Effects of oral moisturizing gel containing propolis following head and neck radiotherapy: randomized controlled pilot trial

Aim Topical administration of oral gel may reduce radiotherapy-related oral complications. The aim of this study was to examine clinical and microbiological effects of self-administration of different gel formulations to oral mucosa in head and neck cancer patients. Materials and methods Twenty-seven subjects were recruited from outpatients who underwent radiotherapy of at least 50 Gy to the head and neck area. They were randomly assigned to oral gel with the following different ingredients: placebo, chlorhexidine, curry leaf, propolis, and turmeric. Before and after intervention, oral symptoms were evaluated, and nine oral pathogens in saliva were also quantified using real-time PCR. Results Twenty-five subjects completed the study and their data were analyzed. The number of Porphyromonas gingivalis in saliva significantly decreased after treatment with propolis gel, but not after any other treatments. Propolis gel treatment also relieved oral pain in all subjects who had oral pain at the baseline. Conclusions Topical administration with propolis gel may not only reduce P. gingivalis carriage in saliva, but also relieve oral pain. Discussion A future larger-scale clinical trial of oral propolis gel is needed to determine its clinical efficacy in radiotherapy-related oral complications of head and neck cancer patients.


Supplementary text for "Materials and methods" section.
19

Patient selection 20
Patient selection for the trial was performed according to the following inclusion 21 criteria: outpatients who had undergone radiotherapy with a dose of at least 50 Gy to the 22 head and neck area, not less than 20 years old, and relatively good performance status 23 ranging from a score of 0 to 2 according to the Eastern Cooperative Oncology Group 24 (ECOG) 1 . Those who had received antimicrobial treatment in the three months before 25 the study; or were smokers; edentulous in both jaws, or allergic to the components used 26 in the treatments were excluded. 27 28

Preparation of gels with and without natural products or antimicrobial biocide 29
Propolis used in this study was collected in a region of Minas Gerais State, Brazil 30 during the rainy season from October to May. Baccharis dracunculifolia was the 31 botanical origin of this Brazilian propolis, which is classified as type-12 propolis 2 . 32 Crude propolis was pulverized with a homogenizer and extracted in ethanol. After 33 stirring at room temperature for 12 hours, the extract was evaporated until the solid 34 content reached 55% (Yamada Bee Company, Inc., Okayama Japan), which was then 35 standardized to contain a minimum of 8.0% artepillin C. The propolis concentrate was 36 diluted with ethanol to yield 25 mg/mL (w/v, dry weight) of ethanol-extracted propolis. 37 The curry leaf used in the present study was collected in a region of North Central 38 Province, Sri Lanka (S&B Foods Inc., Tokyo, Japan). Turmeric used in the present 39 study was collected in India (House Foods Inc., Tokyo, Japan). Ethanol-extracted curry 40 leaf or the turmeric was obtained as a supernatant after stirring 1 g of curry leaf powder 41 or 1 g of turmeric powder (House Foods Inc., Tokyo, Japan) into 20 mL of ethanol 42 using a magnetic stirrer for 2 hours. Ethanol-extracted curry leaf was evaporated until 43 the solid content reached 45 mg/mL (w/v, dry weight) to yield ethanol-extracted curry 44 leaf , while ethanol-extracted turmeric was evaporated until the solid content reached 16 45 mg/mL (w/v, dry weight) to yield ethanol-extracted turmeric. The placebo gel was 46 prepared by mixing 1 mL of food grade ethanol and 100 g of oral moisturizing gel with 47 a homogenizing mixer. A low concentration of the food grade ethanol (1%) was added 48 to the placebo gel instead of an ethanol-extracted natural product without expectation of 49 any antimicrobial activity, for standardizing all the gel contents except for the individual 50 natural product extracts. Propolis 250 µg/mL (w/v) gel was prepared using the same 51 method as for the placebo gel, except 1 mL of 25 mg/mL (w/v) ethanol-extracted 52 propolis was added instead of food-grade ethanol. Curry leaf 450 µg/mL (w/v) gel was 53 prepared using the same method as that for the placebo gel, except 1 mL of 45 mg/mL 54 (w/v) of ethanol-extracted curry leaf was added instead of food-grade ethanol. Turmeric 55 160 µg/mL (w/v) gel was prepared using the same method as that for the placebo gel, 56 except 1 mL of 16 mg/mL (w/v) of ethanol-extracted turmeric was added instead of 57 food-grade ethanol. CHX 0.05% (w/v) gel was prepared by mixing with 1 mL of 5 % 58 (w/v) CHX (Nichi-Iko Pharmaceutical Co., Ltd., Toyama, Japan) instead of 1 mL of 59 food-grade ethanol in placebo gel. All five kinds of gels were packaged in 50-mL 60 laminated tubes (Kansai Tube Co., Ltd., Osaka, Japan) and sealed by heating the open 61 end, then stored at 4 °C before use. 62 63

