twitter

Sunday, 10 June 2018

A medicinal herb-based natural health product improves the condition of a canine natural osteoarthritis model: A randomized placebo-controlled trial

Research in Veterinary Science Volume 97, Issue 3, December 2014, Pages 574-581 Author links open overlay panelMaximMoreauabBertrandLussierabJean-PierrePelletierbJohanneMartel-PelletierbChristianBédardcDominiqueGauvinabEricTroncyab a Research Group in Animal Pharmacology of Quebec (GREPAQ), Department of Veterinary Biomedical Sciences – Faculty of Veterinary Medicine, Université de Montréal, P.O. Box 5000, Saint-Hyacinthe, Quebec J2S 7C6, Canada b Osteoarthritis Research Unit, University of Montreal Hospital Research Centre (CRCHUM), Tour Viger, 900 St-Denis Street, Montreal, Quebec H2X 0A9, Canada c Department of Pathology and Microbiology – Faculty of Veterinary Medicine, Université de Montréal, P.O. Box 5000, Saint-Hyacinthe, Quebec J2S 7C6, Canada Received 17 February 2014, Accepted 15 September 2014, Available online 28 September 2014. crossmark-logo https://doi.org/10.1016/j.rvsc.2014.09.011 Get rights and content Highlights • We evaluated a herb-based natural health product in osteoarthritis-afflicted dogs. • We performed a randomized placebo-controlled trial. • We examined changes in the functional ability over a duration of 8 weeks. • Administration of a herb-based natural health product improved the dog condition. Abstract An oral herb-based natural health product (NHP) was evaluated in the canine natural osteoarthritis model. At baseline, the peak vertical force (PVF, primary endpoint) and case-specific outcome measure of disability (CSOM) were recorded in privately-owned dogs. Dogs (16/group) were randomized to receive NHP formulations or a negative control. The PVF was measured at week (W) 4 and W8. Daily locomotor activity was recorded using accelerometer. The CSOMs were assessed bi-weekly by the owner. The NHP-treated dogs (n = 13) had higher PVF at W4 (p = 0.020) and W8 (p <0.001) when compared to baseline. The changes at W8 were higher than control dogs (n = 14, p <0.027) and consistent with Cohen's d effect size of 0.7 (95% confidence interval: 0.0–1.5). The NHP-treated dogs had higher locomotor activity at W8 (p = 0.025) when compared to baseline. No significant change was observed for the CSOM. The NHP improved the clinical signs of osteoarthritis in this model. Previous article Next article Keywords Force platform Locomotor activity Phytochemical compounds Naturally-occurring osteoarthritis Canine model 1. Introduction Osteoarthritis (OA) is by far the most common human musculoskeletal disease, affecting millions worldwide (Lawrence et al., 2008). The prevalence of OA in dogs is also high, particularly in geriatric animals, being estimated to be five times that observed in mature adults (Shearer, 2011). In dogs, OA results mainly from traumatic insults to the cranial cruciate ligament (CCL), and hip or elbow dysplasia (McLaughlin, 2001; Roush, 2001). Cascades of biological and biomechanical events then merge to induce and perpetuate structural changes at the level of the entire joint, which, as in humans, lead to crippling pain, disability and poor quality of life (Cook, 2010; Johnston, 2001; Madsen and Svalastoga, 1994; Martinez, 1997; Martinez and Coronado, 1997). Naturally-occurring models of OA have been proposed to accelerate the development of human therapeutics (Pelletier et al., 2010), and a recent review of experimental data underlined the high translationability of outcomes obtained from canine OA models, in particular the response to treatment (Moreau et al., 2013). Undertaking a trial in privately-owned dogs afflicted by natural OA would provide preclinical data and additional evidence on the therapeutic potential of new compounds under development. Of note, the potential of several therapeutic approaches has been tested in different randomized controlled trials (RCTs) in the canine natural OA model using force platform gait analysis as an outcome measure of pain-related functional impairment. These tested compounds include non steroidal anti-inflammatory drugs (NSAIDs) (Budsberg et al., 1999; Moreau et al., 2003, 2007), therapeutic diets (Moreau et al., 2012b; Rialland et al., 2013; Roush et al., 2010) as well as natural substances (naturaceuticals) used to restore or maintain good health status (Hielm-Bjorkman et al., 2009; Moreau et al., 2004, 2012a). The latter therapeutic class is considered by the authors as natural health products (NHPs) which originate from plants, fruits and vegetables, animals, microorganisms and marine sources. Currently, no effective therapy seems able to alleviate the clinical signs of OA in humans or dogs. As relief of pain and the preservation of joint structure cannot be claimed with certainty for currently approved treatments, there is a need for effective strategies to improve the condition of afflicted patients. Medicinal herbs have long been used in traditional medicine and there is considerable evidence that such NHP and their derivatives may play beneficial roles in OA (Mobasheri, 2012). Harpagophytum procumbens, also known as devil's claw, is a South African plant which includes harpagoside as one of its major biologically active phytochemical compounds. A large body of evidence supports the efficacy of harpagoside and related extracts in alleviating symptoms of OA in humans (Gagnier et al., 2004). Resin extracts from the Boswellia serrata tree have been demonstrated to be effective in alleviating the clinical signs of OA in humans (Kimmatkar et al., 2003) and dogs (Reichling et al., 2004). Active phytochemical compounds isolated from Ribes nigrum leaves showed anti-inflammatory properties in-vivo in chrondrocyte assays (Garbacki et al., 2002), while its seed oil was an effective treatment for active rheumatoid arthritis (Leventhal et al., 1994). Salix alba extracts have recently been reported to have in-vitro chondroprotective properties in primary canine articular chondrocyte culture (Shakibaei et al., 2012). These extracts seem also to be potent in counteracting low back pain in humans (Gagnier et al., 2007). In rodent models of inflammation, an extract from Tanacetum parthenium demonstrated antinociceptive and anti-inflammatory effects (Jain and Kulkarni, 1999). Classified as a herb, bromelain is a digestive enzyme found in the stem and the fruit of Ananas comosus. This herb has been shown to have anti-inflammatory properties mediated through prostaglandin synthesis (Lotz-Winter, 1990). Finally, curcumin, which is the main biologically