Method for quantifying plaque in pet animals

Abstract

The present invention relates to methods for detecting and quantifying the plaque levels and/or lesions in companion animals, and the use of such information during trials of oral health products. The methods disclosed enable the trialling of companion animals for a shorter period of time, without the need for long and expensive trials on oral health products to be conducted. The methods described are conducted on conscious pet animals.

Claims

1. A method of determining the efficacy of a test composition in reducing or preventing oral substrate in a companion animal, comprising: (i) obtaining one or more images of one or more teeth of a conscious test companion animal and a conscious control companion animal at the start of the trial (day 0) using an image taking device that is capable of detecting fluorescence; (ii) analysing the images to quantify the substrate coverage on the one or more teeth of each companion animal at the day 0; (iii) administering the test companion animal with a test composition and the control companion animal with a control composition for the duration of the trial; (iv) obtaining one or more images of the same one or more teeth of step (i) of each of the test companion animal and the control companion animal at pre-determined intervals during the trial; (v) analysing the images to determine and quantify the substrate coverage and size on each tooth of each companion animal and compare with the images at the start of the trial and or at each stage of the trial; (vi) comparing the substrate coverage of the one or more teeth of the test companion animal and the control companion animal; and (vii) determining the efficacy of the test composition in reducing or preventing oral substrate in the companion animal Wherein the substrate coverage is qualitatively scored or quantified using image analysis software and wherein the companion animal has been trained to remain still for the duration of the obtaining the one or more images.

2. The method of claim 1, wherein the one or more teeth are located in the upper jaw of the companion animal.

3. The method of claim 2, wherein eight or fewer teeth are imaged.

4. The method of claim 1, wherein the substrate is selected from dental caries lesions, dental plaque, bacteria, calculus, staining, or any combination thereof.

5. The method of claim 1, wherein the companion animal is a dog or cat.

6. The method of claim 1, wherein the image taking device detects the fluorescence.

7. The method of claim 1, wherein the test composition is selected from a pet food stuff, pet treat pet chew, or a combination of these.

8. The method of claim 1, wherein the analysis is carried out by Quantitative Light-induced Fluorescence.

9. The method of claim 1, wherein the trial period is from 3-7 days.

Description

FIGURES

(1) FIG. 1: Shows a representation of the trial design followed (duration of 21 days).

(2) FIG. 2: Shows a representation of a dog's entire mouth (upper and lower jaw) and the respective numerical system used for labelling the teeth in the mouth.

(3) FIG. 3: Shows the average mouth results of the percentage plaque coverage per dog per day.

(4) FIG. 4: Shows the individual teeth results of the percentage plaque coverage per tooth over time for each repeat for two example dogs.

(5) FIG. 5: Shows the reproducibility of undisclosed individual teeth results for 103 to 108 evidencing the percentage plaque coverage by tooth, dog and photographer.

(6) FIG. 6: Shows the reproducibility of undisclosed individual teeth results for 203 to 208 evidencing the percentage plaque coverage by tooth, dog and photographer.

(7) FIG. 7: Shows the percentage plaque coverage (whole mouth average: I3, C, P3, P4) as determined by QLF on undisclosed teeth per day, by dog and photographer.

(8) FIG. 8: Variability plot of percentage plaque coverage (whole mouth average including maxillary P1, P2, P3, P4), as determined by QLF on undisclosed teeth by dog, day and repetition.

(9) FIG. 9: Shows the whole mouth average percentage plaque (maxillary I3, C, P3 and P4) at 1, 2, 3 and 4 weeks in conscious dogs fed an oral care chew (circles) compared to no chew (triangles). Means are shown as solid shapes with 95% confidence intervals.

(10) FIG. 10: Visual representation of the plaque identified by five human scorers marking plaque in Photoshop and plaque identified by QLF software, on four sample disclosed teeth.

(11) FIG. 11: Variability chart of average percentage plaque coverage identified by five human scorers marking plaque in Photoshop (triangles and squares) and QLF software (the circles) on maxillary 3.sup.rd incisors, maxillary and mandibular canines and 3.sup.rd and 4.sup.th premolars (disclosed teeth) [103, 104, 107, 108 and 109 and 404, 407, 408 and 409].

(12) FIG. 12: Shows the power curves to detect a 15% reduction in plaque accumulation when fed an oral care chew compared to no chew in a two-way crossover trial (clean mouth model).

(13) FIG. 13: Represents a chart of the average plaque coverage of maxillary jaw only (I3, C, P3, P4) of conscious dogs (circles) and maxillary (I3, C, P3, P4, M1) and mandibular (C, P3, P4, M1) jaw of unconscious dogs (triangles) when fed an oral care chew vs. no chew. Means are shown as solid shapes with 95% confidence intervals.

(14) FIG. 14: Shows mean % plaque coverage against day by tooth type with and without oral care chew respectively (LSD=least significant difference, bar represents the median difference between means needed for a significant difference at 5%), where FIG. 14A is without baseline adjustment and FIG. 14B is with baseline (day 0) adjustment.

(15) FIG. 15: Shows mean % plaque coverage against day by dogs fed daily oral care chew (squares) and those receiving no chew (circles), with 95% confidence intervals.

(16) FIG. 16: Shows mean difference in % plaque coverage from baseline, with 95% confidence intervals, by dogs fed no chew and those fed a standard chew, using three different methods for quantifying mouth plaque coverage.

(17) FIG. 17: (cat repeatability data) shows the average tooth percentage plaque coverage by cat for each repeat photograph.

