| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Ann Thorac Surg 2004;78:1042-1051
© 2004 The Society of Thoracic Surgeons
a Departments of Cardiothoracic Surgery Weill Medical College of Cornell University, New York, New York, USA
b Department of Genetic Medicine Weill Medical College of Cornell University, New York, New York, USA
c Department of Medicine, Weill Medical College of Cornell University, New York, New York, USA
Accepted for publication March 8, 2004.
* Address reprint requests to Dr Korst, Department of Cardiothoracic Surgery, M 404, Weill Medical College of Cornell University, 525 East 68th St, New York, NY, USA 10021
rjk2002{at}med.cornell.edu
Presented at the Fortieth Annual Meeting of The Society of Thoracic Surgeons, San Antonio, TX, Jan 26–28, 2004.
 |
Abstract |
|---|
 Top Abstract IntroductionMaterial and methodsResultsCommentDiscussionAcknowledgmentsReferences |
|---|
METHODS: A syngeneic murine mesothelioma flank tumor model (AB-12) was established in BALB/c mice. Separate adenovirus vectors expressing the cDNAs for human pigment epithelium-derived factor (AdPEDF) and a soluble form of the human vascular endothelial growth factor receptor-1 (Adsflt-1) were administered intratumorally. End points measured included tumor size, animal survival, and microvessel density using CD31 immunohistochemistry. An orthotopic model of mesothelioma was established by implanting AB-12 cells into the murine pleural cavity. Simultaneously, AdPEDF and Adsflt-1 were instilled intrapleurally and tumor burden and survival were recorded. The development of pulmonary emphysema was also assessed by calculating the mean linear intercept (a measure of interalveolar septal distance) in histologic lung sections from tumor-free mice after vector administration.
RESULTS: In the flank tumor model, the combination of AdPEDF and Adsflt-1 inhibited tumor growth, prolonged survival, and decreased microvessel density more profoundly compared with either AdPEDF or Adsflt-1 alone. In the orthotopic model, the combination was also more effective in prolonging survival. Intrapleural AdPEDF or Adsflt-1 did not increase the mean linear intercept compared with controls in tumor-free mice.
CONCLUSIONS: In this murine model, inhibiting multiple mechanisms of angiogenesis using two agents is a more effective antineoplastic strategy than using either agent alone. In addition, instillation of antiangiogenic gene transfer vectors into the pleural space does not result in histologic evidence of pulmonary emphysema.
|
Introduction |
|---|
|
TopAbstractIntroductionMaterial and methodsResultsCommentDiscussionAcknowledgmentsReferences |
|---|
| Dr Crystal discloses that he has a financial relationship with GenVec, Inc.
|
Malignant pleural mesothelioma (MPM) is an invasive tumor of the pleura afflicting approximately 2500 patients every year in the United States [1]. Conventional antineoplastic therapies (chemotherapy, radiotherapy, surgical resection) have had limited effects on survival from this disease, underscoring the need for additional and novel strategies [2]. In this regard, inhibition of tumor angiogenesis may prove worthy of investigation in patients with MPM, because serum levels of vascular endothelial growth factor (VEGF), an important proangiogenic mediator, are elevated in patients with MPM [3], and VEGF may serve as an autocrine growth factor for this disease [4].
Antiangiogenesis as an antineoplastic strategy is based on the observation that the growth of many types of solid tumors to larger than 0.5 cm requires neovascularization, a multistep process in which VEGF plays a pivotal role [5, 6]. Specifically, VEGF promotes not only proliferation and chemotaxis of endothelial cells, but also enhances vascular permeability, an all-important feature for blood vessel and subsequent tumor growth [7]. Vascular endothelial growth factor blockade, achieved by a variety of strategies, has proven effective in retarding tumor growth in preclinical models [8, 9]. One such approach has been the intratumoral injection of an adenovirus (Ad) gene transfer vector encoding the soluble extracellular domain of the human VEGF-1 receptor (Adsflt-1), which may function to sequester VEGF molecules in the tumor bed [10]. Although these preclinical results have been encouraging, the clinical use of VEGF blockade may be problematic, because data from some recent reports suggests that systemic block of VEGF receptor signaling results in the development of pulmonary emphysema in rodents [11].
