Metabolic tumor volume – hype or hope?

Michel Meignan, Université Paris-Est Créteil, and colleagues published findings in the Journal of Clinical Oncology in August 2016 aiming to determine the prognostic impact of the Total Metabolic Tumor Volume (TMTV) measured at baseline with [¹⁸F]FDG/PET-CT scans and its added value to these models.¹ In this pooled analysis, a total of 185 patients with Follicular Lymphoma (FL) had baseline scans suitable for TMTV calculation: 106 from the LYSA PET-FOL trial², 38 from the LYSA PRIMA trial³, and 41 from the FIL FOLL05 trial⁴. Progression Free Survival (PFS) and Overall Survival (OS) data were available for 181 patients, with a median follow-up of 63.5 months from registration. Of the 181 patients, 159 had a post-induction PET for response assessment reported with Deauville criteria.

• Median TMTV = 297cm³ (IQR: 135–567cm³). 510cm³ TMTV was selected and validated as optimal cut-off for PFS and OS value.
• TMTV >510cm³ was significantly associated with stage III–IV disease; greater nodal, extranodal, and bone marrow involvement; higher FLIPI and FLIPI2 scores; and increased β₂m and LDH
• Patients with TMTV >510cm³ (29%) had a significantly inferior outcome compared to lower TMTV patients: 5-year PFS 32.7% vs 65.1% (HR 2.90, P = <.001), 5-year OS 84.8% vs 94.7% (HR 3.45, P = .013)
• On multivariate analysis, TMTV (HR 2.3, P = .002) and FLIPI2 score (HR 2.2, P = .002) were independent predictors of PFS
• Three risk categories were devised, with significantly different PFS rates, based on presence or absence of TMTV >510cm³ or FLIPI2 score of 3–5
  • Group 1: low TMTV and low FLIPI2 score (n = 88); PFS = 69%
  • Group 2: high TMTV or intermediate–high FLIPI2 (n = 55); PFS = 46%
  • Group 3: high TMTV and intermediate–high FLIPI2 score (n = 24); PFS = 20%

The authors stated that no specific study thus far has identified the prognostic benefit of TMTV in identifying patients with FL at high-risk of disease progression and early death after current immunochemotherapeutic strategies. The group concluded that TMTV is a robust pre-treatment predictor of outcome in high-tumor-burden FL patients and TMTV is a strong predictor of early progression within the first or second year after starting treatment. Combining TMTV with FLIPI2 score can identify patients at high risk of progression and so warrants further investigation and validation as a biomarker for development of first-line PET-adapted approaches in FL, contributing to developing risk-adapted individualized treatment approaches and thereby avoiding under- and over-treatment of patients with FL.

Heiko Schöder and Craig Moskowitz of the Memorial Sloan Kettering Cancer Center, New York, provided an accompanying editorial, published in the Journal of Clinical Oncology in September 2016⁵, noting some shortcomings the results obtained by Meignan et al.

• Meignan et al. used three different approached for defining the optimal TMTV cut-off for PFS: X-tile analysis, Receiver Operating Curve (ROC) analysis, and restricted cubic spline. Schöder and Moskowitz noted that in the ROC analysis, follow-up time was ignored
• The retrospective nature of the analysis is a huge limitation: most patients (1,634/1,819) from the two LYSA trials and one FIL trial were ineligible and it can be assumed that the PET scanners used were not cross-calibrated to ensure the same SUV was measured and the same lesion was assessed
• It was observed that PET scans were performed on three different generations of scanner from three different vendors, which potentially affects the calculation of Standardized Uptake Values (SUVs) and therefore TMTV
• Uptake times for FDG were not standardized, which can lead to variable FDG SUV not due to differences in tumor biology but due to technical reasons. Moreover, no range or median of uptake times were provided

Schöder and Moskowitz state that is it probably too early to call TMTV a new prognostic biomarker in lymphoma. They also mentioned that TMTV combined with treatment response assessed on interim PET imaging, rather than FLIPI2 score, leads to better segregation of prognostic groups⁶. In the future, we need to determine the biologic reasons for why TMTV is associated with patient outcome and ask how these data can improve clinical practice. Schöder and Moskowitz also hypothesize that the most accurate prognostic tool will combine clinical, imaging and biologic factors and so it will be some time for a specific test to be standardized and accepted into clinical practice. Until this time, hypothesis-testing is needed for better risk stratification and treatment selection of lymphoma patients in studies exploring imaging techniques and potential biomarkers. The editorial concludes by stating that “the data provided by Meignan et al. moves us one step along in this direction.”