| Sample | Periodontal Disease |
Tooth Resorption |
Bad Breath |
|---|---|---|---|
| ~15x WGS 3-14-2019 |
High Risk (7.35) |
High Risk (9.98) |
Low Risk (0.34) |
| Dental Health (mostly lick) 6-7-2021 |
High Risk (7.89) |
High Risk (9.58) |
Low Risk (0.05) |
| Dental Health (mostly lick) 6-8-2021 |
High Risk (6.68) |
High Risk (9.76) |
Low Risk (0.02) |
| Dental Health (mostly lick) 6-9-2021 |
High Risk (7.73) |
High Risk (9.65) |
Low Risk (0.18) |
| Dental Health (vet assisted) 8-16-2021 |
High Risk (8.87) |
High Risk (9.97) |
Low Risk (0.04) |
| Whole Genome Sequencing (not vet assisted) 4-21-2022 |
High Risk (8.7) |
High Risk (9.93) |
Low Risk (0.19) |
| Sample | Periodontal Disease |
Tooth Resorption |
|---|---|---|
| ~15x WGS 3-14-2019 |
~~6/5/2021 Report~~ Moraxella bovoculi Corynebacterium sp. ATCC 6931 Actinomyces howellii ~~6/15/2022 Report~~ Kocuria indica Moraxella bovoculi Corynebacterium sp. ATCC 6931 |
~~6/5/2021 Report~~ Corynebacterium xerosis Parabacteroides distasonis Moraxella bovoculi ~~6/15/2022 Report~~ Corynebacterium xerosis Parabacteroides distasonis Moraxella bovoculi |
| Dental Health (mostly lick) 6-7-2021 |
Moraxella cuniculi Wolinella succinogenes Corynebacterium xerosis |
Moraxella cuniculi Corynebacterium xerosis Parabacteroides distasonis |
| Dental Health (mostly lick) 6-8-2021 |
Moraxella catarrhalis Wolinella succinogenes Corynebacterium xerosis |
Moraxella catarrhalis Corynebacterium xerosis Wolinella succinogenes |
| Dental Health (mostly lick) 6-9-2021 |
Moraxella cuniculi Wolinella succinogenes Capnocytophaga sputigena |
Moraxella cuniculi Parabacteroides distasonis Wolinella succinogenes |
| Dental Health (vet assisted) 8-16-2021 |
Moraxella bovoculi Wolinella succinogenes Actinomyces sp. Chiba101 |
Parabacteroides distasonis Moraxella bovoculi Wolinella succinogenes |
| Whole Genome Sequencing (not vet assisted) 4-21-2022 |
Moraxella bovoculi Moraxella osloensis Moraxella cuniculi |
Corynebacterium xerosis Moraxella bovoculi Moraxella cuniculi |
basepaws (mouth swab, ~15x Whole Genome Sequencing, with raw data as FASTQ+BAM+gVCF):
Activated Kit / Collected Sample on 3/14/2019 (ID 31001812151645)
I noticed that a report is now provided for my earlier Whole Genome Sequencing sample, which I have uploaded.
While raw data is not available for the separate Dental Health Test, you can see the raw data for my regular coverage WGS here:
basepaws 15x WGS Read1: https://storage.googleapis.com/bastu-cat-genome/HCWGS0003.23.HCWGS0003_1.fastq.gz
basepaws 15x WGS Read2: https://storage.googleapis.com/bastu-cat-genome/HCWGS0003.23.HCWGS0003_2.fastq.gz
Because the raw data is available, I added the following scripts (uploaded in this folder, along with output reformated as Excel files):
a) run_Kracken2_Bracken-FASTQ-PE.py - runs Kraken2 and Bracken
b) filter_Bracken.R - filters Braken species-level assignment for those listed in the Supplemental Table of (Kao et al. 2021). Please note this script does not currently work as intended.
basepaws (Dental Health Test, $69-$79 each, 4 samples, Order #31272):
Samples were ordered on 5/30/2021.
All 4 kits arrived on 6/7/2021.
I collected the 1st sample (ID 31201053202314) on 6/7/2021. Bastu mostly licked the foam collector, so hopefully that is OK (versus having more direct contact with her teeth). The prior training helped keep her from running away, but she hasn't quite figured out what is earning her the treat afterwards.
I used baby oral cleaners (such as here) for the earlier training. I either directly applied Churu treat or I gave her a separate treat after I could get the oral cleaner without anything on it in her mouth (if even for a second). So, I think that is why she did not immediately run away when collecting these dental health test samples.
