Abstract
Objectives
The Stopping Cavities Trial investigated effectiveness and safety of 38% silver diamine fluoride in arresting caries lesions.
Materials and methods
The study was a double-blind randomized placebo-controlled superiority trial with 2 parallel groups. The sites were Oregon preschools. Sixty-six preschool children with ≥1 lesion were enrolled. Silver diamine fluoride (38%) or placebo (blue-tinted water), applied topically to the lesion. The primary endpoint was caries arrest (lesion inactivity, Nyvad criteria) 14–21 days post intervention. Dental plaque was collected from all children, and microbial composition was assessed by RNA sequencing from 2 lesions and 1 unaffected surface before treatment and at follow-up for 3 children from each group.
Results and conclusion
Average proportion of arrested caries lesions in the silver diamine fluoride group was higher (0.72; 95% CI; 0.55, 0.84) than in the placebo group (0.05; 95% CI; 0.00, 0.16). Confirmatory analysis using generalized estimating equation log-linear regression, based on the number of arrested lesions and accounting for the number of treated surfaces and length of follow-up, indicates the risk of arrested caries was significantly higher in the treatment group (relative risk, 17.3; 95% CI: 4.3 to 69.4). No harms were observed. RNA sequencing analysis identified no consistent changes in relative abundance of caries-associated microbes, nor emergence of antibiotic or metal resistance gene expression. Topical 38% silver diamine fluoride is effective and safe in arresting cavities in preschool children. Clinical Significance: The treatment is applicable to primary care practice and may reduce the burden of untreated tooth decay in the population.
1
Introduction
Dental and medical care providers are increasingly involved in early interventions to prevent tooth decay in young children . The main preventive service is the topical application of fluoride varnish every 3 to 6 months. The rationale is that fluoride treatments plus anticipatory guidance may prevent the onset of tooth decay at the time of greatest risk. This is important because tooth decay negatively impacts quality of life , particularly for children living in poverty, who have less access to dentists. Despite increased attention on young children, not all receive preventive treatments and large numbers still experience tooth decay. Primary care providers have had little option other than invasive and costly specialist treatment in the hospital under general anesthesia.
In this context, there is interest in simple treatments to halt progress of cavities (caries lesions) after tooth decay onset. Topical silver diamine fluoride is a clear liquid that is painted on the active lesion surface in milligram amounts and arrests the lesion. Cleared by the Food and Drug Administration in 2014 as a treatment for sensitive teeth and used off-label for the treatment of cavities in the United States since 2015, 12 clinical trials outside the United States have documented caries arrest . Moreover, preventive benefits extend to unaffected teeth . Serum concentrations of fluoride and silver after topical application revealed no potential toxicity .
Silver ions are assumed to be primarily responsible for the antimicrobial action of silver diamine fluoride. Silver ions inhibit the growth of all tested oral bacteria, and denature enzymes that would breakdown collagenous dentin. Streptococcus mutans , a primary pathogen in dental caries, is less able to form a biofilm on teeth treated ex vivo . Fluoride promotes deposition of fluoroapatite, which is more resistant to acidic degradation than normal tooth structure .
The purpose of this randomized controlled trial was to investigate the short-term effectiveness and safety of topically applied 38% silver diamine fluoride in arresting caries lesions in primary teeth.
2
Methods
The study design was a double-blind randomized placebo-controlled superiority trial with two parallel groups. The study setting was three Head Start programs (preschools) in the U.S. state of Oregon. The participants were children 24 to 72 months of age who had at least one untreated cavitated active caries lesion with dentin exposed based on the Nyvad criteria (level 3: “Enamel/dentin cavity easily visible with the naked eye where the surface of cavity feels soft or leathery on gentle probing.”) . The children were healthy by medical history. The exclusion criteria were weight <15 kg, known sensitivity to silver or other heavy-metal ions, presence of any gingival or perioral ulceration or stomatitis or presence of a tooth abscess. The weight exclusion was to keep the applied dose in the range of the minimum risk dose for chronic exposure to silver determined by the U.S. Centers For Disease Control and Prevention. The participant flow is outlined in Fig. 1 . Recruitment began February 25, 2016 and the final child was evaluated on May 10, 2016.
The study was conducted under IND 124808 from the Food and Drug Administration. The Western Institutional Review Board approved the protocol and the informed consent of participants was obtained. A $50 gift card was provided after follow-up.
