Circulating endothelial cells as a therapeutic marker for thalidomide in combined therapy with chemotherapy drugs in a human prostate cancer model

Authors

  • Haiqing Li,

    1. Molecular Pharmacology Section, Medical Oncology Branch, Center for Cancer Research, National Cancer Institute, NIH, Bethesda, MD, USA, and Divisions of Hematology-Oncology, Department of Medicine, European Institute of Oncology, Milan, Italy
    Search for more papers by this author
  • Valentina Raia,

    1. Molecular Pharmacology Section, Medical Oncology Branch, Center for Cancer Research, National Cancer Institute, NIH, Bethesda, MD, USA, and Divisions of Hematology-Oncology, Department of Medicine, European Institute of Oncology, Milan, Italy
    Search for more papers by this author
  • Francesco Bertolini,

    1. Molecular Pharmacology Section, Medical Oncology Branch, Center for Cancer Research, National Cancer Institute, NIH, Bethesda, MD, USA, and Divisions of Hematology-Oncology, Department of Medicine, European Institute of Oncology, Milan, Italy
    Search for more papers by this author
  • Douglas K. Price,

    1. Molecular Pharmacology Section, Medical Oncology Branch, Center for Cancer Research, National Cancer Institute, NIH, Bethesda, MD, USA, and Divisions of Hematology-Oncology, Department of Medicine, European Institute of Oncology, Milan, Italy
    Search for more papers by this author
  • William D. Figg

    1. Molecular Pharmacology Section, Medical Oncology Branch, Center for Cancer Research, National Cancer Institute, NIH, Bethesda, MD, USA, and Divisions of Hematology-Oncology, Department of Medicine, European Institute of Oncology, Milan, Italy
    Search for more papers by this author

William D. Figg, National Cancer Institute, Building 10/Room 5A01, 9000 Rockville Pike, Bethesda, MD 20892, USA. e-mail: wdfigg@helix.nih.gov

Abstract

OBJECTIVE

To investigate how thalidomide confers its survival benefit in prostate cancer, by assessing its effect on circulating endothelial cells (CECs) and progenitors (CEPs) in a combined therapy of thalidomide and chemotherapy drugs in a human prostate cancer xenograft model, as in clinical trials patients treated with both thalidomide and docetaxel had a >50% decrease in prostate-specific antigen (PSA) levels and longer median overall survival than those treated with docetaxel monotherapy.

MATERIALS AND METHODS

A human prostate cancer xenograft model was used to evaluate the effect of either thalidomide, docetaxel or a combination of the two drugs on circulating ECs. Drug treatment was continued for 17 days, and tumours were measured two or three times a week. Blood samples were taken at three different time points: before the treatments, 4 days and 17 days into the treatments, and CECs and CEPs were measured by flow cytometry analysis.

RESULTS

There was an increased level of apoptotic/dead CECs shortly after the intravenous injection of docetaxel, and the addition of thalidomide further increased the apoptotic/dead CEC level, showing that thalidomide enhances the cytotoxicity of docetaxel against tumour vascular ECs.

CONCLUSION

Thalidomide increased the apoptotic/dead CEC level and enhanced the cytotoxicity of docetaxel against tumour vascular ECs, confirming its antiangiogenic property in vivo in combined anticancer treatments. In addition, there was a correlation between the increased apoptotic/dead CEC levels early in the treatment and antitumour efficacy later, suggesting that the apoptotic/dead CEC level could be used as a marker, at an early stage, to predict tumour response to antiangiogenic therapies.

Abbreviations
(C)EC

(circulating) endothelial cell

(C)EP

(circulating) endothelial progenitor

FITC

fluorescein isothiocyanate

Flk1

fetal liver kinase 1/vascular endothelial growth factor receptor 2

CD117

c-kit

7AAD

7-aminoactinomycin-D.

INTRODUCTION

Prostate cancer is the most common cancer among American males [1]. For patients with progressive disease after hormonal therapy, most current treatments are palliative with the exception of docetaxel, which is one of the few agents to show a survival benefit [2]. Angiogenesis is a crucial factor for tumour progression in prostate cancer, and one of the markers for tumour angiogenesis is microvessel counts. In patients with invasive prostate carcinoma, the incidence of metastastic disease increases as microvessel counts increase [3]. The 5-year recurrence-free survival was shown to be significantly higher for patients with a lower microvessel count, compared with those with a higher microvessel count [4]. Furthermore, both cancer-specific survival and overall survival were significantly longer for patients with a lower microvessel count [5], again confirming the importance of tumour angiogenesis in prostate cancer.

