Introduction
Oropharyngeal candidiasis (OPC) is the most common opportunistic fungal infection in patients infected with HIV, occurring in as many as 90% patients at some point during the course of HIV disease [1]. Furthermore, OPC is an important clinical marker in HIV-infected patients predicting increasing immunosuppression [1-4]. Consequently, OPC has been included as an important criterion in most clinical staging systems for HIV infection [5,6].
Candida spp. are commensal pathogens of the oral flora found in the oropharynx of approximately 30% of healthy subjects [7]. The precise defence mechanisms that limit candidal proliferation and mucosal disease are largely undefined. It has been suggested that several factors such as blood group secretor status, salivary flow rates, epithelial barrier, antimicrobial constituents of saliva, lysozyme and lactoferrin release, presence of normal bacterial flora and local immunity are important in protecting the host against oral yeast proliferation [8-11]. Apart from these local factors, it is likely that an intact cell-mediated immunity plays a critical role defending the host against mucosal disease caused by Candida spp. HIV infection is associated with an increase in the proportion of persons colonized with Candida; this occurs even before overt immunosuppression. As HIV infection progresses towards more advanced stages of immunosuppression, the frequency of isolation of Candida spp. from the oral cavity increases; OPC relapses become more frequent and azole-resistant OPC can develop as a significant clinical problem [12,13].
Protease inhibitor-based therapy has been associated with a dramatic decrease in the rates of HIV-related opportunistic infections [14]. It appears that the observed decrease in opportunistic infections is related to the improvement in immunological function experienced by HIV-infected patients receiving this therapy. Indeed, highly potent antiretroviral therapy including protease inhibitors induces a substantial regeneration of naive and memory T cells, together with the recovery of lymphoproliferative responses against agents such as cytomegalovirus and Mycobacterium tuberculosis[15]. We postulated that oral candidal infection and colonization might function as a biological marker of immune function. In order to evaluate the impact of protease-inhibitor therapy on HIV-related OPC, we have performed a prospective study evaluating the prevalence of OPC and oral candidal colonization in a cohort of severely immunosuppressed HIV-infected patients who started therapy containing the protease inhibitor ritonavir.
Methods
Patients with documented HIV infection were eligible for the study if they were to start ritonavir therapy as participants in the ritonavir expanded access programme and consented to participate. At the time of entry in the study, the only nucleoside transcriptase inhibitor drugs available in Spain were zidovudine, zalcitabine and didanosine. Patients were studied at baseline (72 h before starting ritonavir therapy) and then at 8, 16 and 48 weeks after starting ritonavir. Changes in antiretroviral therapy were not guided by the protocol and were made at the discretion of the clinician taking care of the patient.
At each study visit, an oral examination was performed. In addition, CD4 cell count, HIV viral load, oral fungal cultures (quantitative and qualitative) and skin testing for delayed type hypersensitivity (DTH) responses were measured. Patients who took antifungal drugs during the 4 weeks prior to any of the scheduled study visits were not included in the analysis of the evolution of the oral examination and were not sampled for oral fungal cultures but were studied for DTH responses. Oral cultures were not performed in patients who were taking antibiotics at the time of the study visit. The study was approved by the local ethics committee. Written consent was obtained from all patients.
Oral examination
The extent of OPC lesions was evaluated as absent, single-localized, multiple-localized or extensive/confluent, as previously described [16].
CD4 cell count
CD4 cell counts were obtained from peripheral blood samples collected in ethylenediamine tetraacetic acid as anticoagulant, prepared by a whole blood lysis technique and analyzed on a FACScan flow cytometer (Becton Dickinson, San Jose, California, USA) using a two-colour monoclonal antibody panel.
HIV RNA measurement in plasma
Plasma was separated within 2 h of blood sampling and immediately stored in portions at -80°C. HIV RNA levels were measured by a quantitative reverse transcription polymerase chain reaction assay (Amplicor HIV-1 monitor, Roche Molecular Systems, Branchburg, New Jersey, USA) with a lower limit of detection of 200 copies/ml.
Oral fungal cultures
Qualitative assessment of oral candidal colonization was made by swabbed the oral cavity in four predetermined areas and plating the swab in Sabouraud media supplemented with chloramphenicol. Yeasts were identified to the species level.