active phytochemical compound of Curcuma longa, showed inhibitory actions against major inflammatory mediators (Aggarwal et al., 2013; Henrotin et al., 2013; Mathy-Hartert et al., 2009; Mobasheri et al., 2012) while being effective in reducing pain in OA knee patients (Kuptniratsaikul et al., 2009; Madhu et al., 2013). In agreement with those findings, a recent Cochrane systematic review concluded to potential benefits of oral herbal medicines, being more effective than placebo (Cameron and Chrubasik, 2014). However, as also highlighted, further high quality, fully powered studies are required to gain insight in the therapeutic potential of medicinal plants as well for other NHPs (Vandeweerd et al., 2012). These studies suggest that NHP formulations containing the aforementioned medicinal herbs as principal ingredients might be useful in the management of OA. Whether or not such formulations are effective against the functional impairment that prevails in a model of natural OA needs to be scrutinized rigorously. With the scope of providing strong evidence-based findings, the aim of this RCT was to assess NHP formulations in the canine natural OA model when compared with dogs receiving a placebo over an 8-week duration. 2. Materials and methods 2.1. Design and subject selection This study was a randomized, double-blind, parallel-group, placebo-controlled trial. Dogs were evaluated over either 56 or 61 days depending on the balanced attribution of locomotor activity recording (see Section 2.3). The trial was conducted under the approbation of the Institutional Animal Care and Use Committee (#Rech-1437) in accordance with the guidelines of the Canadian Council on Animal Care. All owners provided written informed consent. Adult dogs weighed more than 20 kg and had radiographic evidence of OA exclusively at the hip or stifle joints. Radiographs (hips, stifles, and elbows) were obtained under sedation as previously described (Moreau et al., 2010). Hind limb lameness in association with the presence of OA was confirmed by veterinary surgeons. At the time of screening, all dogs were free of any compound purported to relieve the clinical signs of OA according to washout periods ranging between 4 and 12 weeks. Hence, a 4-week washout period was respected for oral NSAIDs and a 6-week period for NHPs including fatty acid supplements, OA therapeutic diets or treats. Dogs having received injectable pentosan polysulfate sodium or corticosteroid 1 year before the screening visit were not eligible. A 12-week washout period was requested for injectable polysulfated glycosaminoglycan and hyaluronan, and for oral or topical corticosteroid. During the study, dogs were exempted from the administration of any type of medication except those prescribed for exo- and endoparasite control. Additional exclusion criteria were as follows: dogs with surgical repair of the cranial cruciate ligament within 1 year prior to study initiation, dogs suffering from neurologic or other musculoskeletal lesions, dogs that underwent orthopedic surgery within the past year and dogs with CCL disease having gross instability (positive drawer motion upon orthopedic examination). 2.2. Complete blood count and biochemistry panel To ensure that some parameters were within normal limits during the study, each dog underwent routine blood hematological and biochemical analyses in order to evaluate health status at study initiation (baseline, day 0) as well as at week 4 (day 28) and week 8 (day 56). A veterinary clinical pathologist examined all blood counts and biochemistry panels. Many herbs can increase the risk of bleeding through anti-platelet properties (Samuels, 2005). The buccal mucosal bleeding time is a simple test commonly used in the clinical setting to detect platelet dysfunction in dogs (Callan and Giger, 2001). Each dog underwent a buccal mucosal bleeding time procedure at baseline and at week 8. Mucosal punctures were performed on the upper labial mucosa, using a disposable, fully automated incision device (Surgicutt® Bleeding Time device, International Technidyne Corporation, USA). This device provided a controlled incision of 1.0 mm (depth) per 3.5 mm (length). The time of incision was noted, and circular filter paper (Whatman®, USA) was held 1–2 mm away from the incision to blot the blood, taking care not to disrupt the clot, or to allow blood to drip into the dog's mouth. The end point was when the incision stopped bleeding. Normal buccal mucosal bleeding time is defined to be less than 3 minutes. 2.3. Randomization, blinding and therapy regimen Thirty-two privately-owned dogs were randomly allocated in two equal groups (placebo or NHP) according to a permuted-block randomization procedure, which included six blocks of four treatment possibilities (A or B) distributed in a 1-to-1 ratio (i.e. AABB, ABAB, ABBA, BBAA, BABA and BAAB). Among those blocks, eight were randomly selected using random integers to define the treatment allocation sequence. Also, seven blocks were randomly selected using random integers to allocate seven motor activity recordings to treatment A and seven others to treatment B. The 32 treatment allocations (with or without locomotor activity recording) were transcribed on individual cards in sequentially numbered, sealed, opaque envelopes to ensure concealment. A third party was responsible for the randomization process and for the treatment preparation. At the trial site, both treatments were labeled exclusively as treatment A or treatment B and were encapsulated identically. The trialists, the animal health technicians and all dog owners were blinded to which treatment (A or B) was given to each dog. The key code revealing what referred to treatments A and B remained confidential with the third party and was revealed only after study completion and preliminary analyses. The ingredients of the NHP formulations are described in Table 1. Dogs allocated to the NHP formulations received the Alpha formulation from day 1 to day 29, and then received the Beta formulation from day 29 to day 56. The dosing regimen was as follows: one capsule for dogs <25.0 kg; two capsules for dogs 25.0–39.9 kg; three capsules for dogs 40.0–49.9 kg; four capsules for dogs 50.0–59.9 kg and five capsules for dogs >59.9 kg. Dogs allocated to the negative (placebo) control received capsules filled of excipient to match the amount of the NHP formulations. The negative control (placebo) was given under the same dosing regimen as for the NHP formulations. Table 1. Ingredients include in each natural health product formulations. Ingredients (mg/capsule) Formulations Minimal contents Alpha Beta Medicinal herbs  Harpagophytum procumbens 240.0 60.0 Harpagosides 2.7%  Boswellia serrata 240.0 180.0 Boswellic acid 79.2%  Ribes nigrum 60.0 60.0 Rutines 1%  Salix alba 50.0 – Salicin 1%  Tanacetum parthenium 50.0 – Parthenolide 0.2%  Ananas comosus – 40.0 2000–2500 GDU  Curcuma longa – 35.0 Curcuminoids 95% Omega-3 PUFA  Total 40.0 40.0  Eicosapentaenoic acid 0.4 0.4  Docosahexaenoic acid 9.0 9.0 Others  Glucosamine sulfate – 300.0  Methylsulfonylmethane – 90.0  Chondroitin sulfate – 60.0  L-glutamine – 30.0  Hyaluronic acid – 15.0 Excipient 228.0 280.0 Total weight/capsule 908.0 1190.0 GDU, Gelatin digesting unit; PUFA, Polyunsaturated fatty acids. 