| HOME | HELP | CONTACT US | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Cancer Research UK, Department of Medical Oncology, Manchester, United Kingdom
Correspondence: S. Rehman, Cancer Research UK, Department of Medical Oncology, Christie Hospital NHS Trust, Wilmslow Road, Manchester, M20 4BX, United Kingdom. Telephone: 0161-446-3606; Fax: 0161-446-3461; e-mail: Shazza.Rehman{at}christie-tr.nwest.nhs.uk
Access and take the CME test online and receive 1 hour of AMA PRA category 1 credit at CME.TheOncologist.com
LEARNING OBJECTIVES
Top
Learning Objectives
Abstract
Introduction
Imaging Strategies in Clinical...
Discussion
Summary
References
After completing this course, the reader will be able to:
|
ABSTRACT |
|---|
|
TopLearning ObjectivesAbstractIntroductionImaging Strategies in Clinical...DiscussionSummaryReferences |
|---|
There is increasing awareness of the need to obtain evidence of drug activity through the use of surrogate markers of the biologic mechanism of action during early clinical trials, in addition to determining the pharmacokinetics, toxicity profile, and maximum-tolerated dose. One of the major impediments to the rapid development of antiangiogenic agents in the past has been the lack of validated assays capable of measuring an antiangiogenic effect directly in patients. Recently, dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI) has emerged as a useful technique for noninvasive imaging of tumor vasculature in preclinical and clinical models.
The problem of tumor heterogeneity remains to be addressed. The major challenge is the standardization of the technique worldwide for the purpose of early clinical studies that are likely to be multicenter. Convincing data on correlations between changes observed through molecular imaging and changes in tumor angiogenesis, and hence tumor biology, are still lacking. Whether this would translate into a survival advantage remains to be seen.
The ultimate test of the surrogate biological end points determined by molecular imaging will occur in randomized phase III trials. Results of the first randomized trial that showed a survival advantage in favor of antiangiogenic agents were released at the American Society of Clinical Oncology meeting in 2003. There it was reported that the combination of 5-fluorouracil, leucovorin, and irinotecan (Camptosar®; Pfizer Pharmaceuticals; New York, NY) with anti–vascular endothelial growth factor antibody (bevacizumab—Avastin®; Genentech, Inc.; South San Francisco, CA) was superior to the chemotherapy regimen alone when used to treat patients with metastatic colorectal cancer. However, until further phase III clinical trials confirm these results, surrogate end points of clinical efficacy of the newer agents are urgently needed so that development of ineffective drugs can be halted early. This review briefly discusses the role of molecular imaging in general, and DCE-MRI in particular, in relation to treatment with antiangiogenic agents and highlights some of the difficulties encountered in this area.
Key Words. Molecular imaging • Antiangiogenic agents • DCE-MRI • Surrogate end points
|
INTRODUCTION |
|---|
|
TopLearning ObjectivesAbstractIntroductionImaging Strategies in Clinical...DiscussionSummaryReferences |
|---|
Clinical studies have demonstrated a clear correlation between angiogenic factors and prognosis in a range of tumors, including breast cancer, gastric cancer, colorectal cancer, non-small cell lung cancer, and melanoma [4]. There is now a better understanding of the molecular pathways underlying angiogenesis [5, 6], which has led to the identification of a range of targets and candidate drugs [7–11]. Pharmacologic and toxicologic studies have demonstrated that most of the tested antiangiogenic compounds are safe and selectively target tumor vessels [2, 6, 11]. Furthermore, clinical dose-finding studies have shown some tumor responses with first-generation angiogenesis inhibitors, especially those that target the vascular endothelial growth factor (VEGF) pathway [4]. A recent randomized study in metastatic colon cancer has shown for the first time that bevacizumab (Avastin®; Genentech, Inc.; South San Francisco, CA), a monoclonal anti-VEGF antibody, in combination with standard chemotherapy (5-fluorouracil [5-FU], leucovorin, and irinotecan [Camptosar®; Pfizer Pharmaceuticals; New York]) was associated with a median survival time of 20.3 months while patients treated with chemotherapy alone survived for a median of 15.6 months [12]. Initial reports suggested that the experimental arm was associated with a slightly greater risk for gastrointestinal perforation in patients who had not undergone surgical resection of the primary disease. This is a particularly important result, as it supports the antiangiogenic strategy for the first time in a randomized trial setting.
