Upload
others
View
0
Download
0
Embed Size (px)
Citation preview
See discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/20352626
Serologic confirmation of human T-lymphotropic virus type I infection in
healthy blood and plasma donors
Article in Blood · December 1989
DOI: 10.1182/blood.V74.7.2585.bloodjournal7472585 · Source: PubMed
CITATIONS
82READS
643
10 authors, including:
Some of the authors of this publication are also working on these related projects:
Structure function of discrete peptide domains View project
David W. Anderson
St. Luke's University Health Network
52 PUBLICATIONS 2,324 CITATIONS
SEE PROFILE
Carl Saxinger
National Cancer Institute (USA)
150 PUBLICATIONS 6,806 CITATIONS
SEE PROFILE
All content following this page was uploaded by Carl Saxinger on 22 May 2014.
The user has requested enhancement of the downloaded file.
1989 74: 2585-2591
BJ Poiesz and LT PierikDW Anderson, JS Epstein, TH Lee, MD Lairmore, C Saxinger, VS Kalyanaraman, D Slamon, W Parks, healthy blood and plasma donorsSerological confirmation of human T-lymphotropic virus type I infection in
http://bloodjournal.hematologylibrary.org/site/misc/rights.xhtml#repub_requestsInformation about reproducing this article in parts or in its entirety may be found online at:
http://bloodjournal.hematologylibrary.org/site/misc/rights.xhtml#reprintsInformation about ordering reprints may be found online at:
http://bloodjournal.hematologylibrary.org/site/subscriptions/index.xhtmlInformation about subscriptions and ASH membership may be found online at:
Copyright 2011 by The American Society of Hematology; all rights reserved.20036.the American Society of Hematology, 2021 L St, NW, Suite 900, Washington DC Blood (print ISSN 0006-4971, online ISSN 1528-0020), is published weekly by
For personal use only. by guest on July 10, 2011. bloodjournal.hematologylibrary.orgFrom
Blood, Vol 74. No 7 (November 15), 1989: pp 2585-259 1 2585
Serological Confirmation of Human T-Lymphotropic Virus Type I Infection inHealthy Blood and Plasma Donors
By David W. Anderson, Jay S. Epstein, Tun-Hou Lee, Michael D. Lairmore, Carl Saxinger, V.S. Kalyanaraman,
Dennis Slamon, Wade Parks, Bernard J. Poiesz, Lauren T. Pierik, Helen Lee, Richard Montagna,
Patrick A. Roche, Alan Williams, and William Blattner
We wished to develop criteria for serological confirmation
of human T-lymphotropic virus type I (HTLV-l) infection in
healthy donors. Selected serum or plasma samples reac-
tive by HTLV-l enzyme immunosorbent assay or gel-
agglutination assays with at least one viral-specific band on
Western immunoblot (WIB) were tested in six laboratories
by four WIBs and four radioimmunoprecipitation assays
(RIPAs) for antibodies to HTLV-l proteins encoded by gag
(p19 and p24). env (gp46 and/or gp6l). and tax (p41?)
genes. One hundred forty-two donor sera were obtained
from 38 Japanese. 69 American, and 35 Caribbean blood or
plasma donors. Among these samples. WIB assays
appeared more sensitive to p24 antibodies. whereas RIPAs
were significantly more sensitive to gp6l antibodies. All
sera (1 37) with gp6l antibodies had p24 antibodies. Of the
H UMAN T-CELL lymphotropic virus type I (HTLV-I),
the first known human retrovirus, is associated with an
aggressive lymphoma, adult T-cell leukemia/lymphoma
(ATL), and a degenerative demyelinating neurological dis-
ease called tropical spastic paraparesis (TSP) in the Carib-
bean and HTLV-I-associated myelopathy (HAM) in
Japan.”2 Development of AlL may require years to decades
after infection with HTLV-I.3’4 Preliminary data from Japa-
nese patients with HAM and a history of blood transfusion
suggest a mean interval of 4 years between transfusion and
development of neurologic symptoms (Osame M, unpub-
lished observations, November, 1988). HTLV-I endemic
areas have been identified in southern Japan, the Caribbean,
and parts of Africa.5 HTLV-lI is a closely related retrovirus
which shares considerable genomic homology with HTLV-I
and is believed to be indistinguishable from HTLV-I with
current serologic techniques. Little is known about the
epidemiology and clinical outcomes of HTLV-II infections
except for their apparent high frequency among drug abusers
and occasional isolation from patients with T-hairy cell
leukemia and some B-cell lymphoid malignancies. In the
absence of virus isolation or characterization of gene
sequences by polymerase chain reaction (PCR), the virus
type of seropositives in the United States is best classified as
HTLV-l/ll. Antibodies to HTLV-I/II were recently
detected in 10 (0.025%) of 39,898 random blood donors in
eight US cities,6 and HTLV-I/lI seroconversion in multiply
transfused US hospital patients has been documented.7
On November 29, 1988, enzyme immunoassay (EIA)
screening tests designed to detect antibodies to HTLV-l were
licensed by the US Food and Drug Administration (FDA).8
Implementation of HTLV-I EIA screening ofdonated whole
blood and cellular components is currently underway in
many blood establishments nationwide. An estimated I 3,600
US blood donors may test positive in the first year of
HTLV-I EIA screening. (The number of blood donors test-
ing positive [repeatably reactive] in the first year of HTLV-I
EIA screening was estimated as the product of 8 x 106 blood
donors times 0.17%6 repeatably reactive rate). Although
1 37 sera positive for p24 and gp6l antibodies. p19
antibodies were detected in 1 29 sera, and p41? antibodies
were detected in 1 08. In sera with p1 9 antibodies and
antibodies to env- or tax-encoded proteins, p24 antibodies
were always present. Antibodies to p4O� were not found in
the absence of gp6l antibodies. Virological evidence of
infection was found in seven American donors by lympho-
cyte coculture (one HTLV-l. one HTLV-ll) or by polymerase
chain reaction (three HTLV-l, two HTLV-ll). Sera from all
seven donors showed p24 and gp46 and/or gp6l antibod-
ies. We suggest that seroreactivity to both p24 and gp46
and/or gp6l by WIB or RIPA or both are suitable criteria to
confirm but not to distinguish HTLV-l and HTLV-ll infec-
tions.
