Noninvasive measurement of tracer extraction efficiency in tissue, illustrated with Tc-99m-MAG3.

AIM: The aim of this study is to develop a noninvasive technique for measuring tissue tracer extraction efficiency ( E ) and illustrate it for Tc-99m-mercaptoacetyltriglycine (MAG3) and kidney. METHODS: E was measured in 10 patients with normal MAG3 renography. E is the ratio of tissue clearance-to-blood flow ( Ki/F ). For single-photon tracers, attenuation constants are unknown, so Ki and F cannot be separately measured. However, by deriving attenuation-uncorrected Ki' and F' from the same regions of interests (ROIs), these constants cancel out, giving E . Using a lung ROI for blood activity, F was measured from first-pass and Ki' from Gjedde-Patlak-Rutland (GPR) analysis up to 130 s. Because of interference from right ventricle, a left ventricular ROI (LV) is unsuitable for F' but was used in GPR analysis, making an adjustment for the ratio of respective blood pool signals arising from lung and LV ROIs in early frames (60-90 s). RESULTS: A lung ROI underestimates F' by 4% at normal LV function. Chest wall interstitial activity ( I ), which does not affect F' , amounted to 53 and 30% of the lung and LV signals at 20 min, and 12 and 6% at 130 s, resulting in underestimations of Ki of 4 and 2%, respectively. Ignoring these opposing errors, E based on lung ROI for left and right kidneys was 43.5 (SD 8)% and 47.3 (9)%, and based on LV ROI for GPR analysis was 44.5 (10.9)% and 48.3 (10.6)%. CONCLUSION: E can be measured by combining blood flow from first-pass with clearance from GPR analysis, and has potential value both clinically and in clinical research.

Errors in measurement of glomerular filtration rate using the slope-intercept technique and their identification.

BACKGROUND: GFR measured from plasma sampling may be expressed as slope-intercept GFR (SI-GFR) and scaled to body surface area (mGFR/BSA) or as GFR per unit extracellular fluid volume (mGFR/ECV), which is based only on half-time. Measurement errors comprise 3 categories. Pre-injection error arises from error in administered marker and is suspected when mGFR/BSA and mGFR/ECV disagree. Injection errors include 'tissued' injections. Post-injection errors include inaccurate sample timing, inaccurate pipetting, sample haemolysis and sampling through long IV lines through which marker was administered. The aim of the study was to evaluate the impact of errors on mGFR. METHODS: We compared mGFR/BSA with mGFR/ECV in 898 patients undergoing routine investigation. To investigate post-injection error, we took two further patient datasets with r values (correlation coefficient of the 3-sample fit) of 1.0 and introduced errors, in isolation, into each of the 3 recorded sample values, as follows: pipetting (volume) errors of -20%, -10%, -5%, 5%, 10% and 20%, and timing errors of -15 min, -10 min, -5 min, 5 min, 10 min and 15 min. RESULTS: The correlation between mGFR/BSA and mGFR/ECV was close and independent of r. Post-injection error depended on the time of the sample in which it occurred. r correlated poorly with error magnitude for both volume and timing errors. When a 'rogue' sample is suspected its error needed to be substantial for it to be identified by single sample estimates applied to the other samples. CONCLUSION: SI-GFR is resistant to post-injection timing and volume errors but not to pre-injection error.

Quantitative lymphoscintigraphy of the lower limbs for the diagnosis of phlebolymphoedema.

