Staged tendon grafts and soft tissue coverage.

The objective of the two-staged flexor tendon method is to improve the predictability of final results in difficult problems dealing with tendon reconstruction. This article reviews the evolution and benefits of this procedure. It also considers the use of the technique to help deal with problems requiring pulley and skin reconstruction simultaneously with re-constituting the flexor tendon system.

INTRODUCTION

The ideal treatment of flexor tendon injuries under almost every circumstance is primary repair. I believe this to be true whether the injury is a simple laceration of the finger or a replantation, as is easier to rehabilitate a finger over several months following injury with the finger flexing actively than being flexed passively. However, this is not always possible. A percentage of primary repairs may also either become adherent to their surrounds, gap and become secondarily adherent to their surrounds or rupture. All of these circumstances may lead to a need for secondary flexor tendon grafting, with or without pulley reconstruction and skin replacement over the tendons. Although tendon grafting may give comparable results to primary flexor tendon repair in the hands of some experts, general opinion would support the view that primary repair is less difficult for most surgeons and more likely to be more successful, even if this has never been proved by comparative trial. For most of us, tendon grafting is more difficult, therapy is more vital to success and the results are usually less near to normal. The reported results are also generally worse than those of primary surgery.

Wherever one practices, the patients included in the group ‘secondary flexor tendon surgery’ will include some of your ‘worst’ injuries and ‘worst’ patients. These cases are more likely to do badly whatever one does and may do no better after secondary surgery than they did the first time, if they have already undergone the primary surgery. Unfortunately, conventional one-stage secondary flexor tendon surgery is not always followed by satisfactory return of flexor function. Two-stage flexor tendon grafting was introduced with the intention of achieving better results under circumstances where the likelihood of poor results can be identified. The perceived advantage over single-stage grafting is that mobilisation of the tendon graft following the second operation is started with the graft moving in a smooth-walled pseudo-sheath, created by the silicone rod over several months after insertion at a first operation, and in a less traumatised, less painful and more supple hand. Despite a very small incidence of reaction to silicone rods and a small incidence of infection and/or extrusion of the rods, the two-stage procedure appears at least to reduce the influence of the ‘scarring’ factors on the result.[1]

HISTORY OF TWO-STAGE TENDON GRAFTING

The primary, and most significant, problem of flexor tendon surgery is the formation of adhesions between injured tendons and their surrounds, particularly the tendon sheath within the digits and the distal palm. Overcoming adhesion of repairs to the sheath has been the driving force for most research in this field over the last 100 years. It has been the impetus to the development of internal suturing of the tendons, to increasing the strength of this suturing, to the introduction of earlier and increasingly more aggressive rehabilitation, and to various attempts to diminish the possibility of adhesion formation. Mayer and Ransohoff (1936), writing in the era when delayed flexor tendon grafting was the routine,[2] described how ‘these adhesions extend from the point of division of the tendons down to the distal end of the digital theca (tendon sheath). They destroy the smooth cells on the surface of the tendon and the sheath. The gap between the tendon ends is filled with scar tissue which replace the tendon sheath. It is obvious that the normal gliding mechanism of the tendon has been completely destroyed and that, consequently, the conditions for re-establishing free gliding of a transplanted tendon are so unfavourable that only in exceptional instances can an implanted tendon perform its normal function.’ Mayer was the first surgeon to try to re-establish the milieu required for free gliding of the tendon within the sheath by insertion of an inert rod, and then, at a later stage, replace this with a tendon graft. He used celloidin tubes but had to abandon his experiment to create a ‘pseudosynovial’ sheath as these were too rigid. Twenty-five years later, Bassett and Carroll (1963) repeated the work using silicone rods.[3] James Hunter carried this work forward through the 1960s and 1970s to establish the technique clinically.[14-10] The successors to the rods of woven Dacron covered by silicon developed by him for use in the staged flexor tendon grafting are known to us all today simply as ‘Hunter’ rods. Curiously, Hunter's first use of a silicone rod was to replace an extensor tendon. The rod was sutured to the extensor tendon with a fine wire. It became clear that the rod was being stretched and acting as an elastic band to pull the flexed finger back into extension when the flexor tendon relaxed. Had the wire suture not protruded through the skin, there would have been no second stage. On exploration of the protruding wire, it was clear to Hunter that a shiny mesothelial-lined membrane had formed around the rod. The rod was replaced by a tendon graft and the finger went through rehabilitation to achieve the acceptable function. Hunter's initial intention was to design an artificial tendon for permanent use and he pursued both the concept of the pseudosheath and the idea of a permanent artificial tendon replacement in his research. While the former has become part of the armentarium of most hand surgeons, a permanent artificial tendon remains ‘experimental’, largely because of the problems of achieving a permanent bond between the rod and the biological tissues to which it must be attached proximally and distally.

