Oral care

Diseases
and (2002)
treatment
8 (Suppl.
protocols
2), 136–143
in HIV
© 2002 Blackwell Munksgaard All rights reserved
PJ Shirlaw
1601–5665
et al.

136 www.blackwellmunksgaard.dk

Oral and dental care and treatment protocols for

the management of HIV-infected patients
PJ Shirlaw1, U Chikte2, L MacPhail3, A Schmidt-Westhausen4, D Croser5, P Reichart6
1
Department of Oral Medicine and Pathology, Guy’s Hospital, London, UK; 2Department of Community Dentistry, University of
Stellenbosch, South Africa; 3Department of Stomatology, University of California, San Francisco, CA, USA; 4Medizinsche Fakultat,
Humboldt University, Berlin, Germany; 5Waterside Point, London, UK

This paper describes the workings of the workshop 6 Can we provide clinical treatment that has a scientific
dedicated to oral and dental care and treatment basis rather than being trial based? – L. McPhail
protocols for the management of HIV-infected 7 Is ddI + hydroxy-urea an effective African alternative to
patients. The questions addressed were: 1) What HAART? – P. Reichart
are the current ethical issues in dental care of HIV 8 What is the influence of protease inhibitors and HAART
patients, do they need to be addressed? 2) Do we on the excretion of HIV in saliva? – A. Schmidt-Westhausen
need to modify the dental care we give HIV-positive 9 What is the effect of anti-HIV therapy on the oral mucosa
patients? 3) When is it necessary to give antibiotic and oral health? – A. Schmidt-Westhausen
prophylaxis to HIV-positive patients? 4) What is the
evidence for the effective treatment of oral lesions
associated with HIV? 5) What is the most successful
Introduction
palliative treatment for KS? 6) Can we provide Two separate topics were considered during this work-
clinical treatment that has a scientific basis rather shop: oral and dental care, and treatment protocols for the
being trial based? 7) Is ddI + hydroxy-urea an effect- management of HIV-infected patients. The effect of anti-
ive African alternative to HAART? 8) What is the retroviral therapy (ART) or highly active antiretroviral
influence of protease inhibitors and HAART on the therapy (HAART) has influenced oral and dental care. The
excretion of HlV in saliva? 9) What is the effect morbidity and mortality of patients receiving ART and
of anti-HIV therapy on the oral mucosa and oral HAART has been markedly reduced. The availability of
health? This workshop did not fully cover the issue ART and HAART, socio-economic, geographical, ethnic,
of ddI and hydroxy-urea as an alternative HIV gender, risk-group and age differences play a role in the
therapy as this was considered to be the remit of spectrum of oral diseases that may be encountered. Some
general physicians caring for patients with HIV and diseases, for example, penicilliosis or tuberculosis, may be
AIDS rather than that of oral health care workers. seen in some parts of the world but rarely observed in
others. Apart from these differences, drugs to treat HIV, as
Keywords: dentistry; oral lesions; HIV; treatment protocols;
well as basic analgesics and antimycotics, may not be avail-
oral mucosa
able in developing countries.
Dental care in HIV and AIDS patients does not differ from
the general population. They require the highest standard
of dental care available. The ethical problems that were
Questions addressed by the workshop
seen at the beginning of the AIDS epidemic are no longer
1 What are the current ethical issues in dental care of HIV-
considered important.
positive patients, do they need to be addressed? – U. Chikte
Oral care of the soft tissues has focused on the oral
2 Do we need to modify the dental care we give HIV-
manifestations classified by the EC Clearinghouse on prob-
positive patients? – D. Croser
lems related to HIV-infection and AIDS (EC-Clearinghouse
3 When is it necessary to give antibiotic prophylaxis to
1993). Opportunistic infections, including bacterial, viral
HIV-positive patients? – A. Schmidt-Westhausen
and mycotic infections, have to be treated adequately.
4 What is the evidence for the effective treatment of oral
The clinical management of selected oral fungal and viral
lesions associated with HIV? – L. McPhail
infections has been reviewed recently (Reichart 1999).
5 What is the most successful palliative treatment for
Within the framework of oral care and treatment of op-
Kaposi’s sarcoma? – P. Reichart
portunistic infections and neoplasias, two principal question
arise: what is an effective treatment in immunodeficiency,
and what is the effect of immune-reconstitution on oral
Correspondence: P Shirlaw, Department of Oral Medicine and Patho-
logy, GKT Dental Institute, Floor 28, Guy’s Tower, Guy’s Hospital, London health? The Workshop addressed some of these questions.
Bridge, London SE1 9RT, UK. Tel: +44 (0)20 79554289, Fax: +44 (0)20 The second topic on which the Workshop was asked
79558790, E-mail: [email protected] to focus was treatment protocols for the management of

Oral Diseases

ODIC25 136 6/12/02, 9:44

 Oral care and treatment protocols in HIV
PJ Shirlaw et al.

