AIDS RESEARCH AND HUMAN RETROVIRUSES Volume 8, Number 8, 1992 Mary Ann Liebert, Inc., Publishers
Summary: ZOLLA-PAZNER,1
SUSAN
Clinical
BONNIE J.
Immunology Working Group
MATHIESON,2
INTRODUCTION
MARY CLARE
WALKER,2
and BRUCE
WALKER3
that it also was desirable to include ELlSAs based series of peptides representing defined B-cell epitopes, including the V3 loop of gpl20, to evaluate the specificity of vaccinées' sera. It was suggested that these assays be done at selected investigational laboratories that could include collaboratores outside the AVCT network. (4) Titration of sera was deemed preferable to generation of qualitative information or of optical densities (OD) at standard serum dilutions. (5) Efforts should be initiated to develop isotype-specific ELISAs, particularly for IgA, for evaluation of vaccinée sera, and other fluid a consensus
Downloaded by East Carolina University from www.liebertpub.com at 01/11/19. For personal use only.
on a
Clinical Immunology Working Group was convened priorities for the evaluation of the humoral and cellular immune responses of volunteers in clinical trials undergoing vaccination with prototype AIDS vaccines. The difficulty of this task was acknowledged by the participants from the outset, given that the correlates of protective immunity in human immunodeficiency virus (HIV-1) infection are not yet identified. In the absence of more precise information about which immune responses are expected to confer protection, it was deemed impossible to identify, with certainty, responses to vaccines that would correlate with protection and translate into efficacy. A list of assays that could be expected to measure immunogenicity and the induction of functional forms of immunity was compiled. The assays were prioritized for performance: (a) on site at the five AIDS Vaccine Evaluation Units (AVEU), (b) at the Central Immunology Laboratory (CIL) (at Duke University), of the AIDS Vaccine Clinical Trials (AVCT) Network, or (c) by selected collaborating investigational laboratories. Five categories of evaluations were identified:
The
to define the
ELISA tests
enzyme-linked
immunosorbent assay
(ELISA) were discussed. ( 1 ) It was felt that commercial ELISAs
lysates were of limited value and should only be at a few periodic intervals during the vaccine trials of seronegative volunteers at the AVEUs. (2) To study vaccine-induced antibody responses, an ELISA incorporating the specific vaccine immunogen needs to be performed with appropriate control sera and antigens that include the same viral gene product derived from different HIV-1 isolates. It should be performed both at the AVEUs and the CIL. There was general accord that this immunogen-based ELISA had the highest priority of the serological assays. (3) Further, there was based
on
HIV
performed
on
entry and
Western blots These assays were considered of little use beyond screening of potential volunteers at entry for immunization with vaccines composed of HIV envelope proteins because of the ambiguity in reactivity for gp41 and gp 120 components. Western blots can be useful for evaluation of multigene product AIDS vaccines but, again, because titration of sera is needed and Western blots are expensive, this type of assay probably is suboptimal for evaluation of vaccine responses.
Neutralization assays
ASSAYS OF HUMORAL IMMUNITY
Several types of
specimens.
Here, it was agreed the testing should first be against the HIV isolate from which the vaccine immunogen was derived. However, a second level of assessment to evaluate cross-neutralization was considered essential and should include other HIV-1 isolates and specifically primary isolates. Neutralization of HIV-1 in the presence of human complement was urged in addition to neutralization in the absence of complement. Fusion inhibition assays should augment the HIV neutralization tests since they measure an independent functional parameter. For consistency, comparability, and conservation of scarce serum resources, the majority of these assays should be performed at the CIL. Because of the relative cost and technical limitations of HIV neutralization assays based on p24 antigen or reverse transcriptase activity readouts, alternative methodology should
'Laboratory Service, New York Veterans Affairs Medical Center, New York, NY 10010 and Department of Pathology, New York University Medical Center, New York, NY 10016. 2Vaccine Research and Development Branch. Basic Research and Development Program, Division of AIDS, National Institute of Allergy and Infectious Diseases, National Institutes of Health, Rockville. MD 20892. 'Department of Medicine, Harvard Medical School, Boston. MA 02114. 1441
1442
ZOLLA-PAZNER ET AL.
