PRIMARY FLEXOR TENDON REPAIR – OPERATIVE REPAIR, PULLEY MANAGEMENT AND REHABILITATION

PRIMARY FLEXOR TENDON REPAIR – OPERATIVE REPAIR, PULLEY MANAGEMENT AND REHABILITATION

INVITED PERSONAL VIEW PRIMARY FLEXOR TENDON REPAIR – OPERATIVE REPAIR, PULLEY MANAGEMENT AND REHABILITATION DAVID ELLIOT From the Hand Surgery Depart...

283KB Sizes 1 Downloads 115 Views

INVITED PERSONAL VIEW

PRIMARY FLEXOR TENDON REPAIR – OPERATIVE REPAIR, PULLEY MANAGEMENT AND REHABILITATION DAVID ELLIOT From the Hand Surgery Department, St Andrew’s Centre for Plastic Surgery, Broomfield Hospital, Chelmsford, Essex, UK

The central tenet of modern flexor tendon surgery is to repair and move divided flexor tendons within a few days of injury. While all flexor tendon surgery is complicated, it is simplest in the newly injured and unscarred digit and the results of correctly rehabilitated primary repairs are likely to be the best attainable. Nevertheless, repair of the divided flexor tendon to achieve normal or near-normal function consistently remains a problem which has not yet been solved. Over and above the actual technical difficulties of repairing tendons, the complications of rupture and adherence of repairs during healing continue to trouble us to an extent that the result of every primary flexor tendon repair still remains uncertain. Healing the flexor tendon takes about 3 months, a period which is sometimes longer than that for which the hand can be kept free of activities or accidents liable to snap the repair. In any healing area, a glue of fibrin-loaded oedema is formed which later converts to scar tissue to achieve a very durable bond. Unfortunately, the body does not limit this healing process to those structures which are injured. Everything in the vicinity becomes involved in the healing process, with the unwanted result that all the tissues become ‘spot-welded’ together by scar adhesions. The devastation this can cause, not only to the flexors but also to other structures of the hand, is the cause of a great deal of the morbidity of hand injury and the source of much of secondary hand surgery. This ‘spot-welding’ can occur anywhere along the length of a flexor tendon but is a particular problem in the digits, where the flexors move within a system as finely bored as the pistons in an engine. For 50 years, most of the drive in this field has been to create a mechanical system which allows us to keep the tendon repair moving after surgery, in the belief that this will prevent adhesions. Early mobilization does not, of course, prevent adhesions entirely, but it does create a form of scarring which allows us to regain much of the range of movement and, sometimes, even return function to normal. Because rupture defeats this aim, there has been a need to create sutures and suture techniques strong enough to allow this movement. Current debate is largely concerned with the fine detail of ‘best’ ways to repair and rehabilitate. An adjunct, or even alternative, to the mechanical approach of strong suture techniques and early mobilization is to try to minimize the formation of adhesions by chemical means. This has been attempted with a variety of drugs, including cytotoxics, hyaluronidase

and most recently Adcon, a proteoglycan. Prior to the release of Adcon, this approach had remained in the province of laboratory-based experiments, because of nervousness about the effect on healing of the tendon repair and the overlying tissues. Despite reasonable scientific evidence that Adcon can be of benefit in reducing adhesions following spinal disc surgery, evidence to support the use of this drug in hand surgery remains anecdotal.

REPAIR OF THE TENDON Primary repair of the flexor tendons should be performed as early as possible after the injury. However, there is now a body of evidence which suggests that a delay of 24–72 h before surgery is not followed by poorer results, and it is likely that delayed primary repair by an experienced surgeon will achieve a better result than immediate surgery by an inexperienced surgeon. This is only true if the surgery is carried out by adequately trained surgeons and followed by early mobilization of the repaired tendon by competent specialist hand therapists. For the future, emergency services in European countries should be arranged to relocate these injuries appropriately and the training of hand surgeons should be organized to allow maximum exposure to acute flexor tendon surgery as this treatment will deal with 70–80% of all injuries to the flexor tendon system if the referral pattern is effective. Independently, in 1917, Harmer (1917) and Kirchmayr (1917) first recognized the advantage of a primary repair of sufficient strength to allow immediate mobilization. Because of Bunnell’s teaching, it was not until the late 1950s that early mobilization after primary repair appeared again with the work of Kleinert et al. (1967), Young and Harman (1960) and Verdan (1952). Since that time, very many papers have been written recommending different suture types and materials. At the time of writing, there is no ‘best’ suture material or ‘best’ suture technique and the choice of each in any one unit, country or area of the world is more often determined by opinion, historical precedence and availability of particular materials than by science. A landmark paper in the assessment of the strength requisites of these repairs was that of Urbaniak et al. (1975). This paper remains the gold standard for our efforts, although there is a need to repeat this work in the light of subsequent changes of suture materials and 507

