Osteoarthritis (OA) is the most common form of joint disease in dogs, affecting 20 and 90% of dogs older than 1 and 5 years, respectively (14). Progressive degeneration of articular cartilage causes significant pain in dogs, leading to lameness and impaired quality of life. An effective drug for treating the underlying cause of OA has not been established, and the purpose of existing treatments is to relieve pain, prevent further progression, and improve the overall quality of life (7). Nonsteroidal anti-inflammatory drugs (NSAIDs) are the primary pharmaceutical therapies recommended for managing OA pain in dogs (15). However, there is increasing interest in alternative treatment options for OA due to the reported adverse effects of NSAIDs, with a prevalence of 2.6-34% (25). In recent years, nutraceuticals have garnered significant attention in OA research because of their potential long-term use and minimal side effects (14).

Since the 1970s, extracts of Perna canaliculus, commonly known as green-lipped mussels (GLMs), have been used to treat arthralgia in both human and veterinary patients. In vitro and in vivo studies have demonstrated the anti-inflammatory effects of GLM over the last 50 years (23). GLMs contain several omega-3 fatty acids, including eicosatetraenoic acid, eicosapentaenoic acid, and docosahexaenoic acid, which inhibit the 5-lipoxygenase and cyclooxygenase-2 pathways and promote the reduction of pro-inflammatory leukotriene, prostaglandin, and cytokine production in inflammatory cells (19). Other lipids, including pro-resolving lipid mediators and bioactive peptides, may contribute to the beneficial effects demonstrated through different pathways (8). Severe adverse events have been associated with GLMs. The only documented events in human clinical studies are minor, including mild gastrointestinal issues such as reflux and nausea (23).

Early clinical investigations using steam-processed, unstable GLM yielded inconsistent results in the treatment of OA. However, clinical studies employing freeze-dried stable GLM powder have demonstrated a significant reduction in joint pain and stiffness since the introduction of temperature-controlled stabilization in 1986 (11,13,23). This indicated improved efficacy. The GLM oil complex (GLMOC) became available recently. The lipid fraction of stabilized GLM powder is extracted through supercritical fluid extraction and combined with olive oil and vitamin E as antioxidants (28). In an early human comparative study of stabilized powder and the lipid fraction, both GLM oil and powder were equally effective in reducing OA symptoms (12). GLMOC, which is characterized by low protein content, is unlikely to induce allergies. Additionally, it is more suitable for cardiac patients than dried mussel preparations because it is salt-free (28).

Several clinical studies have assessed the effectiveness of GLMOC for OA treatment in dogs (6,15,17,20,26,27). An initial investigation by Mongkon and Soontornvipart (20) demonstrated improvements in the clinical signs in dogs with OA treated with GLMOC. Three comparative studies (15,17,26) assessing various nutraceuticals, including GLMOC, revealed significant improvements in the subjective and objective findings of the GLMOC-treated group, showing its superior efficacy relative to other nutraceuticals. Kwananocha et al. (17) found the most significant improvement when GLMOC was combined with carprofen. Similarly, another study observed notable enhancements in the clinical signs after administering GLMOC alone; however, its combination with firocoxib led to greater improvement (27). While these studies collectively suggest the efficacy of GLMOC in treating canine OA, translating their findings to clinical applications is hindered by variations in dosage and administration techniques.

Therefore, this study aimed to evaluate the treatment effects of GLMOC in dogs with osteoarthritis and investigate the impact of dose and disease severity, which are factors in real-world practice. We hypothesized that administering GLMOC would alleviate clinical signs in dogs with OA and a reduction in dosage or severe OA conditions would attenuate the therapeutic effects.

Study design

This study was designed as a prospective open-label clinical trial involving client-owned dogs and involved two veterinary practices (Seoul, South Korea; Gyeonggi-do, South Korea) and a university veterinary teaching hospital (Seoul, South Korea). The study protocol was reviewed and approved by the Institutional Animal Care and Use Committee of Seoul National University (SNU-201002-1-1). All owners were informed about the study, and written consent was obtained before the dogs were enrolled in the study.