More than a dozen endogenous proteins that act as positive regulators or activators of tumor angiogenesis have been identified. These include VEGF, basic fibroblast growth factor (bFGF), tumor necrosis factor alpha (TNF-
), angiopoietin-1 and angiopoietin-2, interleukin-8 (IL-8), and platelet-derived growth factor beta (PDGF-ß) [13, 14]. There are endogenous angiogenic inhibitors as well, which include angiostatin, endostatin, interferon-, and interferon-ß [13, 14]. The factor that determines whether the angiogenic switch is on or off is the balance of angiogenic activators and inhibitors [5]. It also depends on the presence or absence of receptors.
Morgan et al. reported dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI) data from PTK787/ZK22584 (Novartis; East Hanover, NJ), an inhibitor of the VEGF receptor. These data suggest that there is a dose-response relationship that can be shown through DCE-MRI, but the raw data show that considerable heterogeneity is still a problem [15].
Antiangiogenic and Antivascular Agents
A number of pharmacological targets have been identified in the tumor vasculature. These include the signaling pathways responsible for the growth of new blood vessels together with factors required for the survival and structural integrity of immature endothelium. It is important to distinguish between antivascular agents, which destroy established vessels, and antiangiogenic agents, which prevent the formation of new blood vessels. Because there is co-option of existing vasculature around a tumor, it may be inferred that one would need to destroy existing vasculature as well as prevent new vessel formation. Hence antivascular therapies might be given with antiangiogenic therapies.
The First International Conference on Vascular Targeting focused on vascular targeting agents (VTAs) [16]. VTAs cause a rapid shutdown in the blood supply to the tumor such that it is deprived of oxygen and nutrients. The VTAs that have been evaluated in early clinical studies include the antitubulin drugs, combrestastatin A4 phosphate (CA4P) [17], ZD6126 [16], and 5,6-dimethylxanthenone-4-acetic acid (DMXAA) [18]. These VTAs generally lack the bone marrow and gastrointestinal toxicities associated with cytotoxic chemotherapy (except for vinca alkaloids, which can cause bowel problems), although many do cause neuropathy. As a marker of biological effect, blood flow reductions in tumors treated with these agents have been measured using MRI or positron emission tomography (PET), and single-agent clinical activity has been demonstrated [18, 19]. Most of these trials looked at changes in a number of vascular parameters, including vascular permeability determined by MRI, but the interpretation of this parameter is difficult when vascular shutdown is seen. These agents are now being evaluated in combined-modality treatments. This review, however, focuses mainly on antiangiogenic agents.
Potential advantages of antiangiogenic agents over conventional cytotoxic chemotherapy regimens include easy access to targets within the vasculature, independence of tumor cell resistance mechanisms (since the cancer cells themselves are not targeted), and the broad applicability of antiangiogenic therapy to many tumor types [20]. In addition, as physiological angiogenesis in the adult is only seen in pregnancy and menstruation, antiangiogenic agents should have acceptable toxicities if used appropriately [21].