Evaluation of clinical parameters 64
Symptoms of dryness and pain in the oral cavity were evaluated using a 100-mm visual 65 analogue scale (VAS). We determined the degree of mouth dryness using a Mucus® 66 moisture checker (Yoshida Co., Tokyo, Japan), according to the manufacturer's 67 instructions. Oral hygiene status, oral dryness, gingivitis, and oral candidiasis were 68 clinical parameters judged based on visual inspection by a single, trained dentist who 69 was blinded to the type of treatment received by the subject. 70 71

Evaluation of microbiological parameters 72
Saliva was obtained and 1 mL samples were pretreated with Sputazyme®, an enzymatic 73 reagent, to decrease the sample viscosity, according to the manufacturer's instructions 74 (Kyokuto Co., Tokyo, Japan). Saliva samples were centrifuged for 5 minutes at 17,400 75 x g to obtain a pellet containing bacterial cells, with the pellets used as starting materials real-time polymerase chain reaction (qPCR) assay to quantify the numbers of the 83 following six major periodontopathic bacteria; P. gingivalis 3 , Tannerella forsythia 4 , 84 Treponema denticola 5 , F. nucleatum 6 , and Aggregatibacter actinomycetemcomitans 3 , 85 as well as four opportunistic pathogens; methicillin-resistant Staphylococcus aureus 86 (MRSA) 7 , Escherichia coli 8, 9 , Serratia marcescens 10 , and Candida albicans 11 . The 87 number of total bacteria was quantified as the universal 16S rRNA gene copy number 9 . 88 Real-time qPCR was performed using Premix Ex Taq (Probe qPCR) (Takara Bio, 89 Shiga, Japan) with universal or species-specific primers and probes (Table S1) Improvements in oral hygiene (visual inspection), gingivitis (visual inspection), oral dryness (visual inspection), and oral dryness (Mucus ® ) were compared among the groups. Oral hygiene status was evaluated by visual inspection, according to the following criteria: no plaque accumulation as "Good" (Grade I); small amounts of plaque accumulation in oral cavity as "Moderate" (Grade II); overall plaque accumulation in oral cavity as "Poor" (Grade III). Gingivitis status was evaluated by visual inspection, according to the following criteria: no inflammation as "Good" (Grade I); mild redness or swelling as "Moderate" (Grade II); strong redness, severe swelling, or pus discharge as "Poor" (Grade III). Oral dryness status was evaluated by visual inspection, according to the following criteria: no symptom of dryness as "Good" (Grade I); foamy saliva as "Moderate" (Grade II); increased viscosity of saliva as "Poor" (Grade III); severe oral dryness of the whole mucosal surface as "Very poor" (Grade IV). Oral dryness status was also evaluated by Mucus ® . All values are expressed as the mean ± SD.

Fig. S2
Total bacteria in saliva. Copy numbers of universal 16S rRNA in saliva samples collected from the subjects were quantified by real-time PCR, and shown for each before (black bars) and after (grey bars) intervention. The subjects were divided into the placebo (n = 6), CHX (n = 5), curry leaf (n = 4), propolis (n = 5), and turmeric (n = 6) groups. The distribution of the total bacterial numbers in all subjects (n = 25) is also shown. Horizontal axes in all figures represent copy numbers of universal 16S rRNA (log10).