Flexor Tendon Lacerations

Updated: Oct 18, 2022
Author: Courtney E Stone, MD; Chief Editor: Jorge I de la Torre, MD, FACS 


Practice Essentials

Flexor tendon lacerations are common hand injuries and may have debilitating sequelae. It has been estimated that in the United States, flexor tendon lacerations cost the medical system $240.8 to $409.1 million annually, with a total direct cost of $13,725 per injury and indirect costs ranging from $60,786 to $112,888 per injury.[1]

Early repair has evolved as the mainstay of treatment for flexor tendon lacerations.[2, 3] There is no firm consensus on the optimal repair technique for injuries in each zone.[3, 4] In the setting of a failed early repair or conditions in which early repair is not feasible, a delayed reconstruction remains a viable option to restore function to the affected digit.

As understanding of tendon biology and healing continues to advance, repair techniques and outcomes will also improve. Research is directed toward the development of means by which to limit adhesions and scarring and to promote earlier tendon healing through the use of mesenchymal stem cells, molecular growth factors, and gene therapy.[5, 6, 7]

Anatomy and Physiology

Fibrous flexor sheath

Roughly half of all flexor tendon injuries occur in zone II. The sheath commences at the palmar plate of the metacarpophalangeal (MCP) joint with the A1 pulley. A condensation of the palmar aponeurosis (PA) results in the so-called PA pulley. Where the tendon overlies a joint, the sheath should be sufficiently thin and resilient, resulting in the cruciate (or retinacular) intervals. Where the flexor sheath overlies the phalanges, it is tough and unyielding (anular pulleys A2 and A4).

Additional anular pulleys overlie the palmar plates of the MCP, proximal interphalangeal (PIP), and distal interphalangeal (DIP) joints (A1, A3, and A5 pulleys, respectively). These are continuous with the transverse retinacular ligaments dorsally. (See the images below.)

Flexor tendons with attached vincula. Flexor tendons with attached vincula.
Retinacular portion of flexor tendon sheath. Retinacular portion of flexor tendon sheath.

In the thumb, A1 and A2 pulleys are over the palmar plates, and an oblique pulley is over the proximal phalanx. This passes from proximal ulnar to distal radial; in so doing, it is virtually an extension of the adductor, which inserts into the sesamoids. The sesamoids, into which the two heads of the flexor pollicis brevis (FPB) insert, lie within the substance of the palmar plate.

In the thumb, similar to the A2 and A4 pulleys in the fingers, the oblique pulley is sacrosanct. Because of the obliquity of the oblique ligament and ulnar takeoff, the A1 pulley in the thumb is best divided radially. This is important when surgical release of a trigger thumb is performed. No pulley should be incised during the course of tendon repair, with the exception of the A1, A3, and A5 pulleys. Repair is impossible because of the snug fit and the transverse orientation of the fibers.

Tendon sheath and pulley reconstruction

The issue of sheath reconstruction is controversial, and the decision to undertake this is best individualized after thorough assessment of the patient. The sources of fascia include the adjacent fingers, the dorsal wrist retinaculum, and the foot. For pulley reconstruction, place the tendon graft around the phalanx (sutured to itself), either beneath the extensor tendon for the A2 pulley or superficial to it for A4 pulley reconstruction.

A transverse strip of dorsal wrist retinaculum is harvested via a longitudinal incision. If a Hunter rod reconstruction is being performed, it is often useful to reconstruct the pulley first before placing the rod so as to achieve sufficient tension on the pulley. After the graft has been sutured, it is rotated around so that a synovial surface overlies the tendon. Such grafts have been demonstrated to continue secreting synovial fluid.

For more information, see Flexor Tendon Anatomy.

Tendon nutrition

The tendon derives its nutrition from the following two sources:

  • Diffusion
  • Perfusion

Diffusion takes place via the synovial lining sheath. Canaliculi pass through the tendon to the surface of the tendon. Movement of fluid into these canaliculi has been demonstrated. This effect is enhanced with digital motion.

