UvA-DARE is a service provided by the library of the University of Amsterdam (https://dare.uva.nl)
UvA-DARE (Digital Academic Repository)
Human herpesvirus 8 and Kaposi's sarcoma in the Amsterdam cohort studies.
Disease association, transmission and natural history
Renwick, N.M.
Publication date
2001
Link to publication
Citation for published version (APA):
Renwick, N. M. (2001). Human herpesvirus 8 and Kaposi's sarcoma in the Amsterdam cohort
studies. Disease association, transmission and natural history.
General rights
It is not permitted to download or to forward/distribute the text or part of it without the consent of the author(s)
and/or copyright holder(s), other than for strictly personal, individual use, unless the work is under an open
content license (like Creative Commons).
Disclaimer/Complaints regulations
If you believe that digital publication of certain material infringes any of your rights or (privacy) interests, please
let the Library know, stating your reasons. In case of a legitimate complaint, the Library will make the material
inaccessible and/or remove it from the website. Please Ask the Library: https://uba.uva.nl/en/contact, or a letter
to: Library of the University of Amsterdam, Secretariat, Singel 425, 1012 WP Amsterdam, The Netherlands. You
will be contacted as soon as possible.
Humann herpesvirus 8 infections in the
ÖÖ Amsterdam Cohort Studies (1984 -1997):
analysiss of seroconversions to ORF65 and ORF73
JaapJaap Goudsmit, Neil Renwick, Nicole H. T. M. Dukers, Koel A. Coutinho, Stem Heisierkamp,
MargreetMargreet Rakker, Thomas F. Schulp, Marion Cornelissen, and Gerritj. Weverling
ProcProc Natl Acad Sci USA 2000; 97: 4838-43
Abstract t
Wee have shown previously that human herpesvirus 8 (HHV8) seroconversion for antibodies to the latency-associated nuclear antigenn encoded by ORF73 a n d / o r the lytic capsid antigen (vpl9) encoded by ORF65 is associated with urogenital contact andd is strongly linked to the development of Kaposi's sarcoma among HIV-infected individuals in the Amsterdam Cohort Studies.. Here, we investigate the relationship between seroconversion to these antigens and primary- HHV8 infection. Be-tweenn 1984 and 1997, 215 H H V 8 seroconversions to ORF73 (106 cases or 49%) and/or to ORF65 (159 cases or 74%) were recordedd in the cohort of homosexual men. The HHV8 seroconversion rate among 1 IIY-infected homosexual men (6.2 per 1000 person years) was consistently higher than among HIV-uninfected men (2.6 per 100 person years). In HIV-infected but nott in uninfected individuals, seroconversion to ()RF73/latency-associated nuclear antigen precedes that to O R F 6 5 / v p l 9 . Antibodyy levels to both ORF65- and ORF73-encoded antigens were higher in HIV-infected than in HIV-uninfected men, andd among HI V-seropositives, antibody levels to O R F 6 5 / v p l 9 rise even higher with declining CD4 cell counts and peak with Kaposi'ss sarcoma development, suggesting continuing and increasing viral replication. In 10.3% of HHV8 seroconversions, transientt serum viremia could be demonstrated before or at seroconversion. Together with the previously reported link be-tweenn unprotected orogenital sex and HHV8 seroconversion, our observations suggest that HHV8 seroconversions result fromm primary infections.
Introduction n
T h ee h u m a n h e r p e s v i r u s 8 ( H H V 8 ) o r K a p o s i ' s sar-c o m a - a s s o sar-c i a t e dd h e r p e s v i r u s (KSHV) belongs to the g a m m a 22 or r h a d i n o v i r u s sublineage of t h e G a m m a h e r -p e s v i r i n a ee subfamily t o g e t h e r with the O l d W o r l d m o n k e y viruses,, r h e s u s m o n k e y r h a d i n o v i r u s , and r e t r o p e r i t o n e a l f i b r o m a t o s i s - a s s o c i a t e dd h e r p e s v i r u s e s ( R F H V ) ; t h e N e w W o r l dd m o n k e y viruses, h e r p e s v i r u s saimiri (HVS), and h e r p e s v i r u ss ateles ( H V A ) ; equine herpesvirus type 2 ( E H V 2 ) ;; a n d m u r i n e h e r p e s v i r u s 6 8 ( M H V 6 8 ; refs. 1-6). H H V 88 is strongly associated with K a p o s i ' s s a r c o m a (KS) in H I V - i n f e c t e dd individuals, b o d y cavity-based l y m p h o m a s , a n dd C a s t l e m a n ' s disease (7-10). T h e only o t h e r h u m a n g a m m a h e r p e s v i r u s ,, E p s t e i n - B a r r virus, is associated with l y m p h o m a ss a n d n a s o p h a r y n g e a l carcinoma (11).
T e s t ss for a n t i b o d i e s t o b o t h lytic a n d latent H H V 8 antigens cann identify n o t only m o s t H I V - i n f e c t e d individuals diag-n o s e dd w i t h K S b u t also t h o s e at idiag-ncreased risk to d e v e l o p K SS (12-18). Recently, w e s h o w e d t h a t s e r o c o n v e r s i o n to a r e c o m b i n a n tt H H V 8 lytic-phase capsid antigen, v p l 9 , en-c o d e dd by O R F 6 5 , a n d / o r t h e latent-phase nuen-clear antigen
(LANA)) e n c o d e d by O R F 7 3 , is highly predictive of K S (19).. A m o n g HIV-infected p e r s o n s , t h o s e w h o seroconvert forr H H V 8 after H I V infection are at higher risk to develop KSS than those w h o s e r o c o n v e r t for H H V 8 before H I V in-fection.. T i m e - d e p e n d e n t adjustment for C D 4+ cell c o u n t andd H I V - 1 R N A copy n u m b e r have n o impact on this addi-tionall risk, although the C D 4+ cell c o u n t was an i n d e p e n d
-entt risk factor for K S (19).
