Tendinopathy: sex bias starts from the preclinical development of tendon treatments. A systematic review

The main finding of this study is that preclinical studies performed to assess treatments for tendinopathies currently used in clinical practice largely neglected the importance of sex. None of the published studies analyzed sex-based differences, and only 4% of the studies reported disaggregated data suitable for the analysis of treatment results based on sex. Beside the low number of studies including both sexes, there is an alarming female under-representation in the study subjects, in particular for injective therapies. Despite the growing awareness of the importance of investigating treatments in both males and females, the field proved resistant from properly designing studies, and the lack of sex-representation remains critical.

The notion of the importance of sex and gender is not new to the scientific literature, which defines sex as biological and physiological features characterizing male and female individuals, while gender refers to socially constructed roles, behaviors and identities of female, male, and gender-diverse people [16, 17]. Since the NIH Revitalisation Act came out in 1993, several international organizations encouraged a broader female inclusion in both preclinical and clinical studies [13]. Between 2005 and 2008, the Gender Basic Project in Europe supported scientific studies focusing on sex and gender and on the importance of both sexes’ outcomes. Few years later, the European Association of Science Editors (EASE) established a Gender Policy Committee with the aim to develop a set of guidelines for reporting of Sex and Gender Equity in Research (SAGER) [17]. More recently, in 2012, the Canadian Institute of Health, as well as the German Society of Epidemiology, mandated a justification for any study if only one sex was considered [18, 19]. In 2016, NIH requested all applicants to provide a justification in case of one-sex studies [14]. Considering the several authoritative and pressive guidelines requiring the field to address this key issue, one may expect a literature evolution toward a more balanced sex representation. Surprisingly, this is not the case, and the literature is showing a significant worsening trend. In fact, despite the 2016 NIH request for looking at Sex as a Biological Variable, this systematic review on one of the most common orthopedic pathologies, tendinopathy, showed that the number of male specific studies increased more than the female focused studies, while the percentage of studies including both sexes even decreased.

This is sadly not a surprising result. Bryant et al. in 2018 analyzed four orthopedic journals and found that only 13% of the studies disaggregated data by sex [20]. While this issue is less explored in the orthopedic field, these results echo many non-orthopedic studies denouncing the unrecognized female relevance in the preclinical medical research. Over a decade, more than 79% of animal studies published in the Pain journal were on male animals, while only 4% were explicitly focused on testing sex differences [17]. In cardiovascular research, Ramirez et al. found that only 13% and 15% of the studies included females and both sexes, respectively [21]. Low percentages of female inclusion and data disaggregation can also be found in a long list of other disciplines, including basic science, dermatology, neurosciences, pharmacology, otolaryngology, etc. [17, 22,23,24,25,26]. As emerged from this systematic analysis of the literature, female neglection reaches an alarming level in the research efforts to develop solutions to address tendinopathies, a common and debilitating clinical challenge. Beside the marginal effort on including both females and males in the same experiments, no study recognized the lack of comparative sex-based analysis as a limitation, and 20% did not even considered it useful to report the sex of the studied animals. Only 13% of the studies included both males and females and, among these studies, only 6 disaggregated results by sex, representing only 4% of the total studies, which include an even lower number of animals (2%). In these studies, moreover, the results have been disaggregated due to the description at a subject level but without the analysis on the influence of sex, thus reflecting the fact that not one of these authors consciously looked out for sex-based outcomes.

Men and women are characterized by differences resulting from millions of years of evolution and therefore warrant scientific differentiation and investigation. This applies both to men and women. Men should be better studied with more representative models, since 71% of the large male animals documented in the literature were castrated, which could affect the metabolism and physiological treatment response, hindering the translatability of the study findings in humans. In women, estrogen receptors (ER) have been localized in both ligaments and tendons, and an increase of estrogen concentration has been linked to a decreased collagen synthesis and a lower tendon stiffness [8, 27, 28]. Also, it has been hypothesized that the hormonal fluctuations related to the menstrual cycle and menopause may influence the incidence of musculoskeletal injuries, with an increasing risk in the ovulatory phase and in menopausal women, respectively [29,30,31,32]. Based on these premises, some attempts of investigating sex-based differences in terms of tendon biology have been performed in animal models. In rats, it was demonstrated that ageing and more significantly estrogen deficiency negatively affect tendon metabolism and healing rate. A decrease in fibronectin and elastin, an increase in vascular endothelial growth factor and Metalloproteinase-13, and a low healing rate of microwounds have been found in tenocytes of estrogen-deficient rats when compared to young and old groups [33]. In mice, male and female tendons differ for extracellular matrix proteins and proteoglycans composition, mechanical properties, gene expression, protein composition, resistance to mechanical stress, and response to therapy [34,35,36]. The combined effects of estrogen and mechanical loading may alter the mRNA expression for extracellular matrix components exclusive of females, supporting the higher injury risk in females [37], and a correlation has been suggested between the expression of estrogen receptor-beta and mechanical stress in rat tendinopathy [36]. Despite this evidence, few studies are currently available on this topic and the results are not conclusive, which underlines the importance of further exploring sex-related differences in both etiopathogenetic mechanisms and treatment development.

