|
Advertisement |
|
|
||
 |
|
|||||
|
|||||
|
 |
||||||
|
||||||
|
||||||
|
Originally published as JCO Early Release 10.1200/JCO.2002.05.102 on August 12 2002 © 2002 American Society for Clinical Oncology Development of Biologic Markers of Response and Assessment of Antiangiogenic Activity in a Clinical Trial of Human Recombinant EndostatinByFrom the Departments of Thoracic and Head and Neck Medical Oncology, Cancer Biology, Biostatistics, Diagnostic Radiology, Radiation Oncology, Surgical Oncology, Pathology, Pharmaceutical Sciences, and Gastrointestinal Medical Oncology, The University of Texas M.D. Anderson Cancer Center, and The University of Texas Medical School, Houston, TX; and National Cancer Institute, Bethesda, MD. Address reprint requests to Roy S. Herbst, MD, PhD, The University of Texas M.D. Anderson Cancer Center, 1515 Holcombe Blvd, Box 432, Houston, TX; email: rherbst{at}mdanderson.org
PURPOSE: Angiogenesis is a target for the treatment of cancer and other diseases, and its complex biology suggests that establishing the appropriate dose and schedule for antiangiogenic treatment will require extensive study. We present the initial results of a dose-finding clinical trial of recombinant human endostatin (rh-Endo) that examined potential surrogates for response to antiangiogenic therapy. PATIENTS AND METHODS: Twenty-five patients were treated with escalating doses of rh-Endo. Positron emission tomography (PET) was used to assess tumor blood flow (with [15O]H2O) and metabolism (with [18F]fluorodeoxyglucose) before the start of therapy and then every 4 weeks. To directly assess the effects of rh-Endo on endothelial cells within the tumors, biopsy specimens of tumor tissue were obtained before therapy and again at 8 weeks and evaluated for endothelial cell and tumor cell apoptosis. RESULTS: Tumor blood flow and metabolism as measured by PET scans generally decreased with increasing doses of rh-Endo; however, the effects were complex and in some analyses nonlinear. Tumor biopsy analysis revealed a significant increase in tumor cell apoptosis (P = .027) and endothelial cell apoptosis (P = .027) after 8 weeks of therapy. However, there was no statistically significant relationship between rh-Endo dose and induction of tumor cell or endothelial cell apoptosis. CONCLUSION: These initial data suggest that rh-Endo has measurable effects on tumor blood flow and metabolism and induces endothelial and tumor cell apoptosis even in the absence of demonstrable anticancer effects. Further study and validation of these biomarkers in the context of antiangiogenic therapy will be required.
TUMOR ANGIOGENESIS is essential for neoplastic proliferation, progression, invasion, and metastasis. Attacking the vascular endothelium of tumors has several potential advantages. For example, the endothelium of a tumor is directly accessible through the circulation. Further, because a single capillary supports a large number of cancer cells, there is the potential to amplify an anticancer effect through disruption of the tumor-related vasculature. Finally, since endothelial cells are not themselves malignant, they may be less likely to acquire drug resistance.1 The recognition that cancer cells are dependent on neovascularization2,3 and that the switch to an angiogenic phenotype is a key event in the progression to malignancy4 has led to the development of numerous therapeutic agents with antiangiogenic activity.5 It is now apparent that an impediment to the successful and rapid development and testing of therapeutic agents that target angiogenesis is the lack of validated assays capable of measuring an antiangiogenic effect directly in patients. Given the potentially high therapeutic index of antiangiogenic drugs, documenting an antiangiogenic effect during early clinical studies of these agents is clearly important. Therefore, we examined noninvasive and invasive techniques to assess antiangiogenic activity in the context of a phase I clinical trial of recombinant human endostatin (rh-Endo). Endostatin is a specific inhibitor of angiogenesis that was first isolated from the supernatant of an in vitro culture of a murine hemangioendothelioma cell line (EOMA cells).6 Further characterization revealed that endostatin is a 20-kd C-terminal fragment of collagen XVIII.6 In vitro, both rh-Endo and recombinant murine endostatin specifically inhibit the proliferation and migration of capillary endothelial cells and can induce apoptosis of proliferating endothelial cells. In vivo, rh-Endo was subsequently shown to produce dose-dependent growth inhibition of Lewis lung carcinoma cells.1,6 In this model, prolonged therapy with high doses of endostatin induced a virtual complete blockade of tumor angiogenesis and resulted in the regression of established tumors to microscopic lesions. Examination of these dormant lesions revealed decreased microvascular density, little or no change in tumor cell proliferation rates, and significantly increased tumor cell apoptosis when compared with controls. No resistance to therapy or toxicity was reported even after prolonged therapy with endostatin.1 Studies designed to assess endostatin’s mechanism of action have demonstrated an ability to induce endothelial cell apoptosis7 via Shb (an adaptor protein involved in apoptosis),8 to inhibit endothelial cell migration9,10 and proliferation,6 and to block ex vivo capillary outgrowth.11 Endostatin has also been shown to bind to endothelial cell tropomyosin,12 to interact with integrins,13 to disrupt cell-cell and cell-matrix adhesion of activated endothelial cells,8 and to modulate proteinase activity.14,15 We hypothesized that rh-Endo’s antiangiogenic effects could be more accurately assessed by using a panel of biomarkers designed to provide an early indication of antiangiogenic activity and that measurable effects on angiogenesis might be documented before clinically demonstrable reductions in tumor size. We prospectively selected and imaged primary and metastatic lesions serially using [15O]H2O positron emission tomography (PET) to assess changes in tumor blood flow and [18F]fluorodeoxyglucose ([18F]FDG) PET to assess changes in tumor metabolism. Furthermore, the antiangiogenic effects of rh-Endo were directly examined through the acquisition of tumor biopsy specimens that were analyzed for changes in endothelial cell and tumor cell apoptosis. Taken together, it was hypothesized that these studies would help to identify the biologically active doses of rh-Endo, thereby facilitating selection of an appropriate dose for future phase II trials.