137
HIV-infected patients. This topic was considered to be out- treatment that would be administered to HIV-negative pa-
side the responsibility of health care workers treating oral tients. However, the treatment may be provided more slowly
disease of HIV/AIDS patients. The experts as well as the and carefully and may involve greater protective measures
workshop participants felt that it was not appropriate for (Doyal 1997).
either the dental team or the specialists in oral medicine
to prescribe ART or HAART therapy. Nevertheless, it was Does an OHCW have a professional obligation
recognized that anyone working with patients who receive
ART and HAART should know the presently approved to disclose their own HIV status?
antiretroviral drugs, their efficacy, toxicity and the major There is no professional obligation for OHCW to disclose
treatment guidelines. their seropositive status to the public. It is unethical for
OHCWs who are infected with HIV to put patients at risk
What are the current ethical issues in by failing to seek or act upon appropriate counselling. If a
dental care of HIV-positive patients, doctor counsels an HIV-positive colleague and becomes
aware that their advice to modify professional practice is
do they need to be addressed? not followed, they have a duty to inform an appropriate
The emergence of the HIV pandemic has challenged the regulatory body that the OHCW’s fitness to practice may
ethics of all health care professions. The gravity, infectious be seriously impaired.
nature and social stigma of the illness has led to a range of
ethical dilemmas in management and research into oral and
dental care. The recently issued Joint United Nations Program
Research concerns
on AIDS guidance document, Ethical Considerations in HIV There has been a heated debate whether research in develop-
Preventive Vaccine Research (Bloom 1998; Bayer 2000), ing and developed countries should have different standards
has added to other guidelines and these documents form the (Angell 1997). The application of universal ethical guide-
basis for the best standard of ethics in practice and research. lines across cultural boundaries is problematic especially
with ‘first person informed consent’. Difficulties in obtain-
ing this consent in cross-cultural settings may be caused by
Consent a lack of communication and comprehension due to differ-
The oral health care worker (OHCW) can only examine or ences in social status, language, literacy and the concepts of
treat patients who have given their consent. This consent health and disease of the researchers and participants. Des-
should be based on a patient’s voluntary authorization of a pite these problems, the Joint United Nations Program on
dental procedure after being given the relevant information AIDS reached a consensus that each individual must con-
by an OHCW (Odom and Bowers 1993). It rests on the sent to participation (Bayer 2000). Guidelines on the ethics
principle of respect for autonomy, which acknowledges the of research in cross-cultural settings need to be formulated.
ability to understand information, weigh alternatives and form There is also a need to ensure that an appropriate standard
judgements. Social differences and learning disabilities of of counselling, HIV treatment, care and preventive meas-
patients need to be considered when obtaining consent. ures are provided for the patient irrespective of where the
research is conducted.
Consideration should be given to:
Confidentiality • Design of trials;
Autonomy for the patient also entails confidentiality. OHCWs • Risk and consent;
providing treatment may not disclose HIV-related informa- • Beneficial treatment of subjects;
tion and may only share information with the written consent • Cross cultural perspectives;
of the patient. However, confidentiality is not absolute and • Research ethics in a pandemic.
clinical information may be transmitted between health care During the discussion, the difficulty of providing treatment
professionals. while conducting research was highlighted. Participants felt
that research teams cannot screen and treat at the same time
but they must offer pain and sepsis relief or refer patients
Dental duty of care for treatment. However, in some rural areas there are no
OHCWs are under an ethical obligation to treat patients facilities available for referring participants. The participants
within the scope of their clinical practice. Race, colour, creed, suggested that when research is being planned, some provi-
sexual identity and culture should not impinge on treatment sion must be made to address this issue. If the team were
planning (Rule and Veatch 1993). It is unethical to refuse not able to provide treatment they should offer education.
to treat patients based on their HIV status alone. OHCWs
are not expected to provide treatment beyond their areas of Do we need to modify the dental care
expertise and referrals are appropriate as long they are not
pretexts for discrimination or a refusal to treat. When an we give HIV-positive patients?
OHCW becomes aware of the patient’s risk for HIV infection A consensus meeting of experts (ADA, Chicago 1994) stated
or identifies clinical conditions that may be associated with that it was both safe and desirable to make regular dental care
HIV infection, they should refer the patients for counselling available to HIV-positive patients. Generally speaking, no
and testing. Access to dental care should not be linked to modification of treatment was recommended, with the excep-
testing. HIV-positive patients should always receive the same tions noted in Table 1 that apply to the later stages of AIDS.

Oral Diseases

ODIC25 137 6/12/02, 9:44

 Oral care and treatment protocols in HIV
PJ Shirlaw et al.

