Introduction
Kaposi sarcoma is a common opportunistic disease associated with AIDS and is responsible for severe morbidity and mortality [1,2].
Since the introduction of highly active antiretroviral therapy (HAART), the incidence of Kaposi sarcoma has decreased [3-5] and in some cases, established Kaposi sarcoma goes into remission. Incident cases of Kaposi sarcoma are prevented when treatment is started early, before advanced immune suppression. Immune reconstitution increases the level of the antibodies directed against the viral K8.1 lytic protein and restores the CD8 T-cell response against human herpesvirus 8 (HHV8) epitopes [6]. No direct effect of protease inhibitors on HHV8 has ever been demonstrated [7,8], but indinavir and saquinavir reduce the number and size of Kaposi sarcoma-like angioproliferative lesions in nude mice and inhibit bFGF or VEGF-induced formation of new vessels [7].
Nevertheless, the response of established Kaposi sarcoma to HAART is unpredictable. While many patients do well, others will have progressive disease and will need chemotherapy, which produces only short remissions. Up to 30% of patients still die following Kaposi sarcoma progression [9]. It is therefore important to define risk factors of unfavorable evolution, so as to select high-risk patients who may benefit from more aggressive treatment, while avoiding unnecessary therapy and toxicity in those who are at low risk of progression.
Few studies have tried to identify such prognostic factors and their conclusions are contradictory [9-12]. One study concluded that sex and age, CD4 cell count, level of HIV viremia and HAART-related factors such as whether patients were HAART-naive or HAART-experienced at the time of diagnosis of Kaposi sarcoma, as well as differences in the types of HAART used, did not provide prognostic information [9]. These studies [9-12] did not include information concerning HHV8, which may well be relevant as the inability to detect HHV8 in peripheral blood mononuclear cell (PBMC) predicts clinical response [13]. Specific antibodies and CD8 T-cell responses increase during HAART [14], but their prognostic value has not yet been studied. The type of HAART used may play a role, as suggested by a series of patients with flares of Kaposi sarcoma after the switch from a protease inhibitor to a nonnucleoside reverse transcriptase inhibitor (NNRTI)-based regimen [15].
We describe the incidence and prevalence of Kaposi sarcoma, in the Swiss HIV Cohort Study (SHCS) since 1996 and establish a predictive score of evolution of Kaposi sarcoma, after having assessed prognostic factors among socio-demographic variables, immunovirological factors concerning HIV and HHV8, staging and differences in the type of HAART regimens used.
Methods
Study population, covariates and endpoints
We included patients from the SHCS with a histologically confirmed or presumptive clinical diagnosis of Kaposi sarcoma made between January 1, 1996 and December 31, 2004. When available, frozen plasma samples were collected, at the time of diagnosis and at the end of the follow-up, to quantify HHV8 DNA and anti-HHV8 antibodies. Study endpoints included the initiation of chemotherapy to treat Kaposi sarcoma or death if no chemotherapy was administered before. Patients who fulfilled one or both these endpoints were considered part of the unfavorable outcome group. Potential prognostic factors were assessed among the following variables: sex, age, race, origin, sexual orientation, and intravenous drug use; tumor staging; development of multicentric Castelman's disease (MCD) or body cavity lymphoma (BCL); exposure to antiviral treatment active against herpes viruses; type of antiretroviral therapy (ART), CD4 levels, HIV viremia, quantitative HHV8 DNA, latent and lytic HHV8 antibody titers, at Kaposi sarcoma diagnosis and at the end of the follow-up.
Definitions
According to the AIDS Clinical Trials Group (ACTG) classification [16] tumor stage was classified as T0: Kaposi sarcoma confined to the skin and lymph nodes and minimal oral Kaposi sarcoma; and T1: ulcerated Kaposi sarcoma, Kaposi sarcoma-associated oedema, nodular oral Kaposi sarcoma or Kaposi sarcoma involvement of any visceral organ.
ART was defined as the use of any antiretroviral drug, either as monotherapy, dual therapy, or triple therapy (HAART); HAART as the use of three nucleoside reverse transcriptase inhibitors (NRTIs), or two NRTIs with either a protease inhibitor or a NNRTI, or four antivirals.
The exposure to antiviral treatment active against herpes, during the follow-up, included (val)ganciclovir, (val)aciclovir, adefovir, cidofovir or foscarnet.
