A pilot study examining a proprietary herbal blend for the treatment of canine osteoarthritis pain

Abstract

Osteoarthritis is the most common joint disease in dogs. Despite the use of nonsteroidal anti-inflammatory drugs (NSAIDs), many owners seek natural therapies; either to augment the response to NSAIDs, or as a replacement. Substantial research has been directed to investigation of novel therapies. A randomized, double-blinded, controlled study was conducted to assess the efficacy of a herbal remedy for treatment of canine osteoarthritis pain. Client-owned dogs (N = 24) with osteoarthritis were enrolled between 2 veterinary hospitals. Each dog underwent veterinary and owner assessment at 0, 4, and 8 weeks, using the Canine Brief Pain Inventory and Hudson activity scale. Blood was collected for a complete blood (cell) count (CBC) and serum chemistry analysis. The product was deemed to be safe and well-tolerated at the manufacturer recommended dosage, with no significant changes seen in the CBC or serum biochemical analyses. Aside from1 dog that developed gastrointestinal upset, all other dogs tolerated the supplement without complication. The supplement did not statistically improve clinical signs in dogs based on veterinary or owner assessments of lameness. There was a treatment/time effect when assessing veterinary pain scores; however, post-hoc analysis suggests no observable benefit of treatment compared with the placebo group at any time point.

Introduction

Osteoarthritis (OA) is the most common joint disease in dogs and often results in pain and joint stiffness. Approximately 20% of the North American canine population over 1 year-of-age is affected, to some degree (1); with clinical symptoms related to chronic pain and inflammation, lameness, and loss of joint function and mobility leading to decreased quality of life (2). Given that OA is a chronic inflammatory disease, management is lifelong, often involving a multi-modal approach to control clinical signs related to chronic pain and subsequent lameness (3). Currently, the standard of care for pharmacological management of osteoarthritis is the use of non-steroidal anti-inflammatory drugs (NSAIDs) to decrease pain and inflammation (3). Most patients tolerate NSAIDs well; however, 5 to 10% of patients must discontinue treatment due to adverse gastrointestinal, renal, or hepatic side effects, and 10 to 12% of patients will potentially not respond to therapy (4). Often in these cases, natural remedies are sought out by owners as an alternate means to control pain with fewer perceived side effects. These remedies may work in a similar way to the typical NSAID (e.g., control of inflammation); however, studies are lacking on the efficacy of these remedies, or whether they can be used together with NSAIDs (5). This clinical trial was conducted to assess the safety and efficacy of a herbal remedy from Pet Wellbeing, called Agile Joints (McDonough, Georgia, USA), on the clinical signs associated with naturally occurring osteoarthritis in client-owned dogs. This proprietary herbal blend contains devil’s claw tuber (Harpagophytum procumbens), sarsaparilla root (Smilax officinalis), dandelion root (Taraxacum officinale), yucca root (Yucca glauca), boswellia gum (Boswellia serrata), turmeric root extract standardized to 10% curcuminoids (Curcuma longa), and bromelain as active herbal ingredients: all of which demonstrate mild to moderate anti-inflammatory properties in both in vitro and in vivo human studies (6–13). In particular, devil’s claw, boswellia, curcumin, and bromelain have been studied as to their effective use for management of OA. Devil’s claw has strong anti-inflammatory properties, with effects on cellular release of nitric oxide (NO) and cytokines, as well as the expression of cyclooxygenase (COX)-1 and COX-2 (6). It has been shown to specifically reduce cytokine expression in human osteoarthritic chondrocytes (14). Devil’s claw has also been the subject of in vivo studies in combination with curcumin and bromelain, providing relief of OA pain in humans (15), as well as improving the condition of canine OA patients when used in combination with curcumin and boswellia (16). In clinical trials, boswellia reduced the severity of pain and lameness associated with both canine (5) and human (17) OA when given at doses of 40 mg/kg/d and 10 to 17 mg/kg/d, respectively. Curcumin has been demonstrated to decrease pain-related symptoms in human patients affected with OA when used at a dose of 5 to 8 mg/kg, q8h (18); also by reducing inflammation in canine OA patients when given doses of 4 mg/kg twice daily through decreasing expression of genes involved in the inflammatory response (5). Bromelain has been effective at reducing OA pain in humans used alone (13) or in combination with devil’s claw and curcumin (15). Except for boswellia and curcumin, the pharmacokinetics of the active ingredients of each herbal constituent of this preparation have not been studied clinically in dogs; with only in vitro data suggesting efficacy.

The objective of this clinical trial was to assess the safety and efficacy of a combination herbal remedy on the clinical signs associated with naturally occurring canine osteoarthritis compared with dogs receiving a placebo over an 8-week period. It was based on the hypothesis that an herbal remedy with the above-mentioned, anti-inflammatory herbals as principal ingredients would reduce symptoms of chronic pain associated with osteoarthritis.

