A case report published in this issue of Equine Veterinary Education (Lenoir et al. 2022) documents the radical surgical resection of a portion of the superficial digital flexor tendon in an area that usually shows poor natural healing. This case report is interesting because the gap became filled with tissue with similarities to tendon on ultrasound and enabled the horse to regain soundness with the aid of an orthotic device.

The resection of the tendon was deemed necessary because of suspected infection although this was not possible to confirm with positive culture or histology. While infection is rare in tendon, it does carry a poor prognosis (Kidd et al. 2002) which, therefore, justifies radical debridement. However, this left a large gap between the two ends of the superficial digital flexor tendon. Gap healing of equine tendons is seen following tendon lacerations with the formation of bridging fibrous ‘callus’ (Jordana et al. 2011). However, this can be a protracted process and may ultimately become exaggerated with restrictive adhesion formation. Reflection of the tendon ends can also occur which can delay healing further, and so, there have been many attempts to assist healing, especially for the digital flexor tendons because of their importance in normal weight-bearing limb function. These approaches have included the use of a variety of biologic and synthetic implants to bridge the gap. The choice of a suitable material has progressed from inextensible carbon fibre (Vaughan et al. 1985), through artificial nonabsorbable materials (Gibson et al. 2002; Barrett et al. 2014; Fig 1), to long-lasting absorbable materials (Eliashar et al. 2001; Jenson et al. 2005), and even autologous tendon tissue (Valdes-Vazquez et al. 1996). Their role has been postulated to act as a scaffold to provide a conduit along which cells and healing vasculature can grow as well as maintain the tendon ends in alignment. This is in contrast to the use of scaffold in smaller animals where they can act as a robust mechanical ‘bridge’ because of the smaller forces involved. This is not possible to achieve in horses and, indeed, it has been our observation in the past, especially in hindlimbs supported with a distal limb cast post-operatively which has immobilised the fetlock but allowed hock and stifle flexion, that the implants have become detached from the transected tendon ends. This observation, the poor availability of appropriately sized, long-lasting but absorbable, implants, and the failure to show significant benefit over debridement alone in the few limited studies published, has meant that implants are now rarely used. As a result, clinicians have relied on simple debridement and natural repair while the region is protected from loading using external coaptation, as described in this case report.

What is, however, surprising in this case report is that this gap healing has occurred in a region which is particularly problematical for healing. Plantar to the fetlock, where the tendon is subjected to compression as well as tension, requires a more complex extracellular matrix. Moreover, its intrathecal location means that the tendon is surrounded by synovial fluid and lacks a paratenon. We have recently established that synovial fluid has toxic effects on the cells within tendon which helps explain why many tendon and ligament lesions that communicate with the synovial cavity of a tendon sheath, bursa or joint often fail to heal spontaneously (Garvican et al. 2017). The paratenon is considered to be a major player in extrinsic (coming from surrounding tissues) healing, which is believed to dominate extrathecal tendon repair, at least in experimental animal models (Kajikawa et al. 2007). The paratenon is continuous with the endotenon, or interfascicular matrix, which is believed to be the source of endogenous tendon stem cells and neovasculature (Cauvin 2000; Marr et al. 2017). However, the lack of a paratenon in intrathecal locations, means that, in the absence of adhesion formation, repair relies more heavily on intrinsic (from within the tendon) healing which is less effective and slower. Adhesions will enhance gap healing by providing a route for cells and vasculature to access the lesion, and by isolating the injury from the synovial environment, but they may also result in restriction of tendon movement, causing pain and lameness. Hence, it is impressive that the surgically created gap in the superficial digital flexor tendon within the digital flexor tendon sheath in this case became bridged with tissue without restrictive adhesion formation.

