Utility of Bone Marrow Biopsy and Aspirate for Staging of Diffuse Large B Cell Lymphoma in the Era of Positron Emission Tomography With 2-Deoxy-2-[Fluorine-18]fluoro-deoxyglucose Integrated With Computed Tomography.

BACKGROUND

About one-third of patients with diffuse large B cell lymphoma (DLBCL) have lymphomatous bone marrow involvement (BMI) at the time of diagnosis, and bone marrow aspirate/biopsy (BMAB) is considered the gold standard to detect such involvement. [F] fluorodeoxyglucose positron emission tomography combined with computed tomography (PET-CT), has become standard pretreatment imaging in DLBCL and may be a noninvasive alternative to BMAB to ascertain BMI. Prior studies have suggested that PET-CT scan may obviate the need for BMAB as a component for staging patients with newly diagnosed DLBCL, but this is not yet a standard of practice. The aim of this retrospective study was to determine the accuracy of PET-CT in detecting BMI in DLBCL and to define 2-year and 5-year overall survival based on BMI by BMAB versus PET-CT.

METHODS

We reviewed institutional records of all patients with newly diagnosed DLBCL between January 2004 and December 2013 who underwent pretreatment PET-CT and BMAB. PET-CT images were visually assessed for BMI, including the posterior iliac crest. Patients with primary mediastinal DLBCL, previous history or coexistence of another lymphoma subtype, and those with a nondiagnostic BMAB, and in whom the PET-CT did not show marrow signal abnormality, were excluded from the analysis. Ann Arbor stage was determined using PET-CT with and without the contribution of BMAB, and the proportion of stage IV cases by each method was measured.

RESULTS

Among 99 eligible patients, the median age was 62 years (range, 24-88 years), 62 (63%) were male, 53 (53%) had elevated serum lactate dehydrogenase, and 17 (16%) had an Eastern Community Oncology Group performance status of > 2. Thirteen (12%) patients had more than 1 extra-nodal site of lymphoma involvement. Revised International Prognostic Index score was 1 in 39 (37%) patients, 2 in 42 (40%) patients, 3 in 20 (19%) patients, and 4 in 4 (4%) patients. A total of 38 (36%) patients had BMI established by either PET-CT (n = 24; 24%), BMAB (n = 14; 14%), or by both modalities (n = 12; 12%). Twelve (50%) of the 24 patients with positive PET-CT had BMI by DLBCL, whereas only 2 (3%) of the 75 patients with negative PET-CT showed BMI. BMAB upstaged 1 (2%) of the 53 stage I/II patients to stage IV. The sensitivity and specificity of PET-CT scan to detect BMI by DLBCL was 86% (95% confidence interval, 51.9%-95.7%) and 87% (95% confidence interval, 76%-92%), respectively. Eighty-five (86%) patients had concordant results between lymphomatous BMAB and PET-CT (12 patients were positive for both; 73 patients were negative for both), and 14 (14%) patients had a discordant interpretation (2 patients were positive by BMAB and negative by PET-CT, and 12 patients were negative by BMAB and positive by PET-CT). The positive predictive value of PET-CT was only 50%, whereas the negative predictive value was 98%. The accuracy of PET-CT was 86%. Although patients with positive BMAB had inferior 5-year overall survival estimates compared with those with negative BMAB (66% vs. 85%; P = .08), no such difference was demonstrated between PET-CT-positive and PET-CT-negative patients (79% vs. 83%; P = .30).

CONCLUSIONS

In patients with newly diagnosed DLBCL, PET-CT is accurate in detecting BMI by DLBCL. Although PET-CT has a very high negative predictive value for BMI, it overestimates the number of cases with marrow involvement by DLBCL. In clinical practice, routine BMAB may no longer be necessary for all patients with DLBCL who are staged by PET-CT, unless the results would change both staging and therapy. The prognostic implication of BMI identified by PET-CT compared with BMAB remains unknown. Whether a PET-CT precludes the need for a BMAB in patients with DLBCL remains to be evaluated in a prospective study.