S 1989 by Grune & Stratton, Inc.
blood from these donors will not be used for transfusion, less
than one of six is likely to be confirmed positive with
additional, more specific serological tests.6 Thus, criteria
used to confirm HTLV-I/II seropositivity will affect notifi-
cation and counseling strategies used for thousands of
healthy donors.
As with screening tests for human immunodeficiency
virus-I (HIV-l), a distantly related but not serologically
cross-reactive retrovirus, current approaches to confirmation
of l-ITLV-I/Il seropositivity rely on additional, more specific
tests such as Western immunoblot (WIB) and radioimmuno-
precipitation assay (RIPA).9”#{176} Both WIB and RIPA enable
separate demonstration of antibodies to HTLV-I proteins
encoded by viral genes. However, there are important differ-
ences between WIB and RIPA in terms of viral antigen
From the Center for Biologics Evaluation and Review, FDA.
Bethesda, MD; Department of Cancer Biology. Harvard School of
Public Health, Boston, MA; Laboratory of Tumor Cell Biology and
the Viral Epidemiology Section. National Cancer Institute. Bet hes-
da, MD; Retrovirus Diseases Branch, Center for Disease Control,
Atlanta, GA; Bionetics Research, Rockville, MD; Division of
Hematology/Oncology, UCLA School of Medicine, Los Angeles,
CA; Department of Pediatrics/Immunology. University of Miami.
Miami, FL; Division of Hematology/Oncology. Department of
Medicine, SUNY at Syracuse, Syracuse. NY; Abbott Laboratories,
North Chicago, IL; Cellular Products, Buffalo, NY; E. I. Dupont
De Nemours, Wilmington. DE; and American Red Cross Jerome H.
Holland Laboratory, Rockville. MD.
Submitted March 20, 1989; accepted July i I, /989.
Address reprint requests to Jay S. Epstein. MD. Chief, Labo-
ratory of Retrovirology, Division of Blood and Blood Products,
Center for Biologics Evaluation and Research, Food and Drug
Administration, 8800 Rockville Pike, Bethesda, MD 20892.
The publication costs ofthis article were defrayed in part by page
charge payment. This article must therefore be hereby marked
“advertisement” in accordance with 18 U.S.C. section 1 734 solely to
indicate thisfact.
� I 989 by Grune & Stratton, Inc.
0006-4971/89/7407-0007$3.00/0
For personal use only. by guest on July 10, 2011. bloodjournal.hematologylibrary.orgFrom
2586 ANDERSON ET AL
preparation and the physicochemical milieu of antigen-
antibody interaction. Viral antigen commonly used in WIB
formats is derived from density-gradient purified virions
rather than extracts of infected cells. Viral proteins in WIB
formats therefore consist predominantly of virus particle-
associated structural proteins (gag-encoded proteins p19,
p24, and their precursor p55 and env-encoded proteins
gp2lE, gp46, and their precursor gp6I/68). After boiling,
sodium dodecyl sulfate (SDS) treatment, reduction and
electrophoretic transfer, all of which may modify the anti-
gens, the antigen-antibody interaction occurs on a two-
dimensional nitrocellulose membrane. In contrast, viral anti-
gens commonly used in RIPA formats are extracts from
chronically-infected cells and include structural as well as
nonstructural (tax-encoded p4(Y) regulatory proteins at van-
ous stages of posttranslational modification. The antigen-
antibody interaction in a RIPA occurs in a solution (three-
dimensional) of whole cellular lysate under nondenatuning
conditions.Investigators have used various confirmatory criteria and
differing combinations of WIB and RIPA to identify asymp-
tomatic HTLV-I/II seropositive persons.9”#{176} Indeed, valida-
tion of serological confirmatory criteria in this setting is
complicated by the inability of current assays to distinguish
antibodies to HTLV-I from antibodies to HTLV-II, a lack of
common positive reference sera, and incomplete knowledge
ofvariability in the pattern of HTLV-I-specific antibodies or
HTLV-Il-specific antibodies among persons after infection.