INTRODUCTION: Phlebolymphoedema is caused by the interaction of the venous and lymphatic systems in a state of chronic venous insufficiency in which increased microvascular filtration causes an increased rate of lymph production. Lymphatic drainage rate increases in response, but this is unsustainable and can cause lymphatic failure and oedema. We hypothesise that in phlebolymphoedema we could measure unusually high lymphatic drainage while the lymph system is still fully functional. METHOD: Patients referred for lymphoscintigraphic investigation of swollen legs between April 2021 and December 2022 were reviewed. Quantitative lymphoscintigraphy was performed following the technique of Keramida et al . (2017) and ilio-inguinal nodal uptake (IIQ%) was calculated. The presence of scintigraphic features of increased lymph production was noted for each limb. RESULTS: A total of 39 patients were reviewed (78 limbs, 29F, 10M). Seven limbs were identified with supranormal lymphatic function (IIQ > 30%) plus three borderline. Of these 10 limbs, all had at least two scintigraphic features of increased lymph production. CONCLUSION: Quantitative lymphoscintigraphy, although developed for diagnosing abnormally low lymphatic function, may also have utility at the upper end of the spectrum for identifying chronic venous insufficiency. An IIQ% upper normal limit of 30% could be used to diagnose venous insufficiency as the cause for limb swelling. This is of note for patients of large body habitus in whom venous ultrasound is difficult.

Measuring myocardial blood flow with 82rubidium using Gjedde-Patlak-Rutland graphical analysis.

OBJECTIVE: Myocardial blood flow (MBF) is measured with 82Rb using non-linear, least-squares computerised modelling. The study aim was to explore the feasibility of Gjedde-Patlak-Rutland (GPR) graphical analysis as a simpler method for measuring MBF. METHODS: Patients had myocardial perfusion imaging using adenosine (n = 45) or regadenoson (n = 33) for stressing. Blood 82Rb clearance into myocytes (K1) was measured from Cedar-Sinai QPET software using the modified Crone-Renkin equation of Lortie et al. (K1 = [1-0.77 × e-B/MBF] × MBF) to convert K1 to MBF (ml/min/100 ml), where B (63 ml/min/100 ml) is myocardial permeability-surface area product. Using aorta or left ventricular cavity (LV) to measure arterial blood 82Rb concentration, blood 82Rb clearance into myocardium (Z) was measured from GPR analysis based on data acquired between 1 and 3 min post-injection. As units of K1 and Z are, respectively, ml/min/ml intracellular space and ml/min/ml total tissue including extracellular space, myocardial extracellular fluid volume (ECV) is 1 - [Z/K1]. Using Z/K1 (see Results) to modify its index, the Lortie equation was changed to Z = (1-0.77 × [Formula: see text]e-BZ/MBFZ)*MBFZ, following which MBFZ was calculated from Z. In GPR analysis, spillover of activity from LV to myocardium conveniently 'drops out' in the intercept of the plot. RESULTS: Both agents increased myocardial blood flow almost equally. ECV was ~ 35 ml/100 ml at rest, increasing to ~ 40 ml/100 ml after stress. Z/K1, averaged between stress, rest, stressing agents and arterial ROI, was 0.62, so BZ was taken as 39 (i.e. 0.62 × 63) ml/min/100 ml. Based on LV, MBFZ (y) correlated with MBF (x): y = 0.43x + 22 ml/min/100 ml; r = 0.84; n = 156). Their respective stress/rest ratios showed a moderate correlation (r = 0.64; n = 78). CONCLUSIONS: GPR analysis offers promise as a valid and analytically simpler technique for measuring myocardial blood flow, which, as with any clearance measured from GPR analysis, has units of ml/min/ml total tissue volume, and merits development as a polar map display.

Measurement of Eosinophil Kinetics In Vivo.

Radiolabeled leukocyte scans are used in nuclear medicine to detect sites of infection and inflammation. We have previously demonstrated the use of clinical grade immunomagnetic beads to isolate autologous eosinophils and image their distribution in healthy volunteers. Here we describe the use of radiolabeled eosinophils coupled to single-photon emission computed tomography (SPECT) to quantify eosinophil uptake in the lungs of healthy volunteers, patients with asthma, and patients with focal eosinophilic inflammation.