(18) FIG. 18: (cat reproducibility data) shows the average tooth percentage plaque coverage by cat for each photographer.

(19) FIG. 19: (Cat dental diet efficacy) shows the mean average tooth percentage plaque coverage by diet and cat with 95% confidence intervals.

(20) FIG. 20: shows the mean percentage plaque coverage for each tooth by diet, with 95% confidence intervals. Stars indicate significant differences between diets within tooth, p<0.05.

(21) FIG. 21: shows the subsequent results of Variance components and diet differences for various subsets of teeth.

(22) FIG. 22: sets out the numbering system of teeth in animals (FIG. 21 A) and the VOHC list of recommended teeth in dogs and cats (FIG. 21B).

(23) FIG. 23: sets out the teeth assessed in the examples of the invention and as shown in the figures.

EXAMPLES

(24) The following examples show methods used to validate the use of Quantitative Light-induced Fluorescence (QLF) for quantifying plaque levels in animals by assessing its repeatability, reproducibility and accuracy and comparing it to an established clinical scoring system. Trials were performed using dogs to determine (i) Intra-photographer repeatability, (ii) inter-photographer reproducibility and (iii) comparison with current scoring method, as well as evidencing a 7 day trial with conscious dogs and using the methodology in cats.

Example 1: Intra-Photographer Repeatability

(25) The aim of this example was to determine the intra-photographer repeatability when measuring plaque coverage of dog's teeth using QLF.

(26) A panel size of 11 miniature schnauzer dogs aged between 2.5 and 6.9 years was selected. The dogs all had received a recent scale and polish (within the last month) and had little or no visible calculus. Dogs were tooth brushed daily for approximately one week prior to the start of the trial and were last tooth brushed on day one of the trial prior to their baseline measurement. Dogs received no subsequent tooth brushing for the three week (21 days) duration of the trial. All dogs during the trial were conscious.

(27) The trial phase of 3 weeks (21 days) was determined to be long enough to allow the build-up of high levels of plaque. Dogs were fed a mixed wet and dry diet during the trial and the test group were given a chew on a daily basis (chew vs. no chew).

(28) Images of undisclosed teeth were captured QLF by a single photographer on days 1, 3, 7, 14 and 21. Three repeat images, two in the morning and one in the afternoon, were taken on days 3, 7, 14 and 21 and two on day 1 (see FIG. 1). Taking images on all these days gave a range of different plaque coverages which was important for assessing the intra-photographer variability across the range.

(29) Measures were obtained by taking four images of the upper jaw capturing teeth numbers 103-108 and 203-208. See FIG. 2.

(30) FIG. 2: Shows a representation of a dog's entire mouth (upper and lower jaw) and the respective numerical system used for scoring the teeth in the mouth. It is known that VOHC clinical scoring methods requires teeth to be scored from maxillary 03, 04, 07, 08, 09 and mandibular 04, 07, 08, 09. The amount of plaque coverage was measured across teeth numbers 101 to 108 and 201 to 208 (as indicated).

(31) To reduce the effects of muscle memory on the positioning of the camera for imaging, the trial was designed so that the examiner imaged all the dogs in succession and then went back to the first dog and imaged them all again for the repeat set. After lunch the third set of images were then taken. This meant that the time between the 3 images was about 1 hour between images 1 and 2 and was about 1 hours between images 2 and 3.

(32) A set of images comprised five views around the mouth; two images on both the left hand side and right hand side of the dog's mouth were taken to visualise the maxilla 1.sup.st premolars (P1; 105, 205), 2.sup.nd premolars (P2; 106, 206), 3.sup.rd premolars (P3; 107, 207) and 4.sup.th premolars (P4; 108, 208) and one image from the front to visualise the maxilla and mandibular 1.sup.st incisors (I1; 101, 201, 301, 401), 2.sup.nd incisors (I2; 102, 202, 302, 402), 3.sup.rd incisors (I3; 103, 203, 303, 403) and part of the canines (C; 104, 204, 304, 404). See FIG. 2.

(33) TABLE-US-00002 Teeth numbers assessed Teeth set of images by the images taken Maxilla 1.sup.st premolars (P1) 105, 205 2.sup.nd Premolars (P2) 106, 206 3.sup.rd Premolars (P3) 107, 207 4.sup.th Premolars (P4) 108, 208 Maxilla and 1.sup.st Incisors (I1) 101, 201, 301, 401 2.sup.nd Incisors (I2) 102, 202, 302, 402 3.sup.rd Incisors (I3) 103, 203, 303, 403 Part of canines (C) 104, 204, 304, 404

(34) Statistics

(35) Linear mixed effects models (REML) were used to estimate variance components of the percentage plaque coverage. First a model including all teeth, using repeat nested within time nested within tooth nested within dog as the fitted random effects and then secondly for each tooth type, using repeat nested within time nested within dog. The percentage variability accountable to repeatability and the percentage coefficient of variability (repeatability standard deviation relative to the overall mean of the model) were then calculated.

(36) Results

(37) Images of 1584 undisclosed maxillary teeth (P1, P2, P3, P4) and 198 front views (single image capturing I1, I2, I3 and part of canine) from eleven miniature schnauzers were analysed using QLF image analysis software to quantify plaque coverage (FIGS. 3 and 4).