Dose Optimization of Antiangiogenic Drugs
Antiangiogenic compounds are likely to be cytostatic rather than cytocidal, and thus, tumor stabilization rather than response is a more probable end point. The traditional end point of toxicity may not be relevant in assessing biologic agents, as the dose-response relationship may not be the same as those seen for chemotherapeutic agents. A lack of knowledge on predictive markers for treatment outcome in early clinical studies confounds the selection of the optimum dose and schedule. Although relatively nontoxic, adverse events have been reported in some trials, and other studies have shown severe toxicity for certain subgroups of patients. For instance, fatal pulmonary hemorrhage was seen in some patients with non-small cell lung cancer who were treated with carboplatin (Paraplatin®; Bristol-Myers Squibb; Princeton, NJ), paclitaxel (Taxol®; Bristol-Myers Squibb), and an anti-VEGF monoclonal antibody [22]. However, all the fatal cases seemed to be tumor related, originating from centrally located pulmonary tumors of squamous cell histologies. Thus this trial has later continued in patients with non-squamous cell histologies. In a different drug combination study, when the VEGF receptor 2 inhibitor SU5416 was administered with cisplatin (Platinol®; Bristol-Myers Squibb) and gemcitabine (Gemzar®; Eli Lilly and Company; Indianapolis, IN), a greater than expected number of thromboembolic events, including transient ischemic attacks, cerebrovascular accidents, and deep vein thrombosis were seen [23]. These data suggest that the administration of the maximum-tolerated dose of antiangiogenic drugs with other agents could be associated with substantial toxicity, which was not seen when the agents were administered alone. However, there are several arguments for using antiangiogenic therapy in combination with other drugs. Firstly, the profusion of angiogenic factors that can be produced by the tumors suggests that the inhibition of angiogenesis may require the combined action of several inhibitors. Secondly, traditional chemotherapeutic agents also exert antiangiogenic effects of their own, and these effects can be potentiated by agents such as antibodies to VEGF, which can also be a reason for potentiated toxicity in combination. Thus, it is important to define the minimum and optimum biologically active doses of the antiangiogenic agents in early clinical evaluation, and if there are subgroups of patients who are more prone to toxicity than others. The most optimal sequences of drug combinations should also be addressed in prospective trials (concommitant treatment versus sequential treatment).
There are approximately 75 antiangiogenic agents in clinical trials at present, some of which are shown in Table 1
, mostly in phase I or phase II testing. At least 12 of these agents have entered or completed phase III evaluation.
|
|
IMAGING STRATEGIES IN CLINICAL USE |
|---|
|
TopLearning ObjectivesAbstractIntroductionImaging Strategies in Clinical...DiscussionSummaryReferences |
|---|
DCE-MRI
Introduction
DCE-MRI can monitor the effectiveness of a variety of treatments, including chemotherapy, hormonal manipulation, and radiotherapy, and novel therapeutic approaches, including antiangiogenic drugs [25]. A large number of antiangiogenic drugs have been targeted against the angiogenic cytokine VEGF. This cytokine is a principal mediator of vascular permeability, which has been measured by DCE-MRI and used as a pharmacodynamic end point for the development of these new compounds. DCE-MRI provides a powerful tool for the rapid evaluation of the acute pharmacodynamic effect of these newer agents in clinical trials, most notably in the case of mechanisms that affect tumor perfusion. Evidence is mounting that DCE-MRI measurements correlate with immunohistochemical surrogates of tumor angiogenesis [26].
In DCE-MRI, imaging is performed at the same time as contrast administration, and it has been used in a large number of studies to examine tumor vasculature. It is a noninvasive technique that yields parameters related to tissue perfusion (T2 methods) and permeability (T1 methods) [27–32]. The paramagnetic contrast agent gadopentetate dimeglumine (Gd-DTPA) is injected as a rapid i.v. bolus and, as it passes through tissues, it diffuses out of the blood vessels into the extravascular extracellular space. The signal intensity on T1-weighted images increases as the concentration of Gd-DTPA in the extracellular space increases [31–34]. These changes in signal intensity are recorded by serial images acquired before, during, and after the injection. Relative changes in semiquantitative parameters, such as the maximum gradient of the signal-intensity-time curve, the maximum increase in signal intensity normalized to baseline signal intensity (enhancement), and the area under the initial part of the curve, can be examined and are indirectly related to changes in the physiologic end points of interest: tissue perfusion, vascular permeability, and vessel surface area [27, 30, 35–37]. A couple of reproducibility studies of this technique in humans have been performed, in glioma and in liver tumors, which defined the limits of change in these DCE-MRI parameters that might occur spontaneously between two examinations performed 5 days apart [38–40]. A number of significant issues were revealed during the course of that work, and these issues are reviewed here. The main points include the design of the DCE-MRI acquisition and analysis protocol, the complexity introduced by tumor heterogeneity, the reproducibility of the measurement, and the validity of the measurement as a surrogate for the parameter being investigated.