Dose-Finding (Phase I) Clinical Trial Patients with biopsy-confirmed advanced cancer meeting protocol-specified eligibility requirements were treated with daily intravenous doses of rh-Endo administered over 20 minutes (Fig 1).16 A major criterion for eligibility was the presence of tumor lesions that were accessible for biopsy and greater than 2 cm in one diameter to facilitate imaging analysis. All patients were informed about the investigational nature of the program and gave informed consent according to institutional and federal guidelines. A separate informed consent outlining the risks of the biopsy procedure was obtained before all biopsies. When possible, imaged lesions were not biopsied so as to avoid a potential postbiopsy effect on the imaging analysis.
Patients were sequentially allocated in cohorts of three to increasing dose levels. The schema for the correlative studies performed is shown in Fig 1. Noninvasive imaging using PET was repeated every 28 days; tumor biopsies were performed at baseline and after approximately 8 weeks of therapy. A full description of the eligibility criteria for the study is included in the accompanying article.16
PET Imaging Protocol Tumor blood flow and glucose metabolism were calculated by drawing regions of interest (ROIs) on the index tumors. A separate region, typically the psoas muscle, was drawn as a control for estimating blood flow and metabolism in nontumor tissue. Between one and four ROIs were defined for each patient. ROIs for the follow-up scans were drawn to match the ROIs from the baseline scans. The arterial concentration of the injected dose was estimated by analyzing a region over a large artery, such as the descending aorta. Given the large number of repeat studies, arterial blood gas sampling was not thought to be practical. Blood flow was computed using the one-compartment model published by Yoshida et al.17 The model assumes accumulation of tracer in the tissue with very little egress of the tracer from the tissue during the 2 minutes of data acquisition. Errors associated with using this simple model for measuring blood flow with [15O]H2O have been reported previously and show that the magnitude of the errors in computed blood flow is a function of the scan times and the initial blood flow values.18,19 Glucose metabolism was calculated using the standardized uptake value method, which normalizes the glucose PET value in the region of interest by the dose injected and the weight of the patient.20 Changes in tumor blood flow and glucose metabolism were computed by measuring the fold change in the baseline values compared with measurements at days 28 and 56 after rh-Endo treatment. These time points were chosen to coincide with standard imaging of tumor size. The detailed scientific rational and reproducibility of the PET imaging protocol is outlined in the Appendix.
Tumor Biopsies Immunofluorescence CD31 and terminal deoxynucleotidyl transferase–mediated deoxyuridine triphosphate-biotin nick end labeling (TUNEL). For in vitro studies, human microvascular endothelial cells (HMVECnd; Cascade Biologics, Portland, OR) were exposed to 100 µm of endostatin for 24 hours. Cells were fixed with 4% paraformaldehyde and TUNEL (Promega, Madison, WI) was performed according to the manufacturer’s protocol. Cell nuclei were counterstained with propidium iodide 1 µg/mL. For biopsy analysis, frozen tissue sections (8 µm) were fixed with cold acetone for 5 minutes, acetone plus chloroform (1:1) for 5 minutes, and acetone for 5 minutes. Tissues were washed with phosphate-buffered saline (PBS) for 3 minutes and incubated with protein block (1% normal goat serum and 5% normal horse serum in PBS) for 15 minutes. Protein blocks were drained, and tissues were incubated with a 1/400 dilution of monoclonal antihuman CD31, clone JC/70A (Dako Corp, Carpinteria, CA), in protein block overnight at 4°C. Exposure to light was avoided while tissues were washed with PBS three times for 3 minutes and incubated with a 1/400 dilution of Cy5-conjugated goat antimouse secondary (Jackson ImmunoResearch Laboratories, West Grove, PA) in protein block for 4 hours at 4°C. Specimens were washed two times with PBS containing 0.1% Brij for 3 minutes and washed once with PBS for 3 minutes. TUNEL21 was performed using a commercial kit (Promega). Tissues were fixed with 4% paraformaldehyde for 10 minutes at room temperature. Specimens were washed with PBS two times for 3 minutes and then incubated with 0.2% Triton X-100 for 15 minutes at room temperature. Tissues were washed with PBS two times for 3 minutes and incubated with reaction buffer (from kit) in a humid atmosphere at 37°C for 1 hour, while exposure to light was avoided. Tissues were washed three times for 5 minutes to remove unincorporated fluorescein–deoxyuridine triphosphate. Cell nuclei were counterstained with propidium iodide 1 µg/mL for 5 minutes. Tissues were then washed with PBS two times for 3 minutes and Prolong (Molecular Probes, Eugene, OR) was used to mount coverslips. Images were captured using a 20x objective (Zeiss Plan-Neofluar, Oberkochen, Germany) on an epifluorescence microscope equipped with narrow bandpass excitation filters mounted in a filter wheel to select for green and red fluorescence. Images were processed using Adobe Photoshop software (Adobe Systems, Mountain View, CA).