An additional complication is insufficient data regarding the
comparative performance of WIB and RIPA in detecting
variations in the pattern of antibodies to specific HTL V-I/I!
antigens, particularly when different cell lines are used for
antigen production.
Using a variety of assays in six different laboratories, we
undertook to characterize the pattern of HTLV-I/II-
specific antibodies in healthy blood and plasma donors. Our
investigation represents a consensus approach to confirma-
tion of HTLV-l/II seropositivity and serves as the basis for
current Public Health Service and blood bank guidelines for
confirming positive results of antibody screening. The results
obtained provide a rationale for use of antibodies to certain
gag and env gene products in confirmatory criteria as well as
for assessing the suitability of WIB and RIPA for use in
confirmatory strategies.
MATERIALS AND METHODS
Positive and negative reference sera. The Center for BiologicsEvaluation and Research (CBER), US FDA distributed to each ofsix collaborating laboratories and one in-house laboratory (Table 1)
eight uncoded sera (positive reference sera) from US patientsclinically diagnosed as having ATL (three cases) or TSP (five cases)(B.J. Poiesz, unpublished observations, December, 1987). All caseshad evidence of HTLV-I-specific sequences in peripheral lympho-cyte DNA extracts amplified by PCR,” including six patients withHTLV-l-specific sequences found by Southern blotting of unampli-
fled lymphocyte DNA extracts. Each laboratory chose to studyantibodies to one or several of the gag-encoded core proteins p19,p24, and precursor p55, the env-encoded glycoproteins gp46 or gp6l,gp63, gp65, or gp68 (hereafter called gp6l) or the tax-encodedtranscriptional activator p4#{216}X(Table 1 ). These particular viralproteins were selected for study because, with the exception of gp68,
each represents the product of only one HTLV-I gene.22’23 Uncodedsera from 20 normal US blood donors (negative reference sera) were
also provided. Normal donors included 18 males, mean age 21 ± 4
years (SD) and two females (aged 20 and 34 years) who donatedblood in Maryland and were found nonreactive for HIV-1 antibodies(HIV-1 EIA, Abbott Laboratories, N. Chicago, IL), hepatitis B coreantibodies (Corzyme, Abbott) and hepatitis B surface antigen
(Auszyme, Abbott). For all nine positive reference sera, eachlaboratory found strong seroreactivity to all viral proteins listed for
its assays in Table 1 . No detectable seroreactivity was found by anylaboratory for any of the 20 negative reference sera. To investigatenonspecific interference, 37 sera representing a variety of serologicalconditions that might interfere with detection of HTLV-I/II-
specific antibodies were distributed under code. These sera wereobtained from Boston Biomedica, Medford, MA, and included twosera from multiply transfused hospital patients, one grossly lipemicserum, one serum with antiHLA DR antibodies, two sera with
antibodies to HIV, three sera reactive for rheumatoid factor, andsera with immunoreactivity to each of the following infectious agents
(three sera per agent): Toxoplasma gondii. cytomegalovirus
(CMV), rubella, Herpes simplex, and Epstein-Barr virus (EBV)
early antigen. In addition, three sera were positive for hepatitis Bsurface antigen, one serum was positive for hepatitis B core antibod-
ies, three sera were reactive by the heterophile test, three sera were
reactive with antinuclear antibodies, and three sera were reactive bythe rapid plasma reagin test. No seroreactivity was found by any
laboratory (laboratory 3 performed WIB but not RIPA on thesesamples) in any of the 37 samples. Thus, the performance of each
assay was similar among seven laboratories, and common positiveand negative reference sera were available as standards for assessingreactivities of unknown test sera.
Donor samples. Serum or plasma samples from 153 HTLV-Iseroreactive blood and plasma donors from Japan, the United States,and the Caribbean basin were obtained by CBER from bloodestablishments and manufacturers (Table 2). Samples were
accepted for study if they met the following criteria: (a) volume aSmL, (b) repeatably reactive by HTLV-l EIA or gel-agglutinationscreening tests, (c) seroreactive to at least one HTLV-I viral protein
on WIB or reactive on HTLV-I indirect immunofluorescent assay
(I FA), and (d) remained frozen at less than - 20#{176}Csince blooddrawing. Of I 53 samples received in an 8-month period, I 42 (donorsamples) met the above criteria for study and represented similarly
significant numbers ofJapanese (38), American (69) and Caribbean(35) donors. Of 142 donor samples, 37 were serum and the remain-
der were plasma. Identical aliquots of each donor sample weresimultaneously distributed under code to seven collaborating labora-
tories for investigation by serological assay methods as listed inTable I. All donor samples were found nonreactive by HIV-l EIA(Organon-Teknika, Durham, NC) except two plasma samples from
Martinique, both of which showed no HIV-specific bands by HIV-lWIB (Biotech/DuPont HIV Western Blot Kit, DuPont, Wilming-
ton, DE).Statistical methods. Associations between discrete variables
were evaluated for significance by Fisher’s exact test (corrected for atwo-tailed distribution). Differences in the performance of WIB andRIPA for detection of p24 and gp6l antibodies were evaluated byMcNemar’s paired chi-square test.
