General Principles of Internal Fixation

Throughout history, fractures have been treated with immobilization, traction, amputation, and internal fixation. Immobilization by casting, bracing, or splinting a joint above and below the fracture was used for most long-bone fractures—except femur fractures, for which traction was the mainstay of treatment. In the past, open fractures and ballistic wounds with long-bone fractures were not amenable to standard fracture care because of the associated soft-tissue injury and the difficulty of preventing sepsis; thus, they usually resulted in amputation.

Although the concept of internal fixation dates back to the mid-1800s, Lister introduced open reduction and internal fixation (ORIF) of patella fractures in the 1860s. Use of plates, screws, and wires was first documented in the 1880s and 1890s. Early surgical fixation initially was complicated by many obstacles, such as infection, poorly conceived implants and techniques, metal allergy, and a limited understanding of the biology and mechanics of fracture healing.

During the 1950s, Danis and Muller began to define the principles and techniques of internal fixation. Since then, advances in biologic and mechanical science have led to contemporary fixation theories and techniques.

Disruption of the endosteal and periosteal blood supply occurs with the initial trauma, and maintaining adequate blood supply to the fracture site is essential for healing. Hunter described the four classic stages of natural bone repair, as follows:

• Inflammation
• Soft callus
• Hard callus
• Remodeling

The inflammation stage begins soon after injury and appears clinically as swelling, pain, erythema, and heat. Disrupted local vascular supply at the injured site creates a hematoma and prompts the migration of inflammatory cells, which stimulate angiogenesis and cell proliferation.

After the initial inflammatory phase, the soft callus stage begins with an infiltration of fibrous tissue and chondroblasts surrounding the fracture site. The replacement of hematoma by this structural network adds stability to the fracture site.

Soft callus is then converted into rigid bone, the hard callus stage, by enchondral ossification and intramembranous bone formation.

Once the fracture has united, remodeling begins. Fibrous bone is eventually replaced by lamellar bone. Although this process has been called secondary bone union or indirect fracture repair, it is the natural and expected way fractures heal. Fractures with less than an anatomic reduction and less rigid fixation (ie, those with large gaps and low strain via external fixator, casting, and intramedullary [IM] nailing) heal with callous formation or secondary healing with progression through several different tissue types and eventual remodeling.

Anatomic reduction and absolute stabilization of a fracture by internal fixation alter the biology of fracture healing by diminishing strain (elongation force) on the healing tissue at the fracture site. Absolute stability with no fracture gap (eg, via open reduction and internal fixation using interfragmental compression and plating) presents a low strain and results in primary healing (cutting cone) without the production of callus.

In this model, cutter heads of the osteons reach the fracture and cross it where bone-to-bone contact exists. This produces union by interdigitation of these newly formed osteons bridging the gap. The small gaps between fragments fill with membranous bone, which remodels into cortical bone as long as the strain applied to these tissues does not cause excessive disruption and fibrous tissue develops (nonunion).

This method of bone healing is known as direct bone healing or primary bone union. Essentially, the process of bone remodeling allows bone to respond to the stresses to which it is exposed.

On the basis of the mechanical milieu of the fracture as dictated by the surgeon's choice of internal fixation and the fracture pattern, two patterns of stability can result that determine the type of bone healing that will occur:

• Absolute stability (ie, no motion between fracture fragments) results in direct or primary bone healing (remodeling)
• Relative stability (ie, a certain amount of fragment motion) heals with secondary or indirect bone union

Pins and wires

Kirschner wires (K-wires; 0.6-3.0 mm) and Steinmann pins (3-6 mm) have various uses, from skeletal traction to provisional and definitive fracture fixation. Resistance to bending with wires is minimal; thus, they are usually supplemented with other stabilization methods when used for fracture fixation. Most commonly, wires are utilized as provisional fixation before definitive fixation with a stronger device. Skeletal traction with K-wires is possible with the use of a K-wire tensioner, which, with application, stiffens the wire and allows it to resist bending load. ^[1]

K-wires and Steinmann pins can provide provisional fixation for reconstruction of fractures while incurring minimal bone and soft tissue damage and leaving room for additional hardware placement. Planning pin placement is important to avoid the eventual permanent fixation devices, and if possible, pins should be placed parallel to screws used for fracture compression. Depending on the diameter, pins may also be used as guide wires for cannulated screw fixation.