I collected the 2nd sample (ID 31201053201942) on 6/8/2021. Again, Bastu mostly licked the foam collector.
I collected the 3rd sample (ID 31201053202280) on 6/9/2021. Again, Bastu mostly licking, but I think less than the previous days. Towards the ends, I might have had something closer to the ideal interaction in the mouth, but I then stopped to give a treat (and, hopefully, build off that later).
Accordingly, I have decided not to submit my 4th sample. I was considering using it for training, but I ultimately threw the whole collection device away (not just the purple fluid).
I was able to download all 3 results on 7/9/2021.
basepaws (Dental Health Test, $59 each, 1 sample, Order #32719):
I ordered 1 additional sample on 7/11/2021, which was delivered to my apartment on 7/22/2021. This is sample with ID 31210251700872. This is the sample that I hoped I could get a vet or vet tech to help collect.
I took Bastu to a vet appointment on 8/16/2021. The vet thought Bastu looked to be in good shape, and confirmed that she was healthy enough to go under anesthesia. However, I added her to the wait list to get a dental cleaning, and I think she will get some dental x-rays to be extra cautious. More specifically, she was taken off the wait list in the middle of September, and her appointment for a dental cleaning is 10/5/2021. As of 10/2/2021, I still did not have the report for the vet-assisted sample collected on 8/16/2021.
In the process of providing some materials to the vet, I also noticed a potential point of confusion: the outer wrapper/sleeve correctly describes this as a "Dental Health Test" but they use the same inside box for all tests (and that makes no mention of being a Dental Health Test or recommending collection more centered on the teeth/gumline). So, I think you should keep the outer sleeve, if providing the materials to your vet.
I was able to download the results for this sample on 10/5/2021.
I took Bastu in for her dental cleaning on 10/21/2021. You can see the x-rays taken during the cleaning here. Overall, my understanding is that Bastu is in good health, especially for a cat her age.
I think there was at least 1 earlier bowl movement with a strong smell, but I mostly noticed diarrhea starting on Saturday (10/23). I forget exactly who provided the recommendation, but I had purchased wet food to give Bastu immediately after the cleaning.
On Sunday (10/24), I noticed that Bastu couldn't eat the greenies that I previously had mixed in with her dry food for the automatic feeder. I believe I also noticed some loss of appetite for the regular dry food on Monday (10/25), which is very unusual for her.
I called the vet on Monday (10/25), since this was recommended if there were side effects lasting longer than 48 hours. They suggested she may still have sensitivity, so I switched back to wet food. They provided some options, but I followed the suggestion to purchase Purina FortiFlora to mix into her wet food (for the diarrhea).
I will keep my eye on how things progress back to normal, but I think at least some of this is within the range of the reactions in the detailed information. For example, when I look into the details of Bastu's medical record it says "Refrain from offering any hard toys or treats for 7-10 days after the cleaning to allow any gum tenderness to fade.".
I think there is something about the taste the FortiFlora that Bastu doesn't like. However, if I put the FortiFlora in the middle of the wet food and I re-create a cylindrical-like shape, then she eats a larger fraction of the food.
During most of the 1st week after the cleaning, I offered a small amount of dry food (1 unit from the automatic feeder) for lunch. Towards the end of the week, I believe she consistently ate the dry food offered each day (and, before the cleaning, she would always eat all food as soon as it was available). On Friday 10/29/2021, she was able to eat a few Greenies (which should could not do before). So, on Saturday (10/30/2021), I started using the automatic feeder for dinner (3 units), to see if she consistently eats normal portions of her regular food. I am providing wet food with FortiFlora in the morning.
Bastu had a solid stool on 10/31/2021. This was used for the PetQCheck stool sample. The oder was not optimal, but I will continue with the FortiFlora for some extra time (before switching back to all dry food). I also started mixing back in the Greenies treats.
On 11/2/2021, it was harder to get Bastu to eat the wet food with the FortiFlora added. So, I will test using this as the last day for FortiFlora, turning back on her automatic feeder for 3 units of breakfast on the following day.
As far as I could tell, I think Bastu seemed normal after that. On 11/6/2021, I picked up some Royal Canin Feline Dental Food (dry, by prescription). I mixed that in with the existing food in Bastu's feeder. I also offered her a small sample, and she ate them. So, I think this should go smoothly. However, I will continue to follow up and make sure that I understand the situation (as well as make sure that all of my questions are eventually answered, if at all possible).