Detection and quantification of human herpesvirus 8 DNA
HHV8 viral load in the plasma was evaluated by qualitative and quantitative PCR. DNA was extracted using an EZ1-DNA extraction robot (Qiagen, Hombrechtikon, Switzerland) according to the manufacturer's instruction. For qualitative evaluation, a nested PCR for the HHV8 ORF 26 region was amplified as previously described [17]. For quantitative PCR, HHV8 DNA was amplified by the Sequence Detection System 7500 Fast (TaqMan) (Applied Biosystems, Rotkreuz, Switzerland) amplifying a 147 bp PCR product of ORF 26 using the following primers and probe: Forward primer KS3 5′-AGC CGA AAG GAT TCC ACC AT-3′; Reverse primer KS4 5′-TCC GTG TTG TCT ACG TCC AG-3′; KSProbe 5′-FAM-CGC TAT TCT GCA GCA GCT GTT GGT GTA CCA-TAMRA-3′. TaqMan standard conditions were applied with 40 cycles of denaturation at 95°C for 3 s and primer annealing/extension at 60°C for 30 s. All reactions were performed in duplicate. For the calculation of molecule quantity, serial dilutions of ORF 26 containing HHV8 plasmid DNA were analyzed and a standard curve established as previously described [18]. Results were expressed in copies/ml.
HHV8 serology
To determine the presence and titer of antibodies to the latent nuclear antigen (LANA) and lytic antigens at Kaposi sarcoma diagnosis and at outcome, we used CE-marked immunofluorescence assays (Panbio, Inc., Maryland, USA). Briefly, LANA-antibodies were identified in the nucleus of uninduced HHV8-infected body cavity-based lymphoma-1 (BCBL-1) cells. In induced cells, (with 12-0-tetradecanoylphorbol-13-acetate), we detected anti-HHV8 lytic antibodies, in the nucleus and cytoplasma.
Results were given as titers, ranging from 1: 64 to 1: 32 768. For practical reasons, an arbitrary dilution unit was defined: 1 unit is equivalent to the first titer of 1: 64, 2 units 1: 128, up to 10 units representing 1: 32 768.
Statistical analysis
Patients were followed from the time of Kaposi sarcoma diagnosis to the date of the first day of chemotherapy or the date of their death. Patients who did not reach an endpoint were censored on December 31, 2004 or at their last cohort visit, whichever occurred earlier. We used the Kaplan-Meier method to estimate the survival function and the log-rank test to compare the differences between the subgroups. Differences between subgroups were also tested in a univariate analysis using the Cox proportional hazards model to compute the hazard ratio (HR) and corresponding 95% confidence interval (CI). Covariates that were found significant in the univariate analysis and those found to be biologically plausible, this is to say age and sexual orientation, were entered in a multivariate model, using a forward stepwise procedure. To establish the predictive score we used the variables, present at the time of Kaposi sarcoma diagnosis, which were found significant in the multivariate analysis. We attributed the number of points to each according to their coefficient 'B' in the Cox model. The score ranged from 0 to 4. Poisson regression was also used to explore the changes in Kaposi sarcoma incidence rate from 1996 to 2004.
In all cases, statistical significance was claimed for P ≤ 0.05 (two-sided). Statistical analyses were performed using SPSS 11.0 (SPSS, Chicago, Illinois, USA) and Stata 8.0 softwares (STATA Corp., College Station, Texas, USA).
Results
Two hundred and forty-three patients developed Kaposi sarcoma, during the study period. Two hundred and three were included in the survival analysis. Thirty-four patients were excluded because clinical and virological data on HIV were lacking and six others because of unknown outcome.
The median follow-up was 30 months (range 0-108).
Prevalence and incidence
The prevalence of Kaposi sarcoma ranged from 4.5% in 1996 to 3.8% in 2003. The incidence rate ranged from its highest value of 3.5 per 1000 person-years in 1996 to its lowest value of 0.18 per 1000 person-years in 2003 (Table 1). There was an estimated reduction of 32.18% in the incidence of Kaposi sarcoma with each successive calendar year (95% CI -37.9 to -26.4%; P < 0.0001). This decrease was most important between 1996 and 1999 when it was estimated to be of -53.0% (95% CI -68.8 to -37.2%; P < 0.0001) (Fig. 1).
