Flexor Tendon Repair Techniques

Flexor tendon repair aims to restore digital flexion while minimising adhesion formation. The challenge is balancing mechanical strength, which is what allows early motion, against gliding resistance, the bulkiness of the repair.

How we got here. The history of flexor tendon repair has evolved from "primary repair is impossible" (Bunnell's "No Man's Land") to the current standard of primary repair with robust constructs that allow early active motion (EAM). Three names mark the road:

• Bunnell advocated tendon grafting in Zone II
• Kleinert introduced immediate controlled active motion with rubber bands
• Strickland defined the biomechanical requirements for EAM: work of flexion against repair strength

The tendons. Flexor digitorum superficialis (FDS) and flexor digitorum profundus (FDP) travel together into each finger, FDS splitting around FDP at Camper's chiasm. Distal to the FDS insertion only FDP continues, to the distal phalanx.

View larger

View larger

Verdan zones. Five zones for the fingers:

• Zone I - distal to the FDS insertion, so FDP only. Repair is straightforward (or advancement).
• Zone II - from the A1 pulley to the FDS insertion, where FDS and FDP share the sheath. The most complex zone, and the one with the highest adhesion risk.
• Zone III - the palm, at the lumbrical origin. Good prognosis.
• Zone IV - the carpal tunnel. Crowded.
• Zone V - the distal forearm, proximal to the carpal tunnel. Good prognosis.

The transition from Zone II to Zone III is defined by the distal palmar crease. The thumb is graded separately as T1-T3.

View larger

The pulleys. The annular pulleys (A1-A5) are thick and prevent bowstringing; the cruciate pulleys (C1-C3) are thin and collapsible, allowing flexibility. Two are critical and must be preserved: A2 on the proximal phalanx, which handles the highest load, and A4 on the middle phalanx, which is critical for DIPJ flexion. A pulley can be "vented", partially incised laterally, to let a bulky repair pass. Understanding the sheath roof is what prevents iatrogenic pulley loss.

Nutrition. The tendon is fed two ways. The vincula enter dorsally (VBC and VBP) and supply blood; synovial diffusion (imbibition) is the primary source of nutrition in Zone II. Early motion pumps synovial fluid, enhancing nutrition and healing; without motion the diffusion gradient is insufficient for the repair to heal.

View larger

How a flexor tendon heals determines whether it glides or scars, and is the biological justification for early motion.

Intrinsic healing is mediated by the tenocytes within the tendon itself, nourished by synovial diffusion and the vincula. It restores tendon continuity without binding the tendon to its surroundings, so glide is preserved; early controlled motion promotes this pathway.

Extrinsic healing comes from fibroblasts that invade from the surrounding sheath and soft tissues, laying down scar that bridges the repair but tethers the tendon as adhesions. Immobilisation, rough tissue handling, sheath and pulley damage, and gapping all push healing toward this pathway.

The phases. Healing runs through three overlapping phases, and the strength of the repair follows them:

1. Inflammatory (0 to about 5 days) - the repair is held almost entirely by the suture and tensile strength is at its lowest, the classic early "soft" period
2. Proliferative / fibroblastic (about 5 days to 3-4 weeks) - collagen is laid down and strength climbs, often after an early dip
3. Remodelling (from about 3-4 weeks, continuing for months) - collagen aligns along lines of stress and the repair regains functional strength; controlled loading guides this alignment

Why early motion works. Controlled early motion shifts healing toward the intrinsic pathway and remodels collagen along the line of pull, improving glide and final strength, but only if the core repair is strong enough to survive the loads applied during the weak inflammatory phase. That is the biological reason a strong repair and early motion are inseparable.

A closed avulsion of the FDP from the distal phalanx ("jersey finger", classically the ring finger caught in a jersey) is a distinct Zone I injury graded by how far the tendon retracts and whether a bony fragment is attached. That is the Leddy-Packer classification, and it sets the urgency.

The 1 cm advancement rule. When reattaching an avulsed or lacerated FDP to the distal phalanx, with a pull-out suture over a dorsal button or a bone anchor, the tendon must not be advanced more than about 1 cm. Over-advancing shortens the FDP and, because the profundi share a common muscle belly, produces a quadriga effect, in which the other fingers cannot fully flex, together with a DIP flexion contracture.

