- Email: [email protected]
Saliva and Salivary Gland Alterations in HIV Infection
07714822
P H IL IP O. FOX, D .D .S.
hanges in the m outh may occur early and frequently after a person is infected with HIV-1. Because norm al salivary function is essential to m aintaining oral homeostasis, studies have attem pted to relate the oral changes to secretory function changes.1These investigations have established that salivary glands and salivary function may be altered by HIV infection. This report reviews current knowledge concerning HIV in saliva and salivary glands and examines the effects of HIV infection on salivary function and salivary tissues. H IV -1
IN SALIVA.
There was great concern early in the AIDS epidemic that HIV-l might be spread through saliva or orally. Although the evidence from occupational, family and sexual practice studies showed that the m outh did not transm it HIV-l, m any individuals continued to express fear about contracting the virus orally. HIV-l was detected initially in whole saliva, at frequencies ranging from l percent to 45 percent of patient samples.23 Whole saliva, the mixed fluid contents of the mouth, contains varying am ounts of serum, serum products, cells, microbial products and debris, in addition to saliva. Therefore, the virus origin could not be determ ined in these studies. 46
JADA, Vol. 122, November 1991
ABSTRACT
HIV-l infection may involve saliva and the salivary glands. The virus may be recovered from the m outh and anti-viral antibodies are found in saliva. Salivary dysfunction after in fection ranges from xerostom ia to a Sjogren’s syndrome-like condition with persisten t glandular enlargem ent and marked secretory hypofunction. In addition to the serum in whole saliva, the virus may be in cells within the oral fluids. Indeed, lymphocytes containing HIV were recovered from whole saliva.4 Currently, there has been only one report of infectious virus recovered from saliva collected directly from the excretory duct.5 A very low quantity (< l infectious particle/m illiliter) of virus was found in a single parotid saliva sample of an HIV seropositive individual. Recently, a study has examined whole saliva in HIV-infected individuals with m ultiple sampling in two m onths.6Using the polymerase chain reaction, which amplifies portions of the HIV-l proviral genome, small quantities of the virus can be detected. The
polymerase chain reaction does not, however, determ ine if the virus detected is complete or infectious. The HIV proviral sequences were detected initially in the whole saliva of 10 of 20 patients tested. In another experim ent, two of six patient samples were positive in the initial sampling. Repeated sampling for 60 days, however, detected HIV proviral DNA in each patient at some time in at least a single sample. The presence of virus in the mouth is analogous to serum recovery of HIV-l, where viremia is episodic. In general, the recovery of HIV-l from oral secretions is sporadic and the concentration of virus, when present, is low. Most importantly, although HIV-l can be found in the mouth, saliva and the m outh have not been implicated in the virus transmission. H IV - 1 IN S A L IV A R Y GLANDS
Viruses are frequently recovered from salivary secretions and some, such as hum an herpesvirus 6 and cytomegalovirus, are believed to persist and replicate within salivary glands. These viruses have been identified within the acini and ductal elem ents of major salivary glands.7Investigators have searched for HIV-l in salivary tissues, and have found it in salivary glands of seropositive patients with evidence of salivary
gland disease, usually glandular enlargement. Immunofluorescence and in situ hybridization techniques have detected the virus in macrophages, monocytes and lymphocytes infiltrating both major and minor salivary glands.8'10Importantly, however, HIV-l was not found within the salivary acinar and ductal elements. This implies that infected cells may traffic into the salivary glands in HIV-l infection, but do not reside in the secretory elements. This may, in part, explain the low frequency of virus recovery from glandular secretions. A N T I-V IR A L A C T IV IT Y OF S A LIV A
A num ber of factors may con tribute to the lack of oral HIV-l transm ission. As mentioned, the virus is not found with high frequency in the m outh even in known seropositive patients. Additionally, the concentration of virus, w hen present in the mouth, is low. Even in the presence of oral mucosal lesions or periodontal disease, relatively small quantities of serum are found in the oral secretions. All of these contribute to a low oral infectious load. Saliva itself has inhibited the infectivity of HIV-l in vitro.1112This inhibitory activity is present in whole saliva and in submandib ular/sublingual secretions. When saliva is mixed with virus, the virus’ ability to infect lymphocytes is reduced or completely eliminated. This inhibitory activity has been shown in saliva from healthy men, women and children, as well as men who are HIV-l seropositive.13 The specific inhibitory factor(s) in saliva has not been identified, but this activity may reduce the infectivity of oral secretions in vivo.