2.1
Treatment conditions
The test drug was silver diamine fluoride (Advantage Arrest™, Elevate Oral Care LLC., West Palm Beach, FL). The drug is a transparent blue-tinted solution containing 38 w/v% of silver (Ag) and 5.5 w/v% of fluorine (F). The placebo was non-fluoridated sterile water with a blue tint. The affected tooth surface was gently cleaned and dried with cotton gauze. The gingival tissue of the tooth was protected with petroleum jelly. An applicator was dipped into the agent and 3–4 mg applied to the lesion (1 drop treats 3 to 5 teeth). No rinse was performed or special instructions given following application.
2.2
Plaque sample collection
Before treatment and at the beginning of the follow-up examination, plaque samples were taken from up to 2 caries lesions and one unaffected tooth surface, selected sequentially from a list of tooth identifiers randomized for each patient, by rubbing a 2-mm diameter plastic applicator on the lesion surface for 4 s. The applicator was then placed in 250 μL tryptic soy broth with 15% glycerol and 2% sucrose, inverted twice, and kept at −20° C up to 2 months until moving to −80° C to await processing.
2.3
Allocation
Participants were randomly assigned to groups with a 1:1 allocation using a computer-generated randomization schedule with stratification by Head Start program and randomly permuted blocks of sizes 2 and 4.
2.4
Blinding
The dental providers who applied the treatments, participants, and examiners were blinded. Blinding was maintained by dispensing the test and comparator liquids from similar bottles. The test and placebo bottles were assigned codes A, B, C or D with 2 codes for test and 2 codes for placebo. Four bottles were used to strengthen the blinding of treatment, and were otherwise identical. The biostatistician knew only which two labels were the same, but was blinded to which labels were the test or placebo conditions.
2.5
Outcomes
The primary outcome was caries lesion arrest. The hypothesis was that silver diamine fluoride is more effective than placebo in arresting caries lesions with dentin exposed in primary teeth. A visual/tactile examination was performed before and 14 to 21 days after treatment to identify and assess lesions for caries activity . The teeth were examined with artificial light but without magnification in the preschool setting after drying with cotton gauze. A change from a Nyvad lesion code 3 to 6 (“Enamel/dentin cavity easily visible with the naked eye; surface of cavity may be shiny and feels hard on probing with gentle pressure.” ) was considered a positive outcome: caries arrest. The criteria have been shown to be reliable and valid . The 14 to 21-day follow-up for outcome assessment was based on in vitro studies demonstrating silver diamine fluoride acts to arrest caries within this period .
Six examiners were calibrated. The examiners reviewed the criteria and pictures showing the codes. Each independently examined 6 to 8 children with code 3 and code 6 lesions, including lesions treated with silver diamine fluoride. One investigator served as the index examiner. The observed agreement with the index examiner and 6 examiners ranged from 89.3% to 93.2% and the kappa statistic ranged from 0.68 to 0.79.
The secondary outcome was to measure the change, if any, in relative abundance of microbial species or genus, assessed by metatranscriptomic (RNA) deep sequencing. The hypothesis was that caries-associated bacteria would decrease following treatment. We also hypothesized that prevalence of antibiotic and resistance genes would not change. RNA was used in proxy for vitality because many oral bacteria cannot be cultivated, and presence of RNA signifies production thereof by vital microbes within an hour beforehand. Plaque samples were collected from all participants. One participant in each study arm from each of the 3 study sites was selected from sequential lists of subject identification numbers by the biostatistician for the pilot analyses described here; treatment arm participants were selected from the subset for whom all studied lesions successfully arrested and placebo arm participants were selected from the subset from whom all studied lesions did not arrest.
2.6
Harms
Within 24 to 48 h of treatment, trained staff contacted parent/caregivers by telephone about adverse events. The questions included (Fig. S3): (1) Has your child required medical care since his/her dental visit in the last 48 h? (2) Has your child been to an emergency room, medical doctor, nurse or health care provider? (3) Since receiving the solution on the teeth has your child experienced any of the following? Nausea; Not eating; Vomiting, Difficulty swallowing or breathing; Swelling around the lips or skin of the face; Itchiness around the lips or skin of the face; Hives or rash; Stomach ache; or Diarrhea. At the follow-up visit 14 to 21 days after treatment, a dental provider performed a visual examination to detect gingival or soft tissue stomatitis or ulcerative lesions. Lesions were characterized as mild or severe and either localized or generalized.