The presentation. A flexor tendon injury classically presents after a sharp laceration to the volar finger, hand or wrist with loss of the normal resting flexion cascade: the injured finger lies more extended than its neighbours. The resting cascade and individual FDS/FDP testing are the cornerstone of bedside diagnosis; imaging is rarely needed acutely.

What is lost depends on which tendon is cut:

• FDP division - loss of active DIPJ flexion
• FDS division - loss of active PIPJ flexion
• Both divided (Zone II) - loss of active flexion at both PIPJ and DIPJ
• FPL division - loss of active thumb IP flexion

The partial laceration. Active flexion may be preserved but painful, with weakness, triggering or catching. This is a high-risk trap for missed injury; what to do about it once the tendon is exposed is decided at operation and is set out under Surgical Technique.

Associated injuries. Digital nerve injury, giving altered sensation on the side of the laceration, and digital artery injury are frequently concomitant because both lie adjacent to the tendon, the digital nerves lateral to it.

History. Sharp, clean lacerations have the best prognosis, and the rest of the history is what shapes the plan:

• Mechanism - sharp (knife, glass) or crush (machinery)
• Time - hours since injury
• Contamination - animal bite, soil, marine environment
• Occupation - manual worker, musician, typist
• Hand dominance
• Smoking - a major risk factor for complications
• Diabetes - impaired healing
• Previous hand surgery - scarring, previous tendon injury

Red flags. Each of these requires urgent specialist consultation:

• Vascular injury - a cool, pale digit requires urgent revascularisation
• Compartment syndrome - rare in an isolated tendon injury; suspect it with a crush
• High-pressure injection - paint or grease requires emergent debridement
• Delayed presentation - greater than 3 weeks means muscle contracture and a scarred sheath
• Bite wounds - animal or human bites require thorough washout and antibiotics

Examination. To test FDS, hold the other fingers in extension and ask the patient to flex the PIPJ. To test FDP, hold the PIPJ in extension and ask the patient to flex the DIPJ. Test FDS for each digit individually: some patients have an independent FDP to the index finger (absent FDS). Check the digital nerves, check perfusion through the digital arteries (capillary refill, Allen's test), and assess the wound for depth, contamination and tissue loss.

Ultrasound can locate retracted tendon ends and confirm the diagnosis in a partial tear; it also quantifies sheath fluid and pulley thickening. A radiograph rules out an avulsion fracture (jersey finger) or a foreign body.

View larger

View larger

MRI is rarely needed for an acute laceration but is useful for a chronic rupture, where it shows the location of the tendon stump (retraction), the integrity of the pulley system, and scar tissue or adhesions, which helps in planning tenolysis or staged reconstruction.

A common exam and clinic problem is the digit that will not actively flex. The differential turns on the passive versus active discrepancy, timing, and the integrity of the repair.

Algorithm

View larger

Timing. Primary repair within 24-72 hours is the gold standard, and the limit is 10-14 days: after this the pulleys collapse and the muscle contracts. Beyond 3 weeks a two-stage reconstruction (Hunter rod) may be needed. Immediate repair is always technically easier than delayed repair.

The rehabilitation protocol. A dorsal blocking splint (wrist 30° flexion, MCP 70° flexion, IP joints extended), then one of passive flexion (Duran), place-and-hold (an active hold), or true active flexion (Stark/Manchester), which requires a 4-strand repair. The choice of protocol dictates the strength required of the repair.

Exposure. Zig-zag (Bruner) incisions prevent scar contracture. Raise full-thickness flaps and extend the incisions proximally and distally, because good exposure is key. Window the sheath between the pulleys (a C1 window, for example) to reach the tendon, and do not vertically slice A2 or A4 if it can be avoided.

Retrieval. The proximal end retracts. Retrieve it with a skin hook or a catheter; milking the forearm or flexing the wrist helps. Avoid repeated grabbing with forceps, which crushes the tendon: trauma to the tendon surface (the epitenon) causes adhesions, so be gentle. Once retrieved, pin the tendon in the sheath with a 25G needle to hold it for the repair.

The core suture. The number of core strands is the most important factor: the mechanical strength of the repair is linearly related to the number of strands crossing the repair site, and a 4-strand core is the minimum for early active motion: a 2-strand Kessler is insufficient. Use 3-0 or 4-0 non-absorbable suture (Prolene, Ethibond, FiberWire). Locking loops grab the tendon fibres and prevent pull-out, but can strangle the tendon (ischaemia). Knot placement, inside or outside, is debated; mostly inside, to reduce friction.

