Chondrotoxicity of Intra-Articular Injection Treatment: A Scoping Review

The purpose of this scoping review was to identify possible chondrotoxic effects caused by drugs usually used for intra-articular injections. PubMed, Scopus, Web of Science and Cochrane were searched. Inclusion criteria required randomized controlled trials written in English that evaluate the toxic effect that damages the cartilage. The literature search resulted in 185 unique articles. 133 full-text articles were screened for inclusion, of which 65 were included. Corticosteroids, with the exception of triamcinolone, along with local anaesthetics, potentially excluding ropivacaine and liposomal bupivacaine, and nonsteroidal anti-inflammatory drugs, exhibited insufficient safety profiles to warrant casual use in clinical settings. Hyaluronic acid, on the other hand, appears to demonstrate safety while also mitigating risks associated with concurrent compounds, thereby facilitating therapeutic combinations. Additionally, there remains a paucity of data regarding platelet-rich plasma, necessitating further evaluation of its potential efficacy and safety. Overall, it seems that results are significantly influenced by the dosage and frequency of injections administered, observed in both human and animal studies.


Introduction
Several studies have highlighted the extensive use of intra-articular (IA) injections in treating various forms of arthritis.This includes a specific focus on osteoarthritis (OA), as well as other inflammatory and rheumatologic conditions, such as rheumatoid arthritis and psoriatic arthritis [1].For example, OA is a degenerative disease of the cartilage, associated with important changes in cartilage metabolism.If an imbalance between degradation and synthesis by chondrocytes occurs, OA may appear.In fact, inactivity has been shown to lead to the degradation of cartilage and joint motion has a high significance in maintaining cartilage physiological properties [2].IA injection treatments are not only used to manage diseases in the acute phase but even as chronic therapy, leading to the chance of reducing long-term consequences [3].The most used medications for IA injections are corticosteroids, local anaesthetics, hyaluronic acids (HA), platelet-rich plasma (PRP), nonsteroidal anti-inflammatory drugs (NSAIDs), collagen medical devices and bisphosphonates [4,5].Like all substances used as drugs, these compounds can also have side effects.This is particularly so for the joint environment, as it is particularly delicate.
In this regard, articular cartilage is a balanced tissue of 2-4 mm-thick hyaline cartilage with no nerves and blood or lymphatics vessels, comprising a complex extracellular matrix (ECM) and specialized cellular elements.This sophisticated composition confers upon it the capacity to endure substantial compressive forces while enabling smooth, frictionless articulation of the joints.
In this context, ECM is the main structure of articular cartilage.Water, collagen, proteoglycans, other non-collagenous proteins and glycoproteins are the main ECM constituents [6,7].The amount of water present within ECM is critical to maintain its unique mechanical properties and is determined by the aforementioned structural components [8].The ideal quota of tissue fluid should be between 65% and 80% of the total weight [9].Considering the dry weight of cartilage, collagen alone accounts for about 60% of the total, making it the most represented molecule.Fibers and fibrils of type II collagen (90%) are stabilized by other less frequently appearing collagen types and are associated with proteoglycan [10].Another 10-15% of the dry weight is constituted by proteoglycans, heavily glycosylated protein monomers.The largest and most copious is called aggrecan; it has a peculiar ability to interact with HA to form large aggregates and provide cartilage with osmotic properties, needed to resist compressive loads [11].The remaining dry weight is made of other proteins and glycoproteins (e.g., fibronectin) which may play a role in the organization and maintenance of the structure of the ECM [12].
The resident population of cells is the chondrocytes.Chondrocytes, highly specialized and metabolically active cells, have a distinctive function in the formation, upkeep, and restoration of the ECM.They derive from mesenchymal stem cells and make up approximately 2% of the overall volume of articular cartilage [13].Chondrocytes create specialized microenvironments and regulate the turnover of the ECM in their vicinity.Chondrocytes are trapped within their microenvironment's matrix, hindering migration and direct cellto-cell contacts.However, they can respond to various stimuli (e.g., growth factors and mechanical forces).Chondrocytes have limited replication capacity, meaning that, consequently, cartilage has low healing potential after damage of any kind, and cell survival is dependent on optimal chemical and mechanical conditions [7].
