Management of Complications of Flexor Tendon Surgery

1 Management of Complications of Flexor Tendon Surgery

Peter C. Amadio and Duncan Angus McGrouther


Adhesions are the most common and troubling complication after flexor tendon repair in the hand. Aside from small children, essentially every tendon repair is complicated by adhesions that limit motion; the only real question is whether these will be severe enough to require surgical intervention in an attempt to improve function. Pulley loss can be the result either of direct injury or a decision by the surgeon to trim pulleys to improve motion. Usually, if the contiguous loss is less than 2 cm, reconstruction is not needed. Management of flexor tendon repair rupture will be focused by the history, exploration, and forensic examination of the wound to establish causation and identification of preventable factors. Key technique changes for primary or re-repair have been the move toward stronger repairs by multistranded core suture configurations, better anchorage points, and pulley release. Most therapy regimes aim at active mobilization and this is also indicated for a re-repair as the stuck tendon is vulnerable to rupture from later mobilization. The key to managing infection lies in understanding how the time-honored surgical principles of drainage, decompression, and dilution fit into modern understanding of inflammation and bacterial virulence factors. Infection in a tendon repair is uncommon and usually due to inadequate debridement or comorbidities. It is best managed by antibiotics, drainage, and catheter irrigation. If the tendon is necrotic, it should be excised and reconstruction considered.

Keywords: Keywords: tendon, adhesion, bowstringing, pulley, rupture, infection

1.1 Management of Flexor Tendon Adhesions

1.1.1 Definition/Problem: Current Understanding of the Biology of Adhesion Formation

The observation “one wound, one scar” as it pertains to tendon repair is nearly as old as the specialty of hand surgery itself, and of course is a truism that reflects wound healing biology for nearly all tissues in nearly all mammalian species beyond the fetal stage. The functional implications and severity will vary depending on location, mechanism of injury, and the specific tissues that are injured. The flexor tendons in the hand are especially at risk for the negative consequences of this dictum—the gliding tendon fixed in place by scar to adjacent pulley and bone, limiting tendon excursion and impairing function.

The unique anatomy and nutritional arrangement of the flexor tendons in the hand particularly predispose them to adhesion formation because, even normally, the nutritional supply is precarious. The flexor profundus excursion is around 2.5 cm in the adult finger. This long excursion is made possible by a special nutritional arrangement. Instead of a circumferential paratenon to supply nutrients to the tendon, the blood supply of the finger flexors in the fingers is segmental, through the vincula, which arise from the digital arteries at the joint level, and enter the tendons through their dorsal surfaces (Fig. 1‑1). The feeding vessels enter just lateral and anterior to the bone, just proximal and distal to each joint. These feeding vessels must be carefully protected during dissection at the time of repair or tenolysis. If they are cut, even with a physically intact vinculum, the tendon will be effectively devascularized. To supplement this nutritional source, the tendons are surrounded by a synovial sheath also, so that synovial diffusion can provide nutrition as well. Both systems are commonly injured when the flexor tendon is lacerated, the synovial sheath by the same injury that injured the tendon, and the vincula either by that mechanism or by rupture with muscular contraction, pulling the proximal tendon stump out of the finger.

Fig. 1.1 (a-c)Blood supply of the flexor tendons. SV, short vinculum; LV, long vinculum; DA, digital artery.

Unsurprisingly, this loss of nutrition has consequences, and one of the major drivers of adhesion severity is vascularity. Well-vascularized tendons have better motion than poorly vascularized ones, strongly suggesting that a good tendon blood supply is an important factor in reducing adhesion severity.1 This is not only true for severely devascularized tendons, as occurs with amputation/replantation, but also with damage to the vincular system in an otherwise well-perfused finger. Avascular or hypovascular tendons, like any other vascularly impaired tissue, will release cytokines such as vascular endothelial growth factor (VEGF) that will stimulate neovascularization and new vascular ingrowth into the tendon. These new vascular connections, occurring in parts of the tendon normally nourished either by synovial fluid or the vincular system, do good in restoring nutrition to the tendons and aiding tendon healing, but at the same time do harm by binding the tendon to the surrounding tissues and limiting tendon motion. Usually, unfortunately, there is little that can be done to reverse this aspect of the initial injury—though as noted in the next section, there are some things that can be done to try to minimize the impact of tendon hypovascularity on tendon motion.

