Substitution of venous for arterial blood sampling in the determination of regional rates of cerebral protein synthesis with L-[1- 11 C]leucine PET: A validation study

Abstract

We developed and validated a method to estimate input functions for determination of regional rates of cerebral protein synthesis (rCPS) with L-[1-¹¹C]leucine PET without arterial sampling. The method is based on a population-derived input function (PDIF) approach, with venous samples for calibration. Population input functions were constructed from arterial blood data measured in 25 healthy 18–24-year-old males who underwent L-[1-¹¹C]leucine PET scans while awake. To validate the approach, three additional groups of 18–27-year-old males underwent L-[1-¹¹C]leucine PET scans with both arterial and venous blood sampling: 13 awake healthy volunteers, 10 sedated healthy volunteers, and 5 sedated subjects with fragile X syndrome. Rate constants of the L-[1-¹¹C]leucine kinetic model were estimated voxel-wise with measured arterial input functions and with venous-calibrated PDIFs. Venous plasma leucine measurements were used with venous-calibrated PDIFs for rCPS computation. rCPS determined with PDIFs calibrated with 30–60 min venous samples had small errors (RMSE: 4–9%), and no statistically significant differences were found in any group when compared to rCPS determined with arterial input functions. We conclude that in young adult males, PDIFs calibrated with 30–60 min venous samples can be used in place of arterial input functions for determination of rCPS with L-[1-¹¹C]leucine PET.

Keywords

Brain leucine positron emission tomography protein synthesis venous blood sampling

Introduction

Quantitative positron emission tomography (PET) studies typically require measurement of the time course of concentration of PET radiotracer in arterial plasma obtained from serial sampling of arterial blood. Because this approach is invasive for the patient and associated with some risk, the issue of non-invasive estimation of the plasma time-activity curve (TAC) in PET has been extensively addressed. The studies in this regard can be divided into four categories. Wong et al.¹ proposed an approach based on the use of an equation to describe the unknown plasma TAC. Parameters of this equation are estimated simultaneously with the estimation of the parameters of the tissue tracer kinetic model using as data measured tissue TACs in several regions-of-interest (ROIs). This approach is computationally expensive, may not converge, and requires knowledge of the functional form of the plasma TAC.² Others have used an image-derived input function (IDIF) to estimate the plasma TAC from measured PET data (e.g. see literature^3–7). IDIFs, however, are biased due to partial-volume effects and cannot be used when no large arterial blood pool is in the field of view. Their performance is highly dependent on the accuracy of partial volume correction and placement of the arterial ROIs. Furthermore, they require a well-defined relationship between whole blood total activity and plasma parent tracer activity, which is particularly difficult for tracers with significant amounts of labeled metabolites in blood. A third approach, applicable to some receptor-binding tracers, utilizes a reference tissue region in which there is no significant specific binding of the ligand (see, e.g. Wong et al.⁸). The fourth group of methods is based on the use of population-derived input functions (PDIFs) obtained by averaging measured arterial plasma TACs, normalized by injected dose and a factor assumed proportional to total blood volume (usually body weight), over many subjects.^5,7,9,10 For any individual subject, the plasma TAC is then computed from the PDIF using that individual’s injected dose and body weight. IDIFs and PDIFs have frequently been used in conjunction with arterial or venous samples for calibrating individual plasma TACs,^4–7,9 but venous samples cannot provide good calibration for tracers that exhibit substantial differences in arterial and venous activities. Due to arterial-venous differences, Zanotti-Fregonara et al.,⁵ for example, concluded that PDIF and IDIF for the tracer [¹¹C](R)-rolipram provided satisfactory results only when arterial samples were used for calibration.

The suitability of PDIFs and IDIFs depends not only on the kinetics of the specific tracer, but also on the analysis method used to determine the parameter of interest. Many of the studies utilizing PDIFs or IDIFs were performed using graphical methods, e.g. the Patlak^11,12 or Logan plot,¹³ for estimating the macroparameters irreversible uptake rate constant (K_i) or volume of distribution (V_T), respectively; these graphical methods are known to be robust with respect to errors in the input function.¹⁴ Estimation of individual parameters of the tracer kinetic model is more sensitive to input function errors,¹⁴ but some studies have successfully estimated individual kinetic parameters with a calibrated PDIF or IDIF with, e.g., [¹¹C]flumazemil,³ [carbonyl-¹¹C]WAY-100635,⁴ and [¹⁸F]fluoride.⁷

In the current study, we developed and validated a method to determine regional rates of cerebral protein synthesis (rCPS) with the L-[1-¹¹C]leucine PET method without the need for arterial sampling. In the case of L-[1-¹¹C]leucine, three input functions, i.e. time courses of the concentrations of arterial plasma [¹¹C]leucine, arterial blood total ¹¹C, and arterial blood ¹¹CO₂, are required for estimation of individual parameters of the kinetic model for leucine. In addition, the concentration of unlabeled leucine in arterial plasma is required for determination of rCPS.¹⁵ We tested a PDIF-based approach with venous samples for calibration to estimate kinetic parameters at the voxel level. rCPS was then computed with the concentration of unlabeled leucine in venous plasma substituting for arterial measurement. In order to assess accuracy of the method, rCPS values thus obtained were compared with those estimated with the measured arterial data, the gold standard for L-[1-¹¹C]leucine PET analysis.

Materials and methods

Kinetic model

The kinetic model used to describe the behavior of leucine in brain (Figure 1) has been described in detail¹⁶ and is summarized briefly below.

Figure 1.

Homogeneous tissue kinetic model for L-[1-¹¹C]leucine. Total concentration of ¹¹C (C_T*) includes free L-[1-¹¹C]leucine and L-[1-¹¹C]leucine incorporated into tissue protein (C_E* and P*, respectively), plus activity in blood (V_bC_b*, where V_b is the fraction of the volume occupied by blood and C_b* is the concentration of activity in whole blood). Total ¹¹C also includes labeled products of L-[1-¹¹C]leucine metabolism: ¹¹CO₂ and products of ¹¹CO₂ fixation. K₁ and k₂ are the rate constants for transport of leucine from plasma to tissue and back, respectively. k₃ is the rate constant for the first two steps in leucine catabolism, transamination and decarboxylation, which yield ¹¹CO₂. k₄ and k₅ are the rate constants for leucine incorporation into protein and release of free leucine from proteolysis, respectively. It is assumed that (i) there is no significant breakdown of labeled product during the experimental interval, i.e. k₅P*∼0; (ii) fixation of ¹¹CO₂ during the experimental period is negligible;^30,31 and (iii) equilibration of diffusible ¹¹CO₂ between brain and blood is rapid.³⁰ The concentration of ¹¹CO₂ in tissue is approximated by V_DC_c*, where C_c* is the ¹¹CO₂ concentration in whole blood and V_D is the brain:blood equilibrium distribution volume of ¹¹CO₂. V_D was fixed at 0.41, the value measured in rhesus monkeys,²⁴ which is in agreement with the mean whole brain:plasma equilibrium distribution volume determined from ¹¹CO₂ studies in humans.²⁵ The blood:plasma equilibrium distribution volume is close to unity.²⁶ The model assumes the tissue voxel is homogeneous with respect to concentrations of amino acids, blood flow, rates of transport and metabolism of amino acids, and rates of incorporation into protein.¹⁵ The model used to describe labeled leucine also describes unlabeled leucine (bottom row), except that unlabeled leucine and protein are in steady state, and the steady-state breakdown of unlabeled protein (k₅P) is greater than zero. Under the assumption that there is no isotope effect, the rate constants for labeled and unlabeled leucine are identical.

