740 MBq/ml) or a regimen including combined 213Bi-DTPA-F3 and paclitaxel (n = 2 ± SD)

Full size image

Combined treatment of OVCAR-3 cells with 213Bi-DTPA-F3 + paclitaxel results in enhanced induction of apoptosis and G2/M phase arrest

Quantification of cell death and cell cycle arrest in 213Bi-DTPA-F3- and/or paclitaxel-treated OVCAR-3 cells was performed by flow cytometry. Combined 213Bi-DTPA-F3 + paclitaxel treatment resulted in enhanced induction of apoptosis and necrosis as compared to single treatments with either 213Bi-DTPA-F3 or paclitaxel after 48 h (Fig. 3a, b). Induction of apoptosis and necrosis by any of the treatments applied could be inhibited by the membrane-permeable caspase-3 inhibitor Z-VAD-FMK indicating that occurrence of necrosis is secondary to apoptosis. Caspase-3 activation, indicative of induction of apoptosis, as determined by Western blotting correlated well with the flow cytometry results (Fig. 3c). Combined 213Bi-DTPA-F3 + paclitaxel treatment resulted in the strongest caspase-3 activation of all treatments after 48 h. Moreover, 213Bi-DTPA-F3 and/or paclitaxel treatments induced G2/M phase arrest (Fig. 4a, b). While 213Bi-DTPA-F3 triggered a transient increase in G2/M phase cells peaking at 24 h that returned to the control level at 72 h, paclitaxel treatment induced less but sustained increase in G2/M phase cells. Combined 213Bi-DTPA-F3 + paclitaxel treatment induced the most pronounced increase in G2/M phase cells also peaking at 24 h, however dropping to the paclitaxel level at 72 h.

Fig. 3
figure 3

Characterization of cell death in OVCAR-3 cells induced by treatment with paclitaxel and 213Bi-DTPA-F3. OVCAR-3 cells were treated with paclitaxel (60 ng/ml), 213Bi-DTPA-F3 (0.37 MBq/ml) or combined paclitaxel + 213Bi-DTPA-F3 with or without the caspase inhibitor Z-VAD-FMK (50 μM); 48 h after treatment cells were incubated with annexin V-FITC and PI and subjected to flow cytometry. a Percentages of apoptotic and necrotic cells (n = 3 ± SD). b Flow cytometric distribution of cells. Representative dot plots of triplicates are shown. Areas of necrotic (rectangle) and apoptotic (ellipse) cells are marked. c Western blot detection of active caspase-3 in cell lysates. Blots were re-probed with an anti-β-actin antibody

Full size image
Fig. 4
figure 4

Cell cycle analysis of OVCAR-3 cells after treatment with paclitaxel, 213Bi-DTPA-F3 or combined paclitaxel + 213Bi-DTPA-F3. At 24, 48 and 72 h after treatment with paclitaxel (60 ng/ml), 213Bi-DTPA-F3 (0.37 MBq/ml) or combined paclitaxel + 213Bi-DTPA-F3, OVCAR-3 cells were fixed, stained with PI and subjected to single-parameter flow cytometry. a Percentages of OVCAR-3 cells in G2/M phase (n = 3 ± SD). b Representative histograms showing the proportions of cells in G1, S and G2/M phase

Full size image

Combined 213Bi-DTPA-F3 + paclitaxel treatment results in an enhanced inhibition of tumour growth