Patient characteristics
Patient characteristics are summarized in Table 2.
The median age at Kaposi sarcoma diagnosis was 41 years, ranging from 19 to 65; the majority were white men having sex with men (MSM). The overall mortality rate was 21%, and 27% of patients started chemotherapy. Fifty percent of these endpoints occurred during the first 3 months after diagnosis. Eighteen (33%) patients who were started on chemotherapy ultimately died.
Seven patients developed MCD and two BCL. Eight belonged to the unfavorable outcome group.
Regarding Kaposi sarcoma, 78% of the patients had a T0 disease. The median CD4 cell count at baseline was 112 cells/μl (range 0-1267). At the end of the follow-up, the median CD4 cell count increased to 284 (92-516). The median HIV viral load was 76 710 copies/ml (<50-2 385 000); by the end of the follow-up, it was 57 (<50-19 117). Thirty-nine percent of patients were using ART at the time of Kaposi sarcoma diagnosis and by the end of the follow-up, this proportion had increased to 76%. When ART was started after Kaposi sarcoma diagnosis, the median time to introduction was 42 days (1-1282). During the follow-up, 28% of patients used an antiviral drug active against herpes viruses, mainly valaciclovir.
Plasma samples for quantification of HHV8 DNA, at baseline, were available for 147 patients (72%, see Table 2). Thirty-eight (26%) of these presented a positive HHV8 viral load at baseline. The mean value was 803 copies/ml (SD 3468).
At the end of the follow-up, 107 plasma samples (53%) were available for HHV8 quantification. Twenty-three (22%) were positive, the mean value had increased to 3348 copies/ml (SD 22 728).
Baseline latent and lytic titers were analyzed in 136 patients (67%) (Table 2). One hundred and eleven (82%) and 126 (93%) of these had, respectively, a positive latent and lytic value. The median dilution unit was 3 (titer 1: 256) for the latent and 4 (titer 1: 512) for the lytic.
At the end of the follow-up, 108 samples (53%) were analyzed. Ninety-four (88%) and 104 (96%) had positive latent and lytic titers, respectively. The median titers increased in both (4 and 5, respectively) and the lytic values remained higher than the latent.
Dual reactivity was found in 78% of cases at Kaposi sarcoma and 86% at the end of the follow-up.
Prognostic factors
Univariate analysis
Table 3 summarizes the hazard ratios associated with each factor.
We found that evolution toward death or the necessity of chemotherapy was significantly influenced, at baseline, by tumor staging, CD4 cell count, HHV8 DNA detection and finally HHV8 lytic titers. Patients who had a T1 disease were four times more likely to present an unfavorable outcome (HR 4.23; 95% CI 2.68-6.68). Those presenting CD4 cell counts less than 200 cells/μl were at increased risk of unfavorable outcome (1.74; 1-2.95). The detection of HHV8 DNA was strongly associated with an unfavorable outcome (2.25; 1.30-3.91) and a high number of copies (>3000) even more so (7.17; 3.15-16.31). The absence of lytic antibodies was associated with a two-fold increase in risk of unfavorable outcome (2.34; 0.99-5.5) and each increase in the dilution unit reduced that risk (0.84; 0.74-0.96).
Age, sex, origin, intravenous drug use, HIV viral load, ART and latent antibody titers were not associated with a significant change in outcome.
Taking into account the evolution of certain factors during the follow-up, we noted that the risk of unfavorable outcome was significantly increased by the development of MCD or BCL (3.66; 1.59-8.47); a CD4 cell count below 200 cells/μl at the end of the follow-up (9.26; 5.47-15.7); less than a 50 CD4 cell increase from baseline (2.42; 1.37-4.27); HIV viral load values ≥5000 copies/ml at end of the follow-up (4.15; 2.62-6.57), becoming or remaining HHV8 DNA detectable (9.68; 4.01-23.36) and the absence of lytic antibodies at the end of the follow-up (19.87; 6.36-62.08). Patients without ART, at the end of the follow-up, were two times more likely to present an unfavorable outcome (2.37; 1.49-3.77). The use of a protease inhibitor was associated with an increased risk of unfavorable outcome (1.70; 1.09-2.65), as was the use of antiviral drugs active against herpes virus (2.01; 1.28-3.15).