82 Rb tissue kinetics in humans.

AIM: The study aim was to compare the kinetics of the potassium analogue, 82 Rb, between spleen, liver and kidney. METHODS: Patients had myocardial stress/rest perfusion imaging using adenosine (n = 45) or regadenoson (n = 33) for stressing. Hepatic arterial (HAP), splenic (SP) and renal (RP) perfusions were measured from first-pass and blood 82 Rb clearances (Ki) from Gjedde-Patlak-Rutland graphical analysis of data between 1 and 2 min postinjection, using regions of interest over left ventricular cavity or abdominal aorta to monitor arterial concentration. Tissue 82 Rb extraction efficiency (E) was calculated as [Ki/perfusion]*100. Tissue extracellular fluid volume (ECV) was derived from the GPR plot intercept. RESULTS: SP (24%) and RP (23%) increased after regadenoson but decreased (-41% and -19%) after adenosine. HAP increased after adenosine (91%) and regadenoson (68%). Resting E was high in kidney (69%) and low in spleen (26%). After adenosine, it increased to 91% in kidney and 49% in spleen. Assuming an arterial contribution of 25% to hepatic blood flow, resting E in liver was estimated as 23%. Relationships between Ki and perfusion in spleen and kidney were consistent with the Crone-Renkin equation (Ki = [1 - A.e-B/perfusion ]*perfusion), with respective values of A of 0.95 and 0.94 and B of 31 and 186 ml/min/100 ml. Splenic ECV decreased following adenosine from 62 to 39 ml/100 ml and showed a logarithmic correlation with SP. CONCLUSION: Kidney, spleen and liver display contrasting tissue kinetics. E is high in kidney and low in spleen and liver. Spleen is erectile, collapsing when perfusion decreases.

New gender-specific formulae for estimating extracellular fluid volume from height and weight in adults.

AIMS: First, to derive gender-specific formulae for estimation of extracellular fluid volume (eECV) and second, compare eECV as a scaling metric for slope-intercept glomerular filtration rate (GFR) with estimated body surface area (eBSA), lean body mass (eLBM) and total body water (eTBW). METHODS: GFR and 'slope-only' GFR (GFR/ECV), both single compartment-corrected, were measured in a previously published multicentre database of healthy potential kidney transplant donors. Measured ECV (mECV) was obtained as ratio GFR-to-GFR/ECV. Formulae for eECV in men and women were derived from the relationship of mECV with height and weight and expressed as eECV = a.weight.height. In a population of prospective kidney transplant donors from a single centre, eECV was compared with mECV. GFR was scaled to eECV, eBSA, eLBM and eTBW, estimated from previously published formulae. RESULTS: In men and women, respectively, a was 0.0755 and 0.0399, x was 0.6185 and 0.6065 and y was 0.4982 and 0.6217. In the single centre, biases (±precisions) of eECV against mECV in men and women were 0.26 (±1.68) and 0.31 (±1.67) l. Mean GFR/eBSA was higher in men but mean GFR/eLBM and GFR/eTBW were higher in women. Mean GFR/ECV and mean GFR/eECV were very similar between the two genders. GFR/ECV and GFR/eECV showed correlations with each other that were almost identical between men and women. CONCLUSIONS: New formulae are described for estimating eECV. Scaling GFR to eECV is more physiological than scaling to eBSA and accounts for gender. eECV used for measuring GFR from a single blood sample should be gender-specific.

Correction of slope-intercept glomerular filtration rate measurement without scaling for body size.

AIM: The aim of this study was to evaluate a slope-intercept glomerular filtration rate (GFR) one-compartment correction method based exclusively on the rate constant (α2) of the exponential between 2 and 4 h post-injection that requires no scaling for BSA. METHODS: The correction factor is 1/([C.α2]+1). C depends on the difference between one-compartment-corrected and uncorrected GFR, so varies with different correction procedures. Patients were in four groups: group 1 (Cr-EDTA; n = 141) and group 2 (Tc-DTPA; n = 47) had sampling at 2, 3 and 4 h. Groups 3A (Tc-DTPA; n = 168) and 3B (Tc-DTPA; n = 361) gave nine samples up to 480 min. C was calculated from GFR corrected using Brochner-Mortensen (BM) without prior BSA-scaling (CBM; GFRBM), after BSA-scaling then reverse-scaling as per British Nuclear Medicine Society (BNMS) guidelines (CBNMS; GFRBNMS), and after correction using the equations containing 'f' described by Fleming (CFlem; GFRFlem) and Jodal and Brochner-Mortensen (CJBM; GFRJBM). In group 3A, C (C9) was determined from GFR measured from all nine samples (GFR9) and from seven samples (C7) up to 240 min. In 3B, GFRC, corrected using 1/([C9.α2]+1), was compared with GFRBM, GFRBNMS, GFRFlem and GFRJBM against GFR9 (gold-standard). RESULTS: C derived from these one-compartment correction formulae ranged from 25 to 32 min. In group 3, C7 and C9 were 28 ± 11 and 38 ± 14 min (P < 0.0001). Biases of GFRBM, GFRBNMS, GFRJBM, GFRFlem and GFRC against GFR9 were 2.7, 1.5, 4.2, 3.4 and 0.4 ml/min. Corresponding precisions were 9.3, 7.3, 7.0, 6.7 and 7.6 ml/min. CONCLUSION: Correction using α2 avoids BSA scaling, has a low bias against gold-standard GFR and does not over-correct at high GFR.

