Signs and Symptoms
Imaging and Other Diagnostic Studies
Introduction to Distal Phalanx (P3) Fractures
Pathoanatomy and Applied Anatomy
Dorsal Base Fractures—CRIF
Dorsal Base Fractures—ORIF
Volar Base Fractures
Author’s Preferred Treatment
Dorsal Base Fractures
Introduction to Distal Interphalangeal (DIP) and Thumb Interphalangeal (IP) Joint Dislocations
Introduction to Middle Phalanx (P2) Fractures
Dynamic Extension Block Splinting
Condylar Fractures of the Head
Volar Base Fractures—Closed Treatment
Volar Base Fractures—ORIF
Volar Base Fractures—Osteochondral Reconstruction
Introduction to Proximal Interphalangeal (PIP) Joint Dislocations
Dorsal Dislocations—Nonoperative Management
Pure Volar Dislocations—Nonoperative Management
Rotatory Volar Dislocations—Nonoperative Management
Introduction to Proximal Phalanx (P1) Fractures
Local Soft Tissue Relationships
Biomechanics of Fixation
Proximal Phalanx (P1) Fracture Treatment Options
Closed Reduction Internal Fixation
Open Reduction and Internal Fixation with Lag Screws
Plates at the Phalangeal Level
Introduction to Metacarpophalangeal (MP) Joint Dislocations
Open Reduction of Finger MP Dislocations
Thumb MP Collateral Ligament Repair
Free Tendon Graft Reconstruction of the Thumb UCL
Introduction to Metacarpal Fractures
Closed Reduction and Internal Fixation
Open Reduction and Internal Fixation
Introduction to Carpometacarpal (CMC) Joint Dislocations and Fracture Dislocations
Finger CMC Joints
Thumb CMC Joint
Finger CMC Joint Pure Dislocations
Finger CMC Joint Fracture-dislocations
Thumb CMC Joint Pure Dislocations
Thumb CMC Joint Fracture-dislocations
Management of Expected Adverse Outcomes and Unexpected Complications
Malunion and Deformity
Summary, Controversies, and Future Directions
Managing Skeletal Loss
Many tuft fractures can be splinted in a simple aluminum and foam splint for a duration determined by the patient’s symptoms alone. The time course for healing of the associated soft tissue injury may well determine the total duration of disability far more than that of the fracture itself. When the seal of the nail plate with the hyponychium has been broken and the tuft fracture is displaced, this represents an open fracture that should be treated on the day of injury with debridement followed by direct nail matrix repair. If the distal fragment is of substantial size, the dorsal cortex of the distal phalanx that supports the nail matrix will provide a more level surface if pinned with one or more 0.028-inch K-wires for 4 to 6 weeks.
Longitudinal sagittal plane shaft fractures of the distal phalanx can be treated entirely nonoperatively if minimally displaced or with CRIF using oblique 0.028- to 0.035-inch K-wires for the rare displaced fracture. Two or more wires should be used to prevent sliding of the fracture along the smooth surface of a single wire. Care should be taken to avoid penetration of the nail matrix with the wire. If the fracture is at midshaft level or more distal, the wire will provide enough stability if driven to the subchondral base of the distal phalanx only. Fractures occurring at the metadiaphyseal junction may need to have the wire passed across the DIP joint to achieve sufficient stability.
Dorsal base intra-articular shear fractures produce a triangular dorsal fragment that is extended and translated by the pull of the terminal tendon. With proper collateral ligament damage, the larger articular fragment that is in continuity with the remainder of the phalanx may sublux volarly. ORIF adds excessive surgical trauma to this delicate set of tissues and the dorsal fragment is usually too small to accommodate fixation devices passing directly through it without experiencing comminution. The injury is best addressed by extension block pinning. The DIP joint is hyperflexed, drawing the dorsal fragment volarly to reach its natural position in relation to the head of P2. A 0.045-inch K-wire is then inserted at the dorsal margin of the fragment (but not through the fragment) to block it from returning to the retracted position under the influence of the terminal extensor tendon (Fig. 30-14). The remainder of the distal phalanx (consisting of the volar articular fragment and shaft) is then extended to meet the blocked smaller fragment and restore articular congruity. A second 0.045-inch K-wire is passed from the larger P3 fragment across the DIP joint into P2. The wires are retained for 4 weeks. Upon removal, passive extension exercises further compress the two fragments and assist in the final stages of cancellous bone healing. The treatment can still be executed up to 4 to 5 weeks after the initial injury, but the early callus that has formed between the two fragments must be dispersed or satisfactory approximation will not be achieved (Table 30-4).