2.7
Analysis plan
The primary outcome was the proportion of treated surfaces with arrested lesions at follow-up (code 6). The hypothesis was evaluated according to the intent-to-treat principle and based on all children who completed follow-up (97.0%). An exact two-sample permutation test was used to compare the average proportion of arrested lesions between the two treatment conditions . Bias-corrected bootstrap confidence intervals using 10,000 replications were also constructed for the average proportion of arrested lesions by treatment condition and for the difference in average proportion of arrested lesions between the 2 treatment conditions . Log-linear regression, implemented using generalized estimating equation (GEE) methods and robust standard errors, was used as a confirmatory analysis, comparing the risk of arrested lesions between the 2 treatment conditions. The outcome variable was the number of arrested lesions (subject-level outcome), and the number of treated surfaces and length of follow-up was accounted for by including the product of these 2 quantities as an offset term in the count regression analysis . In addition, the proportion of children with 100% of caries arrested was compared between the 2 treatment conditions using Fisher’s exact test, and exact 95% confidence intervals were constructed for the proportion by treatment condition and for the difference in proportions between the 2 treatment conditions . A two-sided 5% significance level was used for all statistical inference. Statistical analyses were performed using SAS, Version 4 (SAS Institute Inc., Cary, NC, USA) and R, Version 3.3.0 (R Core Team, Vienna, Austria).
Based on a two-sample t -test, for 80% power it was estimated 79 to 100 participants per group were required to demonstrate an effect size 0.40 to 0.45 in the average proportion of arrested caries lesions (i.e., a difference of 0.40 to 0.45 standard deviations (SDs) between the two group averages). If there were no arrested caries in the control group and 25% in the silver diamine fluoride group, it was estimated this difference would correspond to a difference of 0.40 to 0.45 SDs. Due to a delay in initiating the study and the need to complete follow-up by the end of the school year, only 66 participants were enrolled. However, the observed difference in the average proportion of arrested caries was much higher (2.4 SDs) than used in the sample size determination.
2.8
Metatranscriptomic sequencing library preparation
Total RNA was extracted by adding the applicator and 100 μL holding media to 400 μL Trizol, bashing with 2 mm ceramic beads for 3 min at 150 Hz with a TissueLyser (Qiagen, Hilden, Germany), and processing with an RNA Clean and Concentrator kit (Zymo Research, Irvine, CA, USA). Comparable amounts of RNA were obtained from each sample (1.5 ± 1.1 μg per sample; mean ± standard deviation). Holding media, and that spiked with Saccharomyces cerevisiae RNA were used as controls.
5 μg RNA was reverse transcribed to single-stranded complementary DNA (cDNA) using random hexamer primers and amplified to double-stranded (ds) cDNA (73 ± 221 ng per sample) with New England Biolabs’ (NEB, Ipswitch, MA, USA) Ultra v1.5 Directional RNA Library Prep Kit , then purified and size selected with 1.8 x Ampure XP reagent (Beckman-Coulter Life Sciences, Indianapolis, IN, USA). The cDNA was converted to Illumina libraries using the NEBNext Ultra II DNA Library Prep Kit (E7645) according to the manufacturer’s recommendation with the addition of USER enzyme (NEB) when adding Illumina adapters, and purification steps with 0.9 x Ampure. Dual index barcodes were added during 10 steps of PCR library enrichment. BioAnalyzer traces revealed high abundance 300–550 long oligonucleotides. Each sample was amplified again with dual-indexing primers on an Opticon qPCR machine (MJ Research, Waltham, MA, USA) using a Kapa Library Amplification Kit (Kapa Biosystems) until the exponential portion of the quantitative PCR signal was found, then cleaned with 0.8 x Ampure. Samples were then quantified by qPCR (Kapa Biosystems, Wilmington, MA, USA) and pooled for each participant, then re-quantified with a ddPCR Library Quantification Kit (Bio-Rad, Hercules, CA, USA) and pooled again. Sequencing was performed on one lane in a HiSeq 4000 (Illumina, San Diego, CA, USA) using 135 bp paired-end sequencing.
