BACKGROUND: Among marginal zone lymphomas (MZLs), bone marrow (BM) involvement features are well established in the splenic marginal zone lymphoma (SMZL); few data are available for extranodal marginal zone lymphoma (EMZL) and nodal marginal zone lymphoma (NMZL). PATIENTS AND METHODS: Incidence and patterns of histologic BM involvement are studied in 120 MZL patients (48 SMZL, 59 EMZL, 13 NMZL) at onset and during follow-up; relationships between clinical features, BM histology and flow cytometry (FC) are analyzed. RESULTS: At diagnosis, BM involvement occurs in 90% SMZL, 22% EMZL and 54% NMZL (P<0.0001); at reevaluation, incidence raises to 96% in SMZL and 34% in EMZL. Concordance between histology and FC is found in 87% of cases; most discordant cases have positive histology but negative FC. SMZL and EMZL show a nodular BM infiltration; the interstitial pattern is frequent in NMZL (P<0.0001); sinusoidal localization is typical of SMZL, frequent in NMZL and occasional in EMZL (P=0.0001). Stage, leukemic disease, B symptoms, more than one extranodal involved site, splenomegaly, elevated beta2-microglobulin, serum monoclonal component, International Prognostic Index (IPI) and age-adjusted IPI are directly related to BM infiltration. CONCLUSIONS: The different prevalence of BM involvement in MZL subtypes reflects their heterogeneous dissemination modalities; histology seems more sensible than FC to detect BM infiltration; development of BM involvement during follow-up is typical of EMZL.
BACKGROUND: Among marginal zone lymphomas (MZLs), bone marrow (BM) involvement features are well established in the splenic marginal zone lymphoma (SMZL); few data are available for extranodal marginal zone lymphoma (EMZL) and nodal marginal zone lymphoma (NMZL). PATIENTS AND METHODS: Incidence and patterns of histologic BM involvement are studied in 120 MZL patients (48 SMZL, 59 EMZL, 13 NMZL) at onset and during follow-up; relationships between clinical features, BM histology and flow cytometry (FC) are analyzed. RESULTS: At diagnosis, BM involvement occurs in 90% SMZL, 22% EMZL and 54% NMZL (P<0.0001); at reevaluation, incidence raises to 96% in SMZL and 34% in EMZL. Concordance between histology and FC is found in 87% of cases; most discordant cases have positive histology but negative FC. SMZL and EMZL show a nodular BM infiltration; the interstitial pattern is frequent in NMZL (P<0.0001); sinusoidal localization is typical of SMZL, frequent in NMZL and occasional in EMZL (P=0.0001). Stage, leukemic disease, B symptoms, more than one extranodal involved site, splenomegaly, elevated beta2-microglobulin, serum monoclonal component, International Prognostic Index (IPI) and age-adjusted IPI are directly related to BM infiltration. CONCLUSIONS: The different prevalence of BM involvement in MZL subtypes reflects their heterogeneous dissemination modalities; histology seems more sensible than FC to detect BM infiltration; development of BM involvement during follow-up is typical of EMZL.