Comparison of the effect of marine-derived omega-3 fatty acids (n-3 FAs) as an adjunct to a non-steroidal inflammatory drug (NSAID) therapy vs NSAID therapy alone, for dogs with osteoarthritis

a Knowledge Summary by

Lok Yee Stephanie Wong DVM candidate ^1*

Merran Govendir PhD BVSc MEd(Higher Ed) FHERDSA ¹

¹Sydney School of Veterinary Science, The University of Sydney, Australia
^*Corresponding Author (lwon6099@uni.sydney.edu.au)

Clinical scenario

A 4 year old Rottweiler (male, neutered) is being treated with NSAIDs to manage osteoarthritis (OA) of the hip joints. The owners want to try omega-3 fatty acids (n-3 FAs) supplementation, and you (the clinician) are unaware about the evidence-basis for this.

The evidence

Two prospective, block-randomised clinical trials were found to match the PICO. Both studies assessed the effects of administering the same n-3 FAs with NSAID therapy compared to NSAID therapy alone. The n-3 FAs was a green-lipped mussel extract referred to as PCSO-524 marketed as Antinol® (Pharmalink International Ltd [Hong Kong]). Pharmalink provided financial support to both studies, and the same investigators performed both studies.

Both studies recruited canine hospital outpatients with a diagnosis of OA. Patients were randomly assigned to treatment groups. One treatment group was treated with a veterinary registered NSAID, and another was treated with the same veterinary registered NSAID and n-3 FAs. Both groups were administered their respective medications twice daily for 4 weeks. Patients underwent force plate gait analysis, gait observation and an orthopaedic examination that resulted in an OAS (Moreau et al., 2003) at week 0, and week 2 and week 4 post-treatment (i.e. weeks 6 and 8 of both studies).

Both studies found no significant benefit in the supplementation of n-3 FAs with NSAIDs, versus NSAIDs alone with respect to increasing the peak vertical force (PVF) generated by the arthritic limb, nor any improvement in the OAS. There are study design limitations in both studies such as inadequate sample size, and other factors such as an insufficient dosage, insufficient duration of dosage and brief study duration.

Summary of the evidence

Appraisal, application and reflection

Osteoarthritis is a prevalent condition in dogs, with reports of up to 20% of all dogs over 1 year old, in North America affected (Anderson et al., 2020). While dogs are typically managed with a multimodal approach involving weight control, exercise moderation and analgesics/anti-inflammatories such as NSAIDs; (Anderson et al., 2020; and Belshaw et al., 2016) nutraceuticals such n-3 FAs supplements are marketed as an adjunctive therapy to NSAIDs, for both humans and companion animals (Beale, 2004; and Johnson et al., 2020). PCSO-524® is a source of n-3 FAs derived from the New Zealand green-lipped mussel, Perna canaliculus (Kean et al., 2013). It is reported that ingested long-chain n-3 polyunsaturated fatty acids (PUFAs) are incorporated into inflammatory cell phospholipids resulting in a decrease in the amount of arachidonic acid available for the production of AA-derived eicosanoids, such as inflammatory prostaglandins (Calder, 2009). Additionally, n-3 PUFAs may influence inflammatory cytokine production and transcription factors that regulate inflammatory gene expression (Calder, 2009). NSAID therapy is considered a mainstay therapy in the management of canine OA (Innes et al., 2010), and the efficacy of NSAID supplementation with n-3 FAs for the treatment of canine OA does have a theoretical justification. 

It is noteworthy that despite a wide search strategy, only two studies were applicable to this PICO. Both studies used block classification for the severity of OA to ensure even distribution of mild/moderate and severe OA amongst treatment groups adapted from Moreau et al. (2003).  As the difference in OA severity is a source of variation between patients, blocking aims to remove some of this variability to emphasise treatment effects (Krzywinski & Altman, 2014). However, blocking partitions the number of subjects and as the number of treatment groups increases, more subjects are required. Blocking may also interfere with maintaining homogeneity between the treatment groups (Casler, 2018).  Classification of OA severity as either mild/moderate or severe as undertaken in these studies provides some additional areas of concern. Firstly, this scoring system is not validated (Moreau et al., 2003). Secondly, the OAS was performed by a single veterinarian in Vijarnsorn et al. (2019), and it is unclear whether the OAS was performed by a single veterinarian or multiple evaluators in Kwananocha et al. (2016).  Thirdly, the OAS has many subjective parameters, such as assessment of articular pain for the affected joints. The subjective assessment of articular pain can be problematic as it is reported that variability exists between veterinarians in their perception of the level of pain in their canine patients (Gruen et al., 2020).