2.9
Sequencing read analysis
Following CASAVA quality filtering (Illumina, San Diego, CA, USA), human sequences (v38) were removed using the STAR alignment tool . Low-quality sequences were removed with PriceSeqFilter . Redundant reads were removed using CD-HIT-DUP . Low-complexity sequences (Lempel-Ziv-Welch ratio of less than 0.42) were removed. Illumina adapter and ϕX174 control sequences were removed using BowTie2 . A random subset of 300,000 reads was searched against a repeat-masked subversion of the NCBI nonredundant nucleotide (nt) database (version July 2015) using GSNAPL , and processed with internal scripts. The subset was searched against the BacMet database using RapSearch2 to assess levels of metal- and antibiotic-resistance genes.
High quality sequencing reads were obtained in abundance (2.9 ± 1.9*10^6 reads) for all but one sample (56,835 from the unaffected surface of participant 3058 at follow-up). 21.7 ± 5.0% of filtered reads mapped to nt, which is typical for metatranscriptomic sequencing. The media negative control showed low background (937 reads), verifying the clinical source of reads for the other samples. The spiked positive control resulted in nearly all reads matching S. cerevisiae (89%) or fungal yeasts (3%), as expected.
2.10
Microbiology
Taxonomic and resistance gene counts were analyzed for significant changes using LME4 (R Core Team). Nested grouping was used to represent the experimental design of paired before-after samples and paired lesion-plaque samples from the same patients. The regression design was: ∼treatment + time + treatment:time, where the ß coefficient for treatment:time was assessed for changes. Significance was determined by a Benjamini-Hochberg false discovery rate (FDR) of less than 0.01.
2
Methods
The study design was a double-blind randomized placebo-controlled superiority trial with two parallel groups. The study setting was three Head Start programs (preschools) in the U.S. state of Oregon. The participants were children 24 to 72 months of age who had at least one untreated cavitated active caries lesion with dentin exposed based on the Nyvad criteria (level 3: “Enamel/dentin cavity easily visible with the naked eye where the surface of cavity feels soft or leathery on gentle probing.”) . The children were healthy by medical history. The exclusion criteria were weight <15 kg, known sensitivity to silver or other heavy-metal ions, presence of any gingival or perioral ulceration or stomatitis or presence of a tooth abscess. The weight exclusion was to keep the applied dose in the range of the minimum risk dose for chronic exposure to silver determined by the U.S. Centers For Disease Control and Prevention. The participant flow is outlined in Fig. 1 . Recruitment began February 25, 2016 and the final child was evaluated on May 10, 2016.
The study was conducted under IND 124808 from the Food and Drug Administration. The Western Institutional Review Board approved the protocol and the informed consent of participants was obtained. A $50 gift card was provided after follow-up.
2.1
Treatment conditions
The test drug was silver diamine fluoride (Advantage Arrest™, Elevate Oral Care LLC., West Palm Beach, FL). The drug is a transparent blue-tinted solution containing 38 w/v% of silver (Ag) and 5.5 w/v% of fluorine (F). The placebo was non-fluoridated sterile water with a blue tint. The affected tooth surface was gently cleaned and dried with cotton gauze. The gingival tissue of the tooth was protected with petroleum jelly. An applicator was dipped into the agent and 3–4 mg applied to the lesion (1 drop treats 3 to 5 teeth). No rinse was performed or special instructions given following application.
2.2
Plaque sample collection
Before treatment and at the beginning of the follow-up examination, plaque samples were taken from up to 2 caries lesions and one unaffected tooth surface, selected sequentially from a list of tooth identifiers randomized for each patient, by rubbing a 2-mm diameter plastic applicator on the lesion surface for 4 s. The applicator was then placed in 250 μL tryptic soy broth with 15% glycerol and 2% sucrose, inverted twice, and kept at −20° C up to 2 months until moving to −80° C to await processing.
2.3
Allocation
Participants were randomly assigned to groups with a 1:1 allocation using a computer-generated randomization schedule with stratification by Head Start program and randomly permuted blocks of sizes 2 and 4.
2.4
Blinding
The dental providers who applied the treatments, participants, and examiners were blinded. Blinding was maintained by dispensing the test and comparator liquids from similar bottles. The test and placebo bottles were assigned codes A, B, C or D with 2 codes for test and 2 codes for placebo. Four bottles were used to strengthen the blinding of treatment, and were otherwise identical. The biostatistician knew only which two labels were the same, but was blinded to which labels were the test or placebo conditions.