Permanent fixation options include fractures in which loading is minimal or protected with other stabilization devices, such as external fixators, plates, and braces. Pin or wire fixation is often used for fractures of the phalanges, metacarpals, metatarsals, proximal humeri, and wrists. ^{[2, 3]} K-wires commonly supplement tension-band wire constructs at olecranon, patella, and medial malleolus fractures.

K-wires can be fully threaded or nonthreaded and have diamond or trocar points that are simple in design and have limited ability to cut hard bone, a process that can lead to overheating. Thus, when power equipment is used, they should be inserted slowly to avoid thermal necrosis. Image intensifiers may be used for optimal positioning, especially when percutaneous insertion is combined with closed reduction. The pins may have points at both ends, facilitating antegrade-retrograde fixation; however, they are a potential hazard and should be used with caution.

Steinmann pins are larger, may be threaded or unthreaded, and are currently used primarily for long bone traction in conjunction with a Böhler traction stirrup. Early techniques of fracture treatment consisting of pins for skeletal traction and incorporation into a cast were fraught with pin infections, loosening, and loss of reduction. This technique has been replaced with more advanced external fixation devices, internal fixation methods, and minimally invasive plating and intramedullary (IM) devices.

Guide wires for cannulated screws are employed at times for definitive fixation; they are terminally threaded, allowing fixation on the opposite cortex. An example of this would be the closed reduction and percutaneous pinning technique for proximal humeral fractures.

Screws

Bone screws are a basic part of modern internal fixation and can be used independently or in combination with particular types of implants. ^{[4, 5]} The common design (see the image below) consists of a tip, shaft, thread, and head.

A round screw tip requires pretapping, whereas a fluted tip is self-tapping. The screw shaft is located between the head and the threaded portion of the screw. The screw thread is defined by the following variables (see the image below):

• Major or outside (thread) diameter
• Minor or root (shaft) diameter
• Pitch
• Lead
• Number of threads

The root diameter determines the screw's resistance to breakage (tensile strength). Screws are referred to by their outer thread diameters, bone type for intended use (cortical or cancellous, determined by pitch and major/minor diameters), and proportion of thread (partially or fully threaded).

Pitch, the distance between adjacent threads, affects purchase strength in bone. Increasing the pitch increases bone material between the threads but decreases the number of threads per unit of distance. The lead is the distance a screw advances with a complete turn. Lead is the same as pitch if the screw is single threaded, and lead is twice the pitch if the screw is double-threaded (faster screw insertion).

Screw pullout strength can be affected by several factors. Bone composition (density) is the primary determinant of screw fixation. The total surface area of thread contact to bone (root area) is another factor in pullout resistance.

Pretapping the screw hole theoretically reduces microfracture at the thread-bone interface but requires an extra step for insertion. Self-tapping screws have been shown to have no clinical difference from pretapped screws for fracture or plate fixation, eliminate the tapping step, and are now the industry standard. Because the fluted portion of the screw tip has less thread contact with the bone, slight protrusion at the opposite cortex is recommended.

The industry standard for the screw head is a hexagonal recess, which provides a large contact surface between the screw head and screwdriver and allows for optimal transmission of torque. A cross-type screw head is used on some screws in the 2.0 and smaller (minifragment) sets. The star design (or Torx) found in industry has been adapted to the screw heads for the Association for the Study of Internal Fixation (AO/ASIF) locking plates and has been shown to be superior for torque and resistance to stripping.

Several forces are involved with screw insertion and tightening. Torque is applied through the screwdriver to the screw head in a clockwise rotation to advance the screw in the predrilled path or, in the case of a cannulated screw, over a guide wire; this advancement produces a circumferential force along the thread. For cortical screws, the drill diameter is slightly larger than the root (shaft) diameter of the screw.

Axial tension is created with impingement of the screw head on the cortex or plate, generating tension through the screw. To optimize these forces, screws should ideally be inserted at 80% of the torque needed to cause them to strip. An estimated 2500-3000 newtons of axial compression force can be applied to the average screw.

Over time, the amount of compressive force decreases slowly as the living bone remodels to the stress; however, the fracture healing time is usually shorter than the time it takes for substantial loss of compression and fixation.