The role of chondrocytes in ECM and articular cartilage is essential.They can synthesize ECM components and several enzymes responsible for its remodelling.As vascularization is lacking in articular cartilage, and thus oxygen and nutrients are scarce, chondrocytes have a primarily anaerobic metabolism.Their metabolic activity can be altered by a variety of factors, such as cytokines, growth factors, regulatory peptides, biomechanical forces, joint motion and load [14].
Considering all of this, it is clear that articular cartilage is a particularly delicate tissue and one that is very difficult to repair.The use of drugs with healing purpose must therefore be free of additional risks of tissue compromise.It is therefore of paramount importance to understand which drug class poses the least threat in this regard.
Different compounds are approved to be used for IA injections; however, it is still not clear if and how much those compounds are associated with chondrotoxicity.In this paper we aimed to review studies published regarding the possible chondrotoxic effects caused by drugs usually used for IA injections.

Materials and Methods
An extensive literature search was performed using the following MeSH terms on PubMed, Scopus and Web of Science: "chondrotoxicity", "intraarticular injection", "corticosteroids", "steroids", "hyaluronate or hyaluronic acid", "non-steroidal anti-inflammatory drug", "anaesthetic", "platelet rich plasma", "collagen medical devices" and "bisphosphonates".The search strategy was the following: "chondrotoxicity" OR ("chondrotoxicity" AND "intraarticular injection") OR ("chondrotoxicity" AND "corticosteroids") OR ("chondrotoxicity" AND "steroids") OR ("chondrotoxicity" AND "hyaluronate OR hyaluronic acid") OR ("chondrotoxicity" AND "non-steroidal anti-inflammatory drug") OR ("chondrotoxicity" AND "anaesthetic") OR ("chondrotoxicity" AND "platelet rich plasma") OR ("chondrotoxicity" AND "collagen medical devices") OR ("chondrotoxicity" AND "bisphosphonates").The search was focused on articles published between 1975 and April 2024.Studies were included if they (1) were in vivo or in vitro studies, (2) evaluated chondrotoxicity, and (3) were published in the English language.The search was extended through the reference lists of the selected papers.The exclusion criteria encompassed (1) studies focusing on systemic delivered devices; (2) case reports, reviews and commentaries; and (3) papers not published in the English language.Articles were divided according to the delivered drug and whether they were conducted in humans or animal models.The authors looked for papers discussing the chondrotoxicity of different drugs after their IA injection.Two reviewers (C.P. and A.S.) independently selected the papers by reading titles and abstracts.A third reviewer finalized the selection in case of disagreement (A.M.).After the selection, each title/abstract/full text was independently assessed by each of the authors.The study flowchart is shown in Figure 1.
rich plasma") OR ("chondrotoxicity" AND "collagen medical devices") OR ("chondrotoxicity" AND "bisphosphonates").The search was focused on articles published between 1975 and April 2024.Studies were included if they (1) were in vivo or in vitro studies, (2) evaluated chondrotoxicity, and (3) were published in the English language.The search was extended through the reference lists of the selected papers.The exclusion criteria encompassed (1) studies focusing on systemic delivered devices; (2) case reports, reviews and commentaries; and (3) papers not published in the English language.Articles were divided according to the delivered drug and whether they were conducted in humans or animal models.The authors looked for papers discussing the chondrotoxicity of different drugs after their IA injection.Two reviewers (C.P. and A.S.) independently selected the papers by reading titles and abstracts.A third reviewer finalized the selection in case of disagreement (A.M.).After the selection, each title/abstract/full text was independently assessed by each of the authors.The study flowchart is shown in Figure 1.

Results
In total, 65 studies were included in the review, encompassing research on 5 different compounds: corticosteroids, local anaesthetics, nonsteroidal anti-inflammatory drugs (NSAIDs), hyaluronic acids (HA) and platelet-rich plasma (PRP).

