HIV-associated oral lesions; immunologic, virologic salivary parameters

Charles E. Barr\ Marta R. Lopez\ Ana Rua-Dobles\ Lorraine K. Miller^ Usha Mathur-Wagh' and Livia R. Turgeon' Departments of ^Dental Medicine. ^Medicine Beth Israel tuledical Center. ^Division ol Biostatistics. Columbia University. School of Public Health. New York. USA

Barr CE, Lopez MR, Rua-Dobles A. Miller LK, Mathur-Wagh U, Turgeon LR: HIV-associated oral lesions: immunologic, virologic and salivary parameters. J Oral Pathol Med 1992; 21: 295-8. There are nutnerous reports of oral lesions in HIV-infected individuals. However, few correlate the oral lesious with laboratory parameters. This study examined oral candidiasis and hairy leukoplakia, the two most cotntnon HIV-associated oral lesions, in relation to T-ccll counts, p24 core antigen levels and salivary fiow rates. Oral muco.sal exatninations. immunologic and virologic studies and stimulated whole and parotid saliva flow rates were conducted on 135 (HIV-f = 102, H I V - =33) homosexual or bisexual men. Results indicate that, among HIVinfected subjects, the odds of having oral caudidiasis is 6 times (95% Cl = 0.6-56.6) greater for subjects with T4 counts between 200-399 per mm', and 23 times (95% CI = 2.8-193.0) greater for subjects with T4 counts less than 200/ mrn' compared to subjects with T4 counts of 400/tntTt' or greater. Subjects had an equal likelihood of having hairy leukoplakia at different levels of itnimunocompetencc. The prevalence of oral candidiasis and hairy leukoplakia was higher among subjects with itifcctious virus in their serum, but was only statistically significant for hairy leukoplakia (p = 0.0\).

Many different oral lesions have been noted in itidividuals infected with Human Immunodeficiency Virus (HIVI) (1, 2. 7) and their reported prevalence has varied among scveial studies siuce the epidemic was recognized (3-6). But. the prevalence of the various lesions among representative samples of HIVinfected individuals is unknown. Reports arc often on HIV-itifcctcd individuals who have usually been referred for the evaluation and treatment of oral diseases. Candidiasis and hairy leukoplakia are the two most often observed oral manifestations in HlV-1 infection (8). KLHIN ct al. concluded that the presence of oral caudidiasis is iclatcd to the dcveloptnenl of AIDS iu those itidividuals exhibiting high risk behavior (9). Hairy leukoplakia has also been reported as a precursor of progression to AIDS (10. 11). It was originally believed to be present only in those who were HIV-infected. but later it was diagnosed in case reports associated with other initnunosuppressive disorders not associated with HIV infection (12. 13). Nutnetous reports of oral lesions have also been based on clinical assessments with little or no information about the laboratory status of the patient (1-5). However. MoNiAci et al. (14) did addre.ss the issue

of the relation of oral lesions to circulating T4 lymphocytes as well as p24 antigctiemia iu a sample of individuals exhibititig various risk factors. This report presents initial findings from a longituditial study of a cohort of 102 HIV-infected and 33 non-infected homosexual or bisexual men who were followed over a 2-yr period and evaluated evety 4 months, clinically and by laboratory methods. The specific aim of this phase of the study was to determine whether or not the prevalence of oral tnucosal lesions, at the poiut of study entry, was associated with immutic status, the presence of virus in scrum or saliva production.

Key words: HIV: oral lesions: T4 cells: saliva: viral cultures; p24 core antigen. C. E. Barr. Department of Dental tvledicine. Beth Israel Medical Center. First Avenue at 16th Street. New York. NY. 1003. USA. Accepted tor publication January 26, 1992.

subjects should know their HIV status, be homosexual or bisexual, not have a history of intravenous drug use. and reside in the rnelropolitan New York area. Those selected had not presented to a health facility for oral problems, or for routine dental treatment, since choosing those known to have oral probletns would result in a sample highly biased towards disease. The sample was 85% White. 4% Black, and 11% Hispanic. The subjects ranged in age from 24 to 63. with a median age of 33. Years of formal education ranged from 11 to 18. with a median of 16. All subjects were ambulatory and functional although matiy had symptoms of HIV itifection.