At any time t after tracer injection, the concentration of [¹¹C]leucine in arterial plasma is $C_{p}^{*} (t)$ . Total concentration of ¹¹C in the volume under examination, $C_{T}^{*} (t)$ , includes free [¹¹C]leucine and [¹¹C]leucine incorporated into protein in tissue ( $C_{E}^{*} (t)$ and P*(t), respectively) as well as activity in blood (V_b $C_{b}^{*} (t)$ , where V_b is the fraction of the volume occupied by blood and $C_{b}^{*} (t)$ is the total ¹¹C in whole blood). $C_{T}^{*} (t)$ also includes the labeled products of [¹¹C]leucine metabolism: ¹¹CO₂, and products of ¹¹CO₂ fixation. K₁ and k₂ are rate constants for transport of leucine from plasma to tissue and back, respectively. k₃ is the rate constant for the first two steps in leucine catabolism, transamination and decarboxylation, which yield ¹¹CO₂. k₄ and k₅ are rate constants for leucine incorporation into protein and for release of free leucine from proteolysis, respectively. Because of the long average half-life (>3 days) of brain protein, the rate of breakdown of labeled tissue protein, k₅P*(t), is small. We therefore assume k₅P*(t) = 0 during the experimental interval. We also assume negligible fixation of ¹¹CO₂ during the experimental period and that diffusible ¹¹CO₂ in brain rapidly equilibrates with arterial blood, i.e. its concentration can be approximated by $V_{D} C_{CO 2}^{*} (t)$ , where $C_{CO 2}^{*} (t)$ is the ¹¹CO₂ activity in whole blood and V_D is the brain:blood equilibrium distribution volume of ¹¹CO₂.

Therefore

\begin{matrix} C_{T}^{*} (t) = (1 - V_{b}) [C_{E}^{*} (t) + P^{*} (t) + V_{D} C_{CO 2}^{*} (t)] \\ + V_{b} C_{b}^{*} (t) \end{matrix}

(1)

In the standard approach, time courses of $C_{p}^{*} (t)$ , $C_{b}^{*} (t)$ and $C_{CO 2}^{*} (t)$ are measured from arterial blood samples. From the model of Figure 1, the previous assumptions, and equation (1), it follows that $C_{T}^{*} (t)$ can be expressed in terms of the rate constants as

\begin{matrix} C_{T}^{*} (t) = (1 - V_{b}) [(\frac{K_{1} (k_{2} + k_{3})}{k_{2} + k_{3} + k_{4}}) \\ \times \int_{0}^{t} C_{p}^{*} (τ) e^{- (k_{2} + k_{3} + k_{4}) (t - τ)} d τ] \\ + (1 - V_{b}) [(\frac{K_{1} k_{4}}{k_{2} + k_{3} + k_{4}}) \\ \times \int_{0}^{t} C_{p}^{*} (τ) d τ + V_{D} C_{c}^{*} (t)] + V_{b} C_{b}^{*} (t) \end{matrix}

(2)

The physiological parameter of interest, the rate of cerebral protein synthesis rCPS, can be then computed from the estimated rate constants as

rCPS = \frac{K_{1} k_{4}}{k_{2} + k_{3} + k_{4}} C_{p}

(3)

where C_p is the arterial plasma concentration of unlabeled leucine, which is assumed to be in a steady state and, therefore, unchanged during the study.

From the above equations, the input functions we need in order to estimate rate constants of the kinetic model for [¹¹C]leucine are $C_{p}^{*} (t)$ , $C_{b}^{*} (t)$ , and $C_{CO 2}^{*} (t)$ . Of these, the term $C_{p}^{*} (t)$ has the most influence on estimates of the rate constants as it represents the free [¹¹C]leucine in arterial plasma available for transport into brain. The term (1-V_b)V_D $C_{c}^{*} (t)$ describes the ¹¹CO₂ in brain; it represents a small correction to total activity (equation (2)), on the order of a few percentage points.¹⁷ The term V_b $C_{b}^{*} (t)$ describes the contribution of activity in the blood to the total ¹¹C in the region under analysis. It initially comprises a significant fraction of the total, but decreases to <10% by 10 min after injection of radiotracer and remains low for the balance of the study.¹⁷ After rate constants have been estimated, rCPS is determined as the product of the ratio of rate constants shown in equation (3) above and C_p, the concentration of unlabeled leucine in arterial plasma.

PET studies

All studies were conducted according to the Declaration of Helsinki 1975 (and as revised in 1983). All procedures were approved by the National Institutes of Health Combined Neurosciences Institutional Review Board, the National Institutes of Health Radioactive Drug Research Committee, and the National Institutes of Health Radiation Safety Committee. All subjects gave written informed consent. The subjects were monetarily compensated for research-related discomforts and inconveniences.

Subjects were male aged 18–24 years old, except one healthy volunteer test subject who was 27 years old on the date of the scan. Group 1, from whom the PDIFs were created, consisted of 25 healthy volunteer subjects who were scanned in the awake state with arterial blood sampling. Three groups of test subjects, distinct from Group 1, were studied for comparison of venous-calibrated PDIFs with measured arterial input functions. Test subjects consisted of 13 healthy volunteers who were studied awake (Group 2); six of these subjects were also studied on a separate day under dexmedetomidine sedation (Group 3). An additional four healthy volunteers were studied only under dexmedetomidine sedation (Group 3). One additional group of test subjects consisted of five subjects with fragile X syndrome studied under dexmedetomidine sedation (Group 4). In Groups 2, 3 and 4, both arterial and venous samples were drawn and analyzed to assess the utility of the proposed methodology. Subjects with fragile X syndrome were unable to tolerate arterial blood sampling while awake; hence, there were no arterial and venous data comparisons in this group. Physiological variables for each group studied are summarized in Table 1.

Table 1.

Subjects studied.

	Group 1	Group 2	Group 3	Group 4
	Healthy controls	Healthy controls	Healthy controls	Fragile X
	(n = 25)	(n = 13)	(n = 10)	(n = 5)
Gender	Male	Male	Male	Male
Condition	Awake	Awake	Sedated	Sedated
Age (yr)	22.1 ± 1.3	21.9 ± 2.1	21.7 ± 1.5	21.0 ± 2.5
Height (m)	1.80 ± 0.11	1.77 ± 0.07	1.76 ± 0.05	1.78 ± 0.10
Weight (kg)	84.2 ± 16.3	79.2 ± 9.9	76.7 ± 10.2	74.4 ± 15.3
BMI (kg/m²)	25.8 ± 3.2	25.1 ± 2.1	24.8 ± 2.7	23.8 ± 6.4
Blood volume (mL)^a	5109 ± 649	4921 ± 443	4815 ± 291	4953 ± 607
Arterial plasma leucine conc. (nmol/mL)	111 ± 20	130 ± 28	116 ± 11	111 ± 13
Injected dose L-[1-¹¹C]leucine (mCi/kg)	0.324 ± 0.060	0.302 ± 0.008	0.306 ± 0.024	0.290 ± 0.022

Note: Values are mean ± SD for the number of subjects indicated.

Blood volume is calibrated to estimated lean body weight (see Friesen³² and Nadler et al.³³).