Two basic types of screws are available for use in bone of differing density:

• Cortical screws, designed for compact diaphyseal bone
• Cancellous screws, designed for the more trabecular metaphyseal bone, which is softer

Cortical screws have a smaller major (thread) diameter, a smaller pitch, and a shallower thread than cancellous screws do. Standard nonlocking cortical screw diameter choices include 1.5, 2.0, 2.7, 3.5, and 4.5 mm. Cancellous screws typically have a larger major (thread) diameter and pitch and a greater difference between major and minor (shaft) diameters in comparison to cortical screws, providing more surface area for bone purchase. They are available in sizes 4.0 and 6.5 mm, and cannulated sizes vary from 4.0-7.5 mm.

Tapping is not usually necessary in metaphyseal bone, because cancellous bone is porous relative to compact diaphyseal bone and usually requires only the initial pilot hole or cannulated screw guide wire. With subsequent screw insertion, there is compression of the bone along the path of the threads, which increases the local bone density in contact with the thread, thereby potentially increasing screw purchase. Tapping may be considered in strong metaphyseal bone to avoid stripping if advancement of the screw is difficult.

Positional or neutralization screws are to attach an implant, such as a plate, to bone by compression between the plate and bone (see the image below). This function is modified when the screw is used to lag across a fracture through the plate or when used for fracture compression, as with a dynamic compression screw.

For a positional screw, the pilot hole is drilled with the appropriate-size bit (shaft diameter) for the screw to be inserted (eg, a 3.2-mm drill bit for a 4.5-mm screw), with a centering guide used for the plate hole. A depth gauge is used to determine appropriate screw length, and the thread cut is then made with an appropriate tap or without a tap when self-tapping screws are used or screws are placed in metaphyseal bone.

Interfragmentary lag screws provide compression across two bone surfaces using the lag technique. A lag screw is a form of static compression and is applicable to intra-articular fractures to maintain reduction and diaphyseal fractures for stability and alignment. Ideally, lag screw fixation (see the image above and the image below) produces maximum interfragmentary compression when the screw is placed perpendicular to the fracture line.

Most lag fixation techniques require additional stabilization to neutralize the axial bending and rotational forces applied to the bone during functional postoperative care. This is provided by a neutralization or buttress plate or external fixation.

If lag screws are to be used without neutralization plate fixation, especially in long spiral fractures (>2 times the diameter of the involved bone), the ideal inclination of the screw is halfway between the perpendiculars to the fracture plane and to the long axis of the bone. Placing the screw perpendicular to the long axis of the bone can also be considered, because longitudinal or shear compression may cause the screw or screws to tighten.

Interfragmentary screw fixation alone may be appropriate for avulsion injuries in which shear forces generate metaphyseal and epiphyseal intra-articular fractures, provided bone quality is good.

A fully threaded screw can serve as a lag screw (see the image below) with the near cortex overdrilled to the size of the screw's major (thread) diameter (4.5 mm in the example). Once the near cortex is drilled, which provides a gliding hole, a drill sleeve with the outer diameter of the drill bit (4.5 mm) is inserted into the hole and the standard drill bit (3.2 mm shaft diameter) is used to drill the far cortex.

As the screw threads grasp the distal cortex, compressive forces are generated through the axis of the screw to the screw head, causing the fracture fragments to be compressed. This same mechanical effect can be generated by a partially threaded screw, with all threads entirely within the opposite bony fragment.

Cannulated screws are now provided by most trauma manufactures in sizes from minifragment to 7.5 mm, usually with a cancellous thread, but cortical patterns are also available, as they are more commonly used in periarticular/metaphyseal bone. The guidewire is usually placed under fluoroscopic control and allows for initial provisional fixation.

Cannulated screws are amenable to a percutaneous technique, such as is used with hip pinning, ^[6] or may be used with limited open reduction techniques and can help minimize soft tissue dissection and periosteal stripping. Most designs are now self-drilling and self-tapping, but some may require predrilling over the guide wire in areas with dense bone.

The guide wires are usually terminally threaded, though nonthreaded wires are also available. In drilling over the guide wire, it is recommended not to drill over the threaded portion, because the wire may be inadvertently removed along with removal of the drill bit. This could result in difficulty relocating the drill hole through soft tissue or loss of provisional fixation.