RESULTS
Comparison of WIB and RIPA for detection of p24
antibodies. Seroreactivity to the viral core p24 antigen was
assessed by four WIB and four RIPAs performed in different
laboratories (Table 3). If two or more WIB assays were
positive, the sample was designated consensus positive (ie
For personal use only. by guest on July 10, 2011. bloodjournal.hematologylibrary.orgFrom
CONFIRMATION OF HTLV-l/Il SEROPOSITIVITY 2587
Table 1 . Serologic Methods of Collabor ating Laboratories
Method SeroreactivitiesLaboratory Assay Type Antigen Source (Reference No.) Evaluatedt
1 . Retrovirus Diseases Branch WIB MT-2 cells 12 gag-p 1 9, p24, p55
Centers for Disease Con- env-gp46 and gp6 1/68trol RIPA MT-2 cells gag-p24, p55
Atlanta, GA env-gp6 1/68
2. Laboratory of Tumor Cell Bi- Competitive EIA HUT 102 cells 13, 14 Positive/indeterminate/negative
ology WIB HUT 102 cells 15 gag-p19, p24, p55National Cancer Institute env-gp46
Bethesda, MD
3. Department of Cancer Biol- WIB MJ cells 16 gag-p 19,p24,p55ogy RIPA MT-2 cells 1 7 env-gp63/65
Harvard School of Public env-gp6 1/68
Health
Boston, MA
4. Bionetics Research, Inc. WIB HUT 102 cells 18 gag-p19 and p24
Rockville, MD p24 RIPA HUT 102 cells gag-p24
5. Division of Hematology! RIPA SLB-1 cells 6 gag-p24
Oncology env- gp6 1
UCLA School of Medicine tax-p40’
Los Angeles, CA
6. Department of Pediatrics! p24 RIPA HUT 102 cells 19, 20 gag-p24
Immunology
University of Miami
Miami, FL
7 . Laboratory of Retrovirology IFA MT-2 cells 2 1 Positive, indeterminate, negative
CBER, FDA
Bethesda, MD
Laboratories 4 and 6 performed this assay by measuring counts from immunoprecipitable ‘25l-p24 in a y-counter, whereas laboratories 1 , 2, and 5subjected precipitable proteins to gel electrophoresis and autoradiography.
tEach laboratory determined which seroreactivities to report based on previous experience or validation experiments with monoclonal antibodies orboth.
�:50% concordance) by WIB. Likewise, if two or more
RIPAs showed evidence of p24 antibodies, the sample was
consensus positive (ie, a50% concordance) by RIPA. Of 142
donor samples tested, I 39 were consensus positive for p24
antibodies by WIB and included all 1 34 that were consensus
positive by RIPA. In terms of its performance as a group of
assays, WIB appeared to be more sensitive for p24 antibodies
than RIPA, but this difference did not reach statistical
significance (P = 0.074). We considered all 139 donor sam-
ples consensus positive by WIB to have antibodies to p24.
Table 2. Donor Sera Characteristics
Donation Donor n/5ex/Mean Age Initial 5creeningNationality Site Type N hiT) ± SD Test (Manufacturer)
Japanese Japan B 29 21/M/38 ± 1 1
1/F/44 -
HTLV-l gel-agglutination
(Fujirebio�)Japanese Okinawa B 9 5/M/48 ± 1 1
4/F/46 ± 13
HTLV-l EIA (DuPont)
Japanese-American Hawaii B 34 28/M/49 ± 1 1
6/F/50 ± 14
HTLV-l EIA (DuPont)
American United States B 22 14/M/40 ± 15
8/F/40 ± 8HTLV-l EIA (Abbott, 4;
DuPont, 15; CPI. 3)American United States P 13 9/M/39 ± 10
4/F/36 ± 4
HTLV-l EIA (Abbott, 13)
Jamaican Jamaica B 1 1 8/M/3 1 ± 43/F/45 ± 12
HTLV-l EIA (DuPont)
Martinique Martinique B 24 1 2/M/49 ± 8
12/F/53 ± 7
HTLV-l EIA (Abbott)
Abbreviations: B, blood donor; P. plasma donor.
#{149}Ageand sex data could not be obtained for seven Japanese blood donors.
tFuiirebio is located in Tokyo, Japan.
For personal use only. by guest on July 10, 2011. bloodjournal.hematologylibrary.orgFrom
Seroreactivity to p 19 and p24 was determined by consensus among
four independent WIBS. Seroreactivity to gp 6 1 was determined by
consensus among three independent RIPAS.
2588 ANDERSON ET AL
Table 3. Numbers of Donor Samples With Seroreactivity to p24
by WIB and RIPA
No. Positive Testsfor p24 Antibodies RIPA5 a 2 RIPAS < 2
WlBassays�2 134 5
WIB assays <2 0 3
Each sample was tested by four WIBS and four RIPAS. The number of
donor samples is given for each of four groups of assay results. Each
group is arbitrarily defined according to whether a majority (two or more)
of WIBS or RIPAS were positive for p24 antibodies. P = .074 for
detection of p24 antibodies by WIB in more donor samples (ie, 139
samples) than by RIPA (ie. 134 samples).
