Repeatability of SUV measurements in serial PET.

PURPOSE: The standardized uptake value (SUV) is a quantitative measure of FDG tumor uptake frequently used as a tool to monitor therapeutic response. This study aims to (i) assess the reproducibility and uncertainty of SUV max and SUV mean, due to purely statistical, i.e., nonbiological, effects and (ii) to establish the minimum uncertainty below which changes in SUV cannot be expected to be an indicator of physiological changes. METHODS: Three sets of measurements were made using a GE Discovery STE PET/CT Scanner in 3D mode: (1) A uniform 68Ge 20 cm diameter cylindrical phantom was imaged. Thirty serial frames were acquired for durations of 3, 6, 10, 15, and 30 min. (2) Esser flangeless phantom (Data Spectrum, approximately 6.1 L) with fillable thin-walled cylinders inserts (diameters: 8, 12, 16, and 25 mm; height: approximately 3.8 mm) was scanned for five consecutive 3 min runs. The cylinders were filled with 18FDG with a 37 kBq/cc concentration, and with a target-to-background ratio (T/BKG) of 3/1. (3) Eight cancer patients with healthy livers were scanned approximately 1.5 h post injection. Three sequential 3 min scans were performed for one bed position covering the liver, with the patient and bed remaining at the same position for the entire length of the scan. Volumes of interest were drawn on all images using the corresponding CT and then transferred to the PET images. For each study (1-3), the average percent change in SUV mean and SUV max were determined for each run pair. Moreover, the repeatability coefficient was calculated for both the SUV mean and SUV max for each pair of runs. Finally, the overall ROI repeatability coefficient was determined for each pair of runs. RESULTS: For the 68Ge phantom the average percent change in SUV max and SUV mean decrease as a function of increasing acquisition time from 4.7 +/- 3.1 to 1.1 +/- 0.6%, and from 0.14 +/- 0.09 to 0.04 +/- 0.03%, respectively. Similarly, the coefficients of repeatability also decrease between the 3 and 30 min acquisition scans, in the range of 10.9 +/- 3.9% - 2.6 +/- 0.9%, and 0.3 +/- 0.1% - 0.10 +/- 0.04%, for the SUV max and SUV mean, respectively. The overall ROI repeatability decreased from 18.9 +/- 0.2 to 6.0 +/- 0.1% between the 3 and 30 min acquisition scans. For the l8FDG phantom, the average percent change in SUV max and SUV mean decreases with target diameter from 3.6 +/- 2.0 to 1.5 +/- 0.8% and 1.5 +/- 1.3 to 0.26 +/- 0.15%, respectively, for targets from 8-25 mm in diameter and for a region in the background (BKG). The coefficients of repeatability for SUV max and SUV mean also decrease as a function of target diameter from 7.1 +/- 2.5 to 2.4 +/- 0.9 and 4.2 +/- 1.5 to 0.6 +/- 0.2, respectively, for targets from 8 mm to BKG in diameter. Finally, overall ROI repeatability decreased from 12.0 +/- 4.1 to 13.4 +/- 0.5 targets from 8 mm to BKG in diameter. Finally, for the measurements in healthy livers the average percent change in SUVmax and SUV mean were in the range of 0.5 +/- 0.2% - 6.2 +/- 3.9% and 0.4 +/- 0.1 and 1.6 +/- 1%, respectively. The coefficients of repeatability for SUV max and SUV men are in the range of 0.6 +/- 0.7% - 9.5 +/- 12% and 0.6 +/- 0.7% - 2.9 +/- 3.6%, respectively. The overall target repeatability varied between 27.9 +/- 0.5% and 41.1 +/- 1.0%. CONCLUSIONS: The statistical fluctuations of the SUV mean are half as large as those of the SUV max in the absence of biological or physiological effects. In addition, for clinically applicable scan durations (i.e., approximately 3 min) and FDG concentrations, the SUV max and SUV mean have similar amounts of statistical fluctuation for small regions. However, the statistical fluctuations of the SUVmean rapidly decrease with respect tothe SUVmax as the statistical power of the data grows either due to longer scanning times or as the target regions encompass a larger volume.

A novel respiratory tracking system for smart-gated PET acquisition.