Animals

Client-owned dogs of any breed, sex, and bodyweight aged at least 6 months were eligible for enrolment in the study. All dogs were required to have clinical and radiographic evidence of OA in at least one joint of the thoracic or pelvic limb. Clinical evidence of OA was confirmed as the owner-reported disability, and the veterinarian evaluated lameness and joint pain. The radiographic evidence of OA included asymmetric joint space narrowing, subchondral sclerosis, articular surface erosion, and osteophyte formation.

The enrolled dogs were classified into two groups according to the severity of their OA based on orthopedic examination and radiographic signs. The severe group was defined as dogs with an orthopedic assessment score (OAS) of ≥14 and radiographic OA grade of ≥4, based on the Canine Osteoarthritis Staging Tool (COAST). The remaining dogs were classified into mild or moderate groups.

The exclusion criteria were as follows: history of orthopedic surgery within the preceding 1 month; history of cruciate ligament rupture within the preceding 6 months; lameness primarily related to neoplasia, neurologic disorder, infection, or immunologic disorder; and pregnancy or lactation. Dogs treated with any medications that would interfere with the results were required to have a washout period before the study: 7 days for opioids and opioid-like analgesics; 14 days for systemic NSAIDs, oral nutraceuticals, and short-acting oral/injectable corticosteroids; and 4 weeks for long-acting oral/injectable corticosteroids and injectable sodium-pentosan polysulfate.

Treatment administration

The supplement used in this study was GLMOC (Mungmungpalpal^®, Syspang, Seoul, Korea) (3). Each capsule contained GLMO 50 mg, olive oil 100 mg, and vitamin E 5 mg. The recommended dose was 2 capsules daily for dogs weighing less than 15 kg and 4 capsules daily for dogs weighing more than 15 kg (20).

Study schedule

The study had three phases: a 4-week treatment phase, a 2-week reduction phase, and a 2-week withdrawal phase (Fig. 1). The dogs were administered the recommended dose during the treatment phase and half of the initial dose during the reduction phase. No supplements were administered during the withdrawal phase. The dogs were allowed to exit the study at any time for welfare reasons. Appropriate rescue treatment was approved by the veterinarian.

Figure 1. Summary of the study protocol.

Each dog and owner visited the hospital for a total of five visits (days 0, 2, 4, 6, and 8). On day 0, an orthopedic examination and radiography were performed based on the owner’s complaint to confirm whether the dog met the inclusion criteria. The dogs were classified into the mild/moderate or severe groups. For all dogs, owners completed the Canine Brief Pain Inventory (CBPI), and veterinarians determined the Orthopedic Assessment Score (OAS) at every visit. Gait analysis was also performed at every time point, except for some dogs.

Owner assessment: canine brief pain inventory

At every visit, the owners were asked to provide a paper copy of the CBPI, a questionnaire that measures the severity and impact of chronic pain in dogs. The CBPI contains 11 questions separated into two sections. The pain severity score (PSS) is calculated as the mean of the scores for four questions about the severity of pain on a scale of 0 (no pain) to 10 (extreme pain). The pain interference score (PIS) is calculated as the mean of the scores for six questions about how pain interferes with the typical activities of the dog on a scale of 0 (does not interfere) to 10 (completely interferes). The scores for all 11 questions were summed to obtain the CBPI score. Improvement in clinical signs was indicated by a decrease in the CBPI score. Treatment success for each dog at each time point was also assessed. The CBPI-based treatment success is defined as a reduction of ≥1 in PSS and ≥2 in PIS (1).

Veterinarian assessment: orthopedic assessment score

At every visit, an orthopedic examination was performed by a veterinarian. The subjective scoring system used in this study was originally described by Edamura et al. (9) and modified by Lee et al. (18) (Table 1). Standing posture, lameness during walking and trotting, weight-bearing, joint mobility, pain on palpation, and clinical condition were assessed using 1-5 descriptive scales, and the scores were summed to obtain the OAS. Improvement in the clinical signs was indicated by a decrease in the OAS score.