(1) It was considered that radioimmunoprecipitation assays (RIPA), although tedious and expensive, should be done on
CD44 CTL induction by all HIV-1 vaccine candidates tested to date by the AVCT network. Although the role of CD4+ CTL cells as a protective host immune defense has not been established, in vitro tests indicate that these cells are able to kill HIV-infected CD4+ lymphoblasts. Studies to determine whether CD4+ CTL can be induced in vitro from recently vaccinated individuals by mitogen plus cytokine stimulation were recommended. It was agreed that CD4+ CTL studies should be performed on a subset of subjects from each of the AIDS vaccine trials at specialized laboratories with particular expertise in this area. In addition, this Working Group felt that further studies were necessary to identify the potential role and mechanism of action of CD4+ CTL in infected persons as a host defense against retroviruses.
Antibody-dependent cell-mediated cytotoxicity
HIV-1-specific CD8+ CTL
(ADCC) assays
In vitro studies have suggested that CD8 f CTL may provide a protective defense against HIV-1 infection. It was agreed that CD8 + major histocompatibility complex (MHC) class I-restricted CTL responses to vaccination should be monitored. Many current vaccine products/vaccine strategies being explored may generate only low-level cellular immune responses at best. Therefore, standard assays, developed to evaluate CMI in HIV seropositive persons, must be adapted for use in vaccine recipients. In particular, most of the data indicate that specific in vitro stimulation strategies may be required to detect these HIV-specific CTL in uninfected, vaccine recipients. Use of cultures infected with recombinant HIV-1 pox viruses or HIV1-infected autologous lymphocyte strategies were discussed. Specific targets cell strategies were also discussed. To accommodate the HLA requirements for CTL recognition and to avoid the limitations imposed by HLA polymorphism, it was felt that autologous cells generally should be used as target cells. Additionally, assays should ultimately allow for quantitative assessment of cytolytic responses, but it was agreed that most current protocols must be considered to provide only qualitative assessments of such activities. Recommendations were also made to improve upon the available expression systems used to generate both stimulator and target cells. In particular, it was suggested that the widely employed recombinant vaccinia-HIV-1 method of envelope expression in target cells needs to be improved, since this system does not differentiate between MHC-restricted and non-MHC-restricted (ADCC, NK) cytolytic responses to the envelope protein using bulk cells. Where sufficient levels of CTL activity are detected, the surface phenotype of the CTL effector population should be defined.
be pursued and standardized. Additional efforts developing neutralization of HIV infection of monocytes/macrophages should be encouraged.
Assays for conformational epitopes
Downloaded by East Carolina University from www.liebertpub.com at 01/11/19. For personal use only.
selected serum samples. These assays might be especially useful to evaluate antigen-antibody interactions in the field phase. (2) There was strong feeling that it was necessary to evaluate conformational epitopes involved in CD4 binding. Many such assays are under development in specialized laboratories and it was recommended that vaccinée sera be made available to these laboratories.
ADCC
was
considered
a
well-established assay of immuno-
genicity and a potential measure of in vivo biological activity. However, since immunogenicity would be well-documented by the ELISAs, this test was not a high priority and it would be sufficient to perform it on selected sets of sera. It was proposed
that in vivo validation of ADCC as a correlate of protection should be ascertained with selected monoclonal antibodies in the
chimpanzee model.
Antibody-dependent enhancement (ADE) autoantibody assays
and
There was general accord that AIDS vaccinée sera should be screened for the presence of potentially adverse antibody responses. These tests can be performed as batched assays in labs with specialized expertise.
Complement-mediated assays Complement-mediated, antibody-dependent cytolysis and virolysis should also be carried out in specialized labs on selected vaccinée serum samples. ASSAYS OF CELL-MEDIATED
IMMUNITY (CMI) In general, the participants felt that it was extremely important also monitor parameters of cellular immune function in vaccine recipients. Although some animal studies have suggested that antibodies alone may confer protection against experimental challenge with HIV-1, HIV-2. or SIV. it was stressed that such studies have protected against intravenous challenge with cell-free virus and should not lead to the conclusion that cellular immunity would be unnecessary or ineffective as a host defense against naturally acquired infection. The following parameters of CMI were discussed. to
HIV-1-specific CD4+ cytotoxic
T
lymphocytes (CTL)
CD4+ CTL were detected when lymphocytes from recipients of the MicroGeneSys gpl60 subunit vaccine were stimulated in vitro with gpl60, and it has been possible to demonstrate
,
Lymphocyte proliferation/activation The consensus among the Working Group was that, although all HIV vaccines tested thus far have been shown to induce HIV antigen-specific, T-cell proliferation/activation, the relevance of such responses has not been established. It was felt that these studies should be continued as a general measure of T-cell memory. Several investigators underscored the caveat that the HLA genotype of the antigen presenting cell (APC) in vitro must be identical to that of the APC in vivo to insure functional readout in vitro. It was noted that a range of concentration of HIV antigen(s) should be employed routinely in lymphoprolif-
SUMMARY: CLINICAL IMMUNOLOGY WORKING GROUP eration assays. It was also stressed that results should be presented in a form that allows some means of comparison from study to study. In this regard, reporting raw numbers as well as stimulation indices was considered to be essential.