THE PSEUDOSHEATH

After implantation, the silicone rod holds the adjacent tissues apart. Microscopic studies show that the pseudosheath is not simply a tube of scar.[11] The tissues adjacent to the rod organise over a few days into a mesothelial-like layer. In the chicken experiments described by Salisbury and his colleagues, this ‘intima’ consisted predominantly of one layer of flattened fibroblasts with a surface made irregular by pleats and pores, whereas the normal tendon sheath has an intima of several layers of thicker, cuboidal cells. The new intima also contained macrophages, intermediate cells, mast cells and Schwann cells. By 4–6 weeks, the deeper tissues have formed a second layer of loose, well-vascularised connective tissue and the pseudosheath has a basic appearance similar to normal synovial sheath. The collagen in the outer layer was of variable thickness and contained collagen, reticulum, elastic and unmyelinated nerve fibres. The collagen was orientated along the axis of the implant much as it is orientated along the lines of stress in a tendon sheath.

CHOICE BETWEEN SINGLE-STAGE AND TWO-STAGE GRAFTING

Although much of the flexor tendon surgery in Western Europe is primary surgery, we do have to undertake a surprising amount of secondary flexor surgery, some of which is because of delayed presentation and some of which is because the flexor tendon injury quite frequently occurs in association with injuries to other tissues. Nevertheless, our secondary surgery is mostly that of the complications of primary repair, namely ruptured and adherent primary repairs. In the best units, the failures constitute about 10% of all primary repairs, not a small number and something of an indictment of our present techniques of primary flexor tendon surgery. The cases which come to secondary surgery are mostly either the result of more severe injuries or have occurred in patients who make excessive amounts of scar tissue or have not co-operated with therapy because of low pain thresholds, social circumstances or stupidity.[12] So the cases needing secondary surgery can be considered under the headings of either ‘bad injuries’ or ‘bad patient’. Elsewhere in the world, many patients will only get to an appropriate surgeon at a time when secondary repair of the tendon using grafts is the only option because of proximal tendon retraction. This, by definition, becomes ‘secondary flexor surgery′, although the problem is simply an extended finger with good passive but no active flexion, but now no longer amenable to direct repair. However, among the patients presenting after delay are a group with much more complicated problems, sometimes as a result of injuries to the other structures of the digits and, sometimes, as a result of the unaided healing process within a digit in the presence of an inactive flexor system. The problems in these cases are not simply those of being unable to get the tendon ends together. So, in terms of the pathologies in the digits themselves, the problems we all face in the rather heterogeneous mix of cases which we call ‘secondary flexor tendon surgery’ are not so different, namely flexor tendons which are not intact and flexor tendons which are stuck in scar tissue, variably associated with divided pulleys, skin deficit on the flexor aspect of palm and digits, stiff fingers and injuries to other, adjacent structures in the fingers, hand or forearm.