(38) FIG. 3: Shows the mean percentage plaque coverage for 11 dogs (A-K) receiving a chew compared to no chew. Three sets of images (1-3) were acquired on days 0, 3, 7, 14 & 21. Measures of percentage plaque coverage were obtained from the computerised images for teeth 105, 106, 107, 108, 205, 206, 207 and 208 individually and collectively and for front teeth (101, 102, 103, partial 104, 201, 202, 203 and partial 204). This study demonstrates that images of teeth can be obtained in conscious dogs and that the amount of plaque accumulation can be measured over a 3-21 day time-frame. It can also be seen that dogs that received a chew generally had lower levels of plaque than dogs that received no chew.

(39) Variance components analysis was used to quantify the intra-photographer repeatability across days, teeth and dogs. The repeatability coefficient of variation (100*standard deviation relative to the mean plaque coverage) for the maxillary premolars was 7.5%.

(40) The repeatability component of variability was also calculated for each tooth, across days and dogs (see table 2 and FIG. 4).

(41) FIG. 4: Shows an example of the percentage plaque coverage for each of the maxillary premolars (105 to 108 and 205 to 208) and the front teeth (mean of maxillary and mandibular incisors and partial canines) over time (days 0, 3, 7, 14, 21) for two dogs and the three repeat sets of images (1-3). QLF enables the analysis of plaque coverage in individual teeth over time which is of great benefit for products targeted to specific teeth.

(42) The QLF method showed good repeatability with the majority of teeth assessed accounting for <1.4% of the total variability in the data, with the exception of tooth 206 (5.2%) and the front teeth (27.9%). Making the variance components relative to the tooth means showed that the percentage coefficient of variation (% CV) ranged from 6% to 16% for most of the teeth analyzed with the exception of 206 (30%) and the front teeth (79%). These teeth had the lowest average percentage plaque coverage, 7.7% and 0.68% for tooth 206 and front teeth respectively which may have contributed to the increased relative variability. However, the high % CV on these teeth is of little consequence if using this technique to support product efficacy claims as the P3 and P4 were the only teeth evaluated in this trial that are required by the VOHC for claims validation trials and QLF showed the highest repeatability on these teeth with % CVs of <9%. By comparison Modified L&B repeatability is estimated to be >29% based on the info from Hennet et al. 2006.

(43) TABLE-US-00003 TABLE 2 Variability and % CV for eight individual teeth and the front teeth Variance Tooth Component SD % Variability Mean % CV 105 2.02 1.42 0.87 9.643 14.74 106 1.85 1.36 1.36 9.597 14.16 107 2.91 1.71 0.83 25.77 6.62 108 3.18 1.78 0.69 28.73 6.20 205 2.60 1.61 1.13 10.29 15.68 206 5.04 2.24 6.19 7.386 30.39 207 3.53 1.88 0.94 20.83 9.02 208 6.24 2.50 1.54 20.47 12.20 Front 0.29 0.54 28.47 0.6535 81.98

(44) One of the objectives of this trial was to assess how to use the technique. In subsequent trials, the front teeth were not imaged and more images were taken from the side to ensure all VOHC teeth were photographed (where visible).

(45) In summary, this study showed that images can be acquired of the maxillary teeth in conscious dogs. QLF is repeatable and the changes in the quantities of plaque can be measured over time frames of between 3 and 21 days. The levels of plaque coverage can be determined for individual teeth, which is beneficial for products targeted at specific teeth.

Example 2Inter-User Reproducibility

(46) The aim of this example was to evaluate the reproducibility between five photographers when measuring plaque coverage in conscious dogs using QLF.

(47) Five photographers took images of teeth in conscious miniature schnauzers (N=12) fed on the same dry diet. Dogs were tooth brushed every other day prior to the start of trial. Tooth brushing was ceased 1, 10 and 21 days prior to images being taken to assess different levels of plaque and calculus. Images were taken of teeth 103-108 and 203-208 by each photographer within a 2 hour period (FIG. 5).

(48) Linear mixed models (REML) were used to estimate variance components of the percentage plaque coverage, with photographer nested in dog as the random effects. The percentage variability accountable to the photographer and the % CV (reproducibility standard deviation relative to the overall mean of the model) were then calculated.

(49) The percentage plaque coverage was determined for 480 undisclosed maxillary teeth (I3, C, P3 and P4), 96 per photographer. The teeth selected were based on the teeth scored using the modified Logan & Boyce method, as defined by the VOHC standard product testing protocols. FIGS. 5 and 6: Shows the percentage plaque coverage of individual teeth (FIG. 5 shows 103 to 108 teeth and FIG. 6 shows teeth 203 to 208) for 12 dogs (A-L) and 5 photographers (1-5). The coefficient of variation (deviation relative to mean plaque coverage) for an average tooth was 10.9% CV. This data shows that QLF can be used to reproducibly measure the quantity of plaque on individual teeth.

(50) The mouth averages ranged from 1.2% to 41.2% plaque coverage (FIG. 7) and the inter-photographer reproducibility coefficient of variability for a mouth average was 3.21%. FIG. 7: Percentage plaque coverage (whole mouth average: I3, C, P3, P4) as determined by QLF on undisclosed teeth, by dog and examiner. The figure shows the reproducibility of average plaque coverage from five photographers (1-5) taking images of 12 dogs (A-L). The variation relative to the mean percentage coverage (% CV) between the different photographers was 3.2%. As compared to standard Modified Logan & Boyce (Based on Hennet et al., 2006) having 8.8% CV.

(51) In summary, QLF is reproducible and the percentage plaque coverage can be measured in the individual teeth on the maxillary jaw in conscious dogs. The amount of plaque can be quantified at 1, 10 and 21 days after stopping tooth brushing.