Table 1 . Orthopedic assessment scoring system.

Score criteria
1. Standing posture
[1] Normal stance
[2] Slightly abnormal stance (partial weight-bearing of the limb, but the paw remains firmly in contact with floor)
[3] Markedly abnormal stance (partial weight-bearing of the limb, with minimal contact between the paw and the floor)
[4] Severely abnormal stance (no weight-bearing)
2. Lameness at walk
[1] No lameness; normal weight-bearing on all strides observed
[2] Mild lameness with partial weight-bearing
[3] Obvious lameness with partial weight-bearing
[4] Marked lameness with no weight-bearing
3. Lameness at trot
[1] No lameness; normal weight-bearing on all stridesobserved
[2] Mild lameness with partial weight-bearing
[3] Obvious lameness with partial weight-bearing
[4] Marked lameness with no weight-bearing
4. Willingness to allow the clinician to lift the limb contralateral to the affected limb
[1] Readily acceptscontralateral limb elevation, bears full weight on the affected limb for more than 30 s
[2] Offers mild resistance to contralateral limb elevation, bears full weight on the affected limb for more than 30 s
[3] Offers moderate resistance to contralateral limb elevation and replaces it in less than 30 s
[4] Offers strong resistance to elevation of contralateral limb and replaces it in less than 10 s
[5] Refuses to raise contralateral limb
5. Range of motion (ROM)
[1] Full ROM
[2] Mild decrease (10-20%), with no crepitus
[3] Mild decrease (10-20%), with crepitus
[4] Moderate decrease (20-50%)
[5] Severe decrease (≥50%)
6. Pain at palpation/mobilization
[1] No pain elicited on palpation/mobilization of the affected joint
[2] Mild pain elicited, e.g., turns the head in recognition
[3] Moderate pain elicited, e.g., pulls the limb away
[4] Severe pain elicited, e.g., vocalizes or becomes aggressive
[5] Severe pain elicited, e.g., not allow examiner to palpate/mobilize the joint
7. Evaluation of overall clinical condition
[1] Good
[2] Mildly poor
[3] Moderately poor
[4] Severely poor
[5] Very severely poor

Objective gait analysis: peak vertical force

Ground reaction forces were measured using a 3.25-meter-long pressure-sensitive walkway system (Strideway^TM; Tekscan, South Boston, MA, USA). Before the data collection, the patients were weighed, and the sensors of the walkway were equilibrated and calibrated. The dogs were allowed to acclimatize to the room and walkway for 5 min. They were guided across a pressure-sensitive walkway along a straight line with a loose leash. The velocity was maintained at 0.9-1.1 m/s. Five valid trials per dog were performed and used for data analysis. A trial was considered valid if all four limbs contacted the walkway surface during each gait cycle without the dog turning its head, stepping off the walkway, or pulling on the leash.

Designated research-grade software (Strideway v 7.7; Tekscan Inc., South Boston, MA, USA) was used for data acquisition and analysis. A video camera (Logitech C920; Logitech Europe S.A., Lausanne, Swiss) recorded each pass to ensure that the program appropriately labeled each foot placement. The mislabelled pawprints were manually labeled as left front, left hind, right front, and right hind. The mean value of the peak vertical force (PVF, expressed as a percentage of body weight [% BW]) of each limb for each dog was derived from the average PVF of the first five valid trials. The percent body weight distribution (%BWD) of the four limbs was calculated as follows: (PVF of the index limb/total PVF of the four limbs) × 100 (16). For dogs with multi-joint osteoarthritis, the index limb was identified based on the greatest deviation from normal peak vertical force values (100% BW/40% BW for forelimbs/hindlimbs).