PURPOSE: In this study, the authors validated a novel respiratory tracking device, the multidimensional respiratory tracking (MDRT) system, that was designed to assist in correcting for respiratory motion in PET/CT images. The authors also investigated a novel PET acquisition technique, smart gating (SG), that enables to acquire motion-free PET data prospectively, with minimum user interference and with no additional postprocessing of the PET data. METHODS: MDRT uses visual tracking techniques to track simultaneously the two-dimensional (in the vertical plane) motion of multiple fiducial markers using a standard video camera. A threshold window is set at the breathing amplitude of interest using the MDRT GUI. A trigger is generated at a rate of 250 Hz as long as the breathing signal is within the threshold window. The triggers are fed into the PET scanner to initialize one single bin of a gated acquisition every 4 ms. No triggers are delivered as the breathing signal drifts outside the threshold window. Consequently, PET data are acquired only whenever the breathing signal is confined within the amplitude threshold window, thus resulting into a motion-free image set. The accuracy of MDRT in tracking the breathing signal was assessed (1) by comparing the period of an oscillating phantom, as measured by MDRT, to that measured with a photogate timer and (2) by comparing the MDRT output to that of the real-time position management (RPM) in ten patients. The SG PET/CT acquisition was validated in phantoms and in two stereotactic body radiosurgery (SBRS) lung DIBH-PET/CT patients. RESULTS: MDRT was in agreement with the photogate timer in determining the period of motion to less than 2%. The percent errors between MDRT and RPM in the positions of the peaks and troughs of the ten patients' breathing signals were within 10%. In phantoms, SG technique enables to correct for motion-induced artifacts in the PET images and improve the accuracy of PET quantitation. For the SBRS application, in one patient, the patient's CT lesion was not detected in the corresponding clinical PET images, while it exhibited an SUV of 1.8 in the DIBH image set. In the second patient, DIBH-PET images showed an improved PET-to-CT spatial matching and a 52% increase in the lesion SUV. CONCLUSIONS: MDRT has been shown to be accurate in tracking breathing motion and assisted in implementing a smart-gating PET acquisition technique that allowed to acquire prospectively motion-free PET images.

Implementation and evaluation of patient-specific three-dimensional internal dosimetry.

UNLABELLED: Current methods for calculating the absorbed dose in a target region from a source region rely on a standard "reference man" geometry and assume an uniform distribution of radiolabel. While this approach is acceptable at the low levels of radioisotope administered for most diagnostic purposes, the generality of the calculations is not adequate for doses at the higher levels required for therapy and is not easily extendible to tumor dosimetry. METHODS: We have developed an integrated system which utilizes patient anatomy and radionuclide distribution in the calculation of absorbed dose rate or total dose to any user-defined target region. Images of radionuclide distribution (PET/SPECT) are registered to anatomic images (CT/ MRI) and then entered into a three-dimensional internal dosimetry software system (3D-ID) where regions of interest are defined. Dose calculations are performed by the mathematical convolution between a user-specified, dose-point kernel with the activity in the source volume over the target volume. The resulting dose rate distribution may be scaled by cumulated activity to yield absorbed dose. In addition to calculating the mean dose, dose-volume histograms may be generated which plot absorbed dose with respect to percent of volume. The method was evaluated using selected standard man phantom organs. RESULTS: Dose estimates for two patient studies are included to illustrate differences between patient-specific and MIRD-based calculations. The package provides an alternative approach to image display and three-dimensional internal dose calculations. CONCLUSION: The dose-volume histogram representation of absorbed dose to a target volume provides valuable information in assessing tumor control probability and normal tissue toxicity.

Image registration of SPECT and CT images using an external fiduciary band and three-dimensional surface fitting in metastatic thyroid cancer.

Image registration of 131I SPECT with CT scans was performed in a patient with metastatic thyroid carcinoma using an external fiduciary band and a three-dimensional surface-fitting algorithim. Areas of metastatic disease taking up 131I were accurately localized to the liver, lungs and vertebral bodies; providing information that could not be obtained by planar or SPECT images alone. Based on these findings, further invasive diagnostic procedures were not performed, therefore considerably altering management in this patient. This approach to image registration has immediate clinical utility in the registration and interpretation of SPECT studies with corresponding CT or MRI scans.

Clinical validation of SPECT and CT/MRI image registration in radiolabeled monoclonal antibody studies of colorectal carcinoma.

