The 18F-FDG SUVmax Reduction After Two Cycles of R-CHOP Regimen Predicts Progression Free Survival of Patients with Diffuse Large B-Cell Lymphoma.

PET assessment of early response allows to predict outcome of patients with DLBCL, but the more accurate criteria for early PET interpretation remain to be define. So we evaluated the prognostic value of PET after two cycles of immuno-chemotherapy and compared a semiquantitative approach using FDG uptake reduction, with two methods of visual analysis.

Forty five consecutive patients with newly diagnosed DLBCL treated in a single institution with rituximab and CHOP or CHOP-like regimen underwent ¹⁸F-FDG PET at baseline (PET0) and after 2 cycles of induction treatment (PET2). Images were interpreted visually according to two separate methods: - Juweid criteria (JCO 2007; 25: 571) (visual analysis-1) – A visual comparison of the FDG uptake between the residual mass and the normal hepatic tissue, a PET being considered as positive when the uptake of at least one residual mass was found greater than the liver (visual analysis-2). The quantitative approach was based on the lymphoma FDG uptake estimated by the maximal standardized uptake value (SUVmax) corrected to body weight. The SUV reduction between PET0 and PET2 (ΔSUVmax) was calculated for each patient. The ΔSUVmax cut-off was estimated to 65% by ROC analysis.

Patient median age was 50 years (range, 24 – 79) and 37 patients (73%) were younger than 61 years. The age-adjusted IPI score was, 2 or 3, 1 and 0 in 26 (49%), 16 (36%) and 7 (16%) patients respectively. With a median follow-up of 25 months, 6 (13%) out of 45 patients progressed or relapsed after treatment and 4 died from progressive disease. According to the visual analysis-1, PET2 was interpreted as negative in 16 (36%) patients and positive in the 29 remaining patients. Using the visual analysis-2, PET2 was negative in 25 cases (56%) and positive in 20 cases. The quantitative analysis showed a ΔSUVmax < 65% in 9 patients and a ΔSUVmax >= 65% in 36 patients (80%). The estimated efficiency of the three methods to predict patient outcome is detailed below:

The probability of 2-years progression-free survival (PFS) for patients with a negative PET2 according to the visual analysis was slightly higher than those with a positive PET2 when using as well the Juweid criteria (93% vs 83%; p = 0.3), as the visual analysis-2 (95% vs 75%: p = 0.04).

However, the quantitative aproach allows to better identify and split up the population of patients with a good outcome from those with a poor outcome, since the 2-years PFS was 56% in patients with a ΔSUVmax reduction less than 65% compared to a 94% 2-years PFS probability in patients with a ΔSUVmax higher than 65% (p=0.0009). A multivariate analysis was performed including the IPI score and the ΔSUVmax reduction as explanatory variables for PFS, showing that the PET2 result assessed by the ΔSUVmax reduction remains the only independant prognosis factor for PFS (p = 0.008; RR = 10).

SUV-based assessment of PET after two courses of immuno-chemotherapy is more reliable to patient outcome than visual analysis. The SUVmax reduction is an early prognostic factor for DLBCL patients that may help to reduce false positive interpretations, and provides a useful tool to guide risk-adapted treatment.

No relevant conflicts of interest to declare.