Closed reduction and splinting is the preferred treatment for most injuries (Fig. 30-20). Should added pin stabilization prove necessary because of recurrent instability, a single longitudinal 0.045-inch K-wire is sufficient. Closed reduction may seem to be impossible. Interposed tissue is usually the cause and may include volar plate, collateral ligament, or tendon. Longitudinal traction rarely is successful in overcoming the blockade. Instead, proximal joint positioning to relax the involved tendons and gentle rotation may allow the interposed tissue to slip out of the joint.
Should open reduction prove necessary, my preferred incision for the DIP/IP joint is dorsal and transverse. The most distal of the major extensor creases corresponds to the joint level. Proximal extensions of 5 mm made in the midaxial lines create a small trapdoor effect that gives ample exposure for any procedure. The terminal extensor tendon or extensor pollicis longus (EPL) should be protected. Using a single prong skin hook is a gentle method to control the tendon without grasping and crushing its fibers with forceps while working to achieve reduction. One must search for small chondral or osteochondral injuries primarily for the purpose of removing the fragments from the joint to prevent subsequent third body wear.
Stable fractures are preferably treated by limited digital splints for 3 weeks or less and protected early motion thereafter with side strapping to an adjacent digit until clinically healed. Unstable but not comminuted fractures of the shaft can be treated well by temporary (3 weeks) closed pinning (Fig. 30-33). There are a few spiral fractures for which closed reduction will not achieve satisfactory control of rotation such that lag screw fixation with 1.2-mm screws is preferable to closed pinning techniques. These treatment strategies are also used in proximal phalanx fractures and more details may be found in that subsequent section of this chapter.
When a dorsal base fracture presents early, extension block pinning is an excellent treatment. The principles are all the same as described above for extension block pinning of dorsal base fractures in the distal phalanx. At the base of P2, the larger dorsal fragment (compared with the base of P3) is easier to work with and manipulate, but the PIP joint (compared with the DIP joint) imposes greater demands for a perfectly congruent joint reduction because of its more important role in overall digital function. The volar articular and shaft fragment is almost always subluxed proximally and volarly. When more than 10 to 14 days have passed since injury, it can be quite difficult (because of early soft tissue contracture) to achieve a closed reduction of this fragment relative to the head of P1. It is for these reasons that late-presenting dorsal base fractures are often better managed with ORIF to ensure the clearance of consolidating hematoma from between the fragments and exact approximation of the articular reduction (Fig. 30-34). In this setting, fixation with two 1.2-mm lag screws usually affords enough stability to pursue early motion. Use of the countersink tap is important to minimize dorsal prominence of the screw heads and to avoid pressure concentration that might comminute the still relatively small dorsal fragment. Even though the surgical procedure occurs distal to extensor zone IV, a priority still must be placed on active extensor tendon excursion during rehabilitation to avoid a long-term extensor lag. In select cases, additional support may be needed in the form of a buttress plate (Fig. 30-34). Intraoperative assessment of the stability of the fixation will guide the progression of rehabilitation to ensure against fixation failure, recognizing the small size of the thread purchase in cancellous rather than cortical bone at the metaphyseal base of P2.
Volar base fractures constituting less than 25% to 30% of the joint surface rarely require surgery unless presenting late with an incongruent joint. When seen acutely, these fractures are well managed with extension block splinting that begins at around 40 degrees and advances 10 to 15 degrees per week for the first 3 weeks. If the extension block splint cannot be eliminated in 3 weeks’ time, this treatment strategy may not be appropriate. Fractures constituting more than 25% but less than 40% of the joint surface pose a difficulty in treatment planning as they are an intermediate group where the disadvantages of the two primary options are relatively well matched. It is difficult to predict in advance how the disadvantages will play out over the course of treatment for an individual patient. The disadvantage of extension block splinting or pinning is that with a greater amount of joint surface involved, the blocking must begin at a higher angle and it will take longer to achieve full extension. A permanent fixed flexion contracture is the consequence to be avoided. This must be compared with the overall tendency for loss of joint motion associated with ORIF or open reconstruction.
