The volumetric imaging parameters of 18F-FDG PET/CT have been shown to be useful not only in the prognostic evaluation of patients with DLBCL, but also in other types of lymphomas such as Hodgkin lymphoma  and peripheral T-cell lymphoma. In the present study, we assessed the prognostic value of 18FDG PET/CT volume-based PET parameters (TMTV) prior to pretreatment in patients with PGI-DLBCL. The results indicated that patients with a TMTV greater than 415.0 cm3 had lower survival and that TMTV was an independent predictor of survival outcomes after multivariate analysis. This finding is consistent with the results of Song et al. In the study of Song et al, TMTV was a valid prognostic indicator of survival for Ann Arbor Stage IE/IIE patients greater than 160.1 cm3 with PGI-DLBCL. The discrepancy between the optimal thresholds in this previous study compared to those in our present study may be partially explained by the different stages of patients included; Stage I/II patients were included in the study by Song et al, while Stage I–IV patients were included in our present study. In addition, SUV is currently a popular semiquantitative, easy-to-calculate index of the FDG metabolic rate. However, TMTV, as a volume metric quantitatively presenting total tumor burden, has been found to be a better predictor than SUVmax for survival in our present study and in the study by Song et al. These results support TMTV as a useful volumetric imaging parameter for developing better treatment plans by providing more accurate prognostic information.
In recent years, with the progress of molecular genetic research, it has been shown that molecular heterogeneity of tumors is directly correlated with treatment responses and prognoses and is crucial for allowing the possibility of individualized risk-adapted therapy . A series of studies have shown that DLBCL patients with concurrent expression of MYC and BCL2 proteins have a significantly poorer survival outcome.[20-22] In the current study, DE accounted for 30.8% of PGI-DLBCL patients which is consistent with the results of PGI-DLBCL patients reported by Xia et al(30%)  but was higher in all of the DLBCL patients (21%) included in Johnson et al, suggesting that concurrent expression of MYC and BCL-2 is more frequently found in PGI-DLBCL patients. Moreover, our present study, as well as that of Xia et al, indicated that PGI-DLBCL patients with DE have significantly poorer survival outcomes. In our present study, BCL2 protein expression alone did not significantly predict PFS or OS, which is contrast to the findings of Cakmak et al . The reason for this discrepancy may be explained by the different IHC-sample positive criteria, in which 30% of cells were stained with the antibody in our study while only 10% of cells were stained in Cakmak et al. In our present study, MYC and BCL-2 expression determined via IHC—as a convenient, rapid and low-cost method to detect protein expression—was found to have clinical significance in predicting prognoses in PGI-DLBCL patients.
PGI-DLBCL, characterized as highly heterogeneous, has been treated with various modalities in the past, but optimal treatment strategies have remained controversial. The IPI has been the most widely used metric for determining prognoses among patients with DLBCL since it was first published in 1993. However, because of the addition of rituximab (R) to conventional CHOP or CHOP-like regimens, the capacity of the IPI model to discriminate between risk groups has been questioned in the rituximab era.[25-26] Moreover,the combination of molecular and imaging characteristics at diagnosis has been demonstrated to improve DLBCL patient stratification in several studies.[27-28] Accordingly, the desire for more efficient risk stratification urged us to further assess a monogram comprised with TMTV and DE performance as prognostic indicators in patients with PGI-DLBCL. In present study, our results suggest that the monogram (PFS:χ2=32.178, P<0.001; OS:χ2=23.091, P<0.001) could yield a better stratification than IPI (PFS:χ2=8.738, P=0.003; OS:χ2=8.216, P=0.004) for PGI-DLBCL patients. Taken together, TMTV combined with DE could be better for risk stratification and selecting patients for tailor treatments.
This single-institution retrospective study was constrained by several limitations. The differences in thresholds used for delineating tumors when calculate TMTV may have resulted in inconsistencies among our results. An isocontour threshold of Liver SUVmean + SD × 2.00 and a fixed threshold of 42% of the intratumoral SUVmax were chosen by Mixue Rosj Gallicchio et al., respectively . Although the limits of different tumor-delineation techniques have been reported ,  most of these studies have shown that, independent of the segmentation approach used, a higher TMTV in DLBCL patients is more strongly associated with a higher risk of unfavorable survival outcomes compared to that of SUVmax.[16,30] Additionally, because of the limited number of patients in our present study, a selection bias may have confounded the analysis of our results. Therefore, a prospective clinical trial with a larger sample size of PGI-DLBCL patients is needed to provide a more reliable prediction of survival in such patients.
TMTV and DE identified a subset of PGI-DLBCL patients with poor survival outcomes. Furthermore, our findings suggest that TMTV combined with DE may assist clinicians in more accurately stratifying PGI-DLBCL patients and providing them with more risk-tailored therapies.