Time sensitivity-corrected retention index: an enhanced metabolic index from 18F-FDG PET to differentiate between benign and malignant pulmonary nodules.

The aim of the study was to evaluate pulmonary nodules (PNs) by incorporating time sensitivity (S) factor in the retention index (RI) and compare with the traditional fixed interval method. After obtaining approval from the Human Investigations Committee, 97 PNs from 81 patients (age=70±11) referred for dual-time fluorine-18 fluorodeoxyglucose PET (16.1±1.9 mCi) with definite pathological diagnosis or 1-year computed tomography follow-up were retrospectively studied. S=d{ln[SUV]]/d{ln[T]} was obtained by logarithmic regression using scan times, T (0, 1, 2), and standard uptake value (SUV) (0, 1, 2). This time-corrected RI, RIs=[(T2/T1)-1]×100%, was compared with traditional fixed time interval RI, RIx=[(SUV2/SUV1)-1]×100%, by means of receiver operating characteristic curve analysis. The mean±SD of T1 and T2 (72.3±14.0 and 134.9±17.6 min, respectively) skewed markedly from the intended time of PET scans (skewness=2.076 and 1.356, respectively). There were 27 benign tumors, 37 cases of non-small-cell lung cancer, 15 other types of cancer, and 18 stable lesions by 1-year computed tomography follow-up. There were significant differences between the nonmalignant group (NM, n=45) and the cancer group (CA, n=52) in time sensitivity (0.186±0.161 vs. 0.483±0.180, P<0.0005) and RIs (12.7±12.5 vs. 37.4±17.5%, P<0.0005). The RIx showed wider variation than RIs, although the difference between NM and CA was also significant (18.0±28.8 vs. 37.8±32.0%, P=0.002). The RIs and RIx were only weakly correlated (r=0.257, P=0.011). Receiver operating characteristic curve analysis performed for the CA or NM groups revealed a significant improvement in the diagnostic accuracy for malignancy by RIs (area under the curve=0.880±0.035, P<0.0005) compared with RIx (area under the curve=0.694±0.054, P=0.001). Incorporating the time sensitivity factor improves the diagnostic performance of RI for malignant PNs by using additional biologic information from the variation in fluorine-18 fluorodeoxyglucose uptake times and rates.

Fluoro-2-deoxy-D-glucose positron emission tomography/computed tomography predicts extrahepatic metastatic potential of colorectal metastasis: a practical guide for yttrium-90 microsphere liver-directed therapy.

The objective of this retrospective study was to assess the likelihood of extrahepatic metastases based on tumor metabolic load index (TMLI) for patients with colorectal liver metastases to determine the potential intermediate endpoint of yttrium-90 (Y-90) microsphere liver-directed therapy. Forty-eight (48) patients with colorectal metastatic cancer of the liver who were referred for Y-90 microsphere therapy and F-18 fluoro-2-deoxy-D-glucose positron emission tomography (PET) imaging were included. All patients had baseline computed tomography, hepatic angiography, and planning intra-arterial technetium-99m macro-aggregated albumin scans. Pretreatment PET images were analyzed by visual inspection of extrahepatic metastases and by computer quantification of total liver tumor metabolism. For each patient, regions of interest were drawn along the liver edge to measure total liver standard uptake value on axial images, covering the entire span of the liver. The total liver standard uptake value was then converted by logarithm in equivalent volumes of liver mass to obtain TMLI for comparison. A Levene test for equality of variances and t-tests were used for comparing pretreatment TMLIs of patients with or without extrahepatic metastasis. Discriminant and receiver operating characteristic (ROC) analyses were used to obtain a cutoff value with highest specificity in predicting negative extrahepatic metastasis. There were 21 and 27 patients identified as negative and positive for extrahepatic metastasis, respectively. The TMLI of the group with negative extrahepatic metastasis was significantly lower than that with positive extrahepatic metastasis (10.22 + 0.32 versus 10.74 + 0.57, p < 0.0005). The cutoff TMLI with 100% specificity was found to be 10.65. There was a significant difference in liver tumor load with respect to the presence or absence of an extrahepatic metastatic tumor as evaluated objectively with PET. This leads to the identification of TMLI threshold, below which extrahepatic metastases are unlikely and thus may provide guidance for Y-90 therapy.