When the volar fragment(s) constitute greater than 40% of the joint surface, an open procedure offers the greatest assurance of achieving a congruent joint as a final result. The distinction between the need for ORIF for one or two relatively large fragments or open reconstruction for highly comminuted multiple fragments often cannot be made until the time of surgery. One should always be prepared for both possibilities in the preoperative planning discussions with the patient. Dorsal base fractures usually provide a single fragment of reasonable size for direct lag screw fixation. Volar base fractures are not so easy. One or two large fragments that facilitate lag screw fixation are the exception rather than the rule. In this case, two 1.2-mm lag screws are appropriate. Placement is side by side with one screw in the radial half of the base fragment and the other in the ulnar half. If two separate radial and ulnar volar base fragments are found, this strategy is still acceptable provided that the fragment diameter is at least three times the screw diameter and compression can be achieved without causing fragment comminution. The countersink tap is useful in this regard. With increasing comminution and loss of metaphyseal support for the articular fragments, the original bone may still be salvaged by a volar buttress plate to avoid progressing to the next rung of the reconstructive ladder, osteochondral reconstruction.23
A Bruner incision is made using one limb over P1 and a second over P2. The flexor tendon sheath is reflected as a single rectangular flap hinging on its lateral margin between the distal margin of the A2 pulley and the proximal margin of the A4 pulley. The FDS and FDP are retracted laterally, one to either side, and the collateral ligament origins are dissected as a sleeve from the lateral surfaces of the head of P1. Release of the volar plate allows complete hyperextension of the PIP joint and presentation of both joint surfaces toward the surgeon. This is the so-called “shotgun” approach, and its variations center on the management of the volar plate. This approach is also used for volar plate arthroplasty and ORIF. In the former procedure, the volar plate is released distally so that it may be advanced to replace the defect in the volar articular surface. In the latter, it should remain attached to the fragments as an important source of blood supply. When performing a reconstruction of irreparable comminution, the volar plate no longer has an anatomic connection to the volar base of P2, and complete excision will facilitate restoration of PIP flexion after graft reconstruction. The defect in the volar articular surface may range anywhere from 40% up to almost 90%, often with irregular margins. A small saw or burr should be used to straighten the irregular margins into sharp orthogonal cuts that define a clear bed of cancellous bone in the metaphysis that can be accurately measured for reconstruction. The articular surface at the base of P2 has a sagittally oriented ridge that interdigitates with the groove between the two condyles at the head of P1. This relationship is important not only for preserving joint congruence but also for maintaining stability in the setting of collateral ligament releases. An excellent geometric match has been found in the distal articular surface of the hamate at the ridge that separates the ring from the small finger CMC joints. The measurements taken from the defect at the base of P2 are transposed to the hamate and a small saw and osteotomes are used to remove the osteochondral graft from its donor site. The graft is then exactly trimmed to match the defect and secured with two 1.2-mm lag screws (Fig. 30-35). The joint is checked clinically and radiographically for maintenance of congruence through a full ROM. The flexor sheath is reapproximated with 6–0 monofilament sutures and the PIP joint splinted for protection. Immediate active motion rehabilitation is begun within days of surgery. These same techniques can be extrapolated for use in alternative sites such as the thumb IP joint with the use of a partial FDP-reflecting approach (Fig. 30-36).
Complete articular fractures of the base of P2 may be treated by entirely closed reduction and stabilization. If significant metaphyseal bone loss is present or if the articular fragments at the base of P2 do not reduce sufficiently with traction alone, a small incision can be made through which cancellous bone graft can be added to fill the metaphyseal void and to assist in supporting a reduction of the articular fragments. Transverse 0.035-inch K-wires may be placed at the subchondral level to maintain the articular relationships. The fracture must then be reduced at the metaphyseal level and undergo stabilization sufficient to withstand the rigors of early motion that must accompany the rehabilitation of articular fractures. It is at this point that the significant variations in technique arise along with different devices available for stabilization. My previous preference was for an off-the-shelf unilateral hinged external fixator. The device, which is no longer available, allowed free AROM with a gear disengaged or passive range of motion (PROM) with the gear engaged and the ability to hold and stretch the end points of motion (Fig. 30-37). External fixators for “pilon” fractures are not well received by patients who tend to refuse to move the PIP joint much while the device is in place. For these reasons, my current preference is to treat “pilon” fractures with ORIF, a transition that has been aided by the increasing availability of small locking plates. This is a well-received option by patients, provided that stable fixation is achieved and the result maintained during the stress of therapy (Fig. 30-38).