(52) FIG. 8. Variability plot of percentage plaque coverage (whole mouth average: maxillary P1, P2, P3, P4), as determined by QLF on undisclosed teeth by dog (A-K), day (0, 3, 7, 14, 21) and repetition (1-3).

(53) The percentage of plaque coverage can be measured on each of the maxillary premolars 3, 7, 14 and 21 days after stopping tooth brushing in conscious dogs.

Example 3: Longitudinal Assessment of Plaque AccumulationDetermining the Timeframes Over which Plaque Coverage can be Measured

(54) The aim of this example was to determine whether the levels of plaque coverage could be measured over time and at what stage it is possible to see a significant difference between treatment effects.

(55) The trial demonstrates that product efficacy can be measured in conscious dogs over shorter time frames than the normal 28 day test phase.

(56) This trial was designed to assess whether QLF could measure a product difference in conscious dogs at weeks 1, 2, 3 and 4 during the product test phase. 10 dogs were imaged consciously once a week for the duration of each 28 day test phase in the two phase cross over study. Only images of the maxillary I3, C, P3 and P4 were captured due to difficulties accessing the mandibular teeth in conscious dogs.

(57) Statistics

(58) The percentage plaque coverage, as measured by QLF, was analyzed using a linear mixed effect model, with weighting by week specific variability to allow for increasing variance over time. Dog was included as a random effect to account for repeated measures on a dog and chew type, week and their interactions were included as fixed effects. Contrasts were performed between chew types at each week using a family wise controlled error rate of 5%. (R v3.02 using libraries nlme and multcomp).

(59) Results

(60) A significant difference was found between chew and no chew at weeks 1, 2, 3, and 4 (see FIG. 9. and Table 3) in whole mouth average plaque coverage (VOHC teeth I3, C, P3, P4).

(61) After one week, a reduction in plaque coverage of 74% was observed (95% upper and lower confidence intervals of 54.3% and 93.6% respectively) when comparing chew to no chew (see Table 3).

(62) TABLE-US-00004 TABLE 3 Percentage plaque reduction and significance when dogs received an oral care chew compared to no chew. Mean % reduction in Lower 95% Upper 95% plaque confidence confidence Week coverage interval interval P value 1 74 93.6 54.3 <0.001 2 75.3 102.9 47.8 <0.001 3 64.2 104 24.5 <0.001 4 66.4 107.6 25.2 <0.001

(63) FIG. 9. Whole mouth average percentage plaque (maxillary I3, C, P3 and P4) at 1, 2, 3 and 4 weeks in dogs fed an oral care chew (circles) compared to no chew (triangles). Solid shapes illustrate the means and bars depict 95% confidence intervals.

(64) In conclusion, this study showed that QLF images can be acquired of the maxillary teeth in conscious dogs and that a significant difference in the percentage reduction in plaque accumulation, between dogs fed an oral care chew and those receiving no chew, can be determined in time-frames as short as seven days.

Example 4Assessing the Accuracy of the QLF Software

(65) The ability of the QLF software to identify plaque correctly was determined by comparison with plaque coverage levels determined by five human scorers manually marking plaque on QLF acquired image.

(66) QLF software results were compared to those from five trained human plaque scorers using photographs of 50 teeth with a range of plaque coverages. See FIG. 10. FIG. 10: Visual representation of the plaque identified by five human scorers marking plaque in Photoshop and plaque identified by QLF software, on four sample disclosed teeth.

(67) Whole mouth scores from nine dogs, as assessed by five human scorers, were analysed by a linear mixed model with scorer nested in dog fitted as the random effects. The variance estimates were then used to inform a simulation of 1000 scorers (with an average plaque coverage as found from the five human scorers). The probability of the QLF software results falling within the distribution of the human scorers' results was calculated by the percentage of simulated scorers with an average less than the average QLF software score. A test level of 5% was used.

(68) There was a high level of agreement in identification of plaque seen between the five scorers and QLF software across the entire range of plaque coverage from 0.6% to 100% (min, max). Simulations of the variance of the five scorers showed the QLF software was not significantly different to the human scorers with 10% of simulated human scorers having lower average percentage plaque coverage than the QLF software (p=0.1). See FIG. 11. FIG. 11: Variability chart of percentage plaque coverage identified by five human scorers marking plaque in Photoshop (the triangles and squares) and QLF software (the circles) on maxillary 3.sup.rd incisors, maxillary and mandibular canines and 3.sup.rd and 4.sup.th premolars (disclosed teeth). FIG. 11a shows the results from the teeth analyzed in the upper jaw (103, 104, 107, 108 and 109) and FIG. 11 b shows the results from the teeth analyzed in the lower jaw (404, 407, 408, 409). The teeth analyzed are the mandibular and maxillary VOHC teeth used in standard clinical scoring studies.

Example 5Comparison of QLF to Modified Logan and Boyce

(69) This study compared QLF to the established plaque scoring method modified Logan and Boyce (Hennet et al., 2006).

(70) Modified Logan & Boyce uses disclosed plaque and dissects tooth into gingival and coronal halves, and a score is then assigned between 0 and 4 dependant on the percentage plaque coverage on that half and a score between 1 and 3 dependent on the thickness of the plaque. The overall plaque score for each tooth half (gingival and coronal) was calculated by multiplying the coverage and thickness scores. Gingival and coronal scores were then added to give the total tooth score. The mean of all tooth scores provided the mouth score. The following teeth were included in the assessments: Maxillary I3, C, P2, P3, P4 and 1.sup.st molar (M1; 109,209), and mandibular C, P2, P3, P4 and M1 (309,409).