138
Table 1 Circumstances when routine dental treatment may need to be Individual national resources dictate that every country
modified develops its own response to HIV, while the dental profes-
sion has an obligation to shape and assist with that response
Circumstances necessitating modification from the routine norm: at a local level.
• Low CD4 lymphocyte levels predispose to oral lesions requiring
specific treatment
• Reduced platelet levels below 60 000 cells mm−3 (normal 150 000– When is it necessary to give antibiotic
400 000) effect clotting time
• Reduced neutrophil levels below 500 cells mm−3 (normal 2500–7500) prophylaxis to HIV-positive patients?
may require antibiotic prophylaxis
• Patients with late stage AIDS may require a rolling treatment plan
There are no evidence-based data supporting the need for
with regular reviews of ability to attend and withstand treatment routine antibiotic cover to prevent bacteremia and septicemia
arising from routine dental procedures. Indications for anti-
biotic prophylaxis should not only include CD4+ lymphocyte
count, but in cases where the count is less than 200 µl−1,
The consensus guidelines also advocated a preventive the patient’s history should be reviewed thoroughly prior
approach for the HIV-positive patient: to oral treatment. Antibiotic cover prior to oral surgery or
• Optimize oral hygiene; periodontal treatment is recommended for patients whose
• Establish regular review periods; polymorph (PMNL) count drops below 500 µl−1. Neutro-
• Screen for HIV-related oral lesions and treat if necessary; penia frequently complicates HIV infection and has been
• Screen for xerostomia as a possible symptom of HIV or reported to occur in 35–75% of patients with AIDS (Doweiko
as side-effect of HAART. 1993). Decreased numbers of neutrophils impair defence
By the third Oral AIDS Workshop (in London in 1996) mechanisms against microbial invaders, so neutropenia is
many national dental associations had adopted guidelines a significant risk factor for bacterial infection. However,
advising dentists that a patient living with HIV could be administration of HAART results in a significant increase
safely treated in general practice, provided that universal in circulating hematopoietic lineages, including total white
infection control was observed. Some countries embodied blood cells, lymphocytes, PMNLs and platelets (Huang
this principle in their legislation and made discrimination et al. 2000), and endogenous cytokines (G-CSF and GM-
against an HIV-positive patient an offence. A review of the CSF) that have been found to regulate neutrophil count and
literature published since 1996 supports the earlier views. function (Kuritzkes 2000).
There are reports in the literature of delayed healing fol- When indicated by medical history, antibiotics should
lowing extractions (Robinson et al. 1992; Asseri et al. 1997; be prescribed according to the guidelines of the American
Dobson 1997); poor response to periodontal treatment Heart Association (AHA) to prevent endocarditis (Dajani
(Robinson 1992; Glick et al. 1994); and endodontic failures et al. 1997) or those guidelines recommended by the
(Cooper 1993; Glick 1994). country in which treatment is taking place, if such guide-
As the incidence of these suboptimal results in HIV- lines exist. Patients infected with HIV via intravenous
positive patients is not significantly greater than in the popu- drug use are at risk for developing endocarditis. Additional
lation at large, HIV status alone would not be grounds to antibacterial mouthrinse and scaling may be useful, prior
modify treatment. All the problems that arose responded to to and following any dental procedure (especially oral sur-
subsequent remedial treatment. gery). Until now, no scientific evidence suggests an in-
Concerns about local anaesthetic competing with a creased risk for postsurgical local infections. When they do
patient’s HAART medications for breakdown in the liver occur, oral systemic antibiotics may be prescribed; how-
remain a hypothetical consideration. The benefit of effect- ever, symptoms of postsurgical infections may be altered in
ive pain control would seem to justify the minimal risk of patients with HIV infection, i.e. reduced inflammation and
toxicity. no purulence.
Over the last 15 years the profile of this patient popula- The discussion on this question highlighted the need to
tion has evolved from the chronically sick, frequently dying liaise with the physcians treating the patients, and the need
from their disease, to the present mix of long-term sur- to stabilize the dentition ahead of definitive treatment.
vivors and the chronically well. Access to HAART means
that many HIV-positive patients now present with a func- What is the evidence for the effective
tional immune system and a viral load that is undetect- treatment of oral lesions associated
able. Dental care for these patients requires no particular
modification. In poor communities where HAART is un- with HIV?
available and laboratory tests infrequent, the dentist may not Many oral lesions are associated with HIV/AIDS and
know how ill the HIV-positive patient is. A blood analysis multiple treatment protocols have been developed for each.
can assist in planning treatment. In its absence, generally Oral candidiasis was selected for discussion because it may
the more ill the patient, the more likely the need for one occur in 50% of HIV-infected subjects and 90% of AIDS
of the modifications listed in Table 1. Treatment can be patients (Sweet 1997). The US Food and Drug Admin-
prioritized thus: istration (FDA) have approved nystatin, amphotericin B,
• Relieve pain and treat infection; clotrimazole, ketoconazole, fluconazole and itraconazole,
• Restore function; and each has oral formulations available for outpatient use.
• Prevent further disease; Summarized in Table 2 are results of selected examples
• Consider aesthetics. of studies that assessed the efficacy of these antifungals.

Oral Diseases

ODIC25 138 6/12/02, 9:44

ODIC25
Table 2 Summary of antifungal therapy studies

Study design, drug Study arms Dose Treatment time Results: % success Follow-up time Conclusions
(author, date) (patient number) Clinical (Culture)

Amphotericin B Nystatin 22 000–200 000 U Assessed at 48 h Clear 45 Variable Both were safe and reasonably effective for treatment of
Six small studies (n = 60) 3 or 4 td* Treatment for Improve 37 candidiasis in infants and children
Infants/children 2–21 days Fail 18 Patients re-treated High frequency of
Open label Amphotericins 40 mg Assessed at 48 h Clear 33 relapse due to Candida persistence (Huang et al. 1957–58)
No controls (Amphotericin B, 80 mg Treatment for Improve 39 upon relapse