The acquisition protocol was identical to that previously described,^15,16,18 except for the addition of venous blood sampling. We summarize it here briefly. Arterial lines were placed in the radial artery of the non-dominant arm and in Groups 2–4, intravenous lines also were placed in the contralateral antecubital fossa. Studies were performed on the ECAT high resolution research tomograph (HRRT) (CPS Innovations, Knoxville, TN). After optimal positioning of subjects within the field of view, a 6-min transmission scan was obtained for attenuation correction. A 90-min emission scan was initiated coincident with the intravenous infusion of 14–31 mCi of L-[1-¹¹C]leucine administered by a computer-controlled infusion pump at a continuous rate over 2 min. Data were acquired in list mode and reconstructed by means of the motion-compensated 3D ordinary Poisson ordered subset expectation maximization (OSEM) algorithm (30 subsets, 2 iterations).¹⁹ Spatial resolution after reconstruction was approximately 2.6 mm full width at half maximum (FWHM) in radial and transverse directions.¹⁹ Three-dimensional data were reconstructed to 207 slices 1.23 mm thick with a pixel size of 1.21 × 1.21 mm. Images were reconstructed as 42 frames of data (16 × 15 s, 4 × 30 s, 4 × 60 s, 4 × 150 s, 14 × 300 s). With this reconstruction, motion-correction was based on position data collected throughout the scan with the Polaris system,²⁰ and the data were corrected for attenuation based on a single attenuation-correction map (MuMap) applied to all frames of data. To correct for any residual motion, we proceeded as follows: (1) Summed frames 1–17 (0–4.5 min) and aligned the summed image to the MuMap; (2) Determined transformation matrices for frames 18–42 (4.5–90 min post injection) by aligning each frame to the summed image; (3) Applied the inverse of each transformation from the previous step to the original MuMap to obtain individual MuMaps for frames 18–42; (4) Reconstructed the data again utilizing the original motion-correction together with individual MuMaps for each frame; (5) Aligned frames 18–42 with the sum of the first 17 frames using the transformations defined in Step 2. These steps assured not only that the data from each frame were aligned to the initial frames, but also that a correct attenuation correction was applied.

Arterial blood sampling was initiated concurrently with the start of the [¹¹C]leucine infusion. Arterial blood samples were hand drawn continuously (∼one sample every 9–10 s) for the first 4 min, and at increasing intervals thereafter for a total of ∼40 samples per study. Concentrations of unlabeled and labeled leucine in arterial plasma and total ¹¹C and ¹¹CO₂ activities in arterial blood were measured according to the methods detailed previously.¹⁷ For subjects in Groups 2–4, venous samples were withdrawn and the concentrations of unlabeled and labeled leucine in venous plasma and total ¹¹C activity in venous blood were measured according to the methods detailed previously.¹⁷ Venous samples were collected 15, 30, 45 and 60 min after tracer injection. Control venous samples were also taken prior to tracer injection to measure the initial concentration of plasma leucine.

PDIFs

In order to determine PDIFs, time courses C_p*(t) and C_b*(t), measured in arterial blood samples of the subjects of Group 1 (the “population” group), were linearly interpolated between samples to provide values at 1 s intervals. The interpolated data for each subject were time shifted so that the maximum value of C_p*(t) was achieved at 140 s, the most frequently occurring peak time in our subjects. Each time course was then divided by the individual’s injected dose of L-[1-¹¹C]leucine in mCi/kg body weight. We also calculated the time course of the ratio

R^{*} (t) = C_{{CO}_{2}}^{*} (t) / C_{b}^{*} (t)

(4)

in order to express ¹¹CO₂ in arterial blood as a fraction of the total activity in arterial blood. Values of C_p*(t), C_b*(t), and R*(t) were then averaged over all the subjects in the group to create the PDIFs C_p* ^PDIF (t), C_b* ^PDIF (t), and R* ^PDIF (t). Note that each of the values in the PDIFs is based on measured arterial blood data in the population subjects. A flowchart of the procedure is given in Supplemental Figure 1.

Venous calibration of the PDIFs

The PDIFs described above were then applied to the kinetic analyses of subjects in the test groups for assessment of efficacy in determination of rCPS. Timed venous blood samples were taken during the PET scan of each test subject, as described above, and venous concentrations of plasma [¹¹C]leucine and total ¹¹C in blood were measured. Denote these concentrations by C_pv*(t_i) and C_bv*(t_i), respectively, where the subscript v stands for venous and t_i is the time at which the venous blood sample was drawn. Procedures for obtaining venous-calibrated PDIFs are described below and a flowchart is given in Supplemental Figure 2. The steps are as follows.

Figure 2.

Plasma L-[1-¹¹C]leucine concentrations. (a) Measured arterial plasma [¹¹C]leucine concentrations in 25 healthy control subjects in Group 1, normalized (divided by) injected dose of L-[1-¹¹C]leucine/kg body weight. Solid lines represent individual subjects. Dashed black line is the mean value for the group, i.e. the PDIF C_p* ^PDIF (t). Tracer was administered at a constant rate over the first two minutes of the study. Individual subjects’ curves were time shifted so that the peak arterial plasma concentration of [¹¹C]leucine occurs at 140 s. (b–c) Measured arterial (red circles) and venous (blue squares) plasma [¹¹C]leucine concentrations, VcPDIF1 (purple line), and VcPDIF2 (green line) for two subjects in Group 2. These subjects are identified as HM7A and HM2A in the Supplemental Data. (b) Both VcPDIF1 and VcPDIF2 agree well with the arterial measurements. Whole brain rCPS was 1.38, 1.40 and 1.34 nmol/g/min when determined with the arterial input function, VcPDIF1 and VcPDIF2, respectively. (c) VcPDIF1 underestimates the peak of the arterial input function, but agrees well with the arterial measurements from about 5 min to the end of the study. VcPDIF2 agrees well with the arterial input function during the first 5 min, but then somewhat overestimates it in the middle part of the scan, approximately 5–30 min. Whole brain rCPS was 1.71, 1.95 and 1.70 nmol/g/min when determined with the arterial input function, VcPDIF1 and VcPDIF2, respectively. The first min are shown on enlarged time scale in the left column. Note the different scale on the ordinate of graphs in the left and right columns.

The first step is to scale the PDIF functions describing arterial plasma [¹¹C]leucine concentration, C_p* ^PDIF (t), and arterial blood total ¹¹C concentration, C_b* ^PDIF (t). Each is multiplied by the individual test subject’s injected dose per kg body weight (S)

S = injecteddose (mCi) / kgbodyweight

(5)

We now have the scaled PDIF functions SC_p* ^PDIF (t) and SC_b* ^PDIF (t).

The second step is to examine the relationship between the test subject’s measured venous plasma [¹¹C]leucine concentrations, i.e. C_pv*(t_i), and the scaled population arterial plasma [¹¹C]leucine concentration, i.e. SC_p* ^PDIF (t). This is done by calculating their ratio (r_i)

r_{i} (t_{i}) = C_{pv}^{*} (t_{i}) / [{SC}_{p}^{* PDIF} (t_{i})], i = 1, 2, \dots

(6)

at each time t_i for which a venous sample was withdrawn. A linear least squares estimation is then used to fit a straight line to the ratios r_i (t_i), i.e. the parameters α₀ and α₁ of the function

f (t) = α_{1} t + α_{0}

(7)

are estimated from data points (t₁,r₁), (t₂,r₂),… Finally, the line is used to calibrate the scaled PDIF to the test subject’s ratios for all values of t between 0 and 60 min. This yields the estimated venous-calibrated PDIF of arterial plasma [¹¹C]leucine, denoted by C_p*^VcPDIF(t), i.e.