(71) Dogs were maintained on a commercial adult dry diet and a product efficacy study was conducted as previously described. The plaque on the dog's teeth was visualized by the use of a disclosing solution and the amount of plaque determined using modified Logan & Boyce (Hennet et al., 2006) and QLF. Data were analyzed by linear mixed effects models, with dog as a random effect to account for repeated measures and product as a fixed effect. Products were compared at the 5% test level.

(72) Good agreement between QLF and the modified Logan and Boyce method was demonstrated in the percentage reduction of plaque accumulation between dogs fed an oral care chew versus no chew (Table 4). The methods gave similar results with product efficacy measured as a 22% reduction in plaque build-up by modified Logan and Boyce and a 19% reduction as measured by QLF when comparing dogs fed an oral care chew versus no chew. The reduction in plaque accumulation as measured by QLF is within the 95% confidence intervals of the Logan and Boyce result, and the width of the confidence interval for the reduction seen with QLF is 52% of that with Logan and Boyce indicating less variability with QLF.

(73) TABLE-US-00005 TABLE 4 Comparison of product efficacy measured by QLF and modified Logan and Boyce methods. Mean plaque measure (95% Confidence Interval) Percentage Data set OC Chew No chew reduction P value Modified 9.79 12.57 22.13 <0.001 Logan & (8.83, 10.75) (11.54, 13.59) (12.64, 31.62) Boyce index QLF 54.78 67.73 19.12 <0.001 disclosed (51.72, 57.85) (64.48, 70.98) (14.09, 24.14) percentage

(74) A retrospective power analysis showed fewer dogs would be required to measure a 15% reduction in plaque accumulation using QLF compared to using the modified Logan & Boyce methodology, with at least 90% power.

(75) FIG. 12 shows the power curves to detect a 15% reduction in plaque accumulation when fed an oral care chew compared to no chew in a two-way crossover trial (26 dogs, clean mouth model). Solid line depicts QLF (disclosed teeth), dashed line modified Logan & Boyce and the dot dashed line QLF (undisclosed teeth). To measure a 15% reduction would require seven dogs with QLF compared to 19 with the Logan & Boyce methodology (90% power).

(76) QLF is less subjective than modified Logan & Boyce. It is also faster; photographers require less training and the images can be stored to provide a permanent database for future use. In addition, fewer animals are required to measure the same size effect in dental product efficacy trials.

(77) In summary, the percentage reduction in plaque accumulation as measured by QLF was in the range of that determined using modified Logan & Boyce. In addition, fewer animals are required to measure the same size effect in dental product efficacy trials. The use of fewer animals and the ability to undertake studies in conscious dogs supports two of the guiding principles underpinning the humane use of animals in scientific research; namely reducing the number of animals used to a minimum and refining the way experiments are carried out to improve animal welfare.

Example 6Conscious Vs Non-Conscious Dogs

(78) A method for measuring plaque in conscious dogs over a short time frame is highly desirable for testing new or improved oral care products (as opposed to the 28 day standard test trials).

(79) Dogs were trained for conscious QLF imaging including: With/without disclosing solution Plastic cheek retractors to view pre-molars Upper jaw only.

(80) QLF images of undisclosed teeth were also taken of dogs consciously at the end of each test phase of the cross-over study prior to being placed under anaesthesia.

(81) The percentage plaque coverage as measured by QLF of undisclosed teeth from conscious dogs (average of upper jaw teeth) and anaesthetised dogs (average of all VOHC teeth), was analysed using linear mixed models, with dog as a random effect to account for repeated measures on a dog and measure type, product and their interactions as fixed effects. The contrast between chew types was compared between measures at the 5% level. This is shown in FIG. 13 which represents a chart of the average plaque coverage (y-axis) of maxillary jaw only (I3, C, P3, P4) of conscious dogs (circles) and maxillary (I3, C, P3, P4, M1) and mandibular (C, P3, P4, M1) jaw of unconscious dogs (triangles) when fed an oral care chew vs. no chew (x-axis). The bars depict 95% confidence intervals.

(82) There was a significant difference between oral care chew and no chew for both dogs imaged consciously (P<0.001) and the same ten dogs imaged unconsciously (P<0.001)

(83) The average plaque coverage for the dogs that were imaged consciously (undisclosed) was 27.7% (22.2%, 33.2%) and 7.6% (2.1%, 13.1%) for no chew and oral care chew respectively which is a 72.6% (54.0%, 91.2%) reduction in plaque accumulation. When the same ten dogs were imaged under anaesthesia (undisclosed) the average plaque coverage was 30.5% (25.0%, 36.1%) for no chew and 9.5% (4.0%, 15.0%) when fed an oral care chew which is a reduction in plaque accumulation of 69.0% (52.1%, 85.8%). This is shown in Table 5 below.

(84) TABLE-US-00006 TABLE 5 showing the statistical results % plaque Data set Oral care chew No chew reduction P value QLF 7.6% 27.7% 72.6% <0.001 conscious (2.1%, 13.1%) 22.2%, 33.2%) 54.0%, 91.2%) (n = 10) QLF 9.5% 30.5% 69.0% <0.001 unconscious 4.0%, 15.0%) 25.0%, 36.1%) 52.1%, 85.8%) (n = 10)

(85) No significant difference was found between conscious and unconscious dogs in the percentage reduction in plaque accumulation between dogs fed the oral care chew and no chew (P=0.984).

(86) In summary, product performance can be differentiated in conscious dogs as well as anesthetized dogs. In addition, there was no significant difference in the percentage reduction in plaque accumulation when dogs were fed an oral care chew compared to no chew when only the maxillary teeth (I3, C, P3, P4) were assessed as opposed to the whole mouth (VOHC teeth).