508

techniques. At present, most surgeons feel that flexor tendon repairs should include a suture within the tendon, the ‘core’ suture, and a continuous ‘circumferential’, or ‘epitendinous’, suture around the edge of the repair. The principle of the core sutures in common use is that the suture grips the tendon at a distance from the cut ends to prevent the suture pulling through the tendon fibres when subject to longitudinal tension during mobilization in the first few weeks after repair, at which time the tendon ends soften (Mason and Allen, 1941). This softening of the tendon ends eliminates any advantage of the different strengths of different suture materials (Urbaniak et al., 1975). The commonest core suture used at present in the United Kingdom is the Tajima modification of the suture originally described by Kirchmayr (1917) and redescribed by Kessler and Nissim (1969). The Tajima modification buries a single knot in the centre of the tendon. The Kirchmayr/Kessler suture is generally made with a 3/0 or 4/0 monofilament polypropylene (Prolene) or braided polyester (Ticron) suture material, with both materials being of adequate strength and each having relative benefits. We use polypropylene sutures, partly out of habit, partly because we find them easier to pull through the tendon and partly because they are less bulky to knot. Following the work of Savage (1985), which showed that a six-strand Kirchmayr/Kessler-type of core suture was very much stronger than the original two-strand repairs, there has followed a decade of intense activity to find a four- or six-strand suture which achieves similar strength but is more easily placed within the cut tendons than the Savage suture. The tendon repair has been commonly completed using a continuous circumferential or epitendinous overand-over suture, usually of 5/0 or 6/0 monofilament nylon (Ethilon) or polypropylene (Prolene) which was originally introduced to tuck in ragged parts of the tendon edges to allow easier gliding. One of the most significant experimental findings in the last 15 years has been that the epitendinous suture has considerable strength and has much more significance than the tidying role originally ascribed to it (Lin et al., 1988; Wade et al., 1986). Several recent papers describe elaborations which increase the strength of this part of the repair to a point where it may be greater than that of the core suture and many times greater than that required to prevent disruption of the repair during early mobilization. All of these new epitendinous sutures act by gripping the tendon in each throw of the suture in much the same way as the core sutures have done, so that the epitendinous suture will grip the tendon on either side of the division with eight, ten or more ‘bites’. A problem of increasing the elaboration of this suture on the surface of the tendon is increasing resistance to free gliding of the tendon within the tendon sheath (Kubota et al., 1996) and it remains to be established which of the new epitendinous sutures provides the most useful balance between additional strength and

THE JOURNAL OF HAND SURGERY VOL. 27B No. 6 DECEMBER 2002

increased resistance to movement. The possibility of dispensing with the core suture may eventually arise, although this stage has not yet been reached. Many new core and epitendinous sutures have been described and tested, mostly in vitro, during the last 10 years. It is evident from comparison of the results in the literature that most reported series of primary flexor tendon repairs in zone 2 of the fingers, which has been the testing ground of flexor tendon surgery for 50 years, include a rupture rate of approximately 5% and a tenolysis rate of 5%, regardless of which methods of core and epitendinous suturing are used. As yet, it remains to be seen whether any of the new core or epitendinous sutures will affect these figures. Although many have been shown in the laboratory to have considerably more strength than the conventional sutures, very little clinical work has yet appeared to justify their use. In 1994, while researching repair of the finger flexors, we reported a much higher rupture rate of repairs of the flexor pollicis longus (FPL) tendon than the 5% occurring in our zone 2 finger flexor repairs. Difficulty carrying out primary repair of the FPL tendon, probably because of the particular tendency of this muscle to retract more than the finger flexor muscles, has been well recognized in the literature from as long ago as 1937 (Murphy, 1937), although this particular flexor has received little attention since the 1950s and 1960s. The higher rupture rate of this repair makes it particularly suited to testing the adequacy of the newer sutures clinically, in preference to the time-honoured zone 2 model. We found that addition of one of the newer epitendinous sutures, described by Silfverskio¨ld and Andersson (1993), to a conventional Kirchmayr/ Kessler suture reduced the rupture rate of the FPL repair considerably (Sirotakova and Elliot, 1999). A more recent and as yet unpublished series of FPL repairs in which a second Kirchmayr/Kessler suture inserted at right angles to the first (one of various new methods of creating a four-strand core suture) has been added to the Silfverskio¨ld epitendinous suture has reduced the rupture rate even further. It remains the work of this decade to identify the ‘best’ of the new suture techniques or, at least, combinations of them which are both very strong and simple enough to be practical. While the new suture techniques undoubtedly provide greater strength to the repair of a divided tendon, it may be at a cost. Like the Savage core suture and its offspring, many of the new and more elaborate epitendinous sutures are not easy to insert neatly and with the degree of precision suggested in line drawings by authors of research papers. Their use makes an already complicated procedure even more so. Bearing in mind that most primary flexor tendon surgery is carried out by trainee hand surgeons world-wide, this may prove a serious disadvantage to their use. Other features of these new sutures, which have received little attention in the laboratory studies but which may make