Once reduced, rotatory volar dislocations, isolated collateral ligament ruptures, and dorsal dislocations congruent in full extension on the lateral radiograph can all begin immediate AROM with adjacent digit strapping. Dorsal dislocations that are subluxed on the extension lateral radiograph require a few weeks of extension block splinting before progressing (however, this is an almost unheard of situation with pure dislocation and no fracture component). Volar dislocations with central slip disruptions require 6 weeks of PIP extension splinting followed by nighttime static extension splinting for 2 additional weeks. The DIP joint should be unsplinted and actively flexed throughout the entire recovery period. Short arc motion of the PIP joint can begin at 4 weeks.
Open dorsal dislocations usually have a transverse rent in the skin at the flexion crease. Debridement of this wound should precede reduction of the dislocation. Any joint debris should be cleared out to prevent third body wear. The “critical corner” warrants particular attention. For closed irreducible joints, unilateral or bilateral midaxial incisions allow excellent access to both volar and dorsal structures without violating the extensor mechanism. Postoperative management follows the same time courses stated above for nonoperative management based on the injury pattern and severity.
CRIF is my preferred treatment for all isolated, closed transverse, and short oblique fractures of the proximal phalanx (Fig. 30-58). Longitudinal pinning with two K-wires passing through the metacarpal head with the MP joint flexed 80 to 90 degrees has yielded reliable results.77 In larger patients, two 0.045-inch K-wires can be fitted into the medullary canal. In smaller patients, one 0.045-inch and one 0.035-inch wire may be more compatible. When rotational interlock is felt between the fragments, one wire can be used. The wires are placed one each on either side of the thick central extensor tendon dorsal to the MP joint, and may be placed just proximal to the sagittal band fibers. The wires are then passed through the base fragment, across the fracture site, and down the distal shaft of the phalanx to the head. Closed pinning is also a valuable technique for nondisplaced fractures at the head of P1. Both a transverse pin connecting the two condyles as well as an oblique pin from the condyle to the opposite diaphyseal cortex should be used for a unicondylar fracture. For bicondylar fractures, two oblique pins are needed. The oblique pins are best cut for retrieval proximally rather than distally as their passage through the periarticular soft tissues will interfere with PIP joint motion. Closed pinning also represents a reasonable treatment option for nondisplaced long oblique or spiral fractures that are suspected of subsequent displacement when subjected to the stress of motion rehabilitation. However, practically I have not found this fracture pattern to exist. I see either truly nondisplaced fractures that I expect to remain stable and treat nonoperatively or displaced long oblique and spiral fractures that I prefer to treat with open reduction.
This is my preferred treatment for long oblique and spiral fractures of the shaft and displaced partial articular fractures (Fig. 30-59). I have found it difficult to correct all the shortening and rotation of long oblique and spiral fractures by closed means alone. There is a natural trade-off between the undeniable added surgical trauma of an open approach and the benefits of an anatomically precise reduction. When lag screws alone are used for fixation, full motion rehabilitation can begin immediately.79,75 This is not the case with K-wires that tether soft tissues, limiting motion, and risking pin tract infection. Performing the open fixation gently and precisely to minimize soft tissue trauma is more easily described than executed. I prefer to operate on closed fractures around postinjury day 3 to 5 when even adult periosteum will thicken dramatically in response to injury and can be surgically manipulated as a tissue flap. A true midaxial incision is in the neutral tension lines of the skin and brings the approach down to the volar leading edge of the intrinsic wing tendon. One of the most important principles in open fixation of a P1 fracture is not to create planes of surgical dissection either superficial or deep to the zone IV extensor tendon. The only dissection that should occur at the subcutaneous level is to identify dorsal cutaneous nerve branches passing obliquely from the proper digital nerves and to mobilize them effectively to avoid neuromas. Other than this, the approach should create a single tissue flap from skin through periosteum to bone.79 A sharp blade is needed to carefully preserve the periosteum for later repair using fine monofilament resorbable sutures. This creates an additional gliding layer of protection for the extensor mechanism. A fine-tipped curette must be used to clear the fracture interface of all clot and soft tissue or a truly anatomic interdigitated reduction will not be possible. In a simple two-fragment diaphyseal fracture, there will be inherent stability between the bone edges once the fracture has been reduced. The role of internal fixation is then to exploit this inherent stability by further compressing the fracture line, which is optimized by interfragmentary lag screws. Although provisional fixation of the fracture with K-wires has been recommended by others, I have found that an absolutely perfect reduction is not well maintained by smooth wires, which invariably allow the reduction to slip a little. This ensures that the drill path will not be exactly in the desired location and that final placement of the screw or plate will thus