Example 7Repeatability Study

(87) The aim of the study was to evaluate the differences in the average plaque coverage over time between dogs receiving an oral care chew and no chew.

(88) The study design was as described for example 1.

(89) Undisclosed by Tooth Type:

(90) A mixed model analyses was performed on the undisclosed data collected on days 0, 3, 7, 14 and 21. The random effects were fitted as day nested in tooth nested in dog. Tooth, day, chew and their interactions were fitted as fixed effects.

(91) The interaction between tooth, day and chew is significant (p<0.001), thus the changes in coverage of plaque over the days changes with tooth type, and this was also different with the addition of an oral care chew (FIG. 14A).

(92) There is a range in the average plaque coverage between tooth types on day 0 of the trial from 0% up to 23%, this range is slightly larger for the oral care chew dogs.

(93) If multiplicity adjustments are made for the number of comparisons between tooth types at each time point (from day 0 for each time profile and between with and without oral care chew for each tooth type at each time point) then no significant changes with time are found with oral care chew presentation. There are significant increases from day 0 for teeth 107, 108, 207 and 208 without oral care chew presentation.

(94) FIG. 14A: Mean % plaque coverage by day and tooth number with and without oral care chew respectively (LSD=least significant difference, bar represents the median difference between means needed for a significant difference at 5%).

(95) To investigate the influence of day 0 variability on the results, the average repeat at day 0 was used as a covariate in a mixed model analyses, with day nested within tooth nested within dog as the random effects and chew, tooth, day and their interaction as fixed effects.

(96) The interaction between tooth, day and chew is significant (p=0.002), thus the changes in coverage of plaque over the days changes with tooth type, and this is also different with the addition of an oral care chew.

(97) FIG. 14B: Mean percentage plaque coverage, adjusted for day 0, by day and tooth type, with and without oral care chew respectively (LSD=least significant difference, bar represents the median difference between means needed for a significant difference at 5%)

(98) In conclusion, images of individual teeth of conscious dogs can be used to measure changes in quantities of plaque levels over time, even when not using a clean mouth model. Individual teeth can be used to differentiate between an oral care product compared to a control group.

(99) Undisclosed Averaged Over Tooth Type (without Front), i.e. Average Mouth:

(100) A mixed model analyses was performed on the undisclosed data collected on days 0, 3, 7, 14 and 21, excluding the front tooth data. The random effects were fitted as day nested in tooth nested in dog. Day, chew and their interaction were fitted as fixed effects.

(101) The interaction between day and chew is significant (p<0.001), thus the changes in coverage of plaque over the days is different if a chew is presented (see FIG. 15).

(102) FIG. 15 shows mean % plaque coverage against day by dogs fed daily oral care chew (square) and those receiving no chew (circle) with 95% confidence intervals.

(103) Using a test level of 0.004 (=0.05/121216 comparisons include the difference from day 0 to each time point within each chew, plus the comparison of the difference from day 0 to each time point between chews chew) it was found that Within chew, there were significant increases from day 0 at day 14 and day 21, p=0.001, where differences were larger than 6.0% (day 7 was significantly different at the unadjusted level, p=0.014) Within no chew, there were significant increases from day 0 at day 7, 14 and 21, p<0.001, where differences ranges from 6.5% up to 18.4% (day 3 was significantly different at the unadjusted level, p=0.03) Between chews, the difference from day 0 was higher at days 3, 7, 14 and 21 for no chew when compared to chew and significantly so at day 21, p<0.001 (day 14 was significantly different at the unadjusted level, p=0.005).)

(104) Thus the interaction is due to the quicker increase in average % plaque in the no chew group. The table below illustrates these results with adjusted 95% confidence intervals for the differences over time with chew and without chew.

(105) TABLE-US-00007 TABLE 6 Comparisons in average % plaque coverage with 95% confidence intervals and p-values [yes = chew and no = no chew]. Comparison (time, Lower Upper chew) Difference 99.69% 99.69% Probability 0 Yes vs 3 Yes 2.005 7.19 3.18 0.156 0 Yes vs 7 Yes 3.84 8.87 1.19 0.014 0 Yes vs 14 Yes 6.392 11.37 1.41 0.001 0 Yes vs 21 Yes 5.957 10.94 0.97 0.001 0 No vs 3 No 2.941 7.52 1.63 0.030 0 No vs 7 No 6.499 11.02 1.98 0.000 0 No vs 14 No 12.541 17.04 8.04 0.000 0 No vs 21 No 18.444 22.95 13.94 0.000 0v3 Yes vs 0v3 No 0.936 6.750 4.881 0.601 0v7 Yes vs 0v7 No 2.658 8.350 3.035 0.151 0v14 Yes vs 0v 14 No 6.150 11.800 0.499 0.005 0v21 Yes vs 0v21 No 12.487 18.140 6.833 <0.001

(106) In conclusion, significant increases in the percentage of plaque accumulation can be measured from as early as three days. Differences in the rate of plaque accumulation can be detected in when dogs are fed an oral care chew compared to no chew.

Example 8: Conscious 7 Day Trial Using QLF Methodology

(107) The aim of this study was to use Quantitative light induced fluorescence on conscious dogs to quantify plaque build-up over each 7 day phase of the trial.

(108) Thirteen miniature schnauzer dogs using a clean mouth model, i.e. they received a single scale and polish at the start of the trial to set their teeth to baseline no plaque or calculus.