Materials and methods

The study population consisted of client-owned dogs presenting to 1 of 2 veterinary hospitals; the University of Florida College of Veterinary Medicine (UFCVM) or Essex Animal Hospital in Essex, Ontario (EAH) for evaluation and treatment of an OA-induced lameness. All owners received a detailed written description of the study, and owner consent was obtained for enrollment. Dogs were candidates for the study if they had radiographic evidence of OA in one or more joints, clinical signs of pain according to owner assessment, a detectable lameness on visual gait assessment, and elicitable joint pain on examination. The study protocol was approved by the Institutional Animal Care and Use Committee (IACUC) for UFCVM and the Animal Care Committee for EAH.

Over the course of the clinical trial, dogs were permitted to receive NSAIDs and other disease modifying supplements (i.e., fish oil, glucosamine/chondroitin sulfate, injectable PSGAGs, prescription joint health diet) without any change in these treatments for 4 wk prior to or during the entirety of the study period. Owners were asked to continue this regimen throughout the study with no changes. All other analgesic medications (including gabapentin, tramadol, pregabalin, and amantadine) were discontinued at least 1 wk prior to enrollment. Dogs were permitted to continue to receive routine physical rehabilitation during the trial at the same frequency at which they had received therapy for 4 wk before enrollment in the study. All dogs received initial baseline diagnostic tests; including a complete blood (cell) count (CBC) and serum chemistry analysis as well as urinalysis. Serum chemistry values of blood urea nitrogen (BUN), carbon dioxide (CO₂), alkaline phosphatase (ALP), albumin (ALB), alanine transaminase (ALT), aspartate aminotransferase (AST), cholesterol (Chol), globulin (Glob), glucose (Glu), magnesium (Mg), phosphorus (Phos), total protein (TP), calcium (Ca), sodium (Na), potassium (K), total bilirubin (Tbili), chloride (Cl), creatinine (Creat), and albumin/globulin ratio (A/G) were evaluated. Samples were analyzed at the University of Florida Veterinary Diagnostic Laboratories, Clinical Pathology Laboratory (Gainesville, Florida) for UFCVM cases and through Antech Diagnostics (Antech Diagnostics Canada, Calgary, Alberta) for EAH. Dogs were excluded by the study veterinarian if there was evidence of renal, hepatic, neurologic, endocrine, or neoplastic disease.

The study was a randomized, placebo-controlled, owner and veterinarian double-blind, clinical trial. After provision of informed owner consent, 24 dogs were randomly assigned to 1 of 2 groups: placebo or Pet Wellbeing Agile Joints glycerol decoction (PWB decoction). The PWB decoction group received ~17 to 19 mg/kg of total herbal extracts every 12 h (based on the manufacturer’s dosing recommendations) for 8 wk. The PWB decoction contained (per mL) the following herbal concentrations:

• 20 mg bromelain,

• 38 mg boswellia gum,

• 96 mg devil’s claw tuber,

• 38 mg sarsaparilla root,

• 39 mg dandelion root,

• 38 mg yucca root, and

• 115 mg turmeric root extract.

A glycerol placebo was provided by Pet Wellbeing, with identical packaging and labeling as the PWB decoction. Randomization of study participants was performed using a commercially available software application (Randomizer; Rob Sammons, Newcastle upon Tyne, UK). Blood was collected to repeat CBCs and serum chemistry analyses at weeks 4 and 8.

Before treatment initiation and at 4 and 8 wk, each dog was evaluated by the study veterinarian at the respective veterinary hospital, who assessed locomotion, joint pain, weight-bearing, and crepitus on a 5-point scale based on a subjective scoring system as previously reported (19). Baseline assessments included patient signalment, use of NSAID medication, and/or disease-modifying supplements (if applicable), presence of unilateral or bilateral lameness, body weight, and body condition score (BCS) on a scale of 1 to 9. Study veterinarians at both locations had extensive training and experience in the specialty of canine rehabilitation, with active or pending Board certification in the American College of Veterinary Sports Medicine and Rehabilitation. Dogs were also assessed at the same time intervals by their owners via the Canine Brief Pain Inventory (CBPI) and Hudson activity scale (HAS). The CBPI measures owner assessment of the severity and impact of chronic pain on their dogs using a combination of a pain sensitivity scale (PSS) and pain interference scale (PIS). The PSS is comprised of 4 questions, grading severity of pain when at its worst, least, and average over the previous 7 d, and at the time of completion of the questionnaire, using a grading scale of 0 (no pain) to 10 (extreme pain). The PIS is comprised of 6 questions grading several life functions (general daily function, enjoyment of life, ability to rise, walk, run, and climb stairs) as to how pain has interfered with the dog’s ability to complete them, again using a scale of 0 (does not interfere) to 10 (completely interferes). The HAS is a visual analog scale questionnaire for use in assessing pain and lameness in dogs, comprised of 11 questions to evaluate attitude, comfort and mobility.