The pullout strength of cannulated 7-mm cancellous screws was tested against that of 7-mm noncannulated screws and 3.5-mm cannulated and noncannulated screws in two studies; no significant difference was noted. However, the studies were specific to those screw designs, and other designs and sizes cannot necessarily be assumed to have similar fixation properties. It should be kept in mind that cannulated screws are often 10 times as costly as similar-sized noncannulated screws; thus, noncannulated screws should be used when technically feasible.

Self-tapping screws have the advantage of eliminating a step during screw insertion, thereby decreasing operative time. The fluted design of the screw cuts a sharp path in the predrilled hole, eliminating the need for tapping. Baumgart et al showed that insertion torque and pullout strength were comparable for tapped and self-tapping screws. Only if the cutting tip did not protrude through the second cortex did they find a reduction of pullout strength of approximately 10%.

Schatzker et al demonstrated that self-tapping screws inserted at 80% of thread-stripping torque and then removed and reinserted 12 times lost no significant holding power. When a self-tapping screw is inserted as a lag screw, care must be taken to avoid missing the opposite cortex; the screw is often at an angle to the diaphyseal shaft, or it may prove hard to advance the screw while also tapping, especially with hard cortical bone. It is reasonable to consider tapping the opposite cortex first to help with alignment and advancement of the lag screw.

Locked screws are incorporated in more recent plate designs and may be inserted as unicortical or bicortical screws, depending on the type of plate and fracture. These screws, with reduced pitch, produce minimal axial force, if any, and provide biomechanical fixation by locking the screw head into the plate with a tapered thread, perpendicular to the plate.

Some newer designs allow for some variable angulation of the locking screws. Biomechanically, locking screws function more like a bolt than like a screw (see the first image below), and the system acts generally like an internal-external fixator (see the second image below). (See Plates below.)

Plate types

Plates are provided in various sizes and shapes for different bones and locations. Dynamic compression plates (DCPs) are available in 3.5 mm and 4.5 mm sizes. The screw holes in a DCP are shaped with an angle of inclination on one side away from the center of the plate. When tightened, the screw head slides down the inclination, causing movement of the bone fragment relative to the plate (see the image below).

As one bone fragment approaches the other at the fracture, compression occurs. The shape of the holes in the plate allows 25° of inclination in the longitudinal plane and 7° inclination in the transverse plane for screw insertion. ^[7]

Limited-contact DCPs (LC-DCPs) were designed to limit possible stress shielding and vascular compromise by decreasing plate-to-bone contact by 50% (see the image below). ^[8] Theoretically, this leads to improved cortical perfusion with increased preservation of the periosteal vascular network and reduces osteoporosis under the plate.

The regular DCP has an area of decreased stiffness located at the plate holes and, with bending, has a tendency to bend at the holes with a segmented pattern, whereas the LC-DCP, with a different geometric design incorporating the holes and plate undersurface, allows gentle bending distributed throughout the plate (see the image below).

Finally, the LC-DCP is designed with plate-hole symmetry, providing the option of dynamic compression from either side of the hole and allowing compression at several levels.

In general, standard DCPs were replaced years ago by most manufacturers with updated designs that were variations on the LC-DCPs, and these plates in turn have been replaced by all manufacturers with plates capable of both locking and nonlocking functions. Some specific nonlocking-style plates are still retained in use because they function well for a variety of specific fractures, such as the one-third tubular plate for lateral malleolar fractures and the 3.5-mm recon plates for periacetabular fixation.

Techniques for the application of both DCPs and LC-DCPs are the same (see the image below). Screws can be inserted in neutral position or a compression position, depending on the desired mechanical result. The DCP uses a green guide to insert a neutral screw, which adds some compression to the fracture owing to the 0.1-mm offset. The gold guide produces a hole 1 mm off center, away from the fracture, and allows for 1 mm of compression at the fracture site with tightening of the screw.

The LC-DCP universal drill guide allows for either neutral or eccentric placement of screws. When creating an eccentric hole to one side or another, the guide is slid to the end of the plate hole without applying pressure and the hole is drilled. By placing pressure against the bone with the drill guide, the spring-loaded mechanism allows centralization of the hole for neutral screws, particularly if the screw must be inserted at an angle to the plate.

