0 1992 Wiley-Liss, Inc.
Cytometry 13:445-447 (1992)
MEETING REPORT
Clinical Applications of Cytometry: 6th Annual Meeting J.V. Giorgi, P.E. Hurtubise, L.S. Cram,J.W. Parker, and M.F. La Via Department of Medicine, UCLA School of Medicine and Jonsson Comprehensive Cancer Center, Los Angeles 90024 (J.V.G.) and Department of Pathology, University of Southern California School of Medicine, Los Angeles, 90032 (J.W.P.), California; Department of Diagnostic Immunology, University of Cincinnati Medical Center, Cincinnati, Ohio, 45267 (P.E.H.); Life Sciences Division, Los Alamos National Laboratory, Los Alamos, New Mexico 87545 (L.S.CJ; and Department of Pathology and Laboratory of Medicine, Medical University of South Carolina, Charleston, South Carolina 29425 (M.F.L.) Received for publication February 3, 1992;accepted February 3, 1992
The Sixth An n u al Clinical Applications of Cytometry Meeting was held September 11-14,1991, in Charleston, SC. Attendance reached a record 470. The meeting provides a f o r u m f o r interactions among investigators w h o utilize cytometry as a tool in their clinical immunology, cell biology, hematology, and cancer investigations. Clinical lab o rato ry directors and their technical staff find the meeting of practical value because of the presentation of new applications that the y can take home to their own laboratories. The emphasis of the meeting is on advances in the application of cytometry to clinical problems. Often, advances result from new d y e s or reagents or improved instru-
The Sixth Annual Clinical Applications of Cytometry Meeting opened with the Distinguished Lecture tha t was delivered this year by Dr. Noel Warner (Becton Dickinson Immunocytometry Systems, San Jose, CA). The highlight of the presentation was a description of results obtained using fluorescence in situ hybridization (FISH). Slides were shown using FISH with both image and flow cytometry. Among other applications, FISH allows precise assessment of the amount of viral nucleic acid in infected cells, e.g., viral load in HIV infected subjects. Other applications are quantitation of cytokine and oncogene RNA. Fluorescencelabeled synthetic oligonucleotides complementary to the sequences of interest are used as probes. Cells are permeabilized, and the relevant probe is hybridized to the RNA. Cells can be simultaneously stained for cell surface immunofluorescence to provide multiparametric information. Issues of sensitivity, specific-
mentation. Sometimes they result f r o m advances in biology that make the studies possible. Occasionally a new w a y of looking at the same data provides a useful answer. In every case, the effort is to provide a reliable, s tra ig h tfo rw a rd way to quantitate biologic information in order to provide improved diagnosis or treatment of human disease. 0 1992 Wiley-Liss,Inc.
Key terms: Cytometry, immunophenotyping, DNA ploidy analysis, bone marrow transplantation, pediatric AIDS, leukemiaflymphoma diagnosis, Diagnostic flow cytometry
ity, standardization, and quantitation are currently foremost in the minds of those developing this novel approach. The next several years will provide the opportunity to collect data on the equally critical issue of whether this new technique will provide information that is clinically useful for patient management. Presentations related to hematopoietic neoplasms and bone marrow transplantation were featured during the first day of the meeting. Major conceptual advances were presented that enhance the interpretation of results from immunophenotyping of acute leukemias (Michael Borowitz, Duke University, Durham, NC) and lymphomas (Susan Wormsley, Cytometry Associates, San Diego, CA). Both speakers emphasized that multiparameter analysis, combining light scatter with surface, cytoplasmic, and nuclear markers, has replaced single-color analysis. Panels of markers are used in dual-color and triple-color custom combinations
446
GIORGI ET AL.