Laser Scanning Cytometry Analysis
Statistical Methods Since many of the scatterplots suggested that the relationships were not monotonic, we also computed piecewise linear regressions for the data.24 Inspection of the plots indicated that the joining point for the two line segments was generally around a dose of 180 mg/m2, so this value was used in our analyses. The P values reported test whether the slopes of the two segments are simultaneously equal to zero. To identify potential problems with this approach, regression diagnostics were performed on these analyses. For biopsy data, several of the biopsy specimens had baseline apoptotic tumor endothelial cell counts of zero; therefore, change rather than fold change was used to analyze the biopsy data. Wilcoxon signed rank tests were used to assess the significance of the change in apoptosis from baseline, and the analytic approaches described previously for the PET data were used to assess relationships between these changes and dose. Finally, we analyzed each lesion as an independent observation even though some tumors were from the same patient. There was generally too little data to fit the hierarchical models necessary to appropriately account for having correlated, multilevel (lesions nested within patients) observations. Furthermore, data from different lesions within patients were not highly correlated.
Clinical Effects Twenty-five patients were treated on this dose-finding (phase I) clinical trial using the treatment schema outlined in Fig 1. All patients were treated according to a prespecified dose escalation schedule with daily intravenous doses of 15, 30, 60, 120, 180, 300, and 600 mg/m2.16 Each cohort consisted of three patients with the exception of the 300-mg/m2/d cohort, which was expanded to seven patients in an effort to increase the statistical power of the pharmacodynamic end points of the study (Table 1). Since there was an absence of significant rh-Endo–related toxicity, intrapatient dose escalations during subsequent courses of therapy were allowed. Five patients received doses higher than their starting dose based on these criteria, but this occurred only after 8 weeks of therapy and, hence, does not affect the PET and biopsy studies described here. The accrual of patients with melanoma and sarcoma reflected referral of cancers with a prominent local/regional component that rendered these patients excellent candidates for the tumor biopsies required during therapy. At all dose levels tested, rh-Endo was well tolerated, supporting the preclinical predictions that a clinically effective dose could not be predicted from its maximum-tolerated dose. While some minor responses and stable disease were observed, no partial tumor responses were seen.16
PET PET scans were performed to estimate the change in index tumor blood flow (ITBF) using radiolabeled water ([15O]H2O) and index tumor metabolism (ITM) using radiolabeled fluorodeoxyglucose ([18F]FDG) over time relative to the dose of endostatin administered (Fig 2, Table 2). With a moving average statistical analysis, Fig 3 summarizes the flow and metabolism data from the baseline scan to the 28-day scan for 21 of 25 patients treated (no PET was done at the 15-mg/m2 level, and one patient died before the second scan was performed). At the lower dose levels (30 mg/m2/d to 60 mg/m2/d), the mean fold change in ITBF for scan 2 increased relative to scan 1, but it fell below baseline at doses  120 mg/m2/d. The absolute magnitude of this effect was modest, averaging a reduction in ITBF of approximately 20% from baseline, but without further decrease in ITBF with increasing rh-Endo dose. Interestingly, ITM continued to increase through the 180-mg/m2/dose level, before decreasing at dose levels of 300 mg/m2/d. The results achieved statistical significance but depended on the inclusion of the three patients treated at 30 mg/m2. [15O]H2O was not available for the baseline scans for these patients, and hence their change represents the difference from cycle 1 to cycle 2 (open circles in Fig 3). In light of this, we analyzed the results with these patients included and then excluded.
Including all patients, relative to rh-Endo dose, the change in ITBF and ITM at 28 days was statistically significant by rank correlation analysis (RCA) (ITBF, P = .009; ITM, P = .005). However, at 56 days (Fig 4), the change in ITBF was no longer significant (RCA, P = .45) while the change in ITM persisted (RCA, P = .025). To further assess the relationship between rh-Endo dose, ITBF, and ITM, the flow and metabolism data were also analyzed using a technique sensitive to biphasic linear relationships (piecewise linear regression analysis [PLA]). At 28 days, these analyses again demonstrated a statistically significant relationship with rh-Endo dose (ITBF, P = .002; ITM, P = .066), but at 56 days, the complexity of the relationships between rh-Endo dose, ITBF, and ITM was suggested by the fact that the change in ITBF was highly significant (PLA, P = .0002), but the change in ITM failed to remain statistically significant (PLA, P = .28). The trends, however, at both time points remained the same.