Perfusion is accomplished via the segmental arterial supply. The blood supply to the tendon enters distally via the bony insertion and proximally via the vincula. Four vincula, designated V1 to V4, are present. V1 and V2 supply the flexor digitorum superficialis (FDS), and V3 and V4 supply the flexor digitorum profundus (FDP). They arise at the necks of the proximal and distal phalanges, respectively. In the thumb, the vincula likewise are termed V1 and V2.

No flow occurs between adjacent territories of vincula. Presumably, this area is sustained by diffusion through the synovial fluid. The vascular plexus within the tendon occupies the dorsal half. This is important at the time of placement of core sutures during flexor tendon repair.

Logically, diffusion occurs in areas of the tendon that are compressed in flexion, while the other areas are perfused. The FDP is more dependent on diffusion than the FDS is.

Tendon healing

A debate persists as to the nature of tendon healing — that is, whether it is extrinsic or intrinsic:

  • Extrinsic - The original theory was that sheath fibroblasts were responsible for peritendinous adhesions, and the tendons were healed by this route; this was the theory behind total flexor sheath excision and prolonged immobilization for tendon repairs
  • Intrinsic - Tendons bathed in synovial fluid were found to heal satisfactorily; the necessary collagen was produced by the tenocytes

Modern thinking is that tendon healing is initiated by the proliferation of epitendinous cells, which migrate into the defect, forming a "callus" equivalent. Somewhat later, the tenocytes or fibroblasts from within the tendon invade the callus, producing further collagen that realigns to produce the strong tendon. Peritendinous adhesions are not necessary for either healing or nutrition.


Various methods of expressing the results of tendon repair or grafting have been developed, including the following:

  • Littler method
  • Boyes technique
  • Total active motion (TAM)
  • Ratio of TAM to total passive motion (TPM)
  • Lister technique
  • Buck-Gramcko technique

In the Littler method, each joint's individual range of motion (ROM) is measured. The Boyes technique uses the pulp-to-midpalmar crease measurement (nail-to-table measurement can be added).

The TAM method incorporates the summed ROM of the interphalangeal (IP) joints minus the extension deficit, as a fraction of 175.[8, 9] Some incorporate MCP measurement. Strickland reported 50% of repairs achieved 50% of TAM. He reported 80% improvement following tenolysis, with 3% rate of rupture. Lister reported 71% improvement but with 21% rupture.

The TAM-to-TPM ratio expresses the relation of the postoperative TAM2 to the preoperative TPM1 for tendon grafts. In flexor tendon repair, the postoperative TAM2/175 is used. Lister believed this to be the best method for reporting results and reported 76% TAM/TPM for grafts. Strickland's formulae were as follows:

  • Flexor tendon repair - TAM2/175 as a percentage
  • Tendon grafting - TAM2/TPM1 as a percentage
  • Tenolysis - 100 – (TPM1 – TAM2)/(TPM1 – TAM1) as a percentage

Both the Lister technique and the Buck-Gramcko technique incorporate the TAM (with MCP) and the distance from pulp to midpalmar crease. These are used widely.

A retrospective review presented a functional outcome score in which tendon function, opposition, intrinsics, deformities, and sensation are assessed to evaluate the results of both tendon and nerve repair in patients with a "spaghetti wrist" combined injury of the flexor tendons, nerves, and vessels at the wrist.[10, 11]

Lister reported 80% good or excellent results in zone II with 85% outside of zone II. The FDS was excised in only 25% of patients with zone II injuries, and only 45% of those 25% achieved good or excellent results. This result is difficult to explain, but it may reflect improved blood supply or greater strength with the intact FDS. Only 12% of patients required tenolysis. Singer and Maloon reported 80% excellent or good results.