Originally developed in the late 1950s as a sedative, it was discovered that thalidomide is a potent teratogen. However, it was re-introduced in clinical practice, albeit under strict regulations, because it is effective in treating patients with erythema nodosum leprosum. The antiangiogenic properties of thalidomide was shown in a rabbit cornea model [6]. Investigators examined the clinical activity of thalidomide, initially as a monotherapy, in patients with androgen-independent prostate cancer; 18% of patients showed a >50% decrease in their PSA levels [7]. When treated with a combined therapy of docetaxel and thalidomide, half of patients showed a 50% decrease in PSA levels, compared with 37% of those treated with docetaxel alone [8]. Although the lengthening in median progression-free survival was not statistically significant, prolonged follow-up showed that median overall survival was significantly longer in patients treated with combined therapy, compared with those treated with docetaxel alone [8]. However, it was not clear how thalidomide conferred its survival benefit.

It has been shown that the circulating endothelial cell (CEC) level is increased in patients with a wide range of cancers, including prostate cancer [9], and that the measurement of CECs and circulating endothelial progenitors (CEPs) has the potential as a surrogate marker for monitoring antiangiogenic treatment drug activity [10]. Thus, we tested the effect on CECs and CEPs of a combined therapy of thalidomide and chemotherapy drugs in a human prostate cancer model, to help ascertain the molecular pharmacology of thalidomide.

MATERIALS AND METHODS

Thalidomide was obtained from Celgene (Summit, NJ, USA). Docetaxel (Taxotere®) was purchased from Aventis (Bridgewater, NJ, USA), estramustine (Emcyt®) from Pharmacia (Kalamazoo, MI, USA), and cisplatin was purchased from Bedford Laboratories (Bedford, OH, USA). Human prostate cancer cell line PC3 was purchased from ATCC (Manassas, VA, USA).

All animal experiments were done in accordance with institutional guidelines for animal welfare. PC3 cells (5 × 106) were injected s.c. into 5–6-week-old male severely combined immunodeficient mice.

For the docetaxel and thalidomide experiments, mice were randomized into four groups (five per group) when the tumour volume reached 50 mm3 for different treatments: vehicle only, thalidomide alone, docetaxel alone and the combined therapy (thalidomide and docetaxel). Mice received either vehicle (0.5% carboxymethyl-cellulose) or thalidomide (100 mg/kg), 5 days a week, administered by i.p. injection. Docetaxel (10 mg/kg) was administered as an i.v. bolus on day 2. Thalidomide was administered at the maximum tolerated doses determined in a pilot study, and docetaxel was administered at a dose that resulted in a comparable peak concentration to that achieved clinically [11].

For the thalidomide, docetaxel and estramustine studies, the mice were randomized into three groups (five per group) when the tumour volume reached 200–300 mm3. Mice received vehicle (0.5% carboxymethylcellulose), estramustine (4 mg/kg) or thalidomide plus estramustine (4 mg/kg), 5 days a week, administered by i.p. injection. Docetaxel (10 mg/kg) was administered in the latter two groups as an i.v. bolus on day 2 and day 9.

For the cisplatin experiments, mice were randomized into two groups (five per group) when the tumour volume reached 200 mm3 for treatments with either vehicle or cisplatin. Mice were administered on day 2 with either vehicle (saline) or cisplatin (5 mg/kg) as an i.v. bolus.

The treatment was continued for 17 days. Tumours were measured two or three times a week, and tumour volumes were calculated using the formula: π/6 × a × b × c; where a is the longest dimension of the tumour, b is the width and c is the depth. Blood samples were taken at three different time points (before the treatments, 4 days and 17 days into the treatments) for flow cytometry analysis.

For measuring CECs and CEPs, mice were bled from the retro-orbital sinus (surviving technique) or through intracardiac injection (lethal technique). Flow cytometry analysis was carried out, using a method adopted from a previously published protocol [12]. Mature CECs were defined as negative for the haematopoietic marker CD45, positive for the murine endothelial marker fetal liver kinase 1/vascular endothelial growth factor receptor 2 (Flk1) and negative for the progenitor marker CD117 (c-kit). CEPs were depicted by the positive expression of CD117. Briefly, after red cell lysis, 200 µL EDTA blood was stained with 6 µL anti-CD45-fluorescein isothiocyanate (FITC), 6 µL anti-Flk1-phycoerythrin, 5 µL anti-CD117-allophycocyanin (all antibodies were purchased from Pharmingen, San Diego, CA, USA) and 10 µL and 7-aminoactinomycin-D (7AAD; 0.1 mg/mL, Sigmal-Aldrich, St. Louis, MO, USA) and measured by a FACSCalibur (Pharmigen) using analysis gates designed to exclude dead cells, platelets and debris. After acquisition of at least 150 000 cells per sample, analyses were considered as informative when adequate numbers of events (i.e. >50, typically 100–200) were collected in the CEC and CEP enumeration gates.