UNLABELLED: Registration methods combine the anatomic localizing ability of CT or MRI with SPECT images of radiolabeled monoclonal antibodies (Mabs), allowing the accurate staging of patients prior to surgery or following treatment. METHODS: Twenty-four patients (15 males and 9 females, mean age 55 yr, range 29-70 yr) were studied with this technique. Ten patients had suspected colorectal cancer recurrence and were infused with 10 mCi of 131I-CC49 prior to staging laparotomy. Fourteen patients treated in a Phase I radioimmunotherapy study with 131I-CC49 were also studied. All patients underwent SPECT imaging of the abdomen and pelvis 5-7 days following infusion of Mab. RESULTS: Phantom studies demonstrated a 3.6-mm surface fitting mean accuracy of datasets for the liver and 1.8 mm for an intrahepatic tumor. In the presurgical group, SPECT and CT/MRI registration allowed more accurate identification of uptake abnormal sites. Areas of metastatic disease > 1 cm confirmed at surgery were found in six of nine patients with liver lesions and in two patients with extrahepatic (including one patient with pelvic) disease. In patients imaged following radioimmunotherapy, all lesions > 1.5 cm seen on CT/MRI were identified, and activity distribution in tumor and normal tissue could be more accurately assessed. CONCLUSIONS: Routine registration of SPECT and CT/MRI images is feasible and allows more accurate anatomic assessment of sites of abnormal uptake in radiolabeled Mab studies.

Three-dimensional dosimetry for radioimmunotherapy treatment planning.

Absorbed-dose calculations for radioimmunotherapy are generally based on tracer imaging studies of the labeled antibody. Such calculations yield estimates of the average dose to normal and target tissues assuming idealized geometries for both the radioactivity source volume and the target volume. This work describes a methodology that integrates functional information obtained from SPECT or PET with anatomical information from CT or MRI. These imaging modalities are used to define the actual shape and position of the radioactivity source volume relative to the patient's anatomy. This information is then used to calculate the spatially varying absorbed dose, depicted in "colorwash" superimposed on the anatomical imaging study. By accounting for individual uptake characteristics of a particular tumor and/or normal tissue volume and superimposing resulting absorbed-dose distribution over patient anatomy, this approach provides a patient-specific assessment of the target-to-surrounding normal tissue absorbed-dose ratio. Such information is particularly important in a treatment planning approach to radioimmunotherapy, wherein a therapeutic administration of antibody is preceded by a tracer imaging study to assess therapeutic benefit.

Evaluation of carbon-14-colchicine biodistribution with whole-body quantitative autoradiography in colchicine-sensitive and -resistant xenografts.

UNLABELLED: Quantitative autoradiography (QAR) with radiolabeled monoclonal antibodies in xenografted animals has been extensively described in the past, either on individual tissues or on the whole body. We applied whole-body QAR to identify multidrug resistant tumors using 14C-colchicine (14C-CHC). METHODS: Two groups of five animals each were xenografted with CHC-sensitive and CHC-resistant human neuroblastoma cells. Animals were injected intravenously with 4 microCi/0.11 mumole 14C-CHC per gram of body weight and sacrificed after 60 min. Whole-body QAR was carried out using 25-microns thick sections. RESULTS: Fusion images allowed direct comparison of 14C-CHC uptake in tumor and nontumor tissues. Mean 14C-CHC distribution in sensitive and resistant tumors was 882.0 +/- 43.6 and 399.6 +/- 157.7 nCi/g corresponding to 24.5 +/- 1.21 and 11.1 +/- 4.38 nmole/g, respectively (p < 0.001), with normal tissue distribution in both groups being similar. Three-dimensional QAR showed that the uptake of 14C-CHC was in the cellular zones of the tumor. This method has potential in biodistribution studies of novel radiopharmaceuticals such as 14C-CHC. CONCLUSION: These studies further suggest that PET imaging of 11C-CHC is feasible to distinguish between sensitive and resistant tumor deposits in vivo.

Quantitative imaging of iodine-124 with PET.

UNLABELLED: PET is potentially very useful for the accurate in vivo quantitation of time-varying biological distributions of radiolabeled antibodies over several days. The short half-lives of most commonly used positron-emitting nuclides make them unsuitable for this purpose. Iodine-124 is a positron emitter with a half-life of 4.2 days and appropriate chemical properties. It has not been widely used because of a complex decay scheme including several high energy gamma rays. However, measurements made under realistic conditions on several different PET scanners have shown that satisfactory imaging and quantitation can be achieved. METHODS: Whole-body and head-optimized scanners with different detectors (discrete BGO, block BGO and BaF2 time-of-flight), different septa and different correction schemes were used. Measurements of resolution, quantitative linearity and the ability to quantitatively image spheres of different sizes and activities in different background activities were made using phantoms. RESULTS: Compared with conventional PET nuclides, resolution and quantitation were only slightly degraded. Sphere detectability was also only slightly worse if imaging time was increased to compensate for the lower positron abundance. CONCLUSION: Quantitative imaging with 124I appears to be possible under realistic conditions with various PET scanners.

