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.
Seroconversionn for human herpesvirus 8 during HIV
infectionn is highly predictive of Kaposi's sarcoma
NeilKenwick,NeilKenwick, Teysir Halaby, GerritJ. Weverting Nicole H.T.M. Dukers, Guy R. Simpson,
KoelKoel A. Coutinho, Joep AL-4. Lange, Thomas F. Schulp and]aap Goudsmit
Abstract t
Background:: The finding of antibodies against human herpesvirus 8 (HHV-8) is associated with the occurrence of Kaposi's sarcomaa in persons infected with HIV. However, the predictive value of HHV-8 antibodies for Kaposi's sarcoma in HIV in-fectionn is unknown.
Methods:: The Amsterdam Cohort Studies on HIV infection and AIDS started in 1984 for homosexual men and in 1985 for injectingg drug users. Serum samples from 1459 homosexual men and 1167 drug users were tested for antibodies to recombi-nantt HHV-8 lytic-phase capsid (ORF65) antigen and latent-phase nuclear (ORF73) antigen. Individuals were retrospectively identifiedd as HHV-8-positive or HHV-8-negative at enrolment or HHV-8 seroconverter during the study. Kaposi's sar-coma-freee survival time was compared between HIV-infected men who were positive for HHV-8 at enrolment and those whoo later seroconverted for HHV-8. Hazard ratios were estimated for Kaposi's sarcoma, lymphoma, and opportunistic infec-tionn according to the HHV-8 serostatus.
Results:: The incidence of HHV-8 seroconversion among drugs users was 0.7 per 100 person-years based on 31 seroconversions,, whereas an incidence of 3.6 was found among homosexual men based on 215 seroconversions. The hazard ratioo for Kaposi's sarcoma was 3.15 (95% CI: 1.89-5.25) in HIV-infected individuals if HHV-8 antibodies were present either att enrolment or at HIV seroconversion. In HIV-infected persons who later seroconverted to HHV-8, Kaposi's sarcoma de-velopedd more rapidly: hazard ratio of 5.04 (95% CI: 2.94-8.64), an additional risk of 1.60 (95% CI: 1.01-2.53; P = 0.04). Time-dependentt adjustment for CD4+ cell count and HIV RNA had no impact on the additional risk, although the CD4+ cell countt was an independent risk factor for Kaposi's sarcoma. HHV-8 infection did not increase the risk of AIDS-related lym-phomaa or opportunistic infections.
Conclusions:: The incidence of HHV-8 infection is higherr in homosexual men than in drug users. The presence of HI IV-8 an-tibodiess in HIV-infected persons increases the risk of Kaposi's sarcoma. Among HIV-infected persons, those who subse-quentlyy seroconvert for HHV-8 are at highest risk. These results strongly confirm the causal role of HHV-8 in Kaposi's sarcomaa and emphasize the clinical relevance of HHV-8 seroconversion before and after the HIV infection.
Introduction n
I nn a l m o s t all tissue s p e c i m e n s of K a p o s i ' s s a r c o m a o f per-s o n per-sper-s infected with H I V , D N A per-sequenceper-s o f the h u m a n h e r p e s v i r u ss 8 ( H H V - 8 ) are d e t e c t e d by S o u t h e r n b l o t analy-siss a n d p o l y m e r a s e chain reaction (PCR) amplification (1). H H V - 88 D N A s e q u e n c e s are also identified in endothelial a n dd spindle cells from K a p o s i ' s s a r c o m a lesions (2) and in p e r i p h e r a ll b l o o d m o n o n u c l e a r cells of A I D S patients with thiss vascular t u m o u r (3). T h e distribution o f K a p o s i ' s sar-c o m aa a m o n g distinsar-ct H I V / A I D S risk g r o u p s distinguishes p o p u l a t i o n ss w i t h a high risk o f K a p o s i ' s s a r c o m a , such as h o m o s e x u a ll m e n , from t h o s e w i t h a lower risk, such as
in-jectingg d r u g users (4). H H V - 8 antibody prevalence rates are significantlyy higher a m o n g the high-risk g r o u p s than the low-riskk g r o u p s , with t h e highest H H V - 8 rates universally reportedd for h o m o s e x u a l m e n with Kaposi's sarcoma (5-12). .