Randomised allocation of study subjects into treatment groups was performed in both studies. However, the method of randomisation was not reported in Kwananocha et al. (2016). As randomisation reduces selection bias during the assignment of treatments (Pannucci & Wilkins, 2010), the lack of description on the manner of randomisation may cast doubt on the experimental design rigor. To evaluate whether the groups were homogenous before the intervention, the studies undertook statistical analysis of parameters (such as body weight, age, body condition score and breed) between groups. This degree of analysis is discouraged by the CONSORT (Consolidated Standards of Reporting Trials) Statement as it interferes with proper randomisation within treatment groups (de Boer et al., 2015).

Force plate gait analysis is considered the current gold standard measure of limb function in humans and animals (Brown et al., 2013). Compared to subjective scoring such as visual observation, it is more sensitive for lameness evaluation (Quinn et al., 2007). Both studies evaluated peak vertical force (PVF) values generated via force plate gait analysis as an outcome. In bilaterally lame dogs, the index limb (to be followed throughout the study) was elected as the limb with the smaller PVF at week 0 in Vijarnsorn et al. (2019), but it was unclear what criteria was used by Kwananocha et al. (2016). Additionally, peak PVF values and OAS are only surrogate indicators and may not accurately reflect improvements in a patient's mobility or whether pain has reduced due to the treatment alone (Administration USFaD, n.d.).

While repeated measurement analysis of PVF values was performed to determine the statistical significance of treatment effects and the effect of time within treatment, only descriptive analyses (i.e. numerical PVF mean values ± SD) were provided for the change in PVF, and no further statistical analyses were reported in both studies (Kwananocha et al., 2016; and Vijarnsorn et al., 2019). The results provided in both Kwananocha et al. (2016) Table 3 and Vijarnsorn et al. (2019) Table 2 must be scrutinised carefully as a greater numerical change from day 28 post-treatment to Day 0 PVF value is seen in the combination groups. However, there is no reporting of between group analysis over this time period and consequently no further conclusions can be made, other than within groups.

There was no determination of any changes of physiological parameters within the plasma and so there was no evidence that the 3-n FAs were absorbed from the gastrointestinal tract. Furthermore, the dosage and duration of treatment was possibly insufficient for n-3FAs to reach adequate therapeutic levels. The anti-inflammatory effects of n-3 FAS are reported dose dependent (Calder, 2009) and studies in humans report a daily dosage three times greater than that used in these canine studies (Dangardt et al., 2010; and Root et al., 2013). Another study suggested that a minimum of 70 days of treatment may be required for glucosamine hydrochloride and chondroitin sulphate to take effect in osteoarthritic dogs (McCarthy et al., 2013). Additionally, n-3FAs were administered daily for 4 weeks to humans and no physiological markers in response to n-3FAs administration were detected (Root et al., 2013). However, when n-3 FAs were administered for 12 weeks to humans, cytokine tumour necrosis factor alpha (TNF-a), interleukin (IL)-1β and IL-6 had significantly decreased (p = 0.008, 0.023 and 0.035, respectively) compared to the placebo treatment group (Dangardt et al., 2010). 

Sample size calculation was performed and reported in Vijarnsorn et al. (2019), but the firocoxib only group did not meet the calculated group size. The insufficient sample size is limiting as it reduces statistical power, which lowers the chance of detecting true treatment effects (Button et al., 2013). Additionally, the investigators did not account for attrition in their sample size calculation, which is 3.7% in Vijarnsorn et al. (2019). This proportion of loss of follow-up is not a concerning source of bias for Vijarnsorn et al. (2019); as it has been suggested that losses between 5–20% may confer bias (Sargeant et al., 2014). Another reason why a larger sample size would be appropriate is to reduce the effect of confounders. This is especially pertinent in Kwananocha et al. (2016), wherein body condition score is a possible confounder and may have biased the results.