Volume-normalized uptake rates with robust transportability from PET dual-time and Patlak analyses.

PURPOSE: The intention here is to enhance the usefulness of the Gjedde-Patlak plot of dynamic positron emission tomography (PET) tracer uptake. Two additional parameters closely related to the physiologically significant and diagnostically useful phosphorylation rate k (3) are therefore studied. Additionally, their inter-institutional transportability is examined. METHODS: The two traditional parameters obtained from a Patlak plot are its slope Ki and its usually ignored tissue/plasma (=Q/Cp) axis intercept V. As a useful result, a normalized uptake rate may be defined as k=Ki /V. This is can be theoretically close to k (3). Similar to this an alternative normalized uptake rate is defined as k (3)' =Ki /V '. Here, V ' would be a composite of model rate constants, reasonably known a priori, and the measured V so as to depend less on errors in the latter. Parameter determination demonstrations utilize data from the 2-deoxy-2-[F-18]fluoro-D-glucose(FDG)-PET literature. RESULTS: Using median k (i) values from 24 FDG dynamic studies and algebraic relationships, on average: k=1.07 k (3)(r=0.97), and k (3)' =0.95k (3) (r=0.91). A skeletal muscle case also demonstrates agreements with k (3). For liver malignancies k and k (3)' can be diagnostically slightly superior to Ki. Unaffected by institutionally dependent Q and Cp calibrations and methods, these can be more robust than Ki in a number of circumstances. CONCLUSION: Two studied physiologically meaningful parameters, close to the diagnostically important k (3), can supplement Ki and enhance Patlak analysis by appropriately utilizing normally ignored information. Hitherto, k (3) was obtainable only by complex nonlinear least squares compartmental model analysis. The additional parameters can have more robust inter-institutional transportability than Ki.

Time sensitivity: a parameter reflecting tumor metabolic kinetics by variable dual-time F-18 FDG PET imaging.

OBJECTIVES: The aim of this study was to define and investigate the time sensitivity of tumors by variable dual-time fluorodeoxyglucose positron emission tomography (FDG PET). METHODS: Variable dual-time (t) protocol (P) FDG PET-computed tomography (CT) scans from 40 patients with pathologically proven head and neck tumors without brain metastasis were analyzed. The first protocol (P.I) consisted of 26 patients with early (E) and delayed (D) PET-CT obtained at 106 +/- 15 and 135 +/- 16 min after injection of 16.3 +/- 1.9 mCi FDG. The second protocol (P.II) recruited 14 patients with E- and D-PET performed at 54 +/- 13 and 151 +/- 28 min after injection of 9.6 +/- 1.7 mCi FDG. The maximum standardized uptake values (SUVs) were measured in the primary tumor (CA1) and the cerebellum (CBL). The time sensitivity (S) was defined as d{ln(SUV)}/d{ln(t)} and its value was obtained by linear regression of ln(D-SUV/E-SUV) vs ln(t (D)/t (E)). Patients with cerebellar variations greater than 30% in SUV between E- and D-PET was excluded from the analysis. RESULTS: Two patients from P.I were excluded due to wide cerebellar SUV variations. D-SUV were significantly higher than E-SUV in CA1 for both P.I (18.9 +/- 6.9 vs 14.8 +/- 5.6, p < 0.0005) and P.II (11.5 +/- 7.9 vs 9.7 +/- 6.9, p = 0.013). The S values for CA1 in P.I and P.II were 0.67 and 0.17, respectively. The D-SUV were also higher than E-SUV in CBL for both P.I (12.5 +/- 1.6 vs 11.6 +/- 1.6, p < 0.0005) and P.II (7.6 +/- 1.6 vs 7.0 +/- 1.6, p = 0.008). The S values for CBL in P.I and P.II were 0.47 and 0.04, respectively, which were over 1.4-fold smaller than that of CA1, suggesting fundamental kinetic differences between CA1 and CBL. CONCLUSIONS: The time sensitivity factor reflects another kinetic parameter of tumor metabolism besides SUV when using variable dual-time FDG PET. It offers another useful diagnostic tool in optimizing choices of dual-time protocols for oncologic PET-CT and in reducing SUV variations due to time interval differences with corrections using S.