2.4. Force platform measurement Peak of the vertically-oriented ground reaction force (PVF) was measured at baseline (day 0), week 4 (day 28) and week 8 (day 56) at the trot (1.9–2.2 m/s) using a force platform, as previously described (Moreau et al., 2010). The PVF was defined as the primary endpoint of the study. Normalized PVF values in percentage of body weight (%BW) from the first five valid trials were used for statistical purposes. To be eligible, dogs must have at least one-hind limb with PVF value lower than 66.0 %BW. This value was consistent to −1 standard deviation (SD) of the PVF value measured in normal dogs (Madore et al., 2007). When unilateral or bilateral lameness was observed, the hind limb having the lowest PVF value determined which one was selected for evaluation. This limb was defined as the most affected limb and was used in the subsequent follow-up of the study. The hind limb selected for evaluation must have been in accordance with orthopedic examination findings, otherwise the dog was excluded. The change in PVF was the mean difference between week 8 − baseline values. 2.5. Locomotor activity recording Accelerometer-based motor activity recording was accomplished using the Actical® system (Bio-Lynx Scientific Equipment Inc., Canada) as described (Rialland et al., 2012). According to the balanced attribution of motor activity recording, collar-mounted accelerometers were worn by 14 dogs for the entire treatment duration (61 days, 24 h/day) which included a baseline period (day −4 to day 0) that preceded the initiation of treatment administration. This period was used to establish baseline level of locomotor activity recording before treatment. Over the 61 days, the motion was continuously recorded every 2 minutes, giving 720 recordings per day. Daily duration of active period (DDAP) referred to the time spent (expressed in hour per day) when the recording exceeded 30 (no unit) in term of intensity. This cut-off value was based on intern data and was used to discern active from inactive period (Moreau et al., 2011; Rialland et al., 2012). Among the 61 days of continuous recording, three periods were predefined: baseline (day −4 to day 0), week 4 (day 26 to day 28) and week 8 (day 54 to day 56). Owners of dogs allocated to the locomotor activity recording were requested to attend an additional fourth appointment. 2.6. Case-specific outcome measure of disability (CSOM) Assessment of at-home functional disability was accomplished using CSOM as previously described (Moreau et al., 2012a; Rialland et al., 2012, 2013). Owners assessed the ability of their dogs to perform two to five activities, and scored on a five-point Likert-type scale for each activity that ranged from no problem (zero) to full incapacity (four). Each activity was selected by the owner according to his/her own perception of what characterized the disability of the dog. Assessments were conducted twice weekly using a specific form that remained in the possession of the owner. For each dog, median of the activities scores was determined for each assessment, giving a total of 17 median CSOM scores over the study. Among all the assessments, three periods were predefined: baseline (assessment on day 0), week 4 (assessments on days 24, 28 and 31) and week 8 (assessments on days 49, 52 and 56). 2.7. Statistical analysis All statistical tests were two-tailed with significance determined by reference to a 5% threshold. Normality of the data was tested using Shapiro–Wilk test. Data were log-transformed when requested to assure transformed data Gaussian distribution. Equality of efficacy was the null hypothesis based on the PVF (primary endpoint) as measured for the hind limb having the lowest value. Per trial log-transformed PVF values were analyzed with a repeated-measures general linear mixed model that included two fixed factors (time and group) and their interaction (time × group interaction), with trials and dogs nested in treatment group as random effects. The change in log-transformed PVF values (week 8 − baseline) were analyzed with a repeated-measures general linear mixed model that included group as fixed factor with trials as random effect. Log-transformed DDAP were analyzed similarly to PVF (period and group as fixed factors) and their interaction (period × group interaction) with days and dogs nested in treatment group as random effects. A repeated-measures generalized linear model was used to analyze median CSOM data under Poisson distribution function using independent working matrix. Fixed factors were period and group and their interaction (period × group interaction) with assessments and dogs nested in treatment group as random effects. Scale factor was estimated by Pearson's chi-square. Covariance structures were defined as recommended (Littell et al., 2000). All post hoc analyses were conducted with appropriate Bonferroni adjustments. Data are presented as mean (SD). 2.8. Sample size calculation According to previous works conducted under similar conditions (Moreau et al., 2007), a sample size of 16 dogs/treatment group ensured that a difference of 4.2%BW in the primary endpoint (PVF) between groups could be detected assuming 75% power, a SD of 4.5 and a 5% significance threshold. 