Shedding of Cytomegalovirus and Herpesviruses 6, 7, and 8 in Saliva of Human Immunodeficiency Virus Type 1 Infected Patients and Healthy Controls

Transcription

1 137 Shedding of Cytomegalovirus and Herpesviruses 6, 7, and 8 in Saliva of Human Immunodeficiency Virus Type 1 Infected Patients and Healthy Controls Erik Lucht, Maria Brytting, Lotte Bjerregaard, Inger Julander, and Annika Linde From the Department of Immunology, Microbiology, Pathology, and Infectious Diseases, Karolinska Institutet, Huddinge University Hospital; and the Department of Virology, Swedish Institute for Infectious Disease Control; and the Microbiology and Tumor Biology Centre, Karolinska Institutet, Stockholm, Sweden We used the polymerase chain reaction to study the presence of DNA from cytomegalovirus (CMV) and human herpesvirus (HHV)-6, HHV-7, and HHV-8 in saliva from 44 human immunodeficiency virus (HIV) type 1 infected patients at different stages of disease and in 15 healthy HIVseronegative controls. CMV DNA, HHV-6 DNA, and HHV-7 DNA were found in all groups, but HHV-8 DNA was found only in symptomatic HIV-1-infected patients (5 [17%] of 29). One of the patients with HHV-8 DNA in saliva had oral Kaposi s sarcoma at the time of sampling, and another later developed the tumor. CMV DNA was found most often in the patients with AIDS. HHV-6 shedding tended to be less frequent among HIV-1-infected patients than among healthy controls. HHV-7 DNA was detected least frequently in the group of patients with AIDS. The presence of viral DNA was not correlated either with antiherpesvirus drug therapy or with oral symptoms, apart from Kaposi s sarcoma. Asymptomatic viral shedding in saliva is a common source [9]. Oral ulcerations have been associated with CMV [10] but of infection for the Herpesviridae, such as Epstein-Barr virus not with HHV-6 or HHV-7 [5]. HHV-8 has been demonstrated (EBV) [1], cytomegalovirus (CMV) [2], and human herpesvi- both in Kaposi s sarcoma at other sites and in the majority of rus (HHV)-6 [3] and HHV-7 [4, 5]. The mode of spread of oral lesions [11]. the Kaposi s sarcoma herpesvirus (KSHV, or HHV-8) is less To further analyze the mode of spread of HHV-8 and the defined. Spread by sexual intercourse is suspected, but the relationship between CMV, HHV-6, HHV-7, and HHV-8 and findings in semen have been controversial [6]. In one study HIV-1-induced immunodeficiency, we used PCR to examine HHV-8 was identified in saliva from 33% (25 of 76) of patients the presence of viral DNA in saliva from HIV-1-infected painfected with HIV-1 [7], indicating a possible spread via saliva. tients and healthy controls. The results were also correlated It is known that shedding of EBV and CMV in saliva is with the degree of immunodeficiency of the patients, the presincreased among patients with severe immunodeficiency [1, ence of oral lesions, and the use of antiherpesvirus drug therapy. 2]. Both HHV-6 [3] and HHV-7 [4] are frequently secreted in saliva in healthy individuals, but studies of the relationship Patients and Methods between the presence of HHV-6 or HHV-7 and HIV-1 have Patients and controls. Unstimulated whole saliva (3 ml) given contradictory results [8]. According to one report, was collected on one occasion from 44 HIV-1-infected patients HHV-7 but not HHV-6 is found more frequently in whole and from 15 healthy HIV-seronegative controls. The subjects saliva from HIV-infected patients than in that from healthy oral statuses were determined by a dentist, and the subjects controls [5]. were followed up for 12 months at 3-month intervals. All of Oral lesions are frequent among HIV-1-infected patients, the patients were also examined for general symptoms and and in most cases, the etiology of these lesions is not known classified according to the Centers for Disease Control and Prevention (CDC) 1993 classification of HIV infection. Fifteen patients were classified as CDC group A (asymptomatic or lymphadenopathy; mean age, 38.6 years), 14 were classified Received 28 August 1997; revised 5 January as CDC group B (opportunistic infections not indicative of This work was presented in part at the 2nd International Conference on Human Herpesviruses 6, 7, and 8 held on 8 11 May 1997 in Pisa, Italy. AIDS; mean age, 42.5 years), and 15 were classified as CDC Financial support: The Committee for the Health and Caring Sciences, Swed- group C (AIDS; mean age, 39.4 years). All medications; the ish Institute for Infectious Disease Control and Dental Health Service, Stockuse of alcohol, snuff, or drugs; and CD4 / and CD8 / cell counts holm County Council, Stockholm, Sweden. Reprints or correspondence: Dr. Erik Lucht, Department of Immunology, in peripheral blood were registered. Forty-three of the patients Microbiology, Pathology, and Infectious Diseases, Karolinska Institutet, Hud- were male. Forty-two of the patients had been infected sexually, dinge University Hospital, I 48, S Huddinge, Sweden. and two of the men had been infected via intravenous drug Clinical Infectious Diseases 1998;27: use. Fourteen HIV-negative, heterosexual males and one HIV by the Infectious Diseases Society of America. All rights reserved /98/ $03.00 negative female served as controls (mean age, 35.7 years).