Pigment epithelium-derived factor (PEDF) is a potent antiangiogenic molecule that actively induces apoptosis in endothelial cells by the Fas/Fas ligand pathway, while also inhibiting endothelial cell migration [12, 13]. Previous work from our laboratory has demonstrated that in vivo administration of an Ad vector encoding the human PEDF cDNA in preclinical models of thoracic malignancies inhibits tumor growth and vascularity, while prolonging animal survival [14].
Given that tumor angiogenesis is a multistep process, combined with the observation that VEGF blockade and intratumoral PEDF delivery may affect different components or mechanisms of this process, we hypothesized that local administration of a combination of both Adsflt-1 and AdPEDF may inhibit tumor growth more effectively than either strategy alone.
|
Material and methods |
|---|
|
TopAbstractIntroductionMaterial and methodsResultsCommentDiscussionAcknowledgmentsReferences |
|---|
Cell culture
The murine mesothelioma cell line AB-12 (H-2d; syngeneic to BALB/c mice) was kindly provided by Dr Jay Kolls of the Louisiana State University School of Medicine (New Orleans, LA). This cell line was originally generated by intrapleural implantation of asbestos fibers in BALB/c mice [15]. The AB-12 cell line was maintained in complete Dulbecco's modified Eagle's medium/F12 (10% fetal bovine serum, 100 µg/mL streptomycin, and 100 U/mL penicillin) at 5% CO2 and 37°C.
Adenovirus vectors
All Ad vectors used in this study were replication-deficient, E1-/E3- vectors based on the adenovirus serotype 5 genome. Adsflt-1 expresses a naturally occurring soluble, secreted form of the human flt-1 receptor, lacking the intracellular and transmembrane domains [10]. Transgene expression is under the control of the cytomegalovirus early-immediate promoter/enhancer [10]. AdPEDF is a similar vector, but contains an expression cassette with the human PEDF cDNA (GenVec Inc, Gaithersburg, MD) [16]. AdNull is a negative control vector that contains no transgene [17]. Ad vectors were propagated in human embryonic kidney cells (293 cells; American Type Culture Collection, Manassas, VA) and purified through two cesium chloride gradient ultracentrifugations as described previously [18, 19]. The viral particle concentration was determined by ultraviolet absorbance at 260 nm [20].
Tumor models and vector delivery
Flank tumors
AB-12 cells (106 cells in 100 µL phosphate-buffered saline [PBS]) were injected subcutaneously into the right flanks of BALB/c mice. Flank tumor size was assessed in situ every 2 to 3 days by measuring the largest perpendicular diameters using microcalipers, and recorded as an average tumor area (mm2). When the tumors reached 35 to 40 mm2 (day 7), intratumoral injection of either PBS, AdNull, AdPEDF, or Adsflt-1 (1011 particles for each vector, suspended in 100 µL PBS) was performed using a 31-gauge needle. An additional group of mice received both AdPEDF and Adsflt-1 (5 x 1010 particle units of each vector).
Orthotopic model
An orthotopic model for MPM was established by instilling AB-12 cells into the right pleural space of BALB/c mice. To perform intrapleural instillation, mice were anesthetized with an intraperitoneal injection of ketamine (100 mg/kg) and xylazine (10 mg/kg). The trachea was cannulated with a 22-gauge angiocatheter (Becton Dickinson, Franklin Lakes, NJ) and mechanical ventilation was initiated with a small animal ventilator (Harvard Apparatus, Holliston, MA). After making a 5- to 10-mm incision through the skin on the right chest, a right anterolateral thoracotomy was performed in the fourth intercostal space, followed by instillation of 5 x 105 AB-12 cells suspended in 100 µL PBS. The thorax was then closed by one pericostal suture and the skin reapproximated with an autoclip (MikRon Precision, Inc, Gardena, CA). Mechanical ventilation was discontinued as spontaneous respiration was observed and the tracheal cannula was removed. This technique led to grossly visible pleural tumor in all mice when assessed 10 days after intrapleural instillation (not shown).