Inn the several p o p u l a t i o n s of n o r t h e r n and western E u r o p e andd the United States, H H V - 8 infection appears to b e infre-quentt c o m p a r e d with s o m e Mediterranean countries and, in particular,, with parts of Africa (5,7,11-15). T h e frequency ratess for classic, o r H I V - n e g a t i v e , Kaposi's sarcoma are
generallyy higher in regions where HHV-8 is more prevalent, butt even in those areas, it rarely occurs in an individual who iss HHV-8 seropositive but otherwise healthy (13,14,16). Thee clinical severity of Kaposi's sarcoma in the context of HIV-11 infection (16,17) and the fact that AIDS-related Kaposi'ss sarcoma appears to be far more often associated withh HIV-1 infection than HIV-2 infection (15) suggests thatt HIV-1 infection represents an important co-factor for thee development of Kaposi's sarcoma in an HHV-8-in-fectedd individual. However, in HIV-1-infected individuals otherr than homosexual or bisexual men (i.e. heterosexually infectedd women, injecting drug users, patients with haemo-philia)) the rate of AIDS-related Kaposi's sarcoma is as low inn HHV-8 endemic countries of southern Europe as it is in countriess where HHV-8 infection is uncommon in the
gen-Materialss <
Studyy participants
Thee Amsterdam Cohort Studies on HIV infection and AIDSS started in October 1984 for homosexual men (19) andd in December 1985 for injecting drug users (20). A total off 1218 drug users and 1666 homosexual men were enrolled inn the cohort studies by 31 December 1996. All participated inn the present study except for 192 homosexual men, for whomm there was only one serum sample as a result of recent enrolmentt or short participation in the cohort studies. In addition,, 15 homosexual men and 51 drug users were ineli-giblee because of a paucity of clinical material. All serum sampless were obtained before 31 December 1996, whereas clinicall data were present until 1 May 1997.
erall population (16). The mode or timing of HHV-8 trans-missionn may therefore determine the risk of Kaposi's sar-comaa (13).
Inn this study, we investigated the incidence and prevalence off HHV-8 infection in injecting drug users and homosexual menn participating in the Amsterdam Cohort Studies on HIVV infection and AIDS. The enzyme immunoassay (EIA) systemm was used to detect IgG antibodies against the HHV-88 lytic-phase capsid antigen (ORF65) (11) and a frag-mentt of the HHV-8 latent-phase nuclear antigen (ORF73) (18).. In addition, we estimated the risk of Kaposi's sarcoma, opportunisticc infection, and lymphoma attributable to the presencee of HHV-8 antibodies among HIV-infected ho-mosexuall men.
II methods
Clinicall follow-up
AIDS-definingg events such as AIDS-associated malignan-ciess and opportunistic infections were diagnosed according too the guidelines in the (1993) revised classification system forr HIV infection and expanded surveillance case definition forr AIDS among adolescents and adults (21).
Enzymee immunoassay method
Antibodiess to HHV-8 were detected by EIA using a modi-fiedfied version of a method that has been described previously (11).. Microtitre plates (Greiner BV, Alphen a/d Rijn, the Netherlands)) were coated with 750 ng recombinant antigen (aaa 86-170) encoded by ORF65.2 and associated with the lyticc stage of HHV-8 infection or with 100 ng of a carboxy-terminall fragment (aa 951-1163) of the latency-as-sociatedd nuclear antigen (LAN A) that is encoded by ORF73
(11,18).. The antigen concentration was determined in the bicinchoninicc acid protein assay (Pierce, Rockford, Illinois, USA).. Because both are fusion proteins with mouse dihydrofolatee reductase that were expressed in the pQE42 expressionn system (Qiagen G m b H , Hilden, Germany), pro-teinn derived from this vector but lacking an HHV-8-specific insertionn was used as the control antigen. Serum samples fromm 40 HIV-seropositive drug users and 186 HIV-sero-positivee men were tested with the control antigen, yielding onlyy a single positive result (1/226; 0.4%), indicating mini-mall fusion protein reactivity. Patient sera were tested in a 1:1000 dilution, and samples were tested for HHV-8 anti-bodiess in a blinded set-up.
Studyy design
T oo determine whether participants were HHV-8 seronega-tivetive throughout the study, HHV-8 seropositive at enrol-ment,, or HHV-8 seroconverters during the study, the most recentlyy obtained serum samples from 1167 drug users and 14599 homosexual men were tested using the EI A system. If aa sample tested negative, the participant was considered to havee had no antibodies against HHV-8 throughout the study.. If a sample tested positive, the enrolment sample was testedd to determine whether an individual was positive at entryy or seroconverted during follow-up. Among serocon-verters,, the year of seroconversion was determined by test-ingg serum samples at yearly intervals and, within the year of seroconversion,, at intervals of 3 to 6 months. The midpoint betweenn the last negative sample and the first positive sam-plee (seroconversion sample) was considered to be the date off H H V - 8 seroconversion.
Studyy definitions
Resultss from the EI A system were considered to be HHV-8 positivee if they showed optical densities of 0.350 or greater
forr ORF65 and 0.375 or greater for ORF73. This cut-off valuee was three times the standard deviation of the mean opticall density from a panel of 40 individuals at low risk for Kaposi'ss sarcoma (unpublished data).
Participantss who tested HHV-8 negative at the end of fol-low-upp were considered to be negative throughout the study.. However, to investigate the potential for false negativity,, the enrolment samples of 200 participants whose mostt recent sample tested negative were evaluated using the EIAA system. A positive result at entry was found for nine of thee 200, yielding a putative false negativity rate of 4.5% (95%% CI: 2.1-8.4).
