Flow Cytometry Diagnoses Hodgkin's in Lymph Nodes With High Sensitivity/Specificity
Flow Cytometry Diagnoses Hodgkin's in Lymph Nodes With High Sensitivity/Specificity
The diagnosis of classical Hodgkin lymphoma (CHL) has been made in tissue sections as attempts to identify neoplastic Hodgkinand Reed Sternberg (HRS) cells in lymph nodes by flow cytometry (FC) have been unsuccessful. However, we have recently demonstratedthat HRS cells can be identified by FC, often present as T-cell–HRS-cell rosettes. In this study, we examined the usefulnessof a novel 9-color (CD95–Pacific blue/CD64–fluorescein isothiocyanate/CD30-phycoerythrin [PE]/CD45-PE–Texas red/CD40–PE cyanine[Cy]5.5/CD20-PECy7/CD15-allophycocyanin [APC]/CD71-APC–AlexaFluor A700/CD5-APC-Cy7), single tube FC assay to diagnose CHLin lymph nodes. We used the FC assay to determine diagnostic sensitivity and specificity in 279 blindly identified and 141selected (for specimen type or cytopreparation morphologic features suggesting CHL) tissues. Of the 53 morphologically defined CHL cases identified (10 in the unselected group; 43 in the selected group), the FC assay diagnostic sensitivity and specificity were 88.7% and 100%, respectively. With the current availability of 8 (or more) color clinical flow cytometers, this assay can now be applied to routinely immunophenotype and confirm a diagnosis of CHL or as an adjunct to immunohistochemical analysis.
Classical Hodgkin lymphoma (CHL) is a unique type of B-cell lymphoma in which the neoplastic Hodgkin and Reed-Sternberg (HRS) cells are infrequent (usually < 1% of the leukocytes in a lymph node), have an unusual immunophenotype (CD30+ and CD15+ and CD3-, CD20-, and CD45-, features often used to confirm the diagnosis in tissue sections), have characteristic morphologic features, and demonstrate rosettes of T cells bound to an HRS cell. The HRS cells are embedded in an inflammatory, non-neoplastic infiltrate that includes lymphocytes, histiocytes, plasma cells, and eosinophils.
Although flow cytometry (FC) is useful in the diagnosis of many hematopoietic neoplasms and can routinely detect very small populations (often < 0.01% of the leukocytes), its usefulness in contributing to the diagnosis of CHL involving lymph nodes has been limited. Although most FC studies of CHL have looked at changes in the composition of reactive lymphocytes, such as the CD4/CD8 ratio of the associated T cells, the HRS cells have generally escaped detection by this technique.
We recently described a method of identifying HRS cells in lymph nodes by FC using a 2-tube, 10-color assay that used a combination of antibodies reactive with antigens on HRS cells (CD15, CD30, CD40, CD71, and CD95), B cells (CD19 and CD20), T cells (CD3 or CD5), macrophages (CD64), and leukocytes (CD45). It is interesting that a subset of the HRS cells in most cases examined demonstrated expression of T-cell antigens (CD3 and CD5) and bright CD45 owing to the presence of HRS-cell–T-cell rosettes (aggregates). Preincubation with unlabeled monoclonal antibodies to the adhesion molecules CD2 and LFA-1 (present on T cells) and CD54 and CD58 (present on HRS cells) resulted in a loss of expression of T-cell antigens and a decrease in (but not absence of) CD45 expression, corresponding to HRS cells that are free of bound T cells (“naked” HRS cells).
The initial study examined antigen expression in 27 cases of CHL, 29 non-CHL large cell neoplasms, and 23 reactive lymph nodes. Although this method enabled HRS cells to be identified in 89% of CHL cases, and no non-CHL neoplasms or reactive lymph nodes demonstrated HRS populations, specimens were chosen for further study (because of a clinical suspicion of CHL, the presence of atypical cells in the lymph node suspensions, or a characteristic inflammatory background) and, consequently, the study may have been subject to selection bias. In addition, a limited number of cases were examined, and no cases of anaplastic large cell lymphoma (ALCL), a neoplasm often in the differential diagnosis of CHL, were present in the previous study. The present study expands on this initial work to define the true sensitivity and specificity of a novel single-tube, 9-color FC assay for the diagnosis of CHL.