be imperfect. I prefer to hold the reduction manually with either a bone clamp specialized for the short tubular bones of the hand or with Brown‑Adson forceps. After reduction and provisional stabilization, the steps are core drilling followed by countersinking the near bone surface. Countersinking not only recesses the screw head but also distributes the force of compression, lessening the chance of propagating a new fracture line. Measuring for screw length is done next, and the time for the scrub technician to procure the correct screw can be used to drill the gliding hole. Self-tapping screws are a little difficult to start into bone as some axial load is necessary to get them to bite, but application of this load off the true axis will toggle the screw. A fine touch must be learned over the course of many cases using these implants. Screws are tightened with a “chuck” pinch on the screwdriver using three fingers, not with the more forceful key pinch that may shear the head off the shaft of the screw. The rules for selecting screw-only fixation include fracture length that is at least two times the bone diameter, and fragment width that is at least three times the screw diameter. Adherence to strict principles is mandatory; multiple drill bit passes are not well tolerated by the phalanx. Screws of 1.2- to 1.5-mm diameters are appropriate for P1. Biomechanically, it is desirable to have at least one screw placed perpendicular to the neutral axis of the bone. The remaining screws should be perpendicular to the fracture plane. In a spiral fracture, one screw can satisfy both of these requirements simultaneously and is termed the “ideal” screw. In an oblique fracture this will not be the case.
Plates at the phalangeal level are not desirable because of their bulk and propensity for tendon adherence, and I avoid using them whenever possible. This is my treatment of choice for fractures with comminution and bone loss and complete articular fractures of the phalangeal head that are unstable. Since the biomechanics of P1 fractures create an apex palmar sagittal plane deformity, the plate would (impossibly so) have to be applied to the volar surface of the bone to have its optimum tension band effect. Lateral placement is then the next most desirable option, and this corresponds well to the surgical access that is least harmful to the soft tissues.79 A midaxial incision is carried volar to the margin of the extensor mechanism, straight through periosteum, and the entire soft tissue sleeve is elevated as a single unit, avoiding any dissection of planes surrounding the extensor tendon. When plates are used, one should attempt to place screws as perpendicular to the surface of the plate as possible (Fig. 30-60). The heads of obliquely placed screws have a prominent edge. Plates should also be painstakingly contoured to ensure both the lowest profile as well as proper biomechanical function (preload, dynamic compression, buttress effect). One must not hesitate to remove a plate and recontour it after the first two screws have been placed if it is clear that the shape is not correct. Application of an incorrectly contoured plate guarantees an imperfect fracture reduction. A common error is with plates ending near the metaphyseal flare that must have a small bend at the last hole to accommodate the curvature of the bone at this level. Small-size locking plates represent a tremendous advantage over the only previously available implant that offered fixed angle stability, the minicondylar plate (Fig. 30-61). The minicondylar plate could not be fully contoured to the bone surface before placing the blade; a locking plate can. If more than 50% cross-sectional area of the bone is comminuted or lost, bone graft may be required.
Careful review of the published literature regarding both finger and thumb MP joint ligament injuries indicates that the clinical assessment of instability is paramount in planning subsequent treatment. Local anesthetic injection into the MP joint allows vigorous stress testing of the ligament to be performed without fighting the patient or causing undue pain. Testing in both extension and flexion reveals the absolute value of deviation as well as the discrepancy compared with the uninjured side. The feel at the end point is also a significant piece of information. A greater than 15-degree difference side to side and a soft end point are stronger indicators of complete ligament disruption than the absolute value of the joint angle when stressed. The integrity of the volar plate should be assessed along with the appearance of rotatory subluxation. I use a combination of the clinical degree of instability and the presence of a palpable Stener lesion to choose direct repair of the thumb UCL, the index RCL, and large bony avulsion injuries. The management of complete RCL ruptures is currently in a state of transition in the field of hand surgery. Subluxation into pronation by the proximal phalanx pivoting on the intact UCL has led to increasing interest in early direct repairs. When I can appreciate rotatory subluxation on the examination, I now prefer RCL repair. Volar dislocations risk late instability if not surgically repaired. When the patient presents late following a complete ligament rupture, direct repair is rarely possible. The simplest reconstruction is then to create a proximally based flap of retracted ligament and advance it back to the anatomic insertion at the volar base of the proximal phalanx. This tissue is not always of sufficient quality. When that is the case, a free tendon graft (plantaris or palmaris longus) can be placed through drill holes to reconstruct the ligament. With appropriate rehabilitation, these patients can still achieve near-normal motion.