HHV-88 seropositivity was defined as a positive EIA result forr one or both antigens, and HHV-8 seroconversion was definedd as a change during follow-up from an HHV-8 nega-tivee to an HHV-8 positive EIA result. The seronegative and seropositivee phases of seroconversion were consistent for 1766 of 215 (82%) HHV-8 seroconverters, but for the re-mainingg 39 (18%) individuals, antibody concentrations oc-casionallyy dropped below the cut-off value of the assay dur-ingg the seropositive period. In those cases in which a posi-tivee sample could be confirmed by the subsequent sample, thee initial positive sample was considered to be the seroconversionn sample.
Statisticall analysis
Incidencess of HHV-8 infection per 100 person-years with 95%% confidence intervals (95% CI) were calculated for in-jectingg drug users and homosexual men at risk. On the basis off the most recentiy obtained serum samples of partici-pants,, HHV-8 seroprevalences with 95% CI were calcu-latedd for the two groups and for AIDS subgroups defined byy AIDS-related Kaposi's sarcoma, opportunistic infection, orr lymphoma. Odds ratios were calculated for the
associa-tionn between HHV-8 seropositivity and these events, but onlyy for the HIV-infected homosexual men. Among the drugg users there was only one single case of Kaposi's sar-coma. .
Kaplan-Meierr techniques were used to estimate the distri-butionn of times from the moment of infection with both HIVV and HHV-8 to the diagnosis of Kaposi's sarcoma, ac-cordingg to the order of infection: HIV infection first and subsequentlyy HHV-8 (HHV-8 seroconverters) or HHV-8 firstfirst and subsequently HIV (HHV-8 positive). In partici-pantss seropositive for both infections at enrolment, HHV-8 seroconversionn was assumed to have occurred first, because HHV-88 was highly prevalent among HIV-uninfected ho-mosexuall participants.
Prospectively,, the risk of Kaposi's sarcoma as predicted by thee presence of HHV-8 antibodies was estimated using a
Amongg 1167 injecting drug users, 89 were seropositive for HHV-88 (7.6%; 95% CI: 6.2-9.3%), whereas among 1459 homosexuall men, 520 were HHV-8 seropositive (36%; 95%% CI: 33-38%). HHV-8 prevalence among drug users wass the same whether they were HIV infected or not (Table
1),, but among homosexual men the HHV-8 prevalence was markedlyy higher in HIVinfected men (Table 2). In addition, HHV-88 infection in drug users was not associated with sex andd education. There was a correlation with age and homo-sexuall activities (in men). In homosexual men, increasing agee was associated with HHV-8 infection (P<0.001). Furthermore,, when HHV-8 seroprevalence was investi-gatedd among subgroups on the basis of AIDS-related eventss (Table 3), it was found highest among individuals
time-dependentt Cox proportional hazards model, which enabledd risk estimation on the basis of HHV-8 antibody sta-tuss over time. Homosexual men were included who were ei-therr HIV-seropositive at enrolment or who became HIV-seropositivee during follow-up. For the latter group, thee HIV seroconversion date was taken as the start of the observationn time. The same procedure was followed to es-tablishh the risk of other AIDS-related events, with opportu-nisticc infection or lymphoma as the end point. To adjust for thee deterioration of the immune system, all measurements off CD4+ cell count were included in the timedependent modell as a continuous variable. The same procedure was followedd for HIV-1 RNA copy number, using the NASBA HIV-11 RNA Q T assays (Organon Teknika, Boxtel, the Netherlands)) with the detection limit at 1000 HIV-1 RNA copies/ml. .
withh Kaposi's sarcoma (87.3%). The odds ratio between HHV-88 positivity and AIDS-related Kaposi's sarcoma was 8.99 (95% CI: 4.3-18.3).
AA total of 31 HHV-8 seroconversions occurred in the co-hortt of drug users, revealing an incidence of HHV-8 infec-tionn of 0.7 per 100 person-years (95% CI: 0.5-0.9), which didd not differ between HIV-infected and HIV-uninfected drugg users (0.6 per 100 personyears (95% CI: 0.2-1.2) and 1.22 (95% CI: 0.4-1.1), respectively). A total of 215 HHV-8 seroconversionss occurred in the homosexual cohort, dem-onstratingg an incidence for HHV-8 infection of 3.6 per 100 personyearss (95% CI: 3.2-4.1). The incidence among HIV-infectedd men was 6.2 (95% CI: 5.1-7.4) compared with 2.66 among HIV-seronegative men (95% CI: 2.1-3.1).
Tablee 1. Prevalence of HHV-8 antibodies in injecting drug users
accordingg to HIV serological status, sex, and sociodemographicc characteristics.
Tablee 2. Prevalence of HHV-8 antibodies in homosexual men
ac-cordingg to HIV serological status, age, and nationality.