to yield the maximum amount of clinically significant information. Conventional analysis of histograms, i.e., gating on the lymphocyteimonocytelight scatter region and reporting the percent positive, is no longer considered acceptable for most leukemia/lymphoma immunophenotype applications. Instead, the patterns of reactivity of the markers with the normal cells in the preparation are used to identify the neoplastic population, and the characteristics of the neoplastic population are described qualitatively and quantitatively. Then, a decision tree is used to assign the neoplastic population to a B, T, or myeloid lineage. The utility of a multi-dimensional approach for diagnosis of leukemias was underscored by Michael Loken (BioLogics Consulting, Los Altos, CA) who used CDllb, CD16, and CD15 to describe a normal pattern of myeloid cell differentiation in the bone marrow. He analyzed the cell surface phenotype of 100 cases of acute myeloid leukemia against this normal background and found that each leukemia had light scatter and cell surface properties that distinguished it from any normal human bone marrow cell counterpart. A distinct cluster of leukemic cells could be distinguished in every leukemic sample. Each leukemia was unique from all the others, with no repeats in the aberrant phenotypes among the 100 cases examined. These findings have important diagnostic and therapeutic implications. Minimal residual disease can be diagnosed or the effects of chemotherapy as well as relapses can be monitored in individual patients, especially if the characteristics of the patient’s neoplasm are known. Identification of the cell surface antigenic phenotype of the human stem cell has long been a goal of investigators in the field of bone marrow transplantation. Cell sorting continues to contribute to accomplishing this goal. Curt Civin (Johns Hopkins Oncology Center, Baltimore, MD) described the CD34 antigen, a molecule expressed on about 1%of adult bone marrow cells, which can be used to positively select progenitors of hematopoietic cells that can be used for transplantation. Greater details regarding the cell surface phenotype of stem cell “candidates” was provided by several speakers in a session organized by Joseph Laver (Medical University of South Carolina). Peter Lansdorp (Terry Fox Laboratory, Vancouver, BC) indicated that these cells have the size of large lymphocytes, stain weakly with antibodies against CD71 (transferrin receptor), are 80% CD45RO+/CD45RA-, and are CD34bright.In vitro, these cells begin to increase their numbers only after 5 weeks of culture. Lansdorp calls them long-term culture initiating cells and they are the precursors for clonogenic cells (CFU) that quickly form myeloid and/or erythroid colonies in semisolid culture medium. Further information on the phenotype, growth factor requirements, and in vivo activity of the putative stem cell was provided by Edward Srour (Indiana University). He defined the putative stem cell as CD34+ HLADR- CD15-. In agreement with Lansdorp, Srour also
showed that the human stem cell exhibits low level staining with the supravital mitochondria1 dye rhodamine 123 (R123), suggesting that the cells have a metabolic state midway between resting and activated. After 10 weeks in vitro, murine mast cell growth factor (c-kit ligand) and a GM-CSMIL-3 fusion protein induced a 1.5 x 105-foldincrease of hematopoietic clonogenic cells. During a fascinating and provocative presentation, Dr. Srour described that in vivo, 20 x lo3 CD34 HLA-DR- CD15- R123d”” cells transplanted into fetal sheep resulted in chimerism (6% mature human leukocyte elements including T, B, and NK cells), thereby supporting the contention that these cells are indeed stem cells. The practical advantages of knowing the detailed phenotype of the human stem cell are numerous. The knowledge that these cells are CD34+ has already been put to use by Salvatore Siena (Instituo Nazionale Tumori, Milan, Italy). He has estimated the number of CD34+ cells in bone marrow and peripheral blood of cancer patients recovering from pancytopenia induced by intensive chemotherapy. Flow cytometric information, in vitro cell function, and actual patient hematopoietic recovery have been correlated. Siena monitors the number of CD34 cells in each patient who is a candidate for autologous bone marrow transplantation to support myeloblastic cancer therapies. He harvests marrow (the best source of stem cells) and peripheral blood (the best source of progenitor cells) when the number of circulating CD34+ cells peak. Here, flow cytometry is being used in a real-time basis in a clinical setting to determine when large-scale collection of these cells by leukapheresis should be performed and how many consecutive leukapheresis procedures are necessary to harvest quantities of circulating hematopoietic progenitor cells sufficient for successful transplant. Dr. Siena also provided phenotypic data to functionally distinguish two types of CD34+ cells. Thus, on one hand, CD34+ CD33- cells are putative stem cells: they are the long-term culture initiating cells described by Lansdorp and are required for longterm patient engraftment and survival. On the other hand, CD34+ CD33+ cells are the hematopoietic progenitor cells, also known as colony forming units (CFU) in soft agar: these cells are necessary for rapid neutrophi1 engraftment and patient survival during the first month of transplant. After intensive chemotherapy for cancer, a threefold expansion of the circulating hematopoietic progenitor cells occurs. The contribution that immunophenotyping, cell sorting, and functional studies have made to the rapid increase in our understanding of the immune system was highlighted during the session on pediatric immunodeficiency. A practical approach to the evaluation of primary immunodeficiency was described by Douglas Barrett (University of Florida, Gainesville, FL) that makes use of dual-color immunofluorescence analysis to enumerate functionally distinct lumphocyte subsets. In addition to functional subsets, the lineage relationship of +
+
CLINICAL APPLICATIONS O F CYTOMETRY: 6TH MEETING
447
subsets of T cells has also been clarified using cell sort- gen is not a marker for cytotoxic T cells. HLA-DR' ing (Loran Clement, UCLA School of Medicine, Los CD8 T cell levels were notably elevated in all HIVAngeles CA). Thus, CD4 CD45RA+/RO- cells (func- infected children, even in those (the majority) who did tionally suppressor inducer cells) have been shown to not have circulating cytotoxic T cells. Thus as noted in be the precursors for CD4 CD45RA-/RO+ cells (func- numerous other clinical situations, immune function is tionally helper cells); CD45RAt/RO- cells can give restricted to phenotypically defined lymphocyte subrise to CD45RA-/RO+ cells via an activation-depen- sets, but the presence of a particular maker on a cell does not guarantee that the cell will be functinal. dent post thymic differentiation pathway. Looking toward the future, Peter Rabinovitch (UniCD45RNRO isoform expression was studied on CD8 lymphocytes in HIV-infected children by Theresa Cal- versity of Washington, Seattle) described remarkable velli (Albert Einstein College of Medicine, Bronx, NY). findings in patients with premalignant gastrointestiElevated levels of CD8 cells that express the CD45RO nal disease. In these premalignant patients, aneuisoform were associated with pediatric HIV disease. Up ploidy was already present in biopsy material. Presto 50% of the lymphocytes in late stage pediatric HIV ence of aneuploidy was a more reliable indicator of risk disease were CD45RO+ CD8 cells compared with 4%of progression toward cancer than changes in histology. in normal infants. Ongoing investigations should an- These data suggest th a t it should be possible to select swer the question of whether or not the elevation in the patients with premalignant conditions who are at the percentage of CD45RO+ CD8 cells will be of value in highest risk for progression to cancer and follow them early diagnosis of pediatric HIV infection. by endoscopic surveillance. The goal is to detect proAdvances that have been made in the standardiza- gression to malignancy when treatment is still likely to tion of immunophenotypic analysis in children as well be successful. Selection of patients in whom the outas adult HIV infected patients was the topic of a pre- come of cytometry studies would influence patient care meeting session organized by Dr. Jonathan Kagan (Di- was also recommended by J i m Bennington (Children's vision of AIDS, HIH, Bethesda, MD). This was the sec- Hospital, San Francisco, CAI. ond year that Dr. Kagan has organized a practical An improved system for interpretation of image analinformation-exchange workshop focusing on cytometry ysis DNA ploidy data (Bennington) and immunophein the management of the HIV-infected patient, and notyping data were featured in presentations during workshops on immunophenotyping in HIV disease the last day of the meeting. Automated programs for promise to be a permanent addition to the annual Clin- analysis of multi-color immunophenotyping data (Roy ical Applications of Cytometry meeting as long as there Overton, Gen Trak, Inc., Plymouth Meeting, PA), and is a need for them. expert systems that utilize cluster analysis (Gary SalzAdvances that have been made in elucidating the man, Los Alamos National Laboratory, Los Alamos, function of CD8 cells in HIV-infected infants were re- NM) were described. These software programs autoported by John L. Sullivan (University of Massachu- mate many of the data analysis decisions that are cursetts Medical Center, Worcester). The HLA-DR' CD8 rently made by the operator who is analyzing the data. cell subset, which is elevated in pediatric and adult They require that the program learn how to make comHIV infection, includes all the cells with MHC re- plex decisions. If successful, they promise to make imstricted cytotoxic activity. In contrast, the CD57 anti- munophenotyping more standardized and reproducible.
