Scientific Papers

Oral health outcomes in an HIV cohort with comorbidities- implementation roadmap for a longitudinal prospective observational study | BMC Oral Health

Study hypothesis

The central hypothesis is that the extent and progression of key oral diseases (e.g., caries, periodontal disease, and oral cavity confections) are associated with ART-induced salivary changes, and is exacerbated in PWH who have been on ART for at least one year and have developed non-communicable diseases (+ NCDs) that include diabetes, cardiovascular diseases, osteoporosis, and hyperlipidemia after ART initiation. PWH without NCDs (-NCDs) serve as control group (Fig. 1). The secondary hypothesis is that patients in the + NCD group will present with higher degrees of salivary gland hypofunction (e.g., dry mouth); greater changes in levels of salivary proteins, as well as higher levels of salivary inflammatory and immune activation cytokines. We further included an exploratory hypothesis, that BMD changes noted on PAN are predictors of osteoporosis risk when compared to the gold-standard DEXA scans.

Fig. 1
figure 1

Study Cohort: Patients with HIV on ART (≥ 1 year). Group A: Participants with no comorbidities. Group B: Participants with ≥ 1 comorbid disease diagnosed after ART initiation

Subject recruitment, timeline and inclusion and exclusion criteria

Target enrollment number is 350, with a study duration of 24 months (2 years). Study visits and data collection is scheduled every 6 months for PWH above the age of 18 years, who have been on ART drugs for at least 12 months. Participants with xerostomia-related autoimmune conditions like Sjögren’s syndrome, and sarcoidosis; head and neck radiation therapy (including radioactive iodine therapy); chemotherapy within the prior year before enrollment in the study; use of anti-osteoporotic agents (Bisphosphonates, Denosumab); and pregnancy at time of enrollment, are excluded.

Participants are categorized into two groups of + NCD and -NCD patients as illustrated in Fig. 1.

Eligibility is confirmed by review of medical/dental records of patients with a diagnosis of HIV and eligible subjects are contacted by telephone or approached during clinical care. Written informed consent is obtained at the baseline study visit for interested participants, after thorough review of all study visit procedures in layman terms and answering questions from participants. The study was carefully designed to include only PWH. Also, PWH who have not commenced ART drugs and HIV-uninfected individuals are excluded, because the study focus is on evaluating the effect of ART on oral and systemic health outcomes. This approach allows for assessment of onset (when possible) and progression of oral and systemic diseases.

Participants who have been on ART for less than one year are also excluded because it is necessary to capture onset and progression of ART related non-communicable conditions. The rationale for this timeline was based on current research suggesting that most NCDs are diagnosed approximately 12–24 months after ART initiation [32].

Subjects are followed-up for 2 years to compare findings on PAN with those on DEXA scans, especially findings that relate to changes in BMD. The goal is to ultimately determine if PAN is a sensitive tool for an early diagnosis of bone disease among PWH in a dental setting.

Study procedures

Clinical data collection

The following demographic data are collected: age, biological sex, gender identity, sexual preferences, ethnicity, and race. Medical history captured includes past and current medical conditions, and currently prescribed and over-the-counter medications.

Medical and oral health questionnaire

Three standardized questionnaires used on this study protocol include: Oral health, xerostomia-related quality of life (XeQoLs) [33, 34] using the National Health and Nutrition Examination Survey (NHANES) 2012 [35] and modified XeQoLS questionnaire [36] respectively; and Social history using the Tobacco, Alcohol, Prescription medication, and other Substance use (TAPS) tool [37].

Standard of Care (SOC) Clinical Labs

CD4+ cell count, viral load, and lipid panel (including high-density lipoproteins (HDL), low-density lipoproteins (LDL), and total cholesterol levels) are abstracted from medical records. Point of care glucose levels is also assessed chair side.

Vital signs and physical measurements

Vital signs obtained include weight, temperature, blood pressure and heart rate. Height is measured at baseline only.

Stimulated saliva collection

Stimulated saliva is preferably obtained in the morning due to the circadian pattern of saliva production. Participants are given a 1.25 × 0.25 inch Parafilm cube from a pre-weighed 50 mL conical tube and instructed to chew continuously and avoid swallowing for 5 min. An autoclaved funnel is placed inside the conical tube, with the tube embedded in wet ice. At the end of each minute, participants are asked to gently spit pooled saliva into the funnel. After 5 min, the remaining saliva and parafilm are expectorated into the funnel. The tube is then transferred to the laboratory for preprocessing in a cool box.

Saliva flow rate is determined based on comparing pre and post saliva collection weights of the conical tube. The saliva is also centrifuged at 4 °C, 2,500 rpm for 25 min to separate supernatant and pellet. Pierce Protease Inhibitor cocktail EDTA-free is added at 1uL/100uL of saliva supernatant for up to 2 mL of saliva. Saliva pellet is also transferred into a cryovial tube containing 500ul of DNA/RNA Shield® Solution. Saliva supernatant with and without protease inhibitor and the saliva pellet are stored at -800C. Expression levels of inflammatory and immune activation cytokines in saliva supernatant will be measured using Polymerase chain reaction (PCR), while pellets retrieved are stored for future microbiome and genome studies.

Blood sample collection

Patients are asked to fast for approximately 8 h prior to collection of blood samples. Patients with diabetes, especially those on insulin, are prioritized for the first visit of the day.