Characteristic c HIVV infection* Uninfected d Infected d Sex* * Male e Female e Agee (years) <30 0 >30-35 5 > 3 5 ^ 0 0 >40-45 5 >45-50 0 >50 0 Nationality*8 8 Dutch h Non-Dutch h Education*** * No o 1-2 2 3-4 4 5-6 6 > 6 6 No.tested d 816 6 351 1 672 2 495 5 301 1 296 6 286 6 180 0 77 7 27 7 812 2 350 0 286 6 135 5 327 7 259 9 156 6 Homosexuall contacts(for men)11
No o Yes s 603 3 69 9 I I anti-HI I 7,7 7 7.4 4 7.7 7 7.5 5 10.0 0 8.8 8 4.9 9 7.8 8 5.2 2 3.7 7 6.7 7 9.7 7 7.7 7 8.9 9 6.1 1 7.7 7 9.6 6 6.8 8 15.9 9 Percent t HV-88 positive3 (6.0-9.8) ) (4.9-10.7) ) (5.7-9.7) ) (5.2-9.8) ) (6.6-13.4) ) (5.6-12.0) ) (2.4-7.4) ) (3.9-11.7) ) (0.2-10.2) ) (0.0-10.8) ) (5.0-8.4) ) (6.6-12.8) ) (4.6-10.8) ) (4.1-13.7) ) (3.5-8.7) ) (4.5-11.0) ) (5.0-14.2) ) (4.8-8.8) ) (7.3-24.5) ) Characteristic c HIVV infection11 Uninfected d Infected d Agee (years)11 <30 0 >30-35 5 >35-40 0 >40-45 5 >45-50 0 >50 0 Nationality** * Dutch h Non-Dutch h aa (95% CI); CI, No.. tested 859 9 600 600 340 0 185 5 261 1 245 5 204 204 223 3 1237 7 190 0 confidencee interval. Percent t anti-HHV-88 positive3 26.44 (23.5-29.4) 48.99 (44.9-52.9) 16.88 (12.8-20.8) 29.77 (23.1-36.3) 43.33 (37.4-49.4) 42.44 (36.2-48.6) 46.66 (39.8-53.5) 43.00 (36.5-49.5) 34.88 (32.2-37.5) 40.55 (33.5-47.5)
** Data were not available for ail individuals in every category. ** P = 0.125 by overall chi-square test of independence. '' P < 0.001 by overall chi-square test for independence,
chi-squaree test for trend for age
(95%% CI); CI, confidence interval Yearss of education since 12 years of age.
DataData were not available for all individuals in every category. .
PP > 0.70 by overall chi-square test of independence, chi-squaree test for trend for education,
PP = 0.08 by overall chi-square test for independence. PP > 0.05 by overall chi-square test for independence, chi-squaree test for trend for age.
HIV-infectedd subjects at enrolment HIVV seroconverters
Years s
Figuree 1. Kaplan-Meier estimates of time from infection with both HIV and HHV-8 to the development of Kaposi's sarcoma.
AA Estimates in participants HIV-infected from time of enrolment, for HHV-8-positive participants (HHV-8+, bold line, 25 Kaposi's sarcomaa events) and HHV-8 seroconverters (HHV-8 sc, 23 Kaposi's sarcoma events, thin line).
BB Estimates in HIV seroconverters according to HHV-8-positive participants (HHV-8+, bold line, 11 Kaposi's sarcoma events) or HHV-88 seroconverters after HIV seroconversion (HHV-8 sc, 7 Kaposi's sarcoma events, thin line). The numbers of participants att risk at different time points are depicted at the bottom of each plot. KS, Kaposi's sarcoma.
Overr 12 years of follow-up, Kaposi's sarcoma developed in 999 of 600 HIV-infected men, and no Kaposi's sarcoma was observedd among the HIV-uninfected participants. There-fore,, only HIV-infected homosexual men are included in thee subsequent analysis. Figure 1, (a), shows the Kaplan-Meierr estimates of time since infection with both HHV-88 and HIV to the development of Kaposi's sarcoma, accordingg to the order of infection. Participants who were infectedd with HIV at enrolment and subsequently seroconvertedd for HHV-8 remained free of Kaposi's sar-comaa for a shorter time than participants who were co-infectedd with HIV and HHV-8 from the start of the studyy (P = 0.0019). In addition, when only HIV seroconverterss were considered (Fig. 1, (b)), again
partici-pantss who seroconverted for HHV-8 after being HIV-infectedd remained free of Kaposi's sarcoma a shorter timee than participants who were HHV-8 positive at the time off HIV seroconversion (P = 0.04).