T h ee current study was designed to investigate the persis-tencee o f antibody r e s p o n s e s t o the lytic-phase capsid (ORF65)) and latent-phase nuclear (ORF73) antigens and t o assesss w h e t h e r s e r o c o n v e r s i o n follows a b u r s t in H H V 8 productionn and is associated w i t h clearance o f serum vire-mia.. In addition, w e studied the i m p a c t o f H I V and KS o n thee antibody response to O R F 6 5 / v p l 9 and O R F 7 3 / L A N AA to identify virus reactivation. Subsequently, we in-vestigatedd the association b e t w e e n H H V 8 seroconversion amongg H I V - s e r o p o s i t i v e a n d H I V - s e r o n e g a t i v e individuals andd the practice o f particular sexual behaviors over the coursee of the H H V 8 epidemic.
Materialss and Methods
Studyy Participants, Clinical
Follow-Up,, and Study Design.
Subjectss for the present study enrolled in the Amsterdam Cohortt Studies: 1,458 homosexual men and 1,167 injecting drugg users as described by Renwick et al. (19). T o determine whetherr participants were HHV8 seronegative or seroposi-tive,, their most recently obtained serum sample was tested byy an enzyme immunoassay (EIA) involving recombinant HHV88 proteins (see below). If a sample tested negative, the individuall was considered to have had no antibodies against HHV88 throughout his or her participation. If a sample testedd positive, the sample taken at enrollment of the cohort studyy was tested to determine whether seroconversion had occurredd during follow-up. If so, the year of seroconversion wass determined by testing serum samples at yearly intervals and,, within the year of seroconversion, at intervals of 3-6 months.. The midpoint between the last negative sample andd the first positive sample (seroconversion sample) was consideredd the date of HHV8 seroconversion. However, to investigatee the potential for false negativity, the enrollment sampless of 200 participants whose most recent sample had testedd negative were evaluated with the EIA system. A posi-tivetive result at entry was found for 9 of the 200, yielding a pu-tativee false negativity rate of 4.5% [95% confidence interval (CI):: 2.1-8.4].
Detectionn of HHV8 Antibodies.
Wee used an EIA format as described earlier (13, 19) by uti-lizingg either recombinant ORF65/vpl9, associated with the lyticc stage of HHV8 infection (13), or a carboxyl-terminal fragmentt of the LANA that is encoded by ORF73 (20). In thee case of HHV8, we deal with imperfect reference
stan-dardss because HHV8 cannot be cultured currendy and HHV88 D N A is found only in 58-67% of peripheral blood mononuclearr cells and 46% of serum from KS patients (21). Forr our analysis, results from the EIA system with an opti-call density of 0.350 or more for ORF65/vpl9 or 0.375 or moree for O R F 7 3 / LANA were considered HHV8-sero-positive.. These cut-off values were three times the SD of thee mean optical density from a panel of 40 HIV-uninfected IVV drug-using women representing low risk for KS. In addi-tion,tion, to evaluate the cut-off points in our population of in-terest,, receiver operator characteristic curves for both anti-bodyy tests were calculated (Fig. 1), showing sensitivity and specificityy according to different optical density cut-off points.. The best theoretical cut-off value is the point on the curvee that is closest to the upper left-hand corner. For this analysis,, we used samples from 71 participants with KS of thee homosexual men cohort as an index for true positivity, andd from all 302 women from the drug user cohort as an in-dexx for true negativity, because no KS had occurred among thesee women in 12 years of follow-up. Based on these pa-rameters,, the sensitivity of the combined assay format (pos-itivee for either O R F 6 5 / v p l 9 or ORF73/LANA or both) wass 81.7% (95% CI: 70.7-89.9) and the specificity was 90.76%% (95%) CI: 86.9-93.8).
Too facilitate comparison of antibody levels to ORF65 and ORF73,, serum-over-cut-off (S/C) ratios were calculated. S/CC ratios greater or equal to 1, representing positive anti-bodyy results, were compared among HIV-infected partici-pantss with KS before and after KS was diagnosed. S/C ra-tiostios were calculated in HIV-infected participants without KSS according to the following five CD4 categories: 401-500,301-400,201-300,101-- 200, and 1-100 c e l l s / m m \
> > c c a a c/5 5 100 0 80 0 60 0 40--20' ' 0-" " 0.350 0 1000 80 60 40 20 Specificityy (%)
B B
e. .
't/i i c c 0) 0) <n <n 100 0 80 0 60 0 40 0 20 0 100 0 80 0 60 0 40 0 20 0 Specificityy (%)Figuree 1. Receiver-operator characteristic curve for assays to de-tectt antibodies against HHV8 lytic ORF65 (A) and latent ORF733 (B) antigens.
Detectionn of HHV8 DNA.
