Age, PS, LDH serum level, CSF protein concentration, and involvement of deep structures of the brain were independent predictors of survival. A prognostic score including these five parameters seems advisable in distinguishing different risk groups in PCNSL patients. The proposed score and its relevance in therapeutic decision deserve to be validated in further studies.

Standardized guidelines for the baseline evaluation and response assessment of primary CNS lymphoma (PCNSL) are critical to ensure comparability among clinical trials for newly diagnosed patients. The relative rarity of this tumor precludes rapid completion of large-scale phase III trials and, therefore, our reliance on the results of well-designed phase II trials is critical. To formulate this recommendation, an international group of experts representing hematologic oncology, medical oncology, neuro-oncology, neurology, radiation oncology, neurosurgery, and ophthalmology met to review current standards of reporting and to formulate a consensus opinion regarding minimum baseline evaluation and common standards for assessing response to therapy. The response guidelines were based on the results of neuroimaging, corticosteroid use, ophthalmologic examination, and CSF cytology. A critical issue that requires additional study is the optimal method to assess the neurocognitive impact of therapy and address the quality of life of PCNSL survivors. We hope that these guidelines will improve communication among investigators and comparability among clinical trials in a way that will allow us to develop better therapies for patients.

Intravascular large B-cell lymphoma (IVLBCL) is a rare form of diffuse LBCL characterized by preferential intravascular growth of malignant lymphocytes, aggressive behavior, and an often fatal course. IVLBCL usually affects elderly patients with poor performance status, elevated lactic dehydrogenase serum levels, anemia, and B symptoms. It displays some differences in clinical presentation among diverse geographical areas, mostly between patients diagnosed in Western countries and Japan. In addition, data from the literature suggest that pathologic diagnostic criteria as well as clinical features of this disease may be broader than described in current classification scheme(s). Under the sponsorship of the International Extranodal Lymphoma Study Group, clinicians and pathologists with interest in IVLBCL, coming from Western and Eastern countries, joined to reach a consensus on defining features as well as to focus on the most urgent unresolved issues in IVLBCL. To this end, a representative group of IVLBCL patients coming from both the aforementioned geographical areas were collectively analyzed. Additional features of IVLBCL were proposed both under clinical and pathologic stand points. At the meeting, it emerged that IVLBCL may have additional histopathologic/cytologic definition criteria with respect to those currently recommended, some clinical features are not randomly distributed worldwide, recent therapeutic approaches, such as anti-CD20-containing regimens, may improve outcome, and kidney, spleen, and liver involvement may show peculiar histopathologic features. Finally, a provisional practical diagnostic approach to hemophagocytosis-associated patients and a proposal for the most useful criteria in the settings of differential diagnosis are included.

A retrospective survey of patients with pathologically reviewed extragastric mucosa-associated lymphoma tissue (MALT) lymphomas from 20 institutions was performed. A total of 180 patients with histologically confirmed diagnosis of extragastric MALT lymphomas were studied. Their median age was 59 years (range, 21-92 years). Ann Arbor stage I disease was present in 115 patients (64%) and stage II disease in 16 (9%). Most cases were in the low or low-intermediate risk groups according to the International Prognostic Index (IPI). Forty-one (23%) patients had involvement of more than one extranodal site at diagnosis and in 24 cases (13%) the lymphoma presented at multiple mucosal sites (9 of them with only mucosal involvement, without bone marrow or nodal disease). Lymph node involvement was present in 21%. Patients were treated with a variety of therapeutic strategies, including chemotherapy in 78 cases. The median overall survival (OS) was not reached; the 5-year OS rate was 90% (95% CI, 82%-94%), the 5-year causespecific survival (CSS) was 94% (95% CI, 87%-97%), and the 5-year progressionfree survival (PFS) was 60% (95% CI, 50%-70%). Multivariate analysis showed that Ann Arbor stage was significantly associated with longer OS, nodal involvement with longer CSS, and favorable IPI score with better PFS. At a median follow-up of 3.4 years, 48 patients (27%; 95% CI, 20%-34%) had a relapse, 6 (3%; 95% CI, 1%-7%) showed histologic transformation, and 18 (10%; 95% CI, 6%-15%) experienced the development of a second tumor. Our data confirm the indolent nature of nongastric MALT lymphomas and the high rate of patients presenting with disseminated disease, which, when limited to mucosal sites, was not associated with a poorer outcome. (Blood. 2003;101: 2489-2495)