3. Results 3.1. Animal description No clinically relevant changes were obtained from hematological and biochemical analyses in the entire study cohort. In addition, abnormal buccal mucosal bleeding times were not observed during the study. The numbers of dogs screened, randomly assigned, and analyzed in each group are detailed in Fig. 1. The NHP dog with persistent diarrhea was diagnosed to have gastrointestinal intolerance. Complete CCL rupture (n = 2) and humeral bone inflammation resulted in acute lameness and consequently, to the withdrawal of these dogs. Download full-size image Fig. 1. Flow chart of the study enrolment, allocation, follow-up and analysis. Baseline characteristics of the dogs stratified per group are presented in Table 2. Groups were well balanced according to the outcomes of interest, as significant difference was not observed for the level of PVF, DDAP and CSOM recorded at baseline. It should be noted that in each group, the dogs did not experience significant change in BW over time. Table 2. Baseline characteristics of the dogs stratified per group. Characteristics Groups (n = 16/group) Placebo Natural health product formulations Age (months) 71.1 (22.6) 70.8 (33.5) Sex (male/female) 7/9 10/6 Body weight (kg) 40.7 (8.5) 39.7 (10.8) Peak vertical force (% body weight) 56.5 (6.2) 56.9 (5.3) Daily duration of active period (h/day) 6.7 (1.7) 6.9 (2.4) Case-specific outcome measure of disability 1.6 (0.6) 1.6 (0.6) Osteoarthritis-afflicted joint (most affected limb)  Hip (count) 3 4  Stifle (count) 6 8  Hip and stifle (count) 7 4 3.2. Peak vertical force measurement The PVF generated by the disabled hind limb was increased in the overall study cohort (time effect; p = 0.016), without significant group effect (p = 0.299) (Fig. 2). Increment in PVF was mostly attributed to the changes observed in the NHP-treated dogs. Hence, a significant time × group interaction (p <0.001) was observed which indicates that groups evolved distinctively from baseline to the end of the study. More specifically, analyses revealed that the PVF of NHP-treated dogs (n = 13) was significantly increased at week 4 [58.9 (5.4)%BW, p = 0.020] and at week 8 [59.8 (6.3)%BW, p <0.001], when compared to baseline 57.3 (4.9)%BW. Placebo dogs (n = 14) did not have significantly different values at week 4 [56.4 (5.8)%BW] or week 8 [56.9 (6.8)%BW] than baseline [57.2 (4.5)%BW]. Both groups did not differ significantly at week 8. Fig. 3 presents the respective individual changes in PVF recorded over the study (i.e., week 8 − baseline) as well as the mean change denoted in each group. The mean changes in PVF values were significantly different between groups (p = 0.027). Download full-size image Fig. 2. Peak vertical force. Individual peak vertical force values recorded in dogs having received either natural health product formulations or a negative (placebo) control. Peak vertical force values are expressed as percentage of body weight. The short horizontal lines denote mean group values. For the natural health product formulations group, values at week 4 and week 8 were significantly different (p <0.05) than baseline. Download full-size image Fig. 3. Changes in peak vertical force. Individual changes in peak vertical force recorded in dogs having received either natural health product formulations or a negative (placebo) control over 8 weeks. Changes are the differences between week 8 minus baseline. Negative changes represent a decrease in peak vertical force values at week 8 (i.e., worsening). Dotted lines delineate responders versus non-responders according to the minimal detectable change at 95% confidence interval (Moreau et al., 2013). 3.3. Locomotor activity recording The analysis of DDAP indicated no significant period (p = 0.862), or group (p = 0.414) effect, but a significant period × group interaction (p < 0.001). Analyses revealed that the week 4 period [7.3 (1.9) h/day] in NHP-treated dogs (n = 7) was not significantly different to the baseline, reaching significant increase for the week 8 period [8.2 (3.4) h/day, p = 0.025] (Fig. 4). The DDAP values of placebo dogs (n = 7) at the week 4 [6.7 (2.1) h/day] and week 8 [6.0 (2.3) h/day] periods were not significantly different than the baseline (Fig. 4). A statistical trend (p = 0.064) was observed for a difference in DDAP values between-groups over the study (i.e., week 8 − baseline). Download full-size image Fig. 4. Locomotor activity recording. Temporal evolution of the locomotor activity recorded in dogs having received either natural health product formulations or a negative (placebo) control over 61-day duration. The daily duration of active period is expressed as mean (h/day) with positive (natural health product formulations) or negative (placebo) standard deviation. Periods are baseline (day4 to day 0), week 4 (day 26 to day 30) and week eight (day 52 to day 56) for the statistical analysis. For the natural health product formulations group, values at week 8 were significantly higher than baseline (p <0.05). 3.4. Case-specific outcome measure The CSOM analysis revealed no significant period (p = 0.053), group (p = 0.960) and period × group (p = 0.524) effect. Fig. 5 presents the evolution of the CSOM recorded over the entire study duration. Download full-size image Fig. 5. Case-specific outcome measures of disability. Temporal evolution of the case-specific outcome measures of disability (CSOM) recorded in dogs having received either natural health product formulations or a negative (placebo) control over 8 weeks. Data are expressed as mean with positive (natural health products formulations) or negative (negative control) standard deviation. Periods are baseline (score on day 0), week 4 (scores on days 24, 28 and 31) and week 8 (scores on days 49, 52 and 56) for the statistical analysis. 4. Discussion and conclusions Current therapeutic approaches used to manage OA-afflicted patients remain largely palliative, NSAIDs being the first line of treatment (Bennell et al., 2012). The effect sizes reported for therapeutic modalities range from small to moderate (Bjordal et al., 2004; Zhang et al., 2007). Therefore, there is an opportunity for novel and effective therapeutics to alleviate pain for the OA-afflicted patient. As naturally-occurring models of OA have recently been proposed to accelerate the development of human therapeutics (Pelletier et al., 2010), and since canine OA models have a high translational value to human OA (Moreau et al., 2013), this randomized, double-blind, placebo-controlled trial was undertaken in the canine natural OA model to assess the efficacy of novel phytotherapeutics for human use. A recent systematic review concluded that NHPs had poor therapeutic potential for the treatment of companion animals affected by OA (Vandeweerd et al., 2012). This disappointing conclusion was largely based on the limited number of rigorous RCTs developed to challenge the proposed therapeutic efficacy of NHP. The quality and quantity of current research studies were also criticized for oral herbal medicines purported to alleviate the clinical signs of human OA (Cameron and Chrubasik, 2014). The present trial was undertaken with the