C_{p}^{* VcPDIF} (t) = {SC}_{p}^{* PDIF} (t) \times f (t)

(8)

The third step is to repeat Step 2 with the venous samples of blood total ¹¹C concentration, C_bv*(t_i), and the scaled population arterial blood total ¹¹C concentration, i.e. SC_b* ^PDIF (t), to estimate the venous-calibrated PDIF of arterial blood total ¹¹C concentration denoted by C_b* ^VcPDIF (t).

The fourth step is to use the population average ratio of ¹¹CO₂ to the total ¹¹C concentration in blood to estimate the test subject’s ¹¹CO₂ concentration in blood, denoted by C_CO₂* ^VcPDIF (t)

C_{CO 2}^{* VcPDIF} (t) = R^{* PDIF} (t) \times C_{b}^{* VcPDIF} (t)

(9)

where R*^PDIF(t) is the population average of R*(t), given in equation (4), and C_b*^VcPDIF(t) was calculated in Step 3.

Plasma leucine concentration

In addition to the rate constants estimated with the venous-calibrated PDIFs, an estimate of the arterial plasma leucine concentration, i.e. C_p, is required for determining rCPS (equation (3); Flowchart in Supplemental Figure 3). As an estimate of C_p, the time-weighted average of the measured concentrations of unlabeled leucine in all venous plasma samples of the test subject, including the control sample taken prior to tracer injection, was used.

Figure 3.

Concentration of ¹¹C and ¹¹CO₂ in whole blood. (a) Measured arterial whole blood total ¹¹C concentrations in 25 healthy control subjects in Group 1, normalized (divided by) injected dose of L-[1-¹¹C]leucine/kg body weight. Solid lines represent individual subjects. Dashed black line is the mean value for the group, i.e. the PDIF C_b* ^PDIF (t). Individual subjects’ curves were time shifted so that the peak arterial plasma concentration of [¹¹C]leucine occurs at 140 s. (b) Concentration of arterial blood ¹¹CO₂ as a fraction of the total concentration of ¹¹C in arterial blood. This is the factor R* ^PDIF (t) of equation (4). (c, e) Measured arterial (red circles) and venous (blue squares) whole blood total ¹¹C concentrations, VcPDIF1 (purple line), and VcPDIF2 (green line) for two subjects in Group 2 (HM7A and HM2A). Measured arterial and venous concentrations, as well as VcPDIF1 and VcPDIF2 have been multiplied by the estimated whole brain blood volume from BFM estimations using the respective input functions. In the subject shown in (c) estimated V_b was 0.05, 0.07, and 0.06 when determined with the arterial input function, VcPDIF1 and VcPDIF2, respectively. Corresponding estimates for the subject shown in (e) were 0.06, 0.13, 0.12, respectively. The higher estimated blood volumes for the inputs VcPDIF1 and VcPDIF2 result from a substantially underestimated whole blood concentrations C_b* ^VcPDIF ¹ (t) and C_b* ^VcPDIF ² (t) during the initial part of the study; the products V_bC_b* ^VcPDIF (t) V_bC_b* ^VcPDIF (t) and V_bC_b*(t), however, are very similar. These products represent the activity in the region that is intravascular. Also shown for scale is the whole brain TAC (solid blue line). (d,f) Concentration of whole blood ¹¹CO₂, i.e. C_CO₂* ^VcPDIF (t), determined by VcPDIF1 (purple line) and VcPDIF2 (green line) for two subjects in Group 2 (HM7A and HM2A). These concentrations have been multiplied by the brain:blood equilibrium distribution volume of ¹¹CO₂, V_D, whose value was fixed at 0.41 as measured in rhesus monkeys.²⁴ The products V_DC_CO₂* ^VcPDIF ¹ (t), V_DC_CO₂* ^VcPDIF ² (t) and V_DC_CO2*(t) represent the activity in the region that is in the form of ¹¹CO₂. The whole brain TAC is shown for scale (solid blue line).

ROI definition

All subjects underwent a T1-weighted MRI of the brain for region of interest (ROI) placement. ROIs were placed on each subject’s MRI by visually identifying anatomic landmarks, manually outlining the region, and constructing a binary mask to identify voxels in the region. Nine ROIs and whole brain were analyzed for each subject. To coregister the MRI and PET images, a three-dimensional volume was constructed from the average of the emission data acquired between 30 and 60 min. This volume was isotropically smoothed with a Gaussian filter (FWHM 3 mm) and aligned to the MRI volume by use of the flexible image registration toolbox³⁴ with a 3D rigid body transformation. The resliced average 30- to 60-min PET image was visually reviewed for correct alignment with the MRI by use of Vinci software (Volume Imaging in Neurological Research, Max-Planck-Institute for Neurological Research, Cologne, Germany). The transformation parameters were then applied to each frame of the PET study (without prior smoothing) to effect their alignment with the MRI volume. After the PET data had been resliced to match the MRI, kinetic model parameters and rCPS were estimated for each voxel within the brain.

Rate constant estimation

For each test subject, the rate constants (K₁, k₂ + k₃, k₄) and the blood volume (V_b) were estimated at the voxel level using the basis function method (BFM)¹⁶ as modified.^21,22 BFM utilizes the kinetic model described above which assumes that the region under analysis is kinetically homogeneous. While kinetic heterogeneity is present in at least some voxels in high-resolution L-[1-¹¹C]leucine PET data,²² its impact on calculated rCPS is negligible under the conditions of our study.²¹ The first 60 min of emission data, i.e. 36 frames of data, were utilized in the estimations.²³ V_D in equation (1) was fixed at the value of 0.41 measured in rhesus monkeys,²⁴ which is in agreement with the mean whole brain:plasma equilibrium distribution volume determined from ¹¹CO₂ studies in humans.²⁵ The blood:plasma equilibrium distribution volume is close to unity.²⁶ The basic idea of BFM is that, if the sum of the parameters β = k₂ + k₃ + k₄ were known, equation (2) would be linear in the coefficients; these coefficients could be quickly calculated using standard weighted linear least squares. For each β, the linear coefficients of equation (2) were estimated and the corresponding weighted residual sum of squares (WRSS) was computed. The value of β and the coefficients at which WRSS is minimal were then used to compute the model parameters. One hundred values equally spaced between β_min = 0.0001 min⁻¹ and β_max = 0.5 min⁻¹ were used as the grid for β, as in Tomasi et al.,¹⁶ as this widely covers the range of physiologic values expected for k₂ + k₃ + k₄.¹⁷

Estimates of the rate constants and rCPS for each ROI were computed by averaging voxel-wise estimates over all voxels in the ROI. rCPS values obtained for each subject and ROI from equation (3) were determined with the measured arterial input functions and with the venous calibrated PDIFs. With VcPDIFs, the time-weighted average of the measured concentrations of unlabeled leucine in all venous plasma samples of the test subject, including the control sample, was used as an estimate of C_p. Rates of cerebral protein synthesis determined with arterial functions were regarded as the “gold standard” against which rCPS determined with venous-calibrated PDIFs were compared.

Statistical analyses

We tested for effects of input function method on computed rCPS by means of repeated measures (RMs) ANOVA with RM on both regions and input function method. Each group/condition was examined separately. In the case of a statistically significant interaction, we further probed for differences by means of paired t-tests with rCPS computed from the measured arterial input function as the “control” and Dunnett’s correction for multiple comparisons. We used SPSS, Version 21 (IBM, Armonk, NY, USA) for all statistical analyses.

Results

PDIFs

Time courses of arterial plasma [¹¹C]leucine concentrations, normalized by injected dose per kg body weight, for the 25 subjects in Group 1 are shown in Figure 2(a). Individual time courses have been time-shifted so that the peak concentration occurs at 140 s. Also shown is the mean of the 25 individual time courses, which is the PDIF for arterial plasma [¹¹C]leucine concentration, i.e. C_p* ^PDIF (t).