2 138 Lucht et al. CID 1998;27 (July) Thirteen patients were receiving single or combination ther- a tendency toward less shedding among HIV-1 infected patients apy for HIV-1 infection. None of the patients were being treated than among controls (figure 1B). with protease inhibitors. Of patients classified as CDC group HHV-7 DNA. HHV-7 DNA was found in all groups (figure C, three were receiving acyclovir; two were receiving foscarnet; 1C). It was detected significantly more often (P õ.05) in saliva and one was receiving acyclovir, foscarnet, and gan- from controls (15 [100%] of 15) and from patients in CDC ciclovir. None of the participants in the control group were group A (14 [93%] of 15) than in patients in CDC group B receiving any medication. (10 [71%] of 14) and group C (8 [53%] of 15). The most Extraction of DNA from saliva samples. Both lysis and DNA pronounced difference was that between controls and patients extraction were used. The lysis method, with use of proteinase in CDC group C (P õ.01). K, was performed as described previously [2]. This method HHV-8 DNA. HHV-8 DNA was detected in patients in CDC results in a 1:2 dilution of saliva. DNA extraction was per- groups B (2 [14%] of 14) and C (3 [20%] of 15) but not in formed by using the QIAamp Blood Kit (Qiagen; Hilden, Germany) CDC group A or in the control group (figure 1D). according to the manufacturer s instructions; however, Relationship between the presence of DNA from the different while 200 ml of saliva were absorbed to the silica membrane, herpesviruses in saliva. The most common finding was DNA the DNA was eluted with only 50 ml of sterile water, resulting from both HHV-6 and HHV-7 (detected in 28 of 59 samples), in a fourfold concentration of DNA. After treatment, the samples followed by DNA from CMV, HHV-6, and HHV-7 (7 of 59), were frozen at 070 C until PCR analysis was performed. and HHV-7 alone (7 of 59). Other combinations occurred in To detect contamination during preparation of a sample, every one-to-four of 59 samples. Certain combinations of the herpesfifth extracted sample was a water control. viruses were never found (i.e., CMV/HHV-6, CMV/HHV-7/ PCR. The nested PCR systems used have been described HHV-8, CMV/HHV-6/HHV-8, HHV-6/HHV-8, and HHV-7/ previously: HHV-6 and HHV-7 [12], HHV-8 [13], and CMV HHV-8). However, no viruses were mutually exclusive, since [14]. The assays were performed with minor modifications. all four viruses occured together in one patient. Two of three Five microliters of the prepared samples were analyzed in the of the saliva samples without any detectable viral DNA were HHV-6 and HHV-7 PCR, and 10 ml were analyzed in the from HIV-1-infected patients in CDC group B, and one was HHV-8 and CMV PCR assays. The PCR products were electro- from a patient in CDC group C (table 1). Increased severity phoresed on agarose gel containing ethidium bromide, and the of HIV-1-induced immunodeficiency was characterized by a results were photographed under ultraviolet light. reduction in HHV-7 shedding and an increase in the frequency The following borderline-positive controls were used in the of CMV and HHV-8, but otherwise there was no discernible respective PCR systems: 5 fg of extracted CMV DNA (corre- pattern of viral combinations. sponding to 20 DNA molecules); a supernatant from HHV- Oral findings. The oral symptoms of the 44 HIV-1-infected 6 strain Z29, containing the equivalent of 30 HHV-6 DNA patients were candidosis (39% of patients), periodontal disease molecules; a diluted