Results
In total, 65 studies were included in the review, encompassing research on 5 different compounds: corticosteroids, local anaesthetics, nonsteroidal anti-inflammatory drugs (NSAIDs), hyaluronic acids (HA) and platelet-rich plasma (PRP).

Corticosteroids
Papers on corticosteroids' chondrotoxicity analysed for the purpose of this review are reported in Table 1.A total of 33 papers regarding corticosteroids were retrieved, 14 of which utilized human subjects, 17 utilized animal subjects, and 2 investigated humans and animals in comparison.Sixteen of the papers report in vivo studies, 16 in vitro and 1 ex vivo.The knee was by far the most represented joint.
Overall, 19 (58%) studies expressed evidence about the possible chondrotoxic effects of multiple corticosteroids, raising concerns about the application of at least some of these in clinical practice.Albano et al. used multiple injections of betamethasone inside the knee of 80 living rabbits and, after 8 days of follow up, a decreased concentration of proteoglycans was found [15].Lutfi et al. focused on betamethasone injections, also on rabbit (n = 12), for 10 weeks and, though the sample was smaller, the authors report evidence of macroscopic damage to articular cartilage [16].
Betamethasone also showed signs of chondrotoxicity in vitro.In a study on dogs' chondrocytes, Sherman et al. used multiple drugs (1% lidocaine, 0.5% lidocaine, 0.25% bupivacaine, 0.125% bupivacaine, 0.0625% bupivacaine, betamethasone acetate, methylprednisolone acetate, and triamcinolone acetonide) finding a severe chondrotoxic effect on all of them, except bupivacaine and triamcinolone [17].Two in vitro studies, in which multiple drugs were used (dexamethasone sodium phosphate, methylprednisolone acetate, betamethasone sodium phosphate and betamethasone acetate, or triamcinolone acetonide in combination with doses of 1% lidocaine or 0.25% bupivacaine) on human cartilage, were conducted by Braun et al. and Farkas et al.In both trials the association between corticosteroids and anaesthetic showed significant time-dependent chondrotoxicity [18,19].Dragoo et al. also had comparable findings, with a 14-day follow-up protocol, showing significant cell death related to betamethasone use [20].Only one study, by Davis et al., showed no harm sign with the use of betamethasone.In their in vitro study on human and animal chondrocytes, no cell death caused by the drug was reported [21].
Methylprednisolone was found to be another widely employed corticosteroid: ten studies that investigated its chondrotoxicity showed generally discouraging results.Sherman et al. [17] and Braun et al. [18] evaluated methylprednisolone among the other drugs used in their studies, confirming its detrimental effect.Sherman also conducted a second study comparing the effects of methylprednisolone and triamcinolone in association with lidocaine on 20 dogs, showing lower viable cell density 7 days after initial treatment [22].In their 1984 study, Ishikawa et al. compared the effects of methylprednisolone with halopredone on rabbit specimens with a 13-week follow-up period and severe cartilage damages was ultimately found.This was the only study we retrieved that involved halopredone [23].Murray et al. [24] and Pelletier et al. [25] conducted two similar studies, one regarding the effects of IA methylprednisolone on horses and the other on dogs.Both studies had an 8-week follow-up period but had the following opposite results: Murray reported alteration of the mechanical integrity of articular cartilage while Pelletier did not find any negative effect.Robion et al. also carried out a study that was analogous to Murray's but with a longer follow up (13 weeks).This study observed the inhibition of procollagen II synthesis and increased release of degradation products of aggrecan from articular cartilage [26].A different approach was used by Seshadri et al. who conducted an in vitro evaluation of the effects of methylprednisolone alone or in association with lidocaine [27].The authors reported dose-and time-dependent decrease in chondrocyte viability after exposure to methylprednisolone alone and a synergistic decrease in chondrocyte survival with exposure when combined with lidocaine.An interesting paper by Chu et al. evaluated the effects of methylprednisolone (alongside naproxen and meloxicam) on an ex vivo sample of human knees.There, the authors report possible down-regulation of the plasminogen activator/plasmin system and gelatinases expression in the early osteoarthritic knee of humans, which could be related to structure-modifying activity [28].No evidence of deleterious effects on cartilage have been reported by Gibson et al., who conducted a protocol applying methylprednisolone at the knee of 10 primates [29].We report a final study by Baumgarten et al. which shows no evidence of chondrotoxicity in any of the 56 patients treated with methylprednisolone and anaesthetic (1% lidocaine and/or either 0.5% or 0.25% bupivacaine) [30].