This could, therefore, be considered an example of tendon ‘regeneration’ at least on a limited functional level. Of course, this is not the same as suggesting that the bridging tissue consists of regenerated normal tendon, but it did demonstrate a similarity to tendon ultrasonographically. ‘Regenerative medicine’ has had the ‘holy grail’ of tissue regeneration as its goal for some time, which was the impetus for the first use of mesenchymal stem cells for the treatment of a tendon injury in 2003 (Smith et al. 2003). However, there has been little evidence that true regeneration has been achieved for any of the ever-expanding list of regenerative medicine products and devices that have become commercially available, even though such claims are often made. To a certain extent, this is because ‘regeneration’ is not well defined. It is important to differentiate normal repair from reformation of ‘normal’ tendon, although tendon does change with ageing. In this author’s opinion, true regeneration should be defined, for tendon at least, by the restoration of three key elements – mechanics, structure and composition – to their pre-injury state, which does give a high bar to reach. Defining these characteristics for musculoskeletal tissues is also not always simple, and a spectrum ranging from nonfunctional repair tissue to regenerated tendon probably exists. The normal healing response in tendon is via a process of scarring with the production of fibrous tissue. This tissue is collagen-rich, like tendon, but with a different ratio between collagen types I and III and is more disorganised in structure, which results in different mechanical properties. It also has higher levels of noncollagenous proteins resulting in a higher tissue content of glycosaminoglycans. However, one of the most striking differences exists at an ultrastructural level where the collagen fibril populations are universally small in scar tissue while they are bimodal (mixed population of small and large fibrils) in normal mature (adult) superficial digital flexor tendon (but not all tendons). This latter characteristic was, therefore, used in an experimental study in horses to assess the ability of mesenchymal stem cells (MSCs) to induce tendon regeneration, but MSC-treated tendons still showed the same unimodal distribution of small collagen fibres similar to controls (Caniglia et al. 2012), indicating a failure of regeneration. However, that does not mean that they cannot still exert a beneficial effect and they have been shown to improve tissue composition and organisation as well as function (Smith et al. 2013). We now believe this effect to be due to modification of the inflammatory process rather than true regeneration which still results in a clinical benefit (reduced reinjury rate; (Godwin et al. 2012) but not by the recreation of normal tendon tissue. In the case described in this edition, not surprisingly, it was not possible to retrieve any of this ‘regenerated’ tissue for a more detailed analysis to determine if true tendon regeneration had occurred. The only possible evaluation of the nature of this tissue, other than through the observation of limb function and soundness, was with ultrasound. This showed an echogenic material similar in echogenicity to tendon with at least some evidence of a degree of longitudinally arranged fibres on longitudinal views (figure 6 in the case report). We have seen this occur previously in extrathecal locations, including in a case when both digital flexor tendon had been lacerated and the gap healed with one scar involving both tendons. Over time, however, two separate tendons, capable of moving independently, were restored (M. Schramme and R. Smith, unpublished observations), showing the remarkable propensity for the restoration of tissue function in some situations.

Mechanical loading of the gap tissue is believed to be an important driver for this functional restoration. However, in the horse, such loads can be both helpful and harmful. It is important to protect the neotendon from high damaging loads when it is forming, but a gradual increase in loading also provides an important stimulus for tissue differentiation. A cast, as used in this case report initially, is the only way to unload the gap in the early stages of healing but this does not allow gradual loading. It also unloads the other weight-bearing tendons and ligaments of the distal limb which weakens them, and has recently been shown to be deleterious to the other musculoskeletal tissues of the limb (Stewart et al. 2020). Therefore, it is important to consider alternative methods of controlling digital flexor tendon (or suspensory ligament) loading. Contrary to what is perceived by the horse-owning public, distal limb bandages fail to provide any significant support to the digital flexor tendons in the adult horse (Smith et al. 2002) and a more robust system is needed. Contoured palmar or plantar splints made from casting tape are effective and can be applied to the palmar/plantar aspect of a clinical bandage. They are useful to apply immediately after removal of the cast when a bandage is still required (Smith et al. 2002; Kuemmerle et al. 2018; Fig 2). However, they are difficult to maintain, can cyclically fail, and are likely to be less effective at controlling loading than specifically designed orthotic devices. The first of these devices (EqueStride™) was developed and tested in the early 2000s (Smith et al. 2002) and is still used for cases requiring fetlock support at the Royal Veterinary College Equine Referral Hospital. The case report in this issue used a newer orthotic device, the FastTrack™ boot designed at Tufts University and manufactured in the USA, to achieve the same goal. The case report shows how these devices can be particularly useful in the rehabilitation of these severe injuries, and clinicians should consider their use as part of a structured rehabilitation programme for cases with a significant loss of fetlock support.

