Metabolic Tumor Volume and Total Lesion Glycolysis Can Predict Response to Very Low Dose Radiotherapy (4 Gy) in Indolent B-Cell Lymphomas

Numerous studies have established the role of very low dose radiotherapy (VLDRT), only 4 Gy in 2 fractions, in the management of indolent non-Hodgkin's lymphoma (NHL). While objective response rates to VLDRT are excellent, there are no widely accepted biomarkers of the depth and the durability of response after VLDRT. Radiosensitivity to VLDRT is not clearly linked to clinical features, such as tumor grade, histology, prior treatments or [18F]-FDG-PET standardized uptake values (SUV). Our group has recently demonstrated that pre-treatment tumor size greater than 6 cm may predict suboptimal response to VLDRT; however, this remains an imperfect predictor, and the majority of patients selected for VLDRT have lesions smaller than 6 cm. [18F]-FDG-PET metrics like baseline metabolic tumor volume (MTV) and total lesion glycolysis (TLG) are shown to stratify patient response to chemotherapy for some NHLs, but neither has been shown to predict radiotherapy response to VLDRT. Therefore, we aimed to determine if these features could predict response to VLDRT in indolent B-cell NHLs.

Between 2005 and 2018, 250 patients with follicular or marginal zone lymphoma were locally irradiated to 299 sites using 2 Gy x2. Response was assessed within 1.5-6 months of VLDRT (n=231), and local failure (LF) was assessed over a median follow-up of 29 months. Out of the 299 treated sites, 254 (85%) had a corresponding [18F]-FDG-PET/CT scan performed prior to radiotherapy. PET scans were imported into MIM, a radiotherapy contouring and planning software program, and pre-radiotherapy disease was manually defined. A uniform expansion of 0.5 cm was performed to define a larger region of interest containing the contoured lesion but accounting for technical variability in defining the tumor edges. Of the 254 lesions, 207 (81%) had a maximum SUV (SUVmax) greater than 2.5 and were considered for further analysis. PET parameters were obtained using an SUV of 2.5 as a cutoff for MTV (MTV2.5) and TLG (TLG2.5), which capture lesion bulk, and using 41% of the SUVmax as a cutoff for MTV (MTV41%) and TLG (TLG41%), which capture the most metabolically active parts of a lesion. Maximum lesion size was analyzed using 4 cm and 6 cm as predetermined cutoffs. Lesion response was analyzed using multivariate logistic LASSO regression to account for predictor multicollinearity, and receiver operating characteristic curve analysis to obtain estimates of the area under the curve (AUC). To analyze time to LF, we performed Cox proportional hazards multivariate regression using stepwise selection. PET parameters were divided in quartiles to aid in interpretability. Estimates are shown with 95% confidence intervals.

A complete response (CR) was seen in 156 (68%) lesions and LF was seen in 81 (27%) lesions after VLDRT. Using size alone had similar predictive value for response (AUC: 0.57(0.49-0.65) compared to MTV2.5 (AUC:0.66 (0.57-0.74)), MTV41% (AUC:0.64 (0.56-0.73)), TLG2.5 (AUC:0.64 (0.54-0.74)), and TLG41% (AUC:0.64 (0.54,0.75)). Through LASSO logistic regression, we determined that only MTV41% > 75 ^th percentile was associated with lower odds of CR (odds ratio (OR):0.21 (0.08-0.53)) versus size ≥4 (OR: 0.94 (0.32, 2.73) ) with no interaction noted. Likewise, in a multivariate model of LF, TLG2.5> 75 ^th percentile was shown to be a better predictor of worse LF (HR: 2.44 (1.32, 4.52)) than size when considering lesions < 6 cm.

In stratifying response and local control for patients receiving VLDRT for indolent lymphomas, MTV and TLG may aid in patient selection, especially for lesions smaller than 6 cm. As PET-based radiotherapy planning is routinely performed in the treatment of indolent lymphomas, these parameters are easily attained after contouring and may have immediate clinical utility. These findings may also be applicable to radiotherapy response with higher doses, and the incorporation of [18F]-FDG-PET metrics into treatment decisions is an area of active research. It remains to be seen in ongoing work whether higher-order PET radiomic features (which for instance capture heterogeneity of tumor glucose metabolism) can further improve the accuracy of [18F]-FDG-PET-based radiotherapy outcome prediction.