Comparison of WIB and RIPA for detection of env-
encoded proteins gp46 and/or gp6I. Seroreactivity to env-
encoded proteins was assessed in different laboratories by
three WIB assays for envelope gp46 and/or precursor gp6l
and by three RIPAs for gp6l (Table 4). Iftwo or more WIB
assays were positive (ie, greater than 50% concordance), the
sample was designated consensus positive by WIB. Likewise,
if two or more RIPAs showed evidence of gp6l antibodies
(ie, greater than 50% concordance), the sample was consen-
sus positive by RIPA. Of 142 donor samples tested, 137 were
consensus positive for gp6I antibodies by RIPA and included
all I 25 samples that were consensus positive by WIB for
either gp46 or gp6l. As a group, RIPAs detected antibodies
to env-encoded proteins in significantly more donor samples
than did WIB (P= .0015). We considered all 137 donor
samples consensus positive by RIPA to have antibodies to
gp6 I.
Correlation ofp24 and gp6l antibodies. It is important
that the 137 donor samples consensus positive by RIPA for
gp6l antibodies were all consensus positive by WIB for p24
antibodies. These I 37 samples were also repeatably reactive
when blindly tested by each of the FDA-licensed l-ITLV-I
EIA tests (Abbott, Dupont, and Cellular Products, Buffalo,
NY). When tested by HTLV-l IFA (Table I), 131 ofthe 137
samples were positive and six were indeterminate owing to
weak reactivity.
Five samples, all from US donors, were not seroreactive by
consensus to gp6I (Table 5) and showed predominantly p19
seroreactivity. Antibodies to p19 and p24 were compared on
four WIB in different laboratories (Table 6). One hundred
thirty-one donor samples were seroreactive by consensus (ie,
two or more WIB assays positive for each p19 and p2’!) to
both p1 9 and p24; I 29 of these were also seroreactive by
consensus to gp6 I . Among the remaining I 1 donor samples,
Table 4. Numbers of Donor Samples With Seroreactivity
togp6l by WIB and RIPA
No. Positive Tests for gp46and/or gp6 1 Antibodies RIPAS a 2 RIPAS < 2
WlBassays�2 125 0
WlBassays<2 12 5
Each sample was tested by three WIBS and three RIPAS. The number
of donor samples is given for each of four groups of assay results. Each
group is arbitrarily defined according to whether a majority (two or more)
of WIBS or RIPAS were positive for gp46 and/or gp6 1 antibodies. P
.00 1 5 for detection of gp6 1 antibodies by RIPA in more donor samples
(ie, 1 37 samples) than by WIB (ie 1 25 samples).
Table 5. HTLV-l/lI-Speciflc Antibodies Detected by WIB and
RIPA in Blood Donor Samples Lacking Consensus
Positivity for Anti-GP61
Antibodies Detected Antibodiesby WIB Detected by RIPA
Viral proteins p 1 9 p24 gp46 gp6 1 p24 p40’ gp6 1
Total number of assays used 4 4 2 2 4 1 3
Sample
1 3 1 0 0 0 0 0
2 3 1 0 0 1 0 0
3 3 1 0 0 1 0 0
4 2 2 0 0 1 0 0
5 4 2 1 0 1 0 0
Number of assays positive.
eight were seroreactive by consensus to p24 and not to p19;
there was a significant association (P = .007) between the
presence of antibodies to p24 and antibodies to gp6I among
these I I samples. In contrast, none of three donor samples
with p19 antibodies in the absence of p24 antibodies was
seroreactive to env-encoded proteins. Antibodies to the tax-
encoded protein p40’ were detected by RIPA in 108 donor
samples, of which 108 were seroreactive to both p24 and
gp6 I.
Virological studies of peripheral blood lympho-
cytes. Follow-up studies of peripheral blood lymphocytes
(PBLs) were performed for seven US blood donors who were
positive for HTLV-I/II antibodies by EIA screening tests
and who were available for further studies (Table 7). Anti-
bodies to gag and env were identified by consensus in all
seven cases. Lymphocyte cocultures in two donors were
positive by endonuclease restriction mapping for HTLV-I in
one case and human T-lymphotropic virus type II (HTLV-
II) in one case. Lymphocyte DNA extracts from five other
donors were positive for HTLV-l-specific sequences but
negative for HTLV-II-specific sequences by PCR in three
cases and positive for HTLV-II-specific sequences but nega-
tive for HTLV-I-specific sequences by PCR in two cases.
The HTLV-I/lI-specific antibody pattern of the four
HTLV-I infected donors was not distinguishable from that of
the three HTLV-II infected donors by WIB or RIPA.