Statistical analysis

Statistical analyses were performed using GraphPad Prism 8.0.1 for Windows (GraphPad Software, La Jolla, CA, USA). Values are expressed as the mean ± standard deviation (SD). The normality of the data was assessed using the Shapiro-Wilk test. To assess the treatment effect over time, differences in each variable (CBPI, OAS, and PVF) between the evaluation time points (days 0, 14, 28, 42, and 56) were calculated using the Friedman test with Dunn’s post-hoc analysis for non-parametric variables and repeated measures one-way ANOVA with post-hoc Tukey’s multiple comparison tests for parametric variables. The same statistical analysis was performed separately for each group (severe and mild/moderate) to assess the association between OA severity and treatment efficacy. For all the analyses, statistical significance was set at p < 0.05.

Animals

Fifteen client-owned dogs were included in the study. The breeds included Maltese (n = 4), Pomeranian (n = 3), Poodle (n = 2), Cocker Spaniel (n = 2), Beagle (n = 1), Welsh Corgi (n = 1), Shiba Inu (n = 1), and Golden Retriever (n = 1). The average age and weight were 7.7 ± 3.9 years and 9.5 ± 9.5 kg (mean ± SD), respectively. The affected joints were the stifle (n = 7), coxofemoral (n = 5), elbow (n = 2), and tarsal (n = 1). Seven dogs had severe OA, and eight dogs had mild-to-moderate OA (Table 2). After all the tests were completed, none of the dogs showed any signs of side effects.

Table 2 . Description of the study population.

	Breed	Sex	Age (year)	Weight (kg)	Most affected joint	Radiographic OA score at D0	OAS at D0	Group
1	Maltese	M	14	7.7	Lt. Stifle	4	14	Severe
2	Maltese	MC	9	2.5	Lt. Stifle	3	13	Mild/moderate
3	Maltese	FS	6	4.7	Rt. Stifle	4	17	Severe
4	Poodle	FS	4	5.7	Rt. Stifle	2	9	Mild/moderate
5	Pomeranian	FS	6	3.7	Rt. Stifle	4	17	Severe
6	Poodle	MC	7	6.5	Lt. Stifle	2	13	Mild/moderate
7	Maltese	FS	3	5	Rt. Coxo	3	14.5	Mild/moderate
8	Pomeranian	FS	10	3.7	Lt. Elbow	4	17	Severe
9	Pomeranian	MC	10	2.1	Lt. Stifle	4	14.5	Severe
10	Beagle	MC	11	15.6	Lt. Coxo	2	15.5	Mild/moderate
11	Cocker Spaniel	FS	15	8	Rt. Tarsal	4	17.5	Severe
12	Cocker Spaniel	MC	9	10	Rt. Coxo	4	19	Severe
13	Welsh Corgis	MC	7	15	Rt. Stifle	3	11.5	Mild/moderate
14	Shiba Inu	FS	1	10	Rt. Coxo	2	9	Mild/moderate
15	Golden Retriever	MC	3	42	Rt. Elbow	3	15	Mild/moderate

Gait analysis was performed in dogs 7-15. M, male; MC, male castrated; F, female; FS, female spayed..

Canine brief pain inventory score

Based on the CBPI data, treatment was considered successful for 40% of the patients at the end of the treatment phase. The treatment success rates decreased to 33.33% and 13.33% at the end of the reduction and withdrawal phases, respectively (Table 3).

Table 3 . CBPI-based treatment success rate at each time point (%).

	Week 2	Week 4	Week 6	Week 8
Total (n = 15)	33.33	40	33.33	13.33
Mild/moderate (n = 8)	37.5	37.5	37.5	12.5
Severe (n = 7)	28.5	42.8	28.5	14.3

After 4 weeks of treatment, the mean sum indexed CBPI score was significantly decreased from the baseline for the entire study population (Fig. 2A). The CBPI score at week eight was significantly higher at week eight than at week four. For the mild/moderate group, the CBPI score had decreased at weeks four and six relative to the pre-treatment values (Fig. 3A). There were no significant differences in the CBPI scores in the severe group (Fig. 4A).

