Comparison of Bone Marrow Biopsy, Flow Cytometry and PCR Assays To Detect Bone Marrow Involvement in B-Cell Non-Hodgkin Lymphomas.

Bone marrow biopsy, (BMB) is essential to detect bone marrow (BM) infiltration in B-cell non-Hodgkin lymphomas (NHLs). Flow cytometry (FC) and PCR for clonal IgH rearrangement are considered as ancillary methods, but there is increasing evidence for their clinical usefulness. Observations dealing with combined use of the three methods still are lacking. Thus we carried out a retrospective study about the usefulness of an integrated approach to detect BM infiltration in NHLs. 193 patients suffering from NHLs (79 at presentation, 114 after chemotherapy alone or with: Rituximab, Campath-1, autologous BM transplantation), who had undergone simultaneous execution of BMB, and FC, and PCR from the myeloaspirate on the same iliac crest, were evaluated. BMB was carried out according to standard methods (infiltration pattern and immunohistochemistry). FC was performed using three-color staining, including CD45, to identify: κ/λ ratio, specific phenotype for CLL, MCL and HCL. PCR included identification of IgH rearrangement (CDR3 and VH families), BCL-1/JH translocation for MCL and BCL-2/JH translocation for follicular lymphoma.

BMB, FC and PCR agreed in 142 cases and showed infiltration in 74 and lack of infiltration in 68. Cases at presentation were characterized by higher percentages of concordance than cases during the post-chemotherapy (84,8% vs 65.6%). Discrepant results were obtained in 51 cases (26.4%), 13 at presentation and 38 after treatment. In 17 specimens (8.8% of all cases, 33.3% of discordant cases), BM infiltration was detected only by PCR. In 12 of these samples (3 untreated and 14 treated) small B-cell percentages (0.50 ± 0.72, mean ± SD; range 0.02–3.00%) were present at FC. The remaining 5 cases (2.6%) were characterized by a lack of surface Ig expression and absence of specific phenotype: BMB was negative but IgH was clonal. 2 other cases with lack of surface Ig expression (for 7 cases in total, 3.6%), BMB-/PCR- were identified.

Conversely, in 10 samples (5.2% of all cases, 19.6% of discordant specimens) PCR failed to detect BM infiltration, which was demonstrated by both FC and BMB. These specimens were characterized by high B-cell percentages (7 ± 8.25, mean ± SD; range 0.3–26.0%) and were obtained from 5 untreated and 5 treated patients.

The remaining discordant cases were: 7 treated cases with BMB+/PCR+ and FC-; 6 cases (3 treated and 3 untreated) with FC+ and BMB-/PCR-; 3 treated cases with BMB- and FC+/PCR+; 3 cases (1 untreated and 2 treated) with BMB+ and FC-/PCR-. In the 3 treated cases with lack of amplification by PCR, the following results were observed: FC+/BMB+ in 1 case; FC-/BMB- in the remaining 2.

Our data show that no single method is able to identify all cases of BM involvement in NHLs. BMB is actually considered the gold standard, however the combination of the three assays can increase the yields of detection of minimal residual disease. In fact, considering the three assays together as gold standard, BMB alone has a sensibility of 82.1%, a specificity of 96,3%, with 1.6% of false positive but 10.4% of false negative. The PCR can increase the sensibility and FC can than be considered as a valid confirmation assay, able to solve the cases when BMB and PCR show discrepancy.

To conclude, the three assays are necessary to evaluate BM infiltration in NHLs, because BMB alone underestimate the BM involvement, especially following treatment.