Optimizing dual-time and serial positron emission tomography and single photon emission computed tomography scans for diagnoses and therapy monitoring.

INTRODUCTION: A region's early and late tracer uptake activities, QE and QL, within a dual-time scan (i.e. using two frames) or in serial scans (as for monitoring therapeutic response), are popular quantitative diagnostic aids, especially in oncology. In this paper, maximum performance is sought from their joint use. METHODS: QL/QnE is introduced as a tumor marker with an empirical n. This generalizes traditional data weighting having n=1 for QL/QE, the retention index (RI), with its associated % difference. Using patient data, iterative guessing finds an optimal n that maximizes a measure of diagnostic performance: D=(difference of normal and abnormal marker means)/(their combined SD), which may be computed from values of QL/QnE, as well as of QL, QE, and RI each used alone. For 2-deoxy-2-[F-18]fluoro-D-glucose(FDG)-positron emission tomography (PET) dual-time protocols, another approach to optimization-selection of scan times-is investigated by simulations using the Sokolov model. RESULTS: A meta-analysis of 12 PET and single photon emission computed tomography (SPECT) studies with various tracers, cancers, and scan classes (dual-time or serial) finds ns from 0.5 to 1.1. The optimal D necessarily exceeds the best (or any) computed using QE, QL, or RI: negligibly to by as much as 0.6 (or 1.5). The increases in optimal receiver operating curve area (Az) over the best (or any) traditional marker range from negligible to 0.07 (or 0.4). QE alone usually has the lowest D and Az. Statistically significant performance improvement of QL/QnE over QE and QL is shown for most studies. Contrasting with an optimal n, another value n0 can also be found where D=0. Occasionally, n0 can be close to 1, and RI then will have a small D and poor performance. Simulation with kinetic modeling of FDG dual-time scans for liver and liver metastases demonstrates worst and best scan times. Indicated for these imaging protocols are QE at very early cellular transport associated times and QL rather late when phosphorylation/dephosphorylation dominate. Benefits from choosing optimal times in dual-time protocols, especially in combination with choosing optimal ns, can be significant. CONCLUSION: A protocol-dependent optimizing parameter n in an improved classification marker can easily be identified in a learning set of scans having normals and abnormals. Finding this parameter below 1.0 in most all studies suggests that a popularly used QL/QE may often overweight early activities. Additionally, QL/QE may sometimes be a poor marker choice and underestimate a protocol's diagnostic capability. Subsequent use of the proposed QL/QnE in settings similar to that of the learning set gives improved diagnostic performance over traditional approaches, although by widely varying amounts. Additionally, a method of seeking optimal scan times is demonstrated and suggests significant gains in dual-time protocol performances are possible.

A weight index for the standardized uptake value in 2-deoxy-2-[F-18]fluoro-D-glucose-positron emission tomography.

INTRODUCTION: Known errors in the standardized uptake value (SUV) caused by variations in subject weights W encountered can be corrected by lean body mass or body surface area (bsa) algorithms replacing W in calculations. However this is infrequently done. The aims of the work here are: quantify sensitivity to W, encourage SUV correction with an approach minimally differing from tradition, and show what improvements in the SUV coefficient of variation (cv) for a population can be expected. METHODS: Selected for analyses were 2-deoxy-2-[F-18]fluoro-D-glucose (FDG) SUV data from positron emission tomography (PET) and PET/computed tomography (CT) scans at the University of Tennessee as well as from the literature. A weight sensitivity index was defined as -n=slope of ln(SUV/W) vs. lnW. The portion of the SUV variability due to this trend is removed by using the defined [formula: see text], or a virtually equal SUVm using [formula: see text], with Q and ID being tissue specific-activity and injected dose. [formula: see text] measures performance. Adapting to animal studies' tradition, [formula: see text] is preferred over the conventional [formula: see text]. RESULTS: For FDG in adults [formula: see text] from averaging over most tissues. In children, however, [formula: see text]. Tissues have the same index if their influx constants are independent of W. Suggested, therefore, is a very simplified [formula: see text], which is dimensionless and keeps the same population averages as traditional SUVs. It achieves [formula: see text]. Hence, for cv's of SUVs below approximately 1/3 improvements over tradition are possible, leading to F's<0.95. Accounting additionally for height, as in SUVbsa, gives very little improvement over the simplified approach here and gives essentially the same F's as SUVm. CONCLUSIONS: Introduced here is a weight index useful in reducing variability and further understanding the SUV. Addressing weight sensitivity is appropriate where the cv of the SUVs is below about 1/3. Proposed is the very simple approach of using an average of an adult patient's weight and approximately 70 kg for FDG SUV calculations. Unlike other approaches the dimensionless population average of SUVms is unchanged from tradition.

