Detection and Direct Typing of Herpes Simplex Virus in Perianal Ulcers of Patients with AIDS by PCR

Abstract

The presence of herpes simplex virus type 1 (HSV-1) and HSV-2 in perianal ulcerations of 41 AIDS patients was assessed by virus culture and a type-specific PCR-based assay. HSV was isolated from the lesion site in 24 of 41 (58.5%) patients, and HSV DNA was detected by PCR in all 24 (100%) of these specimens. Additionally, PCR was used to detect HSV DNA in 12 of 17 (70.5%) HSV culture-negative samples. Thus, HSV genomic sequences could be demonstrated in 36 of 41 (87.8%) perianal ulcers in this series. Full agreement in HSV typing by either immunodot assay or PCR was seen in 24 samples that were positive by both virus culture and PCR. HSV-2 was demonstrated in 35 of 36 (97.2%) HSV-positive samples.

Ulcerated perianal lesions are commonly observed in patients with AIDS. Cytomegalovirus, Treponema pallidum, mycobacteria, and non-Hodgkin’s lymphomas, among others, have been indicated as causes of these lesions (1, 7, 11), but herpes simplex virus (HSV) has been pointed out as the main etiologic agent, causing 22 to 76% of all cases (7, 11). Viral isolation in cell culture remains the definitive diagnostic method (6, 13), although in recent studies amplification of viral DNA by PCR has proved to be more sensitive than the standard culture method for assessing genital herpes (4, 5, 8, 9). In otherwise healthy patients, lesions of the urogenital tract are predominantly caused by HSV type 2 (HSV-2) (12), but there are no studies comparing the prevalence of HSV-1 and HSV-2 in perianal ulcers of AIDS patients.

A type-specific PCR-based assay and viral isolation in cell culture were used in this study to determine the frequency of HSV-1 and HSV-2 in perianal ulcers of AIDS patients. Four hundred AIDS patients hospitalized between May and December 1995 for treatment of various opportunistic infections, irrespective of previous or active HSV infection, were included in the study. All patients were classified as clinical category C according to the 1993 revised classification system for human immunodeficiency virus (HIV) infection from the Centers for Disease Control and Prevention (3). In all patients the presence of antibodies to HIV was demonstrated in two different samples by enzyme immunoassay and Western blotting.

Information regarding history of genital or perianal ulceration and receptive anal intercourse, risk factors for HIV acquisition, CD4-cell counts, and the use of antiviral agents was obtained from each patient. Evaluation of patients included clinical examination with special attention to the perianal region in search of ulcers.

Specimens for HSV culture and PCR were obtained with premoistened cotton-tipped swabs gently rubbed over the perianal ulcer and then placed in 3 ml of Hanks balanced salt solution containing 50 μg of vancomycin hydrochloride per ml, 500 μg of imipenem per ml, 500 μg of amikacin sulfate per ml, and 2 μg of amphotericin B per ml. Each specimen was inoculated into tubes containing confluent Vero cell monolayers in Eagle’s minimal essential medium (Gibco) supplemented with 2% fetal calf serum. The cultures were maintained at 37°C for a minimum of 2 weeks and examined daily for the cytopathic effect that is characteristic of HSV infection. HSV isolates were identified and typed by an in-house immunodot assay. Briefly, culture tube monolayers with more than 75% cytopathic effect were scraped and the cells were washed twice with phosphate-buffered saline (PBS), pH 7.2. The pellet was resuspended in 25 μl of lysis buffer (50 mM Tris, pH 8.0; 0.5% sodium deoxycholate; 0.5% Nonidet P-40; and 20% glycerol) and incubated for 1 h at 4°C. After clarification of the solution by centrifugation in a microcentrifuge (Sorvall MC 12V) for 2 min, 2 μl of the supernatant was applied as a dot to two nitrocellulose strips and allowed to dry for 10 min at 37°C. The strips were then incubated for 2 h at room temperature with blocking buffer (5% defatted milk in PBS). Each strip was incubated with either anti-HSV-1 or anti-HSV-2 monoclonal antibodies (Chemicon) for 1 h. After being washed five times with PBS containing 0.1% Tween 80, the strips were incubated for 1 h with anti-mouse peroxidase conjugate (Sigma). After three additional washes with PBS containing Tween 80, the strips were washed twice with Tris-buffered saline, pH 7.5, containing Tris (20 mM) and NaCl (500 mM) and revealed with chromogenic substrate (Tris-buffered saline containing a solution of 0.06% 3,3′-diaminobenzidinetetrahydrochloride, 0.03% cobalt chloride, and 0.01% H₂O₂).