Figure 2. Mean of CBPI scores (A), OAS (B), and PVF (C) in cohort dogs. Data are presented as mean with standard deviation. *p < 0.05, ****p < 0.00005 compared to D0. ^#p < 0.05 between each time point. CBPI, Canine Brief Pain Inventory; OAS, orthopedic assessment score; PVF, peak vertical force; D, day; W, week.

Figure 3. Mean of CBPI scores (A), OAS (B), and PVF (C) F in mild/moderate group. Data are presented as mean with standard deviation. *p < 0.05, **p < 0.005, ***p < 0.0005, ****p < 0.00005 compared to D0. ^#p < 0.05 between each time point. CBPI, Canine Brief Pain Inventory; OAS, orthopedic assessment score; PVF, peak vertical force; D, day; W, week.

Figure 4. Mean of CBPI scores (A), OAS (B), and PVF (C) in severe group. Data are presented as mean with standard deviation. **p < 0.005 compared to D0. CBPI, Canine Brief Pain Inventory; OAS, orthopedic assessment score; PVF, peak vertical force; D, day; W, week.

Orthopedic assessment scores

A significant decrease in the OAS from the pre-treatment values was observed throughout the study for the entire study population (Fig. 2B). The OAS was significantly greater at week eight than at week four but was still significantly lower than the pre-treatment value. For the mild-to-moderate group, a significant decrease in OAS was first observed at week four and maintained until the end of the study period (Fig. 3B). For the severe group, a significant difference was found between pre-treatment and week four OASs (Fig. 4B).

Peak vertical force

Gait analysis was performed for nine dogs: five were included in the mild-to-moderate group, and four were included in the severe group. After 4 weeks of treatment, a significant increase in the PVF from the pre-treatment value was observed (Fig. 2C). For the mild-to-moderate group, PVF was significantly greater at week four than before treatment and at week eight (Fig. 3C). There were no significant treatment effects on PVF in the severe group (Fig. 4C). The number of dogs with changes in the PVF exceeding 5% from the baseline, which has been suggested to be a clinically important value (5), was also investigated (Table 4).

Table 4 . Percentage of dogs with a PVF change exceeding 5% compared to D0 (%).

	Week 2	Week 4	Week 6	Week 8
Total (n = 9)	55.56	66.67	66.67	33.33
Mild/moderate (n = 5)	80	100	60	40
Severe (n = 4)	25	25	75	25

The clinical signs of osteoarthritis in the dogs treated with GLMOC were alleviated in the present study. After 4 weeks of treatment, a statistically significant improvement was observed in both the subjective assessment (owner-assessed CBPI and veterinarian-assessed OAS) and objective analysis of weight-bearing (PVF). No adverse events were observed.

The CBPI is a validated clinical metrology instrument that is currently the mainstay of pain assessment associated with chronic OA. By comparing the averages of the CBPI scores at each time point, we revealed a significant improvement from the baseline to week 4. We also assessed the treatment responses of each dog at each time point using the criteria of success predefined by Brown et al. (1). OA is a clinically variable disease, and score averaging may not reflect the individual differences in clinical severity. The FDA currently requires more rigorous individual assessments than average scores (24). The conversion of clinical metrology instrument data to success or failure has only been published for the CBPI and has recently been used in a few studies on OA treatment. Treatment success rates were 42% and 24% for carprofen and tramadol after 10 days (2), 45.6% for carprofen after 14 days (1), and 48.1% and 43.5% for grapiprant (24) and bedinvetmab (7) after 28 days. The different study designs make it difficult to compare data across studies, but 40% of GLMOC-treated dogs achieved treatment success after 28 days in this study, which is comparable to the reports of the above-mentioned studies.