Natural killer (NK) cells
Although a number of studies have suggested a possible role for NK cells as an acute antiviral host defense, the consensus of the participants was that little information was likely to be gained by studying NK cell function in uninfected vaccine recipients. Thus, the limited blood samples available from these subjects should be conserved for studies of HI V-specific cellular immunity.
CD8
lymphocyte suppression of virus replication
Downloaded by East Carolina University from www.liebertpub.com at 01/11/19. For personal use only.
CD8+ lymphocytes that suppress the replication of HIV (or
SIV) in vitro have been reported by several laboratories, although the precise mechanism of action (cytolysis of infected cells or factor-mediated suppression of viral replication) has not been determined. It was agreed that more studies were needed to identify the soluble factor that has been postulated to mediate this suppression. Pilot studies to look for a similar response in selected vaccine recipients were suggested and would be expected to have the best chance of success if carried out in laboratories with specific expertise. It was stressed that samples would need to be made available to investigators outside the AVCT network for these studies.
ADCC ADCC was also considered from the standpoint of cellular immunity, particularly the antibody-armed effector cells identified in seropositive persons. It was proposed that further studies should be performed on a limited subset of subjects to determine under what conditions armed effectors are generated and detected.
ASSAY DEVELOPMENT FOR ASSESSMENT OF MUCOSAL IMMUNITY Current AIDS vaccine strategies are not expected to induce mucosal immunity, but it was considered timely and prudent to collect mucosal specimens; e.g., parotid saliva or seminal fluid from a selected subset of vaccinées. Adaptation or development of appropriate assays to detect mucosal immune responses to HIV antigens in vaccinées should be initiated forthwith.
1443
EVALUATION OF NONSPECIFIC MARKERS These tests will be of particular importance in postinfection immunization, but were not felt to be a high priority in seronegative vaccinées. The most prominent of the nonspecific markers is the enumeration of CD4 cells. AIDS vaccinées receiving envelope products that bind to CD4 will continue to be monitored for effects on CD4 T cells, although no adverse effects have been detected with vaccines in current trials. This parameter should be analyzed in terms of absolute counts of CD4 cells, the percentage of CD4 cells and the relative change in CD4 cell numbers from baseline in seropositive trials. But CD4 stabilization should not be the only criterion for vaccine effectiveness. The consensus of the group was that the usefulness of measuring beta-, microglobulin, neopterin, interferons. NKcell activity of delayed-type hypersensitivity to various common antigens, etc., is not established.
VIRAL MARKERS These tests were again felt to be of importance only in the evaluation of seropositive individuals undergoing therapeutic vaccination. Due to time constraints these tests could not be discussed in detail; however, it was noted that the AIDS Clinical Trials Group has developed or is developing standard protocols for quantitative viral cultures and the polymerase chain reaction for detection of HIV. In summary, the Clinical Immunology Working Group participants reached a consensus on the priorities for immunological assessment of individuals undergoing vaccination with prototype AIDS vaccines. The group felt there was a compelling need for more data defining the correlates of protective immunity to HIV. Without such information, it was deemed impossible to fully define what immune responses would need to be generated by an AIDS vaccine to achieve efficacy. Given these constraints, the group emphasized that vaccine strategies against HIV should be aimed at eliciting as many humoral and cellular virus-specific immune responses as possible. Furthermore, the adaptation and development of assays was considered essential to facilitate the measurement of specific parenteral and mucosal immune responses to future candidate AIDS vaccines. Address
reprint requests
to:
Susan Zolla-Pazner
New York
Department of Pathology University Medical Center New York. NY 10016
Summary: ZOLLA-PAZNER,1
SUSAN
Clinical
BONNIE J.