A commercially available software package was used for statistical analysis (JMP version 10; SAS Cary, North Carolina, USA). Signalment data for treatment and placebo groups were assessed using the Student’s t-test (body weight, body condition) for continuous variables. Ordinal data (age, gender, NSAID use, glucosamine/chondroitin use, and fish oil use) was assessed using Fisher’s exact test. All veterinary and client scoring, CBC, and serum chemistry data were assessed using the Shapiro-Wilk test for normality. The 3 dogs completing the study from EAH (2 in the PWB decoction group and 1 in the placebo group) were excluded from statistical analysis of CBC and serum chemistry parameters due to the small numbers, different clinical laboratories used and the lack of reference range standardization between the2 laboratories, thus confounding data interpretation. Most of the veterinary scoring, canine brief pain inventory, CBC, serum chemistry, and scoring data were normally distributed; therefore, a mixed-model of variance was used to analyze these outcomes including the fixed effects of treatment and time with the dog as a random variable. If treatment, time, or treatment × time effect was significant, post-hoc analysis was performed using a Tukey’s post-hoc analysis to determine significance. A value of P ≤ 0.05 was considered significant.

Results

Twenty-four client-owned dogs with clinically and radiographically confirmed OA were recruited: 20 from UFCVM, and 4 from EAH. Twenty-two of these dogs completed the trial and were included in the analyses. One dog (UFCVM) was removed due to diarrhea and vomiting (PWB decoction); 1 dog (EAH) was removed due to worsening symptoms and the client’s desire to resume previous pain medications (PWB decoction), resulting in 11 dogs completing the study in each treatment group. Within the PWB decoction group, 8/11 dogs were receiving NSAIDs before and during the entirety of the study (5 = carprofen, 1 = firocoxib, 2 = grapiprant); within the placebo group, 4/11 dogs were receiving NSAIDs prior to and during the entirety of the study (3 = carprofen, 1 = grapiprant).

The study population ranged in age from 1 to 14 y and ranged in weight from 5.2 to 35.8 kg, with body condition scores ranging from 4 to 8 on a scale of 1 to 9. Breeds included in the study were: shih tzu (n = 1), Labrador retriever (n = 1), English bulldog (n = 1), pit bull (n = 2), bichon frise (n = 1), chow chow (n = 1), golden retriever (n = 4), Australian shepherd (n = 1), cocker spaniel (n = 1), greyhound (n = 1), and mixed breed (n = 8). A summary of patient signalment data is listed in Table 1 showing no differences in age, gender, BCS, weight, NSAID use, or chondroprotective supplement use.

Table 1.

Signalment of dogs in the treatment and placebo groups assessed using Student’s t-test^a or Fisher’s exact testing.^b

	PWB decoction	Placebo	P-value
Weight (kg)	21 ± 10	27 ± 8	0.16
BCS	6 (4–7)	5 (4–8)	0.41
Age (years)	9 ± 3	8 ± 4	0.47
Sex:
MN	6	5	1
FS	7	4
Glucosamine/Chondroitin	5	5	1
Fish Oil/Omega 3 Enriched Diet	3	5	1
NSAID Use	8	4	0.19
Placebo

No differences were noted between groups; BCS — Body condition score; MN — Male neutered; FS — Female spayed; NSAID — Non-steroidal anti-inflammatory drug.

^aWeight, BCS, age.

^bNSAID use, sex, glucosamine/chondroitin, fish oil.

No significant change was seen with CBPI nor HAS during PWB decoction treatment when compared to baseline Week 0, nor between treatment groups over time. Veterinarian assessed lameness, pain, crepitus, and weight-bearing scores also did not show a statistically significant improvement at Weeks 4 or 8 compared to baseline for the PWB decoction. There was a downward trend in the veterinary pain scores over time in the PWB group as compared to placebo, with a P = 0.04 (Table 2); however, Tukey’s post-hoc analysis of PWB decoction differences were 0.06 between time point 0 and 4 wk and 8 wk, with no differences noted between treatments at any time point. Veterinary assessed pain scores increased in the placebo group at Week 4, and decreased in the PWB decoction group at both Weeks 4 and 8 compared with Week 0 (Figure 1). When comparing pain scores in dogs between placebo and PWB decoction, higher pain scores were detected on veterinary examination in the placebo group at 4 and 8 wk than in the PWB decoction group (Figure 1).

Table 2.

Veterinary assessed scores,^a CBPI, and Hudson Scale mean and standard deviation at weeks 0, 4, and 8 for PWB and placebo oils.