The 3.5 one-third tubular plate is 1 mm thick and allows for limited stability (see the image below). The thin design permits easy two-dimensional contouring and is primarily used on the lateral malleolus and, on occasion, the distal ulna, though the locking version may be a better option. The oval holes allow for limited fracture compression with eccentric screw placement.

Improvements by all manufacturers have been made for plates used for almost all areas of the body that require placement of a plate near a joint and over extended areas of diaphyseal bone. The refinement of contour, along with screw head modification, reduces hardware prominence and increases fixation options.

The 95° angled plates are useful in the repair of metaphyseal fractures and reconstruction of the femur (see the image below), providing very rigid fixation. Placement is technically demanding, and proper insertion requires the blade to be inserted with consideration of three dimensions (ie, varus/valgus angulation, anterior/posterior position, and flexion/extension rotation of the plate). The screw barrel devices are considered somewhat easier to insert because the flexion/extension of the plate is correctable after insertion of the screw.

Reconstruction plates are thicker than one-third tubular plates but not quite as thick as DCPs (see the image below). Designed with deep notches between the holes, they can be contoured in three planes to fit complex surfaces (eg, around the pelvis and acetabulum). Reconstruction plates are provided in straight and slightly thicker and stiffer precurved lengths. Like tubular plates, they have oval screw holes, allowing the potential for limited compression.

Cable plates incorporate a large fragment plate with cerclage wires to be used with a tensioning device. These are used primarily with femoral fractures surrounding or adjacent to prosthetics (femoral hip or knee implants). Cortical allograft struts are often incorporated for osteoporotic bone.

Plate functions

Standard plate fixation requires exposure of the fracture site, hematoma evacuation, and reduction of the fracture with possible interfragmentary lag fixation. After a fracture occurs, the periosteal blood supply is dominant, and this network of connective tissue must be preserved to optimize healing. Excessive periosteal stripping and careless soft tissue techniques can impair local blood supply and prolong healing.

Diaphyseal plate fixation associated with an anatomic reduction and interfragmentary compression provides absolute stability. Plates are often indicated in articular fractures to neutralize the axial forces on the interfragmentary screws, compressing cancellous bone to facilitate its healing. ^[9] A fracture anatomically reduced without a gap and fixed with absolute stable fixation will undergo primary healing.

Dead bone at the fracture site is resorbed by osteoclasts of the cutting cones, which traverse the fracture site. The osteoclasts are closely followed by ingrowth of blood vessels and mesenchymal cells and osteoblast infiltration. Stress shielding of the bone is rarely caused by the plate relieving axial load to the bone. Plate-induced osteoporosis is caused by disruption of the local vascularity to the bone cortex secondary to an impediment of centrifugal cortical blood flow by the plate.

Osteoporosis under a plate should be kept in mind after removal of hardware, because the bone also has the mechanical disadvantage of empty screw holes. This vascular-caused cancellization of the cortical bone in diaphyseal areas usually resolves within 2 years of plate application; consequently, it is safe to remove a plate at this time, with the refracture rate being minimal. Plates applied to metaphyseal areas may have the option of earlier removal, depending on the amount of diaphyseal extension and healing.

Bridge plating is used for comminuted unstable fractures in which anatomic restoration and absolute stability cannot be achieved. Minimal exposure and indirect reduction techniques are used to preserve the blood supply to the fracture fragments for healing, and a plate is attached to the two main fragments spanning the area of fracture. The plate is used to provide proper length, axial alignment, and rotation, but it is obviously limited for any load.

With subsequent advances of combining minimally invasive plate techniques utilizing locking plate technology, plate devices act more as an internal fixator. This approach began in 2001 with the Less Invasive Surgical Stabilization (LISS) plate (Synthes, West Chester, PA), which is advanced in the submuscular tissue through a small incision over the periosteum but does not necessarily contact the bone along the length of the plate.

This technique limits the disruption of periosteal blood supply that is seen in conventional plating systems because the fixation is through the locking screws and thus does not necessitate compression to the plate for stability. The early development of this concept with the Point Contact Fixator (PC-Fix) system (Synthes) in the 1990s, and then later with LISS, took advantage of unicortical, self-drilling, self-tapping screws with threaded screw heads that lock into the screw hole of the plate and minimize soft tissue disruption. ^[10]

Once the LISS plate is aligned with the central shaft of the bone, screw placement can be accomplished percutaneously with a radiolucent guide attachment to the plate. Unicortical screws are recommended for use in diaphyseal bone, with longer screws for use in the metaphyseal area, thereby functioning as a fixed-angle device.