Too detect HHV8 D N A in serum, we developed a sensitive nestedd PCR. Primers were designed based on the HHV8 se-quencee in the plasmids used for the expression of the ORF655 and ORF73 antigens (13, 20). Both plasmids were purifiedd and used in in vitro reconstruction experiments. Thee sensitivity of the assay was measured by limiting
dilu-tionn of the purified linearized plasmids (cut with the restric-tionn enzyme BamHl). Both PCRs were optimized to detect reproduciblyy an input of one to five plasmid copies. Nucleic acidss were isolated by using a previously described method startingg with 100 ml of serum (22). One-fifth of the isolated nucleicc acids were amplified with ORF65- or ORF73-spe-cificc primers. The sense primer 5'-ORF65-l (5'-GAGTTTCCGCGGCGTCGGCTTA)) and the anti-sensee primer 3'-ORF65-4 (GGGTTTCCTCGCGTCGG-CCA)) were used in the first PCR of ORF65, and sense primerr 5'-ORF73-l (5'-AGCTAGGCCACAACACATCT) andd the antisense primer 3'-ORF73-4 (5'-ACAATAAGTT-ATGGGCGACT)) were used in the first ORF73 reaction. Amplificationn was performed in our standard PCR mix containingg buffer, dNTPs, Taq, and optimized MgCb. in a totall volume of 50 ml (23). One-twentieth of the amount usedd in the first reaction was subjected to the second round off amplification by using the sense primer 5'-ORF65-2 (5'-GACGTTCACCGTGCCTTCGA)) and antisense primerr 3'-ORF65-3 (5'-GGCCGTTTCCGTCGTGGAT-GA)) for ORF65 detection and the sense primer 5'-ORF73-22 (5'-AGGATGGAAGACGAGATCCA) and thee antisense primer 3'-ORF73-3 (5'TCTGGTCCAGGG-TGGGGCAA)) for ORF73 detection. The amplified frag-mentss (169 bp for ORF65 and 150 bp for ORF73) were identifiedd by gel electrophoresis. Random PCR fragments weree directly sequenced to confirm the origin of the ampli-fiedfied fragments and all belonged to the ORF65- or ORI;73-encodedd genome fragments, respectively.
PCRR testing of serum samples of participants began with thee first sample that tested positive for HHV8 and was ex-tendedd to samples taken 3-6 months before and after the seroconversionn point to determine whether the positive re-actionn was due to a reactivation or due to a primary infec-tion.. Samples were available for 175 of 215 homosexual menn who were identified as seroconverter for HHV8.
Al-thoughh the nested PCR made detection possible to the one-moleculee level, HHV8 D N A could be detected in only 188 of the 175 (10.3%) HHV8 seroconverters.
Statisticall Analysis.
Levelss of antibodies against antigens encoded by ORF65 andd ORF73 were compared by using Student's /test among HIV-infectedd and HIV-uninfected participants who seroconvertedd for HHV8. This comparison was done for participantss who seroconverted for ORF65/vpl9 only, or O R F 7 3 // LAN A only, and for those who seroconverted for bothh antigens. In the latter group the order of seroconversionn was compared by using %z test.
Seroconversionn rates for HIV or HHV8 per calendar year weree calculated and expressed as cases per 100 person years (pyrs)) of follow-up, or incidence density (ID), and 95% CI. Polynomiall smoothing was applied to ID curves to illustrate thee change or stability of the ID over time.
HHV88 Antibody Profiles in
thee Seroconverter Population.
AA total of 215 HHV8 seroconverters were identified among homosexuall men. Of these, 106 (49%) seroconverted for la-tentt antigen ORF73,159 (74%) seroconverted for lytic anti-genn ORF65, and 50 (23%) seroconverted for both antigens. Fig.. 2 shows the seroconversion in six persons with differ-entt antibody patterns. The antibody levels for both ORF65 andd ORF73 antigens are persistently elevated, although withh different heights. In patient D, the seroconversion for HIVV was associated with a temporary increase in antibodies
Too determine the year that HHV8 was introduced in the community,, the estimation of the survival distribution for HHV88 seroconversion was done among persons starting HHV8-- seronegative until the known date of HHV8 seroconversionn or censored for the other statuses or lost to follow-up,, by using the exponential form of this survival distribution.. The validity of the exponential form was checkedd by fitting a Weibull distribution and testing whetherr or not the scale parameter could be assumed to equall 1 (meaning an exponential distribution). After esti-matingg the incidence per year, as a parameter of the expo-nentiall survival distribution, and the HHV8 seroprevalence att the start of the cohort, we computed the year at which the viruss could have been introduced, at the latest. For this pur-posee we used the well established relation between the Pois-sonn distribution and the number of events up to a certain timetime and the gamma distribution (i.e., exponential in this case)) of that time to event (24). Thus, for example, if the in-cidencee is 50 per year and there are 100 cases, 2 years are re-quiredd to reach this prevalence and even longer if some of thee 100 patients died.
againstt ORF65/vpl9. Among the 50 individuals seroconvertingg for both ORF65/vpl9 and O R F 7 3 / LANA,, HIV-seronegative individuals seroconverted in 37.5%% of cases (9 of 24) for O R F 6 5 / v p l 9 before ORF73/LANA,, in 33.3% (8 of 24} for ORF73/LANA be-foree O R F 6 5 / v p l 9 , and simultaneously in 7 cases. In con-trast,, HIV-seropositive individuals seroconverted in only 15.4%,, of cases (4 of 26) for O R F 6 5 / v p l 9 before ORF73/LANA,, in 65.4% (17 of 26) for ORF73/LANA beforee ORF65/vpl9, and in 5 cases simultaneously (P -0.066).. HIV-infected and HIV-uninfected persons seroconvertingg for either ORF65 or ORF73 antigens had
Results s
HIV-seropositive e patientt A HII V-seroconversion 25 ll patient C . HIV-seronegative e 2 AA patient E y e a r ss f r o m c o h o r t e n t r y y e a r ss from HIV-seroconversion 44 6 8 10 12 14
yearss from cohort entry
yearss from cohort entry
!!
patient D
- 6 - 4 - 22 0 2 4 6
yearss from HIV-seroconversion
2
^^ patient F
f , w -
N N
*-*-*-*-*-*-*-*-** *
66 8 10 12 14
yearss from cohort entry
Figuree 2 . Antibody responses against HHV8 lytic O R F 6 5 and latent ORF73 antigens in six homosexual individuals, either HIV-infected
(patientss A and B), HIV seroconverters (patients C and D), or HIV-seronegative (patients D a n d E); dashed line, O D o f ORF65; solidd line, O D of O R F 7 3 ; vertical, dashed tine, KS diagnosis.