Triamcinolone was also deeply analysed and we retrieved 13 papers about it, almost equally divided between those that indicated possible negative and positive effects, with a small imbalance in favour of the latter.In their multi-drugs studies, Braun et al. and Sherman et al. showed chondrotoxicity can also be caused by triamcinolone [17,18,22].Celeste et al. focused only on triamcinolone, applying it via IA injections to horses: after a 13-week follow up, an increase in markers of cartilage matrix degradation and aggrecan turnover was found; interestingly, the authors also report altered articular cartilage and collagen metabolism in treated control joints, signs of a possible systemic effect of the IA injections [31].Among other drugs, Dragoo et al. reported the possible chondrotoxic effect of triamcinolone [20].Two in vitro studies were carried out by Suntiparpluacha et al. and Syed et al., applying triamcinolone on knee-derived human chondrocytes; both showed an induced chondrotoxicity and, in the first study, the authors found an increased oxidative stress and altered expressions of genes involved in cell death [32,33].Several papers showed positive or benign effects associated with triamcinolone.Bolt et al. carried out an in vitro protocol on horses' chondrocytes and reported that triamcinolone seems to support chondrocyte morphology in culture and protects chondrocytes from toxic effects [34].Another study on horses was conducted by Frisbie et al., applying the drug on living animals; after a 6-week follow-up period signs of favourable effects on articular cartilage parameters were found [35].Similarly, Pelletier et al. applied triamcinolone on 12 dogs at knee level, finding no deleterious effects on articular cartilage after 8 weeks [36].A large study by Huppertz et al. evaluated the effect of triamcinolone on 21 humans (2 Ankle, 1 Elbow, 17 Knees).After 13 months, no evidence of toxic effects on cartilage was evident at MRI evaluation [37].Raynauld et al. carried out an even bigger study on 68 patients who were evaluated every 3 months for 2 years after IA knee injections; at placebo comparison no difference in loss of joint space and no deleterious effects on the anatomical structure of the knee were observed [3].Finally, Williams et al. found a marked, dose-dependent protective effect of triamcinolone applied with IA injections to the knee of 43 guinea pigs [38].
Literature is also present about the possible chondrotoxic effects of dexamethasone.We retrieved five papers, four of which reported negative outcomes.Liu [39][40][41][42].Only one paper reported a reduction of inflammatory mediator (nitric oxide) expression after application of dexamethasone on horses' chondrocytes in vitro [43].
Finally, we retrieved three papers regarding hydrocortisone.Salter et al. have reported a deleterious effect on the articular cartilage of 55 rabbits when hydrocortisone was applied at the level of the knee [44].On the other hand Wang et al. conducted two studies with similar characteristics (in vitro, knee-derived chondrocytes from humans) that showed enhanced ability to synthesize ECM macromolecules (aggrecan, type II collagen and fibronectin), inhibition of degenerative enzymes, increased hyaluronan levels, and inhibition of deleterious intracellular protease MMP-1 [45,46].