Multivariate analysis and score
The covariates of 144 patients were entered into the final multivariate model. At baseline, the following factors predicted an unfavorable outcome: Kaposi sarcoma stage T1 (HR 5.22; 95% CI 2.97-9.18), CD4 cell count below 200 cells/μl (HR 2.33; 95% CI 1.22-4.45) and positive HHV8 DNA (HR 2.14; 95% CI 1.18-3.89) (Table 4).
To establish our predictive score of unfavorable outcome, we scaled these four factors according to their B coefficient, giving T1 two points; CD4 cell count below 200 cells/μl and a positive HHV8 DNA, one point each. The score ranged from 0 to 4. The receiver operating characteristic (ROC) curve has an area under the curve (AUC) of 0.77 (95% CI 0.68-0.88). Kaplan-Meier curves showing survival according to score are shown in Fig. 2. Each incremental increase in the score was associated with a HR of 2.26 (95% CI 1.79-2.85).
Discussion
The objective of our study was to establish predictive factors of death and the need of chemotherapy, among easily available clinical data and laboratory values, at the time of Kaposi sarcoma diagnosis, in HIV patients. We identified three factors which are T1 staging, a CD4 cell count below 200 cells/μl and a positive plasmatic HHV8 DNA value and developed a predictive score of unfavorable Kaposi sarcoma evolution in order to identify the patients who may need more aggressive treatment of Kaposi sarcoma.
Our patient characteristics are representative of the Kaposi sarcoma population encountered in occidental HIV clinics: a majority of white homosexual men, of northern European origin with no HIV treatment at the time of Kaposi sarcoma diagnosis [3].
We observed a decreasing prevalence from 4.5 to 3.8% and a reduction of 32% in the incidence of Kaposi sarcoma throughout the study period, with a faster decrease during the early period of the HAART era, from 1996 to 1999. These results are consistent with the ones described in the EuroSIDA Study [3] and studies from Australia [4] and the USA [5]. Several explanations can be put forward such as the detection of HIV seropositivity at an earlier stage before AIDS and the introduction of HAART, which prevent immunodeficiency. The question of a lower seroprevalence of HHV8 remains to be clarified. Nevertheless, the morbidity and mortality of Kaposi sarcoma remain high. In our study, more than a quarter of the patients suffered from invalidating Kaposi sarcoma, necessitating the use of chemotherapy and 21% died during the months following the diagnosis of Kaposi sarcoma, confirming the result of 25% previously published by Nasti et al. [9].
Our multivariate analysis revealed that T1 staging was associated with a five-fold increase in risk of unfavorable outcome, which is the highest among the different variables and that a CD4 level below 200 cells/μl and a positive plasmatic HHV8 DNA increased the risk two-fold. Similar results on the importance of tumor extension and immune status at the time of diagnosis have been reported by Krown et al. [12], whereas Nasti et al. [9] refute the predictive value of the CD4 level. We establish the importance of HHV8 DNA quantification as a predictive factor of bad evolution: not only is the detectability of the DNA a negative factor, but the higher the plasmatic load, the worse the prognosis. We chose to detect HHV8 DNA in the plasma, as it may reflect active lytic replication. Our whole plasma analysis, however, cannot distinguish viremia from free viral DNA, released from blood cells, during preparation. Plasma or serum detection of HHV8 is known to be rarer than PBMC detection [6,19]. In our study, 26% of patients were positive at the time of diagnosis, less than the previously published (41%) results, [19], but our plasma samples could have been collected earlier, during the course of Kaposi sarcoma. Few studies have tackled the importance of HHV8 DNA detection in Kaposi sarcoma, and our findings confirm these: in a study on 21 HIV patients with Kaposi sarcoma patients, Gill et al. [14] showed that nine of the 10 patients whose Kaposi sarcoma responded to HAART had undetectable PBMC Kaposi's sarcoma-associated herpesvirus (KSHV) load. Campbell et al. [20] observed an association between the extent of Kaposi sarcoma disease and the levels of HHV8 DNA in PBMC. Pellet et al. [13], in a study on 26 patients, suggested that an undetectable PBMC KSHV viremia was independently associated with the absence of need of chemotherapy. In non-HIV patients, the likelihood of developing classic Kaposi sarcoma was elevated with detectable PBMC KSHV DNA [21].
Although not relevant to our score, several other factors were significantly associated with an unfavorable outcome, in the univariate analysis.