second intention to provide rigorous evidences regarding the therapeutic potential of medicinal herb-based NHP formulations to alleviate the clinical signs of canine OA, and to identify the occurrence of adverse effects with multi-NHP preparations. According to the present trial, medicinal herb-based NHP formulations improved the functional ability in dogs afflicted by naturally-occurring OA to a higher degree than placebo-control animals. When given once daily, improvements were noted as early as 4 weeks after the initiation of the alpha formulation administration, and were even better when the beta formulation was given for an additional 4-week duration. It has to be noted that the NHP dosing regimen in this trial was not constant across the entire dog's body mass observed (i.e. alpha formulation 58 (10) mg/kg, beta formulation 76 (13) mg/kg). The manufacturer's limitations in producing capsules with variable content in multi-NHP preparations support the necessary use of dosing by intervals. The study primary endpoint was selected as the PVF measured using a force platform. Such an objective evaluation tool was previously used to measure the disability that characterized human OA patients as well as their response to treatment (Detrembleur et al., 2005; Gok et al., 2002; Messier et al., 1992; Schnitzer et al., 1993). Similarly, alterations from normality were detected in OA dogs based on the measurement of the PVF (Madore et al., 2007) while strong improvements in the pain-related limb disuse were reported for several therapeutic approaches including NSAIDs (Budsberg et al., 1999; Moreau et al., 2003), a dual inhibitor of cyclooxygenase and 5-lipoxygenase enzymes (Moreau et al., 2007), therapeutic diets (Moreau et al., 2012b; Rialland et al., 2013; Roush et al., 2010) and NHPs (Hielm-Bjorkman et al., 2009; Moreau et al., 2004, 2012a). The change over the initial condition [i.e., 2.6 (2.1)%BW] provided by the medicinal herb-based NHP formulations is similar to common therapeutic approaches as recently reviewed (Moreau et al., 2012a). It outweighs the 95% minimal detectable change (MDC95), calculated as 2.0%BW for PVF in canine OA (Moreau et al., 2013). The MDC95 can be interpreted as the change magnitude, below which there are more than 95% chances that the change has occurred as a result of measurement error (Kovacs et al., 2008). Outside this cut-off point (i.e., lower than −2.0 or higher than 2.0 %BW), the change does reflect a real difference in the functional impairment toward worsening or improvement, in the canine natural OA model. Establishing such a cut-off point fulfils the requirement to define the magnitude of the measurement that corresponds to a clinically recognizable improvement in the individual animals, as previously criticized in a recent review (Sharkey, 2013). The MDC95 can also serve as a responder criteria, similar to that developed for humans by the OARSI Standing Committee for Clinical Trials Response Criteria Initiative (Pham et al., 2004). According to Fig. 3, 46% (6/13) of the medicinal herb-based NHPs treated dogs were positive responders while negative responders were absent. At the opposite, 36% (5/14) of placebo-control dogs had more severe clinical signs while 36% (5/14) had improved. In the present study, statistical analyses revealed a significant difference between groups according to the changes in PVF values with a statistical power of 60%. The magnitude of the therapeutic benefits was consistent with a moderate Cohen's d effect size of 0.7 (95% confidence interval: 0.0–1.5). The effect size is recognized as a simple and straightforward index to quantify the effects of an intervention relative to a comparator (Coe, 2012). However, effect sizes are not commonly reported in canine models of OA, which compromise comparisons among studies. Nevertheless, the effect size reported herein was similar to other therapeutic approaches including a therapeutic diet rich in omega-3 fatty acids of fish origin (Moreau et al., 2012b) as well as a plant extract from Brachystemma calycinum D don (Moreau et al., 2012a). As previously demonstrated in this model of natural OA (Brown et al., 2010; Moreau et al., 2012a; Rialland et al., 2012, 2013), the usefulness of the continuous monitoring of daily locomotor activity was sustained in the present study. After an 8-week period of treatment with the NHP formulations, the DDAP was increased, reaching more than 1.5 h/day of additional time spent on daily life activities. This finding is consistent with a recent review of experimental data aimed to determine the relationship between the limb function (as reflected by the measurement of the PVF) and the locomotor activity recording (Moreau et al., 2013). Hence, the effect of an additional 54 minutes/day of activity is expected to be mirrored confidently by an increase in PVF measurement exceeding the MDC95 (Moreau et al., 2013). As reported herein, the effects of the medicinal herb-based NHP formulations might have been translated into more active dogs, being able to rehabilitate their pain-related limb disuse toward a better muscular strength. This increase in limb use led to dogs more willing to accentuate their limb support by an average of 1.0 kg. These findings sustain the beneficial role of activity in OA dogs. Nevertheless, the level of activity has to be low to moderate to avoid an exacerbation of lameness as reported after intense running (Beraud et al., 2010). Unlike the objective measures of function, the CSOM did not document an improvement in NHP-treated dogs. The CSOM is a validated proxy method of assessment, which was shown to complement the information provided by the measurement of the PVF (Rialland et al., 2012, 2013). Hence, the CSOM reflects the behavioral aspects of the OA disease affliction as perceived by the owner based on day-to-day environment and situation. The CSOM was used in the present study as an attempt to mirror the dog's quality of life over the 8 weeks. This was done however without knowing the level of functional improvement required to be translated into a better quality of life. The present results suggest at first glance a need for more effective therapy based on owner perception, recognized as less sensitive and more prone to placebo response bias (Conzemius and Evans, 2012; Moreau et al., 2013) or to changes in behavior or perception when being utilized as a proxy assessor. On the other hand, as OA is a lifelong disease, the limb impairment which occurs over several years may have compromised the sensitivity of the owner to detect an