Cytometry 13:445-447 (1992)
MEETING REPORT
Clinical Applications of Cytometry: 6th Annual Meeting J.V. Giorgi, P.E. Hurtubise, L.S. Cram,J.W. Parker, and M.F. La Via Department of Medicine, UCLA School of Medicine and Jonsson Comprehensive Cancer Center, Los Angeles 90024 (J.V.G.) and Department of Pathology, University of Southern California School of Medicine, Los Angeles, 90032 (J.W.P.), California; Department of Diagnostic Immunology, University of Cincinnati Medical Center, Cincinnati, Ohio, 45267 (P.E.H.); Life Sciences Division, Los Alamos National Laboratory, Los Alamos, New Mexico 87545 (L.S.CJ; and Department of Pathology and Laboratory of Medicine, Medical University of South Carolina, Charleston, South Carolina 29425 (M.F.L.) Received for publication February 3, 1992;accepted February 3, 1992
The Sixth An n u al Clinical Applications of Cytometry Meeting was held September 11-14,1991, in Charleston, SC. Attendance reached a record 470. The meeting provides a f o r u m f o r interactions among investigators w h o utilize cytometry as a tool in their clinical immunology, cell biology, hematology, and cancer investigations. Clinical lab o rato ry directors and their technical staff find the meeting of practical value because of the presentation of new applications that the y can take home to their own laboratories. The emphasis of the meeting is on advances in the application of cytometry to clinical problems. Often, advances result from new d y e s or reagents or improved instru-
The Sixth Annual Clinical Applications of Cytometry Meeting opened with the Distinguished Lecture tha t was delivered this year by Dr. Noel Warner (Becton Dickinson Immunocytometry Systems, San Jose, CA). The highlight of the presentation was a description of results obtained using fluorescence in situ hybridization (FISH). Slides were shown using FISH with both image and flow cytometry. Among other applications, FISH allows precise assessment of the amount of viral nucleic acid in infected cells, e.g., viral load in HIV infected subjects. Other applications are quantitation of cytokine and oncogene RNA. Fluorescencelabeled synthetic oligonucleotides complementary to the sequences of interest are used as probes. Cells are permeabilized, and the relevant probe is hybridized to the RNA. Cells can be simultaneously stained for cell surface immunofluorescence to provide multiparametric information. Issues of sensitivity, specific-
mentation. Sometimes they result f r o m advances in biology that make the studies possible. Occasionally a new w a y of looking at the same data provides a useful answer. In every case, the effort is to provide a reliable, s tra ig h tfo rw a rd way to quantitate biologic information in order to provide improved diagnosis or treatment of human disease. 0 1992 Wiley-Liss,Inc.
Key terms: Cytometry, immunophenotyping, DNA ploidy analysis, bone marrow transplantation, pediatric AIDS, leukemiaflymphoma diagnosis, Diagnostic flow cytometry
ity, standardization, and quantitation are currently foremost in the minds of those developing this novel approach. The next several years will provide the opportunity to collect data on the equally critical issue of whether this new technique will provide information that is clinically useful for patient management. Presentations related to hematopoietic neoplasms and bone marrow transplantation were featured during the first day of the meeting. Major conceptual advances were presented that enhance the interpretation of results from immunophenotyping of acute leukemias (Michael Borowitz, Duke University, Durham, NC) and lymphomas (Susan Wormsley, Cytometry Associates, San Diego, CA). Both speakers emphasized that multiparameter analysis, combining light scatter with surface, cytoplasmic, and nuclear markers, has replaced single-color analysis. Panels of markers are used in dual-color and triple-color custom combinations
446
GIORGI ET AL.