Quantitative assessment of Candida spp. used a mouthwash procedure as previously described [16]. The procedure was undertaken preferably before breakfast and before taking medications. All oral fluid (saliva) was swallowed and the mouth was rinsed with 10 ml physiological saline for 10 s without swallowing. This rinsing fluid was expelled into a tube that was closed immediately. The solution was mixed and 0.1 ml was plated in duplicate undiluted and at 1:10 and 1:50 dilution (six plates in total) in Sabourad agar media supplemented with chloramphenicol. Candidal colony-forming units (CFU) were counted, if feasible, in the undiluted plate. If the number was too high for counting in that plate, then the diluted plates were counted. The mean of the counts in the duplicate plates was used for calculations.
Delayed type hypersensitivity responses
All the patients received 5 tuberculin units of Tween-stabilized purified protein (Evans-Medical laboratories, Madrid, Spain) derivative (PPD), applied by the Mantoux method; the tests were read 48-72 h after the application. In addition, patients underwent DTH skin testing with candidal antigen (Leti Laboratories, Barcelona, Spain) and streptokinase antigens (Hoechst Farma Laboratories, Barcelona, Spain) in a 1 : 10 dilution with allergenic extract. At least 5 mm of induration was required for a positive response to PPD and 2 mm for the candidal and streptokinase antigens.
Statistical analysis
All statistical analysis was completed with SPSS for Windows (version 7.5). Continuous variables with an approximately normal distribution were tested with the two-tailed paired sample Student's t test. If the distribution was not normal, a Mann-Whitney U test (unpaired results) and the two-sided Wilcoxon signed rank test (paired results) were used. Dichotomous variables were compared by the two-sided McNemar test of significance. The relations of CD4 cell count and oral candidal load in mouthwashes was assessed using Spearman correlation coefficients for continuous variables. Primary analyses were to determine if candidal colonization rates decreased after starting therapy with ritonavir and to determine if a significant response to ritonavir therapy was associated with a decrease in oral candidal load and oral candidal colonization. A significant response was defined a priori as a increase in CD4 cell count to > 200 × 106 cells/l and/or a decrease of > 0.5 log copies/ml in viral load.
Overall DTH response to treatment was defined as the percentage of patients with at least one skin test positive and as the dichotomous indicator of whether a patient had a positive skin test response to more antigens at each study visit that at baseline. The DTH response to treatment was tested with the two-sided McNemar test of significance changes.
Results
The enrolment period of the study lasted from 8 June 1996 to 9 July 1997. Follow-up was for 48 weeks after enrolment. Of the 99 patients who entered the study, 76 (78%) were male. Median age was 34 years (range 25-65). Risk factors for HIV were intravenous drug use (45%), unprotected heterosexual contact (27%), unprotected homosexual contact (23%), recipient of contaminated blood product (2%) and unknown (2%). Ninety-two patients (93%) had an AIDS diagnosis before entering the study (42% based on an AIDS-defining illness). At study entry, median CD4 cell count was 59 × 106 cells/l (range 0-578), median HIV viral load (measured in 77 patients) was 4.74 log10 copies/ml; 75 patients (76%) had suffered at least one previous episode of OPC and 93 patients (94%) were anergic.
Oral examination at study entry (performed in 83 patients) showed the following distribution of OPC lesions: no lesion in 56 (68%), multiple-localized lesions in 15 (18%), single-localized lesions in 11 (13%), and extensive in 1 (1%). At study entry, oral swab was performed in 85 patients and in 59 (69%) was positive for Candida spp. Mouthwash was performed in 79 patients and showed a mean oral candidal load of 2226 CFU/ml (range 0-65 500).
All 99 patients started treatment with ritonavir. Of these 99 patients, 95 (96%) had been previously treated extensively with nucleoside analogues before entering the study. Only four (4%) patients had received no antiretroviral therapy. Ritonavir was started without adding any new nucleoside analogue drugs in 60 patients (61%), with one new nucleoside analogue in 32 (33%) and with two in seven (7%). Four patients died during the study. Seven patients were lost to follow-up at 48 weeks.
There was a significant increase (P = 0.001, Student's t test for paired samples) in CD4 cell counts and a significant decrease (P = 0.0001, Student's t test for paired samples) in HIV viral load at all study visits compared with baseline (Table 1). The number of patients with undetectable viral loads (< 200 copies/ml HIV RNA) increased from 1% at baseline to 17% at 48 weeks.
Candidal colonization (the number of patients with an oral swab positive for Candida spp.) decreased (but not significantly) during the study (69% at baseline to 55% at 48 weeks;P = 0.0890, McNemar test). However, there was a significant decrease (P = 0.0171, Wilcoxon test) in the number of Candida CFU between baseline (2226 CFU/ml) and last visit (811 CFU/ml); (Fig. 1). There was also a significant decrease in the percentage of patients with single or multiple OPC lesions on oral examination at 16 and 48 weeks after starting the study (P = 0.0176 and < 0.001, respectively, McNemar test).