Rigo et al, in a retrospective review of flexor tendon repair outcomes in zones I, II and III (N = 291; 356 fingers), reported excellent or good function in 95 (30%) of 322 fingers at 8 weeks and 107 (48%) of 225 fingers at the last follow-up (mean, 7 months; range, 3-98 months).[12] Variables determined to be negative outcome predictors included the following:

  • Age
  • Smoking
  • Injury localization between subzones IC and IIC
  • Injury to the little finger
  • Extent of soft-tissue damage
  • Concomitant skeletal injury
  • Delay to surgery
  • Use of a two-strand Kessler repair technique
  • Attempted suture or preservation of the tendon sheath-pulley system
  • Resecting or leaving the concomitant superficial flexor tendon cuts untreated

Edsfeldt et al investigated the effect of concomitant nerve transection, combined FDP-FDS tendon transection, and patient age on digital ROM in 273 patients (311 fingers) more than 1 year after FDP tendon transection and repair in zone I and II.[13]  They found that age greater than 50 years was significantly associated with impaired digital ROM during the first year but not at 2 years. Concomitant nerve transection and combined transection of FDP and FDS were not found to affect digital mobility.



History and Physical Examination

Clinical examination is an essential part of the assessment of any patient who presents with a hand injury. Clinical clues to flexor tendon injury include loss of the normal digital cascade and the tenodesis effect observed with wrist motion. In lacerations of the flexor digitorum profundus (FDP) alone, the distal interphalangeal (DIP) joint may be held in flexion by the intact short vinculum. Isolated flexor digitorum superficialis (FDS) injury results in no overt postural deficit.

Identification of damaged structures is best performed by a hand surgeon; several reports have noted underdiagnosis of hand injuries when the injuries were examined by emergency department (ED) staff alone.[14]  Moreover, diagnoses were missed despite examination by a hand surgeon, making the involvement of a surgeon critical, in that those injuries can be identified and repaired on formal exploration.[15]



Imaging Studies

The use of bedside ultrasonography (US) in the emergency department (ED) is more sensitive and specific than physical examination for detecting tendon lacerations. In one study, the sensitivity, specificity, and accuracy of US were 100%, 95%, and 97%, respectively. Bedside US in the ED takes less time to perform than either traditional wound exploration techniques or magnetic resonance imaging (MRI).[16]



Approach Considerations

Early repair has evolved as the mainstay of treatment for flexor tendon injuries. In the setting of a failed early repair or conditions in which early repair is not feasible, a delayed reconstruction remains a viable option to restore function to the affected digit.[17]

Procedures that involve extension of the tendon are contraindicated for the following two reasons:

  • The quadriga effect on the other digits is invoked
  • An extension deficit of the involved digit is always present

Contraindications for one-stage tendon grafting include the following:

  • Less than full range of motion (ROM)
  • Inadequate skin cover
  • A hostile bed

Choice of Surgical Incision

Surgical incisions are planned so as to maintain the blood supply as completely as possible, to obtain adequate exposure, and to achieve optimal scar contracture. Bunnell developed the midlateral incision as the stationary line, either volar or dorsal to the neurovascular bundle. Littler demonstrated the diamonds of skin-to-skin contact during digital flexion. The Bruner zigzag incision avoids these areas.

Preservation of the vincula is difficult if the dorsal midlateral incision is used; accordingly, either the Bruner incision or the volar midlateral incision is recommended. Midlateral incisions should not be made on surfaces exposed to much contact (eg, the ulnar border of the little finger or the radial border of the index finger). The distal extent of either incision is over the pulp of the phalanx.

Primary Repair

The retinacular portion of the sheath is elevated as an L-shaped flap, leaving a 2-mm cuff of tissue for subsequent repair. Furthermore, this L should be oriented in such a way that a tendon end can be funneled into it. Flexion of the wrist is a useful intraoperative ploy that facilitates mobilization of the proximal tendon stump.

Often, the proximal tendon stump is hidden by a hematoma within the tendon sheath. The chiasm of Camper (flexor digitorum superficialis [FDS] decussation that lies opposite the proximal digital flexion crease) is a useful anatomic marker for tendon repair within zone II.