Saliva also contains antibodies of the IgA and IgG class directed against HIV-l,14which are sensitive indicators of serological status. Saliva collection provides a noninvasive test for HIV infection. These antibodies also may protect the mouth, neutralizing some of the virus. H IV - t E F F E C T S ON S A L IV A R Y G L A N D F U N C T IO N
HIV-l infection also affects salivary gland function. In studies of major gland output, stimulated parotid and subm andibular/sublingual saliva output was dim inished significantly in HIV-infected individuals.1546A recent report found significant declines in stimulated parotid gland flow rates over a one-year period in a large group of seropositive patients.17 Some investigators have reported that as many as 13 percent of HIVpositive patients complain of dry m outh18; medications, however, may induce this xerostom ia in some of these individuals. In addition to flow rate changes, saliva composition is altered after HIV-l infection. Concentrations of sodium, chloride, lysozyme, salivary peroxidase and IgA are elevated consistently.'^171920Other anti-microbial proteins such as lactoferrin and histatins have been elevated in specific glandular secretions. Longitudinal studies confirm the elevations of electrolytes and proteins and show increases over time in concentrations of chloride, lysozyme, albumin, salivary perox idase and lactoferrin.17’21While results from different laboratories vary in regard to specific salivary factors, there is general agreem ent that the salivary glands are affected and that saliva of HIVinfected individuals shows increased anti-microbial activity.
Most patients retain adequate salivary function. A subset of patients (approxi mately 6 percent) develop pronounced salivary gland disease. This condition, term ed HIV-salivary gland disease, is characterized by persistent gland ular (usually parotid) enlargem ent and/or greatly decreased salivary flow.21Symptoms suggest Sjogren’s syndrome, an autoim m une exocrinopathy. The labial m inor salivary glands in HIVSGD have a lymphocytic infiltrate indis tinguishable at Dr. Fox is chief, the light Clinical microscopic Investigations Section, NIDR, NIH, level from the Building 10, Room cellular pattern 1 N -1 1 3 , Bethesda, Md. 2 0 8 9 2 . Address found in requests fo r reprints Sjogren’s to th e author. syndrome.22 Subsequent studies show that in HIV-SGD, the infiltrate is composed predom inantly of lymphocytes of the CD8 (suppressor) subtype, while in Sjogren’s syndrom e the cells are largely of the CD4 (helper) type.23 HIV-SGD has been term ed a ‘CD8 lymphocytosis’ and may represent a distinct host im mune response associated, in blacks, with HLADR5. The circulating CD8 cells and this histocompatibility antigen may influence disease activity following HIV-l infection, leading to a relatively slow progression of immunodeficiency. Additionally, HIV-SGD patients rarely have circulating autoanti bodies characteristic of Sjogren’s syndrome, suggesting that the cause of HIV-SGD is distinct from Sjogren’s syndrome. These find ings support the hypothesis that viruses (in particular, retro viruses) play a role in salivary JADA, Vol. 122, November 1991
47
gland disease and hum an auto immunity. This subset of HIVinfected patients, therefore, may be valuable in furthering our understanding of systemically m ediated salivary gland dysfunctions. SUMMARY
