Practical PERCIST: A Simplified Guide to PET Response Criteria in Solid Tumors 1.0

Abstract

Positron Emission Tomography (PET) Response Criteria in Solid Tumors (PERCIST 1.0) describes in detail methods for controlling the quality of fluorine 18 fluorodeoxyglucose PET imaging conditions to ensure the comparability of PET images from different time points to allow quantitative expression of the changes in PET measurements and assessment of overall treatment response in PET studies. The steps for actual application of PERCIST are summarized. Several issues from PERCIST 1.0 that appear to require clarification, such as measurement of size and definition of unequivocal progression, also are addressed. (©) RSNA, 2016.

Figures

Figure 1a:

(a) Maximum intensity projection image and (b) axial PET/CT fusion image show placement of 3-cm–diameter spherical VOI in right side of liver (blue circle) to measure background.

Figure 1b:

(a) Maximum intensity projection image and (b) axial PET/CT fusion image show placement of 3-cm–diameter spherical VOI in right side of liver (blue circle) to measure background.

Figure 2:

Magnified transverse FDG PET image shows a single large heterogeneous tumor. SULpeak was the mean from a three-dimensional 1-mL sphere approximately 1.2 cm in diameter in tumor area, yielding highest average SUL (represented here as circle instead of sphere). SULpeak does not necessarily contain maximum standardized uptake value (SUV, square), which is the highest single pixel value.

Figure 3a:

Maximum intensity projection images (top row) in two different SUL windows and axial fusion images (bottom row) in a man with sarcoma and multiple distant metastatic lesions. SULpeak was measured from the hottest tumor at each time point and can be different at baseline and follow-up. (a) Hottest tumor SULpeak focus was located in right lower paratracheal lymph node at baseline (arrows). (b) After therapy, subcarinal lymph node was hottest tumor (arrows).

Figure 3b:

Maximum intensity projection images (top row) in two different SUL windows and axial fusion images (bottom row) in a man with sarcoma and multiple distant metastatic lesions. SULpeak was measured from the hottest tumor at each time point and can be different at baseline and follow-up. (a) Hottest tumor SULpeak focus was located in right lower paratracheal lymph node at baseline (arrows). (b) After therapy, subcarinal lymph node was hottest tumor (arrows).

Figure 4a:

Graphs show that quantification of whole or subsegmented pixels results in different area measurements. (a) When the whole pixel is considered, each box represents a single pixel, blue line represents tumor boundary, and only shaded pixels are included for quantification. (b) When pixels are subsegmented to quadrants in this graph, the area included for quantification is different from that in a. Smaller green boxes represent newly included area, and red lines represent pixel quadrants that would be excluded from quantification.

Figure 4b:

Graphs show that quantification of whole or subsegmented pixels results in different area measurements. (a) When the whole pixel is considered, each box represents a single pixel, blue line represents tumor boundary, and only shaded pixels are included for quantification. (b) When pixels are subsegmented to quadrants in this graph, the area included for quantification is different from that in a. Smaller green boxes represent newly included area, and red lines represent pixel quadrants that would be excluded from quantification.

Figure 5a:

Graphs show two-dimensional comparison of two different regions of interest. (a) Region of interest (average standardized uptake value [SUV ], 7.78) is inside tumor boundary (bold black line outside of red squares). (b) Region of interest includes one pixel outside tumor boundary, resulting in higher average value (average SUV, 8.33).

Figure 5b:

Graphs show two-dimensional comparison of two different regions of interest. (a) Region of interest (average standardized uptake value [SUV ], 7.78) is inside tumor boundary (bold black line outside of red squares). (b) Region of interest includes one pixel outside tumor boundary, resulting in higher average value (average SUV, 8.33).