Distal to the midpoint of the proximal phalanx, the deep tendon is the FDS, and the superficial tendon is the flexor digitorum profundus (FDP). Before the repair is begun, it is vital that there be no tension across the tendon ends. The technique of repair is based on the length of distal tendon. Two groups of core suture exist (see the image below):

  • Those that criss-cross the tendon (Bunnell)
  • Those in which the suture lies parallel to the tendon fibers (Kessler, Tajima)
Two-strand repair techniques. (A) Tsuge, (B) modif Two-strand repair techniques. (A) Tsuge, (B) modified grasping Kessler, (C) modified locking Kessler (ie, Pennington modified Kessler), (D) modified Pennington.

Urbaniak et al demonstrated that the former group of sutures tends to strangulate the tendon ends, with a net reduction in tensile strength. The epitenon must never be handled, so that damage predisposing to peritendinous adhesions is minimized. Terminal damage to the epitenon also jeopardizes the quality of subsequent peripheral suture.

The tendon end can be grasped lightly with a toothed forceps. A 4-0 material is used for the core suture. The first pass of the needle must be parallel to the tendon fibers. The length of bite is approximately 1 cm, with the transverse limb of the suture passing superficial to the longitudinal limbs. The core suture should not pull the tendon ends together. This is the value of overlapping the ends with transfixion needles.

At least four throws of the knot should be used, with the ends cut at 2 mm. The protruding suture ends so as not to jeopardize subsequent healing or function. Knot slippage is said to occur in as many as 40% of repairs.

The suture material for the peripheral running suture should be two gauges lighter than the core suture (6-0 nylon for adults). Like the core suture, it may be composed of either monofilament or braided material. Monofilament material generally is preferred for both sutures.

The first pass of the peripheral suture should be performed in a figure-eight fashion to avoid cutting out. Most important, it should engage only the epitenon. A shallow inverting suture is ideal, thus burying the repair. A Lembert suture can be employed if necessary. The passage of this needle should meet minimal resistance if in the correct plane. After tendon repair, little or no bulk should be found in the tendon ends.

The general recommendation is that both the FDS and the FDP should be sutured, even in zone II. Lister recommended the use of two core sutures for the flat tendon of the FDS. A divided FDS derotates through 90°. The management of partial lacerations is controversial. Following the report by Weeks and Wray that partial lacerations should be managed conservatively, Kleinert et al found that failure to repair these results in triggering and delayed rupture. Recommendations for partial tendon injuries are presented in Table 1 below.

Table 1. Recommendations for Partial Tendon Injuries (Open Table in a new window)

Less than 25%

Smooth edge to avoid entrapment


Peripheral running suture

Greater than 50%

Core suture plus running suture

There has been debate as to whether multistrand repairs (see the image below) are beneficial and whether they confer greater early tensile strength and thus permit more rapid rehabilitation than conventional repair. Biomechanical and histologic data have not established definitive differences between two- and four-strand repairs. A systematic review found no differences between the two repairs with regard to either outcomes by all measures or rupture rate.[18] Good results have been achieved with four-strand modified Kessler repair.[19, 20] Nassar et al found that four different four-strand repairs had similar yield force, though differing in operating time, ultimate strength, and resistance to gapping.[21]

Multistrand core suture techniques performed with Multistrand core suture techniques performed with single-stranded suture. (A) double modified locking Kessler, (B) cruciate nonlocked, (C) cruciate cross-stitch locked, (D) 4-strand Savage, (E) augmented Becker (also called MGH repair), (F) 6-strand Savage, (G) modified Savage, (H) triple modified Kessler.

The use of an epitendinous suture (see the image below) gives similar results.[22]

Epitendinous suture techniques. (A) cross-stitch, Epitendinous suture techniques. (A) cross-stitch, (B) Lim, (C) Halsted, (D) horizontal intrafiber, (E) simple running, (F) simple running superficial and simple running deep.

