Main

Human herpesvirus-8 (HHV-8) is a recently described gamma herpesvirus, that was first identified as an etiologic agent in Kaposi sarcoma in 1994 by Chang et al,1 who isolated DNA sequences of HHV-8 from endothelial cells in Kaposi sarcoma lesions from patients with the acquired immune deficiency syndrome (AIDS). The development of antibodies to the HHV-8 latency-associated nuclear antigen is considered the hallmark of latent HHV-8 infection, and seropositivity for antibodies to Latency-associated nuclear antigen (LANA) is found in over 85% of patients with Kaposi sarcoma as compared with only 1–4% of the general population.2, 3 The presence of LANA antibodies in HIV-positive patients strongly correlates with the risk of developing Kaposi sarcoma.4

Monoclonal antibodies to LANA have recently been developed for use in formalin-fixed and frozen tissues, and have been shown to be reliable markers of latent HHV-8 infection in Kaposi sarcoma as well as other HHV-8-associated processes, such as primary effusion lymphoma and multicentric Castleman disease.5, 6 A small number of studies have shown LANA expression to be specific for Kaposi sarcoma among vascular tumors occurring in the general population.6, 7, 8, 9 Evaluation of LANA expression has thus been proposed as an ancillary technique for the discrimination of Kaposi sarcoma from other vascular lesions with which it may be confused, such as stasis-associated vascular proliferations, spindle cell hemangioma, kaposiform hemangioendothelioma, and spindled variants of angiosarcoma. It is currently unknown, however, whether non-Kaposi sarcoma vascular lesions arising in immunocompromised patients might express LANA, possibly reflecting origin from an HHV-8-positive endothelial progenitor cell, or secondary HHV-8 infection. We therefore studied a group of Kaposi sarcoma and non-Kaposi sarcoma vascular tumors arising in known HIV-positive patients, for evidence of HHV-8 coinfection using LANA immunohistochemistry (IHC) and HHV-8 real-time PCR.

Materials and methods

Samples

The Emory University and Grady Memorial Hospital Institutional Review Boards granted approval for this study. We retrospectively examined 37 formalin-fixed, paraffin-embedded samples obtained from 37 patients who had undergone resection or biopsy at Grady Memorial Hospital between 1993 and 2003. All patients were serologically positive for HIV by enzyme-linked immunoassay and Western blot. The patients ranged in age from 22 to 61 years, and included 35 male and two female patients. The non-Kaposi sarcoma lesions (n=13) included 12 capillary hemangiomas and one epithelioid hemangioendothelioma, and represented all of the available non-Kaposi sarcoma vascular tumors occurring in HIV-positive individuals in our institutions. The Kaposi sarcoma cases (n=24) ranged histologically from early patch to nodular stage. Both Kaposi sarcoma and non-Kaposi sarcoma lesions included a broad range of anatomic sites. The diagnoses for all cases were reconfirmed by an experienced soft-tissue pathologist (ALF) prior to inclusion in the study.

Immunohistochemistry

Sections, 5 μm thick, were cut and placed on charged glass slides. The slides were then immunostained with the LANA antibody (Clone 13B10, 1:50, Novocastra) using steam heat-induced epitope retrieval and the Dako Envision system according to the manufacturer's recommendations. Negative controls consisted of substitution of buffer for the primary antibody. The presence of any nuclear staining was scored as ‘positive’.

TaqMan Real-Time PCR

Two 5-μm paraffin-embedded sections were cut from each block and placed in Eppendorf tubes, dewaxed at 98°C for 15 min, and suspended in proteinase K digestion buffer for 1 h at 68°C. This was followed by DNA extraction using a QIAamp kit (Qiagen, Valencia, CA, USA). The samples were then centrifuged, and the supernatant was suspended in TE buffer (10 mM Tris-HCl, pH 8.0). The quality of the extracted DNA was confirmed by β-actin amplification. A body cavity B-cell lymphoma cell line (BCBL) was used as a positive control. Water was used as a negative control. An arbitrary fluorescence threshold value of 30 and cycles-to-threshold value of 50 was used. A value over the fluorescence threshold of 30 and within 50 cycles-to-threshold was considered a positive result.

The PCR reaction was performed on SmartCycler (Cepheid, Sunnyvale, CA, USA) at 50°C for 2 min, 95°C for 10 min. This was followed by 50 cycles at 95°C for 15 s and 60°C for 1 min. The following primers were used for HHV-8 detection: HHV8-1, 5′-AGCCGAAAGGATTCCACCATT-3′; HHV8-2, 5′-TCC GTGTTGTCTACGTCCAGA-3′. The HHV-8 TaqMan probe had the following sequence: 5F-TGCAGCAGC TGTTGGTGTACCACAT-T3′, where F=FAM and T=TAMRA. The following primers were used for β-actin PCR: 5′-TCACCCACACTGTGCCCATCTAC GA-3′; reverse, 5′-CAGCGGAACCGCTCATTGCCA ATGG-3′ and TaqMan probe: 5′F-ATGCCCTCCC CCA TGCCATCCTGCGT-3′, with the quencher TAMRA placed on base position 7 from the 5′ end.