Clinical applications of fusion imaging in oncology.

Recent developments in tumor imaging, made possible by advances in instrumentation and radiopharmaceuticals, has led to an increasing need for accurate anatomic correlation of single photon emission computed tomography (SPECT) and positron emission tomography (PET) images. Fusion imaging permits the functional strengths of SPECT and PET to be combined with the anatomic resolution of computed tomography (CT) and magnetic resonance imaging (MRI). Clinical applications of fusion imaging include the evaluation of brain tumors, lymphoma, hepatic lesions and monoclonal antibody studies. The continued development of these techniques will eventually allow fusion imaging to become a routine part of nuclear medicine practice.

In vivo imaging and specific targeting of P-glycoprotein expression in multidrug resistant nude mice xenografts with [125I]MRK-16 monoclonal antibody.

Multidrug resistance (MDR) in tumors is associated with P-glycoprotein (Pgp) expression. In vivo quantitation of Pgp may allow MDR to be evaluated noninvasively prior to treatment planning. The purpose of this study was to radiolabel MRK-16, a monoclonal antibody that targets an external epitope of P-glycoprotein, and perform in vivo quantitation of P-glycoprotein in a MDR xenograft nude mouse model. MRK-16 was labeled with 125I by the iodogen method, with subsequent purification by size exclusion chromatography. Groups of 10 Balb c mice were each xenografted with colchicine-resistant or sensitive neuroblastoma cell lines, respectively. Whole body clearance and tumor uptake over time was quantitated by gamma camera imaging, and biodistribution studies were performed with [125I]MRK-16 and an isotype matched control antibody, A33. Quantitative autoradiography and immunohistochemistry analysis of tumors was also evaluated to confirm specific targetting of [125I]MRK-16. Peak tumor uptake was at 2-3 days post-injection, and was significantly greater in resistance compared to sensitive tumors (mean % injected dose/g +/- SD) (18.76 +/- 2.94 vs 10.93 +/- 0.96; p < 0.05). Quantitative autoradiography verified these findings (19.13 +/- 0.622 vs 12.08 +/- 0.38, p < 0.05). Specific binding of [125I]MRK-16 was confirmed by comparison to [131I]A33 in biodistribution studies, and localized to cellular components of tissue stroma by comparison of histologic and autoradiographic sections of sensitive and resistant tumors. Immunoblot analysis demonstrated a 4.5-fold difference in P-glycoprotein expression between sensitive and resistant cell lines without colchicine selective pressure. We conclude that in vivo quantitation of P-glycoprotein in MDR tumors can be performed with [125I]MRK-16.(ABSTRACT TRUNCATED AT 250 WORDS)

Detection of bony metastases of androgen-independent prostate cancer by PET-FDG.

Fourteen F-18 fluorodeoxyglucose (FDG) studies were carried out in 13 patients known to have bony metastases from carcinoma of the prostate. One patient was newly diagnosed. The remaining patients had various types of therapy and were considered hormonally resistant. The average age was 67. All patients had extensive bony metastases shown on the conventional Tc99m-MDP bone scans. Only about 18% of bony lesions apparent on the conventional bone scans showed corresponding increase of FDG uptake. Anatomical correlation was performed by using co-registered images of SPECT and PET in the same area. The positive FDG uptake was not related to the duration of illness, level of PSA, previous therapy, and magnitude of disease involvement. It appears that only a small percentage of bony metastases is associated with increased glycolysis. It is possible that other metabolic processes are more important than glycolysis for providing prostate cancer with a source of energy and nutrients.

Resistance of [18f]-fluorodeoxyglucose-avid metastatic thyroid cancer lesions to treatment with high-dose radioactive iodine.

Radioactive iodine (131I) is an important therapeutic option for the treatment of metastatic thyroid carcinoma. Survival in patients with metastases that concentrate radioiodine is better than those whose metastatic lesions do not take up radioiodine. Survival is markedly reduced in patients who have metastatic lesions that concentrate 18F-fluorodeoxyglucose (FDG) on positron emission tomography (PET). In this retrospective study, we evaluated the ability of 131I to destroy FDG-avid metastatic lesions in thyroid cancer patients. Twenty-five patients with positive FDG-PET scans received at least one dose of 131I treatment before a second FDG-PET was performed. The average interval between the two PET scans was 12.9 months. The average interval between the 131I treatment and the follow-up FDG-PET was 10.1 months. We measured posttherapy changes in lesional volume, in standard uptake values (SUV) of FDG, and in serum thyroglobulin (Tg) levels. The total volume of FDG-avid metastases rose significantly (p = 0.036) from a mean of 159 mL to 235 mL after 131I therapy, the maximum SUV rose from 9.3 to 11.9, the median Tg at the time of the second PET scan was 132% of that at baseline. Statistical analyses demonstrated no significant changes in maximum SUV, or serum Tg levels after 131I in the FDG-PET-positive group. In a control group of FDG-PET-negative patients, the serum Tg decreased to 38% of baseline after 131I therapy (p < 0.001). We conclude that high-dose 131I therapy appears to have little or no effect on the viability of metastatic FDG-avid thyroid cancer lesions.

