Bone Marrow Biopsy for the Staging of Non-Hodgkin's Lymphoma: Bilateral or Unilateral Trephine Biopsy?

Abstract

Aim

The occurrence of unilateral involvement in bilateral bone marrow trephine biopsies in non-Hodgkin's lymphomas (NHL) at disease onset (10-20% of cases) has been reported since the early 70s. Therefore, although these studies were based on small series, the use of bilateral bone marrow biopsies has become the rule. However, the clinical value of this procedure has never been clearly established. The aim of the present study was to ascertain the true value of bilateral bone marrow biopsy in the staging of NHL.

Study Design

We examined 368 cases of NHL (A-H according to the Working Formulation) (WF), without leukemic involvement of the peripheral blood, in order to evaluate: 1) the incidence of unilateral bone marrow involvement; 2) the percentage of patients who, as a result of unilateral bone marrow involvement, changed from stages I-II to stage IV; 3) assessment of response to therapy for patients with both bilateral or unilateral bone marrow involvement.

Results

In the A-C NHL groups of WF there was a unilateral bone marrow involvement of 8.8%. Overall, bone marrow involvement induced a change from clinical stages I-II to stage IV in 5.6% of cases, a figure which would correspond to a false negative rate of 2.8%, if unilateral bone marrow biopsy was performed. In the D-F and G, H groups of WF, unilateral involvement was 10.1% and 8.5% respectively; the change in stage from I-II to IV by unilateral bone marrow involvement respectively amounted to 1.4% and 2.8%, which correspond to respective false negative rates of 0.7% and 1.4%.

Conclusions

On the basis of these results and of the present therapeutic strategies, we propose: bilateral bone marrow biopsy for clinical stages I-II of all NHL; no bone marrow biopsy at disease onset for clinical stages III and IV of A to H histologic subtypes of the WF; unilateral bone marrow biopsy (A-C subtypes of the WF) or bilateral (D-H of the WF), after the regression of extramedullary localizations.