A glucometer is used to assess point of care fasting blood glucose. In addition, approximately 22 mL of blood is collected by venipuncture in the following order: A one-time collection of 2 PAXgene™ blood RNA tubes for future genomics studies. This is preceded by collecting approximately 1 mL of blood in a discard tube; one EDTA tube (purple top) for collection of plasma samples for analysis of inflammatory and immune activation markers; and two Serum separator tubes (tiger top) for collection of serum sample for analysis of inflammatory and immune activation markers, and the other for assessment of total insulin levels.

The plasma samples are preprocessed within 30 min of sample collection, while one of the two tubes containing serum is preprocessed after 30 min. Samples are stored at -80 °C. Expression levels of inflammatory and immune activation cytokines are assessed by PCR.

PAXgene™ tubes are stored at room temperature for 2 h, transferred to -200C freezer for 72 h and up to 30 days before long term storage at -800C for future genomic studies.

Oral health assessment

Teeth with decayed, missing and filled surfaces (DMFS) are recorded, and DMFS index is used to determine participants’ caries experience. While decayed teeth show the presence of active caries in the mouth, missing teeth could imply tooth loss due to periodontal disease, from non-restorable caries or other factors.

Periodontal probing depth, clinical attachment loss and bleeding indices are also assessed to evaluate periodontal health status. The American Dental Association and American College of Cardiology guidelines is followed regarding antibiotic prophylaxis for participants [38].

An intraoral soft tissue assessment is performed by calibrated examiners to determine the presence or absence of oral mucosal lesions, such as candidiasis, red/white lesions, warts, and ulcers. The oral cavity is also examined for xerostomia by objective assessment of dry oral cavity, stickiness of dental mirror to the oral mucosa, and reduced or loss of saliva flow from the Stenson and Warton ducts bilaterally. All dental examinations are performed according to SOC guidelines.


Whole body DEXA measurement, which includes femoral neck (hip) and lower spine, is performed using the Horizon A Platform (Hologic Inc.; Bedford, MA) with Apex software v5.5, to estimate BMD. T and Z-scores will be measured using the NHANES database SD [39].

Panoramic Radiographs (PAN)

PAN are captured with Planmeca Promax® 2D S3 series (Promeca Corporation, Helsinki Finland) to identify overall image radiographic density as well as changes associated with osteopenia. Such changes include thinning of the mandibular cortical bone (lower border of the mandible) and presence or absence of the trabecular bone. The goal is to compare PAN data with T and Z-scores of DEXA scans with respect to risk for osteoporosis and/or osteopenia.

Oral swabs

Swabs are taken from the dorsum of the tongue and any other intraoral site with suspected candidiasis, using the Isohelix™ buccal swabs. The cotton end of the swab stick is then placed in a 2 mL tube containing DNA/RNA shield®. Samples are vortexed for 30 s and stored at -80 °C, for future quantitative analysis using PCR to detect the presence of candida species.

Urine pregnancy test

Urine pregnancy test is performed on women of childbearing potential prior to each imaging study, unless they are over the age of 45 and had not had a menstrual period for at least 12 months prior to enrollment. If female participants are noted to be pregnant at the follow up visit which requires imaging, they are allowed to continue with the study visit but imaging studies are excluded.

Visit schedule

A schematic representation of the visit schedule is shown in Fig. 2. SOC laboratory study reports are reviewed at every study visit. If results within 12 months of study visits are unavailable, new tests are requested from the patients’ primary care or infectious disease physician. Confirmation of HIV diagnosis is by SOC blood tests. Screening and baseline visits are usually performed on the same day.

Fig. 2
figure 2

Visit Schedule: Shows all study procedures per study visit

Participants who chose to withdraw from the study are not required to attend an in-person withdrawal visit. The reason for withdrawal, if known, is documented in the source documents and study database. Protocol deviations including missed study visits are well documented in the protocol deviation form, and adverse events are assessed and appropriately documented.

Data capture

Data for our study is captured using Research Electronic Data Capture (REDCap version 13.8.0 developed by Vanderbilt University), an electronic data capture, management and reporting tool, under the management of the Clinical Research Computing Unit (CRCU), School of Biostatistics and Epidemiology University of Pennsylvania, with restricted access to authorized personnel and study team members only.

Sample size and statistical analysis

The power/sample size assessment was based on the primary analysis for comparing prevalence of periodontal diseases between the + NCD and -NCD groups. From the pilot sample, the probability of periodontal disease was 0.203 and 0.373 for the -NCD and + NCD groups respectively, and the estimated odds ratio (OR) for periodontal disease for -NCDs subjects relative to + NCDs subjects was OR = 2.33 [(0.373/(1–0.373)/(0.203/(1–0.203)].

The required sample size was determined based on a primary analysis using logistic regression; a baseline probability of periodontal disease of 0.203 (for -NCD subjects); 50% exposure prevalence; and power of 80% using a significance level of 0.05. A sample size of 302 allows detection of an OR of 2.1 with 80% power. To accommodate an anticipated drop-out rate of approximately 10–12%, the target sample size was increased to n = 350, to allow for a 13.7% dropout rate, while maintaining the afore-described power.

For descriptive analysis, comparison of continuous variables will be based on either t-tests or the Mann–Whitney test, while comparison of binary variables will use either the Chi- Square test or the Fisher’s exact test, depending on assumptions for each test. Primary statistical analysis will use Logistic regression to assess the odds ratio for the outcome for subjects in the + NCD group relative to subjects in the -NCD group. Data analysis will be performed with Stata SE v16 and SAS v9.4.

Source link