There are evident flaws in the study design in both studies, such as lack of statistical analyses on the mean changes in PVF values rather than the baseline values, short treatment period, and small sample size. Additionally, the lack of clarity regarding the randomisation process, absence of sample size calculation and unclear blocking protocols further limits the validity of the results that can be extrapolated from one publication (Kwananocha et al., 2016). While the more recent study resolves some of these shortcomings, systematic faults are still present, hindering drawing definitive conclusions about the benefits of n-3 FAs supplementation in conjunction with NSAID therapy.

The conclusion of this Knowledge Summary does not align with the conclusions in each study. Kwananocha et al. (2016) states: The preliminary results imply the clinical benefits of PCSO-524® in combination with carprofen in the treatment of OA (Kwananocha et al., 2016). Vijarnsorn et al. (2019) states: The results of this study suggested combination of both PCSO-524® and firocoxib is more effective in alleviation of inflammation and improvement of weight bearing ability when compared to the uses of either PCSO-524® or firocoxib alone (Vijarnsorn et al., 2019). Both studies were financially supported by the supplier of the PCS0-524®, which may account for the more optimistic conclusions of both studies. A correction has since been published (Vijarnsorn et al., 2020) that the competing interests of one of the authors were omitted from Vijarnsorn et al. (2019).

Methodology Section

Search Strategy
Databases searched and dates covered:	CAB Abstracts via Web of Science (1910—2021) Scopus (1970–present) Medline via OvidSP (1946—2021)
Search strategy:	(dog or dogs or canine or canines or canis or canid or canids or Canidae) AND (osteoarthritis or osteo-arthritis or arthritis or "joint disease" or "joint diseases" or DJD) AND (carprofen or rimadyl or rimifin or canidryl or "carprodyl F" or dolagis or rycarfa or zenecarp or carprogesic or firocoxib or previcox or NSAID or "non-steroidal" or non-steroidal) AND ("omega 3" or "omega-3" or "omega 3 oil" or "omega-3 oil" or DHA or EPA or "eicosapentaenoic acid" or "docosahexaenoic acid" or "Hexadecatrienoic acid" or HTA or "α-Linolenic acid" or "Stearidonic acid" or "Eicosatrienoic acid" or "Eicosatetraenoic acid" or "Heneicosapentaenoic acid" or "Docosapentaenoic acid" or "Clupanodonic acid" or "Tetracosapentaenoic acid" or "Tetracosahexaenoic acid" or "Nisinic acid" or ALA or SDA or ETE or ETA or HPA or DPA or "green lipped mussel" or "green-lipped mussel" or mussel or GLM or "PCSO-524" or "PCSO 524")
Dates searches performed:	15 Jul 2021

Exclusion / Inclusion Criteria
Exclusion:	Systematic reviews, narrative reviews, in vitro studies, conference papers, book chapters.
Inclusion:	Articles relevant to the PICO question, randomised controlled trials.

Search Outcome
Database	Number of results	Excluded – Irrelevant to PICO questions	Excluded – Systematic review	Excluded – Narrative review	Excluded – In vitro study	Excluded – Conference paper	Excluded – Book chapter	Total relevant papers
CAB Abstracts	27	10	2	8	0	6	0	1
Scopus	28	11	3	10	1	0	1	2
Medline	12	6	1	1	3	0	0	1
Total relevant papers when duplicates removed								2

N.B A correction to Vijarnsorn et al. (2019) has been published (Vijarnsorn et al. 2020)

The authors declare no conflicts of interest.

Author contribution: Lok Yee Stephanie Wong (LYSW) wrote the first draft of the manuscript with feedback from Merran Govendir (MG). LYSW and MG edited the final manuscript and undertook the emendations.