Simultaneous gallium-67 citrate and technetium-99m sestamibi SPET in a case of myocardial lymphoma: comparison with echocardiography before and after chemotherapy.

A patient with diffuse large cell lymphoma involving the interventricular septum and the inferior ventricular wall was imaged with a simultaneous dual-isotope single-photon emission tomography (SPET) acquisition technique, using the radiotracers technetium-99m hexakis 2-methoxyisobutylisonitrile (sestamibi) and gallium-67 citrate, in conjunction with echocardiography, prior to and following the first course of chemotherapy. Simultaneous acquisition--with the advantage of displaying corresponding sets of SPET slices without any need for position correction--, supplemented by echocardiography, increased the accuracy of evaluation of the extent of disease and response to treatment.

The value of thallium and three-phase bone scans in the evaluation of bone and soft tissue sarcomas.

Thirty-seven patients with newly diagnosed or treated sarcomas had 47 sets of sequential thallium scans (TS) followed by three-phase bone scan (TPBS) on the same day. The diagnosis in all patients was verified by biopsy (n = 40) or long-term follow-up studies (n = 7). The sensitivity, specificity, and accuracy of TS and TPBS in detecting sarcomatous lesions was calculated: TS sensitivity was 88%, specificity 69%, and accuracy 83%; blood flow (BF) and blood pool (BP) sensitivity was 91%, specificity 54%, and accuracy 81%; delayed bone scan (DB) sensitivity was 88%, specificity 38%, and accuracy 74%. In 17 studies the flow and blood pool parts of the TPBS and TS demonstrated the soft tissue component of sarcomas, which would have been missed if only the delayed bone scan had been performed. The TS lesion to normal tissue ratio alone was not very helpful in differentiating sarcomas from benign conditions because some benign lesions are highly cellular and vascular while some malignant lesions, such as chondrosarcoma, have poor vascularity and a less cellular chondroid matrix. However, when the thallium ratio was correlated with similar ratios calculated from the BP image, it was found that if the TS lesion to normal tissue ratio exceeded the BP lesion to normal tissue ratio (12 patients), the specificity for detecting sarcomatous lesions was 100%. Nevertheless, the reverse was not true. The positive predictive value of this observation was 100% and the negative predictive value was 37%.

Preselection of patients with high TAG-72 antigen expression leads to targeting of 94% of known metastatic tumor sites with monoclonal antibody I-131-CC49.

We studied 18 consecutive patients with advanced colorectal cancer where primary tumors were preselected for high expression of TAG-72 antigen and who underwent a phase I radioimmunotherapy trial with an intravenously administered monoclonal antibody CC49, 20 mg, labeled with I-131 in amounts varying from 15 mCi/m2 to 75 mCi/m2. Whole-body images and SPECT of the abdomen obtained 1 week after infusion were compared with pretreatment CT scans. A total of 66 lesions were evaluated. SPECT revealed 2/66 lesions (3%) that were not detected by CT; 4/66 were only detected by CT: lungs (1.8 cm and < 1 cm), axilla (1.5 cm), adrenal (2.5 cm). Thus, based on immunohistopathological testing in paraffin-embedded tissue blocks of primary tumors stained for TAG-72 antigen, we have selected a subset of patients (about 70% of referrals) with colorectal cancer for whom I-131-CC49 was shown to target to 62/64 CT positive lesions (97%) and 62/66 (94%) of all known positive lesions. We conclude that in patients with significant TAG-72 tumor expression there is excellent targeting of I-131-CC49 in therapeutic doses to colorectal cancer with respect to lesions detected with CT scanning. It should be noted that this study was not designed as a comparison of the sensitivity of CT versus I-131-CC49 SPECT/planar imaging. Instead, the observed results are consistent with a biological hypothesis that in general, the primary tumor histology vis-à-vis TAG-72 expression reflects the TAG-72 expression of the metastatic sites.