	PWB decoction			Placebo			P-value
	Week 0	Week 4	Week 8	Week 0	Week 4	Week 8	Treatment	Time	tx*time
Lameness (0–5)	2.4 ± 0.7	2.1 ± 0.5	2.3 ± 0.6	2.5 ± 0.5	2.6 ± 0.5	2.5 ± 0.5	0.19	0.71	0.71
Pain (0–5)	2.8 ± 0.6	2.2 ± 0.8	2.1 ± 0.7	2.7 ± 1.2	3.0 ± 1.2	2.8 ± 1.1	0.2	0.07	0.04
Weight-bearing (0–5)	2.2 ± 0.9	2.0 ± 0.4	2.2 ± 0.6	2.3 ± 0.5	2.3 ± 0.5	2.1 ± 0.5	0.61	0.55	0.55
Crepitus (0–5)	1.5 ± 0.9	1.7 ± 0.9	1.6 ± 1.0	1.5 ± 1.0	1.2 ± 1.1	1.5 ± 0.7	0.41	0.69	0.69
CBPI PSS (0–40)	11.8 ± 9.2	8.9 ± 6.1	9.5 ± 9.0	14.8 ± 8.6	13.0 ± 8.2	11.7 ± 8.8	0.33	0.08	0.78
CBPI PIS (0–60)	20.5 ± 10.9	15.2 ± 10.5	15.1 ± 11.9	23.2 ± 15.5	17.9 ± 14.5	21.5 ± 15.8	0.42	0.19	0.48
CBPI total (0–100)	32.3 ± 17.9	23.7 ± 15	24.6 ± 19.7	38 ± 23.4	30.9 ± 21.7	33.3 ± 24.3	0.35	0.12	0.71
Hudson (0–110)	76.9 ± 17.7	80.8 ± 17.5	80.2 ± 20.6	70.6 ± 15.2	79.6 ± 14.0	73.8 ± 17.6	0.51	0.21	0.95

Numbers in bold print represent P-value of statistical significance (≤ 0.05). CBPI — Canine Brief Pain Inventory; PSS — pain sensitivity score; PIS — pain interference score.

^aLameness, pain, weightbearing, and crepitus.

Figure 1.

Veterinary Pain scores in the Pet Wellbeing Agile Joints glycerol decoction.

No significant changes were noted in the measured CBC (Table 3) nor serum chemistry (Table 4) values in either the PWB decoction or placebo treated dogs, with the exception of globulin when analyzed as an effect of treatment × time (P = 0.04), yet no differences were observed with post-hoc analysis.

Table 3.

Complete blood (cell) count, mean and standard deviation at weeks 0, 4, and 8 for PWB and placebo oils.

	PWB decoction			Placebo			P-value
	Week 0	Week 4	Week 8	Week 0	Week 4	Week 8	Treatment	Time	tx*time
WBC (5 to 13 K/μL)	6.886 ± 1.975	7.268 ± 2.104	7.541 ± 2.932	8.228 ± 1.884	8.248 ± 1.874	8.379 ± 2.076	0.27	0.57	0.31
RBC (5.7 to 8.3 M/μL)	7.012 ± 0.845	7.073 ± 0.668	6.897 ± 0.743	7.038 ± 0.994	6.880 ± 1.154	6.891 ± 1.217	0.89	0.53	0.61
HGB (14 to 20 g/dL)	17.09 ± 1.86	17.28 ± 1.66	16.83 ± 1.53	16.69 ± 2.22	16.68 ± 2.83	16.65 ± 2.72	0.69	0.73	0.79
HCT (40 to 56%)	48.19 ± 5.77	48.53 ± 4.62	47.09 ± 4.81	47.93 ± 7.06	46.72 ± 7.35	46.72 ± 8.23	0.77	0.38	0.59
PLT (134 to 396 K/μL)	260.3 ± 73.7	264.7 ± 90.3	259.9 ± 103.1	302.4 ± 155.2	263.0 ± 155.0	276.6 ± 138.0	0.73	0.59	0.48
Fibrinogen (0.1 to 0.4 g/dL)	0.19 ± 0.15	0.14 ± 0.10	0.27 ± 0.22	0.15 ± 0.11	0.26 ± 0.22	0.13 ± 0.14	0.79	0.53	0.46
Neutrophils (2.7 to 8.9 K/μL)	4.522 ± 1.610	4.790 ± 1.489	5.024 ± 2.453	5.101 ± 1.431	5.435 ± 1.224	5.265 ± 1.372	0.79	0.53	0.46
Lymphocytes (0.9 to 3.4 K/μL)	1.562 ± 0.524	1.577 ± 0.556	1.547 ± 0.598	2.125 ± 1.016	1.982 ± 0.835	2.191 ± 0.949	0.13	0.64	0.47
Monocytes (0.1 to 0.8 K/μL)	0.287 ± 0.113	0.357 ± 0.126	0.342 ± 0.220	0.407 ± 0.161	0.399 ± 0.252	0.385 ± 0.104	0.22	0.82	0.67
Eosinophils (0.1 to 0.3 K/μL)	0.494 ± 0.418	0.522 ± 0.257	0.617 ± 0.323	0.566 ± 0.355	0.413 ± 0.353	0.471 ± 0.286	0.67	0.38	0.14
Basophils (0.1 to 1.3 K/μL)	0.014 ± 0.014	0.013 ± 0.012	0.012 ± 0.016	0.019 ± 0.017	0.013 ± 0.015	0.023 ± 0.026	0.41	0.6	0.49