Nine patients (4%) developed another HHV8-associated illness, (MCD or BCL); this was associated with a 3.5-fold increase in risk of unfavorable outcome.
We observed a 2.3-fold increase in risk associated with a negative lytic antibody titer to HHV8, at the time of Kaposi sarcoma diagnosis and each increase in dilution was associated with a significant improvement in outcome. There was only a trend toward the same effect with the latent antibody titers. This is hardly explained by the sensitivity of our test, which yields similar or even higher proportions than those observed in previously published works (96% positive for lytic and 72 or 52% positive for latent, in HIV-positive patients with Kaposi sarcoma) [22,23]. This difference between lytic and latent may suggest that the antilytic KSHV antibodies might be more effective in controlling KSHV pathogenesis than the anti-LANA antibodies, by reducing lytic viral replication [6]. Although several studies have failed to show a predictive value of KSHV seropositivity for the subsequent development of Kaposi sarcoma [22,24], the importance of the immunological response to HHV8 is suggested by the fact that patients infected with HHV8 after HIV carry a higher risk of developing Kaposi sarcoma than the ones already infected at the time of HIV seroconversion [25]. This probably reflects the incapacity of the immunosuppressed patients to mount a valuable serological response. HAART is followed by recovery of the immune response to KSHV [6]. These different findings suggest that even if neutralizing antibodies to HHV8 does not play the primary role in the protection against Kaposi sarcoma, the fact of not mounting a response at all is unfavorable.
Consistent with the previous studies [9,13,26], age, sex, sexual orientation and origin did not influence the outcome. Contrary to Stebbing et al. [26], age was not selected in our score, possibly because our outcome was not only survival, but also the need for chemotherapy. Although we intentionally chose to study the HAART era, only 39% of patients were using ART, at Kaposi sarcoma diagnosis. This is higher than previously published data (15% for Tam et al. [11]; 18% for Nasti et al. [9], 22% for Stebbing et al. [26]) and can probably be explained by the fact that the detection of HIV seropositivity is often made at the same time as the diagnosis of Kaposi sarcoma. Like Nasti et al. [9] and Stebbing et al. [26], we did not find any significant relationship between the use of ART/HAART at Kaposi sarcoma diagnosis and an unfavorable evolution.
We followed the evolution of several factors after Kaposi sarcoma diagnosis. Bearing in mind the study's timeframe limitation, the fact that some patients were not using HAART at the end of their follow-up, enabled us to confirm that starting HAART is a factor of favorable outcome. Several papers have suggested that the use of a protease inhibitors, as opposed to a NNRTI, could have an antiangioproliferative effect and therefore influence Kaposi sarcoma's evolution [7,8]. Our study does not support this assertion. It has been suggested that antiherpes treatments could have an impact on Kaposi sarcoma evolution through a direct effect on HHV8 [27,28]. In our study, however, their use was associated with an unfavorable outcome, suggesting these patients were probably sicker and that these treatments do not appear to help in the management of Kaposi sarcoma. The evolution of the HHV8 serologies and DNA quantifications during the follow-up is also important. The absence of antibodies toward the HHV8 lytic antigen, at the end of the follow-up, was associated with a bad outcome and there is a similar trend for the latent antigen. The positivity of the HHV8 DNA, at the end of follow-up, is also associated with a bad outcome, as is the fact of remaining positive or becoming positive during the study period. Higher levels of HHV8 DNA increase this risk.
Our study presents several limitations: the bias linked with the retrospective analysis of prospectively collected data; the subjectivity associated with the use of chemotherapy as one of our endpoints and all cause mortality for the other. We also chose not to study HHV8-specific T-cell responses, as, for the moment, the technique is not standardized, but it could turn out to be an interesting parameter as Guihot et al. [29] have demonstrated a lower specific response in patients with Kaposi sarcoma, than in asymptomatic HHV8 carriers.
Therefore, although the AUC of our ROC curve is acceptable, our score needs to be confirmed in a different cohort of HIV patients.
In conclusion, few studies have examined the various factors associated with evolution of HIV-associated Kaposi sarcoma and only one has included data on HHV8. We establish that a positive HHV8 DNA is an independent risk factor of bad evolution of Kaposi sarcoma, as is a CD4 level of below 200 cells/μl and the T1 staging. Our score is simple to use at the patient's bedside. Each point increase in the scale is associated with a two-fold increase in the risk of death or need for chemotherapy. We suggest that patients with an index of two or more have a poor risk and should be followed more closely and chemotherapy should be considered, whereas patients with a lower score could be treated with HAART alone.