2.5
Outcomes
The primary outcome was caries lesion arrest. The hypothesis was that silver diamine fluoride is more effective than placebo in arresting caries lesions with dentin exposed in primary teeth. A visual/tactile examination was performed before and 14 to 21 days after treatment to identify and assess lesions for caries activity . The teeth were examined with artificial light but without magnification in the preschool setting after drying with cotton gauze. A change from a Nyvad lesion code 3 to 6 (“Enamel/dentin cavity easily visible with the naked eye; surface of cavity may be shiny and feels hard on probing with gentle pressure.” ) was considered a positive outcome: caries arrest. The criteria have been shown to be reliable and valid . The 14 to 21-day follow-up for outcome assessment was based on in vitro studies demonstrating silver diamine fluoride acts to arrest caries within this period .
Six examiners were calibrated. The examiners reviewed the criteria and pictures showing the codes. Each independently examined 6 to 8 children with code 3 and code 6 lesions, including lesions treated with silver diamine fluoride. One investigator served as the index examiner. The observed agreement with the index examiner and 6 examiners ranged from 89.3% to 93.2% and the kappa statistic ranged from 0.68 to 0.79.
The secondary outcome was to measure the change, if any, in relative abundance of microbial species or genus, assessed by metatranscriptomic (RNA) deep sequencing. The hypothesis was that caries-associated bacteria would decrease following treatment. We also hypothesized that prevalence of antibiotic and resistance genes would not change. RNA was used in proxy for vitality because many oral bacteria cannot be cultivated, and presence of RNA signifies production thereof by vital microbes within an hour beforehand. Plaque samples were collected from all participants. One participant in each study arm from each of the 3 study sites was selected from sequential lists of subject identification numbers by the biostatistician for the pilot analyses described here; treatment arm participants were selected from the subset for whom all studied lesions successfully arrested and placebo arm participants were selected from the subset from whom all studied lesions did not arrest.
2.6
Harms
Within 24 to 48 h of treatment, trained staff contacted parent/caregivers by telephone about adverse events. The questions included (Fig. S3): (1) Has your child required medical care since his/her dental visit in the last 48 h? (2) Has your child been to an emergency room, medical doctor, nurse or health care provider? (3) Since receiving the solution on the teeth has your child experienced any of the following? Nausea; Not eating; Vomiting, Difficulty swallowing or breathing; Swelling around the lips or skin of the face; Itchiness around the lips or skin of the face; Hives or rash; Stomach ache; or Diarrhea. At the follow-up visit 14 to 21 days after treatment, a dental provider performed a visual examination to detect gingival or soft tissue stomatitis or ulcerative lesions. Lesions were characterized as mild or severe and either localized or generalized.
2.7
Analysis plan
The primary outcome was the proportion of treated surfaces with arrested lesions at follow-up (code 6). The hypothesis was evaluated according to the intent-to-treat principle and based on all children who completed follow-up (97.0%). An exact two-sample permutation test was used to compare the average proportion of arrested lesions between the two treatment conditions . Bias-corrected bootstrap confidence intervals using 10,000 replications were also constructed for the average proportion of arrested lesions by treatment condition and for the difference in average proportion of arrested lesions between the 2 treatment conditions . Log-linear regression, implemented using generalized estimating equation (GEE) methods and robust standard errors, was used as a confirmatory analysis, comparing the risk of arrested lesions between the 2 treatment conditions. The outcome variable was the number of arrested lesions (subject-level outcome), and the number of treated surfaces and length of follow-up was accounted for by including the product of these 2 quantities as an offset term in the count regression analysis . In addition, the proportion of children with 100% of caries arrested was compared between the 2 treatment conditions using Fisher’s exact test, and exact 95% confidence intervals were constructed for the proportion by treatment condition and for the difference in proportions between the 2 treatment conditions . A two-sided 5% significance level was used for all statistical inference. Statistical analyses were performed using SAS, Version 4 (SAS Institute Inc., Cary, NC, USA) and R, Version 3.3.0 (R Core Team, Vienna, Austria).