S/CC r a t i o in HIV+ a n d HIV- h o m o s e x u a l m e n
0 R F 6 55 O R F 7 3
x x
rr i
H1V++
HIV-S/CC ratio for ORF 65
i r ^ ^
Wlll
Ï
CD44 category
S/CC ratio for ORF 73
T T
rrj rrj
T T
CD44 category
Figuree 3. S/CC ratios for antibody responses to HHV8 lytic O R F 6 5 and latent ORF73 antigens in HIV-infected and uninfected
homosex-uall m e n (A), in HIV-infected men without KS according to five C D 4 categories (401-500, 301-400, 201-300, 101-200, and 11 -100 cells/mm3
), and in HIV-infected men with KS before and after KS was diagnosed [ORF65 (B) and ORF73 (C)]. Of 71 pa-tientss with K S , 69 had samples taken before and after KS was diagnosed. Of these, 41 and 44 tested positive for antibodies againstt O R F 6 5 before and after KS was diagnosed, respectively (B); 36 and 33 tested positive for antibodies against ORF73 beforee and after KS w a s diagnosed, respectively (C).
comparablee antibody levels at the first sample that tested positive.. However, higher antibody levels against both O R F 6 5 / v p l 99 and ORF73/LANA were observed during follow-upp in HIV-infected persons (Fig. 3A). For O R F 6 5 / v p l 9 ,, the median S/C ratio in HIV-infected individualss is 2.3 vs. 1.6 for nonHIVinfected individuals (P -0.0001).. For ORF73/LANA, these figures are 3.3 and 1.9, respectivelyy (P = 0.0001). To investigate the impact of a KS event,, we compared HHV8 antibody levels in samples takenn before and after KS development in HIV-infected in-dividuals.. The median S/C ratio for antibodies against O R F 6 5 / v p l 99 was 3.0 (interquartile range 1.5-4.7) before andd 4.9 (3.9-5.5) after KS was diagnosed (Fig. 3B; P = 0.004).. Concurrently, the median CD4 cell count was 290 (140-430)) cells/mm1 before and 75 (10-200) cells/mm3 af-terr KS was diagnosed. The median S/C ratio for antibodies againstt ORF73/LANA was 4.8 (2.0-5.3) and 3.7 (1.8-5.0), respectivelyy (Fig. 3C; P = 0.28), and median CD4 cell count att the time samples was obtained among the positive tested personss was 380 (265-635) cells/mm3 before and 30 (10-160)) cells/mm3 after KS was diagnosed.
Forr five CD4 cell count categories in HIV-infected partici-pantss without KS, the median S/C ratios of the positive test resultss were: 1.8, 2.0, 2.3, 2.9, and 3.0 for O R F 6 5 / v p l 9 and 2.8,, 3.1, 3.3, 3.5, and 4.1 for ORF73/LANA in patients with CD44 ranging from 401 to 500, 301 to 400, 201 to 300, 101 too 200, and 1 to 100, respectively (Fig. 3 B and C). The S/C ratioss for ORF65/vpl 9 showed a statistically significant in-creasee when CD4 cell declined, but the increase for ORF73/LANAA S/C ratios was not significant (P = 0.04 and
PP - 0.38, respectively).
HHV88 DNA in Serum During Seroconversion.
Wee detected HHV8 D N A in 18 of 175 HHV8 seroconverterss (10.3%; 95% CI: 5.8-14.8%), with
compara-blee proportions of HIV-infected men in the group with and withoutt demonstrable HHV8 DNA (61% and 54%, respec-tively).tively). The DNA could be measured either before (n =8), att (n 5 9),, or after (n = 1) HHV8 seroconversion (Fig. 4). HHV88 DNA positivity did not persist, although four cases weree positive on two or more consecutive occasions. The codingg region for ORF65, ORF73, and both ORF65 and ORF733 was detected by PCR in 11 of 18 cases, 4 cases, and 33 cases, respectively. The period from the last negative until thee first positive PCR ranged from 63 to 189 days and from thee first positive PCR until the first subsequent negative PCRR ranged from 65 to 182 days, bracketing a maximum periodd of HHV8 serum DNA positivity of 364 days. The windoww phase from HHV8 DNA positivity until HHV8 antibodyy positivity ranged from 0 to 112 days. However, be-causee of the sampling frequency of 3 months, duration of DNAA presence of less than 3 months could not be esti-mated. . H0236 6 H0596 6 H0016 6 H6094 4 H0480 0 H6029 9 H0140 0 H0485 5 H0034 4 H0494 4 H0384 4 HOII 36 H6062 2 HO078 8 H0415 5 H0270 0 H0118 8 H1153 3
or r
iii ii i 00-10000 -4001 1
-300 011 1
It t
II I
II I
II I
II I
-200 0 11 11 1
11 11 1
1 1
11 11 1
1 1
II I
II I II I II I II I II III I
II I
-100 0 ++ II -III ++ II ++ II ++ II ++ 1 ++ 1-1 1
-1 1
-1 1
- 1 1 - 1 1 ++ 1 ++ 1 ++ 1 ++ 1 ++ 1 -- I I ' 0 0II I
II I
II I
II I
II I
II I
II I
II I II I II I II I 1! !II I
II I
100 0 LZTJHHV88 seronegative r~ZTJHHV88 seropositive CZUU negative PCR ++ HIV-positive -- HIV-negativeZ J J
11 II 2000 300 40dayss since HHV8 seroconversion
Figuree 4 HHV8 antibody status and PCR positivity for HHV8 DNA inn serum among HIV-seropositive(+) and HlV-sero-negative(-)) homosexual men.
HHV88 and HIV Seroconversion Rates.