If care is taken with the surgical technique, adhesion formation is not necessarily increased by the use of multistrand techniques or by the placement of an epitendinous suture. More variability is introduced by individual healing response than by an increase in tendon handling by an experienced surgeon.[23]

One report advocated a six-strand figure-eight technique, citing the ability to ensure full active extension of the proximal interphalangeal (PIP) joint in the first 4 weeks after surgery, virtually eliminating flexion contracture at the PIP joint while reducing the rupture rate to 2%.[24, 25]

Giesen et al reported on core suture repairs of lacerated FDP tendons using the six-strand M-Tang method without circumferential sutures in 29 patients (32 digits; five thumbs and 27 fingers).[26]  Pulleys were divided as much as was necessary to allow excursion of the repaired tendons, including complete division of the A4 or A2 pulleys when necessary. Of the 27 fingers, 18 had excellent results, six good results, two fair results, and one poor results. The authors found this approach to be safe for early active motion and to yield good outcomes.

Resistance to failure in different repair techniques varies according to the number of locking junctions with the tendon, as well as the location and orientation of the sutures. These factors influence resistance to failure independently of the number of sutures used.[27]

Experimental evidence in both the canine model and clinical studies suggests a stainless steel device (Teno Fix) may be promising for zone II flexor tendon repairs.[28] One in-vitro study found that resection of one slip of the FDS tendon significantly reduced the work of flexion in repair of zone II injuries, whereas no difference in this interaction was observed among three different suture materials (ie, FiberWire vs Ticron vs Prolene).[29]

Maxwell et al reported their initial experience (N = 81) with the use of the PONTiS Flexor Repair System (PONTiS Orthopaedics, San Francisco, CA), a knotless multifilament stainless-steel crimp system.[30]  They found this system to be rapid, simple to deploy, and advantageous, particularly with multiple traumatic tendon injuries. PONTiS had an overall complication profile comparable to that of traditional suture repair but better rupture and tenolysis rates. The authors suggested that this system should be used with caution in the context of poor soft-tissue coverage to minimize extrusion risk and strongly recommended concurrent use of epitendinous sutures to minimize complications.

The history of flexor tendon repair has been described well by Dr Paul Manske.[31]

Distal flexor digitorum profundus stump

The distal FDP stump usually can be involved in a formal suture repair if it is longer than 1 cm. If the distal stump is shorter than this, the proximal core suture can be inserted into the stump and brought out through two 20-gauge needles inserted through the distal pulp, 3 mm volar to the nail. If no distal stump is present, it should be brought out through the distal phalanx and sutured over a button on the nail.

Avulsion of profundus

Avulsion of the profundus is a common "rugby jersey" injury. It is most common in the ring finger. Three types are described in Table 2 below.

Table 2. Three Types of Avulsion of Profundus (Open Table in a new window)

Type I

Tendon has retracted into palm

Repair only can be performed within 10 days

Type II

Tendon has been tethered by long vinculum

Repair is feasible for up to 3 months

Type III

Large bone fragment, which cannot pass through sheath, has been avulsed

Injury can be repaired at any time

When reinsertion is possible, it should be performed as above, taking the sutures out through the pulp. If this is not possible, then either arthrodesis should be performed on the distal interphalangeal (DIP) joint or flexor tendon reconstruction should be considered.

Repair outside zone II

The epitenon is not as well defined, and the peripheral running suture is unnecessary. The objective in these situations is strength of closure rather than finesse. If the deep carpal ligament must be divided, the wrist should be splinted in slight dorsiflexion. Alternately, the latter may be repaired.

In repairs at the wrist, the best initial maneuver is to excise the glutinous mass of synovium by means of a thorough synovectomy. The distal ends easily are identified by their actions. Caution should be exercised in grasping the cut ends of the tendon (ie, grasp only the cut surface and not the epitenon). The clues to identification of the proximal ends are as follows:

  • Matching the cross-sectional areas
  • Matching the angle of laceration
  • Matching the anatomic layers - Superficial FDS to middle and ring fingers; middle FDS to index and little fingers; deep all FDP (index finger separate)
  • Matching the length to achieve the normal position in repose

Injuries at the musculotendinous junction should be repaired whenever possible. A figure-eight suture with two unequal loops is best.