Design of the DCE-MRI Acquisition and Analysis Protocol
A number of DCE-MRI methodologies have been developed. The more complex algorithms can provide unique insights into biology but are less easy to compare among institutions. Thus, for phase I trials, where one or two institutions participate, complex analytical techniques can be used. For multicenter trials, less-demanding technologies should be used, but it is essential that adequate standardization and quality control are performed among centers if comparable data are to be obtained.
Typical trials of antivascular and antiangiogenic agents require comparison of DCE-MRI before and after therapy over periods of hours to weeks. In order to obtain valid estimates of tumor response, it is important that the same region of tumor is imaged on each patient visit. For this reason, three-dimensional (3D) protocols, which provide data over the whole tumor, are preferable to single-slice protocols. However, current MRI technology does not permit "fast" protocols in 3D, so investigators must choose between a more complete assessment of perfusion from "fast" 2D protocols and a greater confidence from "slow" 3D protocols in which follow-up scans are truly comparable. Furthermore, if imaging is to be performed over a few weeks, then it is appropriate to study tumors that are not growing too rapidly, as in these cases, comparison of spatial data from different time points becomes very difficult.
Whichever parametric analysis is used, it is vital to achieve standardization among participating institutions. Thus, the algorithms employed, particularly for the primary end point, are better declared prospectively, and the use of retrospectively applied threshold values should be avoided. On the other hand, additional secondary end points, derived from different analyses of the data, may also be employed to extract the most value from a trial.
Ideally, the analysis should lead to an end point with a link to local tumor biology (for instance, vascular permeability, flow, etc.). Most effective anticancer therapies will, hopefully, eventually reduce the delivery of contrast agent to the tumor. Increasing analytical complexity is only of value if it provides greater statistical power, greater certainty about the underlying physiological changes, or greater comparability among studies performed at different sites. There is no one optimal approach, and a technique that offers excellent statistical power and cross-center agreement for one type of tumor may be of little value in other tumors.
Spatial Heterogeneity Within the Tumor
Tumor heterogeneity poses considerable difficulties within and among patients and has largely not been taken into account when incorporating imaging strategies into early-phase clinical trials. This contrasts with the standardized measurements of normal organ function that are entry criteria for most phase I trials. The consequence of this is that we are able to see dose-response effects with regard to toxicity but not with regard to biological or pharmacodynamic phenomena.
These problems have manifested themselves in other clinical trials where imaging has been used to study vascular parameters [17, 18, 41]. At present, the best interpretation of the data is that they show a threshold effect rather than a dose-response effect. In other words, researchers are reporting that, as higher doses are given, one eventually encounters a dose at which the majority of patients manifest the desired change in the DCE-MRI end point. Below that dose, the majority of MRI studies do not show the anticipated change. In one phase I trial evaluation of an anti-VEGF antibody, the lowest dose level, 0.3 mg/kg, was significantly inferior, in terms of its effect on the permeability endothelial surface area product per unit volume of tissue (Ktrans), to the three higher doses (1, 3, and 10 mg/kg). However, there was no dose-response relationship in the three higher dose levels. Similarly, in the evaluation of CA4P, the initial report of the data suggested a threshold effect at a minimum of 52 mg/m2 [17]. Thus, to date, we have developed methodologies that appear to identify the minimum effective dose rather than the optimum biologically effective dose. Clearly, if we are to use DCE-MRI to guide us in the development of antiangiogenic and antivascular agents, we need to modify clinical trial design so that the degree of heterogeneity is addressed. This might then allow us to identify dose-response curves and, thus, the optimum biologically active dose for further study. The first evidence of a dose-response relationship employing MRI was reported by Morgan et al. [15].