saliva sample containing HHV-7; and (27%), hairy leukoplakia (14%), dry mouth (11%), ulcers (7%), DNA extracted from a Kaposi s sarcoma biopsy specimen con- and Kaposi s sarcoma (2%; table 2). Many patients had more taining HHV-8. To detect contamination, every third sample than one oral symptom, and the oral symptoms were more examined was a water control or a negative extraction control. frequent among patients with advanced immunodeficiency. Statistical analysis. All results were evaluated by performing One of the HHV-8 shedders in CDC group C presented with Fisher s exact test. an oral Kaposi s sarcoma lesion, and one of the HHV-8 shedders in CDC group B developed Kaposi s sarcoma of the oral cavity and skin 10 months after the first examination. None of Results the patients without HHV-8 DNA in saliva developed Kaposi s sarcoma within a period of 18 months (P õ.05). There was Lysed saliva vs. DNA extraction. There were significantly no correlation between the presence of HHV-6 DNA, HHV-7 more positive results when DNA extraction was used than DNA, or CMV DNA in saliva and any specific oral symptom. when saliva treated only with lysis buffer was examined The overall frequency of oral symptoms was higher among (P õ.01) (figure 1). The figures below are based on the results patients who were PCR-positive for CMV DNA than among obtained after DNA extraction. HIV-1-infected patients who were PCR-negative for CMV CMV DNA. CMV DNA was found in saliva from all groups, DNA in saliva (P õ.01). but it was found significantly more frequently in patients classified Antiviral treatment. Six of 15 patients in CDC group C as CDC group C than in those classified as CDC group A received drugs with a known effect on herpesviruses. In the (P õ.05) or in controls (P õ.01; figure 1A). treated group, 10 herpesviruses were found in saliva (mean HHV-6 DNA. HHV-6 DNA was detected in saliva from all number, 1.7 herpesviruses per patient). In the control group, groups at a frequency of 57% 93%, and there were no significant 28 herpesviruses were found (mean number, 1.9 per patient). differences between the patient groups, although there was The three patients receiving acyclovir shed two herpesviruses

3 CID 1998;27 (July) Shedding of CMV and HSV in Saliva 139 Figure 1. Percentage of examined whole saliva samples, obtained from 44 HIV-infected patients and 15 healthy controls, with demonstrable DNA from cytomegalovirus (A), human herpesvirus (HHV)-6 (B), HHV-7 (C), and HHV-8 (D). The patients were classified according to the 1993 Centers for Disease Control and Prevention (CDC) classification for HIV infection (see text). Å PCR results after DNA extraction; Å DNA lysis without concentration. in saliva (mean number, 0.7 per patient). The three patients Discussion treated with foscarnet shed eight herpesviruses (mean number, 2.7 per patient). None of the differences were statistically sig- The present study demonstrates the increased frequency of nificant. CMV and HHV-8 DNA and decreased frequency of HHV-7 Table 1. Combinations of herpesviruses found in saliva from 15 healthy controls and 44 HIV-1-infected patients, according to the clinical stage of the HIV-1 infection. Herpesvirus CDC group* CMV HHV-6 HHV-7 HHV-8 A (n Å 15) B (n Å 14) C (n Å 15) Controls Total / / / / / / / / / / / 0 / / / / 0 / / 0 / / 0 0 / / / / / / / 0 / / / / / / NOTE. CDC Å Centers for Disease Control and Prevention; CMV Å cytomegalovirus; HHV Å human herpesvirus; /Åpositive; 0Ånegative. * See text for definitions of classifications.