Results

Table 1 summarizes the IHC and PCR data. LANA expression by IHC was present in 22 of 24 (92%) of the Kaposi sarcoma cases, and demonstrated a range of focal to diffuse, dot-like nuclear staining in endothelial or spindle cells (Figure 1a and b). No staining was detected in any of the 13 non-Kaposi sarcoma cases (Figures 2a, b and 3a, b).

Table 1 IHC and real-time RT-PCR results
Full size table
Figure 1
figure 1

Kaposi sarcoma (a), strongly positive for LANA by IHC (b) and for HHV-8 by real-time PCR (c).

Full size image
Figure 2
figure 2

Hemangioma (a), negative for LANA by IHC (b). This case was positive for HHV-8 by real-time PCR (c).

Full size image
Figure 3
figure 3

Epithelioid hemangioendothelioma (a), negative for LANA by IHC (b). This case was HHV-8 negative (not shown).

Full size image

Using real-time PCR, the BCBL-positive control was detectable at 25.63 cycles-to-threshold. HHV-8 DNA was detected in all of the Kaposi sarcoma lesions (Figure 1c), with an average cycles-to-threshold value of 45.48, and a range of 36.45–49.4. HHV-8 DNA was also detected in four of 13 (31%) of the non-Kaposi sarcoma lesions (Figure 2c), all of which were hemangiomas occurring in three males and one female patient. Repeat PCR of these four cases confirmed the presence of HHV-8 DNA at 44, 45, 46, and 47 cycles-to-threshold, respectively.

Discussion

Since Kaposi's seminal description in 1872 of the tumor bearing his name, the nature and etiology of Kaposi sarcoma has been the matter of significant dispute. Formerly a rare tumor, the incidence of Kaposi sarcoma exploded with the onset of the AIDS epidemic, and Kaposi sarcoma is now the most common tumor seen in AIDS patients, with a disproportionate incidence among homosexual and bisexual men.10, 11, 12 For this reason, an infectious etiology has long been suspected, and epidemiological, serologic, and molecular genetic studies over the past 10–15 years have identified a novel herpesvirus, HHV-8, as the likely causative agent of Kaposi sarcoma.10, 11, 12, 13, 14

HHV-8 is known to latently infect endothelial cells, as well as peripheral blood monocytes and B-lymphocytes in patients with Kaposi sarcoma. During latent infection, the virus exists in the nucleus as circular, episomal DNA that expresses a small number of viral genes.15 Although no significant difference in HHV-8 viral loads has been observed in biopsies from HIV-positive vs HIV-negative individuals with Kaposi sarcoma,16 HIV-infected individuals have a more aggressive course with frequent dissemination to skin and visceral organs.

LANA is one of the most highly expressed proteins during latent HHV-8 infection.17 The LANA protein is encoded by the open-reading frame 73 (ORF73) of the HHV-8 genome, where it tethers viral DNA to host heterochromatin and is thereby required for persistence of viral DNA in dividing cells.18, 19 LANA is also capable of interacting with pRb, which plays a regulatory role in the beginning of the S-phase of the cell cycle.20 Additionally, LANA mimics the E6 and E7 viral oncoproteins of human papillomavirus in its ability to inhibit the function of p53.21 LANA has an essential role in maintenance of the episomal DNA during latent infection and cell division, and also regulates gene expression in infected cells.22

A relatively small number of studies have examined LANA expression by IHC. Using a polyclonal antisera, Katano et al23 found LANA expression in 100% of Kaposi sarcoma cases studied. Kellam et al,6 Dupin et al5 and Courville et al,24 using the LNA53 monoclonal antibody, noted LANA expression in well over 90% of Kaposi sarcoma cases, as compared with 0% of non-Kaposi sarcoma controls. Schwartz et al,7 using the same 13B10 clone utilized in the present study, noted LANA expression in all classic and HIV-associated Kaposi sarcoma studied and in 69% of endemic (African) Kaposi sarcoma. All non-Kaposi sarcoma vascular tumors were LANA-negative. Most recently, three large studies, by Cheuk et al,8 Robin et al9 and Patel et al,25 have shown LANA expression in 71 of 72 cases, 50 of 50 cases and 21of 21 of Kaposi sarcoma, respectively, as compared with 0% of non-Kaposi sarcoma potential mimics. All of the non-Kaposi sarcoma vascular tumors in these three studies occurred in HIV-negative individuals, except for four cases of bacillary angiomatosis in the study of Cheuk et al;8 neither study examined HHV-8 positivity by PCR. The present study detected LANA expression in 92% of Kaposi sarcoma cases, a rate comparable to that seen in previous results. Our original speculation that a subset of vascular tumors from HIV-infected patients might express LANA proved not to be the case; as in the general population, we observed no LANA expression in any hemangioma occurring in HIV-positive patients, nor in the one epithelioid hemangioendothelioma. The number of such cases that we were able to accrue was, however, relatively small, despite our very extensive search of the archives of a large county hospital with a large number of HIV-positive patients. Conceivably, a study involving a larger number of cases might conceivably identify rare LANA-positive non-Kaposi sarcoma vascular tumors in immunocompromised patients.