Within this orthotopic model, vectors or PBS were administered simultaneously with the tumor cells into the pleural cavity. The rationale was that this approach represents a minimal, residual tumor model that mimics what might be left in the pleural space after conventional surgical resection. Experimental groups included PBS, AdNull, Adsflt-1, AdPEDF (1011 particles of each vector in 100 µL PBS), or the combination of Adsflt-1 and AdPEDF (5 x 1010 particles of each vector in a total of 100 µL PBS). The mice were monitored on a daily basis and survival was recorded.
Assessment of tumor burden and survival
In the flank tumor model, tumor size represented tumor burden. When the animals appeared moribund or the tumor growth exceeded 15 mm in the largest diameter, the mice were sacrificed and this time point was defined as the date of death for the survival analysis. In the orthotopic model, tumor burden was assessed by sacrificing all mice (CO2 inhalation) 14 days after tumor and vector instillation, and performing necropsies. In a blinded fashion, the thorax was widely opened and photographed. Tumor deposits were then excised and weighed. In the orthotopic model, a parallel experiment was conducted and survival was recorded.
Evaluation of microvessel density in the flank tumor model
To assess the affect of AdPEDF and Adsflt-1 on tumor vascularity, immunohistochemical staining for endothelial cells in established AB-12 flank tumors was performed after vector administration. Briefly, flank tumors were initiated and injected with vectors as described previously. Mice were then sacrificed and the tumors were resected 10 days after vector administration, embedded in Optimal Cutting Temperature compound (Sakura Finetek, Torrance, CA), snap frozen in a 2-methylbutane/dry ice bath, and prepared as 10-µm frozen sections. The frozen tumor sections were blocked with 5% goat serum (Sigma Chemical Co, St. Louis, MO) for 20 minutes, washed with PBS, and incubated with rat anti-mouse CD31 (PECAM-1) monoclonal antibody (Pharmingen; San Diego, CA) for 3 hours. The sections were then washed with PBS, incubated with biotin-conjugated goat antirat IgG antibody (Pharmingen) for 1 hour, washed with PBS, and incubated with avidin-horseradish peroxidase solution (Pharmingen) for 30 minutes. After further washing with PBS, the tumor sections were exposed to 3, 3'-diaminobenzidine chromagen substrate (Pharmingen) for 10 minutes, washed with PBS, and counterstained with hematoxylin. In a blinded fashion, digital images of the tumor sections were then obtained from a light microscope using the Metamorph image analysis software (Universal Imaging Corporation, Downington, PA), an automated system used to assess the quantity and morphology of stained cells [21, 22]. Five 20x fields were randomly selected and the computer software quantified the number of CD31-positive blood vessels in each field. Vessel density was expressed as number of vessels/µm2.
Assessment of pulmonary emphysema after intrapleural vector delivery in normal mice
Normal BALB/c mice underwent thoracotomy and vector instillation into the right pleural space as described previously, but without tumor instillation. Experimental groups included PBS, AdNull, Adsflt-1, and AdPEDF (1011 particles of each vector in 100 µL PBS). Mice were sacrificed 21 days after vector instillation, as this time point had been established as sufficient for the development of pulmonary emphysema induced by systemic VEGF receptor blockade in rodents [11]. Tracheal cannulation was performed using a 22-gauge angiocatheter (Becton Dickinson), and 4% paraformaldehyde was then infused into the lungs at a constant pressure of 10 cm H2O. The lungs were excised and fixed in 4% paraformaldehyde for 48 hours. The right lungs were then embedded in paraffin and 5-µm sections were prepared and stained with hematoxylin and eosin. To serve as a positive control, normal BALB/c mice underwent tracheal instillation of porcine pancreatic elastase (30 µg in 50 µL PBS; Elastin Products Company, Owensville, MO), followed by lung harvest, fixation, and staining 28 days later. To assess lung morphology for emphysematous changes, the mean linear intercept, a measure of interalveolar wall distance, was determined by light microscopy at 100x magnification in a blinded fashion [11, 21]. Six random 100x fields were selected from each animal and the mean linear intercept was calculated by dividing the total length of a line (in micrometers) drawn across the entire field by the total number of alveolar intercepts encountered along the length of the line [11, 23].