Since three patients initiated the PET scanning protocol after starting on rh-Endo, we also analyzed the PET data without these data points and again performed the same analyses using the moving average, rank correlation, and piecewise linear approaches. With the exclusion of these early time points (open circles, Figs 3 and 4) at 28 days, the PET tumor blood flow data did not suggest as significant a relationship with rh-Endo dose for the blood flow (RCA, P = .18), but the statistical significance of the change in metabolism (RCA, P = .024) was preserved. The piecewise linear effects analysis was also sensitive to the exclusion of these early time points, but only for metabolism (ITBF, PLA, P = .005; ITM, PLA, P = .17). At 56 days, the exclusion of these early time points did not substantially alter the statistical relationships between rh-Endo dose, ITBF, and ITM. Analysis of multiple non–tumor-bearing organs, including the liver, lung, and myocardium, did not show any change in blood flow or metabolism relative to rh-Endo dose (data not shown). Both analyses suggested that rh-Endo’s effect on ITBF and ITM was biphasic, with a transition point at approximately 180 mg/m2/d. Finally, measures of rh-Endo exposure (area under the concentration-time curve, trough endostatin plasma concentration) showed no relationship with ITBF (data not shown).
Tumor Biopsy Specimens Endothelial cells were identified using immunofluorescence staining for CD31, and apoptotic cells were identified using the TUNEL assay. The slides prepared from the paired tumor biopsy specimens were analyzed by LSC, a technique designed to make fluorescence-based quantitative measurements on tissue sections or other cellular preparations at the single-cell level. To validate the double-labeling technique, HMVECnd cells were exposed to 100 µm of rh-Endo (Fig 5, A, through C). Apoptotic HMVECnd cells (Fig 5C) were detected by yellow fluorescence of double-labeled cells. A representative double-labeling study performed on patient material after 8 weeks of treatment with rh-Endo is shown in Fig 5, D through F. After the double-labeling procedure, the LSC was used to scan the entire biopsy specimen and to generate scatterplots to estimate the percentage of TUNEL+/CD31- and TUNEL+/CD31+ cells in each specimen (Fig 6). To assess the relationship of rh-Endo dose to endothelial cell (CD31+) and tumor cell (CD31-) apoptosis, the double-labeling data from all 17 patients were assessed using moving average, rank correlation, and piecewise linear statistical analyses. The percentage of apoptotic endothelial cells (Fig 7A) and apoptotic tumor cells (Fig 7B) at baseline and after 8 weeks of rh-Endo therapy is shown. There was a statistically significant increase in the percentage of endothelial cell apoptosis after therapy (median, 0.2% before therapy and 1.1% after therapy; P = .027, Wilcoxon signed rank test) and tumor cell apoptosis (median, 0.5% before therapy and 2.3% after therapy; P = .027, Wilcoxon signed rank test). Figure 7 shows the change in the percentage of TUNEL+/CD31+ (endothelial) cells (Fig 7C) and TUNEL+/CD31- (tumor) cells (Fig 7D) relative to dose. A small increase in the percentage of apoptotic endothelial and tumor cells was seen up to an rh-Endo dose of 180 mg/m2/d, but the absolute magnitude of the effect was slight and was not statistically significant.
The optimal design of early clinical trials of antiangiogenic agents should include appropriate end points to assess biologic effects. Unlike cytotoxic drugs, antiangiogenic monotherapy may not induce immediate tumor-cell kill. There may be a period of significant tumor progression before response, and the maximum-tolerated dose may not predict the effective anticancer dose.25 This study is the first to suggest that endostatin can induce endothelial cell apoptosis and changes in tumor blood flow in some patients with advanced cancer. In addition, these data suggest that there is a complex, possibly nonlinear, relationship between tumor blood flow, tumor metabolism, and rh-Endo dose. The fact that rh-Endo produced measurable effects on tumor blood flow, tumor metabolism, and endothelial and tumor cell apoptosis in the absence of tumor regression strongly suggests that tumor imaging and biopsies are important end points for determining the appropriate dosing of antiangiogenic compounds that either have a delayed onset of antitumor activity or, as single agents, do not induce tumor regression. Alternative potential indicators of biologic activity include measurement of direct effects on molecular targets. However, although endostatin can bind with low affinity to heparin sulfate proteoglycans such as glypican-1 and glypican-4 and tropomyosin,12,26 a high-affinity receptor for endostatin has not yet been identified. In this study, noninvasive imaging using PET techniques was used to evaluate tumor blood flow and metabolism as biologic end points of antitumor activity. The use of PET allowed for the assessment of both tumor-related (ie, metabolism) and endothelial-related (ie, blood flow) parameters and provided information about both the tumor and endothelial compartments of the tumor.27 The use of [15O]-labeled water for PET imaging offers several properties that are desirable for the measurement of blood flow. It is freely diffusible and has a short half-life of 2 minutes and favorable dosimetric properties.28 This technique has been used to measure blood flow in several tumor types, including breast cancer29,30 and brain tumors.31-33 There are, however, potential limitations to this technique. In small tumors, partial volume effects may be significant if the tumor size is less than two times the resolution of the scanner.28 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 aorta, or within areas of relatively high flow, such as liver.28 This may limit the use of PET in the lung, liver, and mediastinum. Additionally, tumors may not have 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. Additionally, it has been shown that at high flow rates, measuring blood flow by PET underestimates blood flow. Three issues relevant to the PET data deserve consideration. First, the fact that the data were sensitive to the exclusion of a small number of patients/observations and whether the observations were made at 28 days or 56 days after the start of rh-Endo emphasizes the complexity of such studies and argues that in phase I trials where exploration of biologic correlates are the primary end points, larger