Kaplan-Meierr estimates could not take into account the du-rationn of HIV exposure before HHV-8 seroconversion in thee HHV-8 seroconverters. To eliminate this potential con-founder,, we therefore used a time-dependent Cox propor-tionall hazards model, in which the duration of HIV infec-tionn exposure time is taken into account. In that model, the riskrisk for developing Kaposi's sarcoma among all 600 HIV-1 infectedd men was 3.15 (95% CI: 1.89-5.25) for participants whoo were HHV-8 seropositive before HIV infection. In
Tablee 3. Prevalence of HHV-8 antibodies in subgroups defined by clinical events. Homosexuall men AIDS S Kaposi'ss sarcoma Opportunisticc infections Lymphoma a No.. tested 600 600 193 3 71 1 154 4 11 1
Percentt anti-HHV-8 positive (95%CI)
48.99 (44.9-52.9) 62.22 (55.3-69.0) 87.33 (77.3-94.0) 61.00 (53.3-68.7) 72.77 (39.0-94.0) Oddss ratio (95%C 2.22 (1.6-3.2) 8.99 (4.3-18.3) 1.99 (1.3-2.8) 2.88 (0.7-10.8)
Oddss ratio of the association between the prevalence and clinical event. Odds ratios are calculated from a total of 600 HIV-infected homosexuall men among whom the events occurred. Events listed are not necessarily the first AIDS-defining events. Individuals can appearr in more than one event category.
Tablee 4. Hazard ratios for the development of Kaposi's sarcoma according to updated information on HHV-8 serological status, CD4+ cell
count,, and HIV RNA copy number.
HHV-88 prevalent HHV-88 seroconversion CD4++ cell countt HIV-11 RNAt Univariate e HR(95%% CI) 3.17(1.81-5.57) ) 5.64(3.16-10.10) ) 0.588 (0.50-0.66) 2.08(1.56-2.78) ) PP value << 0.001 << 0.001 << 0.001 << 0.001 Multivariate e HRR (95% CI) 3.27(1.86-5.77) ) 5.17(2.88-9.27) ) 0.59(0.51-0.68) ) 1.255 (0.93-1.69) PP value << 0.001 << 0.001 << 0.001 0.1388 8 HR,, hazard ratio. CI,, confidence interval.
tt CD4+ cell count: per 100 cells increase.
$$ HIV-1 RNA copies per logi0/ml increase.
thosee who seroconverted for HHV-8 after being HIV in-fected,, the risk was 5.04 (95% CI: 2.94-8.64), which indi-catess an additional risk of 1.60 (95% CI: 1.01-2.53; P=0.04). Forr a subgroup of 505 HIV-1 infected participants, we had accesss to measurements of CD4+ cell count as well as HIV-11 RNA copy numbers taken in the course of fol-low-up.. As no differences in proportions were found with respectt to clinical and virological indicators (data not shown),, this subgroup was considered representative of the entiree group of 600. As shown in Table 4, the relative risk
foundd in the univariate analysis for the development of Kaposi'ss sarcoma in subgroup participants who were HHV-88 positive was similar to the risk found for the com-pletee group of 600 participants: 3.17 (95% CI: 1.81-5.57) (Tablee 4) for HHV-8 positivity by the time of enrolment or beforee being HIV-1 infected, and 5.64 (95% CI: 3.16-10.10) forr individuals who subsequently seroconverted for HHV-88 after HIV infection, which is an additional risk of 1.788 (95% CI: 1.09-2.90; P = 0.021). Interestingly, both CD4++ cell count as well as HIV-1 RNA copy number are
(Tablee 4). However, in the multivariate analysis, HIV-1 RNAA was no longer predictive, whereas the CD4+ cell
countt as well as HHV-8 seropositivity remained predictive andd thus are independent risk factors (Table 4).
Fromm a clinical perspective, a decrease in CD4+ cells is of
moree importance when the CD4+ cell count is relatively low.. In order to accommodate this the square root of the
Thee finding of HHV-8 DNA in peripheral blood has been reportedd to precede (3,22,23) and predict (3) the develop-mentt of Kaposi's sarcoma in HIV-infected individuals. However,, PCR detection of HHV-8 DNA in peripheral bloodd is not very sensitive and identifies only approximately 50%% of the patients with Kaposi's sarcoma (3,22). The asso-ciationn could therefore reflect increased HHV-8 replication, ratherr than HHV-8 infection, predictive of Kaposi's sar-coma.. Here we demonstrate in a large cohort of HIV-in-fectedd individuals that HHV-8 seropositivity, as a reflection off HHV-8 infection, predicts progression to Kaposi's sar-comaa in HIV-infected homosexual men, and that HHV-8 infectionn is a separate risk factor in addition to HIV infec-tion. .