PRIMARY FLEXOR TENDON REPAIR

them unusable clinically, are their effect on the configuration, size and shape of the tendon. Deformation, bunching and buckling of the tendon are easy enough to achieve with simpler sutures and may be more likely with the new and more complex ones. Increase of bulk of the tendon is inevitable, yet this factor has only been addressed by friction testing in a very few of the in vitro studies published recently. It is likely that clinical testing will answer these questions, but at a cost to our patients which could be lessened if these factors were considered in laboratory tests and friction tests were considered de rigeur for submissions for publication to journals on the in-vitro assessment of the biomechanics of tendon repairs. An alternative approach to the use of the core and epitendinous suture combination is popular in the Far East but much less so in Europe. I was first introduced to this approach by Guy Foucher in the early 1980s. In 1975, Tsuge described a single suture repair with a looped double-strand nylon suture which acts to grip the tendon on either side of the division in a manner not unlike a single, large epitendinous suture (Tsuge et al., 1975). This was elaborated by Tang, who suggested using three Tsuge sutures spaced evenly around the circumference of the tendon (Tang et al., 1994). Recently, Tang and his colleagues examined the strength of a number of suturing techniques, including both their own ‘triple-Tsuge’ technique and the combination of a Kessler suture and a Silfverskio¨ld epitendinous suture which we had used in FPL repairs (Tang et al., 2001). While all of the more elaborate sutures used in this study showed greater strength than a simple Kessler suture, and all were sufficiently strong to resist early mobilization, the Tang technique proved the strongest. However, it is not particularly the strength of the Tsuge–Tang approach which is most appealing (Tang and his colleagues still could not avoid a small rupture rate), but its simplicity at a time when other approaches may be becoming too complicated.

THE TENDON SHEATH The management of the tendon sheath after repair of the tendons has been influenced considerably over the years by the changing theories of tendon healing. From the 1940s until the 1960s, when it was thought that the flexors only healed if adhesions were allowed to form between the subcutaneous tissues and the tendons, segments of the sheath were excised and the hand was immobilized after zone 1 and 2 injuries (Mason, 1940; Verdan, 1958). In the 1970s, Lundborg and others (Lundborg, 1976; Lundborg and Rank, 1978) showed that tendon survived and healed in a synovial environment. There followed a period of almost obsessive closure of the tendon sheath although no one could prove any benefit to this practice and many of the best series of results at that time came from units which did