The percentages of stained cells were determined and compared with negative controls stained with 6 µL anti-CD45-FITC antibody only. Positive staining was defined as being greater than nonspecific background staining and 7AAD used to enumerate viable, apoptotic and dead cells. The Wilcoxon rank-sum test was used to analyse all of the CEC or CEP data.

RESULTS

Docetaxel increased the apoptotic/dead CEC level and the combination of docetaxel and thalidomide increased its levels further.

Thalidomide treatment inhibited tumour growth by 46.5% on day 17, compared with vehicle control. Docetaxel induced tumour regression and the combination of docetaxel and thalidomide reduced tumour volume to 3.0% of the vehicle control (Fig. 1A).

Figure 1.

A, PC3 xenograft tumour growth in mice treated with vehicle control (ctrl), thalidomide alone (Tha), docetaxel alone (ctrl Doc) and a combination of docetaxel and thalidomide (Tha + Doc). B, Apoptotic/dead CEC (CD45–, Flk1+ and 7AAD+) levels before treatment, at 4 days and 17 days into the treatment. The mice were treated with vehicle control (Ctrl), thalidomide alone (Tha), docetaxel alone (Doc) and a combination of docetaxel and thalidomide (Tha + Doc). C, Viable CEC (CD45–, Flk1+ and 7AAD–) levels before treatment, at 4 days and 17 days into the treatment.

Docetaxel treatment increased the apoptotic/dead CEC level to 9.97 times compared with the control, 4 days into treatment (Fig. 1B). This is consistent with a previous report that taxanes induced CEC levels in patients after i.v. infusion [9]. The addition of thalidomide as a combined therapy further increased the apoptotic/dead CEC level to 13.48 times (P = 0.016), suggesting that thalidomide enhances the cytotoxicity of docetaxel against tumour vascular ECs. In addition, the increased apoptotic/dead CECs of the combined therapy at day 4 was correlated with its enhanced antitumour efficacy at day 17, showing the potential for the apoptotic/dead CEC level as a therapeutic biomarker for thalidomide in combined therapy with docetaxel. The docetaxel and thalidomide combined treatment also induced the viable CEC level to 2.75 times, compared with the controls at day 4 (P = 0.095) (Fig. 1C), suggesting a trend of increasing mature ECs derived from bone marrow and released into the blood circulation.

There was no statistically significant changes in CEP levels (data not shown), suggesting that the mobilization of bone marrow-derived EPs was not increased significantly, and hence the incorporation of these cells into tumour vasculature was not enhanced.

The combined treatment of docetaxel and estramustine increased the apoptotic/dead CEC level and the addition of thalidomide augmented the increase.

Docetaxel and estramustine inhibited tumour growth and the addition of thalidomide further decreased the mean tumour volume at the end of treatment by 59.5% (Fig. 2A).

Figure 2.

A, PC3 xenograft tumour growth in mice treated with vehicle control (ctrl), docetaxel and estramustine (Doc & Est) and a combination of docetaxel, estramustine and thalidomide (TDE). B, The apoptotic/dead CEC (CD45–, Flk1+ and 7AAD+) levels and viable (CD45–, Flk1+ and 7AAD–) levels after treatment. The mice were treated with vehicle control (ctrl), docetaxel and estramustine (DE) and a combination of docetaxel, estramustine and thalidomide (TDE).

The docetaxel and estramustine treatment increased the apoptotic/dead CEC level by 5.49 times, as compared with the vehicle controls (Fig. 2B). The addition of thalidomide in the combined therapy further increased the apoptotic/dead CEC level to 11.22 times (P = 0.095), showing the trend that thalidomide enhances the cytotoxicity of docetaxel and estramustine against tumour vascular ECs. Furthermore, this increased apoptotic/dead CEC level was correlated with the enhanced antitumour efficacy of the triple combined therapy, confirming that the apoptotic/dead CEC level can serve as a therapeutic biomarker for thalidomide in combined therapy with chemotherapy drugs. The CEP level analysis did not show any significant changes (data not shown), indicating that bone marrow-derived EPs were not increasingly released into blood circulation and subsequently incorporated into tumour blood vessels.

