Is there any scientific evidence for the use of glucosamine in the management of human osteoarthritis?

Glucosamine in its acetylated form is a natural constituent of some glycosaminoglycans (for example, hyaluronic acid and keratan sulfate) in the proteoglycans found in articular cartilage, intervertebral disc and synovial fluid. Glucosamine can be extracted and stabilized by chemical modification and used as a drug or a nutraceutical. It has been approved for the treatment of osteoarthritis (OA) in Europe to promote cartilage and joint health and is sold over the counter as a dietary supplement in the United States. Various formulations of glucosamine have been tested, including glucosamine sulfate and glucosamine hydrochloride. In vitro and in vivo studies have uncovered glucosamine's mechanisms of action on articular tissues (cartilage, synovial membrane and subchondral bone) and justified its efficacy by demonstrating structure-modifying and anti-inflammatory effects at high concentrations. However, results from clinical trials have raised many concerns. Pharmacokinetic studies have shown that glucosamine is easily absorbed, but the current treatment doses (for example, 1,500 mg/day) barely reach the required therapeutic concentration in plasma and tissue. The symptomatic effect size of glucosamine varies greatly depending on the formulation used and the quality of clinical trials. Importantly, the effect size reduces when evidence is accumulated chronologically and evidence for the structure-modifying effects of glucosamine are sparse. Hence, glucosamine was at first recommended by EULAR and OARSI for the management of knee pain and structure improvement in OA patients, but not in the most recent NICE guidelines. Consequently, the published recommendations for the management of OA require revision. Glucosamine is generally safe and although there are concerns about potential allergic and salt-related side effects of some formulations, no major adverse events have been reported so far. This paper examines all the in vitro and in vivo evidence for the mechanism of action of glucosamine as well as reviews the results of clinical trials. The pharmacokinetics, side effects and differences observed with different formulations of glucosamine and combination therapies are also considered. Finally, the importance of study design and criteria of evaluation are highlighted as new compounds represent new interesting options for the management of OA.

studies have detailed diff erent lines of evidence for how glucosamine can act on joint tissues from OA patients. In addition, many clinical trials have demonstrated various degrees of effi cacy for glucosamine in OA patients. Based on the published data, the Osteoarthritis Research Society International (OARSI) [2,3] and the European League Against Rheumatism (EULAR) [4,5] have recommended the use of glucosamine sulfate for the management of knee and hip OA. In contrast, the American College of Rheumatology (ACR) [6] and the UK National Institute for Health and Clinical Excellence (NICE) have not recommended glucosamine in the management of OA ( Table 1). None of the current guidelines have recommended the use of glucosamine hydrochloride, only glucosamine sulfate. Finally, it is important to point out that OARSI recommends that treatment with glusosamine sulfate is discontinued if no symptomatic response is apparent within 6 months of use. OA recommendations are outlined in Table 1. More recently, a change in evidence following a systematic cumulative update of research published through January 2009 has been reported by OARSI's treatment guidelines committee [7]. Th is meta-analysis reported a gradual decrease in the eff ect size (ES) when evidence from randomized controlled trials (RCTs) was chronologically evaluated. Th e study highlighted the controversy surrounding the effi cacy of glucosamine in terms of both pain and structure modifi cations. Th e meta-analysis has highlighted the heterogeneity of outcomes in the diff erent RCTs and the presence of publication bias. From a scientifi c perspective, the new concerns raised by the recent meta-analyses will undoubtedly stimulate a re-evaluation of the mechanistic eff ects of glucosamine.
Th e use of glucosamine in the management of OA remains controversial and its specifi c mechanism of action in OA pain and function modifi cation are still unclear. Th e objective of this review is to address the question raised in the title: is glucosamine still an option in the management of OA? Th is review summarizes the eff ects of glucosamine in OA based on in vitro and in vivo results as well as recent clinical trials. Special attention is given to the pharmacokinetics of glucosamine, its side eff ects and the diff erences observed with diff erent formulations and combination therapies. Finally, based on these observations, a conclusion is drawn on the role of glucosamine in the management of OA in the context of new compounds and new combinations.