A second anatomical feature predisposing finger tendons to adhesions is the fibro-osseous sheath, which holds the tendons close to bone and allows the tendon excursion to drive a remarkable 270 degrees of combined active motion of the finger joints. These narrow confines can easily limit gliding of even the smoothest tendon repair, and provide an extremely short leash for any adhesions that do form. However, unlike vascularity, whose loss is currently irretrievable, there are some things that the surgeon can do to mitigate the impact of the sheath on adhesion formation, as discussed below.

Other anatomic factors predisposing to adhesions relate to associated injuries, which may affect tendon vascularity at a distance, require immobilization or otherwise compromise the physical aspects of tendon rehabilitation (fracture, nerve repair, proximal limb injury), or otherwise limit patient ability to participate in rehabilitation (polytrauma, head injury).

The second main group of factors affecting adhesion formation represents those under the control of the surgeon. These include tendon handling, the tidiness of the repair, decisions regarding pulley preservation, and the important question of postoperative rehabilitation. All these areas have seen notable advances in the past decade. The result has been an important reduction in the severity of postoperative adhesions after tendon injury, and the consequent need for tenolysis. These are discussed below.

1.1.2 Treatment

Surgical Techniques to Minimize Adhesions

It has been known for many years that rough handling of the lacerated tendon in Zone 2 can worsen adhesion formation, by disrupting the smooth gliding surface of the tendon. For this reason it is important to handle the tendon gently, and grasp it only by the cut end. If the tendon has retracted into the palm, it should be retrieved with some sort of tendon carrier or passer, such as a narrow catheter. Any vincula that have survived the initial injury should be carefully preserved; indeed, if immediate repair is not possible then the finger should be splinted in wrist and finger flexion, to minimize the risk of rupturing any remaining vincula, until such time as surgical exploration can proceed. Active motion of the affected digit or digits should also be discouraged during this time.

A tidy repair, with the tendon ends coapted with slight bunching and normal rotational alignment, is critical to the smooth passage of the repaired tendon beneath any pulleys that are preserved. A tidy repair should also be a “low profile repair,” with the least possible amount of suture material on the anterior surface of the tendon. Knots and even suture loops are sources of friction that will initially score the overlying pulley (Fig. 1‑2), and later this scoring will lead to inflammation and adhesions.

Fig. 1.2 Pulley scoring from a tendon suture knot in an animal model of tendon repair.

Initially, pulleys were sometimes resected to allow room for the tendon repair, only to result in bowstringing and flexion contracture. This clinical problem will be discussed in another section of this chapter. To avoid this problem, for many years there was a strong emphasis among hand surgeons to preserve the pulleys, and even to close the sheath completely. Unfortunately, this too led to adhesions and limited tendon gliding, even with well-performed, low-profile, tidy repairs. The problem was that even the best repair could not reproduce the dimensions of an intact tendon, and even a cursory examination of the tendons as they slide under the A2 pulley will confirm that there is no room for any additional bulk at all. Thus, more recently there has been a push again for judicious pulley resection, including, if need be, all the A4 pulley and even part of the A2 pulley.2 Usually the amount of bowstringing after A4 excision is modest, because of the short segment affected (essentially, the length of P2), and the modest concave curvature of the palmar P2 surface. This is true even if, as is often the case with Zone 2 injury, the A3 pulley is not intact. In contrast, the P1 segment is longer, and its concavity deeper, resulting in more important bowing with A2 loss, especially if this is associated with A3 loss. Thus, it is important to preserve at least half of A2. An alternative strategy, which we prefer, is to resect (or excise rather than repair) one slip of the flexor digitorum superficialis (FDS), which creates adequate gliding space under the A2 pulley for even a somewhat bulky flexor digitorum profundus (FDP) repair.3

One may ask, how tidy is tidy enough, or how can one tell if enough pulley has been resected? Here the answer is clear. “Wide-awake” surgery, which allows the patient to actively move the tendon after it is repaired, but before the wound is closed, will reveal any deficiency in technique, including not only catching of a tendon repair on a pulley edge (Fig. 1‑3) but also a sloppy knot that unravels unexpectedly!4

Fig. 1.3 Even the most tidy repair can catch on a pulley edge. The best solution is to make room for the repair, either by pulley excision or resection of one slip of the FDS.