WBC — White blood cells; RBC — Red blood cells; HGB — Hemoglobin; HCT — Hematocrit; PLT — Platelets.

Table 4.

Serum chemistry, mean and standard deviation at weeks 0, 4, and 8 for PWB and placebo oils.

	PWB decoction			Placebo			P-value
	Week 0	Week 4	Week 8	Week 0	Week 4	Week 8	Treatment	Time	tx*time
BUN (8 to 25 mg/dL)	21.7 ± 9.6	21.4 ± 7.9	18.1 ± 4.4	25.3 ± 17.6	26.9 ± 18.3	24.8 ± 17.0	0.06	0.56	0.61
CO2 (18 to 28 mEq/L)	21.2 ± 2.5	21.3 ± 3.2	22.8 ± 1.9	21.8 ± 3.1	21.0 ± 3.5	21.3 ± 3.5	0.61	0.68	0.27
ALP (8 to 114 IU/L)	29.2 ± 15.3	29.9 ± 19.9	27.7 ± 11.7	76.6 ± 90.9	65.6 ± 72.7	52.6 ± 36.4	0.24	0.06	0.33
ALB (2.9 to 3.8 g/dL)	3.020 ± 0.325	3.051 ± 0.304	2.957 ± 0.338	2.969 ± 0.325	3.019 ± 0.315	3.002 ± 0.338	0.53	0.73	0.25
ALT (18 to 64 IU/L)	51.3 ± 42.1	57.1 ± 58.7	52.3 ± 37.0	59.1 ± 45.4	55.3 ± 41.2	56.3 ± 40.0	0.84	0.7	0.42
AST (15 to 52 IU/L)	25.6 ± 8.2	24.4 ± 7.8	25.0 ± 6.2	29.9 ± 10.6	27.1 ± 7.4	26.7 ± 11.5	0.51	0.31	0.88
Chol (124 to 334 mg/dL)	226.1 ± 35.2	241.9 ± 38.3	239.0 ± 45.3	205.3 ± 45.8	214.4 ± 45.6	224.7 ± 30.7	0.17	0.28	0.6
Glob (2.2 to 4.2 g/dL)	2.77 ± 0.21	3.01 ± 0.24	2.84 ± 0.25	2.86 ± 0.37	2.83 ± 0.32	2.98 ± 0.28	0.2	0.48	0.04
Glu (79 to 120 mg/dL)	88.7 ± 9.1	96.4 ± 8.7	96.1 ± 12.9	90.5 ± 11.1	96.5 ± 9.6	93.0 ± 4.9	0.12	0.69	0.31
Mg (1/3 to 2.5 mg/dL)	2.04 ± 0.21	2.10 ± 0.19	2.06 ± 0.13	2.22 ± 0.56	2.29 ± 0.63	2.21 ± 0.58	0.1	0.4	0.72
Phos (2.7 to 5.6 mg/dL)	3.70 ± 0.81	3.94 ± 0.67	3.74 ± 0.57	4.2 ± 1.51	4.48 ± 1.82	4.54 ± 1.43	0.27	0.39	0.66
TP (5.6 to 7.5 g/dL)	5.78 ± 0.39	6.04 ± 0.37	5.82 ± 0.34	5.83 ± 0.44	5.85 ± 0.38	5.97 ± 0.30	0.28	0.76	0.13
Ca (8.9 to 10.8 mg/dL)	9.62 ± 0.28	9.70 ± 0.37	9.77 ± 0.39	9.70 ± 0.39	9.83 ± 0.23	9.68 ± 0.42	0.46	0.63	0.41
Na (142 to 153 mEq/L)	147.08 ± 1.80	147.06 ± 2.80	147.43 ± 1.08	147.550 ± 1.771	147.270 ± 1.944	146.63 ± 1.68	0.94	0.6	0.59
K (3.5 to 5.2 mEq/L)	4.56 ± 0.35	4.57 ± 0.21	4.49 ± 0.28	4.67 ± 0.58	4.73 ± 0.60	4.94 ± 0.55	0.49	0.59	0.08
Tbili (0.1 to 0.4 mg/dL)	0.20 ± 0.07	0.17 ± 0.05	0.19 ± 0.03	0.18 ± 0.04	0.18 ± 0.04	0.20 ± 0.10	0.29	0.47	0.43
Cl (107 to 115 mEq/L)	113.79 ± 2.33	113.82 ± 1.53	113.32 ± 1.56	112.71 ± 2.57	113.00 ± 2.26	112.79 ± 2.69	0.84	0.29	0.85
Creat (0.6 to 1.7 mg/dL)	1.131 ± 0.397	1.016 ± 0.247	1.083 ± 0.249	1.143 ± 0.235	1.184 ± 0.240	1.117 ± 0.183	0.67	0.97	0.23
A/G (0.7 to 1.5)	1.10 ± 0.15	1.03 ± 0.12	1.06 ± 0.17	1.06 ± 0.20	1.09 ± 0.22	1.03 ± 0.20	0.51	0.63	0.21

