Prognostic Value of Baseline PET/CT Imaging in Diffuse Large B-Cell Lymphoma: Does the Largest Distance between Two Lesions Play a Role in Prognosis?

Background: 18F-fluorodeoxyglucose positron emission tomography/computed tomography (18F-FDG-PET/CT) has become the international standard for diffuse large B-cell lymphoma (DLBCL) staging and is widely used for response assessment. In the past few years, the prognostic value of a variety of imaging-derived quantitative parameters has been investigated. One of these parameters is the distance between the 2 lesions that are farthest apart (Dmax). Dmax is a simple feature that captures lymphoma dissemination and has been reported to be a strong prognosticator of shorter progression-free survival (PFS) and overall survival (OS) in DLBCL patients (pts).

Methods: Retrospective analysis of 146 pts with newly diagnosed Ann Arbor stage III-IV DLBCL, NOS, treated with R-CHOP(-like) regimens, between January 2010 and December 2019 in a tertiary center. Dmax was normalized with the body surface area [standardized Dmax (SDmax)]. The receiver operating characteristic curve (ROC) method was applied in resetting the optimal cut-off point for predicting prognosis and treatment outcome. The Kaplan-Meier method was used to describe time-to-event end points. A p-value of 0,05 was used as cut-off for significance.

Results: Dmax was available for 40 pts. The median age at diagnosis was 63 years (24-78); 58% of pts were male. Most pts (69%) presented lactate dehydrogenase (LDH) above the upper limit of normal (ULN); 40% presented lymphopenia (lymphocyte count below 1000/μL), and 26% had albumin levels below 3.2 g/dL. Bone marrow (BM) infiltration occurred in 41% of pts, and 5% had central nervous system (CNS) involvement at diagnosis. According to the international prognostic index (IPI), 74% of pts were high-intermediate (HI) and high (H) risk; according to National Comprehensive Cancer Network (NCCN)-IPI, 68% of pts were HI and H risk. According to the cell of origin (COO) 63% of pts presented non-germinal center DLBCL. A complete metabolic response (CMR) on interim PET/CT was observed in 65% of pts, and 71% presented a CMR on end-of-treatment (EOT) PET/CT. At EOT the overall response rate was 80%. Seven percent of pts presented CNS relapse. The 5-year PFS was 58.5% and the 5-year overall survival (OS) was 69.4%.

Higher SDmax was significantly associated with BM and CNS involvement at diagnosis (p=0.034 and p=0.049, respectively), as well as with CNS relapse (p=0.048). No statistically significant association was found between SDmax and the following features: lymphocyte count of < 1000/μL (p=0.505), albumin levels < 3.2 g/dL (p=0.140), LDH > ULN (p=0.501), IPI risk ≥ HI risk (p=0.719), NCCN-IPI ≥ HI risk (p=0.239) and COO (p=0.493).

There was a tendency, although not significant, towards a lower rate of CMR on interim and EOT PET/CT in pts with higher SDmax (p=0.072 and p=0.077, respectively). A higher SDmax had no statistically significant impact on either PFS (HR 0.456; p=0.500) or OS (HR 0.339; p=0.561).

The combination of NCCN-IPI and SDmax allowed to identify a subgroup of pts with low (L) risk and low-intermediate (LI) risk NCCN-IPI and higher (≥ 4.1cm) SDmax with inferior 5-year OS (87.5% vs 50%; HR 4.64; p=0.031) and with a trend towards an inferior 5-year PFS (75% vs 50%; HR 1.92; p=0.166).

Conclusions: In our cohort, a higher SDmax was associated with some high-risk features, as well as with a trend towards an inferior response to treatment. Additionally, the combination of SDmax and NCCN-IPI allowed to identify a subgroup of L risk and LI risk pts with inferior OS and a trend towards an inferior PFS. Interpretation of these results may be limited by the small sample size. The role of Dmax in DLBCL prognostication may therefore require further validation before widespread clinical use.