Untidy injuries

In oblique injuries, when the injury is clean, there is no reason to render the cut transverse. Judging the correct tension in the core suture is more difficult, and this may be left untied until the peripheral suture is completed partially. Ragged ends should be resected. The tendon can be held with umbilical tape and resected with a razor blade. Double-level repairs do not present a problem if they are more than 3 cm apart. If they are closer, a single core suture should be used.

Secondary Reconstruction

Factors dictating the success of tendon reconstruction include the following:

  • Age - Very young and elderly patients do not fare as well
  • Occupation - Often, those with busy lifestyles cannot afford the time to spend at rehabilitation
  • Mechanism of injury - The more widespread the zone of initial injury, the greater the scarring is likely to be; the amount of scarring is inversely proportional to the ROM subsequently achieved
  • Quality of initial care
  • Previous operations - The likelihood of success lessens as the number of procedures that have preceded the proposed operation increases

The procedure should be deferred until the skin is mobile, the joints are supple, scars mature, adequate perfusion to the digit is present, adequate sensation exists, and a stable skeleton is present. Limitation in ROM may be due to skin, sheath, or tendon on either side of the digit or the palmar plate.

In pure flexor tendon adhesions, passive flexion exceeds active flexion. If extensor tendon adhesions are present, tenolysis on these should be done before flexor tendon reconstruction is contemplated. A useful approach is to examine all other structures first, leaving the flexor tendon until last.

The DIP joint contributes only 3% to the overall total arc of motion of the digit; by way of contrast, the PIP joint contributes 20% of the total arc. Jeopardizing the function of the PIP joint by tendon grafting may not have a favorable risk-to-benefit ratio. Instead, the FDS stump may undergo tenolysis to the middle phalanx, or arthrodesis can be performed on the DIP joint.

McClinton et al reported a 13% failure rate in FDP reconstruction. Therefore, the question remains as to whether benefit is found in delayed FDP reconstruction, with a possible gain of 3% total arc of motion. The risk-to-reward ratio is most favorable for carefully selected young patients in whom arthrodesis may result in growth disturbance.

Indications for Tenolysis

Tenolysis usually is indicated when passive range exceeds active range and all joints have full ROM. No optimal time exists at which to perform tenolysis. It should be performed when the following conditions are met[32] :

  • Soft-tissue scars are mature
  • The benefit from therapy has reached a plateau

Local anesthesia is best, in that it allows the patient to participate in the operation and to visualize the intraoperative gains. A forearm tourniquet may be useful. Immediate mobilization follows surgical release. In patients not operated on under local anesthesia, immobilize the digit in flexion; clot adhesions can be broken more easily by extension than by flexion. Prolonged postoperative anesthesia is optimal.

The tendons are identified both distally and proximally in the palm. Dissection continues from either side to the point of fusion or adhesion to surrounding structures. Care must be taken to avoid damage to the epitenon, which would invariably lead to tendon adhesions at this point. The completeness of tenolysis should be assessed by doing the following:

  • Mark the tendon with ink
  • Observe ROM resulting from tendon excursion (limitation is due either to adhesions or inadequate pulleys)
  • Check the proximal extent of the tendon as far as the musculotendinous junction (this is best performed by asking the patient to demonstrate ROM)

On completion, the tendon may be excessively long or tenuous, or the pulley system may be inadequate. If the tendon is very tenuous, placing a rod alongside it may be worthwhile. If the tendon ruptures, a tendon rod is already in situ. Additional procedures include capsulotomy and skin coverage with a flap. The perfusion of the digit may be poor in the extended position, necessitating a vein graft.

Local administration of corticosteroids may lead to increased rupture, attenuation, and infection. Systemic administration is controversial. Interpositional materials generally are not used.

Postoperative mobilization should be unresisted active exercise. Caution should be exercised regarding the possibility of rupture with unprotected extension. An extensor splint may be necessary for recurrent flexion contracture. However, the risk of rupture is increased by the use of such a splint.