Adhesion Barriers

In addition to techniques of surgical approach and tendon repair, hand surgeons have often attempted to reduce the risk of adhesion formation through the use of various lubricant and adhesion barriers. In theory, these can make useful adjuncts, and, in practice, where there is room for them, protective sheets and films can block adhesions from, for example, an underlying fracture and a tendon repair. Unfortunately, there are two main problems that have limited the usefulness of such barriers: first, their bulk often precludes using them just where they are needed most, between a tendon and its pulley; and second, because, as noted above, adhesions are a quite appropriate physiological response to bring blood supply to a damaged tendon. Thus, blocking adhesions often also means a persistently ischemic tendon, and one more likely to fail by rupture, as covered elsewhere in this chapter.

Liquid lubricants, such as hyaluronic acid, have also been used to block adhesions. These fluids, typically characterized as devices rather than drugs because their effect is mechanical and does not affect cellular processes, may at least allow the diffusion of nutrients, if not vascular ingrowth. In most cases though, these lubricants are either metabolized away or simply moved away by tendon motion, and therefore provide little benefit over the course of tendon healing. Lubricants that can be chemically bonded in molecularly thin layers to the tendon surface may one day overcome these problems, and have shown promise in animal models, but not yet in humans.

Drugs have also been used to block or mechanically weaken adhesions, most notably in the past, beta-amino-propionitrile, which interferes with collagen crosslinking. Unfortunately, it has not been possible to simply block crosslinking in the adhesions and not, for example, in the nearby tendon laceration or overlying skin. Thus these drugs end with unacceptable wound-healing complications; when used in experimental animals over longer terms, even the collagen in large vessels can be affected, leading to catastrophic hemorrhage.


As noted above, advances in tendon surgery over the past two decades have reduced the need for tenolysis considerably, from over 20% in the 1970s and 1980s to under 10% today. Nonetheless, tenolysis is still needed for some patients, and the indications remain as they have always been: functionally disabling loss of active tendon motion in the presence of a supple finger (i.e., a mismatch between active and passive motion), an intact, healed tendon with a normal proximal muscle, and a cooperative patient. If the finger is not supple, there is no mismatch and thus no motion to be gained; if the tendon is not healed or the muscle is damaged then the solution is tendon reconstruction, not tenolysis; and of course without patient cooperation during postoperative rehabilitation the whole effort will be unrewarding.

A second factor in considering tenolysis is the risk/reward equation. What is the current functional limitation, and how bothersome is it? Is it worth jeopardizing current function for a chance at improvement? Neither full extension nor full flexion of a PIP joint is a realistic goal. Indeed, in general only about half of any intraoperative gain in active motion is maintained after tenolysis, so that, too, must be considered when determining if the procedure is worth considering.

In the preoperative assessment of a patient for tenolysis, it is important to determine if any joint contractures can be resolved with splinting, stretching, or serial casting preoperatively. It is also important for the surgeon to know that joint contractures can be the consequence of major pulley loss; in cases where pulley reconstruction and tenolysis are both under consideration, the surgeon should consider instead a staged reconstruction, since pulley reconstruction requires prolonged protection, while tenolysis demands early active mobilization postoperatively, to avoid the re-establishment of adhesions. This is especially true if a raw, tenolysed tendon surface is in contact with a reconstructed pulley. Recurrent adhesions are nearly inevitable. Finally, the surgeon should also verify that the tendon is healed—in addition to clinical assessment, ultrasound or magnetic resonance (MR) imaging is useful to confirm tendon integrity preoperatively, and they can also provide useful information on pulley loss.