The border digits, the index and small fingers, can easily be approached with a midaxial incision that offers all the advantages that are proposed for both volar and dorsal approaches. Cartilage injuries on the metacarpal head can be well visualized, the digital nerves are easily protected, and the volar plate can be guided back into its correct position. For the long and ring fingers I prefer a dorsal transverse incision made at the level of the distal portion of the metacarpal head. This level can reliably be found at the dorsal apex of the sloping V shape of the web commissure. The sagittal bands do not need to be divided but rather they can be retracted distally to access the joint. The volar plate can be reduced without dividing it through a combination of wrist flexion to relax the extrinsic flexor tendons and MP hyperextension. A Freer elevator then guides the volar plate to the distal surface of the metacarpal head before attempting to reduce the joint itself. For the RCL of the index, an absorbable 1.3-mm bone anchor can be used for repair of insertional ruptures and a 4–0 absorbable monofilament suture for midsubstance ruptures. Pinning of the joint is not necessary in fingers as adjacent digit strapping provides enough restraint to excessive coronal plane deviation to protect the healing repair. The exception to this is the rare high-energy volar dislocation that is so unstable as to require 3 weeks of transarticular pin fixation.
The operative technique consists of a chevron incision over the ulnar aspect of the MP joint ensuring adequate volar exposure at the base of the proximal phalanx. Care must be taken with the superficial branches of the radial nerve to avoid neuroma formation. There is usually one large branch passing through the surgical field that is best mobilized dorsally. An incision in the adductor aponeurosis is made just ulnar to the EPL tendon with a cuff being left for repair. Reflection of this layer reveals the joint capsule and torn collateral ligament. Whereas all patterns of disruption have been reported, the most frequent is that of distal avulsion from the base of the proximal phalanx. Often there is a transverse rent in the dorsal capsule and evidence of volar plate injury as well. Direct repair is easiest with an absorbable 1.3-mm bone anchor placed at the true insertion site on the volar lateral tubercle to restore normal anatomy and reduce the rotatory subluxation of the joint. The repair may include a suture through the volar plate margin to re-create the “critical corner.” The joint is pinned with a 0.045-inch K-wire before tying the anchor sutures to prevent inadvertent radial deviation and early rupture of the repair during the first 4 weeks postoperatively. A large bone fragment carrying the point of ligament insertion can be stabilized with one or two lag screws (Fig. 30-67). The IP joint should be left free for motion at all times. Motion at the MP joint can begin in a protected fashion at 4 weeks following pin removal and then in an unprotected fashion by 6 weeks. Power pinch activities that stress the ligament in the coronal plane of the thumb should be avoided for up to 3 months after repair.
The approach to the ulnar base of the thumb is the same as for simple repair. The correct anatomic sites of ligament origin at the metacarpal head and insertion at the phalangeal base should be easily discernable, having remnants of the original ligament fibers. Drill tunnels are made from each of these points obliquely directed away from the joint with a 3-mm bit. Free tendon graft may be harvested by conventional methods from either the palmaris longus (within the operative field) or the plantaris (a more appropriate size match). The tendon is passed through each of the drill holes, tensioned, and secured with 3-mm interference screws (Table 30-12).
Many extra-articular and some intra-articular fractures, which are categorized as stable by virtue of having over 30% normal ROM without motion at the fracture site, can be managed with entirely nonoperative means using temporary splinting. Patients with entirely nondisplaced fractures that have excellent inherent stability do not require any external immobilization at all and can begin immediate AROM, usually with the added protection of adjacent digit strapping. Patients with stable metacarpal shaft fractures can be returned to nearly all light activities in a hand-based splint that is continued for a maximum of 3 weeks. Stable neck and intra-articular head fractures are more effectively protected by support that covers from the PIP level to the forearm with the MP joints in full flexion. At least one adjacent digit is included with the affected ray. IP joint motion should begin immediately with all strategies.