139
and B + A) 1-28 days Fail 28
(n = 28)
Clotrimazole Clotrimazole 10 mg 5 td 2 weeks Clear 60 (60) 3–10 months Clotrimazole is highly effective in treatment of chronic oral
(Kirkpatrick (n = 10) Improve 30 (30) candidiasis
and Alling 1978) Fail 10 (10)
Randomized Placebo Placebo 5 td Clear 10 (10)
Double blind (n = 10) Improve 0 (0)
Placebo control Fail 90 (90)
Ketoconazole Ketoconazole 200 mg 5 td 3 days Clear 70 (39) None No significant difference between ketoconazole and
versus clotrimazole (n = 22) 7 days Clear 100 (64) clotrimazole in treating oropharyngeal candidiasis
(Thamlikitkul Clotrimazole 10 mg 5 td 3 days Clear 73 (36)
et al. 1988) (n = 23) 7 days Clear 100 (64)
Randomized
Double blind
Comparison
Multicentre
Fluconazole Fluconazole 100 mg 1 td Treated for 2 weeks Symptom clear 85 Endoscopy before and Fluconazole significantly more effective than ketoconazole
versus ketoconazole (n = 87) post symptom Endoscopy clear 52 5 days post-treatment for oesophagitis in AIDS patients
(Laine et al. 1992) clearance (Assessed by symptom resolution and endoscopy
Randomized Ketoconazole 200 mg 1 td (3–8 weeks) Symptom clear 85 clearance)
PJ Shirlaw et al.

Double blind (n = 82) Endoscopy clear 52

Comparison
Multicentre
Fluconazole Fluconazole 100 mg 1 td 14 days Clear 87 (60) 28 & 42 days Fluconazole suspension was more effective than nystatin
suspension versus (n = 69) Improve 12 suspension in treating oropharyngeal candidiasis in HIV/AIDS
nystatin suspension Fail 1 patients

6/12/02, 9:44
(Pons et al. 1997) Nystatin 500 000 U 4 td 14 days Clear 52 (6)
Randomized (n = 69) swish/swallow Improve 16
Oral care and treatment protocols in HIV

Comparison Fail 32
Multicentre
Itraconazole oral Itraconazole 200 mg 1 td 7 days Clear 86 (52) 21 & 35 days Itraconazole is as effective as fluconazole in treating
solution versus (n = 60) Clear 97 (88) oropharyngeal candidiasis in HIV/AIDS patients
fluconazole Itraconazole 200 mg 1 td 14 days Clear 87 (77)
(Graybill (n = 59)
et al. 1998) Fluconazole 100 mg 1 td 14 days
Randomized (n = 60)
Third party blind
Comparison
Multicentre

*td = times daily.

139

Oral Diseases
 Oral care and treatment protocols in HIV
PJ Shirlaw et al.

140
They are listed in chronological order and the evolution the treatment of eKS, one has to differentiate between local
of clinical study methods over time is readily apparent. The versus disseminated lesions of eKS.
original studies on nystatin and amphotericin in the 1950s Recently, an extensive review of Kaposi’s sarcoma (KS)
were open-label trials conducted on a few infants and including history, clinical variants, pathogenesis, epide-
children with a wide variety of predisposing medical condi- miology and treatment has been published (Antman and
tions (Huang et al. 1957–58). These studies lacked blinding Chang 2000). Yarchoan (1999) also reviewed therapy for
of investigators, consistent dosing, control or comparative KS, including recent advances and experimental approaches.
groups, well-quantified end-points and statistical analyses. Although eKS still causes considerable morbidity and
The 1978 study on clotrimazole was a randomized, placebo- mortality in individuals with AIDS, improved tumour con-
controlled trial with appropriate statistical analyses, but sub- trol is presently possible. Also, evidence suggests that the
ject numbers were minimal (Kirkpatrick and Alling 1978). development of HAART has contributed to a decline in
The later studies (1988 onwards) were large, multicentre the incidence of eKS in AIDS patients (Yarchoan 1999;
trials, with double or third party blinding, specific dosing Cattelan et al. 2000). Regression of eKS was observed
and treatment/follow-up time, control or comparative groups, following ritonavir therapy (Diz Dios et al. 1998), and AZT
well-defined end-points and appropriate statistical analyses and HAART (Yarchoan 1999; Antman and Chang 2000).
(Thamlikitkul et al. 1988; Laine et al. 1992; Pons et al. In addition, regression of eKS has been observed after