HIV-l can be recovered sporad ically from whole saliva and individual salivary gland secretions, b u t the concentration of virus, w hen present, is low. Saliva possesses anti-viral activity, and although virus m aybe found in the mouth, the mouth is not a transm ission route of HIV-l. After HIV-l infection, salivary gland function is altered. Flow rates may decrease, and certain electrolytes and anti-microbial proteins elevate. A subset of patients develops severe salivary gland disease. Most patients, however, m aintain adequate salivary gland function and retain the ability to produce protective salivary proteins. ■ This pa p e r w as p rese n ted April 25,1991, a t th e Scientific F rontiers in Clinical D entistry Sym posium , National Institu te for D ental R esearch, Bethesda, Md 1. M andel ID. T he role of saliva in m aintaining oral hom eostasis. JADA 1989;119:298-304. 2 G roopm an JE , S alah u d d in SZ, Sarngadharan MG, et al. HTLV-III in saliva of people with AIDS-related complex a nd h ealthy hom osexual m en a t risk for AIDS Science 1984;226:447-9. 3. Ho DD, Byington RE, Schooley RT, Flynn T, Rota TR, Hirsch MS. Infrequency of isolation of HTLV-III virus from saliva in AIDS. N Engl J Med 1985;313:1606. 4. Pekovic D, Ajdukovic D, Tsoukas C, e t al. D etection of hum an im m unosuppressive virus in salivary lym phocytes from d en tal pa tie n ts w ith AIDS. Am J Med 1987;82:188-9. 5. Levy JA, G reenspan D. HIV in saliva. Lancet 1988;ii'1248. 6. Goto Y, Yeh C-K, N otkins AL, Prabhakar BS. Detection of proviral sequences in saliva o f p atien ts infected w ith hum an im m unodeficiency virus type 1. AIDS Res H um an Retrovirol 1991;7.343-7. 7. Fox JD, Briggs M, W ard PA, T edder RS. Hum an herpesvirus 6 in salivary glands. Lancet 1990;336:590-3. 8. Yeh C-K, Fox PC, Fox CH, Travis WD, Lane HC, Baum BJ. Kaposi’s sarcom a o f th e p arotid gland in acquired im m unodeficiency syn d ro m e (AIDS). Oral Surg Oral Med Oral Pathol 1989;67:308-12. 9. B ru n n er JM, C leary KR, Sm ith FB, Batsakis JG. Im m unocytochem ical identification o f HIV (p24) antigen in parotid lym phoid lesions. J Laryngol Otol
48
JADA, Vol. 122, November 1991
1989;1031063-6. 10 Schiodt M, G reenspan D, Levy JA, e t al. Does HIV cause salivary gland disease9 AIDS 1989,3:819-22. 11. Fultz PN C om ponents of saliva inactivate hum an im m unodeficiency virus Lancet 1986;u:1215. 12. Fox PC, Wolff A, Yeh C-K, Atkinson JC, Baum BJ. Saliva inhibits HIV-l infectivity. JADA 1988;116'635-7. 13. Fox PC, Wolff A, Yeh C-K, Atkinson JC, Baum BJ Salivary inhibition o f HIV-l infectivity functional properties a n d distribution in m en, w om en, and children JADA 1989;118:709-11 14. A rchibald DW, B arr CE, T orosian JP, McLane MF, Essex M. Secretory IgA antibodies to h um an im m unodeficiency virus in the parotid saliva of patients with AIDS a n d AIDS-related com plex J Infect Dis 1987,155793-6 15. Yeh C-K, Fox PC, S hip JA, et al. Oral defense m echanism s are im paired early in HIV-l infected patients. J AIDS 1988;1:361-6. 16. A tkinson JC, Yeh C-K, O ppenheim FG, Berm udez D, Baum BJ, Fox PC. Elevation of salivary antim icrobial pro tein s following HIV-l infection. J AIDS 1990;3:41-8. 17. Mandel ID, B arr C, Turgeon L. Longitudinal study of parotid saliva m HIV infection (Special Issue) (Abstract no. 172) J D ent Res 1991,70:287. 18. Roberts MW, B rahim JS, R inne NF Oral m anifestations of AIDS: a study of 84 patients. JADA 1988;116:863-6. 19. M arder MZ, B arrC E , Mandel ID. Cytomegalovirus presence a n d salivary com position in acquired im m unodeficiency syndrom e. Oral Surg Oral Med Oral Pathol 1985;60:372-6. 20 M andel ID, Turgeon L, B arr CE. Parotid fluid flow rate and com position in HIV infection (Special Issue) (A bstract no. 1,538) J D ent Res 1990;69:301. 21. Atkinson JC, Yeh C-K, Bermudez D, Fox PC, Baum BJ. Longitudinal evaluation of m ajor salivary gland function in HIV-l infected patients. J Oral Pathol Mod 1989;18:469-70 22. Ulirsch RC, Jaffe ES. Sjogren’s syndrom e-like illness associated w ith the acquired im m unodeficiency syndrom e-related complex. Hum Pathol 1987;18:1063-8. 23. S chiodt M, G reenspan D, Daniels TE, e t al. Parotid gland en larg em en t and xerostom ia associated w ith labial sialadenitis in HIV-infected patients J Autoim m unity 1989,2:415-25. 24. Itescu S, B rancato LJ, W inchester R. A sicca syndrom e in HIV infection: association w ith HLA-DR5 and CD8 Lymphocytosis. L ancet 1989;ii:466-8.