Transverse pinning to adjacent metacarpals is my treatment of choice for all unstable closed metacarpal fractures except multiple adjacent fractures at the same level that include a border digit. The biomechanics of the transverse pinning strategy is that of external fixation. Four points of control are needed. The two points closest to the fracture site on either side should be as close together as possible. The two farthest from the fracture site should be as far apart as possible. The proximal intermetacarpal and CMC ligaments are stout enough to qualify as the most proximal point of fixation such that only one 0.045-inch K-wire is required proximal to the fracture site. The distalmost pin should avoid transgression of the sagittal bands. This must be titrated clinically against the goal of placing the point of fixation as far from the fracture site as possible. The transverse pinning strategy works equally well for central (long and ring) and border (index and small) metacarpals (Fig. 30-80). If the four finger metacarpals are thought of as occurring in two columns (a radial column for index and long and an ulnar column for ring and small) then most combinations of multiple metacarpal fractures can still be fixed with this strategy, and it can always be used if there is only one fracture per column. If both metacarpals in the column are fractured, but at different levels, they can be used to stabilize each other reciprocally (Fig. 30-81). The specific requirement for reciprocal stabilization to be effective is that there is a zone in the diaphysis of both bones where two pins can be placed with adequate spacing from each other (distal to one fracture site and proximal to the other). At the conclusion of the procedure, one has the choice of leaving the pins protruding through the skin or cutting them off beneath the skin. In previous editions of this chapter I had advocated allowing pins of less than 4 weeks’ duration to be left outside the skin for ease of removal, but I now cut nearly all pins below the skin level given the prevalence of MRSA in the community. The hand is initially splinted in full MP flexion to resist the development of contractures. Early motion can proceed while the pins are still in place.
ORIF is my treatment of choice for open fractures and multiple fractures not meeting the criteria for reciprocal transverse stabilization. When fracture plane interlock between bone spicules is present, intraosseous wiring, composite wiring, screw-only, or screw and plate fixation may all be considered. I prefer lag screw fixation for long-oblique or spiral fractures since CRIF cannot control the reduction of these patterns nearly as well as transverse fractures (Fig. 30-82). To select screw-only fixation, the ratio of the length of the oblique or spiral fracture plane to the bone diameter must be at least 2:1. Furthermore, to avoid comminution, the screws must pass through an area in the bone spike where the screw’s outer diameter is less than one-third the width of the spike. The screw sizes most appropriate for a metacarpal are 1.5 and 1.7 mm. Multiple open transverse or short oblique fractures of the mid-diaphysis from open crushing injuries are nicely managed with intramedullary pins. Rotational control can be supplemented with a composite wire loop. When interfragmentary compression cannot be achieved owing to the presence of comminution or bone loss, plates and screws are indicated.
As with all techniques of internal fixation, it is essential to cover the hardware with periosteal closure to provide a separate gliding layer. I prefer to operate 3 to 5 days following injury so that the periosteum will have thickened in response to injury and can be both dissected as a discrete tissue flap and closed with solid suture purchase. Unlike over the proximal phalanx, the extensor tendons at this level are discrete cords, and placement of the hardware away from them should be possible in most cases. Placement of the plate dorsally puts it on the tension cortex of the bone, but in this position it interferes most directly with the extensor tendons. Placement of the plate in a true lateral position allows sagittal plane forces to be resisted by the width of the plate rather than its thickness, and doing so is almost always possible, just technically more difficult. This is my choice for plate placement unless extenuating circumstances dictate dorsal placement. One such circumstance is fracture comminution extending all the way to the base of the metacarpal. All the technical comments made in the section on proximal phalangeal fractures apply equally here (Table 30-14).
Pure dislocations rarely occur without fracture of either the metacarpal bases or carpal bones of the distal row. However, the absence of such fractures creates an opportunity for successful management by CRIF. The metacarpal bases must be felt to engage their articulations fully and demonstrate complete congruence on radiographs. Only when the x-ray beam passes tangentially through the joint can an accurate assessment be made. K-wires are retained for 6 weeks with an additional 2 weeks of splint protection before initiating wrist and CMC rehabilitation. All other joints remain mobile throughout the postoperative period.