509

not close the sheath with sutures. In the 1990s, Gelberman and others showed that the tendon itself has an adequate healing mechanism. The intrinsic capacity of the tendons to heal had been thought to be very poor for a long time, hence the need to postulate adhesion or synovial healing in zones 1 and 2. However, Gelberman showed blood vessels growing along the epitenon on the surface of the tendon to bring nutrients to the repair site and heal the tendon. At present, the relative roles of synovial fluid and epitendinous vascular ingrowth in healing flexor repairs within the digital tendon sheath is uncertain. Although it seems likely that the formation of adhesions does have a place in the nutrition of the healing tendon outside the tendon sheath, the detail of the healing process in zones 3–5 has not been of particular research interest lately. As we have been moved from radical flexor sheath removal, to obsessive sheath closure and finally to calculated neglect – elevating the sheath as flaps with the minimum disruption and then laying them back as we leave – there have been clinical side-effects. This was particularly so for the almost manic attempts in the 1970s and 1980s to close the tendon sheath at all costs. As repaired tendons are inevitably greater in diameter than the original tendon, they are likely to catch once the sheath has been closed, restricting their free movement. As the writings of Mason, Verdan and the others in the 1940s and 1950s show, surgeons in the adhesion era excised parts of the sheath to allow adhesion formation and, in doing so, were also quite unconcerned about removing sufficient of the sheath to allow free movement of the repairs without catching on the edges of their sheath windows. Fortunately, when it became obvious that the results in studies in which the tendon sheath was only laid back (not sutured) were as good as in those with complete sheath closure, the phase of complete sheath closure passed and, with it, part of this problem. However, even with a policy of simply laying the sheath back, catching of some repairs on the main pulleys still seemed a problem and eminent speakers at meetings in the late 1980s and early 1990s remained polarized as to whether it was reasonable to partially release the A2 or A4 pulleys to allow free movement of repairs. My generation, who were doing this surgery as juniors at that time, sat very quietly in these meetings, knowing fully well that we not only regularly carried out these partial releases but also, on occasion, had to cut the A4 pulley completely to allow our repairs to move! The problem for the speakers was that this practice, which has become known as VENTING the pulleys, contravened another dogma. It had become recognized that one had to preserve or reproduce the A2 and A4 pulleys, as a minimum, when carrying out secondary tenolysis surgery if the mechanical function of the flexor system was to be preserved (Barton, 1969; Doyle and Blythe, 1975). This became translated into an absolute need to maintain the A2 and A4 pulleys in their entirety during primary tendon repair, despite these pulleys

510

being the main cause of repairs catching. Complete preservation of these pulleys had become almost sacrosanct, although work by Savage (1990) had shown that parts of these pulleys, and even the whole A4 pulley, could be removed without significant loss of mechanical function provided the remainder of the sheath was mostly intact. This work has been confirmed more recently by others (Mitsionis et al., 1999; Tomaino et al., 1998). In 1998, we examined 126 zone 2 repairs between the distal edge of the A2 pulley and the zone 1/zone 2 interface, at the distal end of the flexor digitorum superficialis insertion into the middle phalanx (Kwai Ben and Elliot, 1998). The repairs had been done by senior trainees in Plastic Surgery with considerable experience of operative hand surgery. Sixty-four per cent of these repairs required some lateral venting of the A2 or, more commonly, the A4 pulley. Sometimes, partial pulley venting was necessary at an early stage of surgery to place the core suture into the distal tendon end as the transverse pass of this suture has to be placed 0.5–0.75 cm distal to the tendon division and this is sometimes under the A4 pulley in these particular injuries. Sometimes, the pulley had to be partially vented at the end of the tendon repair to get a full free range of passive finger motion without catching of the tendon repair on the pulleys. In the 71 fingers requiring venting of the A4 pulley, the degree of venting varied from 10% to 100%. Complete division of the A4 was necessary in 12 fingers. When we analysed these 12 cases carefully, we realized that it was inevitable that some flexor tendon injuries would be placed such that the A4 pulley either had to be divided completely for one of the two reasons described above or for a combination of both. It has yet to be proven that venting of these pulleys, either partially or completely, has no effect on the long-term results of primary repair. However, over and above these observations of the necessities of clinical practice, venting the pulleys would seem, intuitively, correct unless perfect repair of the tendons, exactly reproducing the original diameter of the undivided tendons, is routinely achieved. Repairs snagging on pulleys in digits treated by early mobilization will either restrict movement of the digit or cause the repair to snap. In reality, this is only a problem of the A4 pulley, as the A2 pulley is of sufficient length that one can excise any one-third of it to allow repair and free movement of the repair and still have a pulley which is one or more centimetres in length, and thus, functional. In that there appears to be no indication in secondary flexor tendon surgery to reconstruct an A4 pulley alone for bowstringing at the DIP joint level, it is unlikely that the A4 pulley, in itself, is vital to flexor tendon function when most of the remainder of the sheath is intact. It is now our practice after zone 2 repairs to simply lay the sheath back over the tendons without suturing, after adequate venting of the A2 and A4 pulleys to allow free running of the repairs. Most zone 1 flexor tendon repairs, except those close to the tendon insertion which

THE JOURNAL OF HAND SURGERY VOL. 27B No. 6 DECEMBER 2002

require reattachment to the bone of the distal phalanx, will require complete division of the A4 pulley to achieve free running of the repairs (Moiemen and Elliot, 2000). Free movement of the repair is then tested for a final time by passive movement of the finger through a full range of motion before closing the skin of the finger.