P H IL IP O. FOX, D .D .S.
hanges in the m outh may occur early and frequently after a person is infected with HIV-1. Because norm al salivary function is essential to m aintaining oral homeostasis, studies have attem pted to relate the oral changes to secretory function changes.1These investigations have established that salivary glands and salivary function may be altered by HIV infection. This report reviews current knowledge concerning HIV in saliva and salivary glands and examines the effects of HIV infection on salivary function and salivary tissues. H IV -1
IN SALIVA.
There was great concern early in the AIDS epidemic that HIV-l might be spread through saliva or orally. Although the evidence from occupational, family and sexual practice studies showed that the m outh did not transm it HIV-l, m any individuals continued to express fear about contracting the virus orally. HIV-l was detected initially in whole saliva, at frequencies ranging from l percent to 45 percent of patient samples.23 Whole saliva, the mixed fluid contents of the mouth, contains varying am ounts of serum, serum products, cells, microbial products and debris, in addition to saliva. Therefore, the virus origin could not be determ ined in these studies. 46
JADA, Vol. 122, November 1991
ABSTRACT
HIV-l infection may involve saliva and the salivary glands. The virus may be recovered from the m outh and anti-viral antibodies are found in saliva. Salivary dysfunction after in fection ranges from xerostom ia to a Sjogren’s syndrome-like condition with persisten t glandular enlargem ent and marked secretory hypofunction. In addition to the serum in whole saliva, the virus may be in cells within the oral fluids. Indeed, lymphocytes containing HIV were recovered from whole saliva.4 Currently, there has been only one report of infectious virus recovered from saliva collected directly from the excretory duct.5 A very low quantity (< l infectious particle/m illiliter) of virus was found in a single parotid saliva sample of an HIV seropositive individual. Recently, a study has examined whole saliva in HIV-infected individuals with m ultiple sampling in two m onths.6Using the polymerase chain reaction, which amplifies portions of the HIV-l proviral genome, small quantities of the virus can be detected. The
polymerase chain reaction does not, however, determ ine if the virus detected is complete or infectious. The HIV proviral sequences were detected initially in the whole saliva of 10 of 20 patients tested. In another experim ent, two of six patient samples were positive in the initial sampling. Repeated sampling for 60 days, however, detected HIV proviral DNA in each patient at some time in at least a single sample. The presence of virus in the mouth is analogous to serum recovery of HIV-l, where viremia is episodic. In general, the recovery of HIV-l from oral secretions is sporadic and the concentration of virus, when present, is low. Most importantly, although HIV-l can be found in the mouth, saliva and the m outh have not been implicated in the virus transmission. H IV - 1 IN S A L IV A R Y GLANDS
Viruses are frequently recovered from salivary secretions and some, such as hum an herpesvirus 6 and cytomegalovirus, are believed to persist and replicate within salivary glands. These viruses have been identified within the acini and ductal elem ents of major salivary glands.7Investigators have searched for HIV-l in salivary tissues, and have found it in salivary glands of seropositive patients with evidence of salivary
gland disease, usually glandular enlargement. Immunofluorescence and in situ hybridization techniques have detected the virus in macrophages, monocytes and lymphocytes infiltrating both major and minor salivary glands.8'10Importantly, however, HIV-l was not found within the salivary acinar and ductal elements. This implies that infected cells may traffic into the salivary glands in HIV-l infection, but do not reside in the secretory elements. This may, in part, explain the low frequency of virus recovery from glandular secretions. A N T I-V IR A L A C T IV IT Y OF S A LIV A
A num ber of factors may con tribute to the lack of oral HIV-l transm ission. As mentioned, the virus is not found with high frequency in the m outh even in known seropositive patients. Additionally, the concentration of virus, w hen present in the mouth, is low. Even in the presence of oral mucosal lesions or periodontal disease, relatively small quantities of serum are found in the oral secretions. All of these contribute to a low oral infectious load. Saliva itself has inhibited the infectivity of HIV-l in vitro.1112This inhibitory activity is present in whole saliva and in submandib ular/sublingual secretions. When saliva is mixed with virus, the virus’ ability to infect lymphocytes is reduced or completely eliminated. This inhibitory activity has been shown in saliva from healthy men, women and children, as well as men who are HIV-l seropositive.13 The specific inhibitory factor(s) in saliva has not been identified, but this activity may reduce the infectivity of oral secretions in vivo.