Withinn the Amsterdam Cohort Studies, the HIV serocon-versionn rate among homosexual men has dropped between 19844 and 1996 from 7.1 to 1.8 per 100 pyrs and among IV drugg users between 1985 and 1996 from 7.1 to 3.1 per 100 pyrss (Fig. 5A). In contrast, the HHV8 seroconversion rate hass remained relatively stable both for homosexual men, at 3.66 per 100 pyrs, and injecting drug users, at 0.7 per 100 pyrs (Fig.. 5B). In the cohort of homosexual men, the HHV8 seroconversionn rate among those with HIV infection is consistentlyy higher (6.0 per 100 pyrs) than in those without HIVV infection (2.6 per 100 pyrs) (Fig. 5C), but this differ-encee is not seen among the men and women in the drug-us-ingg cohort. Survival analysis indicated that HIV seroposi-tivityy negatively influenced HHV8-free survival of homo-sexuall men, but that CD4+ cell counts had no impact on the ratee of HHV8 seroconversion, either in the HIV-sero-positivee or in the HIV-seronegative group. HIV-infected homosexuall men with more than 15,850 HIV-1 RNA cop-iess had no increased risk over men with low HIV-1 RNAcopyy numbers for seroconversion for HHV8 (relative riskk 50.96, 95% CI: 0.58-1.63; P = 0.91).
Too compute the latest point in time at which the HHV8 couldd have been introduced in both the homosexual com-munityy and injecting drug users, an exponential survival curvee was fitted for HIV-seronegative individuals, exclud-ingg the potential effect of HIV. We found the exponential curve,, implying a constant hazard, justified, because the Weibulll curve did not yield a better fit. We estimated the lat-estt year of introduction by dividing the prevalence in a givenn year by the incidence rate. This is justified by the cor-respondencee of the exponential distribution and the Pois-sonn distribution (24). HHV8 was introduced in the commu-nityy of homosexual men no later than 1956 (95% CI for 1951-1961),, and for injecting drug users this was 1983 (95%
homosexuall men IDUs s Calendarr time 1 1 1 1 homosexual men AA IDUs I I
[J J
r r_ _
j-h h
II i
1f f
f* *j j
i** f 855 86 87 88 89 90 91 92 93 94 95 96 Calendarr time1 1
HIV-infected d HIV-uninfected d-+M4 4
II i I I I I1 1
1 1
i i
855 86 87 88 89 90 91 92 93 94 95 96 Calendarr time Figuree 5Figure.. 5. Seroconversion rate (incidence) per 100 pyrs of HIV (A)(A) and HHV8 (S) among homosexual men ( • , dashedd line) and drug users (-••, bold line). (C) HHV8 seroconversionn rate per 100 person years among HIV-infectedd ( • ) and HIV-uninfected (D) homosexual men. .
100, , 80---- 4(H o o 2& & » — 5 — - s s
u u
.j.^..I-5-i-H H
oo oral sex •• anal sex 855 86 87 88 89 90 91 92 93 94 95 96 calendarr time t t 03 3 10 0 8 8 6 6 4 4 2 2 0 0 oo oral sex anal sex T T !! "I < <11H H
L—*t—* * i — < < 55 ( < < 5 5-LULL-. .
i i < < < < > > )) < --' "" T II I I I I I I I I I 1 I 855 86 87 88 89 90 91 92 93 94 95 96 calendarr timeFiguree 6. Unprotected anogenital vs. orogenital sexual behavior over time in the Amsterdam Cohort of homosexual men. AA Percentage of participants with at-risk contacts (95% CI).
BB Number of partners with whom at-risk behavior was practiced (median and interquartile ranges).
CII for 1982-1984). However, it must be stressed that this is ann approximation, as the prevalence is used, instead of an exact,, true cumulative incidence. Thus, the denominator mightt be underestimated especially in cases in which a lon-gerr period exists between the introduction of the virus and startt of the observation (calculated prevalence); for that rea-son,, the estimate for the homosexual men should be re-gardedd with caution.
HHV88 Seroconversion Rate and Sexual Behavior
inn the Course of the AIDS-KS Epidemic.
Ass we have reported previously (25), the one behavioral pa-rameterr associated with HHV8 seroconversion in the ho-mosexuall cohort was the number of partners with whom orogenitall sex was practiced (P = 0.04), whereas no such as-sociationn was found for anogenital or oroanal sex. The odds ratioo for orogenital sex with more than five partners is 4.13 (95%% CI: 2.34-7.31). When adjusted for HIV seropositivity, thiss figure increased in HIV-seronegative individuals to
5.955 (95% CI: 2.88-12.29), as was reported by Dukers et al. (25).. In this study, the risk to seroconvert for HHV8 among HIV-seronegativee individuals increased with the number of sexuall partners with whom orogenital sex was practiced, whereass in HIV-infected men such a dose-response effect wass not observed.