Statistical analysis
All data are reported as mean ± standard error. Statistical significance between the means was determined using analysis of variance (ANOVA) with the extended Tukey test for multiple comparisons. When the n for the groups was unequal, an ANOVA for unequal treatment groups followed by the extended Tukey test as modified by Dunnett was used [24]. Survival evaluation was performed using the Kaplan–Meier analysis (p value determined by log rank test). All p < 0.05 were considered statistically significant.
|
Results |
|---|
|
TopAbstractIntroductionMaterial and methodsResultsCommentDiscussionAcknowledgmentsReferences |
|---|
|
|
|
|
|
|
|
|
|
Comment |
|---|
|
TopAbstractIntroductionMaterial and methodsResultsCommentDiscussionAcknowledgmentsReferences |
|---|
Given that angiogenesis is a multistep process, and the observation that many antiangiogenesis strategies function to merely temporarily delay tumor growth, this study was based on the hypothesis that inhibiting multiple components of the angiogenesis cascade using multiple reagents acting by different mechanisms may have an "additive" effect on tumor growth inhibition, compared with using the individual strategies alone. Consistent with this hypothesis, the data demonstrated that in vivo Ad-mediated transfer of the combination of the human PEDF and sflt-1 cDNAs inhibited tumor growth and prolonged survival compared with either strategy alone in syngeneic murine mesothelioma. The enhanced efficacy of this approach was supported by the decreased blood vessel density found in the treated tumors. Finally, intrapleural delivery of antiangiogenic Ad gene transfer vectors did not appear to alter lung histomorphology with respect to the development of pulmonary emphysema.
Adsflt-1 and AdPEDF combined is more effective as an antineoplastic strategy compared with either vector alone
Although multiple secreted tumor cell products have been reported to induce angiogenesis, VEGF appears to play a central role in this process [7]. In this regard, VEGF blockade using the intratumoral administration of Adsflt-1 has been demonstrated previously to retard tumor growth in a murine model [10]. Three potential mechanisms for the antineoplastic effect of intratumoral production of the soluble flt-1 receptor include sequestering of free VEGF in the local tumor milieu; sflt-1–VEGF complexes may function as a dominant negative inhibitor of full-length flt-1 or flk-1/KDR receptors [10]; and VEGF blockade may interfere with the inhibitory effects of VEGF on dendritic cell maturation with subsequent enhancement of the antitumor immune response [25]. Functions of VEGF that may be inhibited after VEGF blockade by using Adsflt-1 may include the proliferation and chemotaxis of endothelial cells, as well as its effects on vascular permeability.
In contrast to sflt-1, which acts to inhibit a pro-angiogenic factor, PEDF has direct inhibitory effects on endothelial cells, resulting in an antiangiogenic effect shown to be more potent than angiostatin [12, 13, 26]. Although the details of the intracellular signaling pathways invoked by PEDF are currently not clear, an end result appears to be upregulation of Fas ligand by endothelial cells, resulting in apoptosis [13]. Pigment epithelium-derived factor has also been demonstrated to directly inhibit endothelial cell migration [12]. Human pigment epithelium-derived factor has been shown to profoundly inhibit tumor growth in murine models; however, similar to Adsflt-1, this effect seemed to be temporary [14].
The data from the present study demonstrate that the combination of these two reagents is more effective in retarding tumor growth than either strategy alone. This additive effect appears to occur because PEDF and sflt-1 inhibit angiogenesis by separate mechanisms, because the amount of total vector administered in each of the experimental groups was identical. Although the decrease in tumor burden seen when the antiangiogenic vectors were combined compared with either vector alone did not attain statistical significance in the orthotopic model, the survival advantage enjoyed by the combination group implies that this decrease is real, and that larger numbers of animals in the experimental groups may be required to demonstrate a statistically significant difference in tumor burden. However, larger experiments were not possible because five groups of animals were necessary and this model is labor intensive and time consuming, requiring approximately 30 minutes per procedure, per mouse. Despite the encouraging additive antitumor effect of the vector combination, all tumors eventually grew and resulted in mortality. This observation may be explained, in part, by the nature of antiangiogenic agents, which are not true cytotoxins. In addition, relevant in the present study is the finding that Ad vectors result in only transient transgene expression [27], which may allow tumors to grow once the quantity of transgene product declines.