Lesson of the month: novel method to quantify neutrophil uptake in early lung cancer using SPECT-CT.

Neutrophils play an important role in the lung tumour microenvironment. We hypothesised that radiolabelled neutrophils coupled to single-photon emission CT (SPECT) may non-invasively quantify neutrophil uptake in tumours from patients with non-small cell lung cancer. We demonstrated increased uptake of radiolabelled neutrophils from the blood into tumours compared with non-specific uptake using radiolabelled transferrin. Moreover, indium-111-neutrophil activity in the tumour biopsies also correlated with myeloperoxidase (MPO)-positive neutrophils. Our data support the utility of imaging with In-111-labelled neutrophils and SPECT-CT to quantify neutrophil uptake in lung cancer.

FDG PET/CT of the non-malignant liver in an increasingly obese world population.

Because of obesity, non-alcoholic fatty liver disease (NAFLD) is becoming increasingly important. 10% of NAFLD patients develop non-alcoholic steatohepatitis (NASH), which may progress to cirrhosis and is now the leading indication for liver transplantation in the Western world. Prefibrotic NASH can only be reliably diagnosed by biopsy. However, given its success in other inflammatory diseases, PET/CT with 18 F-fluorodeoxyglucose (FDG), although non-specific, may offer a promising approach to diagnosing not only NASH but also other inflammatory liver conditions. In addition, FDG PET has generated pathophysiological information on hepatic glucose metabolism and, diagnostically, used liver for quantification of tumour FDG accumulation (e.g. Deauville scoring). A review of hepatic FDG PET is therefore timely. There are two general approaches to the quantification of hepatic FDG accumulation: firstly, standard uptake value (SUV) and secondly dynamic PET. SUV is a poor index of hepatic metabolic function because most hepatic FDG (~75%) is un-phosphorylated 60-min postinjection. Hepatic fat is increased in NAFLD but accumulates negligible FDG. Because fat distribution is heterogeneous, maximum pixel SUV is therefore preferred to mean pixel SUV. Computer modelling of dynamic PET dissects the transport constants governing hepatic FDG kinetics but is challenged by the liver's dual blood supply. Graphical analysis is less informative but more robust and will be the preferred clinical approach to measurement of hepatic FDG phosphorylation. Previous dynamic PET studies have ignored hepatic fat and therefore potentially underestimated glucose accumulation in patients with hepatic steatosis. Future work should use graphical analysis of dynamic PET and correction for hepatic fat.

Stimulation of the hepatic arterial buffer response using exogenous adenosine: hepatic rest/stress perfusion imaging.