Local Anaesthetics
In our research, 18 studies regarding the chondrotoxic effects of anaesthetics were retrieved (Table 2).Papers concerned both animals and humans, in vivo and in vitro, and focused mainly on the effects of bupivacaine, lidocaine and ropivacaine.Bupivacaine is the anaesthetic for which we found the most abundant literature.Breu et al. evaluated chondrotoxic effects of bupivacaine in comparison with ropivacaine, and mepivacaine.The authors found chondrotoxic effects in time-dependent, concentration-dependent and drug-dependent manners.In particular, chondrotoxicity increases from ropivacaine to mepivacaine to bupivacaine, indicating that chondrotoxic and analgesic potencies do not directly correlate [47].Another in vitro study comparing the effects of bupivacaine with lidocaine and ropivacaine on bovine chondrocytes by Lo et al. reports a dose-and durationdependent detrimental effect on chondrocyte viability [48].An in vitro comparison on human chondrocytes between bupivacaine, ropivacaine, lidocaine and/or vitamin C was established by Tian et al., who found a chondrotoxic effect caused by all of these anaesthetics, with ropivacaine being less detrimental than bupivacaine and lidocaine; however, vitamin C improved chondrocyte viability and decreased apoptosis levels following exposure to anaesthetics [49].Shaw et al. also compared the in vitro effects of bupivacaine, ropivacaine and liposomal Bupivacaine on bovine chondrocytes.They observed a dose-dependent chondrotoxic effect for all three drugs, with the highest chondrocyte viability associated with liposomal bupivacaine [50].Shaw et al. also conducted a second study comparing liposomal bupivacaine and bupivacaine after IA knee injections in the knee chondrocytes of Yorkshire cross piglets and found that liposomal bupivacaine showed a good safe profile for IA injections [51].Mwale et al. compared bupivacaine, levobupivacaine and ropivacaine in vitro on hip and elbow cells of dogs.All of the drugs caused decreased chondrocyte viability, though ropivacaine showed less chondrotoxicity [52].We retrieved three papers by Chu et al., all of which focused on bupivacaine applied at the knee level.Two of these were in vitro studies [53,54], the first on humans and bovine chondrocytes, the second only on bovine cells.The third was an in vivo protocol on rats (n = 48) with a follow-up period of up to three months [55].All three of Chu et al.'s papers found detrimental dose-and time-dependent effects of the drug on articular cartilage, even after only 15 to 30 min of in vitro exposure.Oyadomari et al. also carried out an in vitro study on engineered neocartilage constructs and bovine cells and have reported significant chondrotoxicity in native explants and neocartilage and a significant weakening of the mechanical properties of the neocartilage [56].Rengert et al. evaluated the in vitro effects of bupivacaine on dog chondrocytes and have reported a concentration-dependent chondrotoxic effect [57].Stueber et al. were the only group to compare bupivacaine and dexamethasone on human chondrocytes in vitro and found a concentration-dependent chondrotoxic effect associated with the anaesthetic, though dexamethasone failed to induce cytotoxicity [58].Another peculiar association (bupivacaine or bupivacaine with epinephrine) was tested by Gomoll et al. on 30 live rabbits; after 1 week, significant histopathologic and metabolic changes in articular cartilage, without correlations with the drug used, were found [59].Finally, one paper involving bupivacaine in comparison with articaine and lidocaine was carried out on the temporo-mandibular joint; 24 rabbits were subjected to IA injections and, after 4 weeks, apoptotic effects on chondrocytes and degenerative changes in the joint articular structures were found [60].
Several studies focused only on lidocaine.Karpie et al. evaluated its effects at two different concentrations (1% or 2%) in vitro on bovine knee cells and reported a dose-and time-dependent chondrotoxic effect [61].Maeda T et al. confirmed the chondrotoxicity of lidocaine (that increased in a time-and concentration-dependent manner) in their in vitro evaluation on the knee, hip, and shoulder chondrocytes of rabbits [62].Di Salvo et al. attempted to assess the effect of associating lidocaine plus adrenaline with IA injections in the elbows of 12 dogs.The authors report a dose-and time-dependent chondrotoxic effect of lidocaine on the viability of articular cells, reduced by the application of adrenaline [63].Vrachnis et al. carried out the only study in which treatment did not induce any histological changes in articular cartilage.The authors applied lidocaine or ropivacaine on 32 rats and evaluated the effects after a follow-up period of up to 60 days [64].

Nonsteroidal Anti-Inflammatory Drugs (NSAIDs)
Only five studies were retrieved that focused specifically on the association between NSAIDs and chondrotoxicity (Table 3).