Numbers in bold print represent P-value of statistical significance (≤ 0.05).

BUN — Blood urea nitrogen; CO₂ — Carbon dioxide; ALP — Alkaline phosphatase; ALB — Albumin; ALT — Alanine aminotransferase; AST — Aspartate aminotransferase; Chol — Cholesterol; Glob — Globulin; Glu — Glucose; Mg — Magnesium; Phos — Phosphorus; TP — Total protein; Ca — Calcium; Na — Sodium; K — Potassium; Tbili — Total bilirubin; Cl — Chlorine; Creat — Creatinine; A/G — Albumin/Globulin ratio.

Discussion

The main objective of this study was to perform an owner and veterinary double-blinded, placebo-controlled, clinical study to determine the safety and efficacy of Pet Wellbeing Agile Joints decoction in dogs with clinically apparent osteoarthritis. The product was deemed to be safe and well-tolerated at the manufacturer recommended dosage, with no significant changes seen in either CBC or serum biochemical analyses (Tables 3 and 4). Aside from 1 dog that developed gastrointestinal upset, all other dogs tolerated the supplement without complication, despite the concurrent use of NSAID medications and other nutritional supplements, suggesting that this supplement can be safely used as part of a multimodal management program.

Although the supplement did not statistically improve clinical signs nor lameness in dogs based on owner assessments, veterinary assessed pain scores were lower in the PWB decoction group than in placebo group at both 4 and 8 wk. Interpretation of a pain response is highly subjective for dogs, often relying on changes in behavior that may also indicate stress or anxiety (such as a change in breathing pattern when handled). Therefore, it is possible that lowering of patient stress could confound interpretation of the patient response to palpation of affected joints. Current literature is lacking in studies of the effects of these herbal-based medications on physiological parameters that can reduce anxiety. However, the possibility remains that these herbals may work alone or synergistically to reduce anxiety.

Limitations of this study were the inclusion/exclusion criteria reflecting inclusion of canine patients with multiple joint OA, use of a small sample size, several patients on pre-existing supportive therapies, and inability to standardize pre-existing NSAIDs.

It is also possible that the small number of patients in this study further limited the power of the study to detect subtle improvements. A 2002 study evaluated the use of carprofen in 6 dogs over a 28-day period, with significant improvement noted on subjective assessment at 28 d compared with entry into the study (20). In 2003, Moreau et al (21) demonstrated improvement in subjective owner assessment after 30 d of NSAID use, specifically meloxicam, in a study of similar population size to the current study. In the same study, owners subjective assessments did not reveal improvement in patients receiving carprofen therapy, but both NSAIDs produced statistically significant improvement on subjective veterinary assessment and kinematic gait analysis. It has previously been demonstrated that there is no correlation or concordance between the PSS or PIS score changes in owner subjective assessments and changes in kinematic data (22); therefore, inclusion of kinematic gait analysis in a future study of this proprietary blend may provide additional insight as to how effective this supplement is at decreasing osteoarthritis pain. Since veterinary pain scores were lower in the treated group in the current study, with significant improvements noted as treatment progressed, it is possible that continuing the study for a longer period may also have shown some significant improvement in the treated group. A systematic review of long-term NSAID use in the treatment of canine osteoarthritis concluded that there is evidence of progressive decreases in pain with long-term NSAID use for dogs with chronic OA (20). Given that the herbs in the PWB decoction used in this study have similar mechanisms of action of decreasing inflammation as NSAIDs (6–13), it is possible the same effect would be noted with longer-term use. The fact that the PWB decoction may affect the same mediators of inflammation as NSAIDs may have confounded results, as 55% of dogs in this study were receiving NSAIDs. Removal of NSAIDs would have been a better study as these supplements may recapitulate NSAID use, but the standard of care would have been compromised due to the need for removal of NSAIDs for at least 2 wk to allow for proper drug elimination before enrolling in the study. This approach would allow for more clear observation of the effects of the addition of the supplement from an efficacy perspective, but our results suggest that this product is safe during concurrent NSAID use.