Thee dynamics of unprotected oral and unprotected anal sexuall behavior was studied in the course of the epidemic (Fig.. 6). Whereas the proportion of cohort participants re-portingg unprotected anogenital intercourse declined from 88%% in 1985 to 57% in 1996, no decline was noted in the proportionn of participants reporting unprotected oral sex (93%% in 1986 vs. 98% in 1996) (Fig. 6A). Likewise, the num-berr of partners with whom unprotected oral sex was prac-ticedticed remained stable or increased slightly, whereas a nega-tivetive trend was seen regarding partners in unprotected anal intercoursee (Fig. 6B). In 1985, 50% practiced unprotected anall sex with three or more partners and 25% did so with fivee or more, whereas in 1996, 75% practiced unprotected
anall sex with only a single partner or not at all. Unprotected orogenitall sex was practiced by 50% with two or more peo-plee and was practiced by 2 5 % with four or more in 1986
Inn this study we sought to investigate events around the pointt of HHV8 seroconversion and to provide evidence thatt seroconversion reflects, in general, primary HHV8 in-fection.. Because our EIA, based on two recombinant H H V 88 proteins, has a sensitivity of 81.7%, it is likely that we missedd some seroconverters in our retrospective study. In somee cases, primary infection may have happened some-timee before seroconversion, and two examples for this are shownn in Fig. 4. However, in most seroconverters with de-tectablee viremia (Fig. 4), virus was detectable close to the pointt of seroconversion. The overall persistence of HHV8 antibodiess points to persistent antigen presentation even whenn H H V 8 D N A cannot be detected in blood except in HHV8-seropositivee individuals progressing to or diagnosed withh KS. This suggests that the virus remains present some-wheree in the body in a latent or lytic stage of replication, boostingg the antibody levels continuously, as found previ-ouslyy in the case of the mouse gammaherpesvirus, MHV68 (26).. We have demonstrated that HIV infection boosts HHV88 antibody levels, as reflected in S/C ratios, to both thee lytic cycle capsid protein encoded by ORF65 and the la-tentt nuclear antigen encoded by ORF73. Among HIV-infectedd individuals, CD4 cell decline and KS devel-opmentt are associated with a further rise in ORF65 anti-bodies,, suggesting promotion of active virus replication. H H V 88 antibody levels to ORF65/vp 19 are higher in people developingg KS compared with those with low CD4 levels developingg an opportunistic infection (Fig. 3B). The reason thatt we do not see, on average, an increase in O R F 7 3 / L A N AA antibody titers after the development of
comparedd with 50% with four or more and 25% with nine orr more people 10 years later.
KSS development may be that the S/C ratios of ORF73/LANN A antibodies are already boosted to their peak levelss by HIV infection. This notion is corroborated bv the observationn that among HIV- infected individuals, HHV8 seroconversionn tends to occur to ORF73/LANA before ORF65/vpl99 (Fig. 2). Possible explanations include a lack off T cell-mediated control of ORF73/LANA-expressing, latentlyy infected cells or a role of increased inflammatory cytokinee levels in HIV-infected patients in increasing its ex-pression.. In patients with nasopharyngeal carcinoma, reac-tivationtivation of Epstein-Barr virus is reflected by elevated titers off antibodies to both viral capsid and nuclear antigens that risee with severity7 of the disease (27); the same might be true forr HHV8 during disease progression.
Duringg primary HHV8 infection, the period of virus pro-ductionn before antibody development and the subsequent clearancee of virus from the blood apparently takes only monthss instead of years. This relatively short window phase andd rapid control of viremia is seen also in infections with MHV688 and other rhadinoviruses (28). Serum HHV8 vire-miaa in HHV8-infected individuals is short-lived even dur-ingg the acute phase of infection.
Thee low rate of HHV8 serum viremia seen in this study thereforee may be explained at least partly by the fact that sampless had been collected only in 3 monthly intervals. The levell of viremia may be expected to be somewhat higher in peripherall blood mononuclear cells, although in our experi-encee the copy numbers are still very low, i.e., about 10-100
copiesymll of plasma vs. 100-1,000 per 10f) peripheral blood mononuclearr cells in KS patients (29). This accords with re-centt data reported by LaDuca et al. (30).
Takenn together, our data support the assumption that most, iff not all, HHV8 seroconverters have primary infections. Basedd on this assumption, HHV8 is estimated to have been introducedd in the Dutch homosexual community in about 1956.. Because of the relatively low sensitivity of the assay (81.7%),, an underestimation of the incidence as well as the prevalencee might have occurred. An underestimated inci-dencee results in an underestimation of the risk of HHV8 in-fection.. This potential error together with the underestima-tiontion of the prevalence subsequently will be compensated in thee estimation of the introduction time because this is a ra-tiotio of both underestimated figures. In a separate study (25), wee demonstrated that HHV8 seroprevalence was higher in menn of Southern European origin enrolled in our cohort. Dutchh companies actively recruited Southern Europeans to thee Netherlands in the 1950s. The first to immigrate, from 19499 on, were Italians (31). Only in the 1960s did substan-tiall numbers of individuals immigrate from Spain, Portugal, Greece,, and the former Yugoslavia (32). The earliest immi-grants,, who settled largely in Amsterdam, came from north-ernn Italy and from Sardinia, an island that is noted for HHV88 endemicity (33-35).
Thee data reported here and in previous communications (25)) strongly indicate that the stable HHV8 seroconversion ratee among homosexual men participating in the Amster-damm Cohort Studies is related to the persistence of unpro-tectedd orogenital sex in this group. An earlier report on a co-hortt of Danish homosexual men (17) had reported a markedd decrease in HHV8 incidence between 1981 and 1984,, i.e., before the start of our cohort, which could ex-plainn this apparent difference. Our observations fit with the findingfinding that the oropharynx serves as the source of HHV8, ratherr than the port of entry, taking into account the
sub-stantiallyy higher HHV8 load in saliva compared with semen (30).. The findings are reminiscent of Epstein-Barr virus, for whichh intimate oral contact with the exchange of saliva was firstfirst established as the major route of transmission by Hoaglandd (37) and confirmed by Evans (38). Additional ev-idencee for the oral shedding of infectious virus comes from thee animal rhadinovirus group. MHV68 reaches the highest titerss in the respiratory tract and causes disease after experi-mentall infection by the intranasal route but not by the i.v. routee (39). The highest frequency of isolation of EHV2, an-otherr member of the rhadinovirus group, is from the bron-chiall lymph nodes as well as nasal swabs of naturally in-fectedd horses (40). Infectious virus was isolated from the lungss of mice experimentally infected with EHV2, but no infectiouss virus was found in their blood or lymphoid tis-suess (41). HVS is highly frequent among both captive and ferall squirrel monkeys, in which virus was consis-tendy iso-latedd from oropharyngeal secretions (42).
However,, transmission of HHV8 among homosexual men byy oral sex obviously does not represent the usual route of HHV88 transmission in endemic countries, where most transmissionn appears to occur in childhood (43-46). In this setting,, early studies suggest HHV8 transmission in fami-lies,, perhaps from mother to child or among siblings (43-46).. Such a route of transmission also could involve sa-liva,, and the importance of oral sex among gay men re-portedd here and previously (25) thus could represent an un-usuall way of saliva-mediated transmission. We conclude thatt most if not all cases of HHV8 seroconversion among HIVV seronegatives reflect primary HHV8 infection from orogenitall sex. Seroconversion appears to follow replica-tiontion and seeding of HHV8 in a new host and result in rapid controll of viremia. HIV-induced immunodeficiency may furtherr enhance HHV8 replication, which boosts the anti-bodiess to lytic antigen, an effect that is even more pro-nouncedd after development of KS, and also accelerates the appearancee of antibodies to the latent nuclear antigen.
References s
11 Chang Y, Cesarman K, Pessin MS, et al. Identification of herpcsvirus-likee D N A sequences in AIDS-associated Kaposi's sarcoma.. Science 1994;266:1865-9.
22 Desrosiers RC, Sasseville VG, Czajak SC, et al. A herpesvirus of rhesuss monkeys related to the human Kaposi's sarcoma-associ-atedd her-pesvirus. ƒ I 'irol 1997;71:9764-9.
33 Rose TM, Strand KB, Schultz ER, et al. Identification of two homologss of the Kaposi's sarcoma-associated herpesvirus (hu-mann herpesvirus 8) in retroperitoneal fibromatosis of different macaquee species, ƒ I 7re/1997;71:4138-44,
44 Melendez LV, Daniel M D , Hunt R D , Garcia FG. An apparently neww herpesvirus from primary kidney cultures of the squirrel monkeyy (Saimiri sciureus). LMII Anim Care 1968;18:374-81. 55 Plummer G, Waterson A P. Equine Herpes Viruses. I 'iro/ogy
1963;12:412-6. .
66 Blaskovic D, Stancekova M, Svobodova J, Mistrikova J. Isolation off five strains of herpesviruses from two species of free living ro-dents.. Acta 1 7ro/1980;24:468.
77 Desrosiers RC, Bakker A, Kamine J, et al. A region of the Herpesviruss saimiri genome required for oncogenicity. Science 1985;228:184-7. .
88 Cesarman B, Chang Y, Moore PS, Said |W, Knowles DM. Kaposi'ss sarcoma-associated herpesvirus-like D N A sequences aree present in AlDS-related body cavity based lymphomas. iY
ling/]ling/] Med 1995;332:1186-91.
99 Said J W, Chien K, Takeuchi S, et al. Kaposi's sarcoma-associated herpesviruss (KSHV or HHV8) in primary effusion lymphoma: ultrastructurall demonstration of herpesvirus in lymphoma cells.
BloodBlood 1996;87:4937-43.
100 Soulierj, Grollet I,, Oskenhendler F , et al. Kaposi's sarcoma asso-ciatedd herpesvirus-like D N A sequences in multicentric Castleman'ss disease. Blood 1995;86:1276-80.
111 Rickinson AB, Kieff E. Epstein-Barr Virus. In: Fields Virology, Thirdd Edition. 1996. Edited by BN Fields, D M Knipe, PM II lowley, et al. Lippincott, Philadelphia, USA.
122 Martin J N , Ganem DF,, O s m o n d D H , Page-Shafer KA, Macrae D ,, Kedes D H . Sexual transmission and the natural history' of hu-mann herpesvirus 8 infection. N Hngl] Med 1998;338:948-54. 133 Simpson GR, Schulz T F , Whitby D , et al. Prevalence of Kaposi's
sarcomaa associated herpesvirus infection measured by antibodies too recombinant capsid protein and latent immunofluorescence antigen.. Lancet 1996;349:1133-8.
144 Kedes D H , Operskalski E, Busch M, Kohn R, Flood J, Ganem D. Thee seroepidemiology of human herpesvirus 8 (Kaposi's sar-coma-associatedd herpesvirus): distribution of infection in KS risk groupss and evidence for sexual transmission. Nat Med 1996;2:918-24. .
155 G a o Sj, Kingsley I., Hoover DR, et al. Seroconversion to antibod-iess against Kaposi's sarcoma-associated herpesvirus-related latent nuclearr antigens before the development of Kaposi's sarcoma. A'
lingl]lingl] Med 1996;335:233-41.
166 G a o SJ, Kingsley L. Li M, et al KSHV antibodies among Ameri-cans,, Italians and LJgandans with and without Kaposi's sarcoma.
KatKat Med 1996;2:925-8.
177 Melbye M, Cook PM, Hjalgrim H, et al. Risk factors for Kaposi's sarcoma-associatedd herpesvirus (KSHV/HHV-8) seropositivity inn a cohort of homosexual men, 1981-1996. bit / Cancer 1998;77:543-8. .
188 O'Brien TR, Kedes D, Ganem D , et al. Tividence for concurrent epidemicss of human herpesvirus 8 and human immunodeficiency viruss type 1 in FS homosexual men: rates, risk factors, and rela-tionshipp to Kaposi's sarcoma. / Infect Dis 1999;180:1010-7. 199 Renwick N , I lalaby T, Weverling GJ, et al. Seroconversion for
hu-mann herpesvirus 8 during HIV infection is highly predictive of Kaposi'ss sarcoma. AIDS 1998;12:2481-8.
2(11 Rainbow L, Piatt GM, Simpson GR, et al. The 222- to 234-Kilodaltonn Latent Nuclear Protein (LNA) of Kaposi's sar-coma-associatedd herpesvirus (human herpesvirus 8) is encoded by orf733 and is a component of the latency-associated nuclear anti-gen.. ƒ I Irol 1997;71:5915-21.