DISCUSSION
Serological diagnosis of asymptomatic HTLV-I/II infec-
tion among persons living in the United States presents
several challenges. The epidemiology of HTLV-I and
HTLV-ll in the general US population has not been evalu-
ated systematically. Retroviral infection involves proviral
Table 6. Numb
C
ers of Donor Samples With Seroreactivity by
onsensus to p1 9. p24. and gp6l
Seroreactive
toSeroreactive Not Seroreactivetogp6l togp6l
p19+p24
p24 only
pl9only
129 2
8 0
0 3
For personal use only. by guest on July 10, 2011. bloodjournal.hematologylibrary.orgFrom
CONFIRMATION OF HTLV-I/II SEROPOSITIVITY 2589
Table 7. HTLV-I/Il-Specific Antibodies Detected by WIB and RIPA in Blood Donor Samples
With Virological Evidence of Retroviral Infection
Antibodies DetectedbyWIB
Antibodies Detected
byRIPA
Viral proteins p 1 9 p24 gp46 gp6 1 p24 p40’ gp6 1
Total number of assays used 4 4 2 2 4 1 3Virus identified (technique used)
HTLV-l (Lymphocyte coculturet) 4
HTLV-I (PCR*) 4
4 1
4 2
2
2
2
4
0
1
3
3
HTLV-l (PCR*) 1 4 1 1 3 1 2
HTLV-l (PCR�) 1
HTLV-ll (Lymphocyte coculturet) 1
4 0
4 20
13
41
12
3HTLV-II (PCR#) 1 4 0 0 2 1 3
HTLV-ll (PCR#) 4 4 2 2 4 1 3
Number of assays positive.
tCocultivated lymphocytes were extracted, and whole DNA was digested with the following restriction endonucleases: Xbal , EcoR 1 . and BamHI.Southern blots of digested DNA were probed with an AccI to HincIl fragment from the 3’ end of the HTLV-l genome (bases 7346-8534) and a BamHI toBamHI fragment from the 5’ end of the HTLV-Il genorne (bases 36 1 -5090) (Slamon D, unpublished observations. February, 1989).
*Whole DNA extracts from PBLs were subjected to PCR with the following primer pairs: SK43, 44, a tax gene sequence common to HTLV-I andHTLV-ll; SK54, 55. a pot gene sequence specific for HTLV-l; and SK58, 59, a polgene sequence specific for HTLV-lI.’#{176}
#Foung S, personal communication, January, 1989.
integration; therefore, diagnostic virological methods such as
lymphocyte coculture or PCR should allow direct identifica-
tion of viral sequences in cellular DNA of a person infected
many years before. However, these methods are not yet
available for widespread routine use. Thus, serological tests
will be used both to screen for and confirm most cases of
HTLV-l/ll infection in the immediate future. To examine
serological confirmatory strategies we studied a cross-section
of serum and plasma samples from two foreign HTLV-I
endemic areas and the United States that were repeatably
reactive on HTLV-l screening tests and had evidence of
viral-specific antibodies on WIB or IFA. The presence of
antibodies to specific viral proteins was defined by a consen-
sus ofseroreactivities in both WIB and RIPA formats.
Our data confirm the observations of other investigators6’9
that the presence of antibodies to p24 and env-encoded
proteins are highly correlated in sera from healthy donors.
Antibodies to p19 usually accompanied p24 and gp6l anti-
bodies. Antibodies to gp46 and/or gp6I were detected in 9%
fewer donor samples by WIB assays than by RIPAs
(P = .001 5). The capability of RIPAs to detect antibodies to
env-encoded proteins not apparent on WIB has been
observed by other investigators6’9 for small numbers of sam-
pIes; the physicochemical basis for this remains unclear.
Optimal antibody binding to major epitopes of gp6l may
require the conformational presentation of antigens as in the
milieu ofsolubilized cellular lysate used in a RIPA. Alterna-
tively, ultracentrifugal purification and other preparative
processes commonly used in WIB assays may adversely
affect gp6l epitope presentation. In contrast, WIB appears
to be the more sensitive format for p24 antibody detection,
although this trend did not reach statistical significance
(P = .074).
Several serological confirmatory criteria for HTLV-l were
proposed recently. Agius et al’#{176}required antibodies to both
p19 and p24 by WIB whereas Fang et al9 required antibodies
to two or more gene products by a combination of WIB and
RIPA. We showed that seroreactivity to gp6I could be
detected by RIPA in 129 of 131 (98.5%) of donor samples
with p19 and p24 antibodies by WIB. In this regard, the two
sets of confirmatory criteria are similar. However, as shown
in Table 6, two donor samples without gp6I antibodies
showed p19 and p24 antibodies by consensus and in three
other samples p24 antibodies were detected by one of four
WIB assays. Confirmatory criteria that require antibodies to
at least two independent gene products would not be satisfied
for these five samples and are thus more conservative than a
requirement ofantibodies only to p19 and p24.