Individual time courses of blood total ¹¹C concentrations and the fraction of total ¹¹C in blood that is in the form of ¹¹CO₂ are shown in Figure 3(a) and (b). The means of these time courses are C_b* ^PDIF (t) and R* ^PDIF (t), respectively.

Comparison of arterial and venous blood samples

Agreement between arterial and venous blood samples is summarized in Table 2. All venous samples in the subjects from Groups 2–4 are compared with arterial sample values interpolated to the time at which the venous samples were drawn. The difference between venous and arterial concentrations is considered to be the “error” in the venous samples.

Table 2.

Concentration difference between arterial and venous blood samples.^a

Time	Plasma leucine			Plasma [¹¹C]leucine			Whole blood total ¹¹C
(min)	Mean ± SD	RMSE	n	Mean ± SD	RMSE	n	Mean ± SD	RMSE	n
	Group 2: Male healthy controls scanned awake
0	1% ± 6%	6%	13
15	1% ± 4%	4%	13	−10% ± 7%	12%	13	−11% ± 7%	13%	13
30	1% ± 3%	3%	13	−1% ± 6%	6%	13	−3% ± 5%	6%	13
45	0% ± 5%	5%	13	4% ± 5%	6%	13	−1% ± 4%	4%	13
60	0% ± 4%	4%	13	7% ± 7%	10%	13	1% ± 2%	2%	13
Avg^b	1% ± 4%	4%	13
	Group 3: Male healthy controls scanned under sedation
0	2% ± 8%	8%	5
15	−1% ± 3%	3%	10	−4% ± 8%	9%	10	−12% ± 16%	20%	10
30	−2% ± 2%	3%	11	−1% ± 1%	1%	11	−4% ± 3%	5%	10
45	−1% ± 2%	2%	11	1% ± 5%	5%	11	1% ± 9%	9%	11
60	−2% ± 2%	3%	11	2% ± 6%	6%	11	2% ± 8%	7%	11
Avg^b	0% ± 1%	1%	11
	Group 4: Male fragile X subjects scanned under sedation
0	−4% ± 3%	5%	5
15	1% ± 5%	4%	6	−1% ± 3%	3%	6	−3% ± 3%	4%	6
30	−1% ± 5%	5%	7	3% ± 3%	4%	7	−2% ± 2%	2%	7
45	−2% ± 4%	4%	7	0% ± 5%	4%	7	0% ± 4%	4%	7
60	1% ± 2%	2%	6	5% ± 5%	7%	6	0% ± 3%	2%	5
Avg^b	−1% ± 2%	3%	7

RMSE: root mean squared error.

Mean ± SD of individual subjects' [(Venous conc. – Arterial conc.)/Arterial conc.] × 100%

Time-weighted average, 0–60 min.

In all groups, we found excellent agreement between the time-weighted mean arterial and venous plasma concentrations of leucine (root mean squared error (RMSE) <5%). We also found very good agreement between arterial plasma concentrations of [¹¹C]leucine and venous samples acquired at 15, 30, 45, and 60 min in the sedated groups (Groups 3, 4) (RMSE < 10%). In the awake subjects (Group 2), there was very good agreement between arterial plasma concentrations of [¹¹C]leucine and venous samples acquired at 30, 45, and 60 min (RMSE < 10%), but somewhat higher error in the 15 min samples (RMSE 12%). Agreement between arterial and venous blood total ¹¹C concentration at 15, 30, 45, and 60 min was excellent in Group 4 (RMSE < 5%) and good at 30, 45, and 60 min in Groups 2 and 3 (RMSE < 10%). Errors in the 15 min venous sample concentrations of blood total ¹¹C were somewhat higher in Groups 2 and 3 (RMSE of 13% and 20%, respectively).

Venous-calibrated PDIFs

Because of relatively worse agreement between arterial and venous [¹¹C]leucine in plasma and total ¹¹C in blood in the 15-min samples in Group 2, we considered two venous-calibrated PDIFs. VcPDIF1 utilized venous samples taken 15, 30, 45 and 60 min after tracer injection to calibrate the PDIFs; VcPDIF2 utilized the 30, 45 and 60 min venous samples. Plasma [¹¹C]leucine concentrations for two subjects from Group 2 are illustrated in Figure 2(b) and (c). Shown are measured arterial and venous plasma samples along with the VcPDIF1 and VcPDIF2 functions. In the first subject (Figure 2(b)), differences between arterial and venous samples of plasma [¹¹C]leucine concentrations were <1%, 2%, −3%, and −6%, respectively, at 15, 30, 45, and 60 min. Both VcPDIF1 and VcPDIF2 showed good agreement with measured arterial samples, and rCPS determined with VcPDIF1 and VcPDIF2 also showed good agreement with rCPS determined with the arterial input function. In the second subject (Figure 2(c)), differences between arterial and venous samples in plasma [¹¹C]leucine concentrations were −6%, <1%, <1%, −12%, respectively, at 15, 30, 45, and 60 min. Due to the low venous sample plasma [¹¹C]leucine concentration at 15 min, VcPDIF1 substantially underestimates the arterial curve in the crucial first few minutes around the peak and, as a consequence, rCPS determined with VcPDIF1 is 14% higher in whole brain than rCPS determined with the measured arterial input function. With VcPDIF2, the 15-min sample was not used in calibrating the PDIFs and we found better agreement with measured arterial samples in the initial minutes of the study. Between ∼3 and 30 min, however, VcPDIF2 is somewhat higher than the measured arterial samples, yet rCPS determined with VcPDIF2 is in excellent agreement with that determined with the measured arterial input function. This suggests that rCPS estimation may be less sensitive to errors in the input function during this interval, or, conversely, accuracy of the input function for determination of rCPS is likely to be more critical around the time of peak plasma tracer concentration and in the latter half of the scan.

Graphs of blood total ¹¹C and blood ¹¹CO₂ are shown in Figure 3. Subjects are the same as in the previous figure. Concentrations ¹¹C in blood (Figure 3(c) and (e)) have been multiplied by blood volume, V_b; the product V_bC_b*(t) represents activity in blood in the volume under examination. Concentrations of ¹¹CO₂ (Figure 3(d) and (f)) have been multiplied by the brain:blood equilibrium distribution volume of ¹¹CO₂, V_D; the product V_DC_c*(t) represents extravascular ¹¹CO2 in the tissue. Also shown in these graphs are the measured whole brain TACs to illustrate the relatively small magnitude of the vascular ¹¹C and extravascular ¹¹CO₂ terms compared to the total ¹¹C measured in the brain region. The estimation process is not very sensitive to errors in these terms due to their relatively small contribution to the total ¹¹C. In the first subject (Figure 3(c) and (d)), differences between arterial and venous samples of total ¹¹C in blood were −3% at 15 min and <1% in absolute value at 30, 45, and 60 min. Whole brain V_b was 0.05, 0.07, and 0.06 when estimated with the arterial input function, VcPDIF1 and VcPDIF2, respectively. Contributions to these factors to total activity in brain by VcPDIF1 and VcPDIF2 showed good agreement with that of the measured arterial samples. In the second subject (Figure 3(e) and (f)), the difference between arterial and venous samples of blood total ¹¹C was −8% and −4%, respectively, at 15 and 30 min, and <1% in absolute value at 45 and 60 min. As a consequence of the lower values in the venous samples of the blood total ¹¹C at 15 and 30 min, the VcPDIF1 and VcPDIF2 input functions were substantially underestimated in the first half of the study. This was reflected in the estimates of whole brain V_b which were 0.06, 0.13, and 0.12 when estimated with the arterial input function, VcPDIF1 and VcPDIF2, respectively. In the first several minutes of the scan, the contribution of the activity in blood to total activity in whole brain, however, was very similar with all input functions. Although activity in the blood after the initial peak, and of ¹¹CO₂ in the tissue, was somewhat overestimated when the venous-calibrated PDIFs were used, their contribution to total activity in brain remained small.