Whether we should be carrying out single-stage or two-stage tendon grafting, or a mixture of both, depending on the individual case, is a matter of opinion with little hard fact to support either side of the debate. It is, perhaps, worth remembering that the two-stage graft was introduced because of dissatisfaction with the results of the one-stage procedure, albeit in North America, which has a large Caucasian population. When the scar tissue between the tendon and sheath is fine and diaphanous, we talk of ‘adhesions’ and two-stage grafting seems to be overkill. When the scar is more dense, we talk, not of adhesions, but of scarring! This is an inconsistency of surgical thought as either can prevent movement. It is also the case that no one has worked out how to identify the ‘bad patient’, as defined above, whether in North Europe or elsewhere. Therefore, it could be argued that it is better to stage the grafting in all cases, as a failure of secondary flexor tendon surgery is usually a handicap for life and can even lead the patient to demand amputation.

Two-stage graft surgery can be unrealistic to the economic needs of many patients in many parts of the world and the circumstances of hand practice elsewhere push surgeons more towards single stage grafting. Pragmatism in this respect, and one's early teaching and experience, is probably the major determinant of the preferences of different surgeons for single or two-stage grafting! I almost always use two-stage grafting as I was brought up at a time and place where this was believed to give the best results. Sometimes, it seems that this is overkill, particularly when the sheath is not badly scarred. However, most of my cases for tendon grafting are failures of primary surgery who, as a group, are ‘bad patients’. Although not the reason for doing two-stage grafting when I started, I have come to realise that this technique may, particularly, suit these cases and my practice. Despite the preference for single-stage grafting by many surgeons of the generation who taught me, who may have had great skill in this, and the supportive feeling that the particular suppleness of the hands of certain peoples, as compared to the Caucasian hand, allows single-stage grafting to be effective more often, there will be circumstances where this expedient may be so unlikely to give a good result that the two-stage graft should be considered by even the most ardent supporter of the one-stage graft.

Previously unemphasised reasons for staging grafting, including a variety of circumstances grouped together above under ‘bad injury’ and ‘bad patient’, have become evident in my practice over the last 20 years. Some patients present with ruptures of primary repairs and undergo immediate exploration with a view to re-repair.[13] If it is found at surgery that re-repair is not possible, we routinely put a tendon rod into the finger, with the rod being replaced by a graft when the finger has settled. At this point in time, the finger is often unsuited to a single-stage tendon graft procedure because of its swollen condition, this being the third traumatic/operative for this finger in a short period of time. Others who rupture a primary repair and cannot undergo immediate re-repair for reasons such as skin breakdown and infection and those who re-rupture the re-repair may also be unsuitable for single-stage grafting.

A small group of patients presenting with severe and/or contaminated injuries, sometimes with missing segments of the flexor tendons, are deemed unsuitable for primary repair. While we endeavour to carry out primary repairs now in most of these patients while carrying out the other procedures necessary to their overall hand reconstruction, then mobilise them as early as possible, there are some cases where primary repair is impossible. Whenever possible in such cases, we insert tendon rods into the flexor sheaths to maintain the sheath until the flexor tendons can be reconstituted at a later date.

Most other situations in which the surgeon has to consider the need to graft arise in clinic in preparation for exploration of digits with various deficits of movement on elective surgical lists, whether after delayed presentation or after problems following primary surgery. In clinic, these patients should be told that the surgeon cannot predict whether the need during surgery will simply be to free the (intact) tendon from scarring then move it as early as possible to prevent re-adhesion, with a 2–4 week time off work, or the tendon will be found to have gapped or be so scarred that releasing it intact from the scarring will prove impossible. If grafted immediately, this will require a 2–3 month period without use of this hand while the graft heals fully at each end. The dilemma for the patient, given this information and, often, having just had a long period off work, is that he/she cannot tell his/her employers how long the period off work after surgery will be until after the event. Under these circumstances, many, if given the option of staged grafting, will prefer to have a rod inserted, buddy strap the finger to the adjacent one for use for a few months, and choose an opportune time from the point of view of his/her employment to have a graft inserted and the subsequent 2–3 months of one-handed life.