We have assumed that HTLV-I/Il-specific antibodies to
two independent gene products are prudent confirmatory
criteria. On the basis of the present data, we suggest that
antibodies to p24 and to the env-encoded proteins gp46
and/or gp6l be demonstrated by one or a combination of
serologic tests. This strategy would maintain a requirement
for antibodies to two independent gene products while
excluding the presence of antibodies to p1 9 or p40’ as criteria
to be used in confirmation; ie, we found no samples with p19
antibodies but without p24 antibodies that contained anti-
bodies to gp6I at routine sample dilutions. Furthermore,
antibodies to p405 were present only in a subset (79%) of
samples with p24 and gp6I antibodies. Conversely, we have
no evidence that persons lacking p24 and gp46 and/or gp6l
antibodies are infected with HTLV-I or HTLV-II. All seven
samples for which the donor’s lymphocytes yielded virologi-
cal evidence of HTLV-I or HTLV-II infection (Table 7)
showed seroreactivity to p24 and gp#{243}l.It is of interest that
p19 antibodies were detected by two or more WIBs in only
three of these seven samples whereas antibodies to p40’ were
found in six of seven samples. In the absence of more
virologic data, samples which show antibodies to at least one
suspected HTLV-I gene product (such as p19 only, p19 and
p28, p40’ only) should be regarded as indeterminate to leave
open the question of infection in these cases.
We recommend examination of samples by WIB for p24
and gp46 and/or gp6l antibodies. If antibodies to p24 are
present but antibodies to gp46 and/or gp6I are lacking, a
For personal use only. by guest on July 10, 2011. bloodjournal.hematologylibrary.orgFrom
REFERENCES
2590 ANDERSON ET AL
RIPA for gp6I antibodies is recommended. Application of
this algorithm to our samples with p24 antibodies by WIB
would result in the need for a RIPA to evaluate further 7 of
32 (22%) US donor samples and 6 of 34 (18%) Japanese-
American (Hawaii) donor samples for gp6l antibodies. In
contrast, only I of 29 (3%) Japanese, 0 of I I Jamaican, and 0
of 24 Martinique donor samples with p24 antibodies by WIB
would have required further evaluation by RIPA. These
differences among the populations may relate to higher
measurable antibody titers among persons living in areas of
greater prevalence of HTLV-I infection, as in the general
populations of Japan and Caribbean countries.8 The lower
concordance rate in US samples may reflect differences in
virus type, HTLV-l versus HTLV-II, or differences in
expression of virus in vivo resulting in titer differences to
various antigens, or both. Further studies carefully charac-
terizing the nature of the virus by PCR and culture as well as
epitope mapping of various reactivities should help clarify
these alternatives.
The confirmatory criteria we developed have obvious
limitations. First, as demonstrated by the HTLV-I/II-
specific antibody patterns from US blood donors (Table 7),
infection with HTLV-I can not be distinguished from
HTLV-ll infection. Second, virological evidence of infection
was sought for only 7 of the 142 donor samples studied. Our
lack of knowledge of the demographic distribution of virolog-
ically proven HTLV-I and HTLV-II infection in the United
I. Poiesz BJ, Ruscetti FW, Gazdar AF, Bunn PA, Minna JD,
Gallo RC: Detection and isolation of type-C retrovirus particlesfrom fresh and cultured lymphocytes of a patient with cutaneous
T-cell lymphoma. Proc NatI Acad Sci USA 77:7415, 1980
2. Gessain A, Barin F, Vernant JC, Gout 0, Maurs L, CalenderA, de The G: Antibodies to human T-lymphotropic virus type-I inpatients with tropical spastic paraparesis. Lancet 2:407, 1985
3. Tajima K, Kuroishi T. Estimation of rate of incidence of AlL
among ATLV (HTLV-l) carriers in Kyushu, Japan. Jpn J ClinOncol 15:423, 1985
4. Kondo I, Nonaka H, Miyamoto N, Yoshida R, Matsue Y,
Ohguchi Y, lnouye H, Komada H, Hinuma Y, Hanouka M:
Incidence of adult I-cell leukemia-lymphoma and its familial clus-tering. Int J Cancer 35:749, 1985
5. Blattner WA: Retroviruses, in Evans AS (ed): Viral Infectionsof Humans: Epidemiology and Control, (ed 3) New York, Plenum,
I989
6. Williams AE, Fang CT, Slamon DJ, Poiesz BJ, Sander SG,
Darr WF, Schulman G, McGowan El, Douglas DK, Bowman Ri:Seroprevalence and epidemiological correlates of HTLV-l infectionin U.S. blood donors. Science 240:643, 1988
7. Minamoto GY, Gold JWM, Scheinberg DA, Hardy WD,
Chein N, Zuckerman E, Reich L, Dietz K, Gee T, Hoffer J:Infection with human T-cell leukemia virus type I in patients with
leukemia. N EngI J Med 318:219, 19888. Centers for Disease Control: Licensure of screening tests for
antibody to human T-lymphotropic virus type 1. Morbid Mortal37:736, 1988
9. Fang CT, Williams AE, SandIer 5G. Slamon DJ, Poiesz BJ:Detection of antibodies to human T-lymphotropic virus type I(HTLV-l). Transfusion 28:179, 1988
10. Agius G, Biggar Ri, Alexander SS, Waters DJ, DrummondJE, Murphy EL, Weiss SH, Levine PH, Blattner WA: Human I
States remains a significant obstacle to correct interpretation
ofantibody patterns ofall serological tests (eg, seroreactivity
to p1 9 and other gag-encoded proteins might be found to
reflect cross-reacting antibodies to other, yet uncharacter-
ized human retroviruses). Moreover, cross-reaction of p19
antibodies with thymic epithelial antigens might account for
some false-positive EIA and WIB results.24
Comparison of seroreactivity patterns in donors with viro-
logically defined infections, both longitudinally after sero-
conversion and cross-sectionally in larger numbers of donors
will be necessary to assess any set of serological confirmatory
criteria thoroughly. Options for use of antibodies to p19,
p405’ or other env-encoded proteins such as gp2 I E may well
be justified by such studies. The uncertain seroreactivity of
po/-encoded proteins also deserves greater study. In the
interim, given the low estimates of HTLV-I/II infection
among US blood donors (0.025%), the low attack rates of
associated diseases and the possibility of serological follow-
up of persons with some but not all required antibodies, a
conservative approach to serological confirmation of HTLV-
I/Il infection seems warranted. Sera from thousands of US
blood donors are expected to be found repeatably reactive by
HTLV-l screening assays over the next few years. An
understanding of serological confirmatory criteria and their
limitations will be crucial for the health care professionals
who counsel these persons.