To investigate whether the enhanced effect of combined treatment with 213Bi-DTPA-F3 + paclitaxel as observed in vitro can be translated to tumour treatment in a mouse model, we generated xenograft tumours by intraperitoneal injection of OVCAR-3-Luc-GFP cells into SCID mice. Ten days after tumour cell inoculation mice received 120 μg paclitaxel by intraperitoneal injection. On the following day 1.85 MBq of 213Bi-DTPA-F3 was injected intraperitoneally. Treatments were repeated five times every third or fourth day (Fig. 5a). Control mice were injected with PBS instead of 213Bi-DTPA-F3 and paclitaxel. Tumour development was monitored by bioluminescence imaging (Fig. 5b). On day 7 after tumour cell inoculation, i.e. 3 days before initiation of tumour treatment, most tumours had grown to a detectable size. The PBS-treated control mice displayed a strong increase in the bioluminescence signal over time indicative of uninhibited tumour growth. In the 213Bi-DTPA-F3-treated mice the bioluminescence signal decreased until day 33 after cell inoculation and subsequently increased again as shown at day 40 (Fig. 5b). In the paclitaxel-treated mice the bioluminescence signal remained more or less constant until day 33 after tumour cell inoculation indicative of tumour stasis. However, at day 40 the signal had increased again. In contrast, in the combined treatment group (213Bi-DTPA-F3 + paclitaxel) no signal was detectable from day 19 until day 40 after tumour cell inoculation indicating complete tumour clearance or tumour regression below detection limits.

Fig. 5
figure 5

Efficacy of repeated i.p. treatment of xenograft mice with PBS, paclitaxel, 213Bi-DTPA-F3 or combined paclitaxel + 213Bi-DTPA-F3. Xenografts were generated by i.p. injection of 1 × 107 OVCAR-3-Luc-GFP cells into SCID mice. a Treatment plan for i.p. xenograft mice: animals were injected six times either with paclitaxel (120 μg) or 213Bi-DTPA-F3 (1.85 MBq) every third/fourth day starting at day 10 (paclitaxel) or 11 (213Bi-DTPA-F3) after cell inoculation. Mice receiving combined treatment were injected i.p. six times with paclitaxel (120 μg) and 213Bi-DTPA-F3 (1.85 MBq) on 2 consecutive days. Control mice were i.p. injected with PBS (100 μl). b Monitoring of tumour development by bioluminescence imaging. c, d Survival of i.p. xenograft mice. c Kaplan-Meier plot. d Mean survival ± SEM

Full size image

Combined treatment with 213Bi-DTPA-F3 + paclitaxel results in prolonged survival of xenograft tumour-bearing mice

Inhibition of tumour growth in xenograft mice after combined 213Bi-DTPA-F3 + paclitaxel treatment also resulted in prolonged survival of animals. Xenograft mice that had been treated with combined 213Bi-DTPA-F3 + paclitaxel significantly differed in overall survival from the other three treatment groups (PBS: p = 0.00051; paclitaxel: p = 0.00055; 213Bi-DTPA-F3: p = 0.002). As shown in Fig. 5c, d treatment with paclitaxel did not significantly prolong survival compared to treatment with PBS resulting in a mean survival of approximately 40 days each (p = 0.34). Treatment with 213Bi-DTPA-F3 significantly prolonged survival to 84 days both compared to PBS (p = 0.000074) and paclitaxel (p = 0.000083), while combined 213Bi-DTPA-F3 + paclitaxel treatment increased mean survival to 121 days. Although tumour cells could no longer be detected by bioluminescence imaging up to 40 days after tumour cell inoculation following combined 213Bi-DTPA-F3 + paclitaxel treatment (Fig. 5b), animals of the combined treatment group finally died from intraperitoneal tumour spread. This suggests that single tumour cells that escaped therapy caused fatal tumour growth in the end.

Combined i.p. treatment with 213Bi-DTPA-F3 + paclitaxel efficiently eradicates xenograft tumours

The efficacy of the different treatment protocols (Fig. 5a) was also evaluated via histopathologic analysis. For that purpose the mice subjected to bioluminescence imaging analysis were sacrificed at day 41 after tumour cell inoculation and small and large intestine including mesentery and tumours were dissected and processed for histopathologic analysis. H&E-stained slices of abdominal organs of control mice revealed that intraperitoneal tumours grew as independent nodules with a typical gland-like histology (Fig. 6a). Tumour treatment with paclitaxel or 213Bi-DTPA-F3 resulted in drastic reduction of both the number and size of tumour nodules (Fig. 6b, c). No tumour nodules were detectable in sections from mice that had received combined 213Bi-DTPA-F3 + paclitaxel treatment. Because these animals died approximately 120 days after tumour cell inoculation, single tumour cells that could not be detected histopathologically must have escaped treatment finally causing death of animals due to tumour regrowth.

