Peter A. Chiarelli

A calibration method for quantitative BOLD fMRI based on hyperoxia

The estimation of changes in CMR(O2) using functional MRI involves an essential calibration step using a vasoactive agent to induce an isometabolic change in CBF. This calibration procedure is performed most commonly using hypercapnia as the isometabolic stimulus. However, hypercapnia possesses a number of detrimental side effects. Here, a new method is presented using hyperoxia to perform the same calibration step. This procedure requires independent measurement of Pa(O2), the BOLD signal, and CBF. We demonstrate that this method yields results that are comparable to those derived using other methods. Further, the hyperoxia technique is able to provide an estimate of the calibration constant that has lower overall intersubject and intersession variability compared to the hypercapnia approach.

Cerebral perfusion response to hyperoxia.

Graded levels of supplemental inspired oxygen were investigated for their viability as a noninvasive method of obtaining intravascular magnetic resonance image contrast. Administered hyperoxia has been shown to be effective as a blood oxygenation level-dependent contrast agent for magnetic resonance imaging (MRI); however, it is known that high levels of inspired fraction of oxygen result in regionally decreased perfusion in the brain potentially confounding the possibility of using hyperoxia as a means of measuring blood flow and volume. Although the effects of hypoxia on blood flow have been extensively studied, the hyperoxic regime between normoxia and 100% inspired oxygen has been only intermittently studied. Subjects were studied at four levels of hyperoxia induced during a single session while perfusion was measured using arterial spin labelling MRI. Reductions in regional perfusion of grey matter were found to occur even at moderate levels of hyperoxia; however, perfusion changes at all oxygen levels were relatively mild (less than 10%) supporting the viability of hyperoxia-induced contrast.

Flow-metabolism coupling in human visual, motor, and supplementary motor areas assessed by magnetic resonance imaging.

Combined blood oxygenation level‐dependent (BOLD) and arterial spin labeling (ASL) functional MRI (fMRI) was performed for simultaneous investigation of neurovascular coupling in the primary visual cortex (PVC), primary motor cortex (PMC), and supplementary motor area (SMA). The hypercapnia‐calibrated method was employed to estimate the fractional change in cerebral metabolic rate of oxygen consumption (CMRO2) using both a group‐average and a per‐subject calibration. The group‐averaged calibration showed significantly different CMRO2−CBF coupling ratios in the three regions (PVC: 0.34 ± 0.03; PMC: 0.24 ± 0.03; and SMA: 0.40 ± 0.02). Part of this difference emerges from the calculated values of the hypercapnic calibration constant M in each region (MPVC = 6.6 ± 3.4, MPMC = 4.3 ± 3.5, and MSMA = 7.2 ± 4.1), while a relatively minor part comes from the spread and shape of the sensorimotor BOLD–CBF responses. The averages of the per‐subject calibrated CMRO2−CBF slopes were 0.40 ± 0.04 (PVC), 0.31 ± 0.03 (PMC), and 0.44 ± 0.03 (SMA). These results are 10–30% higher than group‐calibrated values, and are potentially more useful for quantifying individual differences in focal functional responses. The group‐average calibrated motor coupling value is increased to 0.28 ± 0.03 when stimulus‐correlated increases in end‐tidal CO2 are included. Our results support the existence of regional differences in neurovascular coupling, and argue for the importance of achieving optimal accuracy in hypercapnia calibrations to resolve method‐dependent variations in published results. Magn Reson Med 57:538–547, 2007.

Sources of systematic bias in hypercapnia-calibrated functional MRI estimation of oxygen metabolism.

The change in cerebral rate of oxidative metabolism (CMR(O(2))) during neural activation may be estimated from blood oxygenation level-dependent (BOLD) functional magnetic resonance imaging (fMRI) and arterial spin-labeling (ASL) fMRI measurements. The established method relies on an epoch of iso-metabolic blood flow increase, typically induced by CO2 breathing, to calibrate the BOLD-CBF relationship at resting-state CMR(O2). Here, we discuss the systematic bias in CMR(O2)-CBF data that can be introduced depending on the value derived for the calibration constant (M) from the CO2 breathing epoch. We demonstrate that the fidelity of BOLD-CBF data acquired during the neural activation task have low impact on the tightness of CMR(O2)-CBF coupling, as well as the coupling slope, when the derived calibration value is of a relatively moderate amplitude (M in the range of, or greater than, 10-15 at 1.5 T). Via the standard reformulation of a grid in BOLD-CBF space into the CMR(O2)-CBF plane, we demonstrate the non-linear transformation that takes place and the sources of systematic bias that result. We find that the outcome of a neurovascular coupling study may be predicted to a large extent purely from the value of the calibration constant, M, that is used. Our results suggest that the accurate determination of M is of greater importance than thought previously and indicate that BOLD-CBF data must always be supplied when considering CMR(O2)-CBF behavior in a particular brain region.