Options for Tendon Grafting

The main donor options for tendon grafting are as follows:

  • Palmaris longus tendon
  • Extensor digiti quinti tendon
  • Plantaris tendon

Alternatively, the extensor digitorum to the index finger or the extensor digitorum longus to the second, third, or fourth toes may be used. The distal insertion is through, around, or along the distal phalanx. Lister expressed a preference for the last option, on the grounds that the first two carry some risk of damage to the nail bed.

Sasaki et al reported a case of a knife injury to the base of the right ring finger in which the FDP tendon was lacerated at the A2 pulley level.[33] The patient's left second toe flexor tendon was used as a graft to reconstruct the finger flexor tendon from palm to fingertip. Through this procedure, active ROM improved, and subsequent tenolysis resulted in restoration of function to a near-normal level. The authors concluded that the intrasynovial tendon may be used as the source of a graft to restore function in cases of tendon laceration.

If the FDP is divided distal to the lumbrical origin, little, if any, myostatic contraction occurs. Pulling the proximal tendon out to length is important to achieve the benefit of creep. The active excursion varies for each muscle, but generally, it is approximately double the passive excursion achieved on the operating table. For tendons of unequal dimension, a Pulvertaft weave is preferred for the proximal tenorrhaphy. Three slots, each at 90º to each other, are used.

Two-stage grafting

Two-stage reconstruction is indicated when the digit is not suitable for a one-stage graft for any of the following reasons:

  • Skeletal instability
  • Joint requiring capsulectomy
  • Inadequate skin coverage
  • Pulleys requiring reconstruction
  • Scarred graft bed

Scarred graft bed is the most common indication. Some digits may be insensate, have severe flexion contractures, or occur in an uncooperative patient; in these situations, the patients are not candidates for reconstruction.

The decision to proceed with two-stage reconstruction usually is made after failed tenolysis. The FDP is resected, leaving the distal 1 cm, back to a point 2 cm distal to the lumbrical origin. The FDS is preserved from the proximal end of the chiasm of Camper distally to prevent recurvatum deformity of the PIP joint.

The Hunter silicone rod or the Holter-Hauser rod may be used. The Holter-Hauser rod has a screw fixation device distally. A 4-mm rod usually is selected for adults. The proximal end of the rod is brought out proximal to the wrist crease. Complications include synovitis (15-20%), migration, extrusion, flexion contracture, and buckling of the rod. Hunter found no propensity for longitudinal contracture in the primate model. The final outcome is better when wounds are left to mature for a longer time.

Sheath closure

Repair of the sheath is highly recommended. This usually is performed with a continuous 6-0 nylon suture. The finger should be flexed to confirm that the sutures are not tethering the epitenon. Overtight closure may impede the passage of an edematous tendon. The best results reported to date were from a series in which the sheath was not closed.[34]  Strauch and others have advocated the use of autogenous vein patches for sheath closure. Closure of the sheath prior to extension of the digit after tendon repair facilitates delivery of the repair beneath the A3 pulley.

Postoperative Care

As Stirling Bunnell observed, "[a] moderate amount of intermittent movement, with as long an excursion as possible, interspersed by rest, will yield the best results."

The tourniquet should be released after closure of the sheath to achieve hemostasis.

Mason demonstrated that a healing tendon has minimal tensile strength until it is stressed. Immediate mobilization of the flexor tendon repair has been demonstrated to reduce peritendinous adhesions and increase tensile strength, DNA content, eventual excursion, and uptake of synovial fluid. The options are as follows:

  • Controlled passive motion (Duran and Houser technique)
  • Active extension and rubber-band flexion (Kleinert technique)
  • Immediate active motion with limited extension

The rehabilitation protocol must be tailored to the individual patient's status and needs. The clinician should approach each patient individually and progress them with a personalized and tailored approach in close communication with the surgeon and therapist. Functional hand motion and strength are the end results of successful flexor tendon surgery and rehabilitation.[35]

Controlled passive motion

In controlled passive motion, a dorsal block splint is used (in a similar position to the Kleinert splint), and the fingers are immobilized in Velcro straps. The joints are passively flexed three times daily. This technique often is used in children younger than 6 years; an above-elbow cast is applied with the wrist neutral, the metacarpophalangeal (MCP) joints at 90º, and the interphalangeal (IP) joints in full extension.