Reproducibility
If there is to be reliance upon MRI as a pharmacodynamic end point, then it is critical to establish the reproducibility of the measurement. For example, if an experimental agent caused a 20% decline in a vascular parameter measured by MRI but the day-to-day variation in that parameter was 25%, then it would not be possible to say whether that drug was active. Thus, current recommendations are that reproducibility studies for each patient are built into any study that relies upon DCE-MRI as a pharmacodynamic end point. In practice, this has meant performing two baseline studies before administering the drug.
Reproducibility studies have been performed for a number of tumor types. Data show that these studies can be most easily performed in glioma, in which the tumor is contained in the rigid skull, while studies in the upper abdomen and thorax have been compromised, to some extent, by respiratory motion artifacts. The reproducibility of these studies in the abdomen is lower than those of studies in the brain. When liver metastases were studied over an 8-hour period using a first-pass method to measure Ktrans, the coefficient of variation was 11% and the percentage change required to prove drug activity in that setting was 15% [40]. Currently, techniques are under development to measure vascular parameters through DCE-MRI in the thorax, the major hurdle of which is the movement associated with breathing and the heart beat. Therefore, lung nodules and liver metastases can be particularly difficult to image reproducibly.
Validity
It is critical to assess the validity of these parameters as surrogates for the desired biomarker. A number of studies have attempted to make these comparisons, and the data are summarized in Table 2. Although some studies have not confirmed that there is a statistical association between particular MRI parameters and the vascular or tumor measurement, there are a sufficient number that support the provisional incorporation of MRI into the early clinical evaluation of antiangiogenic compounds.
|
PET Scanning
PET is a sensitive and quantitative technique that can be used to monitor the pharmacokinetics and pharmacodynamics of drugs when radiolabeled with positron-emitting radioisotopes [19]. This is of particular relevance in assessing newer molecular-targeted therapy where conventional evaluation criteria like maximum-tolerated dose and tumor shrinkage may be inappropriate. PET has been applied to a wide range of drugs to demonstrate activity in vivo including standard chemotherapy, such as 5-FU [42], and antiangiogenic and antivascular therapy [43, 44]. The 5-FU study looked at drug distribution, that is, in situ pharmacokinetics, while the other studies looked at pharmacodynamic effects. This technique can be used to label drugs and then work out intratumoral pharmacokinetics and also to develop pharmacodynamic markers, for example, 18F-labeled fluorodeoxyglucose (18FDG), or 15O-labeled water (H215O) to measure changes in flow.
PET is being used as a physiological imaging modality to evaluate subclinical tumor response to treatment. Some studies have looked at changes in 18FDG uptake as a pharmacodynamic end point after treatment of gliomas with single-agent temozolomide (Temodar®; Schering-Plough Corporation; Kenilworth, NJ) [45] and after combination chemotherapy in pancreatic cancer [46]. The European Organization for Research and Treatment of Cancer (EORTC) PET study group has defined guidelines for the use of 18FDG-PET in response assessment in oncology [47].
PET has been used in some studies to assess tumor blood flow with H215O and tumor metabolism with 18FDG as biologic end points of response to antiangiogenic agents [41]. PET has also demonstrated the effects of an antivascular agent, CA4P, by measuring tumor and normal tissue perfusion and blood volume [44]. That phase I trial showed a 30%–60% temporary reduction in tumor blood flow at the higher dose level, demonstrating proof of principle. This type of pharmacodynamic assessment could potentially reduce the time involved in early clinical trials of antiangiogenic agents by identifying doses that achieve critical biological changes, for example to blood flow [48].