Redox-Responsive Magnetic Nanoparticle for Targeted Convection-Enhanced Delivery of O6-Benzylguanine to Brain Tumors

Resistance to temozolomide (TMZ) based chemotherapy in glioblastoma multiforme (GBM) has been attributed to the upregulation of the DNA repair protein O6-methylguanine-DNA methyltransferase (MGMT). Inhibition of MGMT using O6-benzylguanine (BG) has shown promise in these patients, but its clinical use is hindered by poor pharmacokinetics that leads to unacceptable toxicity. To improve BG biodistribution and efficacy, we developed superparamagnetic iron oxide nanoparticles (NP) for targeted convection-enhanced delivery (CED) of BG to GBM. The nanoparticles (NPCP-BG-CTX) consist of a magnetic core coated with a redox-responsive, cross-linked, biocompatible chitosan-PEG copolymer surface coating (NPCP). NPCP was modified through covalent attachment of BG and tumor targeting peptide chlorotoxin (CTX). Controlled, localized BG release was achieved under reductive intracellular conditions and NPCP-BG-CTX demonstrated proper trafficking of BG in human GBM cells in vitro. NPCP-BG-CTX treated cells showed a significant reduction in MGMT activity and the potentiation of TMZ toxicity. In vivo, CED of NPCP-BG-CTX produced an excellent volume of distribution (Vd) within the brain of mice bearing orthotopic human primary GBM xenografts. Significantly, concurrent treatment with NPCP-BG-CTX and TMZ showed a 3-fold increase in median overall survival in comparison to NPCP-CTX/TMZ treated and untreated animals. Furthermore, NPCP-BG-CTX mitigated the myelosuppression observed with free BG in wild-type mice when administered concurrently with TMZ. The combination of favorable physicochemical properties, tumor cell specific BG delivery, controlled BG release, and improved in vivo efficacy demonstrates the great potential of these NPs as a treatment option that could lead to improved clinical outcomes.

Modelling vascular reactivity to investigate the basis of the relationship between cerebral blood volume and flow under CO2 manipulation.

Changes in cerebral blood flow (f) and vascular volume (v) are of major interest in mapping cerebral activity and metabolism, but the relation between them currently lacks a sufficient theoretical basis. To address this we considered three models: a uniform reactive tube model (M1); an extension of M1 that includes passive arterial inflow and venous volume (M2); and a more anatomically plausible model (M3) consisting of 19 compartments representing the whole range of vascular sizes and respective CO2 reactivities, derived from literature data. We find that M2 cannot be described as the simple scaling of a tube law, but any divergence from a linear approximation is negligible within the narrow physiological range encountered experimentally. In order to represent correctly the empirically observed slope of the overall v-f relationship, the reactive bed should constitute about half of the total vascular volume, thus including a significant fraction of capillaries and/or veins. Model M3 demonstrates systematic variation of the slope of the v-f relationship between 0.16 and 1.0, depending on the vascular compartment under consideration. This is further complicated when other experimental approaches such as flow velocity are used as substitute measurements. The effect is particularly large in microvascular compartments, but when averaged with larger vessels the variations in slope are contained within 0.25 to 0.55 under conditions typical for imaging methods. We conclude that the v-f relationship is not a fixed function but that both the shape and slope depend on the composition of the reactive volume and the experimental methods used.

Measurement of cerebral blood volume in humans using hyperoxic MRI contrast.

To develop a new method of measuring quantitative regional cerebral blood volume (CBV) using epochs of hyperoxia as an intravenous contrast agent with T2*‐weighted MRI.

Bayesian inference of hemodynamic changes in functional arterial spin labeling data.