Investigating the existence of quantum metabolic values in non-Hodgkin's lymphoma by 2-deoxy-2-[F-18]fluoro-D-glucose positron emission tomography.

OBJECTIVES: To investigate the existence of quantum metabolic values in various subtypes of non-Hodgkin's lymphoma (NHL). METHODS: Fifty-eight patients with newly diagnosed NHL and positron emission tomography (PET) performed within three months of biopsy were included. The standardized uptake value (SUV) from PET over the area of biopsy and serum glucose [Glc] were recorded. The group glucose sensitivity(G) for indolent and aggressive NHL was obtained by linear regression with ln(SUV) = G x ln[Glc] + C, where C is a constant for the group. Finally, the individual's glucose sensitivity (g) was obtained by g = {ln(SUV)-C}/ln[Glc], along with their means in various subtypes of NHL. To further investigate the influence of extreme [Glc] conditions, the SUVs corrected by the individually calculated g at various glucose levels, [Glc'] using SUV' =SUV x {[Glc']/[Glc]}(g), were compared to the original SUVs for both indolent and aggressive NHL. RESULTS: The averaged g (=G) for aggressive was significant different from that for indolent NHL (-0.94 +/- 0.51 vs. +0.13 +/- 0.10, respectively, p < 0.00005). There were significant differences in SUV for [Glc] < 80 or >110 mg/dl for both types of NHL. Unlike overlap among SUVs between NHL subtypes, the g value clearly categorized them into two distinct groups with positive (near-zero) and negative g values (around -1) for the indolent and aggressive NHLs, respectively. CONCLUSIONS: Distinct quantum metabolic values of -1 and 0 were noted in NHL. Aggressive NHL has a more negative value (or higher glucose sensitivity) than that of indolent and, thus, is more susceptible to extreme glucose variation.

Regional yttrium-90 microsphere treatment of surgically unresectable and chemotherapy-refractory metastatic liver carcinoma.

PURPOSE: The aim of this prospective study was to assess the safety and tumor response of intra-arterial Y-90 microspheres for the treatment of surgically unresectable and chemotherapy-refractory liver metastases. MATERIALS AND METHODS: Forty-six (46) patients with metastatic cancer to the liver from various solid tumors, with tumor progression despite polychemotherapy, were included. All patients had baseline computed tomography (CT), 18-Fluoro-2-deoxy-D-glucose-positron emission tomography (F-18 FDG-PET), hepatic angiography, and intra-arterial Tc-99m macroaggregated albumin (MAA) scan for the assessment of extrahepatic aberrant perfusion and lung shunting fraction. Twenty-seven (27) and 19 patients were treated with Y-90 glass- or resin-based microspheres (but not both), respectively, on a lobar basis and were monitored over 3 months after last treatment using dedicated attenuation corrected PET. For each patient, regions of interest (ROIs) were drawn along the liver edge to measure total liver standard uptake value (SUV) on axial images covering the entire liver for comparing pre- and post-treatment total liver SUV change. RESULTS: There was a significant decrement in total liver SUV after treatment by either glass- or resin-based microspheres (p = 0.0013 and 0.028, respectively). There was no significant difference in the amplitudes of the mean percentage reduction of tumor metabolism between these two agents (20% +/- 25% vs. 10% +/- 30% for glass- vs. resin-based microspheres; p = 0.38). None of the patients in the glass-based group developed complications, whereas 3 patients had complications related to hyperbilirubinemia (1 transient and 2 permanent) in the resin-based group. CONCLUSIONS: Results suggest that there is significant mean reduction of hepatic metastatic tumor load (metabolism), as evaluated objectively by PET after Y-90 microsphere, for the treatment of unresectable metastatic disease to the liver. The Y-90 therapy provides encouraging and safe results by arresting the progression of metastatic cancer to the liver with decreasing tumor metabolism.