The use of H215O for PET imaging offers several properties that are desirable for the measurement of blood flow. It is freely diffusible, has a short half-life of 2 minutes, and has favorable dosimetric properties [49]. This technique has been used to measure blood flow in several tumor types, including breast cancer [50, 51] and brain tumors [52, 53]. However, there are potential limitations. In small tumors, partial volume effects may be significant if the tumor size is less than twice the resolution of the scanner [49]. Second, there is a phenomenon called "spill over" or "spill in" of counts from surrounding structures with high blood flow, such as the heart and the aorta, or within areas of relatively high flow, such as the liver [49], thereby limiting the use of PET in the lung, liver, and mediastinum. Further issues are that many imaging modalities, such as CT and PET, use ionizing radiation, limiting the number of studies that can be performed. Although MRI is less quantitative than PET, the limitation concerning ionizing radiation does not apply to MRI. Furthermore, many PET isotopes are short lived, requiring synthesis of the relevant compound either at the patient’s bedside, as in the case of H215O, or within a day of administration.
Tumors may not have a uniform exchange of water between blood and tissue. Necrotic areas within tumors may have a poor exchange between blood and tissue and a lower volume of distribution of tracer. In some cases (for instance, with very high flow rates), measuring blood flow by PET underestimates this parameter. The heterogeneity of delivery of drugs to solid tumors may lead to variability in the results obtained from PET and other imaging modalities. Factors such as nonuniform distribution of blood vessels within tumors, high interstitial pressure within tumors, and decreased diffusibility of large molecules [54] may all make the interpretation of imaging data challenging. These techniques are extremely complicated and, therefore, are best used in phase I drug development programs.
In addition to existing technologies, new PET probes are under development, and coregistration methodologies will allow us to relate DCE-MRI parameters to other tumor measurements, principally through PET. These include measurements of tumor metabolism (18FDG), tumor DNA synthesis (125IUdR iodo-2'-deoxyuridine), and tumor apoptosis (annexin V) [55]. The next critical step is to work out how to relate the images so that, for instance, one could determine the extent to which drug distribution (positron-emitting isotope-labeled drug) influences changes in the vascular parameter (DCE-MRI) and vice versa. Certainly, the software now exists to coregister these studies, and data will emerge shortly.
A further complication that applies to both PET and MRI is that quantification, particularly of PET, requires knowledge of the arterial input into an imaged area. In the case of PET, this can be estimated through arterial cannulation, while MRI arterial input functions are normally related either to large vessels or to standardized data.
Dynamic CT
Dynamic or functional CT can be readily incorporated into routine conventional CT examinations, and the physiological parameters obtained can provide an in vivo marker of angiogenesis in tumors. Using this technique, it is possible to determine absolute values for tissue perfusion, relative blood volume, capillary permeability, and leakage. These parameters provide physiological correlates for the microscopic changes that occur with tumor angiogenesis [56, 57]. Tumor angiogenesis is characterized morphologically by increased numbers of small blood vessels. These microvessels are too small to image directly (<0.1 mm), but their increased density translates in vivo to increased tumor perfusion and blood volume. Dynamic CT has been used by various investigators to evaluate tumor microvessel density (MVD) [58].
Dynamic CT is simple, widely available, and reproducible [59] and has been validated against H215O-PET [60]. Quantification is simpler than for MRI, as the relationship between signal and contrast concentration is much more linear than that seen with MRI, although the sensitivity is less [57]. The incorporation of CT scanners into PET scanners [61] has also led to an increased use of dynamic CT in the evaluation of antiangiogenic and antivascular compounds.
The problem is that early clinical studies of antiangiogenic compounds require multiple imaging assessments of the tumor. As contrast-enhanced dynamic CT necessarily uses ionizing radiation, there is a limit to the number of studies that can be performed in any one patient [57]. As yet, reduction in tumor perfusion by antiangiogenic agents has not been demonstrated by dynamic CT in clinical studies. The main reason for this has probably been the lack of commercially available software to perform the more precise quantitative analyses involved in calculating perfusion, blood volume, and capillary permeability. This situation is likely to improve with the rapid development of CT software packages. Furthermore, with the advent of multislice CT systems, the limitations of the single-slice system will, hopefully, be overcome. In future, 3D assessment of spiral CT may find a role in evaluating angiogenesis and antiangiogenic therapy [62]. At present, there are few data from studies employing this technique. It may be possible to label monoclonal antibodies to VEGF and image in that way. The validation of functional CT parameters as markers of angiogenesis is currently under way at several institutions.