I took Bastu in for her dental cleaning on 9/19/2022. You can see the x-rays taken during the cleaning here. Overall, my understanding is that Bastu is in good health, especially for a cat her age.
I started Bastu on FortiFlora a couple days before the cleaning (as well as a separate preparation for ~1 day a week before that), with the goal of continuing until the morning of 9/23/2022 (always mixed with wet food, before starting her back on her prescription dental diet).
My understanding is that cats are typically expected to be back to normal the next day. This was not true for the first appointment. However, during a similar interval of time, I did not have the same problem with diarrhea or continuing with eating even wet food the next day. So, I don't know the relative influence of the FortiFlora, but this cleaning went much better! I hope that continues in the future.
Prevalence Notes (mostly copied from blog post)
On one of the basepaws websites, it says "It is estimated that 50-90% of ALL adult cats have a dental health problem, with periodontal disease being the number one culprit."
I did see a similar number (50-90%) reported on this page for cats.
For any level of severity, if you count 50-70% as being similar, then that overlaps what is reported here and perhaps pages 10-11 of the 2016 report that can be downloaded here.
However, I think that is different than what a lot of people might think of for "periodontal disease"?
For example, I think the claim about prevalence looks different than I see in this paper on dogs. Likewise, I can also see this paper where the prevalence is more similar to what I saw in cats (10-15% for periodontal disease, which was in fact reported as being the most common specific disease)
Possibly somewhere in between (?), in this paper, there was a fraction of overall "dental problems" being reported in the abstract is ~30%.
I don't see dental problems listed on this ASPCA page, (or, I believe, this Wikipedia page). So, I am not sure what to think about that.
I will continue to gradually look into this more, but the "periodontal disease stage" for dogs (and stage I, for cats) in the 2016 Banfield report with a higher percentage for "dental tartar" (greater than 50%, for ages greater than 1 year) seems like what I was expecting (with values <10% for stage I periodontal disease cats, and varying fractions by age for dog that are also all <10%).
For example, in the (Kao et al. 2021) preprint for the basepaws Dental Health Test, I believe 9.3% (570 / 6110) of the filtered cats / samples had periodontal disease. However, as pointed out in an e-mail to me from basepaws on 10/15/2021, I believe cats with comorbidities were excluded, which can cause an underestimation of the prevalence (all other things being equal).
For Periodontal Disease:
- 66% sensitivity for 570 cases (376 true positive, 194 false positive)
- 70% specificity for 1147 controls (344 false negative, 803 true negative)
| Predicted Medium / High Risk | Predicted Low Risk | |
|---|---|---|
| True Periodontal Disease | 376 | 194 |
| No Periodontal Disease Record | 344 | 803 |
From that, I would calculate the positive predictive value to be 376 / (376 + 344) = 52.2%.
For Tooth Resorption:
- 63% sensitivity for 111 cases (70 true positive, 41 false positive)
- 78% specificity for 1147 controls (252 false negative, 895 true negative)
| Predicted Medium / High Risk | Predicted Low Risk | |
|---|---|---|
| True Tooth Resorption | 70 | 41 |
| No Tooth Resorption Record | 252 | 895 |
From that, I would calculate the positive predictive value to be 70 / (70 + 252) = 21.8%.
After some additional considerations, I think substantially increasing the frequency of cats might be too stringent when trying to estimate the positive predictive values.
I still believe the connection between prevalence and positive preditive values is an important concept. So, you can still see those calulcation attempts here.
However, after taking what I saw in the next section, I decided to place the relatively greatest emphasis on the n=1,717 analysis.
I am sure it is not perfect, but I wanted to get some sense of the frequency of estimated risk for independent sets of customers.
My understanding is that there were reasons why Basepaws had some difficulties answering this question. I would also not use what I am about to describe as the main result.
Nevetheless, if I try to summarize what I noticed with minimial information from reports that are posted either publicly or in the Facebook Basepaws Cat Club, then I can attempt the calculations here. Essentially, this caused me to place less emphasis on the earlier alternative calculation that would have been more pessimistic, but I would say the question of whether they have reasonable prevalance (or if the true sensitivity and specifity might from what is described in the preprint) could still be up for discussion.
One consistent result across all 3 attempts to calculate positive predictive values is that it is always true that the positive predictive value is lower for tooth resorption than periodontal disease.
Additionally, I will include prevalence notes below. However, I wonder if maybe the focus should be shifted to whether something different than standard/recommended care should take place, versus precisely how the prevalence estimates were calculated and how serious the disease needed to be to be diagnosed. For example, my cat had some "high risk" estimates, but the vet said she was in relatively good health for her age and I agree with the indiviudal assessement to follow the recommended general guidelines (versus previously not taking discussing preventative dental health measures with my vet).