Currently, most manufacturers offer locking plate products. ^{[11, 10, 12, 13]} These devices range from standard straight plates of all sizes with locking and standard screws to anatomically specific plates that act as fixed-angle devices. These newer plate designs incorporate improved contour with locking screw options for fixation, offering significant advantages over the conventional designs for certain fractures. ^[14]

Proximal and distal humerus, distal radius, distal femoral, and proximal (bicondylar) and distal tibial fractures are examples of injuries that benefit from this technology, having the improved ability to hold a fracture in its anatomic position and resist applied forces while healing. Conventional plates, which rely on friction forces against the plate from screw fixation and buttressing in metaphyseal and articular fractures, are limited in resisting applied loads versus locking fixation.

In contrast, certain shaft fractures with stable patterns and adequate room for fixation have proven high union rates with conventional plating (humeral shaft, radius, and ulna shaft), and with proper surgical technique, any significant difference between the two techniques is difficult to realize.

Current recommendations are to use locking screws in situations with limited fixation options, osteoporotic bone, or need for fixed-angle support. For example, a simple lateral plateau fracture that requires buttress fixation and with which the bone quality is reasonable can be adequately treated with a conventional nonlocking lateral plate.

Currently, most LC-DCP small and large conventional plate sets have been reduced as utilization of specialty plates has increased with periarticular design and locking capability, the surgeon deciding which screws are locking or nonlocking, depending on the fracture.

Like cannulated screws, locking screws can vary in cost, ranging from eight to 15 times the cost of a conventional screw, and therefore should be used when needed in accordance with the fracture pattern and expected loads. This cost issue is lessened to some degree when the need for revision surgery due to failure of fixation or malunion is taken into account; thus, a balance of usage guided by conventional wisdom, common sense, and biomechanical and outcome studies is recommended.

Plates and other constructs can be used to function as a tension band if an eccentrically loaded bone (eg, the femur) has the device placed on the tension (convex) side of the bone. Using load-strain diagrams, Frederic Pauwels, who first described the tension-band concept, showed that a curved tubular structure placed under an axial load had a tension side and a compression side. With this theory, he described the application of internal fixation on the tension side to convert tensile forces into compressive forces at the fracture site. ^[15]

With static compression applied by the implant (eg, tensioning of wire, compression with plate), dynamic compression then develops with joint flexion, as with a patella or olecranon fracture, or with load, as with lateral femoral plating (see the image below). With this technique, the internal fixation device must have the strength to withstand the tensile distraction forces created by muscles during motion, and the bone on the opposite side of the plate must be able to withstand the compressive forces as a medial buttress. ^[16]

Wires and plates are usually quite strong under pure tension forces, but with bending forces added, fatigue can occur rapidly. If bony support is compromised on the cortex opposite from the tension device (eg, from fragmentation, osteoporosis), bending stresses can develop, causing failure of fixation. Wiring techniques commonly include longitudinal K-wires for rotational and axial alignment control in the case of bone fragmentation.

Conversely, fixation on the concave side of the bone occurs in rare situations, such as medial plating of a femur or anterior plating of the humerus. In these situations, fractures have minimal resistance to bending stresses, and gapping can occur on the convex side, resulting in failure of fixation (see the image below). Therefore, attempts should be made to limit potential bending forces to fixation to prevent fixation failure. The tension-band principle can be applied to wires, cables, suture, plates, and external fixators as long as the basic principles are followed. ^[17]

In the 1930s, Küntscher refined nailing techniques, with the result that IM nails became the standard for femoral shaft fixation. ^{[18, 19, 20]} Later developments resulted in IM devices being options for proximal and distal metaphyseal/articular fractures and for tibial and humeral fractures.

IM nails allow stable fixation of diaphyseal fractures with early mobilization of joints, early ambulation, and weightbearing of extremities. As metallurgy and designs have improved, the indications and techniques for IM devices have increased. Specially designed nails now exist for each bone, different entry portals, and specific fracture patterns. IM nails have advantages over plates and external fixation because the intramedullary location allows for axial alignment and load sharing. ^[21]

The location and type of fracture determines the device to be used, and devices are named accordingly. IM devices can be described in terms of length, diameter, curvature, locking options, cross-sectional geometry, material, and insertion site options as determined by the bone and fracture being addressed.