Contrast-Enhanced Ultrasound
Ultrasound is one of the most widely used imaging modalities (one in three imaging tests worldwide are ultrasound scans) [63] and also the most rapidly evolving technology. While the real-time nature of ultrasound has always lent itself to functional assessment, recent developments have dramatically improved the quality of conventional and Doppler ultrasound. In addition, new quantitative approaches, 3D scanning methods, and the increasing availability of microbubble contrast agents all open exciting new avenues for functional ultrasound imaging. There is emerging evidence that these new methods can give both direct and indirect information on the state of the tumor neovascularity and its response to therapy.
Conventional Doppler imaging is able to directly image flow in vessels down to approximately the millimeter level. It is, thus, best seen as a tool for imaging the macrocirculation, rather than the microcirculation. Recently, several manufacturers have developed commercial real-time ultrasound systems, using ingenious signal processing methods, which can scan in Doppler modes at between 10 MHz and 20 MHz with good tissue penetration. Such systems can detect flow in submillimeter-sized vessels with relative ease, but it seems unlikely that they will be able to image much smaller than approximately 0.1 mm. The principle has been extended to experimental Doppler systems able to perform at frequencies in the 20–100 MHz range, and these have been shown to be able to detect flow in vessels as small as 10–20 mm [64]. Although of potential value in experimental systems (e.g., superficial murine tumor models), the poor penetration of such systems means they are unlikely to be of value in human scanning, apart from possible skin or ocular applications.
Several researchers have studied the relationship between histological indices of angiogenesis, for example, MVD, and indices derived from Doppler ultrasound. Current Doppler methods often perform relatively poorly when directly correlated with measures of tumor angiogenesis, such as MVD [65]. This is potentially attributable to sampling errors in heterogeneous tumors. A much more promising clinical application at the moment appears to be imaging the response of a tumor blood supply to cancer therapy. Estimates based on the fractional vascular volume, a surrogate of quantitative power Doppler signals, show particular promise. A study by Gee et al. using a murine model evaluated the effect of angiogenesis inhibition with IL-12 on a tumor strain, K1735, known to be sensitive and a variant, K1735.N23, engineered to be unresponsive [66]. All K1735 tumors showed decline in power of Doppler signals by 3 weeks of therapy, and this was associated with a reduction in vessel density. The unresponsive K1735.N23 variants, on the other hand, showed a consistent increase in Doppler signals.
Ultrasound, despite its popularity, wide availability, and low cost, has lagged behind other modalities such as MRI and PET in the functional assessment of cancer treatments. This is rapidly changing, with a dramatic improvement in scanner technology, the availability of quantitative methods, and now, microbubbles, which can be used as tracers for functional studies. A number of useful fractional methods are already starting to find increasing applications.
Although ultrasound media have been developed for the assessment of vascularity, these studies have been hampered by the operator dependency of the study and the difficulty in obtaining images when flow rates are very low. This makes measurements of reproducibility difficult, which obscures determination of drug effects. Also, depth of penetration is poor; therefore, organs such as lungs and brain are inaccessible [67].
|
DISCUSSION |
|---|
|
TopLearning ObjectivesAbstractIntroductionImaging Strategies in Clinical...DiscussionSummaryReferences |
|---|
The objectives of drug development in oncology include the determination of delivery of drug to the site of the tumor with minimum exposure to the normal tissues and, hence, the achievement of maximum therapeutic benefit. This requires accurate monitoring of drug pharmacology, including pharmacokinetics (absorption, distribution, metabolism, and elimination) and pharmacodynamics (tumor response, enzyme induction/inhibition receptor binding, tolerance). At present, these parameters are measured by analysis of blood and urine samples and, sometimes, using biopsy specimens of relevant tissues. While this strategy is suitable for most chemotherapy drugs, plasma monitoring may be irrelevant with the next generation of therapies such as antiangiogenics, where treatment is targeted against specific tissues.