Molecular genetic techniques, such as traditional gel-based PCR, in situ PCR, and in situ hybridization, are more sensitive than LANA IHC, and have traditionally been accepted as the ‘gold standard’ for the detection of latent HHV-8 infection. Real-time PCR, as used in this study, offers several advantages over conventional PCR. It is a quantitative assay that generates a signal based on the starting copy number only if there is amplification of the target sequence. Real-time PCR has also been shown to have greater sensitivity than conventional PCR, and is less time-consuming and cumbersome because it obviates the need for postamplification gel electrophoresis.26

Our finding of uniform HHV-8 expression in Kaposi sarcoma, by PCR, is in agreement with numerous prior studies, all of which have found >90% of Kaposi sarcoma of all types to be HHV-8 positive, using both traditional and in situ PCR.16, 27, 28, 29, 30, 31, 32, 33 The 100% detection rate we observed may in part be due to our use of a high cycles-to-threshold limit of 50. We chose to carry reactions out to 50 cycles-to-threshold for the purpose of detecting potentially very low viral loads in paraffin-embedded tissue blocks that had been in storage for up to 10 years. Although one could argue that this is the upper limit of sensitivity for the TaqMan technique, it is our experience that a value over a fluorescence threshold of 30 is evidence of a true infection despite a slightly higher cycle limit.

Our finding of HHV-8 positivity in 4 LANA-negative capillary hemangiomas is intriguing, and in agreement with a small number of studies which have documented HHV-8 positivity in occasional non-Kaposi sarcoma vascular tumors. HHV-8 DNA sequences have been identified by both PCR and Southern blot analysis in a subset of angiosarcomas by both McDonagh et al30 and Gyulai et al,34 and in a single case of pyogenic granuloma.29 None of these cases were reported to have occurred in an immunocompromised individual, and none was studied for LANA expression. The great majority of non-Kaposi sarcoma vascular tumors previously studied have, however, been negative for HHV-8 by PCR.16, 27, 28, 29, 30, 31, 32, 33 Given that the positive cases in the present study all showed levels above our preset high threshold level, we believe our results do indicate the true presence of HHV-8, rather than representing ‘false positives’. Regrettably, we did not have access to frozen tissue from these positive cases, and thus were unable to perform Southern blot confirmation. There are several possible explanations for our higher rate of detection of HHV-8 DNA in non-Kaposi sarcoma vascular tumors. We believe the most likely explanation is that the real-time PCR technique was able to detect HHV-8 within intratumoral blood mononuclear cells. A certain subset of these HIV-positive patients were likely to be coinfected with HHV-8, and probably harbored latent DNA in blood mononuclear cells. Further, the endothelial cells in capillary hemangiomas from these same patients failed to show staining with the LANA antibody. Although there is a small possibility that latent DNA is present in endothelial cells at a level that is undetectable by LANA staining, the similar PCR cycles-to-threshold values seen in the Kaposi sarcoma and non-Kaposi sarcoma cases make this explanation less likely. Lastly, although it is difficult to exclude completely the possibility of a ‘false positive’ due to amplification of a nonspecific PCR reaction product, in the absence of Southern blot confirmation, repeated PCR study of these four cases showed identical results.

We conclude that the IHC demonstration of LANA expression is a highly sensitive and specific marker of Kaposi sarcoma in both the general population, as shown previously, and in HIV-positive patients. Given the importance of distinguishing Kaposi sarcoma from its many potential mimics, it is noteworthy that no non-Kaposi sarcoma vascular tumor to date has been reported to express LANA. This is in stark contrast to both our own and previous HHV-8 PCR studies, which have documented HHV-8 positivity in a small but significant subset of non-Kaposi sarcoma vascular lesions. These observations have important clinical implications inasmuch as they support the use of LANA immunohistochemistry, rather than HHV-8 PCR, as a histomorphologic adjunct in the evaluation of problematic vascular proliferations that include Kaposi sarcoma in the differential diagnosis.