REHABILITATION The fundamental need to mobilize repaired flexor tendons to avoid adhesion and loss of tendon gliding is now generally accepted, as is the fact that healing of the flexor tendon, at least in the digits, does not require the formation of adhesions. Experimental evidence would also suggest that this early movement encourages more rapid tendon healing under the influence of longitudinal forces (Gelberman et al., 1991; Mason and Allen, 1941). At the present time, flexor tendon repairs are mobilized by most surgeons in a dorsal blocking splint as an additional safe-guard against tendon rupture. We have not yet returned to the freedom of post-operative activity without a dorsal blocking splint advocated by the pioneers of 1917, although this may occur once the efficacy of some of the new suturing techniques is established and it is shown that their laboratory strengths translate into clinical practice without the other problems discussed earlier preventing their use. Of the ongoing controversies in flexor surgery, possibly the one most often discussed is which is the best of the mobilization techniques available today: ‘Kleinert Traction’, now often amalgamated with Duran–Houser passive mobilization (Duran and Houser, 1975), or ‘Early Active Mobilization’, known in the United Kingdom as ‘The Belfast Regimen’. This discussion is probably an unproductive exercise. If one looks at both techniques closely, one realizes that both are moving towards freer movement and both are pushing repairs ever harder during the early post-operative period. The series reported by Silfverskio¨ld and May (1994) from Go¨tenburg in Sweden, has the best results reported from a civilian unit in the world. It actually combines features of the Kleinert, Duran–Houser and the Belfast techniques of mobilization. As we all become more aggressive in our mobilization of repaired flexor tendons, the problem is not which regimen of mobilization to use but how far we can go along this track without increasing the rate of tendon rupture. We recently examined our own rupture patients to try to identify a soluble common cause for the problem (Harris et al., 1999). Our study showed the usual rupture rate of 5% in zones 1 and 2 in over 500 repairs in 440 patients. Half of the patients who ruptured their repair were doing something stupid, with one patient rupturing his repair while lifting furniture. Most ruptures in both sexes occurred in young patients who, as an age group, are perhaps more inclined not to listen and do

PRIMARY FLEXOR TENDON REPAIR

inadvisable things with their injured hands. This study does suggest that the only practical way to reduce the rupture rate at present is to continue to strengthen the sutures so that moving furniture during the early postoperative period can be accommodated – even if not recommended! At present, there appears to be a relative consensus of opinion about the length of rehabilitation, although the source of the timing of the various stages of this assisted recovery is obscure. Whichever technique of early mobilization is used, flexor tendon repairs are mobilized in dorsal splints with no active grasping with the fingers for 4–5 weeks. There follows a period of 3–4 weeks of gradual increase of activity. Full use of the hand for light activities and therapy to correct failures of finger extension begins only after 8 weeks, with heavy grasping activities being avoided for 12 weeks. Patients return to sedentary manual activities at 8–10 weeks and to heavy manual labour at 12 weeks after surgery. Although shortening of the period of splinting is being suggested in meetings, there are presently no published data on this. Our finding that all but one patient who ruptured a primary flexor repair in the fingers did so in the first 5 weeks, with three of 23 ruptures (13%) occurring in week 5 and none in week 6 (Harris et al., 1999) would suggest that the current period of restricted activity is close to correct. In 1989, the Belfast surgeons actively moved flexor tendon repairs in zone 2 without any strengthening of the tendon suture technique, effectively performing a Kleinert regimen without the rubber bands (Small et al., 1989). Subsequent reports from units using variants of this technique of mobilization have achieved results similar to those reported using the Kleinert technique (Bainbridge et al., 1994; Baktir et al., 1996; Cullen et al., 1989; Elliot et al., 1994). For the last 12 years, we have used a variant of the Belfast regimen of early active mobilization which we described in 1994 and which we have modified in various ways since then. After skin closure, we place the hand in a padded plaster dorsal splint which prevents finger extension beyond midflexion and elevate the hand overnight. A definitive dorsal thermoplastic splint is applied 24–72 h after surgery. While the interphalangeal joints are invariably allowed to fully extend, the precise angles to which the wrist and MCP joints may extend in these splints vary from unit to unit, regardless of whether the Belfast or a Kleinert regimen of mobilization is being used. The degree of standardization of splint construction possible in clinical practice probably belies such precision in print and the variability of the statements in the literature would suggest that the precise degree to which these joints are allowed to extend may not be of great significance. Nevertheless, our own preference at present is to have the wrist at (approximately!) 201 and the MCP joints at 401, which angles represent a slight modification of the wrist position of 301 and MCP position of 301 which we advocated in 1994. While the precision