1997; Graybill et al. 1998). Another study of note (not administration of foscarnet, ganciclovir (Martin et al. 1999)
listed in Table 2 because of space restraints) was a random- and cidofovir and others (Antman and Chang 2000).
ized, multicentre comparison study in HIV/AIDS patients Radiation therapy with application of 15 Gy has also been
(Powderly et al. 1995). They found no significant differ- recommended (Kirova et al. 1998); however, mucositis
ences in efficacy between fluconazole tablets and clotrim- may occur as a side-effect of this treatment modality. Gen-
azole oral troches. erally, patients with eKS respond to radiation therapy and
Intriguingly, nystatin suspension, which was reported in chemotherapy less favourably than patients with classic KS
the early uncontrolled studies (Huang et al. 1957–58), to (Antman and Chang 2000). Intralesional injection (vinbla-
be ‘safe and reasonably effective’ (82% cleared/improved), stine 0.1 mg cm−2) has successfully been used for oral eKS
was much less impressive in the later, rigorous studies in (Reichart et al. 1993).
adults with HIV/AIDS (62% cleared/improved) (Pons et al. Vinca alkaloids (vincristine or vinblastine) were found to
1997) and in immunocompromized children (51% cleared) be effective with acceptable toxicity. The combination of
(Flynn et al. 1995). Furthermore, when compared to flucon- doxorubicin, bleomycin and vinca alkaloids was also shown
azole suspension (99% cleared/improved in adults and to be effective. More recently, liposomal anthracyclines and
91% in children), nystatin suspension was significantly less paclitaxel were developed. Paclitaxel, causing irreversible
effective. polymerisation of cellular microtubules, was found to in-
In summary, clotrimazole, fluconazole and itraconazole hibit growth of KS-derived spindle cell-lines. The response
were shown to be similarly effective and are recommended rate was reported to be between 49% and 71% (Yarchoan
for outpatient treatment of oropharyngeal candidiasis in HIV/ 1999).
AIDS patients. Ketoconazole was less effective in HIV/ In addition to standard therapies, experimental approaches
AIDS patients, apparently because of decreased absorption to treat KS are being developed. These experimental thera-
due to reduced stomach acidity. If used, it should be taken pies include:
with acid food or drink, and its effectiveness closely mon- 1 Anti-angiogenesis approaches
itored. The use of nystatin suspension in HIV/AIDS pati- 2 Pentosan polysulfate
ents, either children or adults, is not recommended. The 3 Tnp 470 (fumagillin)
importance of basing choice of antifungal therapy on rigor- 4 Thalidomide
ously conducted placebo or comparison studies rather than 5 Interleukin-12 (Il 12)
on poorly controlled open-labelled studies is illustrated by 6 Angiostatin
the studies on nystatin suspension. 7 Interferon-α
8 Retinoic acids
What is the most successful palliative 9 Hormonal approaches
10 KHSV-related antiherpes drugs (Yarchoan 1999).
treatment for Kaposi’s sarcoma?
There are no curative therapies known, and no comparative
studies of different treatment modalities available; however,
Considerations
treatment modalities for epidemic Kaposi’s sarcoma (eKS) Treatment of eKS is complicated by problems of assessment
may be divided into four groups: of tumour response and staging. In addition, the effects of
1 Local measures ART and HAART therapy on eKS have to be studied in
2 Interferon α (INF α) more detail. Discussion on this question highlighted the
3 Cytotoxic chemotherapeutic agents different positions of those clinicians working in areas where
4 HAART (Cattelan et al. 2000) HAART is routinely available, because of the effect on
Local therapies include radiation therapy, cryotherapy, lessening KS.
intralesional injections and topical application of various In resource-poor countries the main options still remain
drugs. Epstein has recently proposed a clinical staging of oral between surgery, radiotherapy or chemotherapy, as liposomal
eKS and its management (Epstein 1997). When considering anthracyclines are far too expensive to be used. Some