The percentage of patients with at least one positive skin test increased significantly (P < 0.05, McNemar test) at all study visits (Table 2) and by 48 weeks 28% of patients had at least one positive skin test result. At 48 weeks, the percentage of patients colonized with Candida was significantly lower (P < 0.001, McNemar test) in the group with a positive skin test for candidal antigen (0/6; 0%) than in the group that remained anergic (23/59; 39%).
Possible correlations between the variables that might predict candidal infection or colonization were also examined. At the last study visit, there was a significant weak inverse correlation (P = 0.005; r = -0.3078 Pearson coefficient) between the CD4 cell count and the candidal load in the mouth. Such a correlation was not present at baseline or at 8 and 16 weeks. There was no correlation between viral load and colonization at any study visit. Candidal oral load was not significantly lower in patients who had a decrease of HIV viral load of > 0.5 log10 copies/ml at 48 weeks (P = 0.4201, Mann-Whitney U test). Patients who had increased their CD4 cell counts to > 200 × 106 cells/l at last study visit had significantly less oral candidal load than patients who did not reach that level (P = 0.0165, Mann-Whitney U test). Additionally, patients who had a CD4 cell count > 200 × 106 cells/l were also significantly more likely to have at least one skin test positive when compared with patients with CD4 cell counts < 200 × 106 cells/l, both at 16 (P = 0.0310, McNemar test) and at 48 weeks (P = 0.0002, McNemar test). Patients who had a CD4 cell count > 200 × 106 cells/l at 48 weeks were more likely to have improved their DTH skin responses than patients who did not reach that level (P = 0.0001, McNemar test).
Discussion
Our study shows that antiretroviral treatment including a protease inhibitor markedly reduces the prevalence of OPC in a cohort of severely immunosuppressed patients with AIDS. During a 48-week follow-up period, the prevalence of OPC lesions on oral examination decreased from 32.5% to 1%. The decrease in the prevalence of OPC was accompanied by a significant decline in the mean oral candida load (from 2226 to 811 CFU/ml). In addition, there was a trend towards a reduction in the percentage of patients colonized with Candida spp. Taken together, these data suggest that, in HIV-infected patients, antiretroviral treatment including a protease inhibitor produces a substantial improvement of the defence mechanisms against oral mucosal disease caused by Candida spp.
In the normal host, an intact cell-mediated immunity protects against the uncontrolled proliferation of Candida in mucosal cavities. When cell-mediated immunity deteriorates, it is likely that a sequential process of more frequent yeast colonization, uncontrolled proliferation and finally mucosal disease occurs. Previous studies [12,17] have shown consistently that oral yeast colonization and development of thrush are associated with a progressive decline in CD4 cell count. Our results show that even in patients with advanced HIV infection the process that leads to OPC might be reversed by antiretroviral therapy including a protease inhibitor, even if such therapy was incompletely controlling HIV replication.
Several findings in our study suggest that the improvement in OPC was related to a better immunological function after starting protease inhibitor therapy. First, there was a significant increase in CD4 cell counts (the mean increase from baseline at 48 weeks was 128 × 106 cells/l). Second, the improvement in OPC appeared to correlate with the degree of CD4 cell recovery. Patients who reached a CD4 cell count > 200 × 106 cells/l at 48 weeks after starting ritonavir had a significantly lower oral candidal load than patients who did not reach that level. There was also a weak significant correlation between CD4 cell count and oral candidal load. Third, patients with positive skin reactions to candidal antigens at 48 weeks were not colonized by Candida. Finally, the percentage of anergic patients decreased from 94% at study entry to 72% after 48 weeks of treatment with a protease inhibitor. The improvement in DTH responses also correlated after 48 weeks of treatment with the degree of CD4 cell count recovery.
Most of the patients included in the study had advanced HIV infection and had been heavily treated with nucleoside analogues before study entry. The majority of patients started ritonavir without adding a new nucleoside analogue; therefore, it cannot be considered that they truly received what is known as highly active antiretroviral therapy. In fact, virological efficacy at 48 weeks was only marginal (mean HIV RNA log10 copies/ml change from baseline -0.88). It is remarkable that despite this suboptimal antiretroviral treatment most patients experienced an improvement in immunological function as measured by a better control of OPC and an improvement in DTH responses. Previous studies [18] have shown that CD4 cell counts can increase significantly despite persistently detectable HIV viraemia. Further follow-up is needed to evaluate the duration of this paradoxical immunological benefit in the setting of failure of virological control.