improvement in their dog. This is also supported by the relatively low value of CSOM at baseline, compared to other similar population samples (Rialland et al., 2012, 2013), inducing a risk of floor effect for CSOM masking the responsiveness to NHP treatment. Therefore, much time may be required by the owner to appreciate a better quality of life concomitantly to a functional improvement, as previously denoted in OA dogs after a 13-week treatment duration (Moreau et al., 2012b). Our results indicate that treating with the medicinal herb-based NHPs did not result in a significant buccal mucosal bleeding time prolongation. This indicates that the platelet function was not affected by the treatment. Moreover, the NHP-treated dogs did not demonstrate clinically significant hematological or biochemical alterations when administered for 8 weeks. This result is encouraging for promoting the clinical use of multi-NHP preparations, but would require further confirmation on larger sample size. Several limitations to this clinical trial study need to be acknowledged. First, the study duration was 8 weeks despite the chronic nature of OA. Second, the content and strength of the NHP capsules were based on empirical evidences (intern data files) suggesting anti-inflammatory and anti-nociceptive potential in rodent models of inflammation and pain. Whether or not the content and strength of the NHP capsule were optimal for dogs afflicted by naturally-occurring OA was unknown. Third, the design of the study did not allow conclusions about the respective potential of each NHP formulation (i.e., alpha versus beta formulation). Therefore the efficacy of the medicinal herb-based NHP formulations should be considered as a whole therapeutic regimen involving alpha followed by the beta formulations. Finally, whether or not the improvement denoted in OA dogs is consistent with disease modifying effects is unknown and should be addressed. Of note licofelone, a dual inhibitor of cyclooxygenase and 5-lipoxygenase enzymes, demonstrated similar functional improvement than the one observed with the NHP formulations in addition to a reduction in the progression of structural changes in experimental dog OA model induced by CCL sectioning (Boileau et al., 2002; Moreau et al., 2006). This RCT provided evidence of the efficacy of a medicinal herb-based NHP in alleviating the clinical signs of canine OA. The present findings provide relevant and new information about the potential of medicinal phytochemical compounds as a therapeutic modality for human OA. Such NHP appears also interesting for the management of canine OA as not only clear benefits were demonstrated on the function, but also this NHP mixture (with low grade dosage of each component) was not associated with any clinical toxicity. Acknowledgments Authors would like to acknowledge Ms. Katherine Bernier and Ms. Anne-Andrée Mignault for their technical support. Authors are also grateful to Dr. Christian Bollinger and Dr. Alexander O El-Warrak for their outstanding orthopedic expertise. This study was funded in part by a grant from ArthroLab Inc., an ongoing New Opportunities Fund grant (#9483) and a Leader Opportunity Fund grant (#24601) from the Canada Foundation for Innovation for the pain/function equipment, a Discovery Grant (#327158-2008; #441651-2013) from the Natural Sciences and Engineering Research Council of Canada for the bio-analyses and salaries, and by the Osteoarthritis Chair of the University of Montreal Hospital Centre, Université de Montréal. Maxim Moreau received a doctoral scholarship from the Canadian Institutes of Health Research (TGF-53914) – Strategic Training Initiative in Health Research program (MENTOR) and a doctoral scholarship from the Fonds de recherche du Québec-Santé. References Aggarwal et al, 2013 Aggarwal B.B., Gupta S.C., Sung B. Curcumin: an orally bioavailable blocker of TNF and other pro-inflammatory biomarkers British Journal of Pharmacology, 169 (2013), pp. 1672-1692 CrossRefView Record in Scopus Bennell et al, 2012 Bennell K.L., Hunter D.J., Hinman R.S. Management of osteoarthritis of the knee British Medical Journal, 345 (2012), p. e4934 CrossRef Beraud et al, 2010 Beraud R., Moreau M., Lussier B. Effect of exercise on kinetic gait analysis of dogs afflicted by osteoarthritis Veterinary and Comparative Orthopaedics and Traumatology, 23 (2010), pp. 87-92 View Record in Scopus Bjordal et al, 2004 Bjordal J.M., Ljunggren A.E., Klovning A., Slordal L. Non-steroidal anti-inflammatory drugs, including cyclo-oxygenase-2 inhibitors, in osteoarthritic knee pain: meta-analysis of randomised placebo controlled trials British Medical Journal, 329 (2004), p. 1317 CrossRefView Record in Scopus Boileau et al, 2002 Boileau C., Martel-Pelletier J., Jouzeau J.Y., Netter P., Moldovan F., Laufer S., et al. Licofelone (ML-3000), a dual inhibitor of 5-lipoxygenase and cyclooxygenase, reduces the level of cartilage chondrocyte death in vivo in experimental dog osteoarthritis: inhibition of pro-apoptotic factors Journal of Rheumatology, 29 (2002), pp. 1446-1453 View Record in Scopus Brown et al, 2010 Brown D.C., Boston R.C., Farrar J.T. Use of an activity monitor to detect response to treatment in dogs with osteoarthritis Journal of the American Veterinary Medical Association, 237 (2010), pp. 66-70 CrossRefView Record in Scopus Budsberg et al, 1999 Budsberg S.C., Johnston S.A., Schwarz P.D., DeCamp C.E., Claxton R. Efficacy of etodolac for the treatment of osteoarthritis of the hip joints in dogs Journal of the American Veterinary Medical Association, 214 (1999), pp. 206-210 View Record in Scopus Callan, Giger, 2001 Callan M.B., Giger U. Assessment of a point-of-care instrument for identification of primary hemostatic disorders in dogs American Journal of Veterinary Research, 62 (2001), pp. 652-658 CrossRefView Record in Scopus Cameron, Chrubasik, 2014 Cameron M., Chrubasik S. Oral herbal therapies for treating osteoarthritis The Cochrane Database of Systematic Reviews (5) (2014), 10.1002/14651858.CD002947.pub2 CD002947 Coe, 2012 Coe R. Effect size Arthur J., Waring M., Coe R., Hedges L.V. (Eds.), Research Methods and Methodologies in Education (first ed.), SAGE Publications, London (2012), pp. 368-377 View Record in Scopus Conzemius, Evans, 2012 Conzemius M.G., Evans R.B. Caregiver placebo effect for dogs with lameness from osteoarthritis Journal of the American Veterinary Medical Association, 241 (2012), pp. 1314-1319 CrossRefView Record in Scopus Cook, 2010 Cook J.L. Cranial cruciate ligament disease in dogs: biology versus biomechanics Veterinary