511

with which these angles are described may not be important, the reasons for the modifications provide a useful vehicle for consideration of some of the finer points about rehabilitation. The degree of flexion of the wrist is probably less significant than previously believed. In fact, the extreme flexion advocated in early papers is reminiscent of the Phalen’s test position and may have the same effect on the median nerve, particularly when associated with the considerable local oedema of a zone 5 injury. Savage (1988) first suggested that flexion of the wrist did not achieve less tension on flexor tendon repairs distal to the wrist because any relaxation of the flexor muscles was countered by increased tension on, and spontaneous firing of, the extensor muscles which applied force to the repairs in the opposite direction. He suggested that the position in which there is least tension on repairs in the fingers and palm is the ‘resting position’ in which we commonly splint hands, with the wrist in slight extension. Several units now splint the wrist in this position and we recently examined 50 patients mobilized with this wrist position. We found no increase in tendon ruptures and the same percentages of good and excellent results as we had reported previously in 1994 with the wrist in the flexed position. Significant loss of PIP joint extension was identified as a particular problem of Kleinert traction in patients who spent long periods resting with the rubber bands relaxed and the PIP joints acutely flexed. Modifications to rubber band traction regimes, such as suggested by Silfverskio¨ld and May (1994), can prevent this. Lesser degrees of loss of full extension of the interphalangeal joints remain a problem of both current splinting regimes. Achieving full extension of the PIP joints and avoiding secondary contracture of the PIP palmar plate ligaments, particularly that of the little finger, is still difficult. As early as the first week, when necessary, our therapists place one of their own fingers behind the proximal phalanges during active extension to lift the fingers away from the splint. Patients are also taught to do this with the uninjured hand. The therapists feel that the recent small increase in MCP joint flexion of our current splint, along with temporarily increasing the flexion of the MCP joints in this way, encourages full PIP joint extension by increasing the action of the extrinsic extensor tendons on the PIP joints. Failure to extend the PIP of the little finger continues to be a particular problem in zone 5 injuries to the tendons of the little finger which occur in conjunction with division of the ulnar nerve, as the intrinsic muscles are then paralysed. This method of increasing MCP flexion has to be used more assiduously in such cases, with the little finger MCP sometimes being flexed to as much as 60–701. In some cases, contour foam is used inside the thermoplastic splint opposite the little finger to achieve this degree of little finger MCP flexion continuously. During the last year, we have introduced a modification of the original splint to further encourage PIP joint

512

THE JOURNAL OF HAND SURGERY VOL. 27B No. 6 DECEMBER 2002

Fig 1 The 2001 modification of our dorsal splint, which has no protective palmar bars, being used on a patient with a zone 2 flexor tendon repair who can be trusted (Elliot et al., 1994). (a) The fingers held in the resting, extended position, with slight extension force applied to the interphalangeal joints by a sleeve of elasticated open weave material. (b) The sleeve of elasticated open weave material has been rolled back into the palm to allow finger exercises. The fingers are shown in full extension. (c) The fingers are seen in the PIP flexion/DIP extension position. (d) The fingers are seen in the PIP flexion/DIP flexion position.

extension. In patients whom the therapists trust, the protective palmar bars are removed from the splint early in rehabilitation and are replaced by a sleeve of elasticated open weave material for the period of continuous splinting (Fig 1). This is worn at all times and exerts a slight extension force on the interphalangeal joints. The sleeve is rolled proximally into the palm when doing exercises. Passive extension exercises and dynamic extension splinting are also now used slightly earlier, during the eighth post-operative week. The use of anti-inflammatory drugs during the early post-operative period is of particular value not only for their general analgesic effect, but also as it allows the therapists to encourage early movement in nervous patients and those with a low pain threshold and/or concomitant painful nerve injuries. Ultrasound is introduced at an earlier stage of healing if finger movements into flexion or extension become more sluggish than expected and it is felt that scarring is already beginning to restrict tendon gliding. These changes are obvious to an experienced therapist as soon as 2–3

weeks after repair. Previous fears that ultrasound had a deleterious effect on the early stages of tendon healing are now thought to be incorrect. Pulsed short wave can be useful to help dispel excess oedema from the digits and palm immediately after surgery. The last 50 years are of note for the reversal of Bunnell’s policy of universal secondary tendon surgery and the recognition that the results after primary or delayed primary flexor tendon repair, that is within a few days of tendon division, are better than after delayed tendon grafting. However, we currently report our results using systems of assessment in which ‘excellent’ may be a result less than normal. For example, in the Strickland I Assessment, ‘excellent’, as defined, may only be 85% of normal function (Strickland and Glogovac, 1980). Bearing this in mind, audit from hand units internationally demonstrates that only 70–80% of results are reported as good or excellent, with two in every ten repairs either rupturing or adhering. This suggests that ‘we could still do better’, especially as the published results of flexor tendon repair

PRIMARY FLEXOR TENDON REPAIR

and rehabilitation are probably better than the national averages.