Update on in vitro data
Th e anabolic eff ects of glucosamine were primarily thought to be attributable to its capacity for providing building blocks for the synthesis of GAGs by chondro cytes [8][9][10]. Other eff ects have also been reported, how ever, such as its anti-catabolic potency, seen by its inhibition of the expression and/or activity of catabolic enzymes such as phospholipase A 2 , matrix metallo protein ases or aggrecanases [11,12]. Another study con fi rmed the potency of glucosamine to inhibit the expres sion and activity of aggrecanase-2 (a disintegrin and metalloproteinase with thrombospondin motifs 5 (ADAMTS-5)) in transiently transfected cell lines [13]. Th e authors suggested that the symptomatic and functional eff ects of glucosamine would be justifi ed by the fact that glucosamine interferes with the matrix metalloproteinases responsible for proteo glycan degradation in OA.
Various properties were demonstrated for glucosamine in the three main tissues involved in OA, cartilage, synovial membrane and subchondral bone. Th e main potencies were shown in articular chondrocytes, where glucosamine was demonstrated to reverse the deleterious eff ects of IL-1β [11,14,15]. In rat and human chondrocytes, the glucosamine eff ect was shown to occur through the inhibition of NF-κB signaling [16,17]. Th e reversal of the eff ects of IL-1β in human chondrocytes also inhibited  [2,3,7] 'Treatment with glucosamine and/or chondroitin sulphate may provide symptomatic benefi t in patients with knee OA. If no response is apparent within 6 months treatment should be discontinued' EULAR [5] Knee OA 'There is growing evidence to support the use of two of these agents for their symptomatic eff ects -namely, glucosamine sulphate (1A) and chondroitin sulphate (1A), but for the others the evidence is weak or absent' [4] Hip OA 'SYSADOA (glucosamine sulphate, chondroitin sulphate, diacerhein, avocado soybean unsaponifi able, and hyaluronic acid) have a symptomatic eff ect and low toxicity, but eff ect sizes are small, suitable patients are not well defi ned, and clinically relevant structure modifi cation and pharmacoeconomic aspects are not well established' ACR [6] 'While a number of studies support the effi cacy of both glucosamine and chondroitin sulfate for palliation of joint pain in patients with knee OA, the subcommittee believes that it is premature to make specifi c recommendations about their use at this time because of methodologic considerations, including lack of standardized case defi nitions and standardized outcome assessments, as well as insuffi cient information about study design in a number of these published reports' NICE [81] 'The use of glucosamine or chondroitin products is not recommended for the treatment of osteoarthritis' infl ammatory enzymes, such as the inducible form of nitric oxide synthase and cyclooxygenase-2 [18]. Th is eff ect of glucosamine was further detailed by Imagawa and co-workers [19], who were able to demonstrate that glucosamine prevents the demethylation of specifi c CpG sites in the IL-1β promoter, consequently preventing the expression of IL-1β. A recent pharmacoproteomic study revealed that glucos amine (10 mM) diff erentially regulates the pattern of expression of IL-1β-induced proteins in human articular chondrocytes [20]. Th e proteins aff ected by glucosamine are mainly involved in the signal transduction pathways, redox and stress response, protein synthesis and protein folding. In addition, glucosamine increased the expression of the GRP78 chaperone protein. Th is observation supports the reported antiinfl am matory eff ect of glucosamine. Th is study also compared glucosamine to chondroitin sulfate; these compounds produced diff erent patterns of protein modifi cation when tested alone or in combination [20]. A synergistic eff ect for the modifi cation of superoxide dismutase expression was demonstrated when cells were exposed to both compounds, implying a potent eff ect on oxidative stress in addition to the modulation of energy production and metabolic pathways produced by chondroitin sulfate.