Likewise, I don't believe anybody recommends less than standard/recommended care because of "low risk" results from Basepaws.
In addition to being outliers to other estimates, the following numbers don't match the preprint above:
-
"Periodontal disease affects up to 80% of the adult feline population." (page 5)
-
"Unfortunately, tooth resorption is relatively common, affecting 20-60% of all cats and over 70% of cats over the age of five." (page 6)
The other estimates for periodontal disease are listed in an earlier section. In the 2016 Banfield report, a <10% frequency of periodontal disease was reported for geriatric cats (>10 years), on pages 10-11.
While I haven't currently looked for as many references for tooth resorption, the frequency in that 2016 report on page 21 is even smaller: "Tooth resorption has increased dramatically in cats from 5.4 cases per 10,000 in 2006 to 91.6 cases per 10,000 in 2015—a 1,587 percent increase." The statistic of "91.6 cases per 10,000" is equivalent to ~0.9%.
So, if you calculated the tooth resorption frequences based upon the basepaws preprint, then my current attempts at an estimate are 8.8% for the 1st calculation (111 divided by 111 + 1,147) or ~1.9% with the 2nd calculation (111 divided by 570 + 173 + 1,147 + 4,109). So, that 2nd calculation is still somewhat higher, but it is within an order of magnitude of the estimate from the Banfield 2016 report (unlike the frequencies in the reports returned to customers, which were "20-60%" or "70%").
I originally thought that the Banfield report seemed representative (sine it seemed like that was true for the Periodontal Disease part).
However, I have collected some additional notes on tooth resorption, after receiving the e-mail reply from basepaws. I think there is something that I still don't quite understand, but here is what I have collected so far:
-
This page from Cornell veterinary school reports a frequency of 20% to 60%
-
This VCA summary mentions frequencies closer to 50%. Type I and Type II tooth resorption are defined.
-
This summary from Jan Bellows briefly mentions 5 stages of tooth resorption.
-
This PDF provides a lot of details on tooth resorption, with a number of photos.
-
Girard et al. 2008 report a higher frequency in purebred cats versus mixed-breed cats, which I would guess could relate to a germline genetic basis for varying risk.
-
This blog post says that they are different from human cavities, but uses the term "cat cavity" to describe feline tooth resorption and mentions 50% as a prevalance for a "mild cat cavity" (as I understand it).
I am not sure if how much is something like the presence of 1 small "cat cavity" versus something needing tooth extraction (or other serious medical procedure). From what I understand, I think it would be odd for the tooth resorption frequency to be higher than the periodontal disease frequency, but I am thinking about disease severe enough to require surgical intervention (and I am not sure if that makes a big difference in the prevlance estimates, possibly to the extent of the difference in the minimal and maximal estimates?).
This page provides information about a variety of dental diseases, and I believe uses the term "Feline resorptive lesions (FRLs)" to refer to tooth resorption (citing a high frequency).
Nevertheless, if quickly looking for a frequency value, these are trending much closer to what was in the basepaws report.
As I collect more data, I will comment on the preprint (Kao et al. 2021). Already, I think the company is supposed to declare a competing interest, because this relates to a product that they are be selling. However, I am waiting to say more about the scientific details (and implications for individuals purchasing the test).
I think there is at least some initial misunderstanding on my part, which I could determine based upon an answer in the Basepaws Cat Club Facebook group. Based upon what I see in another individual's report, I think the plots next to the scores are for individual bacteria (versus cats in the healthy and/or disease training groups)? I am going to follow-up via e-mail. Neverthless, I would like to see what the plot looks like for scores among cats that have been tested, and I am also curious what the score distribution looks like in new, independent samples.
I think I may also ask if the liquid in the collection material has something to reduce or prevent post-collection bacterial "blooms".
My understanding is that age is a risk factor. I like that I don't see any obvious over-fitting in the results (with accuracy that seems more than random, but certainly not perfectly predictive). However, is the microbiome score providing an advantage over known risk factors?
I thought I saw something about early detection mentioned somewhere. However, if that was emphasized, then I think the experimental design would be different for samples collected before the disease onset versus trying to understand the basis of the disease in cats that already have symptoms.
I am not sure if it might also be worth looking more into some of the references for this article.
In general, I have collected what I thought were my most important comments and I posted them on the preprint.