(109) Dogs were fed no chew or oral care chew daily for seven days in a randomised cross over design, i.e. each dog received both treatments.

(110) QLF imaging of plaque was performed QLF image capture software was used to analyse the images.

(111) QLF levels were conducted on conscious dogs with the addition of disclosing solution to visualise thinner immature plaque that doesn't fluoresce under blue light. Thus the method is a hybrid of QLF and planimetry. The addition of colour filters enables disclosed plaque to be automatically calculated.

(112) QLF was performed at the beginning and end of each seven day phase to give a baseline plaque measure to work from. Tooth brushing was used to reset the teeth to baseline plaque at the start of each phase. No general anaesthetic was required between each phase as the short nature of the trial meant limited calculus formed so an unconscious scale and polish was not required.

(113) The method enables the visualisation of 8 teeth on the upper jaw (103, 203, 104, 204, 107, 207, 108 and 208) as indicators of plaque development/removal.

(114) Statistics

(115) Image analysis of plaque levels was performed using analysis software. For each tooth the data takes the form of counts of fluoresced/stained pixels over total tooth area pixels to give percentage plaque coverage for each of the 8 indicator teeth (103, 203, 104, 204, 107, 207, 108 and 208) at each time point. The data set was analysed in three ways: (1) % coverage of individual teeth. For each tooth, the % coverage at baseline was subtracted from the % coverage at the end of phase. This measure, % coverage corrected for baseline, was used as the response in a linear mixed model with treatment as the fixed effect and dog and tooth within dog as the random structure. Visual inspection of the residuals indicated no need for any transformation. A likelihood ratio test was used to assess the necessity of the tooth random effect and this resulted in the removal of this term. (2) Average % coverage across teeth. For each tooth, the % coverage at baseline was subtracted from the % coverage at the end of phase. For each dog, the average of these values was calculated. This measure, the average % coverage corrected for baseline, was used as the response in a linear mixed model with treatment as the fixed effect and dog as the random effect. Visual inspection of the residuals indicated no need for any transformation. (3) % coverage of the mouth. For each dog, the % coverage for the whole mouth was calculated at each measurement occasion by summing the fluoresced pixels across all teeth and dividing by the sum of the total pixels across all teeth, then multiplying by 100. The whole mouth % coverage at baseline was subtracted from the whole mouth % coverage at the end of phase. This measure, the whole mouth coverage corrected for baseline, was used as the response in a linear mixed model with treatment as the fixed effect and dog as the random effect. Visual inspection of the residuals indicated no need for any transformation.

(116) For all three models, estimates with 95% confidence intervals were extracted for each treatment and a between treatment comparison was performed.

(117) All three analyses showed a significant difference between the treatments (see FIG. 16). 16 shows all three results on the same graph. FIG. 16 shows all three results on the same graph with 95% confidence intervals for the mean. This demonstrates that it is possible to show significant differences in plaque coverage, in this case induced using a daily dental chew treatment, over a 7 day treatment phase. Differences were observed using all three different analysis methods showing that the data can be effectively interrogated at the level of single tooth differences through to whole mouth average plaque scores.

(118) The first two methods give identical estimates as mathematically expected but the average across teeth appears more variable. The third method is the most variable and gives a slightly different estimate for the % coverage. This is expected since the first and second methods do not weight teeth by their size.

(119) This study shows that plaque accumulation on individual teeth from a conscious companion animal can provide significant results and so significant differences can be determined in plaque removal when measuring specific indicator teeth. The use of conscious animals enables shorter study phases (as general anaesthesia's are not required between study phases the 4 week gaps recommended between general anaesthesia's are not required). Power analysis of this data shows that less animals are required to identify the same differences in plaque coverage than the current accepted gold standard measure of Modified Logan Boyce unconscious scoring.

(120) In conclusion using QLF on conscious dogs allowed for shorter study treatment phases by removing the need for 4 week recovery times between general anesthesia. The study showed it was possible to show significant differences in plaque accumulation induced by oral care treatments using only indicator teeth on the upper jaw after a 7 day treatment phase.

Example 9Validation of QLF on Unconscious Cats

(121) The aim of this study was to quantify the variability and validate the use of QLF methodology to measure plaque in cats and the use of QLF on unconscious cats to demonstrate efficacy of a dental diet.

(122) QLF images were taken of VOHC teeth (Upper=104, 204, 107, 207, 108, 208 and Lower=304, 404, 307, 407, 308, 408, 309, 409) in twenty four cats using a clean mouth model, ie they received a single scale and polish at the start of each study phase to set their teeth to a baseline of no plaque or calculus. The cats were fed either a standard Adult dry diet or a Dental dry diet daily for 28 days in a randomised cross over design, ie each cat received both treatments. In addition to the QLF the Logan & Boyce (L&B) plaque scoring was also performed at the same time.

(123) The study consisted of two phases, Phase 1Repeatability and Phase 2Reproducibility. For each phase the following measurements were taken (i) percentage plaque on each tooth assessed, (ii) average of the percentage of plaque on each tooth and (iii) weighted mouth100*dividing the total plaque in mouth over total area in mouth.

(124) A single assessor used QLF on cats under general anaesthesia with disclosing solution to quantify plaque accumulation at the end of each 28 day phase of the trial using hardware and image capture software. All 14 VOHC scoring teeth were imaged on upper and lower jaws as indicators of plaque development/removal. ie. 104, 204, 107, 207, 108, 208, 304, 404, 307, 407, 308, 408, 309 and 409. Images were analysed using QLF analysis software.