Th e pro-anabolic eff ects of glucosamine were demonstrated in both human chondrocytes and synovial cells, where glucosamine was shown to induce the production of hyaluronic acid (HA) and to directly enter the GAG biosynthetic pathway (that is, for the production of HA, keratan sulfate and sulfated GAGs) [21]. Th e authors also proposed that the chondroprotective eff ect of glucosamine results from the modulation of enzymes responsible for HA synthesis.
Th e potential of glucosamine to induce HA production in the synovial membrane was previously suggested in a study that used synovial explants [22]. In addition, cationic glucosamine derivatives were shown to produce an anti-infl ammatory eff ect through the inhibition of mitogen-activated protein kinase signaling pathways in lipopolysaccharide-stimulated macrophages [23].
Glucosamine sulfate was also shown to be eff ective in human OA osteoblasts [24]. Glucosamine increased the osteo protegrin/receptor activator of nuclear factor kappa-B ligand (RANKL) ratio and reduced bone resorption. Th is eff ect was increased when glucosamine was used in combination with chondroitin sulfate.
It is important to point out that these studies were performed in diff erent culture systems, with various formulations and concentrations of glucosamine. Th e details of the culture systems and the formulations are provided in Table 2. Th e concentrations used in the in vitro studies were 'super physiological' -in some cases up to 2,000 times higher than the maximal concentration that can realistically be achieved in plasma (10 μM) after oral administration of 1,500 mg of glucosamine sulfate in human subjects. Furthermore, some of these studies compared the eff ects of two formulations of glucosamine in order to provide evidence for the superiority of one or another [21,22,25,26]. It was proposed that the diff erences, if they truly existed, might contribute to the diff erent results observed in clinical trials with various glucosa mine formulations. For example, glucosamine sulfate was shown to be a stronger inhibitor of gene expres sion than glucosamine hydrochloride [25]. Both formulations were commercially available from a lab supplier. Th e same group compared glucosamine hydrochloride to N-acetylglucosamine [22] and observed no eff ect of N-acetylglucosamine on HA production whereas glucosamine hydrochloride appeared to modulate this parameter. Th e same conclusion was reached in the study by Igarashi and colleagues [21]. In contrast, another study compared the eff ect of native glucosamine and Nacetylglucosamine on the metabolic activity of human articular chondrocytes [26]. Both compounds exhibited diff erent potencies on glucose transport, and GAG and HA synthesis. Indeed, N-acetylglucosamine appeared to accelerate the facilitated glucose uptake and increase both GAG and HA synthesis, suggesting that N-acetylglucosamine may be more effi cient than native glucosamine.
Glucosamine has also been tested in combination with chondroitin sulfate in vitro. Some of the results have already been discussed above [20,24]. Th e eff ect of combinations of glucosamine and chondroitin was also reported by Chan and colleagues [27][28][29][30]. Glucosamine hydrochloride was tested in combination with chondroitin sulfate on bovine cartilage explants. Th e combination was shown to inhibit both infl ammatory and catabolic intermediates and was slightly superior to glucosamine alone [27,28].
Finally, some authors have studied the contribution of exogenous glucosamine to the synthesis of chondroitin sulfate in human articular chondrocytes in culture [31]. Th ey concluded that exogenous glucosamine cannot stimulate the synthesis of chondroitin sulfate. Furthermore, they showed that glucose can increase endogenous glucosamine to reach concentrations superior to those achieved after oral administration.
In conclusion, many of the in vitro investigations carried out so far have been performed using high concentrations of glucosamine -concentrations that are unlikely to be achieved in plasma with oral doses of the drug. Several authors have proposed that the therapeutic doses used did not allow the identifi cation of proteoglycan synthesis as a mechanism of action of glucosamine [31,32]. Th erefore, extrapolation of the in vitro data to the in vivo situation should be done with great caution.