Sample selection Male volutueers. 102 HIV-infected and 33 nou-iufected. were consecutively recruited from a variety of hotnosexual organizations in New York City including Gay Men's Health Crisis. ACT-UP and Body Positive. Subjects were enrolled on the basis of their HIV status according to a case-control study design, with HIV-infcctcd individuals being cases and non-infected individuals from the same organizations serving as controls. Criteria for inclusion were that

Methods The oral cavity of each subject was thoroughly examined for tnucosal lesions using established diagnostic criteria (7, 18). The dental examiners were "blinded" to the serostatus of subjects upon ctitry into the study. All of the subjects were examined by the same individual, a dctital hygienist/nur.sc. Diagnoses of oral diseases were eonfirmed by a dentist experienced in oral manifestations of HIV infection. A diagnosis of oral

296

B A R R el

al.

Table 1. Prevalence of T4 cell count range, diagnoses, medications, and positive p24 core antigen assays of saliva and serum among HIV-infected subjects at study entry "(a)

Percent

T4 Count/mm^: 400

32/101 30/101 38/101

32"/. 30'^^. 38"/,

Diagnosis: AIDS defming illness Hairy leukoplakia Oral candidiasis Aphthous ulcers Herpetic ulcers Oral Kaposi's sarcoma

29/102 20/102 17/102 4/102 3/102 2/102

29'/., 20"/. 17"/, 4"/, 3'/. 2y«

Medications: Any medication On antifungals On acyclovir

74/102 17/102 26/102

\1"A 26%

p24 assay: Antigenemia (b) Viremia (c)

20/102 45/99

20"/. 46'/,

Variable

(a) Numerator is the number of subjects within the variable category and denominator is the total number ol subjects evaluated. Denominators do not always equal the total number of subjects ( « = 102) due to missing data. (b) Direct assay of serum samples for viral p24 core antigen. (c) Peripheral blood mononuclear cell (PBMC) cultures inoculated with serum samples to detect infectious HIV.

candidiasis required clinical evidence of infection confirmed by Sabouraud culture as well as a positive KOH smear prep. The type of candidiasis, whether angular cheilitis, erythematous, or pseudomembranous, was difierentiated. Hairy leukoplakia was diagnosed by clinical examination using established diagnostic criteria (I, 10). In addition, a culture and KOH smear from the lesion were performed to rule out oral candidiasis. If the white lesion persisted after appropriate local antifungal treatment, a diagnosis of hairy leukoplakia was considered confirmed. Medical histories, including medication use, and physical examinations were performed by a physician. All subjects gave consent and were confidentially tested using Enzyme-linked immunosorbent assay (ELISA. Abbott) for antibodies to HIV to confirm their HIV status. Stimulated whole and parotid saliva flow rates were recorded and saliva specimens were collected in sterile containers. The flow of whole saliva was stimulated by having the subjects chew on a sterile elastic band. The production of parotid saliva was stimulated by rubbing a 2'V;, citric acid solution on the tongue.

Blood specimens were collected for complete blood counts, T-cell differentiation, viral p24 core antigen assay and serum cultures. T4 (helper) and T8 (suppressor) lymphocyte counts were determined by flow cytometry. p24 core antigen assays were directly performed on serum samples using Dupont HIV-1 p24 Core Profile ELISA kits. Serum samples were considered positive, or antigenemic, if the p24 level was 12 pg/ml or greater. In addition, serum samples were added to peripheral blood mononuclear cell (PBMC) cultures to assay for the presence of cell-free, infectious virus. Cultures were considered positive, or viremic, if the p24 level in the five times concentrated supernatant was 1000 pg/ ml for one of the days assayed or 200 pg/ml for two or more days. Cultures were terminated at 27 to 35 days.

Statistical analysis

Differences in the prevalence of oral lesions between groups categorized by diagnosis, medication use, p24 assay results and T4 cell counts were tested by chi-square analyses; a p-value of less than 0.05 was considered statistically significant (16). Differences between groups, categorized by presence or absence of lesions, in the distribution of TTable 2. Prevalence of oral candidiasis (OC4-) in relation to T4 cell counts, AIDS diagnosis, antifungal use, p24 antigenemia and viremia at study entry Variable T4 Couni/mm': 400

"(a)

OC + /(-value (b)

13/32 3/30 1/39

41% 10% J'A

0.0001

AIDS defining illness: Yes 8/30 No 9/72

ITA 12"/,

0.08

29"/. 14%

0.12

25"/, 14'/(,

0.65

22'/, 13"/.