Rates of cerebral protein synthesis

To assess how the input functions VcPDIF1 and VcPDIF2 affect calculated rates of cerebral protein synthesis, we compared rCPS in whole brain and nine regions computed with measured arterial input functions with those computed with the two venous-calibrated PDIFs (VcPDIF1 and VcPDIF2) in subject Groups 2–4. With both venous-calibrated PDIFs, mean venous plasma concentration of leucine in the individual subject was used as a substitute for that subject’s mean arterial plasma concentration in the computation of rCPS. Results are shown in Figure 4 and in Supplemental Table. Kinetic model parameters and rCPS estimated in individual subjects are provided as Supplemental Data. Figure 5 illustrates parametric images of rCPS determined with each of the input functions for a healthy male subject scanned awake.

Figure 4.

Effects of input function used on computed values of rCPS in whole brain and nine regions. We compared two venous-calibrated PDIFs (VcPDIF1 and VcPDIF2) with arterial input functions in three groups of subjects. VcPDIF1 utilized venous samples taken 15, 30, 45 and 60 min after tracer injection to calibrate the PDIFs; VcPDIF2 utilized 30, 45 and 60 min venous samples for calibration. Both venous-calibrated PDIFs were used together with the individual subjects’ measured venous plasma concentration of unlabeled leucine for determination of rCPS. WB: whole brain; Fcx: frontal cortex; Ccx: cingulate cortex; Pcx: parietal cortex; Cau: caudate; Pu: putamen; Th: thalamus; Hi: hippocampus; Am: amygdala; CR: corona radiata. Points represent means ± SDs for 13 healthy control subjects studied awake, 10 healthy control subjects studied sedated with dexmedetomidine, and 5 FrX subjects studied sedated with dexmedetomidine. Estimates of each parameter were analyzed for statistically significant effects by means of RM ANOVA with region and input function as within-subjects variables. (a) Awake healthy controls. The region × input function interaction (F_2,25 = 2.08, P = 0.029) was statistically significant. Post hoc t-tests indicate that in all regions, rCPS computed with the VcPDIF1 was statistically significantly higher than computed with arterial input function (P < 0.05). There were no statistically significant differences between rCPS computed with VcPDIF2 and the arterial input function. (b) Sedated healthy controls. The region × input function interaction (F_4,35 = 0.763, P = 0.555) was not statistically significant, neither was the main effect of input function (F_2,35 = 0.763, P = 0.231). (c) Sedated fragile X subjects. Neither the region × input function interaction (F_2,8 = 1.80, P = 0.227) nor the main effect of input function (F_1,8 = 1.74, P = 0.247) were statistically significant. As expected, in all groups, the main effect of region was statistically significant.

Figure 5.

Effects of input function on computed values of rCPS in a healthy 23-year-old male scanned while awake. The injected dose of L-[1-¹¹C]leucine was 25 mCi (0.30 mCi/kg) (a) MRI; (b) rCPS computed with the measured arterial input function; (c) rCPS computed with the venous-calibrated population-derived input function that utilizes 15, 30, 45 and 60 min venous samples for calibration together this subject’s measured venous plasma leucine concentration; (d) rCPS computed with the venous-calibrated population-derived input function that utilizes 30, 45 and 60 min venous samples for calibration together this subject’s measured venous plasma leucine concentration. Images are in the transverse plane and the frontal cortex is outlined. Images of rCPS computed with the venous-calibrated population-derived input functions (c, d) are almost indistinguishable from the rCPS image determined from the measured arterial input function (b).

In general, there was good agreement between rCPS determined with the arterial input functions and with the two venous-calibrated PDIFs (Supplemental Table). In Group 2, awake healthy control subjects, RMSE in rCPS ranged from 7–11% in the regions evaluated when determined with VcPDIF1. This was the poorest agreement among the comparisons made. In Group 2, RMSE was smaller (6–9%) when determined with VcPDIF2. In the sedated healthy control subjects of Group 3, RMSE was 6–7% and 4–9% when determined with VcPDIF1 and VcPDIF2, respectively. In the sedated fragile X subjects of Group 4, the corresponding RMSEs were 3–5% and 4–6%.

In the awake group, rCPS determined with both VcPDIF1 and VcPDIF2 tended to have somewhat higher inter-subject variability than rCPS determined with the measured arterial input function. The coefficient of variation in whole brain and cortical regions, for example, was about 10% when determined with the arterial input function, and 13–14% when determined with VcPDIF1 and VcPDIF2 (Supplemental Table). This effect was smaller in the sedated groups.

Estimates of rCPS were analyzed for statistically significant effects by means of RM ANOVA with region and input function as within-subjects variables. In awake healthy controls (Group 2), the region × input function interaction (F_2,25 = 2.08, P = 0.029) was statistically significant. Post hoc t-tests indicate that in all regions rCPS computed with VcPDIF1 was statistically significantly higher than rCPS computed with the measured arterial input function (P < 0.05). There were no statistically significant differences between rCPS computed with VcPDIF2 and the arterial input function. In the sedated healthy controls (Group 3), the region × input function interaction (F_4,35 = 0.763, P = 0.555) was not statistically significant, neither was the main effect of input function (F_2,35 = 0.763, P = 0.231). In the sedated fragile X subjects (Group 4), neither the region × input function interaction (F_2,8 = 1.80, P = 0.227) nor the main effect of input function (F_1,8 = 1.74, P = 0.247) was statistically significant. As expected, in all groups, the main effect of region was statistically significant.

Discussion

We present here validation of venous-calibrated PDIFs for determination of rCPS with L-[1-¹¹C]leucine PET in young adult males. These results, coupled with our earlier study which validated use of a shorter scan duration,²³ expand the utility of this PET method to the study of subjects who previously could only undergo the procedure under sedation. This is a significant advance for the study of protein synthesis in patients with intellectual disabilities. Dysregulation of protein synthesis in neural tissue is implicated in several neurodevelopmental disorders in which intellectual disability is a major symptom. Examples include fragile X syndrome, tuberous sclerosis, and mutations in the phosphatase and tensin homolog (PTEN) gene. It has also been suggested that dysregulation of protein synthesis may be a core phenotype in idiopathic forms of autism.²⁷ Measurements of rCPS in patients with these diseases would help to identify the relevance of protein synthesis dysregulation as a core phenotype, form the basis for novel treatment strategies, and might be used to monitor treatment efficacy in research studies.

Measurements of rCPS in studies of awake age-matched healthy young men with the L-[1-¹¹C]leucine PET method are reproducible and have low variability.¹⁷ In healthy young men, rCPS is also unaffected by propofol anesthesia.²⁸ This latter finding had suggested that subjects unable to tolerate the PET scanning procedures while awake, such as those with neurodegenerative or neurodevelopmental disorders, could be studied under sedation. The sedative agent, however, can have an effect on rCPS in the patient group different from that seen in healthy controls.²⁹ We have now pursued an alternative strategy to increase the likelihood that patients can tolerate scanning without sedation: reducing the total time required for scanning and eliminating the need for arterial blood sampling. The L-[1-¹¹C]leucine PET method as originally described and validated was based on a 90-min scanning interval,²⁴ and we have demonstrated that the scanning interval can be reduced to 60 min without loss of precision in rCPS determinations.²³ In the present study, we have validated a method to eliminate arterial blood sampling and utilize instead a limited number of venous samples for the determination of rCPS.

