Part I: Staged Flexor Tendon Reconstruction
JAMES M. HUNTER *
Indications/Contraindications
* Part I reproduced from Hunter JM: Staged flexor tendon reconstruction. In Strickland JW, editor: Master techniques in orthopaedic surgery, the hand, Philadelphia, 1998, Lippincott-Raven.
Transformation of the scarred, postinjury flexor tendon complex to a gliding, pliable, functional system can be accomplished by the two-stage tendon graft method using the Hunter tendon implant (Wright Medical Technology, Inc., 5677 Airline Road, Arlington, Tennessee 38002, 901-867-9971) at stage I. This method permits the “linchpin” factors to take place by permitting new anatomic reconstruction of pulleys and fibrous sheathing around the tendon implant. The results are improved gliding biomechanics and a fluid flow through a synovial nutrition system that can maintain gliding of active and passive implants and ultimately nourish the stage II tendon graft free of restricting adhesions. This concept permits the surgeon and therapist to achieve maximum functional gains in rehabilitation of the hand because stage II surgery can be delayed for extended periods of time.Staged Tendon Reconstruction
A tendon implant is indicated in the following situations: (1) as a temporary segmental spacer in certain damaging injuries where primary tendon repair is not likely to give a good result; (2) in scarred tendon beds where a one-stage tendon graft can be predicted to fail; and (3) in salvage situations where, despite predicted degrees of stiffness, scarred tendon bed, and reduced nutrition, useful function can be returned. This procedure has been used successfully in extensor tendon reconstruction, reconstruction of the severely mutilated hand, and construction of tendon systems in congenital anomalies with deficient tendon systems. Any tendon transfer (i.e., opponens transferre) that would have to transverse a suboptimal bed should be considered for this procedure. This procedure also may be indicated for grafting of a profundus tendon through an intact superficialis when the profundus tendon bed is scarred. In a Boyes grade V or VI salvage finger, particularly when the adjacent finger has been amputated, the passive or active tendon may be used to create a superficialis finger. In this procedure, arthrodesis is performed on the distal joint, and the distal juncture of the implant is to the middle phalanx.
In acute trauma, the tendon implant also may be used if the wounds have been adequately debrided and rendered surgically clean. When the injury requires simultaneous fracture fixation and flexor and extensor tendon repair, the use of a tendon implant in the flexor system should be considered.
A timely indication for the two-stage procedure has been in replantation surgery. In multiple digital amputations, the use of the implant in the flexor system at the time of replantation may significantly simplify the postoperative rehabilitation. A full-length tendon implant, or even a short spacer, can maintain the fibroosseous canal and regenerate a flexor sheath in damaged areas after the fracture or fusion has healed and the neurovascular status is stabilized. The passive flexion and active extension have helped rehabilitate both the flexor and extensor systems simultaneously and have significantly improved the function of replanted digits. If the replantation is proximal to zone II or at wrist level, it is reasonable to repair the flexor tendons proximally and place extensor implants dorsally, because the wrist is usually flexed to protect the neurovascular repairs. Again, the postoperative rehabilitation is simplified.
Acute infection is an absolute contraindication to this procedure. Appropriate surgical and antimicrobial treatment and subsequent wound healing will allow the procedure to be carried out later without complication. Finally, the digit that has borderline nutrition, bilateral digital nerve injuries, and joint stiffness may be better treated by amputation rather than reconstruction.
Active Tendon Implants
An active tendon or passive tendon implant may be indicated in the following situations: (1) as a temporary segmental spacer in selected primary injuries where conditions are unfavorable or impossible for tendon repair; (2) in scarred tendon beds where a one-stage tendon graft would most likely fail; and (3) in salvage situations where useful functions can be returned despite predicted stiffness, scarred tendon beds, and reduced nutrition. If an injury requires simultaneous fracture fixation and flexor and extensor tendon repair, the implant should be considered for the flexor system. The only absolute contraindication to this procedure is acute infection.