Purchase. Take the core suture 7-10 mm back from the cut end. This is the parameter most often left out, and it matters as much as strand count: a short purchase (under about 7 mm) lets the loops pull through the longitudinal fibres and gap, while going much beyond 10 mm buys no extra strength and strangles more of the tendon's intrinsic blood supply.

Configurations. The 4-strand options include the Adelaide, the cruciate and the double Kessler. The modified Kessler is rectangular and locking, 2 strands per pass; the Adelaide is a 4-strand locking repair developed for EAM; the cruciate is a 4-strand cross pattern; the Tajima uses locking loops to grasp the end; the Lim/Tsai is a 6-strand loop repair.

View larger

The gap. A gap greater than 2 mm at the repair site allows fibrous tissue ingrowth (adhesions) rather than tendon healing, weakens the repair significantly and predisposes to rupture; gapping greater than 3 mm leads to poor outcomes. Pre-tensioning the core suture ensures the faces are opposed, and the epitendinous running suture is critical for resisting gap formation.

The epitendinous suture. A running cross-stitch (Silfverskiold) in 6-0 Prolene, placed 1-2 mm from the cut edge. It streamlines the repair, reducing catching, smooths the edges and adds 10-20% strength.

The partial laceration on the table. What to do once you have exposed it depends on the cross-sectional area divided. The conventional threshold is around 50-60%. Below it, do not place a core suture: trim the ragged flap and, if anything, run a fine epitendinous stitch, because a bulky repair in a partial injury causes triggering and catching within the sheath and can turn a functioning tendon into a blocked one. Above it, treat as a complete division and repair formally, since the residual tendon will not survive early motion.

Test before closing. Always test the finger through a full active arc before closing, whether or not you repaired; the point is to see it glide. WALANT (Wide Awake Local Anaesthesia No Tourniquet) allows this intra-operative testing of glide.

Antibiotics.

• Cefazolin (Kefzol) 2 g IV at induction
• Post-operative oral cephalexin is often given for 5 days because of the length of the procedure and the implant material, though the evidence is debated
• Contaminated wounds: Augmentin

Adhesions are the most common complication: loss of active flexion despite good passive range, and the finger needs a tenolysis.

Rupture is the catastrophic failure. It occurs if rehabilitation is too aggressive or the repair is weak (fewer than 4 strands).

Bowstringing follows loss of the A2 or A4 pulley: the moment arm increases but excursion is lost, so the finger curls but does not fully flex.

Contracture. PIPJ flexion contracture is common if the joint is not splinted in extension.

Quadriga effect. Overtightening the FDP in one finger tethers the others through the shared muscle belly: the patient cannot make a full fist because the repaired finger hits the palm first. It is the reason for the 1 cm limit on advancement.

Infection. Deep space infection is disastrous for tendon gliding.

Four phases, from protection to strengthening:

View larger

View larger

View larger

By zone. Zones I, III and V give excellent results. Zone II results are "fair" to "good": stiffness and adhesions remain a challenge, tenolysis is required in 15-20% of Zone II repairs, and the rupture rate is quoted at 10-15% depending on compliance, against the roughly 4-10% used as an audit benchmark in the Guidelines section below.

By patient. Children have excellent healing potential, but poor compliance with rehabilitation creates a high risk of rupture or adhesion, so cast immobilisation is often used. Smokers have higher rates of complications (infection, poor healing).

Four versus six strands. Six-strand repairs are stronger biomechanically, but the added bulk increases gliding resistance and work of flexion. Most surgeons accept a robust 4-strand repair as the pragmatic standard; the incremental clinical benefit of 6 strands is unproven.

Knot placement. Internal knots reduce surface friction but may bulk the juncture; external knots are simpler but can catch. The evidence does not strongly favour one.

How much to vent. How much of A2 or A4 can be vented before bowstringing matters clinically is debated; small partial venting is widely accepted, but the safe upper limit is not precisely defined.

The rehabilitation protocol. Early active motion improves motion (Level I) but demands a strong repair and skilled therapy; where therapy is unavailable, the trade-off against rupture risk remains contested.

The slightly bulky repair. The Tang/Lalonde approach tolerates a slightly bulky repair plus venting to favour strength; traditionalists prioritise a slim, low-resistance juncture. Both have advocates.

FDS in Zone II. Whether to repair one or both FDS slips, or to excise FDS, when the sheath is tight is individualised, not standardised.