Saliva also contains antibodies of the IgA and IgG class directed against HIV-l,14which are sensitive indicators of serological status. Saliva collection provides a noninvasive test for HIV infection. These antibodies also may protect the mouth, neutralizing some of the virus. H IV - t E F F E C T S ON S A L IV A R Y G L A N D F U N C T IO N
HIV-l infection also affects salivary gland function. In studies of major gland output, stimulated parotid and subm andibular/sublingual saliva output was dim inished significantly in HIV-infected individuals.1546A recent report found significant declines in stimulated parotid gland flow rates over a one-year period in a large group of seropositive patients.17 Some investigators have reported that as many as 13 percent of HIVpositive patients complain of dry m outh18; medications, however, may induce this xerostom ia in some of these individuals. In addition to flow rate changes, saliva composition is altered after HIV-l infection. Concentrations of sodium, chloride, lysozyme, salivary peroxidase and IgA are elevated consistently.'^171920Other anti-microbial proteins such as lactoferrin and histatins have been elevated in specific glandular secretions. Longitudinal studies confirm the elevations of electrolytes and proteins and show increases over time in concentrations of chloride, lysozyme, albumin, salivary perox idase and lactoferrin.17’21While results from different laboratories vary in regard to specific salivary factors, there is general agreem ent that the salivary glands are affected and that saliva of HIVinfected individuals shows increased anti-microbial activity.
Most patients retain adequate salivary function. A subset of patients (approxi mately 6 percent) develop pronounced salivary gland disease. This condition, term ed HIV-salivary gland disease, is characterized by persistent gland ular (usually parotid) enlargem ent and/or greatly decreased salivary flow.21Symptoms suggest Sjogren’s syndrome, an autoim m une exocrinopathy. The labial m inor salivary glands in HIVSGD have a lymphocytic infiltrate indis tinguishable at Dr. Fox is chief, the light Clinical microscopic Investigations Section, NIDR, NIH, level from the Building 10, Room cellular pattern 1 N -1 1 3 , Bethesda, Md. 2 0 8 9 2 . Address found in requests fo r reprints Sjogren’s to th e author. syndrome.22 Subsequent studies show that in HIV-SGD, the infiltrate is composed predom inantly of lymphocytes of the CD8 (suppressor) subtype, while in Sjogren’s syndrom e the cells are largely of the CD4 (helper) type.23 HIV-SGD has been term ed a ‘CD8 lymphocytosis’ and may represent a distinct host im mune response associated, in blacks, with HLADR5. The circulating CD8 cells and this histocompatibility antigen may influence disease activity following HIV-l infection, leading to a relatively slow progression of immunodeficiency. Additionally, HIV-SGD patients rarely have circulating autoanti bodies characteristic of Sjogren’s syndrome, suggesting that the cause of HIV-SGD is distinct from Sjogren’s syndrome. These find ings support the hypothesis that viruses (in particular, retro viruses) play a role in salivary JADA, Vol. 122, November 1991
47
gland disease and hum an auto immunity. This subset of HIVinfected patients, therefore, may be valuable in furthering our understanding of systemically m ediated salivary gland dysfunctions. SUMMARY