Too measure antibodies to HHV-8 we used two previously describedd recombinant HHV-8 proteins, derived from the lytic-phasee antigen (encoded by ORF65) and from the la-tent-phasee nuclear antigen, LAN A (encoded by ORF73) (11,12,18).. An EI A system using these two antigens de-tectedd HHV-8 antibodies in 87% of AIDS-related Kaposi's sarcomaa cases, yielding a sensitivity compared with that of thee latent immunofluorescence assay and ORF65 E I A / W B (7,8,11).. Application of a more sensitive serological test may
CD4++ cell count was entered in the Cox model. This model hadd a statistically increased fit and showed an additional risk off HHV-8 seroconversion of 1.64 (95% CI: 1.00-2.67; P = 0.049)) for Kaposi's sarcoma. N o statistically significant in-creasedd risk was found for HHV-8 positivity and the devel-opmentt of opportunistic infection (226 events) or lym-phomaa (24 events) (relative risk: 1.24 (95% CI: 0.95-1.62; P == 0.11) or 1.31 (95% CI: 0.58-3.57; P = 0.52), respectively).
alterr the prevalence and incidence data and decrease the widthh of the confidence intervals of these estimations. Inn our cohorts, the low incidence of Kaposi's sarcoma in drugg users was paralleled by a low prevalence of HHV-8 seropositivityy and, conversely, the high incidence of Kaposi'ss sarcoma in homosexual men was paralleled by a highh prevalence of HHV-8 seropositivity. This accords with previouss reports (5-12) and is generally taken as evidence forr a sexual mode of HHV-8 transmission, as was con-firmedfirmed recently (24). However, the uneven distribution of HHV-88 antibodies among HIV transmission groups can be consideredd either as evidence for a delay in the start of the HHV-88 epidemic or a difference in transmission efficiency. HHV-88 is clearly transmitted in the absence of HIV, and the increasee of HHV-8 antibody prevalence in the cohort of drugg users over the study period shows a similar rate of spread,, with or without HIV infection, whereas among ho-mosexuall men, seropositivity for HIV tends to increase the riskk of acquiring HHV-8. The increase of HHV-8 seroprevalencee in both of our cohorts, despite different sex-uall and drug use behaviours, suggests that HHV-8 is prefer-entiallyy but not exclusively a sexually transmissible agent in thesee populations. The initial HHV-8 seroprevalence amongg the drug users approximates that of a population
withh a low risk of Kaposi's sarcoma development, and ex-tendss reports that HHV-8 prevalence is relatively low in the generall population of northern and western European countries,, like the Netherlands (11), although HHV-8 has becomee much more common in the general population of southernn Europe (7,11,13,14).
AA strong association between the presence of antibodies to HHV-88 ORF65 a n d / o r ORF73 antigens and the develop-mentt of Kaposi's sarcoma among HIV-1-infected individu-alss (odds ratio 8.9; 95% CI: 4.3-18.3) was confirmed. Cox analvsiss shows that the relative risk of development of Kaposi'ss sarcoma between HHV-8 seropositive and seronegativee individuals is 3.2 in the context of an HIV-1 infection. .
Participantss who were infected by HHV-8 after HIV-1 ap-pearedd to develop Kaposi's sarcoma sooner than those in-fectedd by HIV after HHV-8. The possibility that the HHV-88 seroprevalent group contains people who seroconvertedd for HHV-8 after HIV infection before being enrolledd has been considered. Because risk estimates are basedd on the comparison between two groups, the signifi-cancee of our findings mav be underestimated. This observa-tionn was shown to be true even after adjusting for CD4+ cell countt and HIV RNA level, although the CD4+ cell count remainedd an independent determinant of the development off Kaposi's sarcoma. These data indicate that the develop-mentt of Kaposi's sarcoma requires both a change in the im-munee svstem associated with a decline in CD4+ cell counts andd the presence of HHV-8. This mav also indicate that specificc immunity to HHV-8 is able to prevent HHV-8 in-ductionn of tumour development.
Ann adjustment was made for the use of antiretroviral ther-apyy by entering the date when these medications became availablee in the time-dependent Cox model (data not
shown).. The availability of antiretroviral therapy was associ-atedd with a 68% risk reduction of Kaposi's sarcoma, how-everr it was independent of the HHV-8-associated risk ra-tios. .
Inn HIV-1 transmission groups other than homosexual men (i.e.. heterosexually infected women, injecting drug users, patientss with haemophilia), the proportion of AIDS pa-tientss with Kaposi's sarcoma is no higher in HHV-8 en-demicc countries (16), where HHV-8 may be transmitted in childhoodd (13) (S. Mayama et ah, in preparation) than in countriess where HHV-8 infection is uncommon. Because HHV-88 is sexually transmitted among homosexual men (24,25),, our study participants who were dually infected at enrolmentt probably contracted both infections sexually. HHV-88 seroconverters also acquired HHV-8 sexually but hadd a higher risk of Kaposi's sarcoma, suggesting that the timingg of HHV-8 infection relative to HIV-1 infection ratherr than the route of HHV-8 transmission (sexual among homosexuall men, other horizontal routes in endemic coun-triess (13) (S. Mayama et al., in preparation)) determines the ratee of subsequent progression to Kaposi's sarcoma. It may bee postulated that after HIV infection, specific immunity to HHV-88 develops poorly or not at all, with unhampered HHV-88 expression and spread as a consequence.