The following are indications for an active tendon rather than a passive tendon: (1) a patient with the proper motivation for compliance with the rehabilitation protocol, (2) an extensor system that functions well enough to balance the predictive flexion of the finger, and (3) patient understanding that a superficialis finger reconstruction with distal interphalangeal (DIP) joint arthrodesis may be needed if the DIP joint extensor tendon function is not adequate or if more than two pulleys will require reconstruction. The active tendon, when used in the hand injury requiring multiple reconstructive procedures, can be an important choice for patient morale.
Superficialis Finger Reconstruction
The superficialis finger concept is based on the principle of a one-tendon/two-joint finger. The distal fixation of the implant is on the proximal aspect of the middle phalanx, and the proximal implant juncture is in the forearm, as previously described. In addition, tenodesis or fusion of the DIP joint is done, and pulley reconstruction then can be performed at the proximal phalanx ( Fig. 29-1 ). Indications for the superficialis finger include the following: (1) reconstruction of more than two pulleys is needed; (2) extensor tendon control of the DIP joint is poor; (3) arthritis of the DIP joint is present; (4) there are multiple finger injuries in one hand and partial or complete amputations of adjacent fingers; and (5) vascular deficiencies of the finger are present.
Preoperative Planning
Examination: Passive Tendon
The preoperative assessment includes a careful analysis of dismembered parts of the hand and deficient sensibility patterns because these will significantly affect final results. Sensibility loss should be studied carefully before flexor tendon reconstruction, and the nutritional status should be reviewed by the Allen test. The need for skin grafts, which will affect contractures of the fingers, also should be examined carefully at this time. It is important to record ranges of joint motion and the ability to extend digits actively within those ranges and to look at how far the finger pulp comes from the distal palmar crease.
Pencil sketches showing range of motion (ROM) and lack thereof are helpful. The sketches also should include the site of injury and other abnormalities, such as needed skin grafting. Additional sites of injury beyond the finger (i.e., palm and forearm) should be recorded, because these will affect the position of the tendon implant.
Types of Hunter Tendon Implants
Passive Gliding Implant System
Two basic types of Hunter tendon implants have evolved from the experience gained in experimental and clinical trials during the past 25 years: passive implants (Mark I) and active implants (Mark II and III). All the implants have a core of woven Dacron that is pressure molded into a radiopaque medical-grade silicone rubber. The surface finish is smooth, and the cross-sectional design is ovoid to aid optimal tendon sheath development. The term passive gliding implant system implies that the distal end of the implant is fixed securely to bone or tendon while the proximal end glides freely in the proximal palm or forearm ( Figs. 29-2 and 29-3 ). Movement of the implant is produced by active extension and passive flexion of the digit. A new biologic sheath begins to form around the implant during the period of gliding that follows: stage I surgery. The new sheath progresses through a 4-month phase of biologic maturity and develops a fluid system that supports gliding and nutrition gliding for the tendon graft after stage II. Usually, 3 to 4 months after stage I surgery, the sheath is mature and the implant can be electively replaced by a tendon graft ( Figs. 29-4 and 29-5 and Plate 31 ). The passive implant incorporates design characteristics of firmness and flexibility to permit secure distal fixation and to minimize the buckling effect during the passive push phase of gliding.
Silicone rods are not reinforced and therefore tend to coil and buckle during flexion. Distal fixation is a problem with a simple silicone rod; eventual loosening of the rod results in synovitis and proximal migration. All Hunter implants, active and passive, have been designed with a woven Dacron core to eliminate these problems so that results can be predictable.