A nonlocking cloverleaf Küntscher nail is an example of a centromedullary nail, which is inserted in line with the femoral canal and relies on longitudinal interference with bone-to-nail contact at multiple points to maintain axial and rotational stability of the fracture.

Condylocephalic nails such as Ender pins or Rush rods were a significant device in the early years of fracture fixation. These solid devices were smaller in diameter and were inserted in the condyles or the metaphyseal region, advanced across the fracture either antegrade or retrograde, and embedded in the opposite metaphysis for stability. These nails were usually inserted in clusters of two to four for bending stability, but they had limitations with rotational and axial forces. ^[22]

Initial simple IM devices relied on reestablishing bony realignment and contact along with interference fit in the medullary canal for stability. This was enhanced by the cloverleaf designs, which added a dynamic lateral compression within the canal for additional stability. As nail designs progressed, interlocking options were added, which improved the stability and expanded the fracture fixation options, increasing their indications. ^{[23, 24, 25]}

Interlocking screws increase the working length of the nail from a simple interference fit, not attainable with nonisthmal shaft fractures or fractures without stable bony contact, to semirigid fixation at the ends of the nail, which is capable of resisting axial and rotational forces. A nail's working length corresponds to the fracture areas between the sites of fixation and thus can range from several millimeters with a simple transverse fracture to the entire length of nail between the locking screws in fractures with fragmentation or an unstable pattern.

The working length of the nail is increased when the locking screws are located as close to the ends of the nail as is structurally possible, increasing the potential fracture indications. By the 1980s, examples of second-generation interlocking nails included the Grosse-Kempf nail and, later, the Russell-Taylor nail. Currently, all nail manufacturers include basic interlocking screws and other notable features on third-generation nails, such as proximal femoral head/neck screws and dynamic screw slots.

Reconstruction-type nails and gamma-style nails with a reinforced proximal section that allow fixation into the femoral head and neck region are cephalomedullary nails. These nails increase the fixation options for proximal femoral fractures. Recon nails are a variation of a standard piriformis-start antegrade femoral nail, whereas cephalomedullary nails are devices starting at the tip of the greater trochanter, which is not in line with the anatomic axis of the femur; this explains the increased size required to accommodate the larger proximal fixation screw and the stress from the offset position.

Tibial nails have also evolved over the years in a similar fashion. With the introduction of locked femoral nails, the same principles of static and dynamic locking were applied to the tibia. By changing nail design and improving the metallurgy, more configurations for locking were made possible, thus expanding the indications for tibial nailing to the proximal and distal end segment of extra-articular fractures.

Locking configurations can be static or dynamic. A statically locked nail implies the presence of proximal and distal screws in a nonslotted hole, permitting control of axial translation and allowing for rotation, with the nail performing more as a loadbearing implant. This application is appropriate for unstable fracture patterns or locations and is certainly a consideration if immediate full weightbearing is needed, as is sometimes the case in patients with multiple traumatic injuries.

As with any fracture reduction, attention to accurate length restoration and rotation is important for avoiding malreduction and leg-length inequalities. Avoidance of fracture distraction is important to minimize the risk of delayed union or nonunion, especially in the humerus and tibia. ^[26]

Dynamic locking allows the shaft to axially translate several millimeters while rotational control is maintained. This was originally accomplished by leaving the locking screw hole farthest from the fracture empty. This is rarely performed now. Brumback et al demonstrated that dynamic locking leads to malunions, and they recommended static locking for all long bone fractures treated with IM fixation.

Currently, nails are constructed with a slotted locking screw hole, allowing placement of the locking screw so that the nail moves along the slot (~5 mm) while the screw controls rotation. With these improved nails, a dynamic option for fractures with an obviously stable fracture pattern (eg, isthmic location, Winquist fracture pattern II or less) is available to help stimulate healing with axial loading. A statically locked nail may be converted to dynamic lock by removing the static position screws at one end of the nail.