In our study, we did not find a significant correlation between HIV-1 RNA in plasma and oral candidal colonization or oral candidal loads at any study visit. In contrast, Gottfredson et al.[19] recently reported a significant correlation between HIV-1 RNA levels and oral candidal loads but no correlation with CD4 cell counts. In their study, the median CD4 cell counts were 381 and 382 × 106 cells/l for patients not colonized and colonized with Candida spp., respectively. In our series, the median CD4 cell count at baseline was 59 × 106 cells/l. This sharp difference precludes direct comparison between studies but suggests that at higher CD4 cell counts, viral load may be a marker for qualitative immune function. It might also be possible that the defence mechanisms active against oral candidal colonization and proliferation are different in more severely immunosuppressed patients. A number of retrospective studies have evaluated the impact of protease inhibitor therapy in OPC [20-22]. Hoegl et al. reported a trend towards less oral candidal colonization and disease in patients treated with protease inhibitors. In their study, the loss of candidal colonization and clearance of oral candidosis was observed only in patients with elevation of CD4 cells upon use of a protease inhibitor. In a retrospective survey of OPC episodes prior to and following combination therapy including a protease inhibitor, Hood et al.[21] found a significant decrease in the incidence of candidiasis and in the median percentage of time that patients were affected by OPC after initiation of protease inhibitor therapy. The decrease in OPC incidence found by Hood et al. was associated with a significant increase in CD4 cell counts. Finally, there have been reports [23,24] of the resolution of fluconazole-refractory OPC after initiation of protease inhibitor-based combination antiretroviral therapy. Globally these data support the proposition that the use of protease inhibitors has a positive impact on the natural history of OPC.
In a recent case-control study of 93 HIV-infected patients [25] who had recurrent OPC, treatment with protease inhibitors was associated with a marked reduction in the recurrence of OPC. Interestingly, this benefit could not be explained by immune reconstitution alone. An important virulence factor of Candida albicans is the secreted aspartic proteinase, which is assigned to the same class of aspartic proteinases as HIV proteinase [26]. It could be hypothesized that a protease inhibitor drug might improve OPC not only through CD4 recovery but also through direct inhibition of the aspartic proteinase secreted by Candida. Although preliminary data [27] suggest that HIV protease inhibitors might in fact inhibit candidal aspartic proteinase, further study is needed to explore this hypothesis.
In summary, our study shows that even in patients with advanced HIV infection, protease inhibitor-based anti-HIV therapy has a marked beneficial effect on OPC. This benefit is shown by a significant decrease in the prevalence of oral lesions of OPC and in a significant decrease in the oral candidal load. In our study, the beneficial effect induced by protease inhibitors appeared to be related to the immunological improvement achieved despite suboptimal HIV replication control.
Acknowledgements
The authors would like to thank Dr Esther Cabrero for carefully reading the manuscript and Sonia Pariente and Raquel Martin for helping in the realization of the study.
References
1. Feigal DW, Katz MH, Greenspan D. et al. The prevalence of oral lesions in HIV-infected homosexual and bisexual men: three San Francisco epidemiological cohorts. AIDS. 1991, 5: 519 -525.
2. Dodd CL, Greenspan D, Katz MH, Westenhouse JL, Feigal DW, Greenspan JS. Oral candidiasis in HIV infection: pseudomembranous and erythematous candidiasis show similar rates of progression to AIDS. AIDS 1991, 5: 1339 -1343.
3. Katz MH, Greenspan D, Westenhouse J. et al. Progression to AIDS in HIV-infected homosexual and bisexual men with hairy leukoplakia and oral candidiasis. AIDS 1992, 6: 95 -100.
4. Klein RS, Harris CA, Small CB. et al. Oral candidiasis in high-risk patients as the initial manifestation of the acquired immunodeficiency syndrome. N Engl J Med 1984, 311: 354 -358.
5. Centers for Disease Control. 1993 revised classification system for infection and expanded surveillance case definition for AIDS among adolescent and adults. MMWR 1992, 41: 1 -19.
6. World Health Organization. Acquired immunodeficiency syndrome (AIDS). Interim proposal for a WHO staging system for HIV infection and disease. Wkly Epidemiol Rec 1990, 65 : 221-224.