Another limitation of this study is that no objective measurements assessing local or systemic inflammation were performed. Devil’s claw has been the subject of several in-vitro trials exploring its effects on proinflammatory mediators involved in the pathogenesis of OA including nitric oxide (NO), tumor necrosis factor alpha (TNFα), interleukin (IL)-6, IL-1β, and prostaglandin E₂ (PGE₂) (6,14,23). The results of these studies suggest that devil’s claw has significant effect on inhibiting NO and IL-6 release and TNFα synthesis, as well as strongly inhibiting the expression of COX-1 and COX-2, all of which have been demonstrated to be involved in the pathophysiology of canine OA (24,25). In similar fashion, sarsaparilla (7), dandelion (8), yucca (9), turmeric (11,26), and bromelain (12) have been shown to exhibit anti-inflammatory effects on cytokine and/or NO release, suggesting that these could be effective treatments for canine OA patients. In 2005, Secor et al (12) demonstrated anti-inflammatory effects of bromelain using a murine model of allergic airway disease, using doses of 2 and 6 mg/kg twice daily. The current study used doses of ~0.9 mg/kg twice daily. In 2019, Khan et al (7) investigated the anti-inflammatory and analgesic effects of sarsaparilla in mice using 200 and 400 mg/kg; in contrast, the sarsaparilla doses used in this study ranged from 1.7 to 1.9 mg/kg/d. Boswellia has been demonstrated to inhibit activity of the enzyme 5-lipoxygenase (5-LO) (10,27) another important inflammatory mediator in the pathogenesis of OA (28). Future studies of this product including monitoring of effects on cytokine release, NO, COX, and/or 5-LO expression may provide additional information as to the efficacy of this or similar products. In addition, there is a paucity of data on both effective doses of the individual herbal components of this product and the efficacy of similar combination products when used to treat canine OA. The few available canine studies, in combination with studies of the effects of these herbal components in human OA, suggest that the doses of some or all the individual herbs in this product may be sub-optimal. In 2011, a review of the clinical relevance and safety of devil’s claw in managing human OA revealed that studies yielding the best results used standardized extracts containing between 50 to 100 mg of harpagoside daily; the major chemical component thought to be responsible for its anti-inflammatory activity (29). Using an average human weight range of 60 to 100 kg, this equates to ~1 mg/kg of harpagoside. A 1996 study quantified the harpagoside content of devil’s claw, confirming the stipulations of both the Swiss Pharmacopeia and Pharmeuropa that 1.2% harpagoside is a suitable minimum level for standardization of Harpagophytum procumbens (30). Based on this information, the dose of harpagoside administered in this study was 0.05 to 0.06 mg/kg (1.2% of the total dose of 4.3 to 4.8 mg/kg), which is markedly lower than that which has proven effective in human models. Using a rat model of inflammation, Khan et al (7) demonstrated significant anti-inflammatory and analgesic properties to sarsaparilla when dosed at 200 to 400 mg/kg. This dose is higher than the dose of sarsaparilla used in the combination product in this study (3.4 to 3.8 mg/kg/d). Similarly, rat and mouse models were used to evaluate in vivo anti-inflammatory effects of taraxasterol, one of main active components isolated from dandelion root, demonstrating a dose-dependent reduction in inflammation when dosed at 2.5, 5 and 10 mg/kg/d (31). Using a taraxasterol content of 0.0299% as established in 2016 (32), the dose of taraxasterol received by the dogs in this study was only 0.001 mg/kg, with a daily dose range of 3.4 to 3.8 mg/kg/d of dandelion root; again, substantially lower than the dose shown to decrease inflammation. Bromelain has been demonstrated to be an effective treatment for mild to moderate knee OA in humans when dosed at 500 mg/d (~5 to 8 mg/kg) for 16 wk (13); a total daily dose of 1.7 to 1.9 mg/kg of bromelain was used in the current study. In human osteoarthritis of the knee, boswellia doses of 333 mg three times daily (10 to 17 mg/kg/d) for 8 wk were shown to significantly reduce joint pain, improve range of motion, and increase walking distance (17). In 2004, Reichling et al (5) demonstrated good effect at managing canine OA using a standardized boswellia extract (≥ 50% triterpenic acids) at a dose of 400 mg/10 kg of body weight once daily; the dose of boswellia used in the combination product in this study was markedly lower, at 3.4 to 3.8 mg/kg/d. In addition, the concentration of triterpenic acids, particularly 11-keto-β-boswellic acid (KBA) and acetyl-11-keto-β-boswellic acid (AKBA), the boswellic acids known to be the most potent inhibitors of 5-LO (27), is not known in the extract used in this study, further confounding results. A 2020 study exploring the use of curcumin alone or in combination with palmitoyl-glucosamine failed to show significant improvement in a rat model of OA using a dose of 10 mg/kg (33), suggesting that the dose of curcumin in this combination product of 10.4 to 11.6 mg/kg/d may be below the effective therapeutic level. In the human literature, oral administration of 500 mg of curcuminoids 3 times daily (15–25 mg/kg/d) for 12 wk reduced pain-related symptoms in patients with OA (18), also suggesting that the dose in the PWB decoction was lower than that required to effectively reduce OA pain. Curcumin has shown therapeutic effects on OA pain but its use has been hampered by its poor intestinal absorption, low systemic bioavailability, and rapid metabolism (34,35). Various strategies have been proposed to enhance bioavailability and some of these strategies have proven effective (34). In recent studies, doses of 100 to 150 mg of bioavailable curcumin have shown promise in human OA treatment (34). Unfortunately, the formulation and bioavailability of the curcumin in the PWB product are unknown. The current study was also limited by the fact that it did not record whether the supplements were given with or without food, as this can also affect GI absorption; however, overall information on GI absorption in dogs is limited. A pharmacokinetic study of this formula would need to be conducted to determine if an appropriate dose of curcumin was administered, particularly given the rapid metabolism for the dog when compared to humans.

