Abstract
Abstract 1598
Tumor bulk assessed by maximum tumor diameter was previously found to influence the outcome of young patients with low risk (aaIPI 0–1) DLBCL (Pfreundschuh et al Lancet Oncol, 2011; 12: 1013), but the prognostic impact of bulk and more generally tumor volume in high-risk DLBCL remains to be explored. Conversely, maximum SUV reduction calculated between baseline PET and PET performed after 4 cycles of immunochemotherapy (ΔSUVmax0–4) was recently shown to predict outcome in this population (Casasnovas et al Blood 2011; 118: 37).
In this study, we assessed the metabolic tumor volume (MTV) and the tumor bulk at baseline in 121 patients <60 years with an aaIPI2–3 DLBCL enrolled in the LNH07-3B trial (NCT00498043). We then compared the prognosis value of both parameters to that of ΔSUVmax0–4.
MTV was measured by FDG PET/CT with a semi automatic method using various volume shapes and systematic 41% SUVmax thresholding. Tumor mass with a maximum diameter > 10 cm assessed on CT scan was considered as bulky.
Median follow-up was 28 months. Median baseline MTV was 303 cc (17–1488). Baseline MTV (≥625 cc vs <625cc, on the basis of a ROC analysis) significantly predicts 2-year progression free survival (2y-PFS) (57 % vs 83%; p=0.0032) and 2-year overall survival (2y-OS) (60% vs 90%; p=0.002). By contrast bulk >10 cm was not predictive of outcome.
As previously published ΔSUVmax0–4 (> 70% vs ≤70%) found to predict 2y-PFS (88% vs 40%; p<10-4) and 2y-OS (94% vs 59%; p<10-4). Combining baseline MTV to early response assessed by ΔSUVmax0–4 allows to split the group of good responders patients defined on the basis of ΔSUVmax0–4 >70% after 4 cycles of immunochemotherapy (n=101, 88%) into 2 separated prognosis subsets (2y-PFS: 90% vs. 77%, p<10-4; 2y-OS: 96% vs. 77%, p<10-4). So far MTV could not significantly identify subgroups within the 15 patients with ΔSUVmax≤70% even if patients with higher MTV base line had lower PFS and OS.
In our series baseline MTV was predictive of PFS and OS in high risk DLBCL patients while bulk was not. MTV identified different risk categories of DLBCL patients within DLBCL patients with a good interim response (ΔSUVmax0–4 >70%) and may improve the negative predictive value of interim PET.
No relevant conflicts of interest to declare.
Author notes
Asterisk with author names denotes non-ASH members.
This feature is available to Subscribers Only
Sign In or Create an Account Close Modal