7. Odds F. Ecology ofCandidaand epidemiology of candidosis. In: Candida and Candidosis, 2nd edn. Edited by Odds F. London: Baillière Tindall; 1988:68-90.
8. McCarthy G. Host factors associated with HIV-related oral candidiasis. Oral Surg Oral Med Oral Pathol. 1992, 73: 181 -186.
9. Epstein JB, Truelove EL, Izutzu KT. Oral candidiasis: pathogenesis and host defense. Rev Infect Dis 1984, 6: 96 -102.
10. McCarthy GM, Mackie ID, Koval J. et al. Factors associated with increased frequency of HIV-related oral candidiasis. J Oral Pathol Med 1991, 20: 332 -336.
11. Yeh CK, Fox PC, Ship JA. et al. Oral defense mechanisms are impaired early in HIV-1 infected patients. J AIDS 1988, 1: 361 -366.
12. Sangeorzan JA, Bradley SF, He X. et al. Epidemiology of oral candidiasis in HIV-infected patients: colonization, infection, treatment, and emergence of fluconazole resistance. Am J Med. 1994, 97: 339 -346.
13. Powderly WG, Robinson K, Keath EJ. Molecular epidemiology of recurrent oral candidiasis in human immunodeficiency virus-positive patients: evidence for two patterns of recurrence. J Infect Dis 1993, 168: 463 -466.
14. Palella FJ, Delaney KM, Moorman AC. et al. Declining morbidity and mortality among patients with advanced human immunodeficiency virus infection. N Engl J Med 1998, 338: 853 -860.
15. Autran B, Carcelain G, Li T. et al. Positive effects of combined anti-retroviral therapy on CD4+ T cell homeostasis and function in advanced HIV disease. Science 1997, 277: 112 -116.
16. Graybill JR, Vazquez J, Darouiche RO. et al. Randomized trial of itraconazole oral solution for oropharyngeal candidiasis in HIV/AIDS patients. Am J Med. 1998, 104: 33 -39.
17. Schuman P, Sobel JD, Ohmit SE. et al. Mucosal candidal colonization and candidiasis in women with or at risk for human immunodeficiency virus infection. Clin Infect Dis 1998, 27: 1161 -1167.
18. Kaufmann D, Pantaleo G, Sudre P. et al. CD4-cell count in HIV-1-infected individuals remaining viraemic with highly active antiretroviral therapy (HAART). Lancet. 1998, 351: 723 -724.
19. Gottfredsson M, Cox GM, Indridason OS. et al. Association of plasma levels of human immunodeficiency virus type 1 RNA and oropharyngeal candida colonization. J Infect Dis 1999, 180: 534 -537.
20. Hoegl L, Thoma-Greber E, Rocken M, Korting HC. HIV protease inhibitors influence the prevalence of oral candidosis in HIV-infected patients: a 2-year study. Mycoses. 1998, 41: 321 -325.
21. Hood S, Bonington A, Evans J, Denning D. Reduction in oropharyngeal candidiasis following introduction of protease inhibitors. AIDS. 1998, 12: 447 -448.
22. Martins MD, Lozano-Chiu M, Rex JH. Declining rates of oropharyngeal candidiasis and carriage ofCandida albicansassociated with trends towards reduced rates of carriage of fluconazole-resistantC. albicansin human immunodeficiency virus-infected patients. Clin Infect Dis 1998, 27: 1291 -1294.
23. Zingman B. Resolution of refractory AIDS-related mucosal candidiasis after initiation of didanosine plus saquinavir. N Engl J Med 1996, 334: 1674 -1675.
24. Valdez H, Gripshover BM, Salata RA, Lederman MM. Resolution of azole-resistant oropharyngeal candidiasis after initiation of potent combination antiretroviral therapy. AIDS 1998, 12: 538. 538.
25. Cauda R, Tacconelli E, Tumbarello M et al. Role of protease inhibitors in preventing recurrent oral candidosis in patients with HIV infection: a prospective case control study. J Acquir Immune Defic Syndr Hum Retrovirol. 1999, 20-25.
26. Hoegl L, Ollert M, Korting H. The role ofCandida albicanssecreted aspartic proteinase in the development of candidosis. J Mol Med 1996, 74: 135 -142.
27. Cassone A, de Bernardis F, Torosantucci A, Tacconelli E, Tumbarello M, Cauda R. In vitroandin vivoanticandidal activity of human immunodeficiency virus protease inhibitors. J Infect Dis 1999, 180: 448 -453.
© 2000 Lippincott Williams & Wilkins, Inc.