OBJECTIVES: The hepatic arterial buffer response is a mechanism mediated by adenosine whereby hepatic arterial perfusion (HAP) increases when portal flow decreases, and is implicated in liver disease. The first study aim was to measure HAP in patients undergoing myocardial perfusion imaging (MPI), thus developing hepatic arterial rest/stress perfusion imaging (HAPI). The second aim was to compare adenosine-induced changes in splenic perfusion (SP) and HAP with corresponding changes in myocardial blood flow (MBF). METHODS: Patients had MPI with 82Rb PET/CT using adenosine (n = 45) or regadenoson (n = 33) for stressing. SP and HAP were measured using a first-pass technique that gives HAP rather than total hepatic perfusion. Renal perfusion (RP) was also measured. RESULTS: Mean MBF and HAP increased after both adenosine ([stress-rest]/rest 1.1 and 0.8) and regadenoson (1.4 and 0.6), but the respective changes did not correlate. After adenosine, SP (- 0.48) and RP (- 0.26) both decreased. The change in SP correlated positively with the change in MBF (r = 0.36; p = 0.015) but did not correlate with change in HAP. After regadenoson, SP (0.2) and RP (0.2) both increased. The changes in SP correlated with the changes in MBF (r = 0.39; p = 0.025) and HAP (r = 0.39; p = 0.02). Changes in RP correlated with changes in HAP (r = 0.51; p = 0.0008) but not MBF. Resting SP (r = 0.32; p = 0.004), but not resting HAP, correlated with hepatic fat burden. Adenosine-induced change in HAP also correlated with hepatic fat (r = 0.29; p = 0.05). CONCLUSION: HAPI could be a useful new hepatic function test. Neither splenic 'switch-off' nor hepatic arterial 'switch-on' identifies adequacy of stress in MPI. KEY POINTS: • This article describes a new method for assessing arterial perfusion of the liver and its capacity to respond to an infusion of adenosine, a substance that normally 'drives' hepatic arterial flow. • Hepatic arterial flow increased in response to adenosine, sometimes dramatically. Adenosine is already used clinically to stimulate myocardial blood flow in patients with suspected coronary disease, but the increase in flow did not correlate with the corresponding increase in hepatic arterial flow. • Analogous to the use of adenosine in the myocardium, the increase in hepatic arterial flow in response to adenosine has the potential to be a new clinically useful method for the evaluation of hepatic arterial haemodynamics in liver disease.

Whole-body 75SeHCAT retention is determined by entero-hepatic bile acid recycling rate.

AIM: It has previously been proposed that entero-hepatic bile acid recycling frequency is a major determinant of whole-body retention (WBR) of SeHCAT. Hepatocyte to terminal ileum accounts for almost the entire cycle. The study aim was to test this hypothesis by comparing WBR with an estimate of speed of transit of bile acids using Tc-HIDA scintigraphy performed on a separate occasion. METHODS: Using an un-collimated gamma camera and patient-to-camera distance of 1.5 m, WBR at 7 days after oral SeHCAT administration was measured in 14 patients with chronic diarrhoea, of whom 10 had previous cholecystectomy. The distance reached within the intestine of Tc-HIDA at 1 h (n = 14) and 2 h (n = 7) following iv injection was graded as follows: grades 1 and 2: small bowel on left and right sides of abdomen, respectively; and grade 3: colon. Relationships between WBR and grade were assessed using Spearman rank correlation. RESULTS: Interval between studies ranged from 3 to 1219 (median 330) days. Grading correlated with WBR at 1 h (rs = -0.63; P = 0.02) and weakly at 2 h (rs = -0.68; P = 0.09) post-injection of Tc-HIDA. In nine patients in whom Tc-HIDA and SeHCAT scans were performed within 1 year of each other, the correlation remained significant at 1 h (rs = -0.73; P = 0.03). There was no difference in WBR or grading between patients with or without a gall bladder. CONCLUSION: Entero-hepatic bile acid recycling frequency is a major determinant of whole-body SeHCAT retention.

Absence of hepatic activity in lymphoscintigraphy performed with Tc-99m-Nanoscan.