Where it is obvious in clinic that there is either a pulley or skin deficit on the palmar aspect of the finger, or both, I also advise two-stage grafting because (a) I believe it is easier to deal with the deficit(s) surgically without having the reconstructions of pulleys and/or skin mobilised aggressively in the early post-operative period, as is necessary to maintain movement of a tendon graft, and (b) it is easier to rehabilitate the tendon graft after the second operation without restrictions to protect reconstructions of pulley and/or skin deficiencies.

THE CLINICAL PRINCIPLES OF TWO-STAGE TENDON GRAFTING

The first stage

This often begins as the exploration of a digit with a deficit(s) of movement on an elective surgical list. Simple statements, such as ‘Tenolysis’ and ‘Tendon Graft’, which we put on our operating lists are often a gross simplification of the surgery needed and reinforce an underestimation of the problem. Whether secondary surgery is being carried out after delayed primary presentation or for failed primary surgery, all of the tissues on the palmar side of the finger may be scarred to some degree and each layer may require treatment. In respect of the management of the sheath and tendon, the separation of sheath and tendons by surgical dissection requires meticulous technique, time and considerable concentration and can be tediously slow. Great care is required to try to achieve an end point of intact tendons and intact pulleys whenever possible. Where this dissection ends otherwise, both tendons are usually removed in preparation for tendon grafting. If the FDS tendon is intact and functional, this is normally left in situ with a view to replacing the profundus tendon with a graft passed through the chiasma of the FDS, or a decision is made to leave the finger with PIP flexion only, in which case a hyperextending DIP joint may need distal flexor tendon tenodesis or joint fusion.

Two situations will make tendon grafting inevitable. The first is where the tendon is found to be so severely scarred to the sheath that achieving an end point of intact tendons and intact pulleys after surgical dissection is impossible [Figure 1]. The second is finding a gapped tendon with scar in the tendon gap after a previous primary repair. The tendon is now several millimetres too long, and tenolysis is likely to be followed by re-adhesion as the tendon will be moving less than optimally during rehabilitation. It is usually the case that these tendons are considerably scarred within the tendon sheath, possibly because the gapped tendon has moved inadequately for most of period of rehabilitation after the primary surgery.

Figure 1

A flexor tenolysis showing such dense scarring that excision of the flexor tendons leaving a functional pulley system is impossible

If a tendon is so frayed after tenolysis that I do not think it will survive rehabilitation, I replace it with a rod, with a view to two-stage tendon grafting. In this circumstance, I do not use special regimes, such as the Strickland Frayed Tendon Regime, as this places the therapists in the uncomfortable position of using a regime with which they are less familiar with tendons which may snap under their care. This may make their rehabilitation too cautious in a group of patients in whom this is likely to be counterproductive and lead to further stuck tendons.

Under all three of these circumstances, I insert a tendon rod. The distal end of the tendon of the FDS or FDP which will motor the graft is held to physiological length by suturing it, with a non-absorbable 2/0 or 3/0 suture, to either the deep transverse ligament in the palm or the ligamentous structures of the flexor aspect of the wrist. My own preference is to use the palmaris longus whenever possible as the graft, as it is easier to suture than the smaller plantaris, and to graft from the distal phalanx back to the palm, as this only requires harvest of the palmaris tendon in its extra-muscular part. Only occasionally I have found it necessary to graft back to the wrist. A silicone rod of approximately the diameter of one of the patient's flexor tendons is passed through the sheath from the distal end of the proximal motor tendon to the distal phalanx. These rods are manufactured in round and oval shape: either works equally well. We use sizes 3 or 4 most commonly. A new and useful alternative is the ‘universal tendon spacer’, which is a flexible silicon rod whose diameter changes along its length and which is easily accessed on the Internet. This rod is moved through the tendon sheath until the part of appropriate diameter fits comfortably within the sheath, and then the two ends are cut off [Figure 2]. The distal end of the rod is cut obliquely and passed behind the FDP stump distally, to maintain a pocket to aid attachment of the distal end of the tendon graft to the distal phalanx at the second operation. The rod is then sutured with 4/0 nylon to the FDP stump, and to the A4 pulley if suture to the FDP stump is thought to be too weak to hold the rod in position during subsequent finger movements until the second stage. The free proximal end of the rod is not sutured and should reach slightly proximal to the sutured distal end of the motor tendon in the palm or wrist. The hand is then closed. The first stage of a staged tendon graft procedure requires no protection post-operatively. The hand is mobilised early without protective splinting and the patient usually returns to work within 2–3 weeks. Adjunctive procedures, such as pulley reconstructions and skin replacement, are also carried out at this stage (see below).