0.22

On antifungal: 5/17 Yes 12/85 No p24 Antigenemia . (c) 4/20 Positive 13/82 Negative Viremia: (d) Positive Negative

10/45 7/54

Table 3. Prevalence of oral hairy leukoplakia (OHL-f) in relation to T4 cell counts. AIDS diagnosis, acyelovir use, p24 antigenemia and viremia at study entry Variable

" (a)

OHL + /(-value (B)

T4 Count/mm 400 6/39

25'/. 20'/o \5"A

0.60

AIDS defining illness: 7/30 Yes 13/72 No

23'M> 18»/o

0.54

On acvclovir: Yes No

3/26 17/76

ll"/o 22'/.

0.23

p24 antigenemia: (c) Positive 3/20 Negative 17/82

17"/o 20"/>

0.68

Viremia: (d) Positive Negative

31"/. 11"/.

0.01

14/45 6/54

(a) Numerator is number of subjects with oral hairy leukoplakia and denominator is the number of subjects within the variable category. (b) Chi-square statistic, p-valuc. (c) Direct assay of serum samples for viral p24 core antigen. (d) Peripheral blood mononuclear cell (PBMC) cultures inoculated with serum samples to detect infectious HIV.

cell counts, and stimulated parotid and whole saliva flow rates were tested by the nonparametric Wilcoxon 2-Sample Rank Sum W test (16). Differences between types of oral candidiasis in the distribution of T-cell counts were also tested by Wilcoxon 2-sample Rank Sum W test. Odds ratios of oral lesions by T4 cell count categories were determined by logistic regression analysis.

Results

(a) Numerator is number of subjects with oral candidiasis and denominator is the number of subjects within the variable category. (b) Chi-square statistic, p-value. (c) Direct assay of serum samples for viral p24 core antigen. (d) Peripheral blood tnononuclear cell (PBMC) cultures inoculated with serum samples to detect infectious HIV.

One or more oral mucosal lesions, which included hairy leukoplakia, oral candidiasis (i.e. pseudomembranous, erythematous, angular cheilitis), aphthous and herpetic ulcers, and Kaposi's sarcoma, were observed in 36'y!i of the 102 HIV-infected subjects. None of the 33 non-infected control subjects had oral lesions. Therefore, only the HlV-infeeted group was studied to test for associations between the two most common oral lesions, oral candidiasis and hairy leukoplakia, and other clinical and laboratory values. Table I shows the prevalence among the HIV-infected group of T4 cell count ranges, clinical diagnoses, medication use, positive p24 core antigen assays (antigenemia) and positive serutn cultures (viremia) at study entry. Thirty-

HIV-associated oral lesions 297 Table 4. Medians (minimum, maximum) of T4. T8 cell counts. T4/T8 ratios and stimulated parotid and whole saliva How rates by presence (OC-I-) or absence (OC —) of oral candidiasis among HIV-infected subjects at study entry Median value (minimum, maximum) Variable

OC-h

Total («=102)

OC(n = 85)

/'(a)

Lymphocyte counts: (per mm') T4 ' 347 (4. 1012) T8 810 (115. 1929) T4/T8 0.38(0.01.2.67)

93 (17. 403) 754 (115. 1447) 0.17(0.03.0.67)

374(4. 1012) 820 (181. 1929) 0.43 (0.01. 2.67)

0.001 0.24 0.004

Stimulated fiow rate: (ml./min.) Parotid 0.71(0.14.2.17) Whole 1.36(0.43.3.75)

0.62 (0.17. 1.67) 1.54 (0.48. 3.50)

0.71 (0.14. 2.17) 1.25 (0.43. 3.75)

0.13 0.26

(a) Wilcoxon 2-sample rank sum W test. 2-lailed /j-value.

two percent of the subjects were severely immunosuppressed with T4 cell counts below 200/mm\ Twenty-nine percent had a current or past history of an AIDS defining illness (17). Seventythree percent were on various medications, such as antivirals, antifungals. Pneumocystis carinii pneumonia (PCP) prophylaxis, antidepressants, antihistamines. or antidiarrheals. The prevalence of the use of antifungals or acyclovir, which effect the appearanee of oral candidiasis (19) and hairy leukoplakia (20), was 17 and 26%, respectively. Also presented in Table 1 are tesults from direct assays and cultures of serum for p24 core antigen. Twenty subjects (2O'M.) out of a total of 102 subjects assayed had p24 antigenemia, and 45 subjects out of a total of 98 (46%) had viretnia. Tables 2 and 3 cotnpare the prevalence of oral candidiasis and hairy leukoplakia in relation to various variables. The prevalence of oral candidiasis was higher among subjects with lower T4 counts, among those who had an AIDS defining illness, and among tho.se using antifungals; but, only the association with T4 counts was statistically significant. The prevalence of hairy leukoplakia was not significantly different by T4 count. AIDS diagnosis, or u.se of aeyclovir. Since medications did not significantly correlate with the prevalence of lesions, subsequent analyses did not control for tnedication use. Also presented in Tables 2 and 3, are analyses of the prevalence of oral lesions in relation to p24 core antigen assay results. Prevalence of oral lesions and p24 antigenctnia were not significantly associated. The prevalence of oral lesions was higher among subjects with viremia compared to nonviremic subjects, but the diflerencc was only statistically significant for hairy leukoplakia (/? = 0.01. chi-square statistic). Table 4 presents tnedian T4 counts.