Cross-sectional geometry varies widely with manufacture and design and with fracture indication. Nails may be solid, open-section (slotted), or solid-section cannulated of various shapes, including cylindrical, square, triangular, cloverleaf, and multigrooved or multifluted. Solid nail designs may be necessary for smaller-diameter devices, but they are not amenable to insertion over a guide wire and are difficult to extract if broken. Additionally, more recent femoral designs have been replaced with cannulated versions.

Bending and torsion strength is altered by changing wall thickness, materials, and, possibly, the number of (adding) channels or slots. A channel along the length of the nail potentially allows for revascularization, but with the advent of locking screws, the sharp flutes or edges of earlier nail designs were not necessary for rotational control.

Torsion and bending resistance in a cylindrical structure is proportional to the fourth power of its radius. When the radius is increased away from the load axis by a thicker wall or a greater diameter, rigidity increases. Increasing the diameter of an IM nail by 1 mm increases its rigidity by 30-45%, but this would require additional reaming of the canal. Excessive reaming may weaken the diaphyseal bone and increase the possibility of thermal necrosis.

For torsion, rigidity decreases inversely to the working length, and with bending, the stiffness is inversely proportional to the square of the working length; therefore, the shorter the effective working length of the nail fixation and fracture combination, the stiffer the device.

IM implants provide stable fixation, but healing occurs primarily through the formation of periosteal callus. Reaming of the medullary canal increases the working length of an IM implant by elongating the isthmic region with a uniform diameter, thereby increasing the potential implant-to-bone contact. In addition, this allows insertion of a larger-diameter and stronger nail than would be possible without reaming; this often allows larger-diameter locking screws, decreasing potential implant failure.

Reaming of the medullary canal damages the medullary vascular system and increases the IM pressure and temperature, with devitalization and necrosis of the diaphyseal cortical bone. In animal studies of blood flow to long-bone diaphyseal regions, reaming can cause necrosis of the inner half of the cortex, but this is followed by a strong hyperemic response in the periosteal and muscular blood flow. These changes appear to be reversible over a 12-week period. ^[27]

Diaphyseal reaming also weakens the bone, and the recommendation is that the cortex should not be reamed to less than half of its original thickness. Additionally, any instrumentation of the medullary canal, including placement of a guide wire and reaming, embolizes marrow contents to various organs, including the pulmonary system.

IM pressure can be reduced by the presence of a fracture, slowing the rate of reamer insertion, increasing the speed of the reamer, and allowing the reamer tip to incorporate a small shaft relative to the diameter of the reamer, with deep flutes designed for depressurization of the canal. Although this type of embolization is performed in humans undergoing transesophageal echocardiography, its clinical significance is still debated with regard to its effect on pulmonary function in patients with multiple injuries.

Unreamed nailing has been studied as an alternative to reamed nails, and various studies have demonstrated better preservation of endosteal blood supply and more rapid revascularization than is the case with reamed techniques. This advantage is limited. Blood flow rapidly improves with reamed fixation, provided that the soft-tissue envelope is adequate. Clinical studies have revealed better healing rates for both femoral and tibial fractures (except for severe open injuries) with reamed nails than with nonreamed nails.

In North America, the standard practice is to insert reamed IM nails in all closed femoral and tibial diaphyseal fractures. This is contraindicated in patients who have been in shock, have pulmonary compromise, have elevated serum lactate levels, and have abnormal base deficits and also have multiple injuries.

Open fractures are also amenable to reamed nailing. Grade IIIB open fractures may be a relative contraindication. Humeral nailing still presents problems with union, shoulder stiffness, and neurologic injury when inserting locked screws, so it is not as popular as with the other long bones.

Polymers, including polylactic and polyglycolic acids and polydioxanone, are resorbable suture materials that are currently undergoing continued redesign and refinement for use as rods or screws that are reabsorbed with time. These devices offer the theoretical advantage of eventual resorption, eliminating the need for later removal, while allowing stress transfer to the remodeling fracture.

Current bioabsorbable implants do not have mechanical properties to match metallic implants; therefore, their indications are limited, and their fixation usually requires protection from motion or significant loading. Degradation rates vary, and local inflammatory reactions, such as chondrolysis noted with placement in proximity to joints, have been reported with some implants.

These devices are a consideration when fixation of low stress areas is needed and when later removal is anticipated, such as in pediatric patients ^[28] or in medial malleolar fractures, syndesmotic fixation, or osteochondral fractures in adults. ^{[29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39]}