Baseline Staging Evaluation in Lymphoma: The Role of FDG PET, CT, and Bone Marrow Biopsy

The revised response criteria for malignant lymphoma (Cheson et al JCO 25:579 2007) incorporated FDG PET for determination of response. This was strongly recommended for patients with curable lymphomas (diffuse large B-cell lymphoma [DLBCL] and Hodgkin lymphoma [HL]). For incurable lymphomas PET was not recommended unless response was a major trial endpoint. Part of the reservation regarding the use of PET for response evaluation in incurable lymphomas was based on the potential variability in FDG avidity at baseline. This retrospective review was performed to understand the frequency of FDG avidity across a range of lymphoma histologies and compare diagnostic yield to CT. Furthermore, diagnostic utility of bone marrow biopsy was evaluated in patents with DLBCL and HL compared to identification of disease by FDG PET and CT scan.

After obtaining a waiver of authorization from the MSKCC Institutional Review Board patients the DAVInCI data mining tool was used to retrospectively identify patients with the diagnosis of lymphoma who had both a FDG-PET scan and CT at diagnosis available in the PACS. Either the FDG-PET or the CT had to be informative (which excluded patient with CS I resected disease). The sites of disease were recorded for PET (including SUVs) and for CT. The impact on the CS at diagnosis was determined. DLBCL and HL patients with bone (B) or bone marrow (BM) involvement were identified by being positive on any one modality: BM biopsy; FDG-PET; or CT.

Data, including imaging, from 522 incident cases of lymphoma were reviewed. FDG PET performed for lymphoma at initial diagnosis demonstrated FDG-avid disease in 97.3% (508/522) of cases; there was some variability across histologies (Table 1). There was a strong correlation between CT and PET. PET identified more disease in 0–32.3% of cases depending on histology and CT was more informative in 3.2–33.3% of cases. CT tended to be more informative where the median SUV of PET was relatively low (SLL, MZL, MCL). The impact of FDG PET on Ann Arbor stage was modest but was greatest for SLL and T-cell lymphoma (Table 1, Change Stage).

Identification of bone (B) and bone marrow (BM) disease in HL and DLBCL was examined. 57 patients with DLBCL were found to have B/BM involvement by BM, CT, or PET. PET was the most sensitive test for B/BM disease (50/57 88%). CT identified bone disease in 33/37 (58%) of cases but these were strictly a subset of the PET positive cases. However, BM biopsy identified involvement in 7 (15.2%) cases which were negative by PET. For HL, 20 patients were found to have B/BM disease. FDG PET identified all 20 cases, CT 12 cases and BM biopsy only 3.

FDG-PET is positive in baseline in 97% of cases of lymphoma across all histologies. However, there is discordance between CT and PET in 28% of cases. Therefore, for clinical trials baseline contrast enhanced CT and FDG-PET are both necessary at baseline to full identify sites of disease. Outside the setting of a clinical trial, clinical discretion should be used in choosing the appropriate pre-treatment imaging. In the case of B/BM disease FDG-PET was the most informative modality in DLBCL and HL. However, bone marrow biopsy remains an essential part of the diagnostic evaluation of DLBCL since 7/20 bone marrow positive cases were not identified on FDG-PET or CT. The utility of the bone marrow biopsy in HL was questionable as FDG PET identified all cases of B/BM disease.

No relevant conflicts of interest to declare.