T8 counts, T4/T8 ratios and stimulated parotid and whole saliva flow rates for HIV-infected subjects with or without oral candidiasis at entry into the study. Those with oral candidiasis had lower tnedians on all three T-cell measures compared to those without oral candidiasis. Differences in the distributions of the lymphocyte counts between those with or without the lesion were statistically significantly different for T4 eounts and T4/T8 ratios, but not T8 counts. The stimulated parotid saliva flow rate was lower in those with oral candidiasis than those without the lesion (/J = 0. 13. Wilcoxon rank sum W test). For hairy leukoplakia. Table 5 shows that subjects with this lesion had lower medians on all three T-cell measures compared to those without the lesion but the differences were not statistically

significant. Differences in salivary fiow rates between those with lesions and those without were also not significant. In Table 6. median T-cell counts are presented by type of oral candidiasis (i.e. pseudomembranous. erythematous, angular cheilitis). The nonparametric Wilcoxon Rank Sum W Test was used to compare the T-cell count variance of subjects categorized by type of oral candidiasis. This test revealed that the T-cell count distributions were not significantly different (p>0.05. exact 2tailed P) by type of oral candidiasis. A logistic regression analysis of oral candidiasis in relation to T4 counts indicated that as the quantity of T4 cells decreased the odds of having oral candidiasis increased. Using individuals with T4 counts of 400/mm' or greater as a reference group, the odds of having oral candidiasis was 6.0 times greater (95% CI = 0.6-56.6) for subjects with T4 counts between 200-399 and 23.4 times greater (95% CI = 2.8-193.0) for those with counts less than 200/mm'. A logistic regression analysis of hairy leukoplakia in relation to T4 counts indicated that the odds of having hairy leukoplakia did not significantly increase as the T4 cells decreased. Using subjects with T4 counts of 400/mm' or greater as a reference group, the odds of having hairy leukoplakia was 1.4 times greater (95%. CI = 0.4^.9) for subjects with T4 counts between 200-399. and 1.9 times greater (95% CI =0.6-6.2) for subjects with T4 counts less than 200/mm-\

Table 5. Medians (tiiininium. maximutn) of T4. T8 cell counts, T4/T8 ratios and stimulated parotid and whole saliva How rates by presence (OHL-t-) or absence (OHL —) of oral hairy leukoplakia among HIV-infected subjects at study entry Median value (minimum, maximum) Variable

OHL-f

Total (/)=102)

OHL(" = 82)

/"(a)

Lymphocyte counts: (per mm') T4 ' .•'47 (4. 1012) T8 810 (115. 1929) T4/T8 0.38 (0.01. 2.67)

320(4. .581) 731 (181. 1929) 0.29(0.01. 1.03)

349(9. 1012) 814 (115. 1714) 0.38 (0.02. 2.67)

0.30 0.27 0.^9

Stimulated flow rate: (ml./min.) Parotid • 0.71 (0.14. 2.17) Whole 1.36(0.43.3.75)

0.82(0.17.2.17) 1.67 (0.54. 2.69)

0.71 (0.14. 1.71) 1.20(0.43. 3.75)

0.20 0.14

(a) Wilcoxon 2-sample rank sum W test. 2-tailed /»-value.

Table 6. Medians (minimurn. maximum) of T4 and T8 cell counts per mm' and T4/T8 ratio by type of oral candidiasis at study entry Median value (minimum. maximum) Variable

n

T4

T8

T4 T8

Pseudomembranous Lrythematous Angular cheilitis

9 6 2

156 (38. 403) 183 (20. .^91) 62 (17. 108)

786 (2.M. 1447) 660 (115. 1145) 643 (532. 7.54)

0.26 (0.05. 0.70) 0.26 (0.12. 0.45) 0.08 (0.03. 0.14)

298

BARR et al.