Tenolysis is a procedure which carries substantial risks. The tendon and pulley may be deeply entwined in scar. Stray away from the tendon, and pulley loss may condemn the patient to a severe flexion contracture; get too close to the tendon and it may rupture. Because of this risk of intraoperative rupture, any patient consented for tenolysis should also be consented for a possible staged tendon reconstruction as well. At the very least, a plan for dealing with intraoperative rupture should be discussed in detail with the patient preoperatively. Tenolysis is an operation where there is a realistic probability that the result could be worse postoperatively.

The literature on the results of tenolysis is not particularly robust, likely reflecting the wide variety in the underlying cases, as well as the often mediocre results, which may make surgeons less enthusiastic about reporting their results.5 All emphasize that one should expect the final result to be less good than whatever intraoperative gain is noted, and that tendon rupture may occur during or immediately following tenolysis.

1.1.3 Rehabilitation

After tenolysis, early active mobilization is important, but it is also wise not to be aggressive in the first few days, so as not to provoke any bleeding, and to minimize edema. In my experience, a close working relationship between the patient, the hand surgeon, and an experienced hand therapist is essential to a good clinical result after tenolysis. Each patient’s care must be customized to the specifics of the underlying pathology and the details of the tenolysis that was performed. Edema control, passive joint mobilization, and graded active motion are all key elements of a successful rehabilitation program.

Early active flexion, usually through a “short arc” of partial motion that gradually increases over time postrepair, has generally replaced most passive flexion/active extension programs, which were popularized in the 1970s and 1980s. Passive motion and edema control measures still have an important role to play, to maintain a supple soft tissue sleeve and supple joints, but they, in and of themselves, do little to induce tendon gliding, even with various synergistic motion maneuvers. Protective splinting has also changed, as it has become clear that wrist immobilization does little to unload the flexor tendons in the fingers beyond what finger flexion can achieve. Thus, more and more, splinting after flexor tendon repair is hand based, at least after the initial surgical dressing is removed.

Together, the good fortune of a well-vascularized tendon injured by a sharp laceration, combined with modern methods of tendon repair, pulley release, and early active motion therapy, have gone a long way toward eliminating the need for tenolysis, which is often reserved now for complex combined injuries, such as replantation, where the anatomic ground and postoperative milieu are unfavorable.

1.1.4 Tips and Tricks

Each tenolysis case is in many ways unique, but there are still common features of tenolysis that allow the discussion of general principles. One of the most important and fundamental principles is that, since the goal is to restore active motion, it is important to assess active motion during the procedure. For this reason I prefer, and recommend to my patients, the “wide-awake” approach, without sedation, a tourniquet, or muscle-paralyzing agents to interfere with active contraction of the affected muscle tendon units. In many cases, an active contraction will rupture the final adhesions and allow the desired restoration of motion. In other cases, a traction check with the surgeon putting tension on the tendon will show full passive tendon motion through the zone of injury, but no active result when the patient attempts to flex the digit. This is a sign that adhesions may have formed outside the zone of injury. These can be distal, but proximal adhesions can also develop, even as far proximal as the distal forearm, between the profundus or superficialis muscle bellies.

Skin quality is important to surgical planning as well. A scarred digit with atrophic skin is unlikely to tolerate an extensive dissection, and thus represents a contraindication to tenolysis. Supple overlying skin is essential. Dissection must preserve whatever circulation remains, to both the digit and tendon. I prefer to use existing incisions, and to extend them proximally and distally as needed to get areas beyond the zone of injury, so that dissection can begin where anatomy is normal, and move from there into the zone of injury.

When exposing the tendon and sheath, it is important to avoid any further injury to the neurovascular bundles, or to their branches that feed the tendon and sheath, as noted in Fig. 1‑1. The extent of remaining pulleys should be noted. To preserve these pulleys, dissection under them can be done through transverse windows, no closer than 1 cm apart. A variety of special narrow knives and elevators can be used for this purpose; however it is done, it is important to preserve tendon integrity. Just as with digging a tunnel under a mountain from both sides, coaxial alignment of the work is essential, so that when the proximal and distal dissections join they still define a robust tendon in continuity. In some cases with severe adhesions it may be necessary to sacrifice the superficialis tendon and preserve the profundus tendon, especially with severe adhesions under the A2 pulley.