Active extension and rubber-band flexion

In active extension and rubber-band flexion, the wrist is immobilized 20° short of full flexion, and the MCP joints are in 45° flexion. Only the involved finger is incorporated in the rubber band. Free mobility of the other fingers encourages extension of the involved finger. In children older than 6 years, the Kleinert cowl splint is used, but the rubber band is not attached until the child fully understands the technique.

It should be borne in mind that attaching the rubber band in a patient unaware of its significance renders the patient at high risk for a flexion contracture; therefore, great responsibility is associated with the use of the Kleinert band. The Kleinert-Breidenbach splint incorporates a spring-loaded roller bar in the midpalm.

Electromyographic (EMG) studies showed that little flexor inhibition may be present during the extension phase and that opposition to extension serves little purpose. Therefore, the band should be only taut enough to flex the finger. For each 10º of flexion at the DIP joint, the FDP moves only 1 mm. Although passive motion moves joints, whether it moves tendons is unknown. The only known way of moving tendons enough to prevent adhesions is by active mobilization.[36] Modifications of the original Kleinert technique flex the wrist less and the MCP joints more.

Immediate active motion with limited extension

Several reports have advocated immediate active motion with limited extension and have demonstrated good results (see Table 3 below). However, the risk of tendon rupture is inevitably greater.

Table 3. Summary of Modifications of Immediate Active Motion With Limited Extension (Open Table in a new window)

Associated Factor


Nerve or vascular repair

Block full extension appropriately

Palmar plate repair

Block full extension appropriately

Fracture, extensor tendon repair, or replant

Early active and passive mobilization

Reversible cortical deficit, child older than 6 years

Omit rubber band until sensorium clears or child understands

Child younger than 6 years

Duran technique

Patients should be observed at least twice weekly by the surgeon and more frequently by the therapist.

If active ROM is full after 4 weeks, the splint may be removed and the elastic band attached to a wrist strap for a further 2-3 weeks. If the band is discarded at this time, the risk of rupture is significant. It should be borne in mind that patients who are mobilizing best have the greatest risk of rupture. In these patients, the rehabilitation process should be retarded.

If active ROM is poor after 4 weeks, the patient is at risk for a flexion contracture. The band is discarded, and physiology becomes more active. The patient must be warned about the risk of rupture. Dynamic splinting can be employed after 8 weeks without risk of rupture.


The most common complications of tendon grafting include the following:

  • Rupture
  • Adhesion
  • Flexion contracture
  • Bowstringing
  • Recurvatum deformity (swan neck)
  • Lumbrical plus
  • Quadriga syndrome

The rupture rate after tenolysis is high (21% in Lister's series). This is most commonly at the proximal repair site.

A meta-analysis of 29 studies examining the incidence of complications following tendon repair demonstrated a 6% reoperation rate and a 4% rate of tendon rupture and adhesion formation. The study found that use of a modified Kessler technique reduced the development of adhesions by 57% and that the presence of an epitendinous suture decreased the rate of reoperation by 84%.[37]

The causes of flexion contracture are poor rehabilitation, bowstringing of tendons, poor splinting, and failure to create an effective repair. The severity of the contracture is measured by the extension deficit. Bowstringing tendons require additional length to achieve the same degree of flexion. This excursion is not available. The diagnosis is made by asking the patient to flex against resistance.

The swan-neck deformity results from excision of the FDS. It can be prevented by retaining the part of the FDS to which the V2 vinculum attaches. It is corrected by means of either capsulodesis or the construction of a spiral oblique retinacular ligament.

Lumbrical plus occurs when the graft is too long and the tension then is taken up by the lumbrical insertion, resulting in paradoxic extension of the IP joints with forced flexion. Division of the lumbrical corrects this complication. Because the grafted finger reaches the palm before the other fingers, this places a block on further flexion of the other fingers (the quadriga syndrome).