(125) At the end of a second phase of the efficacy study described above, disclosed plaque on the teeth of 12 cats was imaged by 3 photographers using QLF to determine reproducibility of results between different scorers. All 14 VOHC scoring teeth were imaged on upper and lower jaws as indicators of plaque development/removal. i.e., 104, 204, 107, 207, 108, 208, 304, 404, 307, 407, 308, 408, 309 and 409. Images were analyzed using QLF analysis software.

(126) Statistics

(127) Demonstrating Dental Diet Efficacy

(128) The efficacy data were analysed by linear mixed effect models (LMM). The average tooth, weighted mouth and L&B data used chew nested in cat as random effects and chew as a fixed effect. In addition, using the raw % plaque coverage data for each tooth, chew nested in tooth nested in cat were used as random effects and chew by tooth and their interaction as fixed effects. Means and differences between means were estimated with 95% family wise confidence intervals.

(129) Repeatability and Reproducibility

(130) To quantify the repeatability and reproducibility, variance components analyses were performed using LMM for each of the measures. Initially variance components were quantified for each VOHC tooth, by including repeat nested in cat as the random effects. The same model was used to analyze the average tooth and weighted mouth data. In addition, the % plaque on each tooth was analyzed (i.e. all data used rather than averaging prior to analyses) by including repeat nested in tooth nested in cat as the random effects. Variance components were extracted and the percentage of the total variance attributable to repeatability/reproducibility was quantified, along with the coefficient of variation, relative to the overall mean of the respective data.

(131) Results

(132) The repeatability of the average tooth % plaque coverage by cat for each repeat photograph is shown in FIG. 17. The mean repeatability % CV was 2.2%

(133) The reproducibility of the average tooth % plaque coverage by cat for each of the three photographers is shown in FIG. 18. The mean reproducibility % CV was 2.3% The dental diet efficacy results of the LMM analyses of the average tooth % plaque coverage showed a significant reduction in plaque coverage of approximately 14.26% (p<0.001) when fed the Dental diet, as shown in FIG. 19 with the 95% confidence intervals.

(134) The LMM investigating the tooth by diet interaction showed the effect of diet to significantly change by tooth, p<0.0001. This interaction is illustrated in FIG. 20.

(135) Further analyses were performed to investigate QLF in conscious cats, when photographing the following subsets of teeth:

(136) 1. 104, 204, 107, 207, 108, 208, 304, 404, 307, 407, 308, 408, 309, 409

(137) 2. 104 & 204 only

(138) 3. 107 & 207 only

(139) 4. 108 & 208 only

(140) 5. 104, 204, 107, 207

(141) 6. 107, 207, 108, 208

(142) 7. 104, 204, 107, 207, 108, 208

(143) The 7 subsets were then analyzed, by LMM with diet nested in tooth nested in cat as the random effects and diet as a fixed effect. FIG. 21 shows the subsequent results of Variance components and diet differences for various subsets of teeth. All subsets but the 104, 204, 107, 207 combination found a significant difference between diets. It can be seen that the use of only individual teeth 108/208 is sufficient.

(144) In summary the trial demonstrated that using QLF it is possible to reproducibly quantify significant changes in plaque coverage in cats when fed specific dental diets, when using individual teeth (for example fewer than the VPOHC teeth, in particular the upper jaw teeth) as well as whole mouth assessments.

CONCLUSION

(145) The data represents the validated reproducibility and accuracy of QLF and shows that QLF is a suitable technique for measuring plaque levels in dogs and in cats.

(146) The data also shows that QLF can be used on conscious dogs and that the methodology provides significant differences to determine whole mouth assessments in animals by using data obtained on individual teeth (one or more teeth) in the mouth and one or more teeth in one jaw of the mouth (i.e. half mouth), in particular the upper jaw teeth as opposed to obtaining data in each and every tooth of the mouth to obtain the whole mouth assessment or as opposed to obtaining the entire list of recommended VOHC teeth.

(147) The data provides evidence that trials can be conducted over fewer than 28 days.

(148) The data provides evidence of a new rapid product testing methodology with increased accuracy allowing a shorter term for the trial and a lower number of animals tested. The data shows that trials can be conducted with shorter time frames for example, 7 days versus 28 day trials.

(149) The data showed that trials measuring the efficacy of oral products in reducing plaque or tartar showed significant results at shorter time frames, showing evidence that trials can be conducted for at least 7 days as opposed to at least 28 days.

(150) Further the examples of the invention show that accuracy of measuring the plaque and/or tartar in teeth can be achieved not only in shorter time frames (7 days versus 28 days) but also measuring half mouth vs full mouth. This was shown by only measuring the upper jaw of the subjects and then correlating the results to a full mouth percent plaque coverage measurement. The data shows significant results of trial lengths of 3, 7, 14 and 21 days, as well as significant results to conduct such trials using individual teeth (upper jaw) to correlate to full mouth measurements, as well as half mouth measurements correlated to full mouth measurements (i.e. upper jaw vs full mouth).

(151) A particular advantage of the trials is the fact that the dogs were conscious. The results showed that there was no significant difference in the results between conscious and unconscious dogs.

(152) The data shows that QLF is accurate, reproducible and a reliable method to be used on conscious dogs. It is capable of providing comparable results to the established clinical scoring methods, such as Logan & Boyce. Scoring requires much less training and is easily quantifiable using the software, thus less subjective. Fewer animals are required per trial and fewer teeth need to be analysed, therefore providing trials which are shorter and faster to obtain results on product efficacy, and less stressful for the animal.