Oral Diseases

ODIC25 140 6/12/02, 9:44

 Oral care and treatment protocols in HIV
PJ Shirlaw et al.

141
African countries have found reasonable partial remission Principally, a reduction of HIV-1 replication and pre-
rates with single chemotherapeutic agents, for example vention of CD8+ T cell exhaustion occurs. Problems with
vincristine, with no adverse effect and which requires little ddI and HU therapy are toxicity (sensory neuropathy),
monitoring. Palliative care is of importance as advanced pancreatitis and occasionally minor ulceration of the mouth
disease is extremly painful, often accompanied by exten- (Donehower 1992).
sive oedema. Analgesia in resource-poor countries may be
the only treatment available. What is the influence of protease
inhibitors and HAART on the excretion
Can we provide clinical treatment that of HIV in oral fluids?
has a scientific basis rather than being
HIV viral RNA in saliva has only been detected at relatively
trial based? low titres (~102 RNA copies ml−1). Direct comparisons of
The short answer to this question is no. ‘Scientific basis’
involves construction and testing of hypotheses, and when Table 3 Summary of study designs proposed by the American College
the research hypothesis involves clinical treatment for a of Physicians
human disease, a carefully designed trial of the treatment is
inevitable. However, many trials are conducted every year EVIDENCE BASED MEDICINE
throughout the world and different trials of supposedly the The American College of Physicians – American Society of Internal
same hypothesis can have divergent or conflicting out- Medicine (ACP-ASIM). Condensed version of statements at web site
(http://www.acponline.org/journals/ebm/sepoct99/purpproc.htm)
comes. How to evaluate the outcomes from various trials is
Prevention or treatment studies
an ongoing challenge. Since the mid-1990s, meta-analysis • Randomization of participants to different interventions
(Ioannidis and Lau 1999), a systematic and quantitative • Outcomes of clinical importance measured for ≥ 80% of participants
synthesis of data from multiple sources, has been proposed enrolled
as a tool to evaluate published treatment trials with the goal • Analysis consistent with the study design
of making a summary recommendation for clinical practice. Diagnostic test studies
The advantages of meta-analysis are its ability to increase • Clearly identified comparison groups, at least one free of target
disease or disorder
the power of conclusions from smaller or inconclusive stud- • Objective diagnostic standard (machine-produced laboratory result)
ies, to analyse the extent and origins of diversity among the • Clinical diagnostic standard with reproducible criteria for subjectively
studies, and to predict how heterogeneity among patient interpreted components (> chance agreement among interpreters)
groups might affect the outcomes of medical interventions. • Interpretation of test without knowing diagnostic standard result
A fundamental limitation is that meta-analysis cannot im- • Interpretation of diagnostic standard without knowing test result
• Analysis consistent with the study design
prove the basic accuracy or quality of the observations or their
Prognosis studies
reporting. Additionally, the design of the meta-analysis • An inception cohort of persons, all initially free of outcome
itself must account for the diversity among subject groups, • Follow-up of ≥ 80% of patients until major study end point or end of
variations in data quality and potential biases of study design study
or reporting. • Analysis consistent with the study design
During the workshop, Laurie MacPhail presented the sche- Causation studies
mata from the American College of Physicians and American • Clearly identified comparison group for those at risk for, or having,
the outcome of interest (whether from randomized, quasi-randomized,
Society of Internal Medicine on the methodology for trials or non-randomized controlled trials; cohort analytic studies with case-
that will give results that can be used when compiling treat- by-case matching or statistical adjustment to create comparable
ment protocols based on evidence of efficacy (Table 3). groups; or case-control studies)
• Masking of observers of outcomes to exposures (this criterion is
assumed to be met if the outcome is objective [e.g. All-cause mortality
Is ddI + hydroxyurea an effective African or an objective test])
alternative to HAART? • Observers of exposures masked to outcomes for case-control studies
• Masking of subjects to exposure for all other study designs
None of the experts and participants of the workshop had • Analysis consistent with the study design
any experience with the administration of ddI and hydroxy- Economics of health care programmes or interventions studies
urea (HU). It was felt that this question was beyond the • Compare alternative courses of action
expertise of the participants of the workshop. However, oral • Alternative diagnostic or therapeutic services or quality improvement
strategies must be compared on the basis of both the outcomes they
health care providers should be aware of this particular type produce (effectiveness) and the resources they consume (costs)
of antiretroviral therapy (Lori and Lisziewicz 1998; Dybul • Evidence of effectiveness must come from a study (or studies)
et al. 2000; Lori and Lisziewicz 2000). HU is a chemother- meeting above criteria for diagnosis, treatment, quality improvement
apeutic agent inhibiting the DNA synthesis of HIV and pot- • Results must be presented in terms of the incremental or additional
costs and outcomes incurred and realized by one intervention over
entiates nucleotide reverse transcriptase inhibitors (NRTIs). another
There is also synergism with ddI (Lori and Lisziewicz 1998; • Sensitivity analysis must be done where there is uncertainty or
Moore et al. 2000). This particular combination therapy imprecision in the estimates or measurements
results in: Clinical prediction guides
1 Suppression of HIV-1 replication. • Generated in one set of patients (training set)
2 DdI-resistant HIV-1 strains retain sensitivity to combina- • Validated in an independent set of patients (test set)
• Meet the above-noted criteria for treatment, diagnosis, prognosis or
tion of ddI and HU. causation
3 Cytostatic effects on CD4+ and CD8+ lymphocytes.