Surgery, 39 (2010), pp. 270-277 CrossRefView Record in Scopus Detrembleur et al, 2005 Detrembleur C., De Nayer J., van den Hecke A. Celecoxib improves the efficiency of the locomotor mechanism in patients with knee osteoarthritis. A randomised, placebo, double-blind and cross-over trial Osteoarthritis and Cartilage, 13 (2005), pp. 206-210 ArticleDownload PDFView Record in Scopus Gagnier et al, 2004 Gagnier J.J., Chrubasik S., Manheimer E. Harpgophytum procumbens for osteoarthritis and low back pain: a systematic review BMC Complementary and Alternative Medicine, 4 (2004), p. 13 Gagnier et al, 2007 Gagnier J.J., van Tulder M.W., Berman B., Bombardier C. Herbal medicine for low back pain: a Cochrane review Spine, 32 (2007), pp. 82-92 CrossRefView Record in Scopus Garbacki et al, 2002 Garbacki N., Angenot L., Bassleer C., Damas J., Tits M. Effects of prodelphinidins isolated from Ribes nigrum on chondrocyte metabolism and COX activity Naunyn-Schmiedebergs Archives of Pharmacology, 365 (2002), pp. 434-441 CrossRefView Record in Scopus Gok et al, 2002 Gok H., Ergin S., Yavuzer G. Kinetic and kinematic characteristics of gait in patients with medial knee arthrosis Acta Orthopaedica Scandinavica, 73 (2002), pp. 647-652 CrossRefView Record in Scopus Henrotin et al, 2013 Henrotin Y., Priem F., Mobasheri A. Curcumin: a new paradigm and therapeutic opportunity for the treatment of osteoarthritis: curcumin for osteoarthritis management Springerplus, 2 (2013), p. 56 CrossRefView Record in Scopus Hielm-Bjorkman et al, 2009 Hielm-Bjorkman A., Tulamo R.M., Salonen H., Raekallio M. Evaluating complementary therapies for canine osteoarthritis part I: green-lipped mussel (Perna canaliculus) Evidence-Based Complementary and Alternative Medicine: eCAM, 6 (2009), pp. 365-373 CrossRefView Record in Scopus Jain, Kulkarni, 1999 Jain N.K., Kulkarni S.K. Antinociceptive and anti-inflammatory effects of Tanacetum parthenium L. extract in mice and rats Journal of Ethnopharmacology, 68 (1999), pp. 251-259 ArticleDownload PDFView Record in Scopus Johnston, 2001 Johnston S.A. Overview of pain in the lame patient Veterinary Clinics of North America. Small Animal Practice, 31 (2001), pp. 39-53 ArticleDownload PDFView Record in Scopus Kimmatkar et al, 2003 Kimmatkar N., Thawani V., Hingorani L., Khiyani R. Efficacy and tolerability of Boswellia serrata extract in treatment of osteoarthritis of knee – a randomized double blind placebo controlled trial Phytomedicine: International Journal of Phytotherapy and Phytopharmacology, 10 (2003), pp. 3-7 ArticleDownload PDFCrossRefView Record in Scopus Kovacs et al, 2008 Kovacs F.M., Abraira V., Royuela A., Corcoll J., Alegre L., Tomas M., et al. Minimum detectable and minimal clinically important changes for pain in patients with nonspecific neck pain BMC Musculoskeletal Disorders, 9 (2008), p. 43 View Record in Scopus Kuptniratsaikul et al, 2009 Kuptniratsaikul V., Thanakhumtorn S., Chinswangwatanakul P., Wattanamongkonsil L., Thamlikitkul V. Efficacy and safety of Curcuma domestica extracts in patients with knee osteoarthritis Journal of Alternative and Complementary Medicine, 15 (2009), pp. 891-897 CrossRefView Record in Scopus Lawrence et al, 2008 Lawrence R.C., Felson D.T., Helmick C.G., Arnold L.M., Choi H., Deyo R.A., et al. Estimates of the prevalence of arthritis and other rheumatic conditions in the United States. Part II Arthritis and Rheumatism, 58 (2008), pp. 26-35 CrossRefView Record in Scopus Leventhal et al, 1994 Leventhal L.J., Boyce E.G., Zurier R.B. Treatment of rheumatoid arthritis with blackcurrant seed oil British Journal of Rheumatology, 33 (1994), pp. 847-852 CrossRefView Record in Scopus Littell et al, 2000 Littell R.C., Pendergast J., Natarajan R. Modelling covariance structure in the analysis of repeated measures data Statistics in Medicine, 19 (2000), pp. 1793-1819 CrossRefView Record in Scopus Lotz-Winter, 1990 Lotz-Winter H. On the pharmacology of bromelain: an update with special regard to animal studies on dose-dependent effects Planta Medica, 56 (1990), pp. 249-253 CrossRefView Record in Scopus Madhu et al, 2013 Madhu K., Chanda K., Saji M.J. Safety and efficacy of Curcuma longa extract in the treatment of painful knee osteoarthritis: a randomized placebo-controlled trial Inflammopharmacology, 21 (2013), pp. 129-136 CrossRefView Record in Scopus Madore et al, 2007 Madore E., Huneault L., Moreau M., Dupuis J. Comparison of trot kinetics between dogs with stifle or hip arthrosis Veterinary and Comparative Orthopaedics and Traumatology, 20 (2007), pp. 102-107 View Record in Scopus Madsen, Svalastoga, 1994 Madsen J.S., Svalastoga E. Inclination and anteversion of collum femoris in hip dysplasia and coxarthritis Acta Veterinaria Scandinavica, 35 (1994), pp. 115-119 View Record in Scopus Martinez, 1997 Martinez S.A. Congenital conditions that lead to osteoarthritis in the dog Veterinary Clinics of North America. Small Animal Practice, 27 (1997), pp. 735-758 ArticleDownload PDFView Record in Scopus Martinez, Coronado, 1997 Martinez S.A., Coronado G.S. Acquired conditions that lead to osteoarthritis in the dog Veterinary Clinics of North America. Small Animal Practice, 27 (1997), pp. 759-775 ArticleDownload PDFView Record in Scopus Mathy-Hartert et al, 2009 Mathy-Hartert M., Jacquemond-Collet I., Priem F., Sanchez C., Lambert C., Henrotin Y. Curcumin inhibits pro-inflammatory mediators and metalloproteinase-3 production by chondrocytes Inflammation Research, 58 (2009), pp. 899-908 CrossRefView Record in Scopus McLaughlin, 2001 McLaughlin R.M. Hind limb lameness in the young patient Veterinary Clinics of North America. Small Animal Practice, 31 (2001), pp. 101-123 ArticleDownload PDFView Record in Scopus Messier et al, 1992 Messier S.P., Loeser R.F., Hoover J.L., Semble E.L., Wise C.M. Osteoarthritis of the knee: effects on gait, strength, and flexibility Archives of Physical Medicine and Rehabilitation, 73 (1992), pp. 29-36 View Record in Scopus Mobasheri, 2012 Mobasheri A. Intersection of inflammation and herbal medicine in the treatment of osteoarthritis Current Rheumatology Reports, 14 (2012), pp. 604-616 CrossRefView Record in Scopus Mobasheri et al, 2012 Mobasheri A., Henrotin Y., Biesalski H.K., Shakibaei M. Scientific evidence and rationale for the development of curcumin and resveratrol as nutraceuticals for joint health International Journal of Molecular Sciences, 13 (2012), pp. 4202-4232 CrossRefView Record in Scopus Moreau et al, 2003 Moreau M., Dupuis J., Bonneau N.H., Desnoyers M. Clinical evaluation of a nutraceutical, carprofen and meloxicam