P1174: THE COMBINATION OF QUALITATIVE AND QUANTITATIVE METABOLIC PARAMETERS PREDICTS PROGRESSION-FREE SURVIVAL IN PATIENTS WITH LARGE B-CELL LYMPHOMAS (LBCL) TREATED WITH ANTI-CD19 CAR-T CELLS.

Abstract

Background: Autologous anti-CD19 chimeric antigen receptor (CAR) T-cell therapy is an effective therapy for 40% of patients (pts) with relapsed/refractory (R/R) LBCL. Qualitative and quantitative metabolic parameters from the 18-FDG PET-CT scan are tools able to predict outcomes in lymphomas, but their value in the CAR-T setting is still under investigation. Aims: Prognostic role of PET-CT performed before lymphodepletion and 1 month after CAR-T cell infusion, analyzed by qualitative and quantitative parameters Methods: We conducted an observational a prospective study on pts affected by R/R aggressive lymphomas receiving axicabtagene ciloleucel (axi-cel) or tisagenlecleucel (tisa-cel). PET-CT was evaluated before lymphodepletion, at 1 month (M1) and at 3 months (M3) post-infusion using visual assessment with a 5-point Deauville score (DS). Quantitative PET parameters were evaluated with PET-VCAR software (GE Healthcare). The contour of pathological lesions was delineated using semiautomatic contouring systems including a gradient-based algorithm. SUVmax, SUVmean, Metabolic tumor volume (MTV), and Total lesions glycolysis (TLG) were automatically recorded. Total MTV (TMTV) and total TLG (TTLG) were calculated as the sum of all individual lesions. The percentage change in FDG uptake of each parameter between baseline and M1 was defined as Δ=(baseline-M1)/baseline x 100%. Logistic and Cox regression models were applied to evaluate the association between PET parameters at baseline, at M1 and in terms of Δ with respect to clinical response and progression free survival (PFS), respectively. Results: 47 pts were analyzed: median age was 55 years (range, 22-70); histologies included diffuse-large B-cell lymphomas (n=33) and primary mediastinal B-cell lymphomas (n=14); pts received treatment with axi-cel (n=31) or tisa-cel (n=16). Before infusion, 18 (38%) had high LDH, 14 (30%) had extranodal disease and 16 (34%) had bulky disease. Majority of pts [n=41 (84%)] received bridging therapy. At a median follow-up of 12 months (range, 3-29 months), 38 patients are alive and 9 died of disease progression with an estimated 1-year PFS and OS of 46% (95%CI; 30%-60%) and 83% (95%CI; 66%-92%), respectively. Baseline values of SUVmax, but not SUVmean, TTLG, TMTV correlated with poor clinical response [HR 1,49; (95%CI; 1,01-2,2), p=0,04]. By univariate logistic model, a higher ΔSUVmean between baseline and M1 was associated with better response at M3 [odds ratio 3,49, (95%CI; 1,25-9,77), p=0.02]. Univariable Cox analyses showed that a higher ΔSUVmean and ΔTMTV between baseline and M1 were associated with a better PFS [HR 0,63, (95%CI; 0,41-0,96), p=0,03; HR 0,67 (95%CI; 0,48-0,94), p=0,02], respectively. In terms of visual assessment, pts with DS 4-5 at M1 had a significantly higher risk of failure as compared to those with DS 1-3 [HR 2,92, (95%CI; 1,12-7,64), p=0,02]. A combination of M1 DS score and ΔSUV mean showed a better prognosis for patients with DS1-3 (PFS 60%), an intermediate prognosis for those with DS4/5 and high ΔSUVmean (PFS 49%) and a poor prognosis for those with DS4/5 and low ΔSUVmean (PFS 22%), [p=0,009]. In a multivariable analyses, evaluating clinical prognostic factors and PET parameters, a higher risk of failure was observed in pts with high LDH [HR 3,2 (95%CI; 1,30-8,14), p=0,01] whereas a higher ΔSUVmean was associated to a better PFS [HR 0,61 (95%CI, 0,41-0,92), p=0,02]. Summary/Conclusion: Assessment of DS combined with variation of SUVmean at M1 provides informations about early risk of CAR-T failure. A validation group are needed to extend and confirm our observations.