The two passive tendon implants available differ only in their distal juncture. One implant has a stainless steel distal metal end plate that is attached to the distal phalanx by a screw. It provides excellent fixation to bone, eliminating loosening and proximal migration of an implant, the number one cause of sheath synovitis. It also provides the added benefit of shortening the second-stage procedure. The screw hole in the distal phalanx acts as a guide hole that is further enlarged in an oblique fashion for the acceptance of the tendon graft. The Woodruff and the 2-mm AO bone screws are available in various sizes. The length of the screw is determined from the preoperative roentgenogram. A pilot hole is drilled at a 15- to 20-degree angle to the DIP joint with a 1.5 drill. The bone cutter should not be used because self-cutting of the bone is needed. The length of the screw should be sufficient to engage the dorsal cortex of the phalanx, but not pass beyond it, because to do so could result in pain dorsally. The tip of the screw should be proximal to the germinal matrix of the nail so as not to cause a nail deformity. This implant is available in 6-, 5-, 4-, and 3-mm diameters. The lengths are either 23 or 25 cm. These implants can be trimmed proximally to the appropriate length.
The passive tendon implant without a screw-fixation terminal device is held in place with a 4-0 nonabsorbable suture that is woven through the distal end of the implant; this distal end is secured under the profundus stump (see Fig. 29-2 ). Care must be taken to place the suture through the central Dacron core. The distal juncture is also reinforced with two lateral sutures of Dacron. This implant is available in the following sizes: 3 mm × 23 cm, 4 mm × 23 cm, 5 mm × 25 cm, and 6 mm × 25 cm. This implant can also be shortened and should be trimmed at the distal end.
The passive tendon implant is indicated for the young patient with an open epiphyseal plate. A 3-mm tendon implant is often used. The digital end is sutured to the stump of the profundus. Once again, the sutures must be placed in the Dacron core because if the sutures are just in silicone rubber, the distal end will loosen during passive mobilization after the stage I surgery.
Active Tendon Implant Types
Four types of active tendon, one fixed and three adjustable lengths, are available ( Fig. 29-6 ). The fixed-length implant consists of a metal plate distally, a porous-cord silicone flexible shaft, and a preformed porous-cord silicone loop proximally. It is available in 16-, 18-, 20-, and 22-cm lengths and 4-mm overall diameter. The adjustable-length implant is useful when extremely short or long tendon defects are encountered. It allows for length adjustment by pulling the silicone away from the two porous woven polyester cords, allowing for shortening of the implant ( Figs. 29-7 to 29-9 ). Care must be taken not to damage the fine polyester weave while peeling away the silicone. Three designs are available ( Table 29-1 ).
Distal Fixation | Proximal Fixation |
---|---|
Metal plate | Two free porous cords |
Two free porous cords | Porous-cord silicone loop |
Two free porous cords | Two free porous cords |
Surgical Technique
Reconstruction of Scarred Flexor Tendon Bed Using the Basic Dacron Reinforced Implant
The damaged flexor tendons and their scarred sheath are exposed. In the finger, this is done through a midlateral incision or the zigzag incision of Bruner. In the palm, exposure is accomplished through transverse incision or through a proximal continuation of Bruner zigzag incision. In the forearm, an ulnarly curved volar incision is made to expose the proximal portions of the flexor tendons and their musculotendinous junctions (see Fig. 29-2 ). A stump of the profundus tendon, 1 cm long, is left attached to the distal phalanx. Scarred tendons, sheath, and retinaculum then are excised. Contracted or scarred lumbricals always should be excised to prevent the paradoxical motion of the lumbrical that is seen after some tendon grafts. This motion causes the finger to extend rather than flex as the patient attempts to flex the finger completely.
Undamaged portions of the flexor fibroosseous retinaculum that are not contracted are retained. Any portion of the retinaculum that can be dilated instrumentally with a hemostat is also preserved; the remainder is excised. The retinacular pulley system (see Figs. 29-1 and 29-2 ) should be preserved or reconstructed proximal to the axis of motion of each joint; otherwise, normal gliding of the tendon will not be restored. Four pulleys are preferred: one proximal to each of the three finger joints and one at the base of the proximal phalanx.