Type-specific PCR-based assays were used for detection and direct typing of HSV in a manner conceptually similar to that described by Kimura et al. (9). These assays exploit differences in DNA sequence of the DNA polymerase genes of HSV-1 and HSV-2. A commercially available program (Oligo version 7.0; National Biosciences Inc., Plymouth, Minn.) was employed for selection and analysis of the three primers used in the two assays (10). The first (5′-GAGCCACTTCCAGAAGCGCAG), designated HSV-U, serves as the upper primer for both assays. The oligonucleotides used as lower primers were 5′-GTTCGTCCTCGTCCTCCCC, designated HSV-1L, for HSV-1 and 5′-GGGGCCTCCTTGTCGAG, designated HSV-2L, for HSV-2. The HSV-1- and HSV-2-specific PCRs amplify sequences that are 503 and 435 nucleotides in length, respectively. Reaction products of these sizes are easily distinguished visually after electrophoresis in agarose gels. PCR was performed on the same specimens used to inoculate cell cultures for viral isolation (8). DNA extraction was performed as described before (8). The amplification was performed in a Perkin-Elmer Cetus (Norwalk, Conn.) Thermocycler 480 by 35 cycles run at 94°C (1 min), 65°C (1 min), and 72°C (1 min), followed by a final extension at 72°C for 10 min. Amplified DNA was electrophoresed on a 2% agarose gel containing 0.5 μg of ethidium bromide per ml and examined under UV light. Under the conditions described, these primers do not generate a PCR product with other related or unrelated templates, such as human cytomegalovirus DNA (strain AD 169), varicella-zoster virus vaccine strain (Oka-Merck strain), and human papillomavirus type 16 cloned DNA.

Of the 400 AIDS patients who participated in the study, 41 (10.2%) had perianal ulceration. In 36 of 41 (87.8%) cases, HSV was demonstrated at the lesion site by virus isolation or PCR. Of the five patients that were HSV negative by both PCR and virus isolation, one had schistosomiasis, another had Kaposi’s sarcoma (both conditions histopathologically confirmed after biopsy of the lesion), and in the remaining three patients etiologic diagnosis could not be established. One of these three patients had been receiving acyclovir for 8 days, and another had been receiving ganciclovir (5 mg/kg of body weight every 12 h) intravenously for 10 days plus acyclovir (250 mg every 8 h) intravenously for 1 day when the ulcer swabs were taken. HSV was isolated from 24 of 41 (58.5%) perianal ulcers, and HSV DNA was detected by PCR in all 24 (100%) specimens from which HSV was isolated in culture. Additionally, HSV DNA was detected in 12 of the 17 (70.5%) patients whose HSV isolation was negative. In the group of patients with perianal ulcers from whom HSV could be isolated in cell culture, only 3 of 24 (12.5%) were receiving acyclovir, foscarnet, or ganciclovir when the ulcer swab was obtained compared to 10 of 12 (83.3%) of the PCR-positive-and-HSV-culture-negative patients (P = 0.00006, Fisher’s exact test). Thus, HSV could be demonstrated in 36 of 41 (87.8%) perianal ulcers in the present series. Twenty-two of the 24 isolates were typed as HSV-2, and one was typed as HSV-1. The remaining case presented a positive result for both HSV-1 and -2.

In the 24 samples positive by both tests (viral isolation and PCR), there was 100% agreement between HSV immunodot typing and PCR. The 12 samples positive only by PCR were typed as HSV-2. Therefore, in 34 of 36 (94.4%) HSV-positive perianal ulcers, HSV-2 was the only type of HSV detected. Nevertheless, HSV-1 was demonstrated by both virus culture and PCR in 2 of 36 (5.5%) HSV ulcers, alone (n = 1) or combined with HSV-2 (n = 1).

Twenty-one (63.4%) patients reported receptive anal intercourse, 9 (25%) had used drugs intravenously, and 18 (50%) had a history of perianal or genital ulcers. CD4⁺-cell count could be obtained from 30 of the 41 patients at the time they were included in the study. The median CD4⁺-cell count was 25 cells/μl (range, 7 to 102 cells/μl).

In this cross-sectional study, it was demonstrated that the prevalence of perianal ulcers in AIDS patients hospitalized for treatment of other opportunistic infections is high (10.2%) and that HSV could be demonstrated at the lesion site in 87.8% of cases. In previous studies, the incidence of HSV as a cause of perianal ulcers has ranged from 22 to 76% (2, 7, 11). Since a correlation between higher rates of HSV ulceration and CD4⁺-cell counts lower than 50 cells/μl has been previously shown in patients with HIV infection (2), it is important to consider patients’ immunologic status in comparisons of the role of HSV in the etiology of perianal ulcers in AIDS patients in different studies. In the present study, 28 of 30 (93%) patients with perianal ulcers had CD4⁺-cell counts lower than 50 cells/μl (median = 25 cells/μl), indicating advanced HIV-related immunodepression. The fact that HSV-2 was present in 34 of 36 (94.4%) HSV perianal ulcers was somewhat expected, since in immunocompetent patients HSV-2 is more common than HSV-1 in lesions of the genitourinary tract.

The use of a PCR assay, even without the benefit of probe hybridization, substantially increased the detection of HSV from perianal ulcers in the present series of tests (from 24 of 31 by virus culture to 36 of 41 by PCR). The lower sensitivity of viral culture could be due, to some extent, to the use of antiviral therapy, since 10 of 12 virus-culture-negative-and-PCR-positive patients were receiving antiviral drugs potentially active against HSV when the swabs were taken. The use of other cell lines, such as MRC-5 human diploid fibroblasts, could also have improved the rate of HSV isolation in cell cultures. However, in otherwise healthy patients with genital HSV infection, DNA amplification by PCR has been shown to be consistently more sensitive than virus isolation in cell cultures (4, 5, 8, 9).

These results emphasize the high prevalence of HSV-2 in perianal ulcers of AIDS patients with advanced HIV-induced immune depletion. Type-specific PCR-based assay performed directly on specimens taken from ulcerative lesions was found to be a rapid, sensitive, and specific technique for HSV detection and typing.