A statistical investigation of normal regional intra-subject heterogeneity of brain metabolism and perfusion by F-18 FDG and O-15 H2O PET imaging.

BACKGROUND: The definite evaluation of the regional cerebral heterogeneity using perfusion and metabolism by a single modality of PET imaging has not been well addressed. Thus a statistical analysis of voxel variables from identical brain regions on metabolic and perfusion PET images was carried out to determine characteristics of the regional heterogeneity of F-18 FDG and O-15 H2O cerebral uptake in normal subjects. METHODS: Fourteen normal subjects with normal CT and/or MRI and physical examination including MMSE were scanned by both F-18 FDG and O-15 H2O PET within same day with head-holder and facemask. The images were co-registered and each individual voxel counts (Q) were normalized by the global maximal voxel counts (M) as R = Q/M. The voxel counts were also converted to z-score map by z = (Q - mean)/SD. Twelve pairs of ROIs (24 total) were systematically placed on the z-score map at cortical locations 15-degree apart and identically for metabolism and perfusion. Inter- and intra-subject correlation coefficients (r) were computed, both globally and hemispherically, from metabolism and perfusion: between regions for the same tracer and between tracers for the same region. Moments of means and histograms were computed globally along with asymmetric indices as their hemispherical differences. RESULTS: Statistical investigations verified with data showed that, for a given scan, correlation analyses are expectedly alike regardless of variables (Q, R, z) used. The varieties of correlation (r's) of normal subjects, showing symmetry, were mostly around 0.8 and with coefficient of variations near 10%. Analyses of histograms showed non-Gaussian behavior (skew = -0.3 and kurtosis = 0.4) of metabolism on average, in contrast to near Gaussian perfusion. CONCLUSION: The co-registered cerebral metabolism and perfusion z maps demonstrated regional heterogeneity but with attractively low coefficient of variations in the correlation markers.

Addressing glucose sensitivity measured by F-18 FDG PET in lung cancers for radiation treatment planning and monitoring.

PURPOSE: To address glucose sensitivity in lung cancers before and after radiation treatment (Tx). METHODS AND MATERIALS: Twelve patients were each studied with two pre-Tx positron emission tomography (PET) scans and 3 patients each with one post-Tx PET scan, with glucose concentration [Glc] and maximum standard uptake value (SUV) recorded. The pre-Tx glucose sensitivity, g from SUV1/SUV2= {[Glc]1/[Glc]2}g and Tx index, tau from SUVpost-Tx/SUVpre-Tx = {[Glc]post-Tx/[Glc]pre-Tx}tau was calculated by linear regression. Pre-Tx SUVs were corrected to post-Tx Glc with g (SUV'pre-Tx) for a pure Tx effect, R = ln(SUVpost-Tx/SUV'pre-Tx). RESULTS: There were no significant differences in SUV but [Glc] were different (96.4 +/- 10.9 vs. 88.3 +/- 10.5, p = 0.015) between two pre-Tx PET scans. Linear regression yielded g = -0.79 and tau = -1.78 to -2.41 (p < 0.0005 in all). The %DeltaSUV after Tx for 3 patients without vs. with g correction were different by -12%, 0%, and + 7%, suggesting varying effects from glucose. R values were also different and mean R (-0.81 +/- 0.38) was significantly different from zero (p = 0.03), consistent with successful Tx as confirmed by clinico-radiologic follow-up. CONCLUSIONS: The extra dimension of glucose sensitivity, g besides SUV incorporated in the combined Tx-derived tau may be a useful global Tx evaluation index even with differing [Glc].

Glucose-normalized standardized uptake value from (18)F-FDG PET in classifying lymphomas.