The distal end of the prosthesis is sutured beneath the stump of the profundus tendon after resecting all but the most distally attached fibers of the tendon (see Fig. 29-2 ). A figure-of-eight suture of no. 32 and 34 monofilament stainless steel wire (3-0 Ethibond) on an atraumatic taper-cut needle is used. In addition, medial and lateral sutures of no. 35 multifilament wire are passed through the tendon, prosthesis, and fibroperiosteum for further fixation. Any excess of profundus tendon is resected. Traction then is applied on the proximal end of the prosthesis in the forearm to ensure that the attachment of the prosthesis is distal to the DIP joint and its volar plate and that there is no binding of the tendon during flexion and extension. The prosthesis is also observed during passive flexion and extension of the finger (see Fig. 29-2 ) to ensure that it glides freely with no binding or buckling distal to some part of the pulley system that may be too tight. If any portion of the system is tight, it must be removed and replaced with a new pulley constructed from a free tendon graft.
The proximal end of the prosthesis also should be observed during passive flexion and extension to ensure that it glides properly (see Fig. 29-2 ). The proximal end of the prosthesis should be in the forearm so that the newly formed sheath extends to the region of the musculotendinous junction of the motor muscle (see Figs. 29-2 and 29-3 ). The proximal end may be placed superficial or deep to the antebrachial fascia or deep in one of the intermuscular planes. The track for the prosthesis can be fashioned by separating connective tissues and tendon mesenteries with the moistened gloved finger. The track must permit free passive gliding of the prosthesis during passive flexion and extension of the finger. If such a track cannot be established by spreading and adjusting the tissues, the prosthesis should be shortened so that, when the finger is fully extended, the proximal end of the prosthesis lies proximal to the flexion crease at the wrist. When multiple prostheses are threaded through the carpal canal, the superficialis tendons are generally removed from the canal.
Finally, before the wound is closed, traction should be applied to the prosthesis again and the amount of active finger motion determined and recorded (see Fig. 29-3 ). Importantly, if this maneuver does not produce full flexion, it may be necessary to modify the pulley system.
A small amount of barium sulfate is incorporated in Dacron-reinforced active and passive implants so that their function can be checked roentgenographically at 6 weeks and again just before insertion of the tendon graft. Anteroposterior and lateral roentgenograms of the hand and distal half of the forearm are made with the fingers and wrist in full extension and full flexion. These roentgenograms will demonstrate how much the proximal end of the prosthesis moves with respect to the distal end of the radius. If there is full ROM of the wrist and all finger joints, an excision of 5 to 6 cm is not unusual (see Figs. 29-3 and 29-4 ).
Stage II Tendon Grafting Following Active or Passive Implants
The interval between stages I and II should be 2 to 6 months, or long enough to permit maturation of the tendon bed to the point where it can nourish and lubricate the gliding tendon graft and until maximum softening of the tissues and mobilization of the stiff joints has been achieved. Each case must be individualized, and the decision to do the second-stage procedure must be made by the surgeon on the basis of the findings in the hand.
Before stage II surgery, the limits of extension and flexion of the finger must be accurately measured and recorded to establish the base for postoperative care. A short midlateral or Bruner zigzag incision then is made to locate the distal end of the prosthesis, where it is attached to the distal phalanx. This attachment is left intact, and a second ulnarly curved volar incision is made in the forearm through the original stage I incision to expose the proximal end of the prosthesis and the musculotendinous junction of the superficialis or profundus tendon, which is used as a motor for the tendon graft. With the prosthesis still in place, excursion of the proximal end of the prosthesis as the finger is moved from full extension to full flexion also should be measured as an additional check on the amount of excursion that the motor muscle must have to provide full finger motion.