Immunology Working Group
MATHIESON,2
INTRODUCTION
MARY CLARE
WALKER,2
and BRUCE
WALKER3
that it also was desirable to include ELlSAs based series of peptides representing defined B-cell epitopes, including the V3 loop of gpl20, to evaluate the specificity of vaccinées' sera. It was suggested that these assays be done at selected investigational laboratories that could include collaboratores outside the AVCT network. (4) Titration of sera was deemed preferable to generation of qualitative information or of optical densities (OD) at standard serum dilutions. (5) Efforts should be initiated to develop isotype-specific ELISAs, particularly for IgA, for evaluation of vaccinée sera, and other fluid a consensus
Downloaded by East Carolina University from www.liebertpub.com at 01/11/19. For personal use only.
on a
Clinical Immunology Working Group was convened priorities for the evaluation of the humoral and cellular immune responses of volunteers in clinical trials undergoing vaccination with prototype AIDS vaccines. The difficulty of this task was acknowledged by the participants from the outset, given that the correlates of protective immunity in human immunodeficiency virus (HIV-1) infection are not yet identified. In the absence of more precise information about which immune responses are expected to confer protection, it was deemed impossible to identify, with certainty, responses to vaccines that would correlate with protection and translate into efficacy. A list of assays that could be expected to measure immunogenicity and the induction of functional forms of immunity was compiled. The assays were prioritized for performance: (a) on site at the five AIDS Vaccine Evaluation Units (AVEU), (b) at the Central Immunology Laboratory (CIL) (at Duke University), of the AIDS Vaccine Clinical Trials (AVCT) Network, or (c) by selected collaborating investigational laboratories. Five categories of evaluations were identified:
The
to define the
ELISA tests
enzyme-linked
immunosorbent assay
(ELISA) were discussed. ( 1 ) It was felt that commercial ELISAs
lysates were of limited value and should only be at a few periodic intervals during the vaccine trials of seronegative volunteers at the AVEUs. (2) To study vaccine-induced antibody responses, an ELISA incorporating the specific vaccine immunogen needs to be performed with appropriate control sera and antigens that include the same viral gene product derived from different HIV-1 isolates. It should be performed both at the AVEUs and the CIL. There was general accord that this immunogen-based ELISA had the highest priority of the serological assays. (3) Further, there was based
on
HIV
performed
on
entry and
Western blots These assays were considered of little use beyond screening of potential volunteers at entry for immunization with vaccines composed of HIV envelope proteins because of the ambiguity in reactivity for gp41 and gp 120 components. Western blots can be useful for evaluation of multigene product AIDS vaccines but, again, because titration of sera is needed and Western blots are expensive, this type of assay probably is suboptimal for evaluation of vaccine responses.
Neutralization assays
ASSAYS OF HUMORAL IMMUNITY
Several types of
specimens.
Here, it was agreed the testing should first be against the HIV isolate from which the vaccine immunogen was derived. However, a second level of assessment to evaluate cross-neutralization was considered essential and should include other HIV-1 isolates and specifically primary isolates. Neutralization of HIV-1 in the presence of human complement was urged in addition to neutralization in the absence of complement. Fusion inhibition assays should augment the HIV neutralization tests since they measure an independent functional parameter. For consistency, comparability, and conservation of scarce serum resources, the majority of these assays should be performed at the CIL. Because of the relative cost and technical limitations of HIV neutralization assays based on p24 antigen or reverse transcriptase activity readouts, alternative methodology should
'Laboratory Service, New York Veterans Affairs Medical Center, New York, NY 10010 and Department of Pathology, New York University Medical Center, New York, NY 10016. 2Vaccine Research and Development Branch. Basic Research and Development Program, Division of AIDS, National Institute of Allergy and Infectious Diseases, National Institutes of Health, Rockville. MD 20892. 'Department of Medicine, Harvard Medical School, Boston. MA 02114. 1441
1442
ZOLLA-PAZNER ET AL.