Oral Diseases

ODIC25 141 6/12/02, 9:44

 Oral care and treatment protocols in HIV
PJ Shirlaw et al.

142
studies are hampered by variations in saliva source, type of
collection and patient’s treatment. Elevated viral loads in
Significance of these studies for the oral mucosa
saliva have been found in cases of periodontal diseases (Levy As a result of HAART, proper antigen processing and
and Greenspan 1988; Shugars et al. 2000). However, no antigen presenting in APC and transport to the local lymph
correlation was found between periodontal diseases and in- nodes may be achieved, resulting in an activation and
fectious HIV in saliva, due to various antiviral components differentiation of T cells. Also, HAART may restore the
(Barr et al. 1996). function of DCs. Studies on blood DCs show that the pro-
In a cross-sectional study of 26 HIV-positive patients, portion of DCs is in the normal range in treated patients
viral RNA was found in the blood of all 26 patients at a with undetectable viral load. However, incompetent DCs
mean concentration of 1 × 104 ml−1, in the semen of five persist, indicating that HAART might only in part correct
out of 26 patients at a mean concentration of 5 × 102 ml−1 the alterations of circulating DCs (Grassi et al. 1999).
and in the whole saliva of 24 out of 25 patients at a mean
concentration of 1 × 102 ml−1. In this study the viral load Conclusions
in either plasma, semen or saliva was not correlated to
antiretroviral therapy with AZT, but the viral load in plasma It was noted that there is a paucity of randomized controlled
and saliva was correlated to CD4+ lymphocyte count. trials related to dental treatment and treatment of oral lesions
Analysis of HIV viral RNA in whole saliva of 40 HIV- associated with HIV that can be used to synthesize valid
positive patients has been studied (Shugars et al. 2000). evidence-based clinical guidelines. The current guidelines in
Viral RNA was found in 17 out of the 40 patients at a mean use have therefore been synthesized by consensus. Although
concentration of 1 × 103 ml−1. Eleven of the patients were this may be necessary in the face of lack of evidence, the
on ART and 20 were on ART plus a protease inhibitor (PI). validity of such guidelines may be poor and therefore there
Use of at least one antiretroviral medication with or without is a need for high quality clinical research, and the guide-
a protease inhibitor was associated with a significant lower lines of the Cochrane collaboration are to be commended
salivary viral load in comparison with no antiretroviral treat- (http://www.cochrane.org).
ment, and the presence of HIV-associated oral lesions was
not associated with viral load in either saliva or plasma References
(except for periodontal lesions). The effect of HAART on Angell M (1997). The ethics of clinical research in the third world.
the infectivity of saliva has not been studied. N Engl J Med 337: 847–849.
Antman K, Chang Y (2000). Kaposi’s sarcoma. New Engl J Med
What is the effect of anti-HIV therapy on 342: 1027–1038.
Asseri L, Nguvumali HI, Matee MI, et al. (1997). Chronic osteo-
the oral mucosa and oral health? myelitis of the mandible following tooth extraction in HIV
No specific studies on the direct effect of HAART on the infection. Oral Dis 3: 193–195.
normal oral mucosa are available. There are studies on in- Barr CE, Qureshi MN, Qiu Z, Kaim J, Zhang W (1996). Oral
testinal mucosa, and these in vitro studies have shown that HIV-I recovery in the presence of periodontal disease. Oral Dis
2: 198–201.
HIV replication occurs in a mucosal dendritic cell (DC)/T
Bayer R (2000). Ethical challenges of HIV vaccine trials in devel-
cell milieu (an environment in which DCs and T cells can oped nations: conflict and consensus in the international arena.
continously interact). It is known that HIV infection is AIDS 14: 1051–1057.
associated with alterations of the oral mucosal immune sys- Bloom B (1998). The highest attainable standard: ethical issues in
tem: lack of CD4+ cells, and impairment of MHC-class II AIDS vaccines. Science 279: 186–188.
expression of Langerhans cells (LHC) in epithelium and Cattelan AM, Calbrò ML, Gasperini P, et al. (2000). Acquired
lamina propria (Romagnoli et al. 1997; Reichart et al. 2000). immunodefiency syndrome-related Kaposi’s sarcoma regression
after highly active antiretroviral therapy: biologic correlates of
clinical outcome. J Natl Cancer Inst Monographs 28: 44–49.
Studies on the effect of HAART on draining EC-Clearinghouse (1993). EC-Clearinghouse on oral problems
lymph nodes related to HIV infection, WHO Collaborating Centre on oral
manifestations of the human immunodeficiency virus. Classifica-
During periods of diminished or undetectable plasma tion and diagnostic criteria for oral lesions in HIV-infection.
viraemia, lymph nodes are a site of active viral replication. J Oral Pathol Med 22: 289–291.
HIV RNA has been identified in follicular dendritic cells Cooper H (1993). Root canal treatment on patients with HIV
(DC) and lymph node mononuclear cells from persons re- infection. Intl Endodo J 26: 369–371.
ceiving combination NRTI (nucleotide reverse transcriptase Dajani AS, Taubert KA, Wilson W, et al. (1997). Prevention of
inhibitor) therapy, despite undetectable plasma viraemia bacterial endocarditis: recommendations by the American Heart
(Ruiz et al. 1999). Persons receiving ART + PI often have Association. JADA 128: 1142–1151.
levels below detectability of HIV RNA in their lymph nodes Diz Dios P, Hermida AO, Alvarez CM, et al. (1998). Regression
(Tenner-Racz et al. 1998). Virus isolated from lymph node of AIDS-related Kaposi’s sarcoma following ritonavir therapy.
Oral Oncol 34: 236–238.
mononuclear cells of persons receiving NRTI showed evid-
Dobson TB (1997). HIV satus and risk of post-extraction com-
ence of genetic mutations equal to or greater than active plications. J Dent Res 76: 1644–1652.
viral replication. Persons treated with ART + PI or efavirenz Donehower R (1992). An overview of the clinical experience with
(NNRTI, non-nucleotide reverse transcriptase inhibitor) do hydroxyurea. Semin Oncol 19 (Suppl. 9): 11–19.
not have evidence of genetic mutations (Ruiz et al. 1999; Doweiko JP (1993). Management of the haematological manifes-
Dybul et al. 2000). tations of HIV disease. Blood Rev 7: 121–126.

Oral Diseases

ODIC25 142 6/12/02, 9:44

 Oral care and treatment protocols in HIV
PJ Shirlaw et al.