Fig. 6
figure 6

Histopathologic analysis of intraperitoneal tumour spread after repeated i.p. treatment of xenograft mice with PBS, paclitaxel, 213Bi-DTPA-F3 or combined paclitaxel + 213Bi-DTPA-F3. Xenograft mice were treated as outlined in Fig. 5. a H&E-stained sections of mesentery of small and large intestine 41 days after inoculation of OVCAR-3 tumour cells; arrows point at tumour nodules; top overview, scale bars: 2 mm; bottom magnification of indicated areas, scale bars: 50 μM. b Number of tumour nodules per section and c tumour area (mm2) per section (n = 2 ± SD)

Full size image

Discussion

Peritoneal carcinomatosis is characterized by the dissemination of tumour cells throughout the peritoneal cavity. Most commonly intraperitoneal dissemination of tumour cells occurs from gastrointestinal and ovarian malignancies [1]. Prognosis of peritoneal spread of a primary carcinoma is poor because the treatment strategies applied such as cytoreductive surgery together with adjuvant intravenous or intraperitoneal chemotherapy are not efficient. Therefore, new therapeutic options based on tumour targeting are being increasingly pursued [2]. We have investigated efficacy of targeted therapy with 213Bi-DTPA-F3 conjugates that specifically bind to nucleolin on the surface of ovarian tumour cells in an animal model of peritoneal spread. Nucleolin constantly shuttles from the cell surface to the nucleus and back in proliferating tumour cells. Upon binding of the 213Bi-DTPA-F3 conjugates to nucleolin they are co-transported to the nucleus, thus increasing the probability of induction of lethal DNA double-strand breaks by the α-emitter 213Bi. Efficacy of 213Bi-DTPA-F3 treatment was compared with treatment with the cytostatic drug paclitaxel and with combined 213Bi-DTPA-F3 + paclitaxel treatment. Repeated therapy with 213Bi-DTPA-F3 significantly prolonged survival of mice compared to animals that were mock-treated (PBS) or received paclitaxel. Moreover, combined repeated intraperitoneal application of 213Bi-DTPA-F3 and paclitaxel again improved therapeutic efficacy compared to 213Bi-DTPA-F3 therapy (Fig. 5). This suggests that paclitaxel can potentiate cytotoxic effects on tumour cells induced by 213Bi-DTPA-F3.

Nevertheless, even repeated combined treatment with 213Bi-DTPA-F3 and paclitaxel did not result in complete remission of tumour nodules. As shown in Fig. 6 the histopathologic investigation of intraperitoneal tumour spread at day 41 after tumour cell inoculation did not reveal any signs of tumour cell infiltration on mesenteries of small and large intestine after combined 213Bi-DTPA-F3 and paclitaxel treatment. However, these animals died from tumour regrowth approximately 120 days after tumour cell inoculation (Fig. 5d). This means that single tumour cells must have survived the therapeutic regimen though not detectable by histopathologic investigation. In future trials the number of treatment cycles could be increased because no signs of significant toxicity have been detected both after application of 213Bi-F3 conjugates [24] and paclitaxel (data not shown).

Another discrepancy with respect to the results of the different therapeutic regimens obviously affects mock-treated (PBS) and paclitaxel-treated animals. As shown in Fig. 6, treatment with paclitaxel significantly reduced both the number of tumour nodules and the tumour area compared to mock treatment. In contrast, data on mean survival of these two treatment groups show almost identical results (Fig. 5d). Consequently, treatment with paclitaxel did not significantly prolong survival compared to mock treatment. Because paclitaxel efficiently eradicated tumour cells—in contrast to PBS (Fig. 6)—this could mean that treatment with paclitaxel possibly created aggressive tumour cells showing fast proliferation finally triggering death of animals at almost the same time point as it occurred in the control group. On the other hand, death of these animals might have been caused by small tumour nodules that induced internal haemorrhage of the aorta combined with the development of ascites. Such haemorrhages may also happen during tumour shrinkage or may be even caused by intra-abdominal mechanical changes due to reduction of tumour size.