References Bainbridge LC, Robertson C, Gillies D, Elliot D (1994). A comparison of postoperative mobilization of flexor tendon repairs with ‘‘passive flexion– active extension’’ and ‘‘controlled active motion’’ techniques. Journal of Hand Surgery, 19B: 517–521. Baktir A, Tu¨rk CY, Kabak S, Sahin V, Kardas Y (1996). Flexor tendon repair in zone 2 followed by early active mobilization. Journal of Hand Surgery, 21B: 624–628. Barton NJ (1969). Experimental study of optimal location of flexor tendon pulleys. Plastic and Reconstructive Surgery, 43: 125–129. Cullen KW, Tolhurst P, Lang D, Page RE (1989). Flexor tendon repair in zone 2 followed by controlled active mobilization. Journal of Hand Surgery, 14B: 392–395. Doyle JR, Blythe W. The finger flexor tendon sheath and pulleys: anatomy and reconstruction. In: Hunter JM, Schneider LH (Eds) American Academy of Orthopaedic Surgeons Symposium on Tendon Surgery in the Hand. St. Louis, CV Mosby, 1975: 81–87. Duran RH, Houser RG. Controlled passive motion following flexor tendon repairs in zones II and III. In: Hunter JM, Schneider LH (Eds) American Academy of Orthopaedic Surgeons Symposium on Flexor Tendon Surgery in the Hand. St. Louis, CV Mosby, 1975: 105–114. Elliot D, Moiemen NS, Flemming AFS, Harris SB, Foster AJ (1994). The rupture rate of acute flexor tendon repairs mobilized by the controlled active motion regimen. Journal of Hand Surgery, 19B: 607–612. Gelberman RH, Steinberg D, Amiel D, Akeson W (1991). Fibroblast chemotaxis after tendon repair. Journal of Hand Surgery, 16A: 686–693. Harmer TW (1917). Tendon suture. Boston Medical and Surgical Journal, 177: 808–810. Harris SB, Harris D, Foster AJ, Elliot D (1999). The aetiology of acute rupture of flexor tendon repairs in zones 1 and 2 of the fingers during early mobilization. Journal of Hand Surgery, 24B: 275–280. Kessler I, Nissim F (1969). Primary repair without immobilization of flexor tendon division within the digital sheath. Acta Orthopaedica Scandinavica, 40: 587–601. Kirchmayr L (1917). Zur Technik der Sehnennaht. Zentralblatt fu¨r Chirurgie, 40: 906–907. Kleinert HE, Kutz JE, Ashbell T, Martinez E (1967). Primary repair of lacerated flexor tendons in ‘‘No Man’s Land’’. Proceedings, American Society for Surgery of the Hand. Journal of Bone and Joint Surgery, 49A: 577. Kubota H, Aoki M, Pruitt DL, Manske PR (1996). Mechanical properties of various circumferential tendon suture techniques. Journal of Hand Surgery, 21B: 474–480. Kwai Ben I, Elliot D (1998). ‘‘Venting’’ or partial lateral release of the A2 and A4 pulleys after repair of zone 2 flexor tendon injuries. Journal of Hand Surgery, 23B: 649–654. Lin GT, An KN, Amadio PC, Cooney WP (1988). Biomechanical studies of running suture for flexor tendon repair in dogs. Journal of Hand Surgery, 13A: 553–558. Lundborg G (1976). Experimental flexor tendon healing without adhesion formation: a new concept of tendon nutrition and intrinsic healing mechanisms. The Hand, 8: 235–238. Lundborg G, Rank F (1978). Experimental intrinsic healing of flexor tendons based upon synovial fluid nutrition. Journal of Hand Surgery, 3: 21–31. Mason ML (1940). Primary and secondary tendon suture. A discussion of the significance of technique in tendon surgery. Surgery, Gynecology and Obstetrics, 392–402. Mason ML, Allen HS (1941). The rate of healing of tendons. An experimental study of tensile strength. Annals of Surgery, 113: 424–459.