CD31 was chosen as the marker for endothelial cells in the present study because it resulted in darker and what appeared to be more specific staining than CD34. Despite this, it is well recognized that CD31 (as well as other markers including CD34 and factor VIII) is not specific to vascular endothelium, being expressed on several types of circulating hematologic cells [28]. However, given the lack of a completely specific marker for tumor blood vessel endothelium, CD31 was used.
In the present study, vectors were administered at the same time as tumor initiation. This model was used for two reasons. First, this strategy seemed to mimic the clinical "minimal disease state" induced by surgical debulking in patients with MPM. Second, performing multiple thoracotomies on the mice was not practical due to marginally acceptable procedure-related mortality. Further investigation will need to determine the effects of this approach in the bulky disease state.
Intrapleural instillation of antiangiogenic adenovirus vectors does not induce pulmonary emphysema in normal mice
In addition to, and as a result of its proangiogenic properties, some recent lines of evidence suggest that VEGF signaling is necessary for maintenance of normal lung alveolar structures. Vascular endothelial growth factor is abundantly expressed in the lung [29, 30], and expression is reduced in the sputum of patients with emphysema [31]. Interestingly, treatment of normal rats for 3 weeks with the VEGF receptor blocker, 3-[(2,4-di-methylpyrrol-5-yl)methylidenyl]-indolin 2-one, causes alveolar cell apoptosis-dependent emphysema [11]. Given these observations, it is important that the effect of intrapleural instillation of antiangiogenic Ad vectors on lung morphology be determined. In the present study, intrapleural instillation of Adsflt-1 or AdPEDF had no effect on alveolar architecture 3 weeks after vector dosing. Although further studies will be necessary, these preliminary data suggest that intrapleural antiangiogenesis gene therapy may be safe in the clinical setting.
Clinical implications and potential limitations of antiangiogenesis gene therapy in patients with malignant pleural mesothelioma
Vascular endothelial growth factor signaling appears to be important for the progression and prognosis of MPM in the clinical setting [27]. Patients with MPM have significantly higher serum VEGF levels than patients with most other solid tumors [3], and VEGF expression correlates with microvessel density MPM, which may correlate with prognosis [32, 33]. Finally, VEGF also appears to serve as an autocrine growth factor for MPM, with mesothelioma cells both secreting and proliferating in response to VEGF [4]. Given these observations, antiangiogenesis represents a logical choice for clinical investigation for patients with this disease. Indeed, several molecular agents that block VEGF are currently under investigation in clinical trials [2].
Intrapleural gene therapy, as described in the present study, represents an attractive approach for antiangiogenesis in MPM, and may have advantages over protein administration for several reasons that may be important in a disease such as MPM, in which local failure after treatment is common. First, gene delivery to the local milieu allows for potentially long-term expression of the antiangiogenic factor(s). In the present study, relatively transient expression by Ad vectors was achieved; however, other gene transfer vectors, including adeno-associated virus, allow for much longer-term transgene expression [34]. Clearly studies with these other vectors that allow for more persistent transgene expression are warranted, including their effects on alveolar architecture. Second, gene transfer allows for higher local concentrations of the antiangiogenic factor(s), potentially abrogating the toxicity of systemically administered drugs. Finally, delivery of the gene transfer vector may be easily achieved at the time of surgical debulking procedures, including pleurectomy/decortication and extrapleural pneumonectomy.
Potential limitations of this approach clearly exist. First, the orthotopic model chosen for evaluation in this study was an attempt to mimic the "minimal disease state" created by surgical resection. Antiangiogenesis as a strategy for patients with bulky disease, however, may not be as efficacious given that these approaches are not inherently cytotoxic. Second, some growth patterns of solid tumors may not be dependent on angiogenesis. Examples include the bronchioloalveolar variant of non–small cell lung cancer, as well as some pulmonary metastases [35]. Whether antiangiogenesis represents a valid strategy for these types of tumors remains to be investigated.
|
Discussion |
|---|
|
TopAbstractIntroductionMaterial and methodsResultsCommentDiscussionAcknowledgmentsReferences |
|---|
DR KORST: We did not evaluate whether preexisting vessels were eliminated, or if this represented true inhibition of angiogenesis.