Ibuprofen
In vitro Shoulder Animals (sheep) 3 Up to 13 days Ibuprofen did not reduce cell viability and protein content on chondrocyte monolayers.Ibuprofen reduced synoviocyte viability.Overall, ibuprofen showed limited toxicity.
Four studies were carried out in vitro, with the only exception being the study by Sagir et al. which focused on Dexketoprofen.The authors conducted a protocol involving 35 rats with IA injections in the knee and, after a follow-up period of up to 21 days, inhibition of cell proliferation was found despite the lack of any signs of significant histopathologic effects [65].
Regarding the in vitro studies, Abrams et al. evaluated the effects on human chondrocytes related to different compound exposures and, for NSAIDs, ketorolac was associated with significantly increased cell death [66].Alaseem et al. focused their paper on the synthesis of type X collagen (COL X), a marker of late-stage chondrocyte hypertrophy expressed in the mesenchymal stem cells (MSCs) of OA patients, their results show that naproxen seems to induce type X collagen gene (COL10A1) expression in bone-marrow-derived MSCs from healthy and OA donors, to which the authors relate as a sign of possible chondrocyte differentiation towards an undesirable degenerative phenotype [67].In another in vitro assessment, Bèdouet et al. conclude that the limited toxicity of ibuprofen at low and high concentration in sheep joint shoulder makes this enantiomer a promising drug candidate for the loading of intra-articular DDS [68].However, Dingle et al. report opposite results for their evaluation of 845 human cartilage samples.Here, ibuprofen, indomethacin, aspirin, naproxen and nimezulide showed substantial and significant inhibition of glycosaminoglycan synthesis while diclofenac, piroxicam, nabumetone and paracetamol had no significant effects [69].

Hyaluronic Acids (HAs)
For the purpose of this review, seven studies regarding the possible role of HAs on chondrotoxicity were retrieved (Table 4).
All of the studies were conducted with in vitro models and focused on assessing HA efficacy in reducing chondrotoxicity caused by other drugs.Two papers evaluated the association between HA and carprofen, both based on dog chondrocytes: in the first, Euppayo et al. report that HA alone preserved chondrocyte survival but, when in combination with carprofen, could not reduce its chondrotoxicity [70].The second was carried out by Nganvongpanit et al., who found that HA was able to decrease chondrocyte apoptosis [71].Euppayo et al. also evaluated HA in association with a corticosteroid, triamcinolone.The authors applied the drugs in vitro on dog knee and elbow-derived chondrocytes and have reported that the combination increased the percentage of cell viability in normal chondrocytes [72].
Four studies analysed the association of HA with anaesthetics.Its association with lidocaine was assessed by Lee et al.The authors report that HA seems to suppress lidocaineinduced apoptosis of human chondrocytes in vitro [73].Liu et al. evaluated its association with bupivacaine.In their paper, bovine knee-derived chondrocytes were used, with the results showing a reduction of cell death rates [74].Bovine chondrocytes were also used by Onur et al. to test HA effects in association with lidocaine and bupivacaine; HA was able to reduce cytotoxicity caused by the second but not by the first [75].The last paper we retrieved, by Moser et al., was conducted to test the effects of HA in association with anaesthetics (lidocaine, bupivacaine, ropivacaine), and glucocorticoids.Human knee derived chondrocytes were used and the authors report that HA enhanced attachment and branched appearance of the chondrocytes and improved the metabolic activity [76].Few studies have assessed the efficacy and safety of infiltrative treatment with PRP; for this paper we were able to find only two (Table 5).
In fact, in 2015 Al-Ajilouni et al. observed that IA PRP injections significantly improved functional outcome scores in 48 patients with knee osteoarthritis after a follow-up period of 52 weeks.There was no specific association and a specific paragraph on chondrotoxicity was not present, although it is likely that the improvement in functional outcomes could be directly related to a corresponding improvement in articular cartilage qualities [77].In 2013 Beitzel et al. focused their work specifically on PRP-related chondrotoxicity; their study showed that tendon and cartilage cells increased cell viability after an exposure to allogeneic PRP.Proliferation also increased [78].