Abstract and Introduction
Abstract
The diagnosis of classical Hodgkin lymphoma (CHL) has been made in tissue sections as attempts to identify neoplastic Hodgkinand Reed Sternberg (HRS) cells in lymph nodes by flow cytometry (FC) have been unsuccessful. However, we have recently demonstratedthat HRS cells can be identified by FC, often present as T-cell–HRS-cell rosettes. In this study, we examined the usefulnessof a novel 9-color (CD95–Pacific blue/CD64–fluorescein isothiocyanate/CD30-phycoerythrin [PE]/CD45-PE–Texas red/CD40–PE cyanine[Cy]5.5/CD20-PECy7/CD15-allophycocyanin [APC]/CD71-APC–AlexaFluor A700/CD5-APC-Cy7), single tube FC assay to diagnose CHLin lymph nodes. We used the FC assay to determine diagnostic sensitivity and specificity in 279 blindly identified and 141selected (for specimen type or cytopreparation morphologic features suggesting CHL) tissues. Of the 53 morphologically defined CHL cases identified (10 in the unselected group; 43 in the selected group), the FC assay diagnostic sensitivity and specificity were 88.7% and 100%, respectively. With the current availability of 8 (or more) color clinical flow cytometers, this assay can now be applied to routinely immunophenotype and confirm a diagnosis of CHL or as an adjunct to immunohistochemical analysis.
Introduction
Classical Hodgkin lymphoma (CHL) is a unique type of B-cell lymphoma in which the neoplastic Hodgkin and Reed-Sternberg (HRS) cells are infrequent (usually < 1% of the leukocytes in a lymph node), have an unusual immunophenotype (CD30+ and CD15+ and CD3-, CD20-, and CD45-, features often used to confirm the diagnosis in tissue sections), have characteristic morphologic features, and demonstrate rosettes of T cells bound to an HRS cell. The HRS cells are embedded in an inflammatory, non-neoplastic infiltrate that includes lymphocytes, histiocytes, plasma cells, and eosinophils.
Although flow cytometry (FC) is useful in the diagnosis of many hematopoietic neoplasms and can routinely detect very small populations (often < 0.01% of the leukocytes), its usefulness in contributing to the diagnosis of CHL involving lymph nodes has been limited. Although most FC studies of CHL have looked at changes in the composition of reactive lymphocytes, such as the CD4/CD8 ratio of the associated T cells, the HRS cells have generally escaped detection by this technique.
We recently described a method of identifying HRS cells in lymph nodes by FC using a 2-tube, 10-color assay that used a combination of antibodies reactive with antigens on HRS cells (CD15, CD30, CD40, CD71, and CD95), B cells (CD19 and CD20), T cells (CD3 or CD5), macrophages (CD64), and leukocytes (CD45). It is interesting that a subset of the HRS cells in most cases examined demonstrated expression of T-cell antigens (CD3 and CD5) and bright CD45 owing to the presence of HRS-cell–T-cell rosettes (aggregates). Preincubation with unlabeled monoclonal antibodies to the adhesion molecules CD2 and LFA-1 (present on T cells) and CD54 and CD58 (present on HRS cells) resulted in a loss of expression of T-cell antigens and a decrease in (but not absence of) CD45 expression, corresponding to HRS cells that are free of bound T cells (“naked” HRS cells).
The initial study examined antigen expression in 27 cases of CHL, 29 non-CHL large cell neoplasms, and 23 reactive lymph nodes. Although this method enabled HRS cells to be identified in 89% of CHL cases, and no non-CHL neoplasms or reactive lymph nodes demonstrated HRS populations, specimens were chosen for further study (because of a clinical suspicion of CHL, the presence of atypical cells in the lymph node suspensions, or a characteristic inflammatory background) and, consequently, the study may have been subject to selection bias. In addition, a limited number of cases were examined, and no cases of anaplastic large cell lymphoma (ALCL), a neoplasm often in the differential diagnosis of CHL, were present in the previous study. The present study expands on this initial work to define the true sensitivity and specificity of a novel single-tube, 9-color FC assay for the diagnosis of CHL.