DR MARSHALL: In many antiangiogenic experiments, the decrease in blood vessels observed may be due to inhibition of neovascularization or destruction of preexisting vessels. Without additional studies, it is unclear which mechanism is leading to your results.
DR KORST: Yes, these are important questions.
DR THOMAS K. WADDELL (Toronto, Ontario, Canada): Bob, I would like to ask a question. You said that the orthotopic model was a model of postsurgical resection. Is that true in terms of the systemic uptake of these vectors? Presumably after you have exposed all of the raw tissues and delivered intrathoracic vector, you might obtain a much higher systemic uptake, in which case you might have systemic effects, such as emphysema. And in particular, as a positive control, did you try a systemic administration of these vectors rather than the elastase as a positive control?
DR KORST: That is another good point. No data exist on whether a pleural stripping or decortication and intrapleural instillation of these vectors would make a difference on the systemic uptake. This point will need to be studied. With regard to the positive control, an intravenous injection might be a better positive control in terms of the systemic effects of the transgenes on alveolar architecture, but this has yet to be studied. The elastase was used mainly to verify the measuring system that we used.
DR BENNY WEKSLER (Rio de Janeiro, Brazil): Bob, how did you determine what was a full dose? Did you do a dose–response curve previously?
DR KORST: The doses had been established in previous studies. For the purposes of this presentation, I wanted to make the point that we were using half dose and full dose of the vectors and that the total vector quantity was the same for all of the groups.
DR THOMAS A. D'AMICO (Durham, NC): Bob, you demonstrated pretty clearly that we do not expect emphysema to result, but what types of complications might result from either of the two antiangiogenic strategies? We know what happens with lung cancer patients if they have central squamous cell carcinoma. What do you expect to happen with mesothelioma, what complications? Secondly, given the two mechanisms, do you expect any specific class of chemotherapy to be synergistic with your model?
DR KORST: Clinically one would have to look at local and systemic potential complications. Local complications could include wound healing problems, particularly with regard to a bronchial stump if a major lung resection were to be performed. Systemically, we would need to determine how much of the vector is absorbed and what the systemic levels of the transgene products are; we have not made these calculations yet.
|
Acknowledgments |
|---|
|
TopAbstractIntroductionMaterial and methodsResultsCommentDiscussionAcknowledgmentsReferences |
|---|
|
References |
|---|
|
TopAbstractIntroductionMaterial and methodsResultsCommentDiscussionAcknowledgmentsReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  E. T.H. Ek, C. R. Dass, and P. F.M. Choong Pigment epithelium-derived factor: a multimodal tumor inhibitor. Mol. Cancer Ther., July 1, 2006; 5(7): 1641 - 1646. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 B. W. Robinson and R. A. Lake Advances in malignant mesothelioma. N. Engl. J. Med., October 13, 2005; 353(15): 1591 - 1603. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| ANN THORAC SURG | ASIAN CARDIOVASC THORAC ANN | EUR J CARDIOTHORAC SURG |
| J THORAC CARDIOVASC SURG | ICVTS | ALL CTSNet JOURNALS |
, n = 8; 5 x 1010 particles of each vector); Adsflt-1 alone (
, n = 8); AdPEDF alone (
, n = 8); a negative control vector (AdNull, 
, n = 8); or phosphate-buffered saline (PBS, 
, n = 8). All animals received intratumoral injections and the total vector dose per mouse was 1011 particles suspended in 100 µL PBS for all groups in which vectors were administered. The tumor area was assessed every 2 to 3 days with microcalipers, and statistical analysis was performed 25 days after tumor initiation. The mice were sacrificed when they appeared moribund or when the largest tumor diameter reached 15 mm. Arrows indicate time point of vector (or PBS) administration. (A) Tumor area for AB-12 mesothelioma flank tumors in BALB/c mice. The data points represent the mean tumor area ± SEM. (B) Survival for BALB/c mice bearing AB-12 mesothelioma flank tumors. (AdPEDF = adenovirus vector encoding the human pigment epithelium-derived factor; Adsflt-1 = adenovirus vector encoding the human soluble vascular endothelial growth factor receptor-1.)