Figure 2

A ‘universal tendon spacer’ passed through the tendon sheath until the part of appropriate diameter fits comfortably within the sheath, then the two ends will be cut off

Usually between 3 and 6 months is an adequate delay before carrying out the second-stage tendon grafting procedure. Impatience to complete the second stage by the patient or the surgeon is counterproductive as the final result is generally much better if the healing process of the first, and larger, operation is completely over.

Correction of skin deficiencies at the first stage

Scarring of the skin and subcutaneous soft tissues at the time of primary injury may cause longitudinal skin shortening in the finger or thumb. In fact, during dissection of the digit to reach the tendon sheath, if one examines the under surface of the subcutaneous fat, one will see that this almost never escapes some scar deposition with resultant tightening of the overlying skin [Figure 3]. I routinely open these cases using a mid-lateral incision to allow this incision to be extended into the distal palm as a ‘V’ [Figure 4]. This allows one to advance the ‘V’ of skin from the palm into the finger to compensate for the subcutaneous scarring in the finger. We first reported use of this technique for reconstruction of skin deficits of the palmar aspect of the fingers.[14] This advancement of palmar skin is usually enough to deal with skin shortage where the skin injury has been a simple cut. We usually do not close the V as a Y, but allow it to epithelialise under a moist antiseptic dressing done by the patient during the first few post-operative weeks of mobilisation in the way described after the McCash open palm technique in Dupuytren's surgery. Preoperatively, it is usually obvious from the nature of the primary injury and/or the appearance of the finger if more significant skin shortage is present and more skin will have to be incorporated onto the palmar aspect of the finger to achieve full extension [Figure 5a]. Use of a cross-finger flap is a simple way of incorporating more skin onto the palmar aspect of the finger for moderate cases of skin deficiency. The finger is opened through a mid-lateral incision on the same side as the intended donor finger for the cross finger flap [Figure 5b]. The mobilised skin is split at the PIP level and a cross finger flap incorporated into the gap [Figures 5c–e].

Figure 3

The under surface of the skin and subcutaneous fat of a typical case of flexor tenolysis, showing the scarring of the subcutaneous fat which causes a longitudinal skin deficiency

Figure 4

(a) Marking of the typical skin excisions used by the author for secondary exploration of the digital flexors. These include distal palmar ‘V’ incisions. (b) These allow advancement of the distal palmar skin into the digits to counter the typical skin shortage due to subcutaneous scarring shown in Figure 3, with the palmar wounds being allowed to heal by secondary intention under moist antiseptic dressings during early post-operative mobilization

Figure 5

(a) A case with more extensive shortage of skin, for which the palmar V incision is inadequate. (b) The same incision is made initially, then the flap is split at the proximal interphalangeal joint level (shown by the green line). (c) A cross finger flap is inset into the split. (d and e) the finger is mobilised normally postoperatively, then the cross finger flap pedicle divided

Cases with greater skin deficiency need more extensive flap reconstruction. There are many options of free and distant pedicled flaps available. Most will swell significantly and require secondary thinning of the flap later. Guimberteau (2001) described a simple means of resurfacing the whole palmar aspect of a finger using a distal ulnar artery pedicled fasciocutaneous flap from the distal part of the ulnar aspect of the flexor surface of the forearm.[15] Unlike distally based radial artery based forearm flaps, the ulnar artery can be dissected free into the middle of the palm, so the flap reaches the tip of the finger comfortably.