UNLABELLED: Our objective was to derive the best glucose sensitivity factor (g-value) and the most discriminating standardized uptake value (SUV) normalized to glucose for classifying indolent and aggressive lymphomas. METHODS: The maximum SUV obtained from (18)F-FDG PET over the area of biopsy in 102 patients was normalized by serum glucose ([Glc]) to a standard of 100 mg/dL. Discriminant analysis was performed by using each SUV(100) (SUV x {100/[Glc]}(g), calculated using various g-values ranging from -3.0 to 0, one at a time) as a variable against the lymphoma grades, and plotting the percentage of correct classifications against g (g-plot) to search for the best g-value in normalizing SUV(100) for classifying grades. To address the influence of the extreme glucose conditions, we repeated the same analyses in 12 patients with [Glc] < or = 70 mg/dL or [Glc] > or = 110 mg/dL. RESULTS: SUV(100) correctly classified lymphoma grades ranging from 62% to 73% (P < 0.0005), depending on the g-value, with a maximum at a g-value of -0.5. For the subgroup with extreme glucose values, the g-plot also revealed higher and more optimal discrimination at a g-value of -0.5 (92%) than at a g-value of 0 (83%) (P = 0.03). The discrimination deteriorated at g < -1 in both analyses. The box plot for all cases using a g-value of -0.5 showed little overlap in classifying lymphoma grades. For a visually selected threshold SUV(100) of 7.25, the sensitivity, specificity, and accuracy of identifying aggressive grades were 82%, 79%, and 81%, respectively. CONCLUSION: The results suggest that metabolic discrimination between lymphoma grades using a glucose-normalized SUV from (18)F-FDG PET is improved by introducing g-value as an extra degree of freedom.

2-deoxy-2-[F-18]fluoro-D-glucose-positron emission tomography sensitivity to serum glucose: a survey and diagnostic applications.

OBJECTIVE: The positron emission tomography (PET) clinical utility of the sensitivity (gamma) of uptake (Q) to a change in plasma glucose concentration (C) is investigated. METHODS: Gamma is obtained from data as [ln(Q (2)/Q (1))] / [ln(C(2)/C(1))], using previously published intrapatient studies varying C within a single patient and some interpatient ones. It can be theoretically related to the half-saturation constant in the Michaelis-Menten quantification of competitive uptake. One of its uses is making uptake corrections for desired vs. actual C using Q(2) = Q(1) (C(2)/C(1))(gamma). RESULTS: Intrapatient studies proved to be preferable to interpatient ones, and a 2-deoxy-2-[F-18]fluoro-D-glucose (FDG)-PET survey with analyses for gamma yielded the following result: usually the gamma values of tumors and brain tissues were near -1, whereas those of other noncerebral tissues were near 0. Regarding correcting uptakes for C, instead of a universally assumed and applied gamma = -1, corrections should be for a single tissue using its known gamma. An advantageous use of gamma is predicting how C affects image contrast, including where glucose loading is sometimes preferable to fasting. CONCLUSIONS: A potentially useful quantifier of uptake sensitivity to plasma glucose has been defined and values obtained. Correcting uptakes to some standard C requires special care. gamma can help PET clinicians select fasting or loading to achieve glucose levels for optimum contrast.

Degree of fractal behavior as a diagnostic aid in quantitative image analyses.

PURPOSE: Multiple strategies in diagnoses of different diseases from images can include their histogram analyses. Any fractal behavior in the latter is to be quantified as to extent here, with a view toward contributing to a diagnostic process. PROCEDURE: One tool in quantitative image analyses is the fractal dimension D of the pixel histogram, a measure of self-similarity over various scales in a fitted power-law behavior of pixel intensity cumulative probability distribution. Proposed and developed here as diagnostic markers are features of its determination process that indicate to what extent there is fractal behavior. One of these is the curvature c that exists in log-log plots used for extracting the fractal exponent D of power-law behavior. RESULTS: Specific implementations are given both for a general lognormal pixel intensity distribution and for lung images. Both Ds and cs are determined for: normals, pulmonary embolism, cystic fibrosis, as well as a theoretical lognormal distribution. It is shown that D and heterogeneity described by a standard deviation are reciprocally related and not typically independent markers. The added independent information from c has possibilities of assisting in discrimination of normal and pathologic conditions, such as in lung diseases. CONCLUSION: In addition to a histogram's fractal dimension itself, there are indications that measures of the degree of fractal behavior may also hold promise in image diagnoses.