The study of brain function using MRI relies on acquisition techniques that are sensitive to different aspects of the hemodynamic response contiguous to areas of neuronal activity. For this purpose different contrasts such as arterial spin labeling (ASL) and blood oxygenation level dependent (BOLD) functional MRI techniques have been developed to investigate cerebral blood flow (CBF) and blood oxygenation, respectively. Analysis of such data typically proceeds by separate, linear modeling of the appropriate CBF or BOLD time courses. In this work an approach is developed that provides simultaneous inference on hemodynamic changes via a nonlinear physiological model of ASL data acquired at multiple echo times. Importantly, this includes a significant contribution by changes in the static magnetization, M, to the ASL signal. Inference is carried out in a Bayesian framework. This is able to extract, from dual‐echo ASL data, probabilistic estimates of percentage changes of CBF, R  2* , and the static magnetization, M. This approach provides increased sensitivity in inferring CBF changes and reduced contamination in inferring BOLD changes when compared with general linear model approaches on single‐echo ASL data. We also consider how the static magnetization, M, might be related to changes in CBV by assuming the same mechanism for water exchange as in vascular space occupancy. Magn Reson Med, 2006.

Polymer–surfactant complexation in polyelectrolyte multilayer assemblies

Layer-by-layer self-assembly can be used to incorporate amphiphilic molecules into multilayered polyelectrolyte architectures. This review examines equilibrium LbL assemblies constructed by direct adsorption from aqueous solution. LbL systems have not only provided fundamental insight into the nature of polyion-surfactant complexation, but have also yielded functional materials with useful surface, optical, and electronic properties.

Glypican-3–Targeting F(ab′)2 for 89Zr PET of Hepatocellular Carcinoma

Hepatocellular carcinoma (HCC) is an increasingly lethal malignancy for which management is critically dependent on accurate imaging. Glypican-3 (GPC3) is a cell surface receptor overexpressed in most HCCs and provides a unique target for molecular diagnostics. The use of monoclonal antibodies (mAbs) that target GPC3 (αGPC3) in PET imaging has shown promise but comes with inherent limitations associated with mAbs such as long circulation times. This study used 89Zr-conjugated F(ab′)2 fragments directed against GPC3 (89Zr-αGPC3-F(ab′)2) to evaluate the feasibility of the fragments as a diagnostic immuno-PET imaging probe. Methods: Immobilized ficin was used to digest αGPC3, creating αGPC3-F(ab′)2 fragments subsequently conjugated to 89Zr. In vivo biodistribution and PET studies were performed on GPC3-expressing HepG2 and GPC3-nonexpressing RH7777 orthotopic xenografts. Results: Reliable αGPC3-F(ab′)2 production via immobilized ficin digestion was verified by high-performance liquid chromatography and sodium dodecyl sulfate polyacrylamide gel electrophoresis. 89Zr-αGPC3-F(ab′)2 demonstrated F(ab′)2-dependent, antigen-specific cell binding. HepG2 tumor uptake was higher than any other tissue, peaking at 100 ± 21 percentage injected dose per gram (%ID/g) 24 h after injection, a value 33- to 38-fold higher than GPC3-nonexpressing RH7777 tumors. The blood half-life of the 89Zr-αGPC3-F(ab′)2 conjugate was approximately 11 h, compared with approximately 115 h for historic mAb controls. This shorter half-life enabled clear tumor visualization on PET 4 h after administration, with a resultant peak tumor-to-liver contrast ratio of 23.3. Blocking antigen-expressing tumors with an excess of nonradiolabeled αGPC3 resulted in decreased tumor uptake similar to native liver. The kidneys exhibited high tissue uptake, peaking at 24 h with 83 ± 12 %ID/g. HepG2 tumors ranging from 1.5 to 7 mm were clearly visible on PET, whereas larger RH7777 tumors displayed signal lower than background liver tissue. Conclusion: This study demonstrates the feasibility of using 89Zr-αGPC3-F(ab′)2 for intrahepatic tumor localization with small-animal PET. Faster blood clearance and lower background liver uptake enable excellent signal-to-noise ratios at early time points. Increased renal uptake is similar to that as has been seen with clinical radioactive peptide imaging. 89Zr-αGPC3-F(ab′)2 addresses some of the shortcomings of whole-antibody immuno-PET probes. Further optimization is warranted to maximize probe sensitivity and specificity in the process of clinical translation.