for the treatment of dogs with osteoarthritis Veterinary Record, 152 (2003), pp. 323-329 CrossRefView Record in Scopus Moreau et al, 2004 Moreau M., Dupuis J., Bonneau N.H., Lecuyer M. Clinical evaluation of a powder of quality elk velvet antler for the treatment of osteoarthrosis in dogs Canadian Veterinary Journal, 45 (2004), pp. 133-139 View Record in Scopus Moreau et al, 2006 Moreau M., Boileau C., Martel-Pelletier J., Brunet J., Laufer S., Pelletier J.P. Licofelone reduces progression of structural changes in a canine model of osteoarthritis under curative conditions: effect on protease expression and activity Journal of Rheumatology, 33 (2006), pp. 1176-1183 View Record in Scopus Moreau et al, 2007 Moreau M., Lussier B., Doucet M., Vincent G., Martel-Pelletier J., Pelletier J.P. Efficacy of licofelone in dogs with clinical osteoarthritis Veterinary Record, 160 (2007), pp. 584-588 CrossRefView Record in Scopus Moreau et al, 2010 Moreau M., Troncy E., Bichot S., Lussier B. Influence of changes in body weight on peak vertical force in osteoarthritic dogs: a possible bias in study outcome Veterinary Surgery, 39 (2010), pp. 43-47 CrossRefView Record in Scopus Moreau et al, 2011 Moreau M., Rialland P., Pelletier J.P., Martel-Pelletier J., Lajeunesse D., Boileau C., et al. Tiludronate treatment improves structural changes and symptoms of osteoarthritis in the canine anterior cruciate ligament model Arthritis Research & Therapy, 13 (2011), p. R98 CrossRef Moreau et al, 2012a Moreau M., Lussier B., Pelletier J.P., Martel-Pelletier J., Bedard C., Gauvin D., et al. Brachystemma calycinum D. Don effectively reduces the locomotor disability in dogs with naturally occurring osteoarthritis: a randomized placebo-controlled trial Evidence-Based Complementary and Alternative Medicine: eCAM, 2012 (2012), p. 646191 Moreau et al, 2012b Moreau M., Troncy E., Del Castillo J.R., Bedard C., Gauvin D., Lussier B. Effects of feeding a high omega-3 fatty acids diet in dogs with naturally occurring osteoarthritis Journal of Animal Physiology and Animal Nutrition, 97 (2012), pp. 830-837 Moreau et al, 2013 Moreau M., Pelletier J.P., Lussier B., d'Anjou M.A., Blond L., Pelletier J.M., et al. A posteriori comparison of natural and surgical destabilization models of canine osteoarthritis BioMed Research International, 2013 (2013), p. 180453 Pelletier et al, 2010 Pelletier J.P., Boileau C., Altman R.D., Martel-Pelletier J. Animal models of osteoarthritis Hochberg M.C., Silman A.J., Smolen J.S., Weinblatt M.E., Weisman M.H. (Eds.), Rheumatology (fifth ed.), Mosby Elsevier, Philadelphia (2010), pp. 1731-1739 View Record in Scopus Pham et al, 2004 Pham T., van der Heijde D., Altman R.D., Anderson J.J., Bellamy N., Hochberg M., et al. OMERACT-OARSI initiative: osteoarthritis research society international set of responder criteria for osteoarthritis clinical trials revisited Osteoarthritis and Cartilage, 12 (2004), pp. 389-399 ArticleDownload PDFView Record in Scopus Reichling et al, 2004 Reichling J., Schmokel H., Fitzi J., Bucher S., Saller R. Dietary support with Boswellia resin in canine inflammatory joint and spinal disease Schweizer Archiv fur Tierheilkunde, 146 (2004), pp. 71-79 CrossRefView Record in Scopus Rialland et al, 2012 Rialland P., Bichot S., Moreau M., Guillot M., Lussier B., Gauvin D., et al. Clinical validity of outcome pain measures in naturally occurring canine osteoarthritis BMC Veterinary Research, 8 (2012), p. 162 CrossRef Rialland et al, 2013 Rialland P., Bichot S., Lussier B., Moreau M., Beaudry F., del Castillo J.R.E., et al. Effect of a green-lipped mussel-enriched diet on pain behaviours and functioning in dogs with clinical osteoarthritis Canadian Journal of Veterinary Research, 77 (2013), pp. 66-74 View Record in Scopus Roush, 2001 Roush J.K. Hind limb lameness in the mature dog Veterinary Clinics of North America. Small Animal Practice, 31 (2001), pp. 125-141 ArticleDownload PDFView Record in Scopus Roush et al, 2010 Roush J.K., Cross A.R., Renberg W.C., Dodd C.E., Sixby K.A., Fritsch D.A., et al. Evaluation of the effects of dietary supplementation with fish oil omega-3 fatty acids on weight bearing in dogs with osteoarthritis Journal of the American Veterinary Medical Association, 236 (2010), pp. 67-73 CrossRefView Record in Scopus Samuels, 2005 Samuels N. Herbal remedies and anticoagulant therapy Thrombosis and Haemostasis, 93 (2005), pp. 3-7 View Record in Scopus Schnitzer et al, 1993 Schnitzer T.J., Popovich J.M., Andersson G.B., Andriacchi T.P. Effect of piroxicam on gait in patients with osteoarthritis of the knee Arthritis and Rheumatism, 36 (1993), pp. 1207-1213 CrossRefView Record in Scopus Shakibaei et al, 2012 Shakibaei M., Allaway D., Nebrich S., Mobasheri A. Botanical extracts from rosehip (Rosa canina), willow bark (Salix alba), and nettle leaf (Urtica dioica) suppress IL-1beta-induced NF-kappaB activation in canine articular chondrocytes Evidence-Based Complementary and Alternative Medicine: eCAM, 2012 (2012), p. 509383 Sharkey, 2013 Sharkey M. The challenges of assessing osteoarthritis and postoperative pain in dogs The American Association of Pharmaceutical Scientists Journal, 15 (2013), pp. 598-607 CrossRefView Record in Scopus Shearer, 2011 Shearer P. Epidemiology of orthopedic disease Veterinary Focus, 21 (2011), pp. 24-25 CrossRefView Record in Scopus Vandeweerd et al, 2012 Vandeweerd J.M., Coisnon C., Clegg P., Cambier C., Pierson A., Hontoir F., et al. Systematic review of efficacy of nutraceuticals to alleviate clinical signs of osteoarthritis Journal of Veterinary Internal Medicine, 26 (2012), pp. 448-456 CrossRefView Record in Scopus Zhang et al, 2007 Zhang W., Moskowitz R.W., Nuki G., Abramson S., Altman R.D., Arden N., et al. OARSI recommendations for the management of hip and knee osteoarthritis, part I: critical appraisal of existing treatment guidelines and systematic review of current research evidence Osteoarthritis and Cartilage, 15 (2007), pp. 981-1000 ArticleDownload PDFView Record in Scopus Address of the institution where the work was carried out: This study was conducted at the Research Group in Animal Pharmacology of Quebec (GREPAQ), Department of Veterinary Biomedical Sciences – Faculty of Veterinary Medicine, Université de Montréal, P.O. Box 5000, Saint-Hyacinthe, Quebec J2S 7C6, Canada, in collaboration with the Osteoarthritis Research Unit, University of Montreal Hospital Research Centre (CRCHUM), Tour Viger, 900 St-Denis Street, Montreal, Quebec H2X 0A9, Canada. Copyright © 2014 Elsevier Ltd. All rights reserved.

No comments:

Post a Comment