Update on in vivo data
Several recent in vivo studies using diff erent animal models have demonstrated varying potencies for glucosamine in OA. Glucosamine sulfate (200 and 400 mg/kg/day) was tested in the STR/ort mouse model of spon taneous OA [33]. Glucosamine was shown to delay the progression and severity of OA cartilage lesions. Another study reported the eff ect of glucosamine sulfate (250 mg/kg/ day) administered to rats after anterior cruciate ligament transection (ACLT) [34]. Th e glucosamine treatment group showed a lower level of cartilage degradation and synovial infl ammation compared to the control group. Glucosamine was also shown to modify nociception in OA rats. Mechanical allodynia and weight-bearing distribution were signifi cantly improved by the treatment with glucosamine. In addition, the authors showed that glucosamine aff ected mitogen-activated protein kinase signaling in articular chondro cytes by inhibiting p38 and JNK expression while increas ing Erk1/2 expression.
Th e eff ect of glucosamine on cartilage degradation, synovial infl ammation and bone resorption was recently tested in a model of collagenase-induced OA [35]. Th is study also compared the effi cacies of glucosamine hydrochloride and glucosamine sulfate. Th e authors showed a better effi cacy for glucosamine hydrochloride but did not provide any explanation for this diff erence. Indeed, glucosamine sulfate demonstrated no eff ect on the histological score and the formation of osteophytes whereas glucosamine hydrochloride produced a significant reduction of these parameters; glucosamine sulfate was not further evaluated. However, the results showed that glucosamine hydrochloride administered orally to OA mice (20 mg/kg/day) inhibited not only the loss of GAGs and proteoglycans in articular cartilage but also

Glucosamine hydrochloride
Transfected cell lines 10 mM Inhibition of ADAMTS-5 expression and activity [13] Human OA synovium explants 0.5-5 mM Induction of HA production [22] No eff ect of N-acetylglucosamine bone resorption by the inhibition of RANKL through downregulation of bone morphogenic protein-2, transform ing growth factor-β3 and pSMAD2 and the upregulation of Dickkopf-1 protein levels in the joint. In the same study, glucosamine hydrochloride also reduced osteophyte formation and was shown to inhibit the production of the pro-infl ammatory cytokine IL-6 and to upregulate the production of the anti-infl ammatory cytokine IL-10 by the synovial membrane. Glucosamine hydrochloride was also tested in the rat ACLT model of OA [36]. Glucosamine administered at a dose of 1,000 mg/kg/day produced a chondroprotective eff ect and reduced the serum level of the collagen degradation biomarker CTX-II. It was also tested on equine synovial infl ammation induced by lipopolysaccharide injection [37]. Glucosamine hydrochloride was administered by nasogastric gavage at a dose of 20 mg/kg. Th is study revealed that glucosamine levels were increased in the synovium during infl ammation in comparison to the healthy joint. Th e authors did not conclude if glucosamine produced a therapeutic eff ect or not. Th e same group has previously published results comparing the pharmacokinetics of glucosamine sulfate to glucosamine hydrochloride in horses [38]. Th ey measured higher concentrations of glucosamine in synovial fl uid after glucosamine sulfate administration than after glucosamine hydrochloride administration and concluded that glucosamine sulfate is better absorbed by the horse. Glucosamine hydrochloride was also tested in OA dogs in combination with chondroitin sulfate [39]. Th e combination of the two compounds reduced pain and improved weight bearing as well as reduced disease progression.

D(+)-Glucosamine
Glucosamine in the form of N-acetylglucosamine was tested in rabbits undergoing the ACLT model of OA [40]. N-acetylglucosamine was administered once a week intraarticularly at a dose of 150 mg/injection. Th is study suggested that glucosamine may preserve the integrity of articular cartilage but no statistical evidence was provided to support this suggestion.
Finally, glucosamine sulfate (500 mg/kg) and a combination of it with chondroitin sulfate (400 mg/kg) were compared in the rat ACLT model of OA [41]. Th e results showed the superiority of the combination over glucosamine alone in the prevention of biochemical and histological cartilage modifi cations that occur in the rat OA model.
Glucosamine has produced various eff ects in diff erent in vivo models. It has been tested under diff erent conditions with diff erent formulations and dosages, tending to produce a protective eff ect on articular tissues. At this stage it is impossible to conclude whether one formulation is superior to another. Th erefore, further studies are clearly warranted.