在本期 Equine Veterinary Education (Lenoir et al . 2022 ) 上发表的一份病例报告记录了在通常显示较差自然愈合的区域中对部分浅表指屈肌腱进行根治性手术切除。这个案例报告很有趣，因为在超声检查中，间隙充满了与肌腱相似的组织，并使马在矫形装置的帮助下恢复了健康。

由于疑似感染，肌腱切除被认为是必要的，尽管这无法通过阳性培养或组织学证实。虽然肌腱感染很少见，但它的预后确实很差（Kidd et al . 2002），因此有理由进行彻底的清创。然而，这在浅表指屈肌腱的两端之间留下了很大的间隙。马肌腱裂口愈合可见于肌腱撕裂后形成桥接纤维“愈伤组织”（Jordana et al . 2011）。然而，这可能是一个漫长的过程，最终可能会因限制性粘附形成而变得夸大。也可能发生肌腱末端的反射，这会进一步延迟愈合，因此，已经进行了许多帮助愈合的尝试，特别是对于指屈肌腱，因为它们在正常负重肢体功能中的重要性。这些方法包括使用各种生物和合成植入物来弥合差距。合适材料的选择已经从不可拉伸的碳纤维 (Vaughan et al . 1985 ) 到人工不可吸收材料 (Gibson et al . 2002 ; Barrett et al . 2014 ; Fig 1)、长效可吸收材料 (Eliashar et al . 2001 ; Jenson et al . 2005 )，甚至是自体肌腱组织 (Valdes-Vazquez et al . 1996)）。假定它们的作用是充当支架，以提供通道，细胞和正在愈合的脉管系统可以沿着该通道生长，并保持肌腱末端对齐。这与在较小的动物中使用脚手架形成对比，因为所涉及的力较小，它们可以充当坚固的机械“桥梁”。这在马身上是不可能实现的，事实上，我们过去曾观察到，尤其是在术后用远端肢体石膏支撑的后肢中，该后肢固定了蹄锁，但允许飞节和膝关节屈曲，植入物已经成为从横断的肌腱末端分离。这一观察结果表明，适当大小、持久但可吸收的植入物的可用性较差，并且在已发表的少数有限研究中未能显示出比单纯清创术更显着的益处，这意味着植入物现在很少使用。因此，临床医生依赖于简单的清创和自然修复，同时使用外部接合保护该区域免受负荷，如本病例报告中所述。

然而，在这个案例报告中令人惊讶的是，这种间隙愈合发生在一个对愈合特别成问题的区域。足底到 fetlock，肌腱受到压缩和拉伸，需要更复杂的细胞外基质。此外，它的鞘内位置意味着肌腱被滑液包围并且缺乏腱旁。我们最近确定滑液对肌腱内的细胞具有毒性作用，这有助于解释为什么许多与腱鞘、滑囊或关节的滑膜腔相通的肌腱和韧带病变通常无法自发愈合（Garvican et al . 2017）。腱旁被认为是外在（来自周围组织）愈合的主要参与者，至少在实验动物模型中，这被认为主导鞘外肌腱修复（Kajikawa et al . 2007）。腱旁与内腱或束间基质是连续的，这被认为是内源性肌腱干细胞和新血管系统的来源（Cauvin 2000; 马尔等人。2017）。然而，鞘内位置缺乏腱旁，意味着在没有粘连形成的情况下，修复更多地依赖于内在（来自肌腱内）愈合，这种愈合效果较差且较慢。粘连将通过为细胞和脉管系统提供进入病灶的途径以及通过将损伤与滑膜环境隔离来增强间隙愈合，但它们也可能导致肌腱运动受限，从而引起疼痛和跛行。因此，令人印象深刻的是，在这种情况下，在指屈肌腱鞘内的浅表指屈肌腱中通过手术产生的间隙与组织桥接，而没有形成限制性粘连。