1. Administration USFaD. (2018). Surrogate Endpoint Resources for Drug and Biologic Development. [online]. Available at: https://www.fda.gov/drugs/development-resources/surrogate-endpoint-resources-drug-and-biologic-development [Accessed 25 Jul 2021].
2. Anderson, K.L., Zulch, H., O'Neill, D.G., Meeson, R.L. & Collins, L.M. (2020). Risk Factors for Canine Osteoarthritis and its Predisposing Arthropathies: A Systematic Review. Frontiers in Veterinary Science. 7, 220. DOI: http://dx.doi.org/10.3389/fvets.2020.00220
3. Beale, B.S. (2004). Use of nutraceuticals and chondroprotectants in osteoarthritic dogs and cats. Veterinary Clinics of North America: Small Animal Practice. 34(1), 271–289. DOI: http://dx.doi.org/10.1016/j.cvsm.2003.09.008
4. Belshaw, Z., Asher, L. & Dean, R.S. (2016). Systematic Review of Outcome Measures Reported in Clinical Canine Osteoarthritis Research. Veterinary Surgery. 45(4), 480–487. DOI: http://dx.doi.org/10.1111/vsu.12479
5. Brown, D.C., Boston, R.C., Coyne, J.C. & Farrar, J.T. (2008). Ability of the Canine Brief Pain Inventory to detect response to treatment in dogs with osteoarthritis. Journal of the American Veterinary Medical Association. 233(8), 1278–1283. DOI: http://dx.doi.org/10.2460/javma.233.8.1278
6. Brown, D., Boston, R. & Farrar, J. (2013). Comparison of Force Plate Gait Analysis and Owner Assessment of Pain Using the Canine Brief Pain Inventory in Dogs with Osteoarthritis. Journal of Veterinary Internal Medicine 2013;. 27(1), 22–30. DOI: http://dx.doi.org/10.1111/jvim.12004
7. Button, K.S., Ioannidis, J.P., Mokrysz, C., Nosek, B.A., Flint, J., Robinson, E.S. & Munafò, M.R. (2013). Power failure: why small sample size undermines the reliability of neuroscience. Nature Reviews Neuroscience. 14, 365–376. DOI: http://dx.doi.org/10.1038/nrn3475
8. Calder, P.C. (2009). Polyunsaturated fatty acids and inflammatory processes: New twists in an old tale. Biochimie. 91(6), 791–795. DOI: http://dx.doi.org/10.1016/j.biochi.2009.01.008
9. Casler, M.D. (2018). Blocking Principles for Biological Experiments. Applied Statistics in Agricultural, Biological, and Environmental Sciences. 53–72. DOI: http://dx.doi.org/10.2134/appliedstatistics.2015.0074.c3
10. Dangardt, F., Osika, W., Chen, Y., Nilsson, U., Gan, L-M., Gronowitz, E., Strandvik, B. & Friberg, P. (2010). Omega-3 fatty acid supplementation improves vascular function and reduces inflammation in obese adolescents. Atherosclerosis. 212(2), 580–585. DOI: http://dx.doi.org/10.1016/j.atherosclerosis.2010.06.046
11. de Boer, M.R., Waterlander, W.E., Kuijper, L.D., Steenhuis, I.H. & Twisk, J.W. (2015). Testing for baseline differences in randomized controlled trials: an unhealthy research behavior that is hard to eradicate. International Journal of Behavioral Nutrition and Physical Activity. 12(4). DOI: http://dx.doi.org/10.1186/s12966-015-0162-z
12. Gruen, M.E., White, P. & Hare, B. (2020). Do dog breeds differ in pain sensitivity? Veterinarians and the public believe they do. PloS One. 15:e0230315. DOI: http://dx.doi.org/10.1371/journal.pone.0230315
13. Innes, J.F., Clayton, J. & Lascelles, B.D.X. (2010). Review of the safety and efficacy of long‐term NSAID use in the treatment of canine osteoarthritis. Veterinary Record. 166(8), 226–230. DOI: http://dx.doi.org/10.1136/vr.c97
14. Johnson, K.A., Lee, A.H. & Swanson, K.S. (2020). Nutrition and nutraceuticals in the changing management of osteoarthritis for dogs and cats. Journal of the American Veterinary Medical Association. 256(12), 1335–1341. DOI: http://dx.doi.org/10.2460/javma.256.12.1335