to yield the maximum amount of clinically significant information. Conventional analysis of histograms, i.e., gating on the lymphocyteimonocytelight scatter region and reporting the percent positive, is no longer considered acceptable for most leukemia/lymphoma immunophenotype applications. Instead, the patterns of reactivity of the markers with the normal cells in the preparation are used to identify the neoplastic population, and the characteristics of the neoplastic population are described qualitatively and quantitatively. Then, a decision tree is used to assign the neoplastic population to a B, T, or myeloid lineage. The utility of a multi-dimensional approach for diagnosis of leukemias was underscored by Michael Loken (BioLogics Consulting, Los Altos, CA) who used CDllb, CD16, and CD15 to describe a normal pattern of myeloid cell differentiation in the bone marrow. He analyzed the cell surface phenotype of 100 cases of acute myeloid leukemia against this normal background and found that each leukemia had light scatter and cell surface properties that distinguished it from any normal human bone marrow cell counterpart. A distinct cluster of leukemic cells could be distinguished in every leukemic sample. Each leukemia was unique from all the others, with no repeats in the aberrant phenotypes among the 100 cases examined. These findings have important diagnostic and therapeutic implications. Minimal residual disease can be diagnosed or the effects of chemotherapy as well as relapses can be monitored in individual patients, especially if the characteristics of the patient’s neoplasm are known. Identification of the cell surface antigenic phenotype of the human stem cell has long been a goal of investigators in the field of bone marrow transplantation. Cell sorting continues to contribute to accomplishing this goal. Curt Civin (Johns Hopkins Oncology Center, Baltimore, MD) described the CD34 antigen, a molecule expressed on about 1%of adult bone marrow cells, which can be used to positively select progenitors of hematopoietic cells that can be used for transplantation. Greater details regarding the cell surface phenotype of stem cell “candidates” was provided by several speakers in a session organized by Joseph Laver (Medical University of South Carolina). Peter Lansdorp (Terry Fox Laboratory, Vancouver, BC) indicated that these cells have the size of large lymphocytes, stain weakly with antibodies against CD71 (transferrin receptor), are 80% CD45RO+/CD45RA-, and are CD34bright.In vitro, these cells begin to increase their numbers only after 5 weeks of culture. Lansdorp calls them long-term culture initiating cells and they are the precursors for clonogenic cells (CFU) that quickly form myeloid and/or erythroid colonies in semisolid culture medium. Further information on the phenotype, growth factor requirements, and in vivo activity of the putative stem cell was provided by Edward Srour (Indiana University). He defined the putative stem cell as CD34+ HLADR- CD15-. In agreement with Lansdorp, Srour also
showed that the human stem cell exhibits low level staining with the supravital mitochondria1 dye rhodamine 123 (R123), suggesting that the cells have a metabolic state midway between resting and activated. After 10 weeks in vitro, murine mast cell growth factor (c-kit ligand) and a GM-CSMIL-3 fusion protein induced a 1.5 x 105-foldincrease of hematopoietic clonogenic cells. During a fascinating and provocative presentation, Dr. Srour described that in vivo, 20 x lo3 CD34 HLA-DR- CD15- R123d”” cells transplanted into fetal sheep resulted in chimerism (6% mature human leukocyte elements including T, B, and NK cells), thereby supporting the contention that these cells are indeed stem cells. The practical advantages of knowing the detailed phenotype of the human stem cell are numerous. The knowledge that these cells are CD34+ has already been put to use by Salvatore Siena (Instituo Nazionale Tumori, Milan, Italy). He has estimated the number of CD34+ cells in bone marrow and peripheral blood of cancer patients recovering from pancytopenia induced by intensive chemotherapy. Flow cytometric information, in vitro cell function, and actual patient hematopoietic recovery have been correlated. Siena monitors the number of CD34 cells in each patient who is a candidate for autologous bone marrow transplantation to support myeloblastic cancer therapies. He harvests marrow (the best source of stem cells) and peripheral blood (the best source of progenitor cells) when the number of circulating CD34+ cells peak. Here, flow cytometry is being used in a real-time basis in a clinical setting to determine when large-scale collection of these cells by leukapheresis should be performed and how many consecutive leukapheresis procedures are necessary to harvest quantities of circulating hematopoietic progenitor cells sufficient for successful transplant. Dr. Siena also provided phenotypic data to functionally distinguish two types of CD34+ cells. Thus, on one hand, CD34+ CD33- cells are putative stem cells: they are the long-term culture initiating cells described by Lansdorp and are required for longterm patient engraftment and survival. On the other hand, CD34+ CD33+ cells are the hematopoietic progenitor cells, also known as colony forming units (CFU) in soft agar: these cells are necessary for rapid neutrophi1 engraftment and patient survival during the first month of transplant. After intensive chemotherapy for cancer, a threefold expansion of the circulating hematopoietic progenitor cells occurs. The contribution that immunophenotyping, cell sorting, and functional studies have made to the rapid increase in our understanding of the immune system was highlighted during the session on pediatric immunodeficiency. A practical approach to the evaluation of primary immunodeficiency was described by Douglas Barrett (University of Florida, Gainesville, FL) that makes use of dual-color immunofluorescence analysis to enumerate functionally distinct lumphocyte subsets. In addition to functional subsets, the lineage relationship of +
+
CLINICAL APPLICATIONS O F CYTOMETRY: 6TH MEETING
447
subsets of T cells has also been clarified using cell sort- gen is not a marker for cytotoxic T cells. HLA-DR' ing (Loran Clement, UCLA School of Medicine, Los CD8 T cell levels were notably elevated in all HIVAngeles CA). Thus, CD4 CD45RA+/RO- cells (func- infected children, even in those (the majority) who did tionally suppressor inducer cells) have been shown to not have circulating cytotoxic T cells. Thus as noted in be the precursors for CD4 CD45RA-/RO+ cells (func- numerous other clinical situations, immune function is tionally helper cells); CD45RAt/RO- cells can give restricted to phenotypically defined lymphocyte subrise to CD45RA-/RO+ cells via an activation-depen- sets, but the presence of a particular maker on a cell does not guarantee that the cell will be functinal. dent post thymic differentiation pathway. Looking toward the future, Peter Rabinovitch (UniCD45RNRO isoform expression was studied on CD8 lymphocytes in HIV-infected children by Theresa Cal- versity of Washington, Seattle) described remarkable velli (Albert Einstein College of Medicine, Bronx, NY). findings in patients with premalignant gastrointestiElevated levels of CD8 cells that express the CD45RO nal disease. In these premalignant patients, aneuisoform were associated with pediatric HIV disease. Up ploidy was already present in biopsy material. Presto 50% of the lymphocytes in late stage pediatric HIV ence of aneuploidy was a more reliable indicator of risk disease were CD45RO+ CD8 cells compared with 4%of progression toward cancer than changes in histology. in normal infants. Ongoing investigations should an- These data suggest th a t it should be possible to select swer the question of whether or not the elevation in the patients with premalignant conditions who are at the percentage of CD45RO+ CD8 cells will be of value in highest risk for progression to cancer and follow them early diagnosis of pediatric HIV infection. by endoscopic surveillance. The goal is to detect proAdvances that have been made in the standardiza- gression to malignancy when treatment is still likely to tion of immunophenotypic analysis in children as well be successful. Selection of patients in whom the outas adult HIV infected patients was the topic of a pre- come of cytometry studies would influence patient care meeting session organized by Dr. Jonathan Kagan (Di- was also recommended by J i m Bennington (Children's vision of AIDS, HIH, Bethesda, MD). This was the sec- Hospital, San Francisco, CAI. ond year that Dr. Kagan has organized a practical An improved system for interpretation of image analinformation-exchange workshop focusing on cytometry ysis DNA ploidy data (Bennington) and immunophein the management of the HIV-infected patient, and notyping data were featured in presentations during workshops on immunophenotyping in HIV disease the last day of the meeting. Automated programs for promise to be a permanent addition to the annual Clin- analysis of multi-color immunophenotyping data (Roy ical Applications of Cytometry meeting as long as there Overton, Gen Trak, Inc., Plymouth Meeting, PA), and is a need for them. expert systems that utilize cluster analysis (Gary SalzAdvances that have been made in elucidating the man, Los Alamos National Laboratory, Los Alamos, function of CD8 cells in HIV-infected infants were re- NM) were described. These software programs autoported by John L. Sullivan (University of Massachu- mate many of the data analysis decisions that are cursetts Medical Center, Worcester). The HLA-DR' CD8 rently made by the operator who is analyzing the data. cell subset, which is elevated in pediatric and adult They require that the program learn how to make comHIV infection, includes all the cells with MHC re- plex decisions. If successful, they promise to make imstricted cytotoxic activity. In contrast, the CD57 anti- munophenotyping more standardized and reproducible.