To exclude the possibility that the CEC changes are due to nonspecific cytotoxicity of the chemotherapy agent, we tested the effect of cisplatin treatment on CECs. Cisplatin is a widely used chemotherapy drug and has been shown not to have antiangiogenic activity [13]. On day 4 into the cisplatin treatment, there was no detectable variation in the apoptotic/dead CEC levels (Fig. 3B), indicating that it was not changed by nonspecific cytotoxicity of chemotherapy drugs. This is consistent with the negligent efficacy by cisplatin in this model (Fig. 3A), showing again the predictive value of the apoptotic/dead CEC level as a therapeutic marker.

Figure 3.

A, PC3 xenograft tumour growth in mice treated with vehicle control or cisplatin. B, The apoptotic/dead CEC (CD45–, Flk1+ and 7AAD+) levels before treatment and 4 days into treatment. The mice were treated with vehicle control or cisplatin.

DISCUSSION

Since 1971, various angiogenic regulators, both positive and negative, have been identified [14,15]. The discovery of angiogenic factors has led to the development of novel cancer therapies based on targeting a tumour’s vascular supply. A major advance in the field of antiangiogenic therapy occurred when the USA Food and Drug Administration approved Avastin (bevacizumab) in 2004, the first selective antiangiogenesis therapy approved for treatment of human cancer [16]. Because antiangiogenic agents do not target tumour cells directly, the efficacy of these drugs, especially when used as a single agent, often are difficult to detect using the traditional therapeutic endpoints utilized for cytotoxic regimens. Hence, a good therapeutic biomarker for antiangiognic drugs is urgently needed.

The CEC level has been tested for this purpose and has shown promise [10]. An increased level of CECs was found in patients with a wide range of cancers, such as lymphoma [17], acute myeloid leukaemia [18], breast cancer [19], head and neck cancer, RCC, ovarian cancer, colon cancer and prostate cancer [9], compared with healthy controls. In addition, the analysis of CEPs could provide a clue as to whether the treatment also affects bone marrow-derived ECs.

However, there are conflicting reports about the effects of drug treatment on CEC levels. Some drugs, such as bevacizumab [20] and ABT-751 [21], have been shown to decrease CEC levels in patients with cancer. By contrast, other antiangiogenic drugs, such as ZD6474 and ZD6126, have increased CEC levels, accompanied by a decrease in tumour microvessel density and preceded reduction in tumour volume in either a preclinical model [22] or in patients [23].

Many traditional chemotherapy drugs are also cytotoxic to ECs. In the case of taxanes (paclitaxel and docetaxel), ECs have been shown to be 10–100-fold more sensitive to these drugs than tumour cells [24]. Angiogenesis was blocked by both drugs primarily via inhibition of proliferation and differentiation, and induction of cell death. In fact, even 5 nm docetaxel can significantly inhibit microvessel growth in a rat aortic ring assay (data not shown). The present results showed that an increased level of apoptotic/dead CECs occurred shortly after the i.v. injection of docetaxel, confirming the capability of docetaxel to induce apoptosis of tumour EC death.

The addition of thalidomide further increased the apoptotic/dead CEC level, showing that thalidomide enhances the cytotoxicity of docetaxel against tumour vascular ECs. Similarly, the addition of thalidomide to chemotherapy treatment of docetaxel/estramustine showed the trend to further increasing the level of the apoptotic/dead CEC level, confirming the antiangiogenic function of thalidomide in combined therapies and the capability of the apoptotic/dead CEC level as a biomarker to detect it. However, the apoptotic/dead CEC levels decreased at a later time point, indicating that its change was a dynamic process and it is important to include multiple time points to clinical studies when studying the apoptotic/dead CEC level as a therapeutic biomarker.

Radiation therapy has also been shown to increase apoptotic CEC levels in patients [25], indicating that the CEC level has the potential as a surrogate marker for the combined therapy of antiangiogenic drugs and radiation therapy.

Antiangiogenic therapy is often thought to have low toxicity and could be potentially used in patients for long-term treatment. The present data indicated a correlation between the increased apoptotic/dead CEC levels on day 4 and treatment efficacy on day 17, suggesting it could be used as a marker at an early stage to predict the response to antiangiogenic therapy later. In a breast preclinical model, increased apoptotic CEC levels have also been shown to precede tumour inhibition, consistent with the present results [26]. Furthermore, increased CEC levels have been associated with a longer progression-free survival and an improved overall survival [26], further confirming its value as a prognostic marker.

ACKNOWLEDGEMENTS

We would like to acknowledge Dr Seth Steinberg from NCI for his statistical support and CCR ETIB Flow Cytometry Core Laboratory for their technical support.

CONFLICT OF INTEREST

None declared. Source of funding: Intramural NIH.

Ancillary