EQ•PET Brings SUV Comparison to Daily Routine
Greg Freiherr | 2018-01-17
Harmonizing SUVs with EQ•PET allows a physician to reconcile data acquired on different systems and quantitatively assess patient treatment response.
Download your print version here 1.4 MB
Oncologists welcome a more objective way to evaluate their patients, especially when gauging patients’ response to therapy. With its standardized uptake values (SUVs), PET/CT seems ideally suited to the task. SUVs, which indicate maximum, mean, and peak tracer accumulations in soft tissue, gain recognition due to their ability to aid in the differentiation between benign and malignant lesions. Additionally, SUVs assist in the prognosis of patients and help evaluate their response to therapy. The key to monitoring therapy is in the comparison of these measures acquired prior to, during, and after treatment.
SUVs are comparable from one exam to the next when certain variables are controlled. Inherent differences in the scanners used during a patient’s course of care can be especially problematic. Such differences stem from variations between vendors, advancements between scanner generations, and variances in reconstruction methods. For example, the increased sensitivity of a newer-generation scanner may produce higher values than earlier models, leading to data that cannot be consistently compared. But there is a solution.
At the University Hospital of Cologne, research led by Alexander Drzezga, MD, and Carsten Kobe, MD, models such a solution. Their research demonstrates that EQ•PET, a feature within syngo®.via’s MM Oncology software application, reliably harmonizes data obtained from scans performed on different generations of Biograph™ PET/CT systems. Such harmonization of data does not notably alter the image quality or appearance to which readers are accustomed.
“It was very simple to compare the values obtained with our new scanner, which are much higher, with those from the old one,” says Kobe, professor within the Department of Nuclear Medicine. Professors Drzezga and Kobe corrected SUVs for about 300 follow-up patient examinations and compared values from these exams to ones obtained during prior examinations. The data provided objective measures for adjusting therapies for oncology patients. The conversions were accomplished, according to Drzezga, without impacting the image quality achieved with their Biograph mCT.
“Without EQ•PET, it would have been a lot more difficult, if not impossible,” says Drzezga, professor and director of the Department of Nuclear Medicine. The physicians initially recognized the value of EQ•PET when evaluating patients with Hodgkin lymphoma, as well as lung and prostate cancers. “It is not limited to a certain type of cancer though,” Kobe says, “ we can broadly use EQ•PET to follow patients for long periods and make decisions along the way based on how they are truly responding.”
The drawback of reference regions
Both Kobe and Drzezga emphasize the importance of controlling other variables that can affect the comparability of SUVs, namely variances in the scan acquisition settings and data analysis methods. They note that following international guidelines can substantially reduce the chance that errors extend into patient evaluations. But, aside from the use of EQ•PET, they agree there are realistically very few viable options for harmonizing data differences due to varying sensitivities.
Research by the pair calls in to question the use of liver tissue as a reference for normalizing SUVs. Their work contradicts the belief that the liver can effectively harmonize data gathered on equipment made by different manufacturers. The researchers examined PET/CT data from 40 lung cancer patients and normalized SUVs in the hottest lung lesions by utilizing SUVs in the liver as references. They evaluated the liver tissue SUVs for their utility as a means for standardizing various values from different scanners.
“We found out that normalization to the liver is of no help,” Kobe states. “Many nuclear medicine physicians do not yet realize that this is a problem.” According to Drzezga, the wide-spread belief that a reference region—such as the liver—allows the semi-quantitative analysis of data, regardless of the type of scanner or reconstruction, is based on the assumption that measurements taken throughout the body reflect the sensitivity of a scanner. Once the sensitivity of the scanner is calculated for a reference region, this calculation should allow for the conversion of data acquired elsewhere in the body.
“But this is not the case,” he states. “Changes are not homogenous across the entire patient but instead are heterogeneous. So even by using a reference region, you would not automatically be able to correct for differences in scanner type or injected dose of the radiotracer.” EQ•PET circumvents the problem of using a reference region, allowing physicians to look only at the SUVs that characterize a specific lesion. “Here, EQ•PET brings the standardization that we are aiming for,” Drzezga says.
Potential for multicenter trials
Kobe collaborates with more than 60 PET/CT sites in Germany and throughout Europe on a large-scale Hodgkin lymphoma study. But rather than use SUVs, the researchers in this multicenter project depend on the visual analysis of images. “EQ•PET would be an ideal tool to record differences in SUVs along with the visual analysis, so we could do quantitative analysis,” adds Kobe. To compare SUVs accurately the researchers would need EQ-filter data, which is acquired via scans of phantoms. Also necessary would be the data acquisition and reconstruction settings from each scanner used in the study.
With this, “EQ•PET could easily integrate the information and calculate conversions across the entire acquisition set,” he explains. “With that you could normalize the data across scanner types, harmonizing values for radiotracer uptake by lesions that you want to quantify, as well as harmonizing uptake in the background. That would not only correct to the absolute SUV signal, but you would also correct the relative or the ratio based semi-quantitative analysis.”
Kobe emphasizes there is a need for such a tool in multicenter trials. “We predict that EQ•PET is going to be used more and more,” Kobe expresses. “It is obvious there is a need for it, and I would say that its value has been established for longitudinal follow-up examinations at our center.”
About the Author
Greg Freiherr has reported on developments in radiology since 1983. He runs the consulting service The Freiherr Group.
 Kuhnert G, Boellaard R, Sterzer S, et.al. Impact of PET/CT image reconstruction methods and liver uptake normalization strategies on quantitative image analysis. EJNMMI 2016 43(2): 249-258.
The statements by Siemens Healthineers’ customers described herein are based on results that were achieved in the customer’s unique setting. Since there is no “typical” hospital and many variables exist (e.g., hospital size, case mix, level of IT adoption) there can be no guarantee that other customers will achieve the same results.