Based on a two-sample t -test, for 80% power it was estimated 79 to 100 participants per group were required to demonstrate an effect size 0.40 to 0.45 in the average proportion of arrested caries lesions (i.e., a difference of 0.40 to 0.45 standard deviations (SDs) between the two group averages). If there were no arrested caries in the control group and 25% in the silver diamine fluoride group, it was estimated this difference would correspond to a difference of 0.40 to 0.45 SDs. Due to a delay in initiating the study and the need to complete follow-up by the end of the school year, only 66 participants were enrolled. However, the observed difference in the average proportion of arrested caries was much higher (2.4 SDs) than used in the sample size determination.
2.8
Metatranscriptomic sequencing library preparation
Total RNA was extracted by adding the applicator and 100 μL holding media to 400 μL Trizol, bashing with 2 mm ceramic beads for 3 min at 150 Hz with a TissueLyser (Qiagen, Hilden, Germany), and processing with an RNA Clean and Concentrator kit (Zymo Research, Irvine, CA, USA). Comparable amounts of RNA were obtained from each sample (1.5 ± 1.1 μg per sample; mean ± standard deviation). Holding media, and that spiked with Saccharomyces cerevisiae RNA were used as controls.
5 μg RNA was reverse transcribed to single-stranded complementary DNA (cDNA) using random hexamer primers and amplified to double-stranded (ds) cDNA (73 ± 221 ng per sample) with New England Biolabs’ (NEB, Ipswitch, MA, USA) Ultra v1.5 Directional RNA Library Prep Kit , then purified and size selected with 1.8 x Ampure XP reagent (Beckman-Coulter Life Sciences, Indianapolis, IN, USA). The cDNA was converted to Illumina libraries using the NEBNext Ultra II DNA Library Prep Kit (E7645) according to the manufacturer’s recommendation with the addition of USER enzyme (NEB) when adding Illumina adapters, and purification steps with 0.9 x Ampure. Dual index barcodes were added during 10 steps of PCR library enrichment. BioAnalyzer traces revealed high abundance 300–550 long oligonucleotides. Each sample was amplified again with dual-indexing primers on an Opticon qPCR machine (MJ Research, Waltham, MA, USA) using a Kapa Library Amplification Kit (Kapa Biosystems) until the exponential portion of the quantitative PCR signal was found, then cleaned with 0.8 x Ampure. Samples were then quantified by qPCR (Kapa Biosystems, Wilmington, MA, USA) and pooled for each participant, then re-quantified with a ddPCR Library Quantification Kit (Bio-Rad, Hercules, CA, USA) and pooled again. Sequencing was performed on one lane in a HiSeq 4000 (Illumina, San Diego, CA, USA) using 135 bp paired-end sequencing.
2.9
Sequencing read analysis
Following CASAVA quality filtering (Illumina, San Diego, CA, USA), human sequences (v38) were removed using the STAR alignment tool . Low-quality sequences were removed with PriceSeqFilter . Redundant reads were removed using CD-HIT-DUP . Low-complexity sequences (Lempel-Ziv-Welch ratio of less than 0.42) were removed. Illumina adapter and ϕX174 control sequences were removed using BowTie2 . A random subset of 300,000 reads was searched against a repeat-masked subversion of the NCBI nonredundant nucleotide (nt) database (version July 2015) using GSNAPL , and processed with internal scripts. The subset was searched against the BacMet database using RapSearch2 to assess levels of metal- and antibiotic-resistance genes.
High quality sequencing reads were obtained in abundance (2.9 ± 1.9*10^6 reads) for all but one sample (56,835 from the unaffected surface of participant 3058 at follow-up). 21.7 ± 5.0% of filtered reads mapped to nt, which is typical for metatranscriptomic sequencing. The media negative control showed low background (937 reads), verifying the clinical source of reads for the other samples. The spiked positive control resulted in nearly all reads matching S. cerevisiae (89%) or fungal yeasts (3%), as expected.
2.10
Microbiology
Taxonomic and resistance gene counts were analyzed for significant changes using LME4 (R Core Team). Nested grouping was used to represent the experimental design of paired before-after samples and paired lesion-plaque samples from the same patients. The regression design was: ∼treatment + time + treatment:time, where the ß coefficient for treatment:time was assessed for changes. Significance was determined by a Benjamini-Hochberg false discovery rate (FDR) of less than 0.01.
Stay updated, free dental videos. Join our Telegram channel
VIDEdental - Online dental courses