Collagen Medical Devices, Biphosphonates
No papers were retrieved regarding collagen medical devices (CMDs) or for bisphosphonates.

Discussion
A wide variety of protocols were found for all samples and settings, with heterogeneous results.This lack of homogeneity makes it complicated to perform comparative analyses either between studies involving the same molecule or, even more, between different compounds.Thus, all the possible conclusions we could draw are weaker and the answer to the question posed at the beginning still seems unanswered.
Corticosteroids are among the most used drugs for IA injection in consideration of their well-known anti-inflammatory properties (i.e., capacity to reduce the inflammatory cells inside joints, to prevent phagocytosis, lysosomal enzyme release, and to inhibit proinflammatory cytokines).Thus, IA corticosteroids are able to reduce pain and swelling and increase joint range of motion (ROM) and functionality by decreasing inflammation [79,80].
These pharmacological classes can improve functional outcomes for degenerative diseases or after an injury and can reduce the time to return to a good quality of life (QoL) for people affected by different painful conditions [81][82][83][84].However, an abundance of data are available in the literature about their possible toxic effects on articular cartilage.Both direct and indirect evidence for reduced articular cartilage matrix synthesis and/or proteoglycan degradation have been provided both by in vitro and in vivo studies [85].Although there is general agreement on their dose-or time-dependent toxicity, they are still widely used for IA injection.If, on one hand, ACR guidelines [86] ". ..recommends intraarticular corticosteroid injections for the knee and hip. ..", on the other hand both OARSI and AAOS underline how intra-articular steroid therapy is advised with a low level of evidence [4,87].
Studies have been conducted on both animal and humans, both in vivo and in vitro.In 2015 Wernecke et al. published a systematic review on the effect of intra-articular corticosteroids on articular cartilage.Forty scientific papers were selected out of 1929 publications on the topic and the collected data were divided on the basis of the investigated steroid formulation.The authors concluded that the time-and dose-dependent deleterious effects on articular cartilage are widely supported by the basic scientific literature.[88].Our results seem to support these findings; most of the papers we assessed reported risks of chondrotoxicity, with triamcinolone appearing to be the safest intra-articular corticosteroids, though only because the other analysed seemed to be even worse.
Almost all of the papers about local anaesthetics showed signs of chondrotoxicity, and investigated them via conventional histologic analysis, cell quantification (confocal, immunofluorescence), nuclear morphology changes, and metabolic sulphate uptake assessment.These studies were mainly focused on the effects of bupivacaine, lidocaine and ropivacaine.Local anaesthetics are widely used to achieve analgesia in painful joints.Anaesthetics are usually injected inside the knee, shoulder or other joints, often combined with other agents, such as corticosteroids, after injury or for the treatment of degenerative disease.Local anaesthetics express their effects by inhibiting voltage-gated sodium channels on nerve cell membranes, thereby preventing development of an action potential and blocking nerve transmission [89].Usually, a single dose of an anaesthetic is enough to achieve pain control after IA injections, though this also depends on the properties of the single molecule (e.g., bupivacaine, ropivacaine, lidocaine) [90,91].Achieving a rapid reduction of pain is crucial for the success of an individual rehabilitation project (IRP).Local anaesthetics seem to have an overall detrimental effect on articular cartilage, with a dose-and time-dependent modality.Ropivacaine appears to be the safer of these, alongside liposomal bupivacaine.The association with vitamin C seems to reduce the negative effects but further research is needed.