(1) It was considered that radioimmunoprecipitation assays (RIPA), although tedious and expensive, should be done on
CD44 CTL induction by all HIV-1 vaccine candidates tested to date by the AVCT network. Although the role of CD4+ CTL cells as a protective host immune defense has not been established, in vitro tests indicate that these cells are able to kill HIV-infected CD4+ lymphoblasts. Studies to determine whether CD4+ CTL can be induced in vitro from recently vaccinated individuals by mitogen plus cytokine stimulation were recommended. It was agreed that CD4+ CTL studies should be performed on a subset of subjects from each of the AIDS vaccine trials at specialized laboratories with particular expertise in this area. In addition, this Working Group felt that further studies were necessary to identify the potential role and mechanism of action of CD4+ CTL in infected persons as a host defense against retroviruses.
Antibody-dependent cell-mediated cytotoxicity
HIV-1-specific CD8+ CTL
(ADCC) assays
In vitro studies have suggested that CD8 f CTL may provide a protective defense against HIV-1 infection. It was agreed that CD8 + major histocompatibility complex (MHC) class I-restricted CTL responses to vaccination should be monitored. Many current vaccine products/vaccine strategies being explored may generate only low-level cellular immune responses at best. Therefore, standard assays, developed to evaluate CMI in HIV seropositive persons, must be adapted for use in vaccine recipients. In particular, most of the data indicate that specific in vitro stimulation strategies may be required to detect these HIV-specific CTL in uninfected, vaccine recipients. Use of cultures infected with recombinant HIV-1 pox viruses or HIV1-infected autologous lymphocyte strategies were discussed. Specific targets cell strategies were also discussed. To accommodate the HLA requirements for CTL recognition and to avoid the limitations imposed by HLA polymorphism, it was felt that autologous cells generally should be used as target cells. Additionally, assays should ultimately allow for quantitative assessment of cytolytic responses, but it was agreed that most current protocols must be considered to provide only qualitative assessments of such activities. Recommendations were also made to improve upon the available expression systems used to generate both stimulator and target cells. In particular, it was suggested that the widely employed recombinant vaccinia-HIV-1 method of envelope expression in target cells needs to be improved, since this system does not differentiate between MHC-restricted and non-MHC-restricted (ADCC, NK) cytolytic responses to the envelope protein using bulk cells. Where sufficient levels of CTL activity are detected, the surface phenotype of the CTL effector population should be defined.
be pursued and standardized. Additional efforts developing neutralization of HIV infection of monocytes/macrophages should be encouraged.
Assays for conformational epitopes
Downloaded by East Carolina University from www.liebertpub.com at 01/11/19. For personal use only.
selected serum samples. These assays might be especially useful to evaluate antigen-antibody interactions in the field phase. (2) There was strong feeling that it was necessary to evaluate conformational epitopes involved in CD4 binding. Many such assays are under development in specialized laboratories and it was recommended that vaccinée sera be made available to these laboratories.
ADCC
was
considered
a
well-established assay of immuno-
genicity and a potential measure of in vivo biological activity. However, since immunogenicity would be well-documented by the ELISAs, this test was not a high priority and it would be sufficient to perform it on selected sets of sera. It was proposed
that in vivo validation of ADCC as a correlate of protection should be ascertained with selected monoclonal antibodies in the
chimpanzee model.
Antibody-dependent enhancement (ADE) autoantibody assays
and
There was general accord that AIDS vaccinée sera should be screened for the presence of potentially adverse antibody responses. These tests can be performed as batched assays in labs with specialized expertise.
Complement-mediated assays Complement-mediated, antibody-dependent cytolysis and virolysis should also be carried out in specialized labs on selected vaccinée serum samples. ASSAYS OF CELL-MEDIATED
IMMUNITY (CMI) In general, the participants felt that it was extremely important also monitor parameters of cellular immune function in vaccine recipients. Although some animal studies have suggested that antibodies alone may confer protection against experimental challenge with HIV-1, HIV-2. or SIV. it was stressed that such studies have protected against intravenous challenge with cell-free virus and should not lead to the conclusion that cellular immunity would be unnecessary or ineffective as a host defense against naturally acquired infection. The following parameters of CMI were discussed. to
HIV-1-specific CD4+ cytotoxic
T
lymphocytes (CTL)
CD4+ CTL were detected when lymphocytes from recipients of the MicroGeneSys gpl60 subunit vaccine were stimulated in vitro with gpl60, and it has been possible to demonstrate
,
Lymphocyte proliferation/activation The consensus among the Working Group was that, although all HIV vaccines tested thus far have been shown to induce HIV antigen-specific, T-cell proliferation/activation, the relevance of such responses has not been established. It was felt that these studies should be continued as a general measure of T-cell memory. Several investigators underscored the caveat that the HLA genotype of the antigen presenting cell (APC) in vitro must be identical to that of the APC in vivo to insure functional readout in vitro. It was noted that a range of concentration of HIV antigen(s) should be employed routinely in lymphoprolif-
SUMMARY: CLINICAL IMMUNOLOGY WORKING GROUP eration assays. It was also stressed that results should be presented in a form that allows some means of comparison from study to study. In this regard, reporting raw numbers as well as stimulation indices was considered to be essential.