Conventional imaging modalities for tumor response assessment rely on changes in tumor volume and mass or size that may not be apparent for several months after treatment and cannot easily distinguish between necrotic cells and viable, proliferating cancer cells. Early clinical trials have indicated that conventional imaging strategies are not suitable for monitoring the effects of antiangiogenic agents [24]. Most of these agents are not directly cytotoxic to tumor cells and produce disease stabilization rather than tumor regression. Thus, evaluation of tumor size alone is ineffective. Nonimaging approaches, such as measurement of growth factors in serum or urine, are yet to be fully validated, and serial tumor biopsies with determination of MVD are invasive, prone to sampling error, and impractical for a large number of patients. Thus, imaging strategies are required that are capable of monitoring tumor angiogenesis in vivo in a sensitive and specific manner.
The area where molecular imaging is helping is in the demonstration of the "proof of principle" by in vivo pharmacokinetic and pharmacodynamic measurements of antiangiogenic agents. For example, using functional imaging, it has been possible to image a therapeutic target in tumor quantitatively, demonstrating changes during and following therapy, and image responses to antivascular agents by measuring tumor blood flow. However, the imaging techniques discussed above have their advantages and disadvantages. 18FDG-PET is not a specific technique, and problems may occur due to antiangiogenesis drug–induced uncoupling of tumor perfusion and other aspects of tumor physiology [69]. CT involves exposure to ionizing radiation and has not yet demonstrated a reduction in tumor perfusion by antiangiogenic therapies. Until recently, the broader application of CT has been hindered by the lack of commercially available software to perform precise quantitative analyses. Validation of this technique is currently under way by comparing functional CT measurements with histological assessments of angiogenesis in various tumor types.
|
SUMMARY |
|---|
|
TopLearning ObjectivesAbstractIntroductionImaging Strategies in Clinical...DiscussionSummaryReferences |
|---|
|
REFERENCES |
|---|
|
TopLearning ObjectivesAbstractIntroductionImaging Strategies in Clinical...DiscussionSummaryReferences |
|---|
vß3 in endothelial cells. J Biol Chem 2001;276:39562–39568.This article has been cited by other articles:
 |
|
 |
|
  A. A.M. van der Veldt, M. R. Meijerink, A. J.M. van den Eertwegh, A. Bex, G. de Gast, J. B.A.G. Haanen, and E. Boven Sunitinib for Treatment of Advanced Renal Cell Cancer: Primary Tumor Response Clin. Cancer Res., April 15, 2008; 14(8): 2431 - 2436. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. C. Jaffe Response Assessment in Clinical Trials: Implications for Sarcoma Clinical Trial Design Oncologist, April 1, 2008; 13(suppl_2): 14 - 18. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J Heidemann, D G Binion, W Domschke, and T Kucharzik Antiangiogenic therapy in human gastrointestinal malignancies. Gut, October 1, 2006; 55(10): 1497 - 1511. [Full Text] [PDF]
|
|
|
|
 |
|
 N. Hylton Dynamic Contrast-Enhanced Magnetic Resonance Imaging As an Imaging Biomarker J. Clin. Oncol., July 10, 2006; 24(20): 3293 - 3298. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. D. Young, M. Whissell, J. C.S. Noble, P. O. Cano, P. G. Lopez, and C. J. Germond Phase II Clinical Trial Results Involving Treatment with Low-Dose Daily Oral Cyclophosphamide, Weekly Vinblastine, and Rofecoxib in Patients with Advanced Solid Tumors. Clin. Cancer Res., May 15, 2006; 12(10): 3092 - 3098. [Abstract] [Full Text] [PDF]
|
|
|
|