If an accurate closed reduction can be achieved, CRIF is an excellent choice. Cases seen weeks after injury or those with tissue interposition will likely require ORIF to achieve accurate reduction. The approach may be dictated by the presence of an open traumatic wound. Branches of the superficial radial nerve and dorsal cutaneous branch of the ulnar nerve must be identified and protected not only from the surgical approach but also during pin placement. The common extensor tendons overlie the central metacarpal bases, and the wrist extensor tendons insert on the border metacarpals. Incision of the extensor retinaculum increases the lateral mobility of these tendons, allowing the surgeon to work around them. Bone and cartilage fragments too small for fixation but large enough to create third body wear in the joint should be removed. Fixation is founded upon 0.045-inch K-wire passage from the metacarpal bases across the CMC joints into the distal carpal row. If adjacent metacarpals are stable without CMC joint injury, transverse stabilization between metacarpal bases is an excellent addition. Evaluation of the dorsal cortices of the hamate and capitate should be performed on each case as these are also often fractured. Large bone fragments should be restored to their cancellous beds and fixed with countersunk compression screws. Small bone fragments should be excised.
The literature simply does not support CRIF as a valid treatment for this injury despite basic principles that should allow this method to produce satisfactory results. Although some articles suggest that one needs to perform immediate free tendon graft reconstruction of complete thumb CMC dislocations, this has not been my experience. I have consistently found open ligament repair to produce stable and pain-free motion. The reproducible surgical findings are a sleeve-like avulsion of the deep anterior oblique ligament from the volar surface of the metacarpal and a rupture of the dorsoradial and posterior oblique ligaments. The rupture has usually been distal from the metacarpal insertion. The procedure is easily accomplished by inserting a series of 1.3-mm bone anchors around the margin of the metacarpal base and using the sutures for anatomic repair of the dorsal ligaments. The joint should be pinned in a reduced position before tying down the dorsal ligaments. The deep anterior oblique ligament comes to lie flush with the metacarpal surface when the joint is reduced and stabilized. Pins are retained for 6 weeks with the thumb CMC joint motion instituted at that time. Light pinch is also allowed with progression to power pinch by 3 months postoperatively.
The majority of these can be treated with CRIF, most of the remainder augmented with small openings to control small articular fragments or pack bone graft into the metaphysis, and the final minority with full ORIF. The soft tissue anatomy of this region should be taken into account when placing pins. The drill should not be activated until the pin is solidly placed down to bone. Pins may be placed from the main thumb metacarpal fragment into the small volar fragment, the trapezium, and the index metacarpal in variable combinations based on the unique fracture characteristics of each patient.
The goals of treatment in a Rolando fracture are different. The primary aim is to provide distraction to allow healing through the often-comminuted metaphyseal zone. This is best accomplished by pinning the thumb metacarpal (two 0.062-inch K-wires) to the index metacarpal rather than to the trapezium. It is in these cases of complete articular comminution that making a small opening to place an elevator into the metaphysis may prove useful. The articular fragments can be molded against the distal surface of the trapezium and kept there by either packing bone graft in behind them or with additional smaller caliber (0.035-inch) pins placed transversely at the subchondral level to maintain articular congruity. The advantage of plate and screw stabilization of an intra-articular fracture in general is usually to allow early motion of the joint for the sake of cartilage nutrition and preservation of long-term ROM. The small fragments at the base of the thumb metacarpal are more at risk of devascularization with a widely open procedure that includes periosteal stripping to place a small titanium plate. I have not experienced that long-term loss of motion is a problem at the trapeziometacarpal joint following 6 weeks of pin immobilization for these fractures but I have observed that the presence of the plate results in adherence of the EPL and EPB tendons and can cause long-term loss of motion of both the MP and IP joint, which is a clinically relevant problem.
If ORIF is chosen for a select case, a Wagner incision along the glabrous/nonglabrous border of the thumb base may be curved in a volar and transverse direction to expose the thenar muscle group. Reflection of these muscles reveals the joint capsule volar to the insertion of the APL. Arthrotomy reveals the intra-articular fracture, and subperiosteal dissection along the shaft allows for placement of a plate. Stable internal fixation of Rolando fractures is only possible when the fragments are large enough to accept the purchase of individual screws. My current fixation of choice in this situation is a titanium locking condylar plate (either 1.7 mm or 2.3 mm depending on the size of the patient). Eccentric drilling of the condylar holes can add transverse compression between the articular base fragments. If ORIF is chosen for a Bennett fracture, a smaller version of the same approach is used to allow sufficient access to compress the reduction and place an interfragmentary lag screw from dorsal to volar. Micro-sized variable pitch headless screws are also well suited to the fragment sizes seen in a Bennett fracture (Table 30-16).