7. PtNOHORc; JJ. Classiftcation of oral that the prevalence of oral candidiasis lesions associated with HIV infection. significantly increased as quantitative Oral Surg Oral Med Oral Pathol 1989; This study presents findings on oral levels of T4 cells decreased, especially 67: 292-5. lesions that serve as indicators of HIV- below 200/mm', and that hairy leuko8. FiKGKL DW. OvhRBY GL, GREENSPAN D, plakia was significantly associated with infection and relates them to laboratory ('/ al. Oral lesions and immune function findings. Most studies have involved viremia, it is recommended that dentists with and without HIV infection. J Dent dental clinic populations referred for the and other health care professionals may Res 1989; 68: Abstr. No. 65: 190. evaluation and treatment of oral serve an important diagnostic function 9. Kt.i:tN RS, HARRIS CA, SMALL CB, lesions. The present study, however, ex- by carefully examining for presence of Moi.i. B, LI:.SSI:R M , FR/KDLAND GH. amined subjects recruited on the basis Candida and hairy leukoplakia in the Oral candidiasis in high risk patients as the initial manifestation of the acquired of HIV serostatus from homosexual or- oral cavity. Also, that the availability immunodeficiency syndrome. A' Engt J ganizations making prevalence esti- of quantitative laboratory data such as Med 1984; 311: 354-8. mates of lesions in relation to HIV immunologic status as indicated by T4 10. GRI;I;NSI>AN D , GRi:iiNSi'AN J. HIJARST N . cells, T8 cells, T4/T8 ratios and viral status and other factors possible. Relation of oral hairy leukoplakia to inAt study entry, the only significant p24 antigenemia, as well as salivary flow fection with the human immunodeficienrates make reports of HIV-associated oral indicator of immunosuppression in cy virus and the risk of developing AIDS. this study was the presence of oral can- oral lesions more comparable and J Infect Dis 1987; 155: 475-81. didiasis, which increased with decreas- meaningful both in clinical and research 11. ScHiBT M, RiNOUM J, Scuiiinm, E, PtNt> noRc; J. Correlation between hairy leukoing immunocompetence. This finding is endeavors. Since the cohort in the presplakia and the immunological status in consistent with other studies (9, 14, 19). ent study is being examined prospectively. the appearance of new lesions in the HIV infected Danish hemophiliacs. PHAIR et al. reported that T4 counts of AIDS 1987; 1: 191 2. 200/mm' or less significantly enhanced spectrum of HIV infection will be mean12. GRI:HNSPAN D . GRI:I;NSPAN J. DI-SOUZA the risk of developing Pneumocystis car- ingful when evaluated in relation to T4 Y. LKVY JA. UNGAR A M . Oral hairy leuinii pneumonia (PCP). the most com- cells counts and other clinical and labokoplakia in an HIV-negativc renal transratory factors. mon cause of death in HlV-infected inplant patient. J Oral Pathol Med 1989; dividuals (15). They noted that oral can18: n^. didia.sis infections of at least 2 wk Acknowledgment - This study was supported 13. ITIN P. RUH.1 T G . RUDLINGI:R R, el al. duration was significantly related to the by NIDR Grant DEO856O. Oral hairy leukoplakia in an HlV-negaincreased risk of PCP among those with tive renal transplant patient; a marker for immunosuppression? Dermatologia low T4 counts of 200/mm' or less, but References 1988; 177: 126 8. that the presence of hairy leukoplakia 14. MoNiA( t D. GRK( o D. Fi.i-rriA G, RAIT1. SlI.VKRMAN S JR. MiGI.lORAn C A . LOZAwas not significantly associated with the ERi R. SiNii:ro A. Epidemiology, clinical tM-NuR F. GRHHNSPAN D , CONANT M A . risk of developing PCP. features and prognostic values of HIV-1 Oral fmdings in people with or at high In this report, hairy leukoplakia was related oral lesions. ./ Oral Pathol Med risk for AIDS: a study of 375 homosexunot associated with increasing levels of 1990; 19: 477 81. al males. J Am Dent A.

HIV-associated oral lesions; immunologic, virologic and salivary parameters.

There are numerous reports of oral lesions in HIV-infected individuals. However, few correlate the oral lesions with laboratory parameters. This study...
4MB Sizes 0 Downloads 0 Views