numbers of patients should be treated at each dose level and more observations per patient should be made. The fact that the statistical significance of our analysis was influenced by the inclusion/exclusion of three patients treated at the lowest dose of rh-Endo and the timing of the scans points to the preliminary nature of these observations. Second, the fact that many of the changes in ITBF and ITM persisted through day 56 argues that inclusion of the three patients imaged at day 28 and day 56 was valid, although further experience with rh-Endo will be needed to verify this opinion. Third, the absence of clinical anticancer effects could argue that the changes in ITBF and ITM simply reflect the natural progression of cancer metastases and are not specific to an rh-Endo–mediated antiangiogenic effect. Given the ethical issues raised in designing trials in which patients would be serially scanned without therapeutic intervention, there are limited data available to refute this point. However, on a follow-up trial, we have had an opportunity to observe two patients not receiving therapy where [18F]FDG scans were performed at baseline and again approximately 4 weeks later. In these patients, a 20% to 30% increase in tumor blood flow and metabolism and flow was seen. Thus, our small experience with patients undergoing serial PET scans without receiving intervening therapy suggests that the changes in ITBF and ITM documented in this trial are secondary to rh-Endo’s antiangiogenic effect. Other general issues in tumor biology and drug development may affect these results and their interpretation. The heterogeneity of delivery of drugs to solid tumors may lead to variability in 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 molecules34,35 may all make the interpretation of imaging data challenging. Taken together, these data suggest that it may be prudent to use several different strategies to noninvasively evaluate the antitumor and antiangiogenic effects of angiogenesis inhibitors in future clinical trials. Intriguingly, the tumor biopsy analysis in this study suggests that endostatin induces apoptosis of both endothelial cells and tumor cells, although there was no significant relationship with dose. Other investigators have also demonstrated that antiangiogenic agents induce apoptosis. Shaheen et al36,37 demonstrated that tyrosine kinase inhibitors induce tumor cell and endothelial cell apoptosis, and other investigators have demonstrated that endostatin induces apoptosis in endothelial cell lines.8,38 Further studies should address the exact temporal nature of this effect. The absence of a relationship with dose may simply reflect the small number of patients studied or may suggest that there is greater intrapatient and interpatient heterogeneity to the effect of antiangiogenic therapy than initially presumed. This possibility is suggested by the work of Arap and Pasqualini.39,40 Similar end points have been studied by Griffon-Etienne et al,41 who showed in an animal model that paclitaxel and docetaxel induced endothelial cell apoptosis, with the maximum increase in the number of apoptotic cells observed within the first 24 hours. This has interesting implications for future studies, as the period of maximal induction of apoptosis may occur earlier than 8 weeks after the beginning of treatment and earlier biopsies may provide more accurate assessment of apoptosis induction. In conclusion, the noninvasive and invasive biomarkers used in this trial suggest that rh-Endo has an antiangiogenic effect in patients. Noninvasive imaging with PET scans showed complex but generally dose-dependent changes in both tumor blood flow and metabolism. Serial tumor biopsies showed evidence of induction of endothelial cell and tumor cell apoptosis, although dose dependence was not established. These studies provide a starting point for developing sensitive and specific biomarkers of antiangiogenic therapy and suggest that daily bolus doses of approximately 180 mg/m2/d be evaluated further. Understanding the role of these biomarkers will facilitate the rational development of antiangiogenic drug therapy. However, it remains challenging to identify clinically relevant doses of mechanistically novel agents based solely on the identification of a biologically effective dose established by a surrogate end point.
APPENDIX Mathematical Model for Measurement of Blood Flow With Oxygen-15–Labeled Water Oxygen-15–labeled water is used with positron emission tomography (PET) to measure regional tumor blood flow. Water is freely diffusable in tissue; therefore, the distribution of the labeled water is distributed in the tumor as a function of blood flow to the tumor. Once the labeled water enters the tumor tissue, it egresses slowly from the tissue as a function of blood flow and is governed by the partition coefficient. The simplicity of the tracer exchange between the blood and the tumor permits the use of a one-compartment tracer kinetic model to express the dynamics of the tracer. From the mass conservation equation, we can state that the rate of uptake of the tracer in the tumor is a function of the rate of delivery of the tracer minus the rate of egress of the tracer, as shown in the differential equation in Equation 1, where U(t) is the uptake of labeled water in the tumor at a time t, Ca and Cv are the arterial and venous concentrations of the tracer in the blood vessels at a time t, and F is the blood flow to the tumor.
The venous egress, Cv(t), is difficult to measure; therefore, it is approximated by assuming that the concentration of the venous blood is a function of the partition coefficient of tumor and blood and the uptake of the tracer in the tumor. Equation 1 can therefore be modified to the form below:
where lambda is the partition coefficient for tumor tissue and blood. The solution to Equation 2 is a convolution of the arterial input function and the egress from the venous side and is expressed by
For short data acquisition times relative to the rate of egress of the tracer, the venous egress is small. The solution in Equation 3 can be simplified by assuming that the major egress can be expressed as a first-order expansion of the exponential term. Therefore, a simpler solution for measuring blood flow is often expressed as the following equation, where K is the fraction of the tracer that has egressed from tumor. For the case where K = 0, the equation reduces to the simple form that is used in the Ketty-Schmidt model.
