The role of imaging is still quite limited in CLL. In routine clinical practice, CT imaging might be used to assess a symptom (eg, shortness of breath), but there is no role for routine scanning and the risks of overscanning are very real. 

For the particular application of working up suspected CLL transformation, however, advanced imaging with ¹⁸F-FDG PET/CT imaging can be very helpful. When CLL transforms, typically from RT to diffuse large B-cell lymphoma, that transformation is generally associated with greater metabolic activity. Let us say that you have a patient with a rapidly growing area of lymphadenopathy that is discordant from other sites following chemoimmunotherapy (CIT), or some B symptoms, or an elevated lactate dehydrogenase, and you suspect transformation. PET imaging can be used in this scenario to identify a potential target for biopsy, to confirm the clinical suspicion. Standardized uptake values (SUVs) of less than 5 are typical of CLL, whereas values of greater than 10 are suggestive of RT and values between 5 and 10 represent a gray area. 

PET/CT is an extremely useful tool for suspected RT, but it is never diagnostic. Biopsy is always required—ideally, an incisional or excisional biopsy, possibly a core needle biopsy, but never a fine needle aspiration, which destroys the architecture that is diagnostic of RT. Additionally, ¹⁸F-FDG PET/CT might not be as powerful for identifying transformation in the novel treatment era. In a recent post-hoc analysis of a phase 2 study with venetoclax, our data suggested that CLL progression might be more metabolically active on PET/CT following B-cell receptor signaling pathway inhibitors as compared with CIT.  A maximum SUV (SUVmax) cutoff of 10 was able to distinguish suspected RT from CLL progression at screening, with a sensitivity of 71% and a specificity of 50%, testing characteristics that are modest as compared with past reports in chemotherapy- or CIT-treated patients. Nevertheless, PET/CT is still the standard of care when targeting sites for biopsy in suspected RT. The other interesting finding from that study was that, on the pre-venetoclax screening scans, SUVmax seemed to predict clinical outcomes. Response rate to venetoclax was similar for SUVmax (˂10 vs 10), but median progression-free survival was longer with an SUVmax of less than 10 (24.7 months vs 15.4 months). This prompts an interesting research question that requires further study: Does PET/CT have the potential to identify populations that would benefit from therapies other than venetoclax and/or combinations?