The palmaris longus tendon works for short tendon grafting: thumb, fifth finger, and sublimus fingers. Longer grafts may be required for longer fingers. A long tendon graft is obtained from one leg (preferably, the plantaris) but if this is missing, a long-toe extensor tendon may be used. If a toe extensor must be used, the graft is obtained using a modified Brand tendon stripper and two or more incisions so that the portion of the tendon proximal to the retinaculum is obtained. Any attached fat or muscle is removed, and one end of the graft is sutured to the proximal end of the passive implant with a catgut or polyester suture (see Fig. 29-3, A ). Leaving the distal end of the prosthesis attached to the distal phalanx, the remainder of the prosthesis with the attached tendon graft is pulled distally, thereby treading the graft through the new sheath. The prosthesis then is removed and discarded. Free motion of the graft in the sheath can be confirmed by grasping each end of the graft with a hemostat and pulling it proximally and distally.
The tendon graft is secured to the distal phalanx using a Bunnell-type wire suture, with the button on the fingernail, and medial and lateral reinforcing sutures through the profundus tendon stump (see Fig. 29-3 ). Traction is applied to the proximal end of the graft, and the predicted range of active flexion, measured as the distance of the finger pulp from the distal palmar crease, is determined. After this maneuver, attachment of the graft to the distal phalanx is inspected to check the security of the fixation.
Scar tissue from previous surgery, antebrachial fascia, and muscle fascia are excised to minimize motion-restricting adhesions. When the firm fascia is carefully dissected away from the newly formed tendon sheath, the sheath is found to be soft, with loose mesentery-like attachments to the surrounding tissues. The sheath should be resected far enough distally that there will be no scar in the region of the tendon suture of the anastomosis; this is not always possible, however. In this event, the sheath is either dissected away completely so that there is no contact between the anastomosis and the sheath, or the sheath may be left open so that one side of the anastomosis glides on the sheath (see Fig. 29-4 ).
For the index finger, when either the superficialis or the profundus muscle is available as a motor, the graft is anastomosed to the proximal segment of the motor tendon according to the method of Pulvertaft (see Fig. 29-4 ). However, for the long, ring, and little fingers, the graft is woven through the oblique stab incisions in the common profundus tendon, securing the different tendons together as one tendon unit (see Fig. 29-4 ).
It is essential to adjust the length of the graft accurately. The excursion of the prosthesis during flexion and extension has already been determined. The excursion of the tendon graft should be checked by pulling on the graft, starting with the finger in full extension (see Fig. 29-4 ). Having determined the excursion necessary to produce a full range of flexion, the excursions of the available motors then are determined, and the one with the requisite excursion is selected.
The tension of the graft is adjusted so that, with the wrist in neutral, the involved finger rests in slightly more flexion than that of the adjacent fingers (see Fig. 29-22 ). When the anastomosis has been completed, the tension is checked with the wrist in both flexion and extension to assess the tenodesis effect and to ensure that the tension of the graft is correct. If the patient is under local anesthesia, after the distal anastomosis is completed and the distal wound is closed, the graft is sutured tentatively to the motor tendon and, after the tourniquet has been deflated for 10 to 15 minutes, the patient is asked to flex and extend the finger. If the predicted amount of active flexion is not achieved, the tension of the graft is readjusted or a motor with more excursion is selected. When the best possible function has been achieved, the anastomosis is completed, and the wound is closed and dressed.
Stage I: The Active Tendon Implant
This technique is divided into two stages, although stage II may be delayed for months or years. Stage I is similar to stage II using the passive tendon implant, and the same indications are used. The major differences are the exact type of implant and the placement of the proximal junction to an active motor tendon for early function.
Implant Positioning
The distal component is passed from the palm through the finger pulleys using a no-touch technique, which can be facilitated by moistening the device with Ringer’s solution. The implants with porous cords distally or proximally are easily passed, whereas those with a distal metal plate are more difficult. Wet umbilical tape or heavy nonabsorbable, nonmetallic suture is threaded through the screw hole. Both ends of the tape or suture are used to guide the distal component through the A 1 and A 2 pulleys. In most cases, the A 4 retinaculum needs to be cut along the periosteal rim on the middle phalanx to allow the implant to pass. The A 4 pulley should be repaired with multiple sutures using small drill holes if necessary.