Arterial Concentration Measurements To compute blood flow, the arterial concentration of the tracer is required as a function of time. There are two ways of obtaining this arterial concentration: direct and indirect. The direct method requires that samples of arterial blood be acquired by an arterial catheterization and withdrawal of arterial blood samples at regular time intervals. The concentration of the oxygen-15 label in the blood is then measured by assaying the radioactivity in the blood for each sample. The indirect method measures the blood concentration of the tracer in anatomic structures that are known to be composed primarily of blood, such as the left ventricle or larger arteries, such as the aorta. Indirect sampling of arterial blood concentration by PET is fast and easy to achieve and does not requires special additional equipment or tedious blood draws. However, indirect sampling requires corrections for partial volume errors due to the smaller size of the aorta compared with the resolution of the PET cameras. Corrections for partial volume errors can be achieved by determining the size of the blood vessel being monitored as described below. Partial Volume Correction Partial volume errors are caused by the finite resolution of the PET camera as compared with the object being imaged. Objects that are smaller than two times the resolution of the PET camera will result in reduced uptake of tracer and an underestimation of concentrations. Small objects need to have a partial volume correction applied for quantitative PET imaging. These corrections can be applied to the measured data if the size of the object is known. A modification of the blood flow equations incorporating partial volume corrections is shown in Equation 5 below, where PVA is the partial volume correction factor for the arterial blood sampling from the aorta, and PVT is the partial volume correction factor for the smaller sized tumors. The PVT correction for tumor partial volume error can be minimized if the tumors to be studied are greater than two times the resolution of the PET camera.
Measurements of Change in Blood Flow for Sequential Scans Study The use of PET to measure changes in blood flow with sequential scans in the same patient and tumor, obtained during a short time duration, greatly simplify the equations for measuring blood flow. The fractional or fold change in tumor flow from scan 1 to 2 can be expressed as difference in flow from scan 1 to 2 divided by the flow from scan 1, as shown in Equation 6.
Certain assumptions and approximations can be made that simplify the calculations in Equation 6. The first assumption is that the partial volume corrections for tumors (PVT) can be kept small if tumors of larger than two times the resolution of the scanner are chosen. In this trial, we chose tumors that were greater than 2 cm in size to minimize the partial volume errors in the index tumors selected for the measurements of blood flow and glucose metabolism. Additionally, since the same tumor is studied for the measurements of changes from scan 1 to scan 2, the size of the tumor will not change greatly from scan 1 to scan 2, and the partial volume error is minimized. The second assumption is that the partial volume corrections for the aorta (PVA) will be approximately the same for scan 1 and scan 2, if the same area of the aorta is sampled for the two scans. The third assumption is that the egress of the radiolabel from the tumor is small for the scan time chosen and that this fractional egress (K) is approximately the same for scan 1 and scan 2. The above assumptions are presented below in equation form.
Applying these to Equation 6 allows the egress and partial volume errors to cancel out, therefore leaving a simpler equation for the measurement of changes in blood flow in tumors following a short duration of treatment.
Data Acquisition and Analysis Methods There are two methods of data acquisition and analysis for the measurement of tumor blood flow. The first is a dynamic acquisition method that collects sequential images every 3 to 5 seconds and then analyzes the images using a curve-fitting method. This acquisition model produces the best results, but it is more susceptible to statistical noise in the data and does require greater expertise for data analysis. The second method is called the single data acquisition method and is sometimes referred to as the autoradiographic method. It is simpler to use and requires much less analysis time. In some older PET scanners, this is the only mode of data acquisition available since the dynamic acquisition mode is a recent addition to PET. The single data acquisition mode has a fixed scan time that is kept short such that the egress of the tracer is small relative to the blood flow. Data are acquired for one to two minutes and a snap shot of the blood flow image is created. At high blood flows, there is an underestimation of the absolute blood flow but the error due to egress of the tracer from the tumor is minimized if measuring relative change in blood flow. Glucose Metabolism Measurements Glucose metabolism measurements can also be simplified to a one-compartment tracer kinetic model. Glucose is not highly extracted into the tumor; therefore, the tracer kinetic equation is modified to include E, where E is the extraction fraction of glucose in the tumor. There is also a lumped constant, L, to account for fraction of labeled to free glucose in the blood. A simpler solution to the glucose uptake equation is to normalize the uptake of glucose by the body weight and the dose injected shown in Equation 8 below. This normalized uptake of glucose is referred to as the standardized uptake value (SUV) and is proportional to the glucose metabolic rate.
