PET stands for positron emission tomography, and this generally uses a safe radioactive tracer molecule called 18F-fluoro-2-deoxy-D-glucose (FDG). Fortunately, that’s not on the quiz — you don’t need to know it. All that is worth knowing is that PET scans offer “metabolic imaging”, which detects differences in the metabolism of tissues. The most metabolically active tissues have the greatest needs for sugar from the bloodstream, and when the sugar is labelled, the scans detect these areas as “PET-avid”. The objective measurement is a number called a “standard uptake value”, or SUV, where higher numbers mean a higher metabolic rate. While metabolic activity can be increased with inflammation, infection, and normal body activities (organs like heart, brain, and bowel have uptake normally from regular ongoing activity — this is NOT a bad thing) the reason we all care is that PET scans tend to pick up cancer, because cancer cells generally have greater metabolic activity and are dividing faster than most normal tissue.
(click to enlarge)
Some slower growing cancers, such as bronchioloalveolar carcinomas(BAC), are much less consistently reported to appear on PET scans than other NSCLC tumors. And there is some evidence that higher SUV activity of lung tumors correlates with faster progression and worse prognosis.
Now, the first area where PET scans earned their place in oncology is in initial staging of cancer. This is very routine now, and it’s clear that CT (for computerized tomography, but it’s always referred to just a CT or CAT scan) is very good at assessing shape and size of organs and nodules. However, if a lymph node is not enlarged or there’s a spot in the liver that has the same density as the regular liver tissue, or there’s some other spot that looks like it may be a benign cyst, CT scans have a hard time detecting cancer. They’re great for size and shape of things, but they can’t say if something is very metabolically active or not. PET scans are great at determining whether something is metabolically active or not, but they generally can’t pinpoint size very well. Also, the SUV needs to be considered relative to the size of the lesion. A nodule of 1 cm, or these days probably down to 6-8 mm, can likely be detected on a PET scan, but a tiny 2-3 mm nodule probably doesn’t have enough cells, even if they’re very metabolically active, to light up on a PET scan. So PET scans use a combination of how metabolically active the cells are, and how many there are in an area. They don’t tend to be very reliable at determining whether pleural fluid has cancer cells, because those cells aren’t concentrated into a metabolically active cancer ball (not a technical term), but are diluted in fluid that is not very metabolically active.
So CT scans and PET scans both have strengths and weaknesses. We can get useful information by having patients get both types of scans and holding them side by side, seeing if an ambiguous lymph node or spot in the liver on CT is “hot” on PET. That’s good, but over the last few years, the newest machines are now able to perform a fused PET and CT scan and overlay the results on top of each other, so that you can directly see what area in the chest or abdomen is lighting up on PET (note also the heart lighting up red, which is just normal activity):
Without the PET, the slight thickening on the left chest wall would be pretty questionable, but seeing that it’s also metabolically active on PET makes it nearly certain that this person has progression of her cancer inside the chest wall (she had previously had this found at surgery, and then responded well to chemo afterward, with no residual nodules on her scans until this one).
On the other hand, we still have plenty of situations where even with CT and PET scans, or a PET/CT fusion scan, we can’t figure out what is residual cancer, what is inflammation after treatment (particularly radiation or chemo and radiation), and what is infection. This scan is of a patient who underwent chemo and radiation for a stage III NSCLC, and months later we see some collapsed lung and a lot of scarring, but we can’t tell where inflammation stops and cancer begins (heart again lighting up in center of picture):
This is worrisome, but after chemo and radiation we often see very ambiguous changes that even can light up on PET and then subside over time.
Overall, we’re still learning more and more about the value of PET scans and PET/CT fusion scans in following patients over time. At my own center, we actively debate whether the added information we gain from post-treatment PET scans is reliable enough to incorportate into treatment decisions. New technology can be remarkably promising, but we are struggling to have our knowledge and experience catch up with the capabilities of the new imaging machines. In a few subsequent posts, I’ll go into more details on emerging data for using PET scan imaging in following patients on or after treatment.
Posted on January 9, 2007 at 7:19 pm
Dr. West,
I have an appointment with you on 1/18. I just completed both a pet/CT on Monday 1/8. Got radiology report today and somehow radiologist made no mention of secondary lesion in left lung that last measured 2.1 X1.1 cm. I will fax your office this radiology report as soon as I get this omission corrected. Slight growth in primary left lung lesion after 4 maintainence cycles of avastin. Primary lesion had reduced in size after 3 cycles of carbo, taxol and avastin and was stable after 6. I have most of my CT and pet scans on CD and could fedex you a copy before appointment if that would be of value. Pet scans are jpg with an avi thrown in while CT scans use DICOM viewer. Of course I do not know how to interpret these scans, but being a MAC user OsiriX seems to be a better viewer