In the case of the L-[1-¹¹C]leucine PET method, arterial blood samples were required to determine the input functions plasma [¹¹C]leucine, blood total ¹¹C, and blood ¹¹CO₂ concentration TACs. In addition, the arterial plasma leucine concentration was needed to determine rCPS. In the present study, we compared measurements in sampled blood from the radial artery of the non-dominant arm and from intravenous lines placed in the contralateral antecubital fossa. In general, arterial and venous samples agreed very well except for the 15 min samples of plasma [¹¹C]leucine concentration and blood total ¹¹C in the healthy control groups. The reason for higher errors in the venous samples acquired at 15 min in these two groups is unclear, and further data are needed to ascertain whether our findings were due to measurement error or a true difference in arterial and venous concentrations at that time point. With the possible exception of samples acquired at 15 min, however, these data support the use of venous samples to calibrate PDIFs and to compute rCPS.

We generated PDIFs from a group of heathy male young men who were scanned in the awake state, and tested the efficacy of using the venous-calibrated PDIFs for determining rCPS in three additional groups of subjects. The population group and the three test groups were all male and matched for age, but not for state (awake or sedated) or diagnosis (healthy control or fragile X syndrome). Despite differences in diagnoses and states, rCPS determined in the test groups with PDIFs calibrated with venous samples drawn between 30 and 60 min agreed well with rCPS determined with the measured arterial input functions. To date, we do not have sufficient data in young female subjects to determine whether the population input functions from male subjects can be utilized. Nor do we have data in older subjects or data acquired following different tracer administration protocols to examine population input functions in these groups. Clearly, before applying this method to measuring rCPS without arterial sampling, population input functions relevant to the group under study must be validated. We further note that the method developed for L-[1-¹¹C]leucine does not apply to other PET tracers that have different relationships between the arterial and venous time courses, e.g. time to reach arteriovenous equilibrium or significant metabolism of the tracer in tissue of the arm and hand.

Finally, in the current study in which we had both arterial and venous blood samples, we noted that there were some sporadic venous samples in which the plasma [¹¹C]leucine concentration differed more than usual from the arterial value. The arterial concentration in these cases was consistent with the more densely sampled arterial curve, so the larger difference was more likely due to some unknown sampling or measurement error in the venous sample. Nevertheless, we used all venous samples in our data analyses since the errors would not have been apparent without the complete arterial data. More frequently sampled venous data and/or use of replicate measurements would lessen the impact of erroneous venous data points in calibrating the PDIFs.

We conclude that PDIFs, calibrated by venous samples taken between 30 and 60 min, together with venous plasma leucine concentration measurements in each individual subject, can be used in place of measured arterial input functions for determination of rCPS with the L-[1-¹¹C]leucine PET method in young adult males. Because of sensitivity of venous-calibrated PDIFs to errors in venous measurements, we would recommend that the number of venous samples in this interval be increased and/or replicate measurements be considered in future studies.

Supplemental Material

Supplemental material for Substitution of venous for arterial blood sampling in the determination of regional rates of cerebral protein synthesis with L-[1-¹¹C]leucine PET: A validation study

Supplemental material for Substitution of venous for arterial blood sampling in the determination of regional rates of cerebral protein synthesis with L-[1-¹¹C]leucine PET: A validation study by Giampaolo Tomasi, Mattia Veronese, Alessandra Bertoldo, Carolyn B Smith and Kathleen C Schmidt in Journal of Cerebral Blood Flow & Metabolism

Supplemental Material

Supplementary Table -Supplemental material for Substitution of venous for arterial blood sampling in the determination of regional rates of cerebral protein synthesis with L-[1-¹¹C]leucine PET: A validation study

Supplemental material, Supplementary Table for Substitution of venous for arterial blood sampling in the determination of regional rates of cerebral protein synthesis with L-[1-¹¹C]leucine PET: A validation study by Giampaolo Tomasi, Mattia Veronese, Alessandra Bertoldo, Carolyn B Smith and Kathleen C Schmidt in Journal of Cerebral Blood Flow & Metabolism

Supplemental Material

Suppplementary Figures Flowcharts -Supplemental material for Substitution of venous for arterial blood sampling in the determination of regional rates of cerebral protein synthesis with L-[1-¹¹C]leucine PET: A validation study

Supplemental material, Suppplementary Figures Flowcharts for Substitution of venous for arterial blood sampling in the determination of regional rates of cerebral protein synthesis with L-[1-¹¹C]leucine PET: A validation study by Giampaolo Tomasi, Mattia Veronese, Alessandra Bertoldo, Carolyn B Smith and Kathleen C Schmidt in Journal of Cerebral Blood Flow & Metabolism

Footnotes

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This research was supported by the Intramural Research Program, National Institute of Mental Health and the research grant from the Wellcome Trust “Strategic Award: Inflammation in AD and MDD”. MV is also supported by the National Institute for Health Research (NIHR) Biomedical Research Centre at South London and Maudsley NHS Foundation Trust and King’s College London.

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Authors’ contributions

GT, MV, AB, CBS and KCS made substantial contribution to the study design and data interpretation, and critically revised the article. GT, MV and KCS made substantial contribution to the data analysis. GT, CBS and KCS drafted the manuscript. All authors approved the final draft submitted for publication.

Supplementary material

Supplementary material for this paper can be found at the journal website:

References

Wong

Feng

Meikle

, et al. Simultaneous estimation of physiological parameters and the input function – in vivo PET data. IEEE Transac Inform Technol Biomed 2001; 5: 67–76.

Ogden

Zanderigo

Choy

, et al. Simultaneous estimation of input functions: an empirical study. J Cereb Blood Flow Metab 2010; 30: 816–826.

Mourik

Lubberink

Klumpers

, et al. Partial volume corrected image derived input functions for dynamic PET brain studies: methodology and validation for [11C]flumazenil. Neuroimage 2008; 39: 1041–1050.

Hahn

Nics

Baldinger

, et al. Combining image-derived and venous input functions enables quantification of serotonin-1A receptors with [carbonyl-¹¹C] WAY-100635 independent of arterial sampling. Neuroimage 2012; 62: 199–206.

Zanotti-Fregonara

Hines

Zoghbi

, et al. Population-based input function and image-derived input function for [(11)C](R)-rolipram PET imaging: methodology, validation and application to the study of major depressive disorder. Neuroimage 2012; 63: 1532–1541.

Chen

Bandy

Reiman

, et al. Noninvasive quantification of the cerebral metabolic rate for glucose using positron emission tomography, 18F-fluoro-2-deoxyglucose, the Patlak method, and an image-derived input function. J Cereb Blood Flow Metab 1998; 18: 716–723.

Cook

Lodge

Marsden

, et al. Non-invasive assessment of skeletal kinetics using fluorine-18 fluoride positron emission tomography: evaluation of image and population-derived arterial input functions. Eur J Nucl Med 1999; 26: 1424–1429.

Wong

Gjedde

Wagner

. Quantification of neuroreceptors in the living brain. I. Irreversible binding of ligands. J Cereb Blood Flow Metab 1986; 6: 137–146.