For wound closure after tenolysis, hemostasis is another critical factor; another reason to favor the wide-awake approach, since the epinephrine that allows prolonged local anesthesia also provides excellent hemostasis. The skin must be closed carefully and without tension, so that early motion will not jeopardize wound healing.

1.1.5 Conclusions

In summary, tendon adhesions are an important but fortunately increasingly infrequent complication following flexor tendon injury. When tenolysis is required, care is necessary to be sure that the potential benefits of surgery outweigh the risks. Surgery using wide-awake anesthesia can facilitate complete release of the limiting adhesions. Postoperatively, a coordinated team of patient, therapist, and surgeon is required to optimize outcomes.

1.2 Bowstringing

1.2.1 Definition/Problem

Tendon bowstringing, strictly speaking, occurs any time a tendon loses its close contact with the underlying bone. This is a particular aspect of the anatomy of finger flexor tendons, because of the curved anterior surface of the phalanges. Without the bony A2 and A4 pulleys, the flexor tendons will naturally bow away from the curved phalanx, regardless of joint angle.6 This is more so for the proximal phalanx, because it is both more bowed and longer than the middle phalanx. Bowstringing is exacerbated with joint flexion if the joint related A1, A3, or A5 pulleys are also affected. Basically, the longer the affected segment the greater the problem. Of note, the palmar condyles of the phalanges create a palmar bowing of the tendon at the joint so that the path of the tendon at the proximal interphalangeal (PIP) joint, as seen from a lateral view, does not actually become straight until the joint is flexed about 45 degrees; joint related bowstringing, even in the absence of pulleys, only occurs after this point. The more joint motion is restricted, then, by arthritis, intra-articular injury, or periarticular fibrosis, the less bowstringing is actually possible, and, by extension, the less critical it is to restore pulley function. The converse though is also true: the more a patient requires full flexion at the PIP joint to perform important functions, the more important it is to have an intact pulley system.

Why Do We Have Pulleys?

Pulleys are necessary to keep the tendon close to the bone, especially in places where large angular joint motion is needed. Tendon excursion is fixed, based on the ability of actin and myosin molecules in the muscle to contract. To get more angular motion with a fixed excursion, the tendon must be as close to the axis of motion of the joint as possible. Remember that the circumference of a circle is two times the radius, so if a rope (or tendon) is turning a circular joint with a radius of 5 mm, for every 5 mm of tendon excursion the joint angle will move 360/2 degrees, or about 60 degrees. If the tendon is bowing away from the joint by an additional 5 mm, then the same 5 mm excursion will only move the joint around 30 degrees. Because of the increased lever arm (10 mm vs. 5 mm) the tendon will exert twice as much force on the joint, but the joint will only move half as far. This is the tradeoff in pulley preservation or reconstruction, and, by the way, underlines one of the beauties of pulley anatomy—the A2 and A4 pulleys are fixed, but the A1, A3, and A5 pulleys connect to the volar plates, so they naturally allow some bowstringing with joint flexion, reducing potential joint angular motion a little but adding some strength. A very important point that we did not fully appreciate earlier in my career is that while normal pulley anatomy is necessary for perfect finger motion, in essentially all cases of pulley loss we are not realistically trying for normal. Usually a bit less motion and a bit more strength is a fair tradeoff for some residual bowing, and of course in the context of flexor tendon laceration normal motion is anyway almost never achieved in adults—the loss of a few degrees of active motion due to bowstringing caused by pulley trimming is well worth if the alternative is a stuck tendon, and no motion at all. This imbalance will nearly always result in a flexion contracture of some degree, but unless pulley loss is extensive, again, this may well be an acceptable tradeoff, compared to striving for perfect reconstruction, with its attendant risks of adhesions and stiffness.