Given the sex-specific incidence of tendinopathy, with sex hormones affecting tendon metabolism, structure, biomechanical properties and injury risk, and the interplay with age-related tendon modifications, sex differences should be studied across the entire lifespan to gain insight into disease pathogenesis and identify treatment targets for different sexes and times of life. The incidence of tendon injuries increases dramatically with age-related changes in tendon structure, composition, mechanical function, and injury risk which appears to be sex-dependent, with the incidence of tendinopathy rising following menopause [38, 39]. All these aspects are largely overlooked in the preclinical literature. For example, no animals in physiologic menopause were studied, and only 2% of the female animals were neutered. Commonly used animal model species do not naturally undergo menopause with its associated decrease in estrogen levels and ER expression. Ageing rodents fail to consistently replicate the low estrogen concentrations characteristic of human menopause [40, 41]. Traditionally, methods to induce menopause in animals have focused on ovariectomy, which yields a substantially different hormone profile, with a sudden loss of all ovarian steroids rather than continued release of androgens and low levels of other steroids, as well as an altered hypothalamic–pituitary–gonadal axis compared to post-menopausal women [41]. The common use of sexually immature animals, which distorts the effect of ovariectomy, lends further emphasis to the necessity for fit-for-purpose animal models in general and in specific for research into sex- and age-specific pathogenetic and reparative mechanisms [41, 42].

Neglecting a sex-based analysis in preclinical studies might lead to a great bias, with a relevant impact on the translational research. The improper sex-representation in the preclinical research of tendinopathy has been often justified by the fear of a more complex model due to hormonal fluctuations. This led to the overall lower inclusions of female animals, mainly limited to the clinical veterinarian studies in dogs and horses, while experimental studies overly draw their attention to male small animal models. In rodents, the exclusion of female seems to be systematic, even though no justification for this selection bias is provided by scientific evidence [43,44,45,46]. Indeed, there is the improper belief that females are subjected to a greater variability, due to the confounding effect of the estrus cycle, making them unstable and unsuitable for their use as preclinical models. This myth has been questioned several times. Prendergast et al. in 2014 analyzed 293 studies on biomedical research and were able to prove not only that females do not express more variability, but also that male mice could be even more variable, due for example to the group-housing conditions that can lead to fight and consequent hormonal pathways activation [43]. In 2016 another meta-analysis of neuroscience studies confirmed that female rats exhibited the same, or even less, variability than males [47]. This was true for behavioral, electrophysiological, neurochemical, and histological measures. Thus, the authors concluded that power analyses based on variance in male measures are sufficient to yield accurate numbers for females as well when designing experiments to include both male and female rats. To challenge the assumption of inherently greater female variability, Itoh et al. analyzed a large microarray data set measuring gene expression in various tissues of both mice and humans, comprising the analysis of more than 5 million probes [44]. On average, male gene expression was slightly more variable than that of females, reaching again the same conclusion of no evidence for greater variability in females than in males. Thus, the scientific evidence does not justify male selection in preclinical biomedical research.

The development of treatments requires the study of both sexes. Ignoring sex-disaggregation and female specific effects in a preclinical phase can lead, in the best-case scenario, to missing the opportunity of investigating such effects in clinical studies. The consequences might be not trivial, as for the adverse events in women, for instance, with the case of Zolpidem in 1992. This drug was approved and commercialized with the same dosage for both sexes and, 10 years later, it was halved by FDA in women due to sex-specific severe side effects [48, 49]. Between 1997 and 2001, ten drugs have been withdrawn from the US market and eight of them were more harmful to women [50]. While the importance of sex-related differences goes across all fields in biomedical research, the implementation of studies with a proper study design is severely lagging, as vividly portrayed by the current meta-analysis on one of the most common orthopedic diseases. At best, the importance of comparing results of both sexes has been underestimated. Experimental models used to develop the main conservative treatments for tendinopathy focused their attention of only one sex, most commonly on male animals. This is surprisingly even more true for the newest treatments introduced in the clinical practice, the injective approaches. In fact, innovative blood derivatives and cell-based approaches have been tested and implemented in clinical practice without considering potential sex-related risks and possible sex-related differences in terms of healing potential, even though orthobiological solutions could be particularly influenced by the biological differences characterizing males and females [7, 51, 52].

This literature analysis was focused on a common pathology, tendinopathy. Still, a limitation is that different types of tendinopathies have been grouped together for the purpose of this study. Thus, different tendons and treatments may present more or less bias in terms of depth of investigation. However, this goes beyond the purpose of this study, which was focused on a broader concept of sex-representativeness in the overall research field on the development of tendon treatments. To this aim, this study encompassed the entire preclinical literature on conservative tendon treatments, reporting on 150 studies and 8231 animals, and unquestionably identified a critical field limitation due to a severe sex bias, even though it should be underlined that a minority of them on horses and dogs were not intended to develop new treatments, being per se clinical veterinary studies, therefore being outside the translational preclinical context. The enrollment of both sexes should be pursued in future studies. Females do not have a substantial increase in outcome variance, irrespective of the cycle state [27, 45, 53]. Also, if a specific hormonal influence is suspected to be a further complex study variable, the enrollment of both sexes is even more strongly recommended, to better understand how to properly address tendinopathies in women.

Proper gender-balanced studies are needed. Still, this may prove difficult. Preclinical research presents a delicate balance of several aspects which weigh in defining the study design. Having both males and females in every experiment faces impediments both on ethics (increased animal numbers vs the need for “reduction”) and finances (increased budgets required due to higher number of animals and increase in purchase, housing costs, etc.), as well as practical limitations to properly implement environmental/phenotypic study setting (for instance male goats together will be much more active fighting than a female group in normal housing conditions) [54, 55]. Finding the proper balance would require guidelines on how to properly power studies detecting a sex difference based on the specific animal model and study target, while also giving guidance on the most suitable settings to derive sex-based results, especially in terms of preclinical data to translate effective musculoskeletal treatments in women.

Women are more at risk of sustaining tendon and ligament injuries [10, 56, 57], and their growing sport participation urges a decisive change of direction in preclinical tendon studies to provide specific data to develop more suitable treatments for both men and women affected by tendinopathies.

留言 (0)

沒有登入
gif