However, the outcomes of this study require careful interpretation because it is open-label, and placebo controls were not used. To mitigate this drawback, ground reaction forces were measured using a pressure-sensitive walkway, which is an objective method for assessing limb function and joint pain in dogs with OA (10). Statistically significant improvement was observed at the end of the treatment phase. The change in the PVF from the baseline to the end of the trial demonstrated an inverted U-shaped pattern, indicating the alleviation of clinical signs during the treatment phase and gradual deterioration during the reduction and withdrawal phases. We also investigated the percentage change in PVF over time in each dog for further analysis. Previous studies have suggested that a PVF shift within 5-10% is unlikely to occur at random (4,5). Consequently, assessing the number of dogs in each group with changes in the ground reaction force exceeding 5% is regarded as a valuable outcome measure for canine OA studies (15). At the end of the treatment phase, we observed that six of nine dogs, all of which were in the mild-to-moderate group, exhibited PVF changes exceeding 5%. This was an objective indication of the alleviation of the clinical signs of OA.

A standard dose for GLM has not been established, and this necessitates further research (22). In this study, we used three phases with varying dosages to investigate the potential dose-related variations in treatment efficacy. Significant improvements in the OAS were observed throughout the study, but CBPI and PVF showed significant improvements only at the end of the treatment phase. The improvements in CBPI and PVF from the baseline persisted during the reduction and withdrawal phases, but they did not reach statistical significance. These results suggest that short-term reduction or withdrawal of SYSGLMO does not exacerbate OA signs, but at least 4 weeks of administration at the recommended dose is necessary for alleviation. The CBPI score and OAS were significantly increased at the end of the withdrawal phase than at the end of the treatment phase; however, caution is warranted because of the unblinded nature of the study and the subjective nature of both assessments.

While assessing the changes in each outcome variable over time in the cohort of dogs, we categorized them into mild, moderate, and severe groups to investigate the relationship between disease severity and treatment outcomes. The results indicated notable differences between the groups. For the mild-to-moderate group, statistically significant enhancements relative to the baseline were identified at week 4 for PVF, weeks 4 and 6 for the CBPI score, and weeks 4, 6, and 8 for the OAS. In contrast, a significant difference in the OAS was detected only at week 4 for the severe group. These findings align with those of a previous human study demonstrating a positive response of mild-to-moderate knee OA to GLM treatment, whereas severe cases did not show significant improvement (13). Dogs with severe osteoarthritis may require an adjusted GLMOC dosage for effectiveness comparable to that observed in dogs with mild-to-moderate osteoarthritis. A combination of oral or injectable drugs may also be a more effective treatment option for dogs with severe OA (17,27). However, further clinical studies are required to confirm these findings.

The present study has limitations. A placebo group was not included due to ethical and owner-related concerns associated with administering placebo for a painful disease (21). Future studies may improve the validity of these findings by introducing a medication with established effectiveness as a positive control. In addition, the durations of the phases were disparate. Corral et al. (7) emphasized that chronic conditions such as canine OA, especially under field conditions, may require treatments for more than 28 days. For a more thorough investigation of the effects of dose reduction and withdrawal, subsequent studies should consider extended reduction and withdrawal phases.

The administration of the recommended dose of GLMOC for 4 weeks resulted in significant improvements in the CBPI score, OAS, and PVF of dogs with osteoarthritis, and no adverse effects were observed. The short-term reduction or discontinuation of GLMOC did not intensify the osteoarthritis signs compared to the baseline, but it diminished the therapeutic efficacy. In patients with severe osteoarthritis, adjusting the GLMOC dose or combining GLMOC with other medications may improve the outcomes.

This work was supported by the National Research Foundation of Korea (NRF) grant funded by the Korea government (MSIT) (No. 2023R1A2C1003001). The funders had no role in the design of the study; collection, analysis, and interpretation of data; decision to publish; and in writing the manuscript. The authors would also like to thank Syspang Co. Ltd, Seoul, for supplying the GLMOC used in this study.

The authors have no conflicting interests.