Etiology of Pulley Loss

Pulley injury may be closed or open, and open injury may be due to direct trauma or be the result of a surgical decision to trim or vent pulleys to aid in tendon repair or tenolysis. Closed injuries are frequently the result of forceful gripping, and are a common problem in rock climbers. Usually these are small (distal A2, isolated A3, or A4) and can be managed with taping and early return to sport (Fig. 1‑4). Even a complete A2 rupture, if isolated, can often be managed without surgery, using custom thermoplastic ring splints that allow some functional activity. Even in idealized cadaver models, a complete loss of A2 only results in a loss of around 20 degrees of motion, for a normal FDP excursion and normal joints. Such a loss is far smaller than the typical loss of motion, compared to normal, after a tendon repair in Zone 2. Ruptures of multiple pulleys, though, will require surgery. The situation is the same with open pulley injuries—short segments of loss can be left unrepaired or reconstructed, but longer segments will require surgical attention, because the consequence otherwise is marked bowstringing and, usually, a severe flexion contracture. As noted above, the length of pulley loss that can be tolerated without important functional impairment depends upon both patient needs and the mobility of the PIP joint, but there will be few patients who can tolerate combined loss of A2, A3, and A4 without substantial loss of function, usually due to a fixed flexion contracture, as the power of the flexors, augmented by the leverage provided by the bowing, overcomes the ability of the extensors to counterbalance, and extend the joint.

Fig. 1.4 A patient with rupture of the distal A2 pulley has a small but obvious amount of bowing (arrow). This amount of bowing is unlikely to require pulley reconstruction. Nonoperative methods are likely to suffice. Note however that the space posterior to the tendon (star) is not empty—it will fill with a wedge of scar, which must be excised before the tendon can be returned to its proper location adjacent to the phalanx, and an appropriate diameter pulley reconstruction performed.

1.2.2 Treatment

Nonoperative Treatment

As noted above, short segments of pulley loss (typically < 1 cm), whether due to rupture, laceration, or surgical trimming, can be safely managed by fairly simple means. Even a loss of the entire A4 pulley can be managed in this way, provided that the A3 and A5 pulleys remain intact. Schoffl and Schoffl7 have developed a useful classification of pulley injuries in climbers, which is relevant to all pulley injuries. Grade 1 is a sprain, with local tenderness but no loss of pulley integrity as demonstrated by either ultrasound or MR imaging. These can be managed symptomatically, and do not require immobilization. Taping the affected phalangeal segment and a graduated return to full function over the next few months is usually all that is required. Grade 2 injuries represent the shorter segment complete injuries (<1 cm) mentioned above, either partial A2, or A3, or A4, which can be managed similarly, with the addition of a few weeks of initial immobilization to rest the injured area. Grade 3 is basically a complete loss of the A2 pulley, which is around 15 mm long in the adult finger. Schoffl and Schoffl include also an isolated loss of A3 as a grade 3 injury, but we have not seen an isolated loss of A3 to cause clinical difficulties in my own experience, and again it is important to remember that A3 does not even come into play functionally until the PIP joint is flexed beyond 45 degrees. For isolated complete A2 injuries, taping is not sufficient to restrain bowstringing until the pulley heals, and in such cases a custom thermoplastic ring splint may be helpful. More prolonged protection is needed before returning to full use, usually 3 months or more, and the thermoplastic splint may be needed during activity for many months after that. Nonetheless, many such patients can function well without the need for pulley reconstruction.

Surgical Treatment

In contrast to short segments of pulley loss, grade 4 injuries, with complete loss of adjacent pulleys, will often require surgical reconstruction, tempered somewhat by the caveats mentioned above (Fig. 1‑5).

Fig. 1.5 Loss of multiple pulleys is more likely to require reconstruction. (a) A patient with massive pulley loss, undergoing staged tendon and pulley reconstruction. (b) The most important pulley to reconstruct is the A2 pulley, and for this pulley, a tendon graft wrapped around the phalanx three times is the strongest construct. This also creates more pulley width. This is my preferred method of reconstruction at this level. A palmaris longus tendon is preferred, if available. Because pulley reconstruction is usually done in the context of tendon reconstruction, remnants of the flexor tendons can also be used, to preserve the palmaris longus for use as a later tendon graft.

Only gold members can continue reading. Log In or Register to continue

Stay updated, free articles. Join our Telegram channel

Apr 6, 2024 | Posted by in ORTHOPEDIC | Comments Off on Management of Complications of Flexor Tendon Surgery

Full access? Get Clinical Tree

Get Clinical Tree app for offline access