因此，至少在有限的功能水平上，这可以被认为是肌腱“再生”的一个例子。当然，这与暗示桥接组织由再生的正常肌腱组成并不相同，但它确实在超声检查中证明了与肌腱的相似性。一段时间以来，“再生医学”一直将组织再生的“圣杯”作为其目标，这也是 2003 年首次使用间充质干细胞治疗肌腱损伤的动力（Smith et al . 2003）。然而，几乎没有证据表明已经实现真正再生的再生医学产品和设备清单不断扩大，这些产品和设备已经上市，即使经常提出这样的声明。在一定程度上，这是因为“再生”没有得到很好的定义。重要的是区分正常修复和“正常”肌腱的改造，尽管肌腱确实会随着年龄的增长而改变。在这位作者看来，真正的再生应该被定义为，至少对于肌腱，通过恢复三个关键要素——力学、结构和组成——恢复到他们受伤前的状态，这确实给了很高的门槛。为肌肉骨骼组织定义这些特征也并不总是简单的，并且可能存在从无功能修复组织到再生肌腱的范围。肌腱的正常愈合反应是通过产生纤维组织的瘢痕形成过程。这种组织像肌腱一样富含胶原蛋白，但 I 型和 III 型胶原蛋白的比例不同，结构更加杂乱无章，从而导致不同的机械性能。它还具有更高水平的非胶原蛋白，导致组织中的糖胺聚糖含量更高。然而，最显着的差异之一存在于超微结构水平，其中胶原原纤维群体在瘢痕组织中普遍较小，而在正常成熟（成人）浅表屈肌腱（但不是所有肌腱）中它们是双峰的（大小纤维的混合群体） ）。因此，后一种特征被用于马的一项实验研究，以评估间充质干细胞 (MSCs) 诱导肌腱再生的能力，但 MSC 处理的肌腱仍显示出与对照组相似的小胶原纤维的单峰分布 (Caniglia等人。2012），表示再生失败。然而，这并不意味着它们仍然不能发挥有益作用，并且它们已被证明可以改善组织组成和组织以及功能（Smith等人，2013 年）。我们现在认为这种效应是由于炎症过程的改变而不是真正的再生，这仍然会带来临床益处（降低再损伤率；（Godwin et al . 2012) 但不是通过恢复正常的肌腱组织。在本版中描述的案例中，毫不奇怪，不可能检索任何这种“再生”组织进行更详细的分析，以确定是否发生了真正的肌腱再生。除了通过观察肢体功能和健全性之外，对该组织性质的唯一可能评估是使用超声波。这显示了一种回声材料与肌腱的回声相似，在纵向视图上至少有一些证据表明纤维纵向排列有一定程度（病例报告中的图 6）。我们之前已经在鞘外位置看到过这种情况，包括在两个手指屈肌腱都被撕裂并且间隙愈合并且涉及两个肌腱的疤痕的情况下。然而，随着时间的推移，两条独立的肌腱，

间隙组织的机械负荷被认为是这种功能恢复的重要驱动力。然而，在马身上，这样的负荷既有益又有害。在新肌腱形成时保护它免受高破坏性负荷很重要，但负荷的逐渐增加也为组织分化提供了重要的刺激。最初在本病例报告中使用的石膏是在愈合早期阶段卸载间隙的唯一方法，但这不允许逐渐加载。它还会卸载远端肢体的其他负重肌腱和韧带，从而削弱它们，并且最近已证明对肢体的其他肌肉骨骼组织有害（Stewart et al . 2020）。因此，重要的是要考虑控制指屈肌腱（或悬韧带）负荷的替代方法。与拥有马的公众的看法相反，远端肢体绷带无法为成年马的指屈肌腱提供任何显着的支撑（Smith等人，2002 年），因此需要更强大的系统。由铸带制成的轮廓手掌或足底夹板是有效的，可应用于临床绷带的手掌/足底方面。当仍然需要绷带（史密斯它们去除投后立即申请实用等人。2002 ; Kuemmerle等人。2018 ;图 2）。然而，它们难以维护，可能会周期性地失效，并且在控制负荷方面可能不如专门设计的矫形装置有效。这些设备中的第一个 (EqueStride™) 是在 2000 年代初开发和测试的 (Smith et al . 2002) 并且仍然用于在皇家兽医学院马科转诊医院需要胎儿锁支持的病例。本期的案例报告使用了一种较新的矫形装置，即塔夫茨大学设计并在美国制造的 FastTrack™ 靴，以实现相同的目标。病例报告显示了这些设备如何在这些严重损伤的康复中特别有用，临床医生应考虑将其作为结构化康复计划的一部分，用于严重丧失胎锁支持的病例。