143
Doyal L (1997). Good ethical practice in the dental treatment of Moore RD, Wong WM, Keruly JC (2000). Incidence of neuropathy
patients with HIV/AIDS. Oral Dis 3: S214–S220. in HIV-infected patients on monotherapy versus those on com-
Dybul M, Chun TW, Douglas J, et al. (2000). Evaluation of lymph bination therapy with didanosine, stavudine and hydroxyurea.
node virus burden in human immunodeficiency virus-infected AIDS 14: 273–278.
patients receiving efavirenz-based protease inhibitor-sparing Odom JG, Bowers DF (1993). Informed consent and refusal. In:
highly active antiretroviral therapy. J Infect Dis 181: 1276– Weinstein, B, ed. Dental Ethics. Lea & Febiger, Philadelphia,
1279. pp. 65–79.
Epstein JB (1997). Management of oral Kaposi’s sarcoma and a Pons V, Greenspan D, Lozada-Nur F, McPhail L, Gallant JE,
proposal for clinical staging. Oral Dis 3 (Suppl. 1): 124–128. Tunkel A, et al. (1997). Oropharyngeal candidiasis in patients
Flynn PM, Cunningham CK, Kerkering T, et al. (1995). Oro- with aids: randomized comparison of fluconazole versus nystatin
pharyngeal candidiasis in immunocompromised children: a oral suspensions. Clin Infect Dis 24: 1204–1207.
randomized, multicenter study of orally administered fluconazole Powderly WG, Finkelstein D, Feinberg J, et al. (1995). A
suspension versus nystatin. The multicenter fluconazole study randomized trial comparing fluconazole with clotrimazole troches
group. J Pediatr 127: 322–328. for the prevention of fungal infections in patients with advanced
Glick M (1994). Dental Management of Patients with HIV. Quin- human immunodeficiency virus infection. Niaid aids clinical
tessence Publishing Co. Ltd, Carol Stream, IL. trials. N Engl J Med 332: 700–705.
Glick M, Abel S, Muzyka B, Delorenzo M (1994). Dental complica- Reichart PA (1999). Clinical management of selected oral fungal and
tions after treating patients with AIDS. JADA 125: 296–301. viral infections during HIV-disease. Int Dent J 49: 251–259.
Grassi F, Hosmalin A, McIlroy D, et al. (1999). Depletion in Reichart PA, Langford A, Pohle HD (1993). Epidemic oro-facial
blood CD11c-positive dendritic cells from HIV-infected patients. Kaposi’s sarcoma (eKS) – report on 124 cases. Oral Oncol, Eur
AIDS 13: 759–766. J Cancer 29: 187–189.
Graybill JR, Vazquez J, Darouiche RO, et al. (1998). Randomized Reichart PA, Samaranayake LP, Philpsen HP (2000). Pathology
trial of itraconazole oral solution for oropharyngeal Candidiasis and clinical correlates in oral candidiasis and its variants: a
in HIV/AIDS patients. Am J Med 104: 33–39. review. Oral Dis 6: 85–91.
Huang SS, Barbour JD, Deeks SG, et al. (2000). Reversal of human Robinson P (1992). Periodontal disease and HIV infection: a
immunodeficiency virus type 1-associated hematosuppression review of the literature. J Clin Periodont 19: 609–614.
by effective antiretroviral therapy. Clin Inf Dis 30: 504–510. Robinson P, Cooper H, Hatt J (1992). Healing after dental extrac-
Huang NN, Sarria A, High RH (1957–58). Therapeutic evalua- tions in men with HIV infection. Oral Surg Oral Med Oral
tion of nystatin and amphotericin in oral moniliasis in infants Pathol 74: 426–430.
and children. Antibiotics Annual 00: 59–64. Romagnoli P, Pimpinelli N, Mori M, et al. (1997). Immuno-
Ioannidis JP, Lau J (1999). Pooling research results: Benefits and competent cells in oral candidiasis of HIV-infected patients:
limitations of meta-analysis. Jt Comm J Qual Improv 25: 462– an immunohistochemical and electron microscopical study. Oral
469. Dis 3: 99–105.
Kirkpatrick CH, Alling DW (1978). Treatment of chronic oral Ruiz L, van Lunzen J, Arno A, et al. (1999). Protease inhibitor
candidiasis with clotrimazole troches. A controlled clinical trial. containing regimens compared with nucleoside analogues alone
N Engl J Med 299: 1201–1203. in the suppression of persistent HIV-1 replication in lymphoid
Kirova YM, Belembaogo E, Frikha H, et al. (1998). Radiotherapy in tissue. AIDS 13: F1–F8.
the management of epidemic Kaposi’s sarcoma: a retrospective Rule J, Veatch R (1993). Ethical Questions in Dentistry. Quintes-
study of 643 cases. Radiother Oncol 46: 19–22. sence Publishing Co. Ltd, Chicago, pp. 151–162.
Kuritzkes DR (2000). Neutropenia, neutrophil dysfunction, and Shugars DC, Slade GD, Patton LL, Fiscus SA (2000). Oral and
bacterial infection in patients with human immunodeficiency systemic factors associated with increased levels of human
virus disease: the role of granulocyte colony-stimulating factor. immunodeficiency virus type 1 RNA in saliva. Oral Surg Oral
Clin Inf Dis 30: 256–260. Med Oral Pathol Oral Radiol 342: 432–440.
Laine L, Dretler RH, Conteas CN, Tuazon C, Koster FM, Sattler Sweet SP (1997). Selection and pathogenicity of Candida albicans
F, et al. (1992). Fluconazole compared with ketoconazole for in HIV infection. Oral Dis 3 (Suppl. 1): S88–S95.
the treatment of candida esophagitis in AIDS. A randomized Tenner-Racz. K, Stellbrink HJ, van Lunzen J, et al. (1998). The
trial. Ann Intern Med 117: 655–660. unenlarged lymph nodes of HIV-1-infected, asymptomatic pati-
Levy JA, Greenspan D (1988). HIV in saliva. Lancet 2: 1248. ents with highCD4 T cell counts are sites for virus replication
Lori F, Lisziewicz J (1998). Hydroxyurea: mechanisms of HIV-1 and CD4 T cell proliferation. The impact of highly active
inhibition. Antivir Ther 3 (Suppl. 4): 25–33. antiretroviral therapy. J Exp Med 187: 949–959.
Lori F, Lisziewicz J (2000). Rationale for the use of hydroxyurea Thamlikitkul V, Danpakdi K, Visuthisakchai S, Kulpreedarat T
as an anti-human immunodeficiency virus drug. Clin Infect Dis (1988). Randomized controlled trial of clotrimazole troche and
30: 193–197. ketoconazole for oropharyngeal candidiasis. J Med Assoc Thai
Martin DF, Baruch D, Wolitz RA et al. (1999). Oral ganciclovir 71: 654–658.
for patients with cytomegalovirus retinitis treated with a gan- Yarchoan R (1999). Therapy for Kaposi’s sarcoma: recent advances
ciclovir implant. N Engl J Med 340: 1063–1070. and experimental approaches. JAIDS 21: 566–573.

Oral Diseases

ODIC25 143 6/12/02, 9:44