Acknowledgements
This study was mainly financed by a grant from the Swiss HIV Cohort Study and the Swiss Cancer Ligue. The project received The Roche HIV fellowship 2005.
The authors would like to address many thanks to Professor P. Y. Dietrich for his support of the project and to Dr T. Perneger and Dr M. Boulvain for many fruitful discussions on statistics.
References
1. Cathomas G. Kaposi's sarcoma-associated herpesvirus (KSHV)/human herpesvirus 8 (HHV-8) as a tumour virus. Herpes 2003; 10:72-77.
2. Chang Y, Cesarman E, Pessin MS, Lee F, Culpepper J, Knowles DM, et al. Identification of herpesvirus-like DNA sequences in AIDS-associated Kaposi's sarcoma. Science 1994; 266:1865-1869.
3. Mocroft A, Kirk O, Clumeck N, Gargalianos-Kakolyris P, Trocha H, Chentsova N, et al. The changing pattern of Kaposi sarcoma in patients with HIV, 1994-2003. Cancer 2004; 100:2644-2654.
4. Grulich AE, Yueming L, McDonald AM, Correll PK, Law MG, Kaldor JM. Decreasing rates of Kaposi's sarcoma and non-Hodgkin's lymphoma in the era of potent combination antiretroviral therapy. AIDS 2001; 15:629-633.
5. Jones JL, Hanson DL, Dworkin MS, Jaffe HW, and the Adult/Adolescent Spectrum of Disease Project Group. Incidence and trends in Kaposi's sarcoma in the era of effective antiretroviral therapy. J Acquir Immune Defic Syndr 2000; 24:270-274.
6. Bourboulia D, Aldam D, Lagos D, Allen E, Williams I, Cornforth D, et al. Short- and long-term effects of highly active antiretroviral therapy on Kaposi sarcoma-associated herpesvirus immune responses and viraemia. AIDS 2004; 18:485-493.
7. Sgadari C, Monini P, Barillari G, Ensoli B. Use of HIV protease inhibitors to block Kaposi's sarcoma and tumour growth. Lancet Oncol 2003; 4:537-547.
8. Sgadari C, Barillari G, Toschi E, Carlei D, Bacigalupo I, Baccarini S, et al. HIV protease inhibitors are potent antiangiogenic molecules and promote regression of Kaposi sarcoma. Nat Med 2002; 8:225-232.
9. Nasti G, Talamini R, Antinori A, Martellotta F, Jacchetti G, Chiodo F, et al. AIDS-related Kaposi's sarcoma: evaluation of potential new prognostic factors and assessment of the AIDS Clinical Trial Group Staging System in the HAART Era - the Italian Cooperative Group on AIDS and Tumors and the Italian Cohort of Patients Naive From Antiretrovirals. J Clin Oncol 2003; 21:2876-2882.
10. Nasti G, Martellotta F, Berretta M, Mena M, Fasan M, Di Perri G, et al. Impact of highly active antiretroviral therapy on the presenting features and outcome of patients with Acquired Immunodefiency Syndrome-related Kaposi. Cancer 2003; 98:2440-2446.
11. Tam HK, Zhang ZF, Jacobson LP, Margolick JB, Chmiel JS, Rinaldo C, et al. Effect of highly active antiretroviral therapy on survival among HIV-infected men with Kaposi sarcoma or non-Hodgkin lymphoma. Int J Cancer 2002; 98:916-922.
12. Krown SE, Testa MA, Huang J. AIDS-related Kaposi's sarcoma: prospective validation of the AIDS Clinical Trials Group staging classification. AIDS Clinical Trials Group Oncology Committee. J Clin Oncol 1997; 15:3085-3092.
13. Pellet C, Chevret S, Blum L, Gauville C, Hurault M, Blanchard G, et al. Virologic and immunologic parameters that predict clinical response of AIDS-associated Kaposi's sarcoma to highly active antiretroviral therapy. J Invest Dermatol 2001; 117:858-863.
14. Gill J, Bourboulia D, Wilkinson J, Hayes P, Cope A, Marcelin AG, et al. Prospective study of the effects of antiretroviral therapy on Kaposi sarcoma-associated herpesvirus infection in patients with and without Kaposi sarcoma. J Acquir Immune Defic Syndr 2002; 31:384-390.