Pharmacokinetic studies
Th e pharmacokinetics of exogenous glucosamine has been diffi cult to study since glucosamine is naturally present in several biological fl uids. Th e origin of endogenous glucosamine, if any, is unknown, although it could be speculated that endogenous glucosamine mainly results from connective tissue metabolism, since glucosa mine is contained in tissues that are not consumed (for example, shellfi sh exoskeleton). Analytical methods have been developed in order to measure glucosamine in human plasma [42][43][44] and synovial fl uid [45]. Most of the available data on the pharmacokinetics of glucosa mine have been obtained with glucosamine sulfate; few studies have been published on the pharmacokinetics of glucosamine hydrochloride in human subjects.
Th e fi rst pharmacokinetic study of glucosamine sulfate was done in 12 healthy volunteers (6 females and 6 males) after repeated oral doses [46]. Glucosamine was rapidly absorbed and available in the systemic circulation. It reached steady state after 3 hours and the maximal plasma concentration was 10 μM after oral intake of a standard dose of 1,500 mg. Th e plasma concentration was shown to remain above baseline for up to 48 hours after oral administration. Th e elimination half-life was estimated at 15 hours, on the basis of a single oral daily dose. Pharmacokinetics were further investigated in 12 OA patients with the same regimen of administration [47]. Th is study revealed that glucosamine was bio available both in plasma and at the site of action, that is, within the joint. Glucosamine reached constant and higher levels (up to 25% higher) in the synovial fl uid compared to plasma. Selective accumulation was also observed in cartilage after repeated administration of glucosamine [48]. Glucosamine was found not to bind to plasma proteins and up to 43 to 47% was associated with blood cells [49]. Th e authors concluded that the pharma cologically active fraction of glucosamine was the same in plasma and at the site of action.
Some information regarding the pharmacokinetics of glucosamine hydrochloride has been recently published [50]. Th e main purpose of this work was to study the infl uence of glucosamine on the pharmacokinetics of chondroitin sulfate, although this study did provide a few pharmacokinetic parameters for glucosamine hydrochloride. Th e authors showed that a single oral administration of 1,500 mg led to a maximal plasma concentration of 492 ± 161 ng/ml (2.28 μM; C max ) reached in 2.31 ± 1.19 h (T max ). After 3 months of multiple dosing, C max was reduced to 211 ± 93.1 ng/ml (0.98 μM) and T max was identical at 2.31 ± 1.19 h. In addition, glucosamine hydrochloride in combination with chondroitin sulfate showed a lower plasmatic level, meaning that chondroitin sulfate inhibits glucosamine hydrochloride absorption and decreases its biodisponibility [50].
Glucosamine hydrochloride and sulfate are identical from a chemical and structural point of view. Th e addition of salt does not justify the diff erence in effi cacy or biological eff ects observed in diff erent studies. Indeed, both glucosamine sulfate and hydrochloride dissociate in the acidic milieu of the stomach, resulting in the release of glucosamine itself [51]. It is important to bear in mind that the pharmacokinetic parameters were determined in two diff erent systems (administration and analytical methods), which could explain the observed diff erences. To date, no published study has directly compared the two formulations and/or established a rigorous methodology for comparison.

