Top Line: What imaging features are predictive of local recurrence after lung SBRT?
The Study: It can be challenging to differentiate treatment effects from early tumor recurrence with both CT and PET/CT following SBRT. Here’s a review of 32 studies analyzing CT and PET/CT features predictive of local recurrence (LR) after SBRT for early stage NSCLC. The rate of recurrence across studies was 13%. Among CT high risk features, the three strongest predictors of LR were craniocaudal growth, bulging margin, and enlarging opacity ≥ 12 months after treatment–the latter being the best single CT predictor of LR. In addition, the development of multiple risk features had a high sensitivity and specificity for LR. For instance, the presence of both bulging margin and craniocaudal growth had an 85% sensitivity and 100% specificity. CT kinetics > 12 months were also predictive, including 3 consecutive rises in size/volume. In the first 12 months after SBRT, a PET feature suspicious for LR was a <5% decrease in SUVmax from pretreatment SUVmax. After 12 months, a SUVmax > 5 or a SUVmax > 4.5 along with a mass-like shape of uptake were suspicious of LR. An important point is that radiographic features suggestive of recurrence really don’t become useful until 6 months and perform best > 12 months after treatment. The study ends with a very nice algorithm for working through post-SBRT imaging to identify local recurrence.
TBL: CT and PET/CT features are helpful in identifying local recurrence after SBRT for NSCLC with most features performing better > 12 months after treatment. | Lee, Int J Radiat Oncol Biol Phys 2021