15. Bani-Sadr F, Fournier S, Molina JM. Relapse of Kaposi's sarcoma in HIV-infected patients switching from a protease inhibitor to a nonnucleoside reverse transcriptase inhibitor-based highly active antiretroviral therapy regimen. AIDS 2003; 17:1580-1581.
16. Krown SE, Metroka C, Wernz JC. Kaposi's sarcoma in the acquired immune deficiency syndrome: a proposal for uniform evaluation, response, and staging criteria. AIDS Clinical Trials Group Oncology Committee. J Clin Oncol 1989; 7:1201-1207.
17. Cathomas G, Stalder A, McGandy CE, Mihatsch MJ. Distribution of human herpesvirus 8 DNA in tumorous and nontumorous tissue of patients with acquired immunodeficiency syndrome with and without Kaposi's sarcoma. Mod Pathol 1998; 11:415-420.
18. Zsikla V, Hailemariam S, Baumann M, Mund MT, Schaub N, Meier R, Cathomas G. Increased rate of Helicobacter pylori infection detected by PCR in biopsies with chronic gastritis. Am J Surg Pathol 2006; 30:242-248.
19. Polstra AM, Van den Burg R, Goudsmit J, Cornelissen M. Human herpesvirus 8 load in matched serum and plasma samples of patients with AIDS-associated Kaposi's sarcoma. J Clin Microbiol 2003; 41:5488-5491.
20. Campbell TB, Borok M, Gwanzura L, MaWhinney S, White IE, Ndemera B, et al. Relationship of human herpesvirus 8 peripheral blood virus and Kaposi's sarcoma clinical stage. AIDS 2000; 14:2282-2290.
21. Brown EE, Whitby D, Vitale F, Marshall V, Mbisa G, Gamache C, et al. Virologic, hematologic and immunologic risk factors for classic Kaposi sarcoma. Cancer 2006; 107:2109-2116.
22. Inoue N, Spira T, Lam L, Corhero JL, Luo W. Comparison of serologic responses between Kaposi' sarcoma-positive and -negative Men who were seropositive for both HHV8 and HIV. J Med Virol 2004; 74:202-206.
23. Lennette ET, Blackbourn DJ, Levy JA. Antibodies to human herpesvirus type 8 in the general population and in Kaposi's sarcoma patients. Lancet 1996; 348:858-861.
24. Quinlivan EB, Wang RX, Stewart PW, Kolmoltri C, Regamey N, Erb P, et al. Longitudinal sero-reactivity to human herpesvirus 8 (KSHV) in the Swiss HIV Cohort 4.7 years before KS. J Med Virol 2001; 64:157-166.
25. Jacobson LP, Jenkins FJ, Springer G, Munoz A, Shah KV, Phair J, et al. Interaction of human immunodeficiency virus type 1 and human herpesvirus type 8 infections on the incidence of Kaposi's sarcoma. J Infect Dis 2000; 181:1940-1949.
26. Stebbing J, Sanitt A, Nelson M, Powels T, Gazzard B, Bower M. A prognostic index for AIDS-associated Kaposi's sarcoma in the era of highly active antiretroviral therapy. Lancet 2006; 367:1495-1502.
27. Hengge UR, Ruzicka T, Tyring SK, Stuschke M, Roggendorf M, Schwartz RA, et al. Update on Kaposi's sarcoma and other HHV8 associated diseases. Part 1: epidemiology, environmental predispositions, clinical manifestations, and therapy. Lancet Infect Dis 2002; 2:281-292.
28. Hengge UR, Ruzicka T, Tyring SK, Stuschke M, Roggendorf M, Schwartz RA, et al. Update on Kaposi's sarcoma and other HHV8 associated diseases. Part 2: pathogenesis, Castleman's disease, and pleural effusion lymphoma. Lancet Infect Dis 2002; 2:344-352.
29. Guihot A, Dupin N, Marcelin AG, Gorin I, Bedin AS, Bossi P, et al. Low T cell responses to human herpesvirus 8 in patients with AIDS-related and classic Kaposi sarcoma. J Infect Dis 2006; 194:1078-1088.
© 2008 Lippincott Williams & Wilkins, Inc.