Clinical effi cacy
Glucosamine is classically administered orally to human subjects at a dose of 1,500 mg/day. In clinical trials involving OA patients it was shown to reduce pain and provide functional improvement [52][53][54] in addition to structure-modifying eff ects [55,56]. Several metaanalyses have summarized the effi cacy of glucosamine as reported from diff erent clinical trials [52,54,[57][58][59]. Th e most recent study from Wandel and colleagues [60] analyzed the results from randomized trials with an average of at least 100 patients with hip or knee OA. Th e authors searched databases through to June 2010, including the most recent trials [61][62][63][64]. Th ey retrieved a total of 10 trials (a total of 3,803 patients) that met their inclusion criteria. Th eir analysis reported ESs of -0.17 (-0.28 to -0.05) for joint pain and -0.16 (-0.25 to 0.00) for joint space width with glucosamine. Th ey concluded that glucosamine produced no clinically relevant eff ect on pain or structure. Another recent meta-analysis concentrated on the structure-modifying eff ect of glucosamine [65]. Th e authors searched databases through 2008. Th ey retrieved two trials meeting their criteria with glucosamine and structural modifi cations in OA patients. Th eir analysis of joint space narrowing and odds ratios revealed that glucosamine sulfate was not eff ective at producing structural modifi cations after one year of treatment, but glucosamine sulfate produced a small to moderate protective eff ect on minimum joint space narrowing after three years. It is important to note that this analysis was performed using just two trials.
Th e systematic review that is the basis for OARSI recommendations for the management of hip and knee OA was recently updated using more recent publications on glucosamine (until January 2009) [7] and highlighted many interesting points. Th e analysis gathered the results from 19 RCTs from a total of 20 retrieved placebocontrolled trials, including the recent ones already mentioned above. Sixteen of them used glucosamine sulfate preparations (thirteen oral, two intra-muscular and one intra-articular) and three of them used glucosamine hydro chloride. Th is review revealed a decrease of ES for pain evaluated without discrimination between glucosamine preparations from 0.61 (0.28 to 0.95) in 2008 [3] to 0.46 (0.23 to 0.69) in 2010, corresponding to a moderate symptomatic effi cacy [7]. Th e ES for pain reduction was 0.58 (0.30 to 0.87) for glucosamine sulfate, whereas it was not signifi cant for glucosamine hydrochloride (-0.02 (-0.15 to 0.11)). However, the analysis revealed that the outcomes of trials conducted with glucosamine hydrochloride were homogenous, while those conducted with glucosamine sulfate were very heterogeneous. When con sider ing only high quality trials, the ES for glucosamine sulfate decreased to 0.29 (0.003 to 0.57), with no publication bias but still heterogeneity outcomes of trials. Th e same analysis revealed a small but signifi cant ES (0.24 (0.04 to 0.43)) for glucosamine sulfate for reduction of joint space loss in the medial compartment of knee OA patients, as reported by two sponsored RCTs [55,56], but a non-signifi cant eff ect on joint space narrowing in hip OA patients after a 24-month treatment either when considering the whole study group [66] or when analyzing predefi ned subgroups for OA severity [64]. Th e same conclusion was reached by the Glucosamine/Chondroitin Arthritis Intervention Trial (GAIT), with a ES of 0.15 (-0.07 to 0.38) [61]. In addition, this trial did not validate the conclusion of a study by Bruyere and colleagues [67] that glucosamine sulfate (1,500 mg/day) taken for at least 12 months reduces by half the incidence of total knee replacement, since the decision for surgical intervention is infl uenced by many factors. Th e structure-modifying eff ect of glucosamine on OA joints remains controversial.
Two clinical trials have been published since the publications of these meta-analyses. Th e results of the fi rst one by Sawitzke and colleagues [68] were in line with the previously mentioned analysis and reported no signifi cant diff erence in the Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC) pain or function score with glucosamine compared to placebo. Th e second one by Petersen and colleagues [69] studied a diff erent outcome: the serum level of cartilage oligomeric matrix protein, a marker of cartilage degradation. Th is studied showed that combined with exercise, glucosamine sulfate treatment reduced the level of serum cartilage oligomeric matrix protein.
Meta-analyses have revealed the heterogeneity of outcome measures and the presence of publication bias [2,3,7]. Th e most recent analysis by Wandel and colleagues [60] confi rms this. Th e analysis reported that the estimated diff erences between placebo and studied supplements were less pronounced in industry-independent trials compared to industry-sponsored trials. However, the retrieved trials in this analysis demonstrated less heterogeneity. Th e quality of clinical trials, including study design, number of patients, outcome measures and publication bias, justifi ed the drastic reduction of the ES observed for glucosamine over the years and trials.