15. Kean, J.D., Camfield, D., Sarris, J., Kras, M., Silberstein, R., Scholey,A. & Stough, C. (2013). A randomized controlled trial investigating the effects of PCSO-524®, a patented oil extract of the New Zealand green lipped mussel (Perna canaliculus), on the behaviour, mood, cognition and neurophysiology of children and adolescents (aged 6–14 years) experiencing clinical and sub-clinical levels of hyperactivity and inattention: study protocol ACTRN12610000978066. Nutrition Journal. 12(100), 1–10. DOI: https://doi.org/10.1186/1475-2891-12-100
16. Krzywinski, M. & Altman, N. (2014). Points of significance: Analysis of variance and blocking. Nature Methods. 11(7), 699–700. DOI: https://doi.org/10.1038/nmeth.3005
17. Kwananocha, I., Vijarnsorn, M., Kashemsant, N. & Lekcharoensuk, C. (2016). Effectiveness of disease modifying osteoarthritis agents and carprofen for treatment of canine osteoarthritis. The Thai Journal of Veterinary Medicine. 46(6), 363–371.
18. McCarthy, G., O’Donovan, J., Jones, B., McAllister, H., Seed, M. & Mooney, C. (2007). Randomised double-blind, positive-controlled trial to assess the efficacy of glucosamine/chondroitin sulfate for the treatment of dogs with osteoarthritis. The Veterinary Journal. 174(1), 54–61. DOI: http://dx.doi.org/10.1016/j.tvjl.2006.02.015
19. Moreau, M., Dupuis, J., Bonneau, N. & Desnoyers, M. (2003). Clinical evaluation of a nutraceutical, carprofen and meloxicam for the treatment of dogs with osteoarthritis. Veterinary Record. 152(11), 323–329. DOI: http://dx.doi.org/10.1136/vr.152.11.323
20. Pannucci, C.J. & Wilkins, E.G. (2010). Identifying and Avoiding Bias in Research. Plastic and Reconstructive Surgery. 126(2), 619–625. DOI: http://dx.doi.org/10.1097/PRS.0b013e3181de24bc
21. Quinn, M.M., Keuler, N.S., Lu, Y., Faria, M.L.E., Muir, P. & Markel, M.D. (2007). Evaluation of Agreement Between Numerical Rating Scales, Visual Analogue Scoring Scales, and Force Plate Gait Analysis in Dogs. Veterinary Surgery. 36(4), 360–367. DOI: http://dx.doi.org/10.1111/j.1532-950X.2007.00276.x
22. Root, M., Collier, S.R., Zwetsloot, K.A., West, K.L. & McGinn, M.C. (2013). A randomized trial of fish oil omega-3 fatty acids on arterial health, inflammation, and metabolic syndrome in a young healthy population. Nutrition Journal. 12(40), 1–6. DOI: http://dx.doi.org/10.1186/1475-2891-12-40
23. Sargeant, J.M., Kelton, D.F. & O'Connor, A.M. (2014). Randomized Controlled Trials and Challenge Trials: Design and Criterion for Validity. Zoonoses and Public Health. 61(S1), 18–27. DOI: http://dx.doi.org/10.1111/zph.12126
24. Vijarnsorn, M., Kwananocha, I., Kashemsant, N., Jarudecha, T., Lekcharoensuk, C., Beale, B., Peirone, B. & Lascelles, B.D.X. (2019). The effectiveness of marine based fatty acid compound (PCSO-524) and firocoxib in the treatment of canine osteoarthritis. BMC Veterinary Research. 15, 1–8. DOI: http://dx.doi.org/10.1186/s12917-019-2110-7
25. Vijarnsorn, M., Kwananocha, I., Kashemsant, N., Jarudecha, T., Lekcharoensuk, C., Beale, B., Peirone, B. & Lascelles, B.D.X. (2020). Correction to: The effectiveness of marine based fatty acid compound (PCSO-524) and firocoxib in the treatment of canine osteoarthritis. BMC Veterinary Research. 16, 1. DOI: http://dx.doi.org/10.1186/s12917-020-02713-9
26. Zawadzki, M., Janosch, C. & Szechinski, J. (2013). Perna canaliculus Lipid Complex PCSO-524™ Demonstrated Pain Relief for Osteoarthritis Patients Benchmarked against Fish Oil, a Randomized Trial, without Placebo Control. Marine Drugs. 11(6), 1920–1935. DOI: http://dx.doi.org/10.3390/md11061920