The SUV is a simpler method of measuring glucose uptake in the tumor and is currently used in clinical applications. The need for arterial sampling of the tracer is removed and replaced with total dose and patient weight. Changes in SUV are similar to changes in glucose metabolism since the lumped constant, extraction, and the partial volumes cancel out if we assume them to be similar for the two scans. Errors in Repeat Measurement of Blood Flow and metabolism by PET Repeat measurements of blood flow and glucose metabolism by PET have different magnitudes of errors depending on several factors, such as the radiation dose injected, type of tracer used, residence time of the radiolabeled tracer in the tumor, and the data acquisition time. The methods and errors associated with blood flow measurements in tumors with oxygen-15–labeled water have been presented by Bacharach et al,28 Beaney et al,29 Wilson et al,30 and Ito et al.33 The test-retest scan error can range from 8% to 15% depending on the data acquisition mode and the sensitivity of the PET camera. We did not do a test-retest scan with multiple injections of the oxygen-15 tracer in the endostatin–treated patients because of radiation dose limitations of the three PET scans in our study. However, we did evaluate the errors associated with the analysis portion of the study. The error for reanalysis of the data using our PET scanner and data acquisition mode was measured at 7%. We therefore expect the total test-retest error for multiple injections of the tracers to be less than 15% for blood flow measurements and less than 10% for the glucose measurements.
Supported by the United States Public Health Service grant nos. U01 CA62461, U54 CA90810, and CCSG CA16672; an American Society of Clinical Oncology Career Development Award (to R.S.H.); and the Golfers Against Cancer. We thank Drs Merrick Ross, Barry Feig, Paul Mansfield, Douglas B. Evans, Garrett L. Walsh, Michael Andreeff, Isaiah J. Fidler, and Waun Ki Hong. We also thank Sonya Dalton and Bich Tran.
N.A.M., D.W.D., and K.R.H. contributed equally to this work. This article was published ahead of print at www.jco.org.
1. Boehm T, Folkman J, Browder T, et al: Antiangiogenic therapy of experimental cancer does not induce acquired drug resistance. Nature 390: 404-407, 1997[CrossRef][Medline] 2. Folkman J: Tumor angiogenesis: Therapeutic implications. N Engl J Med 285: 1182-1186, 1971[Medline] 3. Folkman J: Angiogenesis in cancer, vascular, rheumatoid and other disease. Nat Med 1: 27-31, 1995[CrossRef][Medline] 4. Hanahan D, Folkman J: Patterns and emerging mechanisms of the angiogenic switch during tumorigenesis. Cell 86: 353-364, 1996[CrossRef][Medline] 5. Hamby JM, Showalter HD: Small molecule inhibitors of tumor-promoted angiogenesis, including protein tyrosine kinase inhibitors. Pharmacol Res 82: 169-193, 1999 6. O’Reilly MS, Boehm T, Shing Y, et al: Endostatin: An endogenous inhibitor of angiogenesis and tumor growth. Cell 88: 277-285, 1997[CrossRef][Medline]
7. Dhanabal M, Ramchandran R, Waterman MJ, et al: Endostatin induces endothelial cell apoptosis. J Biol Chem 274: 11721-11726, 1999
8. Dixelius J, Larsson H, Sasaki T, et al: Endostatin-induced tyrosine kinase signaling through the Shb adaptor protein regulates endothelial cell apoptosis. Blood 95: 3403-3411, 2000
9. Shichiri M, Hirata Y: Antiangiogenesis signals by endostatin. FASEB J 15: 1044-1053, 2001 10. Yamaguchi N, Anand-Apte B, Lee M, et al: Endostatin inhibits VEGF-induced endothelial cell migration and tumor growth independently of zinc binding. EMBO J 18: 4414-4423, 1999[CrossRef][Medline] 11. Kruger EA, Duray PH, Tsokos MG, et al: Endostatin inhibits microvessel formation in the ex vivo rat aortic ring angiogenesis assay. Biochem Biophys Res Commun 268: 183-191, 2000[CrossRef][Medline]
12. MacDonald NJ, Shivers WY, Narum DL, et al: Endostatin binds tropomyosin: A potential modulator of the antitumor activity of endostatin. J Biol Chem 276: 25190-25196, 2001
13. Rehn M, Veikkola T, Kukk-Vladre E, et al: Interaction of endostatin with integrins implicated in angiogenesis. Proc Natl Acad Sci USA 98: 1024-1029, 2001
14. Kim YM, Jang JW, Lee OH, et al: Endostatin inhibits endothelial and tumor cellular invasion by blocking the activation and catalytic activity of matrix metalloproteinase. Cancer Res 60: 5410-5413, 2000