Finally, the weak eff ect of glucosamine reported by clinical trials can be attributed to the low concentrations of glucosamine available to joint tissues [51]. Indeed, the competition that exists between glucosamine and glucose for cellular uptake favors transport of glucosamine in the intestine, liver and kidney, leaving little glucosamine available for uptake by the joint.

Side eff ects
Glucosamine is considered to be safe, and no serious or fatal adverse events have ever been reported from RCTs. However, several potential side eff ects have to be kept in mind. Th e fi rst and most evident side eff ect of glucosamine that patients should be aware of is shellfi sh allergy. Indeed, glucosamine is extracted from chitin contained in shellfi sh and could lead to allergic reactions in certain individuals. In addition, glucosamine sulfate is administered as a salt, combined with NaCl. Th is formulation provides up to 30% of the daily intake of salt. Th is matter has to be taken into account since Na + but also Clcould infl uence blood pressure and renal function in patients [70,71]. Finally, some human and animal studies have suggested that glucosamine can aff ect glucose metabolism and it has been shown to induce insulin resistance [72][73][74][75]. Despite the evidence, however, a recent comprehensive review supported by a glucosamine manufacturer has rejected these statements. Th e authors concluded that glucosamine had no eff ect on fasting blood glucose levels, glucose metabolism, or insulin sensitivity at any oral dose level in healthy subjects, individuals with diabetes, or those with impaired glucose metabolism [76]. Th e lack of reported adverse events during clinical trials could be attributable to the short duration of patient exposure to the treatment. Th erefore, long-term treatment could be risky for diabetic and/or hypertensive patients. For these reasons, many authors recommend to monitor blood glucose concentration and blood pressure in these patients [73][74][75]77]. Clearly, further studies are warranted to determine the longer term eff ects of glucosamine on insulin sensitivity in human subjects.

Conclusion
Diff erent formulations of glucosamine tested in various in vitro or in vivo systems have exhibited many potential mechanisms of action on articular structures. Furthermore, two of the major published guidelines recommended glucosamine sulfate in the treatment of OA pain while another integrating more recent data did not consider glucosamine sulfate. At this time, glucosamine hydrochloride cannot be recommended based on the available clinical data. However, there are no clear indications that the eff ects of the two formulations can be distinguished from each other in terms of biological activity, posology or pharmacokinetics. Recent metaanalyses have reminded us of the importance of the quality of the clinical trials and this matter is more and more frequently addressed in the literature [51]. Th e evaluation tools and the design of clinical trials should be standardized and there are many ongoing eff orts in this area [78].
Finally, it is important to note that the majority of the published clinical trials with glucosamine reported a signifi cant ratio of subjects who failed to respond to treatment. Th erefore, the question of the benefi t of glucosa mine treatment remains largely unanswered. One may wonder about the clinical relevance of this treatment and whether to use it with regards to the cost/benefi t ratio. On the one hand, compared to non-pharmacological modalities such as exercise, weight loss or education, glucosamine is not eff ective with regards to pain and function, but the question of its cost compared to nonpharmacological modalities should be explored. On the other hand, glucosamine sulfate shows an ES superior to (or at least equal to) the commonly used analgesic or nonsteroidal anti-infl ammatory drugs, but has no rare or adverse eff ects.
Th erefore, based on the fact that glucosamine has low and rare adverse eff ects, it represents a viable option for the management of OA (as a symptomatic slow acting drug) but its administration should be discontinued if no signifi cant eff ect is reported by the patient. New tools, such as novel biomarkers, are required to discriminate responders and non-responders. Th e use of glucosamine associated with the monitoring of response using an appropriate biomarker could improve the effi ciency and the cost/benefi t ratio of the treatment. Finally, we should consider the use of glucosamine as a combination therapy with other drugs or other nutraceuticals, such as omega-3 fatty acid or manganese ascorbate [79,80]. Th is would open up a new horizon in this fi eld but the combinations would need to be rigorously assessed.