Pulley Reconstruction
As with the passive tendon technique, stage I is the time for pulley reconstruction, if it is necessary. The flexor retinaculum must be reconstructed and of adequate strength to support the strong vector forces of the active flexor tendon implant. A free tendon graft is wrapped around the bone ( Fig. 29-10 ). The tendon is passed under the extensor tendon hood for A 2 and A 4 pulley reconstruction. At least two wraps are recommended for adequate strength to allow early active motion. If possible, four wraps are preferred for A 2 reconstruction. A separate small dorsal skin and extensor tendon incision will allow a retractor to be placed to lift the extensor tendons to facilitate passing the free graft under the extensor tendon hood. After the graft has been wrapped, it is held with sutures connecting each wrap. The tendon graft pulley placed in the dorsal synovial bed under the extensor tendon will remain soft and flat during function.
Proximal Implant Placement
Next, the implant is passed proximally from palm to forearm through the carpal canal. If necessary, a tendon passer can be used (see Fig. 29-19 ). The diameter of the passer should be slightly larger than that of the implant. Again, the implant with porous cords is easily passed proximally, whereas the implant with a silicone loop may be more difficult proximally. A wet umbilical tape can be looped through the implant loop and brought into the forearm on a tendon passer. It is imperative to protect the silicone coating on the loop during this process.
Distal Fixation: Metal Plate (Active or Passive Implants)
The distal component must be fixed to allow strong, durable, immediate juncture. With the joint identified, the base of the distal phalanx is stripped of volar periosteum, and the following steps are performed ( Fig. 29-11 ). The point where the screw should enter the bone is measured. The ideal position of the component is with its proximal edge 3 mm distal to the joint. This position is marked. The plate then is centered over the marked bone ( Fig. 29-11, A ). The dorsum of the patient’s finger is placed on a firm part of the operative table, and a bone impacter is positioned against the plate and “struck firmly,” which drives the four sharp spikes into the bone so that the plate is secure and evenly aligned on the bone. A 1.5-mm drill bit is passed through both cortices of bone at an angle 15 degrees dorsal to the joint line ( Fig. 29-11, B ), and a lateral radiograph is taken. The drill should have missed the joint and base of the nail and should be in the central portion of the cortical bone.
A 2-mm AO screw is recommended ( Fig. 29-11, C ). Holding the joint securely, the proper 1.5-mm-diameter hole is drilled, and the proper screw length is chosen. The distal end is secured by carefully turning the screw through both cortices of bone to thumb tightness. The screw should appear dorsally with less than 1 mm protruding. The screw fit should be firm all the way, and final fixation should be secure. If the final screw turns are loose, the system may not endure cyclic force. If the screw fixation is not firm, it should be either reinserted more distally in bone or fixed with twisted wires (see following discussion). Importantly, the four metal spikes on the metal end of the implant do the hard work, and the screw merely holds the plate securely in place. The profundus tendon stump should be drawn over the distal component and sutured laterally to provide a soft tissue buffer to prevent irritation of the overlying skin ( Fig. 29-11, D ). In the superficialis finger procedure, a strip of superficialis tendon may be drawn over the distal component and sutured laterally at the middle phalanx level. The tourniquet is released, the wounds are irrigated, and the distal incision is closed.
The distal component is also designed to allow fixation by two twisted wires through the bone as well as the screw. If the bone has been fractured or shows osteoporosis, wiring is preferred. Also, if the bone is deformed during preparation for screw insertion, the wire fixation by two drill holes is preferred ( Fig. 29-12 ). The distal metal component with spikes must be held securely to bone to prevent movement and wire fracture. Fixation must be strong enough to prevent the tendency of the plate to lift off the distal phalanx ( Fig. 29-13 ).