Natural killer (NK) cells
Although a number of studies have suggested a possible role for NK cells as an acute antiviral host defense, the consensus of the participants was that little information was likely to be gained by studying NK cell function in uninfected vaccine recipients. Thus, the limited blood samples available from these subjects should be conserved for studies of HI V-specific cellular immunity.
CD8
lymphocyte suppression of virus replication
Downloaded by East Carolina University from www.liebertpub.com at 01/11/19. For personal use only.
CD8+ lymphocytes that suppress the replication of HIV (or
SIV) in vitro have been reported by several laboratories, although the precise mechanism of action (cytolysis of infected cells or factor-mediated suppression of viral replication) has not been determined. It was agreed that more studies were needed to identify the soluble factor that has been postulated to mediate this suppression. Pilot studies to look for a similar response in selected vaccine recipients were suggested and would be expected to have the best chance of success if carried out in laboratories with specific expertise. It was stressed that samples would need to be made available to investigators outside the AVCT network for these studies.
ADCC ADCC was also considered from the standpoint of cellular immunity, particularly the antibody-armed effector cells identified in seropositive persons. It was proposed that further studies should be performed on a limited subset of subjects to determine under what conditions armed effectors are generated and detected.
ASSAY DEVELOPMENT FOR ASSESSMENT OF MUCOSAL IMMUNITY Current AIDS vaccine strategies are not expected to induce mucosal immunity, but it was considered timely and prudent to collect mucosal specimens; e.g., parotid saliva or seminal fluid from a selected subset of vaccinées. Adaptation or development of appropriate assays to detect mucosal immune responses to HIV antigens in vaccinées should be initiated forthwith.
1443
EVALUATION OF NONSPECIFIC MARKERS These tests will be of particular importance in postinfection immunization, but were not felt to be a high priority in seronegative vaccinées. The most prominent of the nonspecific markers is the enumeration of CD4 cells. AIDS vaccinées receiving envelope products that bind to CD4 will continue to be monitored for effects on CD4 T cells, although no adverse effects have been detected with vaccines in current trials. This parameter should be analyzed in terms of absolute counts of CD4 cells, the percentage of CD4 cells and the relative change in CD4 cell numbers from baseline in seropositive trials. But CD4 stabilization should not be the only criterion for vaccine effectiveness. The consensus of the group was that the usefulness of measuring beta-, microglobulin, neopterin, interferons. NKcell activity of delayed-type hypersensitivity to various common antigens, etc., is not established.
VIRAL MARKERS These tests were again felt to be of importance only in the evaluation of seropositive individuals undergoing therapeutic vaccination. Due to time constraints these tests could not be discussed in detail; however, it was noted that the AIDS Clinical Trials Group has developed or is developing standard protocols for quantitative viral cultures and the polymerase chain reaction for detection of HIV. In summary, the Clinical Immunology Working Group participants reached a consensus on the priorities for immunological assessment of individuals undergoing vaccination with prototype AIDS vaccines. The group felt there was a compelling need for more data defining the correlates of protective immunity to HIV. Without such information, it was deemed impossible to fully define what immune responses would need to be generated by an AIDS vaccine to achieve efficacy. Given these constraints, the group emphasized that vaccine strategies against HIV should be aimed at eliciting as many humoral and cellular virus-specific immune responses as possible. Furthermore, the adaptation and development of assays was considered essential to facilitate the measurement of specific parenteral and mucosal immune responses to future candidate AIDS vaccines. Address
reprint requests
to:
Susan Zolla-Pazner
New York
Department of Pathology University Medical Center New York. NY 10016