Targeted therapy with α-emitters of disseminated intraperitoneal disease has been applied successfully in different animal models of gastric [13, 14], colonic [15, 16] and ovarian origin [17, 18]. Therapeutic efficacy of α-emitters predominantly is based on induction of DNA double-strand breaks that trigger persistent cell cycle arrest at G2/M phase finally resulting in cell death. G2/M phase arrest has been demonstrated for HSC45-M2 gastric cancer cells following treatment with 213Bi-d9MAb immunoconjugates [33]. G2/M cell cycle arrest could be confirmed also in tumour cells of mice bearing LS-174T intraperitoneal xenografts after 212Pb radioimmunotherapy [34]. Moreover, the anticancer drug paclitaxel is known to induce G2/M phase arrest. Therefore, combined treatment with α-emitter conjugates and paclitaxel should increase the number of cells arrested in G2/M phase compared to treatment with either α-emitter or paclitaxel alone. Indeed, our results clearly reveal the expected increase in G2/M phase cells as induced by the combined treatment (Fig. 4).

Accordingly, our results clearly demonstrate that the therapeutic effect of α-emitters can be potentiated by additional application of the cytostatic drug paclitaxel. Potentiation of the cytotoxic effects of particle radiation with cytostatic drugs has been described also in previous studies.

In a mouse model with s.c. xenografts of the Ley-positive breast adenocarcinoma cell line MCF-7, i.v. treatment with β-emitter (90Y) or α-emitter (213Bi) hu3S193 immunoconjugates targeting the Ley antigen in combination with paclitaxel significantly prolonged survival compared to treatment with the radioimmunoconjugates only [35, 36]. In a mouse model with s.c. Ley-positive DU145 prostate xenografts combined treatment with 177Lu-hu3S193 immunoconjugates and the epidermal growth factor receptor (EGFR) inhibitor AG1478 or the cytostatic drug docetaxel significantly improved therapeutic efficacy [37].

Experimental treatment of disseminated peritoneal disease with α-emitter immunoconjugates in combination with cytostatic drugs has been investigated in a nude mouse model with i.p. LS-174T human colon cancer xenografts. LS-174T cells are efficiently targeted by trastuzumab due to overexpression of HER2/neu. By development of a treatment regimen combining multiple doses of gemcitabine (50 mg/kg) with 212Pb radioimmunotherapy (370 kBq) median survival increased up to 196.5 days [38]. Therapeutic efficacy of 212Pb-trastuzumab also increased in combination with the chemotherapeutic paclitaxel. In athymic mice bearing 3-day intraperitoneal LS-174T xenografts, 600 μg paclitaxel applied 24 h before radioimmunotherapy extended the median survival from 44 to 171 days [15]. Paclitaxel also provided benefit as to the therapeutic success of 213Bi-trastuzumab treatment of experimental peritoneal carcinomatosis: Median survival increased to 198 days with paclitaxel (600 μg) applied 24 h before 213Bi-trastuzumab (18.5 MBq) followed by another injection of paclitaxel [15]. These results also demonstrate that application of paclitaxel 24 h before 213Bi-trastuzumab is superior to other orders of application.

An enhanced, additive therapeutic benefit was also achieved by dual targeting of two distinct tumour surface molecules with 213Bi-labelled antibodies. Concurrent α-radioimmunotherapy with 213Bi-trastuzumab and 213Bi-HuCC49ΔCH2 [targeting the tumour associated glycoprotein 72 (TAG-72)] significantly prolonged median survival [16]. Additional application of cytostatic drugs could further improve therapeutic efficacy.

In conclusion, multimodality therapy combining targeted radionuclide therapy with administration of cytostatic drugs is a promising therapeutic concept both after i.v. and i.p. application. Therefore, this concept should soon find its way into clinical practice.