Inn HHV-8 infection on top of a pre-existing HIV-1 infec-tion,, HIV-1 may directly facilitate the spread of HHV-8 to endotheliall cells or its initial lytic replication, even before thee establishment of a persistent HHV-8 infection in endo-theliall or spindle cells of the early Kaposi's sarcoma lesions (2,18.26,27)).. In vitro studies show that vascular endothe-liumm is altered during HIV-1 infection (28) and that produc-tivelyy infected macrophages may transport HHV-8 to the sitee of Kaposi's sarcoma lesions (29). Other lines of in vitro evidencee suggest that inflammatory cytokines released dur-ingg HIV-1 infection, including gamma interferon or HIV-1
Tat,, may promote HHV-8 replication or contribute to angiogenesiss (30,31). Our findings suggest that one or an-otherr of these in vitro mechanisms may operate in vivo. Moreover,, the rapid response, at least in some cases, of Kaposi'ss sarcoma lesions to antiretroviral therapy (32), indi-catess that HIV-1 infection is a cofactor in addition to caus-ingg immunosuppression.
Ourr finding that seroconversion to HHV-8 in an HIV-1-in-fectedd individual is strongly associated with progression to
11 Chang Y, Cesarman E, Pessin MS, et al. Identification of herpesvirus-likee D N A sequences in AIDS-associated Kaposi's sar-coma.. Science 1994;266:1865-9.
22 Boshoff C, Schulz T F , Kennedy MM, et al. Kaposi's sarcoma-associ-atedd herpesvirus infects endothelial and spindle cells. Nature Med 1995;1:1274-8. .
33 Whitby D, Howard MR, Tenant-Flowers M, et al. Detection of Kaposi'ss sarcoma associated herpesvirus (KSHV) in peripheral bloodd of HIV-infected individuals and progression to Kaposi's sar-coma,, lancet 1995;346:799-802.
44 Beral V, Peterman TA, Berkelman RL, jaffe HW. Kaposi's sarcoma amongg persons with AIDS: a sexually transmitted infection? lancet 1990;335:123-8. .
55 Lennette P,L, Blackbourn DJ, Levy JA. Antibodies to human herpesviruss 8 in the general population and in Kaposi's sarcoma pa-tients.. I Mncet 1996;348:858-61.
66 Davis DA, Humphrey RW, Newcomb FM, et al. Detection of serum antibodiess to Kaposi's sarcoma-associated herpesvirus-specific pep-t i d e . // Infecpep-t Dis 1997;175:1071-9.
77 Gao SJ, Kingsley L, Li M, et al. KSHV antibodies among Americans, Italianss and Ugandans with and without Kaposi's sarcoma. Nat Med 1996;2:925-8. .
88 Kedes D H , Operskalski E, Busch M, et al. The seroepidemiology of humann herpesvirus 8 (Kaposi's sarcoma-associated herpesvirus): distributionn of infection in Kaposi's sarcoma risk groups and evi-dencee for sexual transmission. Nat Med 1996;2:918-24.
AIDS-relatedd Kaposi's sarcoma suggests that serological testingg in HHV-8 should be used in the management of thesee patients, to decide on the use of herpesviral inhibitors knownn to reduce the development of AIDS-related Kaposi'ss sarcoma (33,34).
99 Miller G, Rigsby MO, Heston L, et al. Antibodies to butyrate-induc-iblee antigens of Kaposi's sarcoma-associated herpesvirus in patients withh HIV-1 infection. N Engl] Med 1996;334:1292-7.
100 Moore PS, Gao SJ, Dominguez G, etal. Primary characterisation of a herpesviruss agent associated with Kaposi's sarcoma. J Virol 1996;70:549-58. .
11 ] Simpson GR, Schulz TF, Whitby D , etal. Prevalence of Kaposi's sar-comaa associated herpesvirus infection measured by antibodies to re-combinantt capsid protein and latent immunofluorescence antigen.
IjmcetIjmcet 1996;349:1133-8.
122 Gao SJ, Kingsley L, Hoover DR, et al. Seroconversion to antibodies againstt Kaposi's sarcoma-associated herpesvirus-related latent nu-clearr antigens before the development of Kaposi's sarcoma. N tinglj
MedMed 1996;335:233-41.
133 Calabró ML, Sheldon J, Favero A, etal. Seroprevalence of Kaposi's sarcoma-associatedd herpesvirus (KSHV/HHV8) in different re-gionss of Italy. J Hum I 'irol 1998;1:207-l 3.
144 Whitby D, Luppi M, Barozzi P, Boshoff C, Weiss RA, Torelli G. Hu-mann herpesvirus 8 seroprevalence in blood donors and patients with lymphomaa from different regions of Italy, ƒ Natl Cancer Inst 1998;90:395-7. .