Vriens

de Geus-Oei

Oyen

, et al. A curve-fitting approach to estimate the arterial plasma input function for the assessment of glucose metabolic rate and response to treatment. J Nucl Med 2009; 50: 1933–1939.

10.

Tahara

Saito

Sugimoto

, et al. AVP-induced mitogenic responses of Chinese hamster ovary cells expressing human V1A or V1B receptors. Pflugers Arch 1999; 437: 219–226.

11.

Patlak

Blasberg

. Graphical evaluation of blood-to-brain transfer constants from multiple-time uptake data: generalizations. J Cereb Blood Flow Metab 1985; 5: 584–590.

12.

Patlak

Blasberg

Fenstermacher

. Graphical evalution of blood-to-brain transfer constants from multiple-time uptake data. J Cereb Blood Flow Metab 1983; 3: 1–7.

13.

Logan

. Graphical analysis of PET data applied to reversible and irreversible tracers. Nucl Med Biol 2000; 27: 661–670.

14.

Zanotti-Fregonara

Chen

Liow

J-S

, et al. Image-derived input function for brain PET studies: many challenges and few opportunities. J Cereb Blood Flow Metab 2011; 31: 1986–1998.

15.

Schmidt

Cook

Qin

, et al. Measurement of regional rates of cerebral protein synthesis with L-[1-¹¹C]leucine and PET with correction for recycling of tissue amino acids: I. Kinetic modeling approach. J Cereb Blood Flow Metab 2005; 25: 617–628.

16.

Tomasi

Bertoldo

Bishu

, et al. Voxel-based estimation of kinetic model parameters of the L-[1-¹¹C]leucine PET method for determination of regional rates of cerebral protein synthesis: validation and comparison with region-of-interest-based methods. J Cereb Blood Flow Metab 2009; 29: 1317–1331.

17.

Bishu

Schmidt

Burlin

, et al. Regional rates of cerebral protein synthesis measured with L-[1-¹¹C]leucine and PET in conscious, young adult men: normal values, variability, and reproducibility. J Cereb Blood Flow Metab 2008; 28: 1502–1513.

18.

Veronese

Bertoldo

Bishu

, et al. A spectral analysis approach for determination of regional rates of cerebral protein synthesis with the L-[1-(11)C]leucine PET method. J Cereb Blood Flow Metab 2010; 30: 1460–1476.

19.

Carson

Barker

Liow

J-S

, et al. Design of a motion-compensation OSEM list-mode algorithm for resolution-recovery reconstruction for the HRRT. IEEE Nucl Sci Symp Conf Record 2003; 5: 3281–3285.

20.

Bloomfield

Spinks

Reed

, et al. The design and implementation of a motion correction scheme for neurological PET. Phys Med Biol 2003; 48: 959–978.

21.

Veronese

Bertoldo

Tomasi

, et al. Impact of tissue kinetic heterogeneity on PET quantification: case study with the L-[1-¹¹C]leucine PET method for cerebral protein synthesis rates. Sci Rep 2018; 8: 931.

22.

Veronese

Schmidt

Smith

, et al. Use of spectral analysis with iterative filter for voxelwise determination of regional rates of cerebral protein synthesis with L-[1-¹¹C]leucine PET. J Cereb Blood Flow Metab 2012; 32: 1073–1085.

23.

Tomasi G, Veronese M, Bertoldo A, et al. Effects of shortened scanning intervals on calculated regional rates of cerebral protein synthesis determined with the L-[1-¹¹C]leucine PET method. PLoS One. Epub ahead of print 16 April 2018. DOI: 10.1371/journal.pone.0195580.

24.

Smith

Schmidt

Qin

, et al. Measurement of regional rates of cerebral protein synthesis with L-[1-¹¹C]leucine and PET with correction for recycling of tissue amino acids: II. Validation in rhesus monkeys. J Cereb Blood Flow Metab 2005; 25: 629–640.

25.

Brooks

Lammertsma

Beaney

, et al. Measurement of regional cerebral pH in human subjects using continuous inhalation of ¹¹CO₂ and positron emission tomography. J Cereb Blood Flow Metab 1984; 4: 458–465.

26.

Gunn

Yap

Wells

, et al. A general method to correct PET data for tissue metabolites using a dual-scan approach. J Nucl Med 2000; 41: 706–711.

27.

Kelleher

Bear

. The autistic neuron: troubled translation? Cell 2008; 135: 401–406.

28.

Bishu

Schmidt

Burlin

, et al. Propofol anesthesia does not alter regional rates of cerebral protein synthesis measured with L-[1-¹¹C]leucine and PET in healthy male subjects. J Cereb Blood Flow Metab 2009; 29: 1035–1047.

29.

Qin

Schmidt

Zametkin

, et al. Altered cerebral protein synthesis in fragile X syndrome: studies in human subjects and knockout mice. J Cereb Blood Flow Metab 2013; 33: 499–507.

30.

Buxton

Alpert

Babikian

, et al. Evaluation of the ¹¹CO₂ positron emission tomographic method for measuring brain pH. I. pH changes measured in states of altered pCO₂. J Cereb Blood Flow Metab 1987; 7: 709–719.

31.

Siesjo

Thompson

. The rate of incorporation of gaseous ¹⁴CO₂ into brain tissue constituents. Experientia 1964; 20: 98–99.

32.

Friesen

. Lean-scaled weight can be used to estimate blood volume for obese patients. Cand J Anesthesiol 2014; 61: 1059–1060.

33.

Nadler

Hidalgo

Bloch

. Prediction of blood volumes in normal human adults. Surgery 1962; 51: 224–232.

34.

Fischer

Modersitzki

. Intensity-based image registration with a guaranteed one-to-one point match. Methods Inf Med. 2004; 43: 327–330.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.11 MB

0.08 MB

0.23 MB

Substitution of venous for arterial blood sampling in the determination of regional rates of cerebral protein synthesis with L-[1- 11 C]leucine PET: A validation study

Abstract

Keywords

Introduction

Materials and methods

Kinetic model

PET studies

PDIFs

Venous calibration of the PDIFs

Plasma leucine concentration

ROI definition

Rate constant estimation

Statistical analyses

Results

PDIFs

Comparison of arterial and venous blood samples

Venous-calibrated PDIFs

Rates of cerebral protein synthesis

Discussion

Supplemental Material

Supplemental material for Substitution of venous for arterial blood sampling in the determination of regional rates of cerebral protein synthesis with L-[1-11C]leucine PET: A validation study

Supplemental Material

Supplementary Table -Supplemental material for Substitution of venous for arterial blood sampling in the determination of regional rates of cerebral protein synthesis with L-[1-11C]leucine PET: A validation study

Supplemental Material

Suppplementary Figures Flowcharts -Supplemental material for Substitution of venous for arterial blood sampling in the determination of regional rates of cerebral protein synthesis with L-[1-11C]leucine PET: A validation study

Footnotes

Funding

Declaration of conflicting interests

Authors’ contributions

Supplementary material

References

Supplementary Material

Supplemental material for Substitution of venous for arterial blood sampling in the determination of regional rates of cerebral protein synthesis with L-[1-¹¹C]leucine PET: A validation study

Supplementary Table -Supplemental material for Substitution of venous for arterial blood sampling in the determination of regional rates of cerebral protein synthesis with L-[1-¹¹C]leucine PET: A validation study

Suppplementary Figures Flowcharts -Supplemental material for Substitution of venous for arterial blood sampling in the determination of regional rates of cerebral protein synthesis with L-[1-¹¹C]leucine PET: A validation study