INTRODUCTION: For assessment of lower extremity swelling using lymphoscintigraphy, several nuclear medicine departments in the UK have recently switched from Tc-Nanocoll to Tc-Nanoscan. The aim of the study was to compare quantitative and semiquantitative features of lymphoscintigraphy between these two tracers. METHODS: Twenty patients received Tc-Nanocoll and 20 received Tc-Nanoscan either side of the switch-over in our department. Tracers were compared with respect to visualisation of the liver, para-aortic lymph nodes and urine (all graded by consensus from 0 to 3; invisible-to-very prominent); and with respect to ilio-inguinal nodal quantification and ratio of liver-to-summed bilateral ilio-inguinal nodal activity at 120 min postinjection (L/N120). Scans were deemed abnormal when there was lymph diversion through skin or deep system, no activity in ilio-inguinal nodes at 60 min or negligible ilio-inguinal activity (<5%, left plus right) at 120 min. RESULTS: Liver was visualised in 18/20 Tc-Nanocoll but only 3/20 Tc-Nanoscan scans. Moreover, para-aortic activity was less prominent after Tc-Nanoscan. Urinary activity was more prominent after Tc-Nanocoll. There were 9/20 patients with stomach activity after Tc-Nanocoll compared with 1/20 after Tc-Nanoscan. Urinary and stomach activities correlated. There was an elevated L/N120, and therefore a suspicion of peripheral lympho-venous communication, in the single Tc-Nanoscan patient who displayed prominent hepatic activity. CONCLUSION: Hepatic activity is the result of accumulation of colloidal degradation products generated in lymph nodes, rather than intact colloid. Tc-Nanoscan gives less hepatic activity than Tc-Nanocoll because it is more resistant to intranodal degradation. Peripheral lymphovenous communication remains a possible alternative route for activity to reach the liver.

Hepatic and splenic 18 F-FDG blood clearance rates (Ki) in hepatic steatosis and diabetes mellitus.

AIM: The study aim was to investigate the relationships of blood glucose level (BGL) with hepatic and splenic blood FDG clearances (Ki) in patients with diabetes mellitus (DM) and/or hepatic steatosis (HS). METHODS: This was a retrospective study of 238 patients, including 92 with type 2 DM (DM2) and 11 with type 1 DM (DM1), having routine whole body FDG PET/CT. Patients with lymphoma were excluded. Patients were divided into the following groups: HS-DM-, HS-DM2+, HS+DM-, HS+DM2+ and 2 DM1 groups (hypoglycaemic and hyperglycaemic). ROI were placed over liver and spleen for measurement of SUVmax , and left ventricular cavity (LV) for measurement of SUVmean . Tissue SUVmax was divided by LV SUVmean . This division, giving Z, eliminates bias from the whole body metric used to calculate SUV and renders SUVmax a closer surrogate of Ki. HS was diagnosed when hepatic unenhanced CT was ≤40 HU. RESULTS: In all patients, individual hepatic Z and individual splenic Z correlated significantly with individual BGL. Highest mean hepatic Z and highest mean BGL were recorded in HS+ DM2+ group and lowest in hypoglycaemic DM1 group. Patients with DM1 and hyperglycaemia showed low hepatic Z in relation to BGL. Hepatic and splenic Z correlated inversely with CT density in patients without DM but not in those with DM2. CONCLUSION: As BGL increases, hepatocyte glucokinase is up-regulated. This includes patients with HS and DM2 but not DM1. We speculate that in HS and DM2, up-regulation results from insulin resistance and hyperinsulinaemia. The data also support a hepato-splenic metabolic axis.

A new single-sample, time-flexible, empirically-derived formula for measuring glomerular filtration rate from a single blood sample.