211 Schulz T F . Epidemiology of Kaposi's sarcoma-associated herpesvirus/humann herpesvirus 8. Adv Cancer Res 1999;76:121-60. .
222 Boom R, Sol CJ, Salimans MM, Jansen CL, Wertheim-van Dillen PM,, van der NoordaaJ. Rapid and simple method for purification off nucleic acids, ƒ Clin Microbiol 1990;28:495-503.
233 Cornelissen M, Mulder-Kampinga G, Veenstra J, et al. Syncytium-inducingg (SI) phenotype suppression at seroconversionn after intramuscular inoculation of a non-syncytium-inducing/SII phenotypically mixed human immu-nodeficiencyy virus population. J t 'irol 1995;69:1810-8.
244 Mood AM, Graybill FA, Boes DC. In: Introduction to the Theory off Statistics. 1974. Edited by D Blackwell and H Solomon. McGraw-Hill,, New York, USA.
255 Dukers NF1, Renwick N , Prins, M, et al. Risk factors for human herpesviruss 8 seropositivity and seroconversion in a cohort of ho-mosexuall men. Amƒ lipidemiol20()0;\5]:2l3-24.
266 Stevenson PG, Doherty PC. Kinetic analysis of the specific host responsee to a murine gammaherpesvirus./ I 'irol 1998;72:943-9. 277 de Schryver A, Klein G, Henle G, et al. EB-virus associated
serol-ogyy in malignant disease: antibody levels to viral capsid antigens (VCA),, membrane antigens (MA) and early antigens (EA) in pa-tientss with various neoplastic conditions. Int J Cancer 1972;9:353-64. .
288 Sarawar SR, Brooks J W, Cardin R D , Mehrpooya M, Doherty PC. Pathogenesiss of murine gammaherpesvirus-68 infection in interleukin-6-deficientt mice. Virology 1998;249:359-66.
299 Wit FW, Sol C J, Renwick N, et al. Regression of AIDS related Kaposi'ss sarcoma associated with clearance of human herpesvirus-88 from peripheral blood mononuclear cells following initiationn of antiretroviral therapy. AIDS 1998;12:218-9. 300 LaDuca JR, Love JL, Abbott LZ, Dube S, Friedman-Kien AH,
Poieszz BJ. Detection of human herpesvirus 8 D N A sequences in tissuestissues and body fluids. J Infect Dis 2000;178:1610-5. 311 Schneider, M. In: (1984) P h D dissertation. University of
Amster-dam,, Amsterdam, T h e Netherlands.
322 U n d o F, Pennings T. In: Zuideuropeanen in Nederland. 1988. Ministeriee van Welzijn, Volksgezondheid en Cultuur, The Neth-erlands. .
333 Calabro ML, Sheldon J, Favero A, et al. Seroprevalence of Kaposi'ss sarcoma-associated herpesvirus/human herpesvirus 8 in severall regions of Italy. J Hum Virol 1998;3:207-13.
344 Angeloni A, Heston L, Uccini S, et al. High prevalence of antibod-iess to Human herpesvirus 8 in relatives of patients with classic Kaposi'ss Sarcoma from Sardinia. J Infect Dis 1998;177:1715-8. 355 Whitby D , Luppi M, Barozzi P, Boshoff C, Weiss RA, Torelli G.
Humann herpesvirus 8 seroprevalence in blood donors and lym-phomaa patients from different regions of Italy. J Natl Cancer Inst 1998;90:395-7. .
366 Beral V, Peterman TA, Berkelman RL, et al. Kaposi's sarcoma amongg persons with AIDS: a sexually transmitted infection?
Lan-cetcet 1990;335:123-8.
377 Hoagland RJ. The transmission of infectious mononucleosis. Am
]] MedSci 1955;229:262-72.
388 Evans AS. Infectious mononucleosis in university of Wisconsin students.. Report of a five-year investigation. Am J Hyg 1960; 71:342-62. .
399 Sunil-Chandra N P , Efstathiou S, Arno J, Nash AA. Virological andd pathological features of mice infected with murine gamma-herpesviruss 68.7 Gen Virol 1992;73:2347-56.
400 Palfi V, Belak S, Molnar T. Isolation of equine herpesvirus type 2 fromm foals, showing respiratory symptoms. Zbl Vet Med 1978;25:165-7. .
411 Rizvi SM, Slater J D , Slade AJ, Field HJ. Transmission of equine herpesviruss 2 to the mouse: characterization of a new laboratory infectionn model. J Gen Virol 1997;78:1119-24.
422 Falk LA, Nigida S, Deinhardt F, Cooper RW, Hernandez-CamachoJI.. Oral excretion of Herpesvirus saimiri in captivee squirrel monkeys and incidence of infection in feral squir-rell monkeys./ Natl Cancer Inst 1973;51:1987-9.
433 Mayama S, Cuevas LE, Sheldon J, etal. Prevalence and transmis-sionn of Kaposi's sarcoma-associated herpesvirus (human herpesviruss 8) in Ugandan children and adolescents. IntJ Cancer 1998;77:817-20. .
444 Sitas F, Newton R, Boshoff C. Increasing probability of mother-to-childd transmission of HHV-8 with increasing maternal antibodyy titer of HHV-8. N Engl J Med 1999;340:1923. 455 Gessain A, Mauclere P, van Beveren M, et al. Human herpesvirus
88 primary infection occurs during childhood in Cameroon, Cen-trall Africa. Int} Cancer 1999;81:189-92.
466 Lyall E G , Patton GS, Sheldon J, etal. Evidence for horizontal and nott vertical transmission of human herpesvirus 8 in children born too human immunodeficiency virus-infected mothers. Pediatr Infect
DisDis J 1999;18:795-9.