NSAIDs are widely used in arthritis and osteoarthritis, mostly to reduce pain caused by the disease and relative functional burdens [92,93].NSAIDs inhibit cyclooxygenase (COX-1 and COX-2), thus preventing the synthesis of inflammation mediators (e.g., prostaglandin, prostacyclin and thromboxane).Some of the most used NSAIDs (i.e., naproxen, diclofenac, aspirin and ibuprofen) non-selectively inhibit both COX-1 and COX-2, though selective COX-2 inhibitors are also available, such as celecoxib and meloxicam [94].This difference is not a mere pharmacological one, it has several clinical implications.COX-1 is expressed in most tissues and has a crucial role in the protection of the gastric mucosa, in the regulation of the renal blood flow and of the vascular homeostasis.COX-2 is less constitutively expressed but higher concentrations can be found near inflamed tissues [95].Numerous ways of administration are available for NSAIDs: oral, intradermal, intravenous and intramuscular; IA injections with NSAIDs are not widely used though this method of delivery might reduce systemic side effects while at the same time increasing local efficacy [96].NSAIDs have also shown a rather low safety profile: 4 of the 5 studies analysed indicated various types of damage to articular cartilage.The one exception indicated encouraging results for only some molecules (aceclofenac, tenidap and tolmetin) and still reported problems with others (ibuprofen, indomethacin, aspirin, naproxen and nimezulide).Data about the NSAIDs use for IA injections and about their chondrotoxicity are, however, too scarce to be relied upon to finalize any conclusion.Broader evaluation must thus be carried out.
A different point can be raised, on the other hand, for HA.HA is one of the most used compounds for IA injections, especially for degenerative diseases such as osteoarthritis, which causes functional disability and reduction of QoL.HA is physiologically present inside joints, both at the cartilage and the synovial fluid level [97].HA has the important role of lubricant and homeostasis regulator inside joints; with disease progression its properties change and mechanical abnormalities in the synovial fluid arise [98].Viscosupplementation with HA could thus be helpful to restore a more physiological microenvironment inside affected joints.Two peculiarities were noticed with regard to the studies we retrieved about HA: all were conducted in vitro and all focused on assessing their efficacy in reducing other compounds' chondrotoxicity.Overall, HA seems to have good efficacy in reducing the damaging effects on articular cartilage caused by other molecules.No study was retrieved about the possible effects of HA alone; chances are that its properties are not related with chondrotoxicity, but research would be needed to confirm this hypothesis.
Finally, of the two studies we assessed regarding PRP, only one was actually focused on chondrotoxicity.PRP is a treatment that uses platelet-enriched blood plasma to promote tissue regeneration.Platelets contain growth factors, which are essential for wound healing and tissue repair.These agents, alongside many others (i.e., transforming growth factor b, fibroblast growth factor, vascular endothelial growth factor, and connective tissue growth factor) account for the therapeutic effects of PRP [99].PRP is obtained by centrifuging the patient's own blood, and then injected into the area to be treated to stimulate healing and regeneration.PRP is emerging as a non-surgical option for degenerative diseases of the musculoskeletal system, as it carries the possibility of reversing cartilage damage, joint space thinning and associated pain [100,101].The risks of adverse reactions related to the treatment are reduced by it being autologous [102].However, the literature evidently lacks papers on this topic; any kind of consideration made at this time seems to be inadequate and more research is needed to clarify PRP potential, both beneficial and harmful.
We found no information regarding the chondrotoxic effects of CMDs and bisphosphonates.This undoubtedly represents a gap in the literature and makes it impossible to draw any conclusions about these compounds.This finding could be a stimulus for future research in this regard.

Conclusions
In conclusion, current research regarding the chondrotoxicity of the most commonly used drugs for IA injection lacks the standardization and numerosity to allow definitive assumptions.Corticosteroids (with the possible exception of triamcinolone), local anaesthetics (with the possible exception of ropivacaine and liposomal bupivacaine), and NSAIDs showed too poor a safety profile to be used lightly in clinical practice.HA appear to be safe and also able to reduce the risks associated with other compounds, thus favouring therapeutic combinations.Moreover, too few data exist with regard to PRP, the potential of which has yet to be evaluated.Finally, it appears that results are heavily dependent on dosage and number of injections used, both on humans and animals.In our opinion, a careful clinical application that falls within the framework of a set of guidelines may prevent side effects and adverse events.
et al. (humans), Song et al. (humans), Su et al. (humans and bovine) and Tu et al. (humans) used dexamethasone on knee articular cartilage samples, reporting increased apoptosis, inhibition of ECM synthesis and inhibition of positive remodelling factors (TGF-β and TIMP-3) associated with its use