lymphotropic virus type I antibody patterns: Evidence of differenceby age and risk group. J Infect Dis 158:1235, 1988
I I . Kwok 5, Ehrlich G, Poiesz B, Kalish R, Sninsky ii: Enzy-
matic amplification of HTLV-I viral sequences from peripheralblood mononuclear cells and infected tissues. Blood 72:1 1 17, 1988
12. Lairmore MD, Jason JM, Hartley TM, Khabbaz RF, Dc B,
Evatt BL: Absence of HTLV-I Coinfection in HIV-infected hemo-philic men. Blood (in press)
I 3. Saxinger WC, Gallo RC: Application of the indirect ELISAmicrotest to the detection and surveillance of human I-cell leuke-
mia-lymphoma virus (HTLV). Lab Invest 49:371, 1983
14. Saxinger W, Blattner WA, Levine PH, Clark J, Biggar R,
Hoh M, Moghassi J, Jacobs P. Wilson L, Jacobson R: Human I-cellleukemia virus (HTLV-I) antibodies in Africa. Science 225:1473,I984
15. Saxinger WC, Levine PH, Dean AG, de The G, Lange-
Wantzin G, Moghassi J, Laurent F, Hoh M, Sarngadharan MG,
Gallo RC: Evidence for exposure to HTLV-III in Uganda prior to
1973. Science 227:1036, 1985
16. Tsang VC, Peralta JM, Simons AR: Enzyme-linked immu-
noelectrotransfer blot techniques for studying the specificities ofantigens and antibodies separated by gel electrophoresis. Methods
Enzymol 92:377, 1983
17. Lee TH, Coligan JE, Homma I, McLane MF, Tachibana N,
Essex M: Human T-cell leukemia virus-associated membrane anti-gens (HTLV-MA): Identity of the major antigens recognized fol-lowing virus infection. Proc NatI Acad Sci USA 8 1 :3856, 1984
18. Schupbach J, Kalyanaraman VS. Sarngadharan MG,Blattner WB, Gallo RC: Antibodies against three purified proteinsof the human type C retrovirus, human I-cell leukemia-lymphomavirus, in adult I-cell leukemia-lymphoma patients and healthy
blacks from the Caribbean. Cancer Res 43:886, 1983
For personal use only. by guest on July 10, 2011. bloodjournal.hematologylibrary.orgFrom
CONFIRMATION OF HTLV-I/II SEROPOSITIVITY 2591
19. Kanner SB, Cheng-Mayer C, Geffin RB, Parks WB, BeltzCA, Arthur LO, Samuel KP, Papas IS: Human retroviral env and
gag polypeptides: Serologic assays to measure infection. J Immunol137:674, 1986
20. Kanner SB, Parks ES, Scott GB, Parks WB: Simultaneousinfection with HTLV-I and HIV. J Infect Dis 155:617, 1987
21. Yamamoto N, Hinuma Y. Antigens in an adult T-cell
leukemia virus-producer cell line: Reactivity with human serum
antibodies. Int J Cancer 30:289, 1982
22. Hattori 5, Kiyokawa T, Imagawa K, Shimizu F, Hashimura
E, Seiki M, Yoshida M: Identification of gag and env gene productsofhuman I-cell leukemia virus (HTLV). Virology 136:338, 1984
23. Slamon Di, Shimotohno K, Cline Mi, Golde DW, Chen IS:Identification of the putative transforming protein of the humanI-cell leukemia viruses HTLV-I and HTLV-II. Science 226:61,I984
24. Haynes BF, Robert-Guroff M, Metzgar RS, Franchini G,Kalynaraman VS, Palker Ti, Gallo RC: Monoclonal antibody
against human T-cell leukemia virus p19 defines a human thymicepithelial antigen acquired during ontogeny. J Exp Med I 57:907,I983
For personal use only. by guest on July 10, 2011. bloodjournal.hematologylibrary.orgFrom
View publication statsView publication stats