HIV-l can be recovered sporad ically from whole saliva and individual salivary gland secretions, b u t the concentration of virus, w hen present, is low. Saliva possesses anti-viral activity, and although virus m aybe found in the mouth, the mouth is not a transm ission route of HIV-l. After HIV-l infection, salivary gland function is altered. Flow rates may decrease, and certain electrolytes and anti-microbial proteins elevate. A subset of patients develops severe salivary gland disease. Most patients, however, m aintain adequate salivary gland function and retain the ability to produce protective salivary proteins. ■ This pa p e r w as p rese n ted April 25,1991, a t th e Scientific F rontiers in Clinical D entistry Sym posium , National Institu te for D ental R esearch, Bethesda, Md 1. M andel ID. T he role of saliva in m aintaining oral hom eostasis. JADA 1989;119:298-304. 2 G roopm an JE , S alah u d d in SZ, Sarngadharan MG, et al. HTLV-III in saliva of people with AIDS-related complex a nd h ealthy hom osexual m en a t risk for AIDS Science 1984;226:447-9. 3. Ho DD, Byington RE, Schooley RT, Flynn T, Rota TR, Hirsch MS. Infrequency of isolation of HTLV-III virus from saliva in AIDS. N Engl J Med 1985;313:1606. 4. Pekovic D, Ajdukovic D, Tsoukas C, e t al. D etection of hum an im m unosuppressive virus in salivary lym phocytes from d en tal pa tie n ts w ith AIDS. Am J Med 1987;82:188-9. 5. Levy JA, G reenspan D. HIV in saliva. Lancet 1988;ii'1248. 6. Goto Y, Yeh C-K, N otkins AL, Prabhakar BS. Detection of proviral sequences in saliva o f p atien ts infected w ith hum an im m unodeficiency virus type 1. AIDS Res H um an Retrovirol 1991;7.343-7. 7. Fox JD, Briggs M, W ard PA, T edder RS. Hum an herpesvirus 6 in salivary glands. Lancet 1990;336:590-3. 8. Yeh C-K, Fox PC, Fox CH, Travis WD, Lane HC, Baum BJ. Kaposi’s sarcom a o f th e p arotid gland in acquired im m unodeficiency syn d ro m e (AIDS). Oral Surg Oral Med Oral Pathol 1989;67:308-12. 9. B ru n n er JM, C leary KR, Sm ith FB, Batsakis JG. Im m unocytochem ical identification o f HIV (p24) antigen in parotid lym phoid lesions. J Laryngol Otol
48
JADA, Vol. 122, November 1991
1989;1031063-6. 10 Schiodt M, G reenspan D, Levy JA, e t al. Does HIV cause salivary gland disease9 AIDS 1989,3:819-22. 11. Fultz PN C om ponents of saliva inactivate hum an im m unodeficiency virus Lancet 1986;u:1215. 12. Fox PC, Wolff A, Yeh C-K, Atkinson JC, Baum BJ. Saliva inhibits HIV-l infectivity. JADA 1988;116'635-7. 13. Fox PC, Wolff A, Yeh C-K, Atkinson JC, Baum BJ Salivary inhibition o f HIV-l infectivity functional properties a n d distribution in m en, w om en, and children JADA 1989;118:709-11 14. A rchibald DW, B arr CE, T orosian JP, McLane MF, Essex M. Secretory IgA antibodies to h um an im m unodeficiency virus in the parotid saliva of patients with AIDS a n d AIDS-related com plex J Infect Dis 1987,155793-6 15. Yeh C-K, Fox PC, S hip JA, et al. Oral defense m echanism s are im paired early in HIV-l infected patients. J AIDS 1988;1:361-6. 16. A tkinson JC, Yeh C-K, O ppenheim FG, Berm udez D, Baum BJ, Fox PC. Elevation of salivary antim icrobial pro tein s following HIV-l infection. J AIDS 1990;3:41-8. 17. Mandel ID, B arr C, Turgeon L. Longitudinal study of parotid saliva m HIV infection (Special Issue) (Abstract no. 172) J D ent Res 1991,70:287. 18. Roberts MW, B rahim JS, R inne NF Oral m anifestations of AIDS: a study of 84 patients. JADA 1988;116:863-6. 19. M arder MZ, B arrC E , Mandel ID. Cytomegalovirus presence a n d salivary com position in acquired im m unodeficiency syndrom e. Oral Surg Oral Med Oral Pathol 1985;60:372-6. 20 M andel ID, Turgeon L, B arr CE. Parotid fluid flow rate and com position in HIV infection (Special Issue) (A bstract no. 1,538) J D ent Res 1990;69:301. 21. Atkinson JC, Yeh C-K, Bermudez D, Fox PC, Baum BJ. Longitudinal evaluation of m ajor salivary gland function in HIV-l infected patients. J Oral Pathol Mod 1989;18:469-70 22. Ulirsch RC, Jaffe ES. Sjogren’s syndrom e-like illness associated w ith the acquired im m unodeficiency syndrom e-related complex. Hum Pathol 1987;18:1063-8. 23. S chiodt M, G reenspan D, Daniels TE, e t al. Parotid gland en larg em en t and xerostom ia associated w ith labial sialadenitis in HIV-infected patients J Autoim m unity 1989,2:415-25. 24. Itescu S, B rancato LJ, W inchester R. A sicca syndrom e in HIV infection: association w ith HLA-DR5 and CD8 Lymphocytosis. L ancet 1989;ii:466-8.