Prognostic significance of architectural patterns in non-Hodgkin lymphomas

The role of clonal selection in the pathogenesis of an autoreactive human B cell lymphoma

To study the association of autoimmunity and human B cell neoplasia, we have established a model of a B cell lymphoma which expresses a pathogenic autoantibody of defined specificity. The Ig VH gene expressed in this neoplasm was analyzed longitudinally using clinical specimens taken from the splenic lymphoma (S) at diagnosis and from lymph node relapses 3 and 4 yr later (N3 and N4). Southern analysis and oligonucleotide hybridization experiments demonstrated that clonally related predominant and minor tumor cell populations were present in S at diagnosis, and that the minor population became the predominant population in the relapse specimens, N3 and N4. Although the Ig specificity and idiotype were the same at diagnosis and at both relapses, analysis of the expressed VH gene sequences showed 14 base changes between S and N3, and 2 further changes at N4. Little sequence heterogeneity was observed at each sampling time, indicating that the ongoing mutation frequency was low. The relevant germline precursor VH gene was determined from autologous germline DNA and compared to the expressed genes. Based on the pattern of shared and unshared mutations, we were able to establish the genealogic relationship of the germline VH gene and the expressed clonotypes of S, N3 and N4. Taken together, the findings from Southern blotting, oligonucleotide hybridization, and sequence analysis permit us to describe a molecular aspect of tumor progression, "clonotypic shift", wherein subpopulations of the malignant clone, marked by different V gene clonotypes, emerge and predominate at different time points in the evolution of the lymphoma. Furthermore, the sequential and nonrandom pattern of the VH mutations, correlated with the observed conservation of autospecificity and idiotype, implies that clonal selection may have influenced the pathogenesis of the lymphoma.

Follicular lymphomas: assessment of prognostic factors in 127 patients followed for 10 years

Response to treatment, histologic progression, and survival of 127 patients with follicular lymphoma were analyzed according to histologic, clinical, and biological parameters. Histologic parameters were percentage of large cells (less than 10%, 41 patients; 10-25%, 38 patients; 25-50%, 11 patients; greater than or equal to 50%, 30 patients), percentage of diffuse areas, presence of intrafollicular proliferation or fibrosis, and mitotic scale. Eighty percent of the patients achieved complete remission (CR) with radiotherapy for localized stages and various chemotherapy regimens for disseminated stages. Three patients did not respond to treatment, and 23 were in partial remission (PR) at the end of treatment. Median survival time was 9.25 years. A constant death rate of 8% per year was observed without plateau. Histologic progression was observed in 32 patients; it occurred at a constant rate during the first six years and plateaued thereafter. Factors associated with low response rate were stage IV, B symptoms, high tumor mass, and two or more extranodal sites. Factors associated with histologic progression were bone marrow involvement and two or more extranodal sites. Factors associated with poor survival were advanced stage, two or more extranodal sites, bone marrow involvement, high lactate dehydrogenase level, and absence of interfollicular fibrosis. The percentages of large cells and diffuse areas had no influence on prognosis, nor had the type of treatment. Median survival has not been reached for CR patients and was four years for PR patients (P less than 0.0001). The LNH-84 prognostic index for aggressive lymphomas, based on tumor mass, number of extranodal sites, stage, and LDH level, is a clear-cut indicator of prognosis in follicular lymphomas too.(ABSTRACT TRUNCATED AT 250 WORDS)

Follicular low-grade non-Hodgkin's lymphoma: long-term outcome with or without tumor progression

Long-term outcome for 127 patients with follicular low-grade lymphoma was investigated. Therapy included radiotherapy (n = 23), low toxicity chemotherapy with or without radiotherapy (n = 76), or more intensive chemotherapy (n = 22). 6 patients had no initial therapy. Complete remission was obtained in 67% of patients. For patients under 60 years of age median survival was 8.7 yr compared with 3.8 yr for older patients, but survival from lymphoma was identical for the two age-groups: 75% at 5 yr, and 58% at 10 yr. The relatively low tumor mortality contrasted with a relapse-free survival of 30% at 10 yr, and relapse 8-9 yr after first remission. Examining the disease topography and the stability of histologic subtype in 78 patients with recurrent lymphoma, two types of relapse with different prognoses were identified: 1) with tumor progression (lymphoma dissemination to atypical extranodal sites and/or histologic conversion to an intermediate/high-grade lymphoma) seen in 56% of patients with a survival from lymphoma of 13% at 10 yr; and 2) without tumor progression (involvement of nodal sites, and unchanged histology) seen in 44% with a survival from lymphoma of 77% at 10 yr. Actuarial risk of tumor progression was 44% at 5 yr, and 67% at 10 yr. Except from the negative impact of a large tumor burden, it was not possible to identify patients with high risk for tumor progression. More important than all pretreatment factors was poor response to initial therapy (p = 0.0001). Due to lack of reliable risk factors, it is recommended that all younger patients be treated with the intention of achieving complete remission; a significant fraction might be curable.

Prognostic significance of differential cell counts in non-Hodgkin's lymphomas

The use of differential counting on H & E stained sections is proposed as a simple means to define low grade malignant and high grade malignant cytologic categories in non-Hodgkin's lymphomas. Differential counts were performed in lymphoma biopsies from 616 cases. In each biopsy we counted 100 lymphoma cells and classified each cell as belonging to a "small", "medium-sized", or "large" cell type. The results indicate the presence of two prognostically distinct cytologic categories: a low grade and a high grade malignant. Lymphomas with less than 10% "large" cells represented low grade malignant cytology. Included in this category were also the, mainly follicular, lymphomas with more then 70% "medium-sized" cells (up to about 25% "large" cells). In addition to the cytologic category, the architectural pattern is of major prognostic importance. We recommend the use of three prognostic categories in non-Hodgkin's lymphomas: I) Favourable architecture + favourable cytology. II) Unfavourable architecture + favourable cytology. III) Unfavourable cytology.

Application of quantifiable criteria in the Lukes and Collins classification of non-Hodgkin's lymphomas

Measurements of nuclear size and differential counts among six lymphoma cell types were performed on H & E stained sections. In differential counting, the definition of cell types was based on nuclear shape, chromatin pattern, and nucleoli. In a pilot study comprising 93 patients we found actual nuclear size inadequate for use in lymphoma classification. This was due to: 1. great overlap among cytological types; 2. no independent prognostic value of mean nuclear area; 3. contradictory terminology; the large cleaved type belonging to the small cell category (mean nuclear area below 40 microns2), and the small non-cleaved type belonging to the large cell category (mean nuclear area above 40 microns2). Differential counting--requiring about 10 min--was an easy way to meet the need for a more objective evaluation of the cellular composition in non-Hodgkin's lymphomas. Quantifiable criteria based on differential counts were applicable in subclassification of three T-cell and seven B-cell types with an intraobserver reproducibility of 80%. More than 25% "large" cell types in a differential count implied an unfavourable prognosis. In test material, using a semi-morphometric classification, a correct prognostic category was obtained in 92% of 461 lymphomas and correct sub-classification obtained in 68%.