A 2018 study (18) demonstrated increased efficacy of a combination of 350 mg of curcumin (7.5 to 12.5 mg/kg/d) and 150 mg of boswellia (4.5 to 7.5 mg/kg/d) 3 times daily for 12 wk in reducing pain-related symptoms in humans with OA compared with curcumin alone. This suggests there may be a synergistic effect. In 2014, Conrozier et al (15) demonstrated a clinically relevant improvement of chronic OA pain in human patients treated with 1300 mg (13 to 22 mg/kg) daily of a combination of devil’s claw, curcumin, and bromelain for 60 d. Also in 2014, a natural health product (NHP) containing devil’s claw, boswellia, and curcumin was demonstrated to improve the functional ability in dogs afflicted by naturally occurring OA, with an improved effect noted at 8 wk compared with 4 wk (16). Dogs in this study received between 9.6 to 20 mg/kg/d of both devil’s claw and boswellia for the first 4 wk of the study as part of a combination product which also contained Ribes nigrum, Salix alba, Tanacetum parthenium, and omega-3 polyunsaturated fatty acids (PUFA). During weeks 4 to 8, dogs received 2.4 to 5 mg/kg/d of devil’s claw, 7.2 to 15 mg/kg/d of boswellia, and 1.4 to 2.9 mg/kg/d of curcumin, as part of a combination product also containing Ribes nigrum, Ananas comosus, omega-3 PUFA, glucosamine sulfate, methylsulfonylmethane, chondroitin sulfate, L-glutamine, and hyaluronic acid. Comparatively, dogs in the current study being administered PWB decoction received 8.6 to 9.6 mg/kg/d of devil’s claw, 3.4 to 3.8 mg/kg/d of boswellia, and 10.4 to 11.6 mg/kg/d of curcumin for the entire 8 wk of study duration. Patients in the current study were allowed to continue to take disease modifying supplements while enrolled; including omega-3 PUFA, which has been documented to improve weight-bearing in dogs with OA (36). However, there was no standardization of this or any other supplements which is acknowledged to be another limitation of the study design. In contrast, the dogs in the 2014 NHP study were absent any compound used for treatment of clinical signs of OA, with washout periods of 4 to 12 wk; however, the NHP product included a consistent dose of omega-3 PUFA for all 8 wk of the study, which could in part account for the more significant improvement seen with that product. Since both the product of the current study, Agile Joints from Pet Wellbeing, and the NHP product in the 2014 study were well-tolerated and shown to be safe at the current dose recommendations, additional studies exploring the safety and efficacy of higher doses of Agile Joints in dogs as a first line of treatment with no other NSAIDs or supplements may be useful in determining its efficacy for management of the signs of canine OA.

In conclusion, this study demonstrated the PWB Agile Joints decoction to be safe and well-tolerated at the current label dose recommendations, with a trend towards reduction in veterinary assessed pain scores noted over the duration of the study, although the clinical significance of this remains unknown as owners did not perceive improvement in treated dogs. Future studies structured at identifying effective doses of each component of this formulation specific to treatment of canine OA, as well as effects on cytokine release, NO, COX, and/or 5-LO expression in canine patients being treated with PWB Agile Joints may provide additional information as to the most effective way to use this combination product to maximize patient response. Further long-term studies of this product with a larger population of dogs with only 1 affected limb/joint, and including kinematic gait analysis are warranted to better understand the utility of PWB Agile Joints in clinical OA as a single modality for treatment. CVJ