Skin deficit in the palm with flexor tendon exposure can often be reconstructed using local flaps, which have the advantage of avoiding the swelling and subsequent unnatural wobbling mobility which is common after distant flap reconstruction of the palm. A longitudinal exposure of a flexor tendon in the palm can be closed quickly and simply using local bipedicled flaps. We first reported this technique as a means of closing the gap in the palm after harvesting a Zancolli reverse digital artery flap for finger tip reconstruction.[16] Two bipedicle flaps are designed, one on either side of the defect [Figure 6a]. The common digital neurovascular structures are retained in the flaps by dissecting under the skin bridges at a deep level, immediately adjacent to the tendons and their sheaths. At the lateral margin of each palmar flap, only the skin is incised and the fibres in the subcutaneous fat broken by blunt scissor dissection to create a much more superficial wound. The palmar flaps then slide in to close the deep defect [Figure 6b] and the superficial lateral wounds epithelialise under moist antiseptic dressings done by the patient during the first few post-operative weeks of mobilisation as in the McCash Open Palm Technique for Dupuytren's surgery [Figure 6c]. We have used the same principle to close longitudinal skin defects on the palmar surface of the fingers exposing the flexor tendons and their sheath.[17]

Figure 6

(a) A patient presenting for secondary flexor surgery with a longitudinal defect of the palm and exposure of the flexor tendons in the mid-palm. The poor quality palmar skin has been excised. (b) Bipedicle flaps have been advanced centrally to close the mid-palmar defect. (c) Late view showing excellent healing of the palm with palmar skin. This case also illustrates multiple pulley reconstructions over a silicone rod

For round, or near-round, defects of the palm, large triangular flaps are useful. This technique was first described by Mathes and his colleagues in 1988 for closure of defects on the sole of the foot.[18] One, or more, skin triangles are designed and incised adjacent to the defect [Figure 7a]. The fibres immediately below the skin incisions are released by pressing down on the subcutaneous fat with a scalpel, then the deeper fat is mobilised by blunt scissor dissection. The blood supply of each flap is from small arterial branches coming up through the underlying pulp from the underlying neurovascular bundles. While a very small flap may have insufficient blood supply beneath the triangle of skin, big flaps are entirely safe. These flaps will slide freely in any direction after being released in this way [Figure 7b]. Several flaps may be designed, each sliding in a different direction, or a very big one right across the palm may be used.

Figure 7

(a) A patient presenting for secondary flexor surgery with a near-round skin deficiency in the distal palm. A palmar triangle flap has been designed and dissected prior to movement laterally into the defect. (b) Late view showing excellent healing of the palm with palmar skin

Skin deficiency exposing flexor tendons in Zone 5 can be easily reconstructed in most cases with local flaps from the forearm, using the V-Y fasciocutaneous technique[19-21] or using fascial flaps with skin graft or fascial flaps carrying a skin island vascularised by the underlying fascia.[22-24] These reconstructions may be based on either the radial or the ulnar artery in the distal forearm.