155 Ariyoshi K, Van der Loeff MS, Corrah T, et al. Kaposi's sarcoma and humann herpesvirus 8 (HHV8) in HIV-1 and HIV-2 infection in the G a m b i a . // Hum I Irol 1998;1:193-9.
166 I ARC Monographs on the evaluation of carcinogenic risks to hu-mans.. Vol 67: Human Immunodeficiency Viruses. 1996. 1ARC Press;; Lyon, France.
177 Bavlev AC, Chiengsong-Popov R, Dalgleish AG, etal. HTLV-I1I se-rology'' distinguishes atypical and endemic Kaposi's sarcoma in Af-rica.rica. iMticet 1985;1:359-61.
188 Rainbow I., Piatt GM, Simpson GR, et al. The 222-to 234 kilodalton latentt nuclear protein (LNA) of Kaposi's sarcoma-associated herpesviruss (human herpesvirus 8) is encoded bv ORF73 and is a componentt of latency-associated nuclear antigen. J I rirol 1997,71:5915-21. .
199 de Wolf F, Lange JMA, Houweling JTM, et al. Numbers of C D 4 + cellss and the level of core antigens of and antibodies to the human immunodeficiencyy virus as predictors of AIDS among seropositive homosexuall men. ƒ Infect Dis 1988;158:615-22.
200 van den Hoek JAR, Coutinho RA, van Haastrecht HJA, van Zadeihofff AW, Goudsmit J. Prevalence and risk factors of HIV in-fectionss among drug users and drug-using prostitutes in Amster-dam.. AIDS 1988;2:55-60.
211 Anonymous. 1993 Revised classification system for HIV infection andd expanded surveillance case definition for AIDS among adoles-centss and adults. MMWK 1992;41:56-8.
222 Moore PS, KJngsley FA, Holmberg SD, et al Kaposi's sarcoma-asso-ciatedd herpesvirus infection prior to onset of Kaposi's sarcoma.
AIDSAIDS 1996,10:175-80.
233 Lefrere JJ, Meyohas MC, Manotti M, Meynard JL, Thauvin M, FF rottier j . Detection of human herpesvirus 8 D N A sequences be-foree the appearance of Kaposi's sarcoma in human immunodefi-ciencyy virus (HIV)-positive subjects with a known date of HIV seroconversion.. / Infect Dis 1996;174:283-7.
244 Martin ]N, G a n e m DF,, O s m o n d D H , Page-Shafer KA, Macrae D , Kedess D H . Sexual transmission and the natural history of human herpesviruss 8 infection. iY Yingl ] Med 1998;338:948-54.
255 Melbye M, Cook PM, Hjalgrim H , et al. Transmission of human herpesviruss 8 (HHV8) among homosexual men follows the pattern off a 'Kaposi's sarcoma agent'. Int] Cancer 1998;12:543-8.
266 Staskus KA, Z h o n g W, Gebhard K, et al. Kaposi's sarcoma-associ-atedd herpesvirus gene expression in endothelial (spindle) tumor cells,, ƒ Virol 1997;71:715-9.
277 Foreman KE, Bacon PF., His E D , Nickoloff BJ. In situ polymerase chainn reaction-based localisation studies support role of human herpesviruss 8 as the cause of two AIDS-related neoplasms: Kaposi's sarcomaa and body cavity l y m p h o m a . / Clin Invert 1997;99:2971-8. 288 Zietz C, Hotz B, Sturzl M, etal. Aortic endothelium in HIV-1
infec-tion:: chronic injun', activation and increased leukocyte adherence.
Am]Am] Pathol 1996;149:1887-98.
299 Blasig C, Zietz C, Haar B, et al. Monocytes in Kaposi's sarcoma le-sionss are productively infected by human herpesvirus 8. J I 'irol 1997;10:7963-8. .
300 Knsoli B, Gendelman R, Markham P, et al. Synergy between basic fibroblastt growth factor and human immunodeficiency virus type 1 Tatt protein in induction of Kaposi's sarcoma. Nature 1994;371:674-80. .
311 Harrington W, Sieczkowski L, Sosa C, etal. Activation of HHV8 by HIV-11 Tat. lancet 1997;349:774.
322 Blum L, Pellet C, Agbalika F, et al. Complete remission of AIDS-re-latedd Kaposi's sarcoma associated with undetectable human herpesviruss 8 sequences during anti-HIV protease therapy. AIDS 1997;11:1653-5. .
333 Mocroft A, Youle M, Gazzard B, et al. Anti-herpesvirus treatment andd risk of Kaposi's sarcoma in HIV infection. Royal Free/Chelsea andd Westminster Hospitals Collaborative Group. AIDS 1996;10:1101-5. .
344 Glesby MJ, Hoover DR, Weng S, et al. F'se of antiherpes drugs and thee risk of Kaposi's sarcoma: data from the Multicenter AIDS Co-hortt Study./ Infect Dis 1996,173:1477-80.
