Abstract
Background
DLBCL is the most common non-Hodgkin lymphoma (NHL), making up about 30%-40% of NHL in the U.S. PET-CT is recommended as the most accurate imaging technique in DLBCL for staging and response assessment. Pretreatment assessment of PET-CT scan derived metrics such as TMTV has been shown to correlate with PFS and/or overall survival (OS) in DLBCL (Sasanelli 2014) We attempted to replicate this finding using EFS at 24 months as a primary endpoint and compare it with pre-treatment TMTV, TLG and cell of origin (COO).
Methods
47 pts with newly diagnosed DLBCL and treated with R-CHOP at our institution between 2014 to 2018 were identified from our electronic medical record system for retrospective analysis after IRB approval. All pts had a pretreatment PET-CT scan available for TMTV measurement. All pts had a pretreatment biopsy which were reviewed along with their clinical information regarding treatment outcome and follow up. Patients were classified as to germinal center B cell (GCB) and non-GCB based on immunochemistry using the Hahn's algorithm. PET-CT scans were reviewed by two nuclear medicine physicians using synovia software, and measurements for TMTV and TLG were recorded. TMTV was calculated using a threshold of 41% of the max pixel value (based on prior studies) to draw the volume of interest (VOI) for a lesion. Pooled t-test was performed to compare TMTV, TLG and COO with EFS at 24 mos. Chi-Square test compared TMTV with COO
Results
Median age of pts was 58 years, with a median duration of follow up of 26 months. There were 33% with limited stage (Stage I or II) and 67% were advanced stage (Stage III or IV). The mean pretreatment TMTV and pretreatment TLG was 295cm3 and 4519 units. 49% were GCB subtype and 47 % non-GCB. Amongst all patients 19.2 % had an event within 24 mos. When TMTV was compared to EFS at 24 months the mean TMTV was 304 for those who had an event versus 294 without (p=0.95). TLG compared to EFS at 24 months showed a mean TLG of 3391 for those who had an event versus 4914 without (P=0.40). GCB and non-GCB had mean TMTV of 264 and 339 respectively with p =0.59. COO when compared to TLG had means of 4365 and 4933 for GCB and non-GBB respectively with p=0.79.Whereas there was no correlation between stage and COO (p=0.4296) TMTV correlated with Ann Arbor staging (p=0.0002).
Conclusion
This retrospective study failed to demonstrate a correlation between pre-treatment TMTV, TLG, COO and EFS at 24 months revealing the lack of prognostic significance of pretreatment PET scan derived metrics in DLBCL. Prior studies with TMTV did not evaluate EFS at 24 months as an endpoint and therefore, longer follow up might be needed to demonstrate prognostic significance of pretreatment TMTV minimizing it clinical significance. The different subtypes of DLBCL based on COO as assessed by Hahns algorithm also did not differ in their disease burden as measured by TMTV.
No relevant conflicts of interest to declare.
Author notes
Asterisk with author names denotes non-ASH members.