BACKGROUND: Measurement of glomerular filtration rate from a single blood sample using filtration markers may be based on empirical or model-based formulae. The former do not require a whole body metric but are sample time-inflexible. The latter include time as a continuous variable but require a whole body metric to estimate extracellular fluid volume. This study describes a new empirical method that is time-flexible and requires no whole body metric. METHODS: Data comprise 982 slope-intercept glomerular filtration rate measurements from a single institution in which the correlation coefficient of the three-sample fit was ≥0.999. In 582 measurements, samples were taken at 118-122, 144-148, 178-182, 202-206 or 238-242 minutes. Within each of these five sample time groups, slope, k(t), of the regression of apparent volume of distribution (Vt) against slope-intercept glomerular filtration rate was recorded following exclusion of outliers. The five sampling times correlated closely and exponentially with k(t) (=5.4.e), allowing calculation of k(t) for any sampling time. In a further 341 measurements, glomerular filtration rate from a single blood sample was calculated as [k(t).V(t)] and compared with slope-intercept glomerular filtration rate. Sample time for glomerular filtration rate from a single blood sample is critical, so the most appropriately timed sample was chosen according to slope-intercept glomerular filtration rate/1.73 m. RESULTS: The current method compared favourably with three previously published model-based methods (Fleming et al., Jacobsson and Gref and Karp), showing better precision, and unlike the other three, no correlation between difference and average on Bland-Altman analysis. CONCLUSION: This completely time-flexible, empirically-derived approach to glomerular filtration rate from a single blood sample requires no whole body metric, is simple, equally applicable to children and adults and superior to three model-based methods.

Intrahepatic fluorine-18-fluorodeoxyglucose kinetics measured by least squares nonlinear computer modelling and Gjedde-Patlak-Rutland graphical analysis.

We aimed to use simple physiological equations to show similarities and differences in blood fluorine-18-fluorodeoxyglucose (F-FDG) clearance in the liver (Ki) and distribution volume (V0) in the liver, respectively, generated from nonlinear least squares computer modelling and Gjedde-Patlak-Rutland graphical analysis of dynamic F-FDG PET. We show theoretically that when, as is usually the case, vascular fraction (fraction of liver occupied by blood, Vb) is included as a parameter in modelling but ignored in graphical analysis, the ratio of Ki values, respectively, generated by graphical anlaysis (Ki) and by modelling (Ki) is equal to 1-Vb. This theoretical prediction was then confirmed from dynamic PET data acquired in a clinical population of patients undergoing routine F-FDG PET/computed tomography and from a review of the literature in which it can be seen that Ki/Ki ranges from 0.47 to 0.98. When Vb is not included as a parameter in modelling, Ki is theoretically equal to Ki and to V0·k3, and V0 is equal to V0. There are several attractions to normalising Ki to V0 with respect to liver. Thus, first, there is no need to correct imaging for photon attenuation. Second, it makes no difference whether uptake constant is expressed as blood or plasma clearance. Third, it circumvents the effects of hepatic fat, which, because it accumulates negligible F-FDG, physically dilutes the F-FDG signal and reduces both uptake constant and distribution volume.

Tissue standardized uptake value is a closer surrogate of blood fluorine-18 fluorodeoxyglucose clearance after division by blood standardized uptake value, illustrated in brain and liver.

The numerator and denominator of the left-hand side of the Gjedde-Patlak-Rutland (GPR) equation for measurement of blood fluorine-18 fluorodeoxyglucose (F-FDG) clearance into tissue (Ki) are the standardized uptake values (SUVs) of tissue and blood, respectively. The extent to which normalized time (NT) in the GPR equation exceeds real time depends on half-time of clearance of F-FDG from blood. A literature review shows that NT is fairly constant, about 100 min at 60 min postinjection of F-FDG, in keeping with our own finding of no significant difference in maximum SUV in blood 60 min postinjection of F-FDG between 39 patients with F-FDG-avid malignancy on routine PET/CT (1.74±0.31) and 21 patients with normal PET/CT (1.79±0.32), and similar blood glucose levels (BGLs). Volume of distribution (V0) in the GPR equation is ∼0.4 ml/ml for brain and ∼0.9 ml/ml for lean liver. Using these values of V0 and an NT of 100 min, we used the GPR equation to calculate Ki from our own published values of SUVliver/SUVblood and SUVbrain/SUVblood at 60 min postinjection, obtaining 0.0045 ml/min/ml for liver and 0.036 ml/min/ml for brain at BGL of 5 mmol/l. These values for Ki at this BGL are close to literature values of Ki, which for liver and brain are ∼0.0033 and ∼0.035 ml/min/ml, respectively. We conclude, therefore, that following division with blood pool SUV, tissue SUV becomes a closer surrogate of Ki. This division also eliminates the controversy over which whole body metric to use in the calculation of SUV.