15. Wickstrom SA, Veikkola T, Rehn M, et al: Endostatin-induced modulation of plasminogen activation with concomitant loss of focal adhesions and actin stress fibers in cultured human endothelial cells. Cancer Res 61: 6511-6516, 2001 16. Herbst RS, Hess KR, Tran HT, et al: Phase I study of recombinant human endostatin in patients with advanced solid tumors. J Clin Oncol 20: 3791-3803, 2002 17. Yoshida K, Mullani N, Gould KL: Coronary flow and flow reserve by PET simplified for clinical applications using rubidium-82 or nitrogen-13-ammonia. J Nucl Med Allied Sci 37: 1701-1712, 1996 18. Kitsukawa S, Yoshida K, Mullani N, et al: Simple and Patlak models for myocardial blood flow measurements with nitrogen-13-ammonia and PET in humans. J Nucl Med Allied Sci 39: 1123-1128, 1998 19. Volkow ND, Mullani N, Gould LK, et al: Sensitivity of measurements of regional brain activation with oxygen-15-water and PET to time of stimulation and period of image reconstruction. J Nucl Med Allied Sci 32: 58-61, 1991 20. Huang SC: Anatomy of SUV: Standardized uptake value. Nucl Med Biol 27: 643-646, 2000[CrossRef][Medline] 21. Gavrieli Y, Sherman Y, Ben-Sasson SA: Identification of programmed cell death in situ via specific labeling of nuclear DNA fragmentation. J Cell Sci 119: 493-501, 1992 22. Cleveland WS: Robust locally weighted regression and smoothing scatterplots. J Am Stat Assoc 74: 829-836, 1979[CrossRef] 23. Spearman C: The proof and measurement of association between two things: By C. Spearman, 1904. Am J Psychol 100: 441-471, 1987[CrossRef][Medline]
24. Vieth E: Fitting piecewise linear regression functions to biological responses. J Appl Physiol 67: 390-396, 1989 25. O’Reilly MS, Holmgren L, Chen C, et al: Angiostatin induces and sustains dormancy of human primary tumors in mice. Nat Med 2: 689-692, 1996[CrossRef][Medline] 26. Karumanchi SA, Jha V, Ramchandran R, et al: Cell surface glypicans are low-affinity endostatin receptors. Mol Cell 7: 811-822, 2001[CrossRef][Medline] 27. Sim BKL, Fogler WE, Zhou XH, et al: Zinc ligand-disrupted recombinant human endostatin: Potent inhibition of tumor growth, safety and pharmacokinetic profile. Angiogenesis 3: 41-51, 1999[CrossRef][Medline] 28. Bacharach SL, Libutti SK, Carrasquillo JA: Measuring tumor blood flow with H(2)(15)O: Practical considerations. Nucl Med Biol 27: 671-676, 2000[CrossRef][Medline] 29. Beaney RP, Lammertsma AA, Jones T, et al: Positron emission tomography for in-vivo measurement of regional blood flow, oxygen utilization, and blood volume in patients with breast carcinoma. Lancet 1: 131-134, 1984[Medline]
30. Wilson CB, Lammertsma AA, McKenzie CG, et al: Measurements of blood flow and exchanging water space in breast tumors using positron emission tomography: A rapid and noninvasive dynamic method. Cancer Res 52: 1592-1597, 1992 31. Rhodes CG, Wise RJ, Gibbs JM, et al: In vivo disturbance of the oxidative metabolism of glucose in human cerebral gliomas. Ann Neurol 14: 614-626, 1983[CrossRef][Medline]
32. Lammertsma AA, Jones T: Low oxygen extraction fraction in tumours measured with the oxygen-15 steady state technique: Effect of tissue heterogeneity. Br J Radiol 65: 697-700, 1992 33. Ito M, Lammertsma AA, Wise RJ, et al: Measurement of regional cerebral blood flow and oxygen utilisation in patients with cerebral tumours using 15O and positron emission tomography: Analytical techniques and preliminary results. Neuroradiology 23: 63-74, 1982[CrossRef][Medline] 34. Jain RK: Normalizing tumor vasculature with anti-angiogenic therapy: A new paradigm for combination therapy. Nat Med 7: 987-989, 2001[CrossRef][Medline] 35. Jain RK: Barriers to drug delivery in solid tumors. Sci Educ Bull 271: 58-65, 1994
36. Shaheen RM, Davis DW, Liu W, et al: Antiangiogenic therapy targeting the tyrosine kinase receptor for vascular endothelial growth factor receptor inhibits the growth of colon cancer liver metastasis and induces tumor and endothelial cell apoptosis. Cancer Res 59: 5412-5416, 1999
37. Shaheen RM, Tseng WW, Davis DW, et al: Tyrosine kinase inhibition of multiple angiogenic growth factor receptors improves survival in mice bearing colon cancer liver metastases by inhibition of endothelial cell survival mechanisms. Cancer Res 61: 1464-1468, 2001 38. Dhanabal M, Volk R, Ramchandran R, et al: Cloning, expression, and in vitro activity of human endostatin. Biochem Biophys Res Commun 258: 345-352, 1999[CrossRef][Medline] 39. Rajotte D, Arap W, Hagedorn M, et al: Molecular heterogeneity of the vascular endothelium revealed by in vivo phage display. J Clin Lab Anal 102: 430-437, 1998 40. Kolonin M, Pasqualini R, Arap W: Molecular addresses in blood vessels as targets for therapy. Curr Opin Chem Biol 5: 308-313, 2001[CrossRef][Medline]
41. Griffon-Etienne G, Boucher Y, Brekken C, et al: Taxane-induced apoptosis decompresses blood vessels and lowers interstitial fluid pressure in solid tumors: Clinical implications. Cancer Res 59: 3776-3782, 1999 Submitted May 17, 2002; accepted July 17, 2002.
This article has been cited by other articles:
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
|
|||||||||||
|
|||||||||||
|
|
Copyright © 2002 by the American Society of Clinical Oncology, Online ISSN: 1527-7755. Print ISSN: 0732-183X
|
TOP
ABSTRACT
INTRODUCTION