513 Mitsionis G, Bastidas JA, Grewal R, Pfaeffle HJ, Fischer KJ, Tomaino MM (1999). Feasibility of partial A2 and A4 pulley excision: effect on finger flexor tendon biomechanics. Journal of Hand Surgery, 24A: 310–314. Moiemen NS, Elliot D (2000). Early active mobilization of primary flexor tendon repairs in zone 1. British and European Journal of Hand Surgery, 25B: 78–84. Murphy FG (1937). Repair of laceration of flexor pollicis longus tendon. Journal of Bone and Joint Surgery, 19: 1121–1123. Savage R (1985). In vitro studies of a new method of flexor tendon repair. Journal of Hand Surgery, 10B: 135–141. Savage R (1988). The influence of wrist position on the minimum force required for active movement of the interphalangeal joints. Journal of Hand Surgery, 13B: 262–268. Savage R (1990). The mechanical effect of partial resection of the digital fibrous flexor sheath. Journal of Hand Surgery, 15B: 435–442. Silfverskio¨ld KL, Andersson CH (1993). Two new methods of tendon repair: an in vitro evaluation of tensile strength and gap formation. Journal of Hand Surgery 18A: 58–65. Silfverskio¨ld KL, May EJ (1994). Flexor tendon repair in zone II with a new suture technique and an early mobilisation program combining passive and active flexion. Journal of Hand Surgery, 19A: 53–60. Sirotakova M, Elliot D (1999). Early active mobilisation of primary repairs of the flexor pollicis longus tendon. Journal of Hand Surgery, 24B: 647–653. Small JO, Brennen MD, Colville J (1989). Early active mobilisation following flexor tendon repair in zone 2. Journal of Hand Surgery, 14B: 383–391. Strickland JW, Glogovac SV (1980). Digital function following flexor tendon repair in Zone II: a comparison of immobilisation and controlled passive motion techniques. Journal of Hand Surgery, 5: 537–543. Tang JB, Gu YT, Rice K, Chen F, Pan CZ (2001). Evaluation of four methods of flexor tendon repair for postoperative active mobilisation. Plastic and Reconstructive Surgery, 107: 742–749. Tang JB, Shi D, Gu YQ, Chen JC, Zhou B (1994). Double and multiple looped suture tendon repair. Journal of Hand Surgery, 17B: 699–703. Tomaino M, Mitsionis G, Basitidas J, Grewal R, Pfaeffle J (1998). The effect of partial excision of the A2 and A4 pulleys on the biomechanics of finger flexion. Journal of Hand Surgery, 23B: 50–52. Tsuge K, Ikuta Y, Matsuishi Y (1975). Intra-tendinous tendon suture in the hand. The Hand, 7: 250–255. Urbaniak JD, Cahill JD, Mortenson RA. Tendon suturing methods: analysis of tensile strengths. In: Hunter JM, Schneider LH (eds) Symposium on Tendon Surgery. St Louis, CV Mosby, 1975: 70–80. Verdan C (1952). Re´paration primaire des fle´chisseurs en dehors des coulisses osteo-fibreuses des doigts. In: Chirurgie Re´paratrice et Fonctionnelle des Tendons de la Main. Paris, L’Expansion Scientifique Franc¸aise: 174–176. Verdan C (1958). La re´paration imme´diate des tendons fle´chisseurs dans le canal digital. Acta Orthopaedica Belgica, 24 (Suppl III): 15–23. Wade PJF, Muir IFK, Hutcheon LL (1986). Primary flexor tendon repair: the mechanical limitations of the modified Kessler technique. Journal of Hand Surgery, 11B: 71–76. Young RES, Harmon JM (1960). Repair of tendon injuries of the hand. Annals of Surgery, 151: 562–566. Received: 14 August 2001 Accepted after revision: 11 April 2002 Mr D Elliot, MA, FRCS, Hand Surgery Department, St Andrew’s Centre for Plastic Surgery, Broomfield Hospital, Court Road, Chelmsford, Essex CM1 7ET, UK. Tel.: +44 1621 857 362; Fax: +44 1621 841 127; E-mail: [email protected] r 2002 The British Society for Surgery of the Hand. Published by Elsevier Science Ltd. All rights reserved. doi: 10.1054/jhsb.2002.0800, available online at http://www.idealibrary.com on