Correction of pulley deficiencies at the first stage

Single-stage tendon grafting in conjunction with pulley reconstruction requires that the pulleys be strong enough to resist the tendon forces when mobilisation is started immediately. The most important pulley deficits requiring reconstruction are those of the A2 at the base of the finger and the A1 at the base of the thumb. By careful preservation of the sheath in the middle part of the digit to avoid distal digital bowstringing, it is usually unnecessary to reconstruct an A4 pulley. Commonly, palmaris longus tendon, the proximal tendon of whichever of the flexor muscles which will not be used to motor the tendon graft or the extensor retinaculum are used as the new pulley at the base of the digit. Whichever material is used, it is passed around the repaired flexor tendons, or flexor tendon graft, and the proximal phalanx and under the extensor tendon two or three times, then sutured to itself. Bearing in mind that many cases undergoing flexor surgery have poor results because the extensor tendons, bathed in fibrin in the oedema of the injury, tether to the underlying skeleton and overlying skin to a greater degree than do the flexor tendons in their sheaths, and are restricting even passive finger flexion,[25] I have reservations about use of techniques of pulley reconstruction which require passage of tendon, or extensor retinaculum, around the phalanges and through the extensor compartment. Attachment of new pulleys to the sides of the phalanx avoids invading the extensor space, but leaves concerns in respect of their strength, unless elaborate bone fixation techniques are used. We mostly have a need for pulley reconstruction when we explore a flexor secondarily and come across a severe ‘mess’ of scarring [Figure 1]. Sometimes, the last cut of the tenolysis, or the removal of a completely welded-in flexor, destroys what was left of a weakened pulley and, sometimes, it is simply impossible to undo the scar tissue. For us, this is a situation which demands use of a two-stage tendon graft. The tendons being removed from the finger, or, if these are too poor, the more proximal part of the tendon of the muscle not intended for use later as the motor for the graft, can be used to reconstruct the pulleys. The tendon is split longitudinally and opened out, then turned through 90° [Figure 8a]. This provides enough material to make as many pulleys as one wants [Figure 8b]. Because there will be no force on them for 3 to 6 months, the new pulleys can be simply sutured to the remnant edges of the sheath, to which they will be strongly bound by the time they have to hold a tendon against the skeleton. This technique avoids harvesting of tendons which may be required for the second stage or the use of extensor retinaculum, which leaves an obvious scar on the very visible dorsum of the wrist.

Figure 8

(a) A pulley reconstruction over a silicone rod using the proximal (FDS) tendon which will not be the eventual motor of the tendon graft. (b) Another case using the same material to entirely replace the pulley system of a finger

Correction of joint deficiencies at the first stage

The deepest problems at exploration of these fingers, and the most likely to give rise to a recurrence of loss of extension, are the ligaments of the underlying joints. In general, this has little bearing on whether one grafts in a single stage or two stages, except that the degree of dissection of the interphalangeal joints usually reflects the degree of general scarring of the fingers. A considerable dissection of the proximal interphalangeal joint, particularly, will also give rise to more inflammation and scarring in the post-operative period and increase the likelihood of failure of the one-stage procedure.

Detail of the second stage

At the second operation, it is only necessary to expose the rod at its two ends through small skin incisions to allow removal of the 4/0 nylon sutures attaching the rod to the stump of the profundus tendon distally, then remove the rod and attach the graft distally and proximally. The harvested tendon graft is sutured to the proximal end of the rod and pulled distally through the new pseudosheath. Before working distally to attach the graft to the distal phalanx, it is advisable to pass a temporary suture through the proximal end of the graft and attach it to the tissues in the palmar or wrist incision to prevent the tendon graft being pulled out of the sheath repeatedly during the distal suturing activities. Numerous methods of attachment of the distal end of the graft to the distal phalanx have been described: we mostly use the technique we described in 1996 which dispenses with the use of a button on the nail of the digit[26] and is simple, effective and cheap. The proximal end of the graft is woven into the tendon of the motor using Pulvertaft's technique, with the tension of the new musculotendinous unit being set in the conventional manner by reference to the finger cascade when the hand is resting on the operating table with the wrist in the neutral position. After the second stage, the graft is mobilised by the modification of the early active mobilisation technique we use for primary flexor tendon repairs.[2728] We use exactly the same regime of immediate mobilisation after tendon grafting as we do after primary repair, which simplifies things for everyone in the unit.