4 140 Lucht et al. CID 1998;27 (July) Table 2. Occurrence of human herpesviruses and oral symptoms in 44 HIV-1 infected patients. No. infected with indicated herpesvirus/no. infected with No. (%) in CDC group HIV (%) No. (%) infected Oral symptom* A (n Å 15) B (n Å 14) C (n Å 15) with HIV CMV HHV-6 HHV-7 HHV-8 Oral candidiais 0 8 (57) 9 (60) 17 (39) 6/17 (35) 12/17 (71) 11/17 (65) 2/17 (12) Periodontal disease 1 (7) 2 (14) 9 (60) 12 (27) 3/12 (25) 9/12 (75) 8/12 (67) 2/12 (17) Hairy leukoplakia 0 1 (7) 5 (33) 6 (14) 2/6 (33) 1/6 (17) 2/6 (33) 2/6 (33) Dry mouth 2 (13) 0 3 (20) 5 (11) 3/5 (60) 3/5 (60) 2/5 (40) 0/5 Oral Kaposi s sarcoma (7) 1 (2) 1/1 (100) 1/1 (100) 1/1 (100) 1/1 (100) Oral ulcers 2 (13) 0 1 (7) 3 (7) 2/3 (67) 2/3 (67) 3/3 (100) 0/3 No oral symptoms 9 (60) 4 (29) 1 (7) 14 (32) 2/14 (14) 7/14 (50) 9/14 (64) 2/14 (14) NOTE. None of 15 non-hiv-infected controls in the study had oral symptoms. DNA in saliva from patients with increasing degrees of HIV- 1-related immunodeficiency. By use of PCR, HHV-8 DNA was recently identified in saliva from 25 (33%) of 76 HIV-infected patients [7]. However, the demographics of the patients and the stages of their HIV infections were not presented. The lower frequency of HHV-8 DNA found in our study (5 [11%] of 44 patients) may be related to less-severe immunodeficiency and to the possibility that Kaposi s sarcoma is less frequent among HIV-1-infected homosexual men in Sweden [2] than in some other countries. The low frequency of HHV-8 genomes found in saliva indicates that spread via saliva between healthy individuals is not a major problem. The two HIV-1-infected patients who had and developed oral Kaposi s sarcoma had HHV-8 DNA in saliva. Although this number is small, it further strengthens the suspicion that there is a causal relationship between HHV-8 and Kaposi s sarcoma [11]. The relationship between HHV-6 and HIV-1 remains contro- versial. HHV-6 has been shown to both promote and inhibit the growth of HIV-1 in vitro, and serology and other studies of HIV-1-infected patients have given discordant results [6]. The tendency toward decreased shedding of HHV-6 among symptomatic HIV-1-infected patients found in the present study, together with case reports of different HHV-6-related symptoms in HIV-infected patients [6], may imply that HHV- 6 is not a major pathogen in HIV-1-infected patients but that sporadic severe infections due to HHV-6 may occur. The number of persons shedding HHV-7 diminished with increasing severity of HIV-1 disease in this study. It has been proposed that HHV-7 uses the CD4 cell receptor for attachment to the cell [15], and the decreased amount of this receptor in patients with AIDS may lead to a decrease in the number of target cells for HHV-7. HIV-1 RNA is detected in saliva more often in asymptomatic persons than in those with AIDS [2], a finding similar to that for HHV-7. Access to many receptorbearing cells in layers of the oral cavity that are relatively protected from immune attacks could allow replication of both viruses. Simultaneous examination for the presence of both HIV-1 RNA and HHV-7 in saliva would be important to further understand the in vivo relationship between the viruses. As previously shown [2], the frequency of CMV secretion in saliva correlated positively with the development of AIDS and a decrease in the number CD4 / cells but not with any specific oral symptom. It is well known that CMV-associated diseases occur most frequently in patients with peripheralblood CD4 / cell counts of õ /L. It has been suggested that the presence of CMV could be related to the oral ulcerations [10] seen in patients with AIDS. Such a relationship was, however, not observed in this study. The presence of herpesviruses found in association with progression to AIDS in this study could not be correlated with antiherpesvirus drug therapy. In conclusion, by using sensitive, nested PCR techniques and strict anticontamination procedures, we showed that the frequency of excretion of HHV-7 DNA in saliva was lower among patients with AIDS than among other groups. HHV-6 DNA excretion was less frequent among HIV-1-infected pa- tients than among healthy controls. Neither the presence of HHV-6 DNA nor HHV-7 DNA could be correlated with oral symptoms in the HIV-1-infected patients. The detection of CMV DNA and HHV-8 DNA in saliva correlated positively with the severity of HIV-1-induced immunodeficiency, but the presence of CMV DNA did not correlate with any specific oral symptom. The presence of HHV-8 DNA correlated with the presence of oral Kaposi s sarcoma, but the rare presence of HHV-8 DNA in saliva makes oral spread less plausible as a common route of infection with this herpesvirus. Acknowledgments The authors thank the Dental Clinic for Infectious Disease; the Department of Immunology, Microbiology, Pathology, and Infec- tious Diseases; the Swedish Institute for Infectious Disease Con- trol; and the Microbiology and Tumor Biology Centre, Karolinska Institutet, Stockholm, Sweden.

5 CID 1998;27 (July) Shedding of CMV and HSV in Saliva 141 References 9. Lucht E, Nord CE. Opportunistic oral infections in HIV-1-infected patients. 1. Lucht E, Biberfeld P, Linde A. Epstein-Barr virus (EBV) DNA in saliva Reviews in Medical Microbiology 1996;7: and EBV serology in HIV-1 infected patients with and without hairy 10. Flaitz CM, Nichols CM, Hicks MJ. Herpesviridae-associated persistent leukoplakia. J Infect 1995;31: mucocutaneous ulcers in acquired immunodeficiency syndrome. A 2. Lucht E, Albert J, Linde A, et al. Human immunodeficiency virus type 1 clinicopathologic study. Oral Surg Oral Med Oral Pathol 1996;81: and cytomegalovirus in saliva. J Med Virol 1993;39: Harnett GB, Farr TJ, Pietroboni GR, Bucens MR. Frequent shedding of 11. Jin Y-T, Tsai S-T, Yan J-J, et al. Presence of human herpesvirus-like DNA human herpesvirus 6 in saliva. J Med Virol 1990;30: sequence in oral Kaposi s sarcoma. Oral Surg Oral Med Oral Pathol 4. Wyatt LS, Frenkel N. Human herpes virus 7 in a constitutive inhabitant 1996;81: of adult human saliva. J Virol 1992;66: Wang F-Z, Dahl H, Linde A, Brytting M, Ehrnst A, Ljungman P. Lympho- 5. Di Luca D, Mirandola P, Ravaioli T, et al. Human herpesviruses 6 and 7 tropic herpesviruses in allogenic bone marrow transplantation. Blood in salivary glands and shedding in saliva of healthy and human immuno- 1996;88: deficiency virus positive individuals. J Med Virol 1995;45: Lin J, Lin S, Mar E, et al. Is Kaposis s sarcoma associated herpesvirus 6. Levy JA. Three new human herpesviruses (HHV6, 7, and 8). Lancet 1997; detectable in semen of HIV-infected homosexual men? Lancet 1995; 349: : Boldogh I, Szaniszlo P, Bresnahan WA, Flaitz CM, Nichols MC, Albrecht 14. Studahl M, Bergström T, Ekeland-Sjöberg K, Ricksten A. Detection of T. Kaposi s sarcoma herpesvirus-like DNA sequences in the saliva of cytomegalovirus DNA in cerebrospinal fluid in immunocompetent paindividuals infected with human immunodeficiency virus. Clin Infect tients as a sign of active infection. J Med Virol 1995;46: Dis 1996;23: Lusso P, Secchiero P, Crowley RW, Garzino-Demo A, Berneman ZN, 8. Frenkel N, Roffman E. Human herpesvirus 7. In: Fields BN, Knipe DM, Gallo RC. CD4 is a critical component of the receptor for human herpes- Howley PM, et al., eds. Virology. 3rd ed. Philadelphia: Lippincott- virus 7: interference with human immunodeficiency virus. Proc Natl Raven Publishers, 1996: Acad Sci USA 1994;91: