Diaphyseal Femur Fractures

Updated: Mar 23, 2023
  • Author: Bart Eastwood, DO; Chief Editor: William L Jaffe, MD  more...
  • Print

Practice Essentials

Orthopedic surgeons often encounter diaphyseal femur fractures. [1]  Because these fractures most often result from high-energy trauma, one must have a high index of suspension for complications or other injuries. The mainstay of treatment has been reamed interlocking intramedullary nailing, but a variety of treatment options now exist for solitary fractures or fractures with associated injury.

Before the 1900s, diaphyseal femur fractures were treated with various types of splinting. However, with the discovery of skeletal radiology near the end of the 19th century came an understanding of the forces acting on fractured bones and a change in the treatment of such injuries. Steinmann in 1907 and Kirschner in 1909 developed the first traction treatment modalities with the use of pins or wires under tension.

Early attempts at internal fixation of such fractures achieved little success until Küntscher developed and utilized the intramedullary nail in 1937. After a short period of disagreement, the nailing method began to spread during World War II in Europe and later in North America. Intramedullary nailing became prominent in the United States in the 1970s. Since the intramedullary nailing technique was introduced in 1939, it has continued to evolve into the antegrade reamed interlocking nails that are the standard today.

Adult nonsurgical treatment options include skin traction, skeletal traction, cast brace, and casting. Nonsurgical options are used infrequently outside of the younger pediatric population. Children have the same options, as well as spica casting for those patients weighing less than 80 lb. [2, 3, 4]

Surgical options in adults include the mainstays of intramedullary nailing, either antegrade or retrograde. Plate fixation and external fixation are used less frequently, but these have a place in the decision-making process for the ideal treatment in certain cases. Pediatric cases may also use flexible rods in addition to the adult options mentioned. However, one must consider the patient's immature bones, open physes, parental care available, and growth potential when forming a treatment plan in children. [2, 3, 4, 5]



The femur is one of the largest and strongest bones in the human body. The femur can be divided into regions consisting of the head, neck, intertrochanteric, subtrochanteric (extending 5 cm distal to the lesser trochanter), shaft, supracondylar, and condylar regions.

The structures of the thigh also can be divided into compartments as follows:

  • Within the anterior compartment lies the quadriceps femoris, the sartorius, the psoas, the iliacus, and the pectineus, as well as the femoral artery, vein, and nerve, along with the lateral femoral cutaneous nerve
  • The medial compartment holds the gracilis, the adductor brevis and longus, the adductor magnus, the obturator externus, the deep femoral artery and vein, and the obturator artery, vein, and nerve
  • The posterior compartment holds the semitendinosus, the semimembranosus, the biceps femoris, portions of the adductor magnus, perforating branches of the deep femoral artery, the sciatic nerve, and the posterior femoral cutaneous nerve

The metaphyseal area begins proximally with the subtrochanteric region and distally with the supracondylar region, with the diaphysis in between. Lying posterior on the femur is the linea aspera, which provides a major attachment for fascia. The femur is not perfectly straight; it has a noted anterior bow. The bow varies in degree from person to person, but its presence explains the need for curved nails in order to hold the reduction.

The femur has an abundant vascular supply, receiving the bulk of the blood from the profunda femoral artery. [6] A nutrient artery usually enters along the linea aspera posteriorly and proximally on the femur and supplies the endosteal circulation. The endosteal circulation supplies the inner two thirds to three fourths of the cortex, making the normal blood flow centrifugal in direction. The periosteal circulation enters posteriorly for the most part along the linea aspera.

The periosteal circulation is almost entirely directed in a circumferential direction, having little or no longitudinal spread. Therefore, small wires may be placed around the femur without the danger of devascularizing an area; however, large bands should be avoided. Periosteal circulation has been estimated to serve only the outer one fourth of the cortex. However, the periosteal circulation is critical to fracture healing in the diaphysis.

When a fracture is displaced, the medullary vessels are disrupted and the periosteal vessels predominate as the vascular supply to the fracture site during early healing. In response to fracture, the periosteal vessels proliferate, while the endosteal circulation is restored much later. Therefore, the use of slotted nails may allow for enhanced return of endosteal neovascularization and a more normal blood flow pattern. The significance of periosteal blood flow in healing also emphasizes the importance of avoiding periosteal stripping especially along the linea aspera.

Depending on the level of the fracture and the insertion and attachment of the different muscles of the thigh, varied deformities result. The proximal segment of the femur is under a valgus force of abduction by the gluteus minimus, medius, and maximus. The short external rotators also exert a force on the proximal segment of fractures. A component of flexion and external rotation also exists due to the attachment of the iliopsoas on the less trochanter.

The adductors span most of the medial femur and produce an axial and varus force on the femur. Some of these medial forces are countered by the tension band effect of the fascia lata. The distal femur is under a flexing influence by the gastrocnemius.


Pathophysiology and Etiology

Fractures most often involve the application of a bending load to the femur, with comminution occurring via higher magnitude forces. Torsional loads, in contrast, form a spiral fracture pattern.

Femoral-shaft fractures are usually the result of trauma. [7, 8] Motor vehicle accidents, pedestrian-versus-vehicle accidents, falls, and gunshot wounds are among the most common causes. Pathologic fractures in adults are most often the result of osteoporosis and metastatic disease. In children, it is also important to consider abuse [9] and underlying neuromuscular disorders and metabolic bone disease as causes of the fracture.

Stress fractures also may develop in the femur shaft, often associated with an increase in physical activity. Low-energy shaft fractures have also been associated with the prolonged use of bisphosphonate drugs for treating osteoporosis. [10]



Fractures of the femoral shaft are among the more common fractures that an orthopedist sees. [11] Injury is most common among persons younger than 25 years and those older than 65 years. Analysis of a statewide discharge database revealed an incidence of 1.33 fractures per 10,000 people. The number of shaft fractures in the elderly is increasing secondary to the growing number of geriatric patients in the general population.

Kim et al studied the characteristics of 147 atypical femoral fractures and concluded that patients with diaphyseal fractures were older, had a lower body mass index (BMI), lower bone mineral density (BMD), and larger lateral and anterior bowing than patients with subtrochanteric fractures. [12]



Antegrade intramedullary nailing in adults

Reamed intramedullary nailing has been shown to have excellent results. In a study of 551 cases of femoral-shaft fractures, Wolinsky et al included the following results [13] :

  • Union - 98.9%
  • Infection - 1.1%
  • Component failure - 2.4% (1 nail and 13 bolts failed)
  • Alignment - No fracture healing with more than 10° of varus/valgus or procurvatum/recurvatum and with less than 5° angulation in 89%

More surprising was that 38% of the patients required some type of hardware removal; the reason in the vast majority was pain. Eighty locking bolts were removed, and 130 nails were removed. Note that nail removal should be done 1-2 years after placement to ensure complete fracture healing.

Winquist et al conducted one of the first large studies, which included 500 patients treated with intramedullary nailing. [14] The study gave much the same results; however, the technology and technique were modified as the study progressed.

A study of 152 middle aged patients with closed femur shaft fracture who underwent closed reduction with intramedullary nailing had a nonunion rate of 10.5%; the majority were oligotrophic nonunion. Three cases of hypertrophic nonunion were reported. Risk factors for nonunion included obesity, hypertension, and diabetes mellitus. [15]

Retrograde nailing in adults

In a study of 45 femoral-shaft fractures, Herscovici et al reported the following results: [16] About 95.5% of the fractures healed after the original procedure. No infections or arthrosis occurred, though seven complications included two cases of nonunion (both had broken nails), one malrotation, one case of reflex sympathetic dystrophy (RSD), one case of skin loss (open knee injury), one ileus, and one deep vein thrombosis (DVT). The average ROM was 129° of flexion at the knee. Eight patients had decreased strength in the thigh.

Ricci et al retrospectively compared 104 femoral-shaft fractures treated with retrograde intramedullary nailing and 94 treated with antegrade intramedullary nailing. [17] The following results were obtained:

  • Union - Retrograde, 88%; antegrade, 89% (no statistical difference)
  • Delayed union - Retrograde, 7%; antegrade 4%
  • Nonunion - Retrograde, 6%; antegrade, 6%
  • Eventual union - Retrograde, 97%, antegrade 99%
  • Malunion - Retrograde, 11%; antegrade, 13% (no statistical difference)
  • Knee pain - Retrograde, 36%; antegrade, 9% (significant difference excluding direct knee injury)
  • Hip pain - Retrograde, 4%; antegrade, 10%
  • Repeat operations - Retrograde, 16%; antegrade, 17%
  • Heterotopic ossification - Retrograde, 0%; antegrade, 26%
  • Locking screw broken - Retrograde, 9%; antegrade, 4%

Ostrum et al completed a prospective study comparing antegrade treatment of 39 shaft fractures with 47 shaft fractures treated with retrograde nailing. [18] At the beginning of the study, 54 patients were treated retrograde and 46, antegrade. Later results were after some deaths and loss to follow-up, leaving 39 patients treated antegrade and 47, retrograde. The following results were obtained:

  • Operative time - Antegrade, 71.3 minutes; retrograde, 68.3 minutes
  • Blood loss - Antegrade, 304 cc; retrograde, 256.5 cc
  • Full return to ROM - Antegrade, 9.1 weeks; retrograde, 16.4 weeks
  • Number of fractures - Antegrade, 39; retrograde, 47
  • Unions, including repeat operations - Retrograde, 46 unions of 47 fractures at an average of 18 weeks to achieve union; antegrade, 39 unions of 39 fractures at an average of 14.4 weeks to achieve union
  • No difference in knee motion was noted; only one patient (with vascular repair) had less than 120° of knee flexion
  • Knee pain - Antegrade, four; retrograde, five
  • Hip/thigh pain - Antegrade, 10; retrograde, two
  • Dynamizations - Antegrade, two; retrograde, 10
  • Symptomatic distal locking screws - Antegrade, four; retrograde, 18

They also found a significant relation between the difference in the canal and nail diameter to time of union.

Plate fixation in adults

Most studies involving plating of femoral-shaft fractures have shown less than desirable results compared with other modalities. In a study of 102 femoral-shaft fractures, Bostman et al [19] demonstrated that 24% had some form of major complication, as follows: 12% mechanical failure, approximately 5% delayed or nonunion, approximately 4% repeat fracture, 7% infection, and 25% second operation.

However, articles exist in which results are better, such as study done on 500 shaft fractures in Slovenia in which 85% of the patients experienced no complications. [20]

In another smaller study by Seligson et al of 15 cases of femoral-shaft fractures, seven complications occurred. Three delayed unions, four nonunions, and two implant failures occurred. [21] No axial deformity greater than 1 cm was observed.

External fixation in adults

Alonso et al treated 24 femoral-shaft fractures with external fixation. [22] Ten of the initial operations were used as the definitive treatment, and the others were changed for other options at a later date. Twenty-one fractures eventually developed solid union. Results included one nonunion and two delayed unions. Eleven patients lost an average of 56° of motion about the knee. Of the 10 patients treated primarily with external fixation, no infections occurred, and two had shortening of greater than 2 cm.

Dabezies treated 20 fractures with definitive external fixation. [23] Nineteen developed solid union in an average of 4.8 months. No cases of chronic osteomyelitis were reported even though 65% of the fractures were open. Fourteen needed cast bracing for an average of 1.5 months after the external fixator was removed. Four pin tract infections and three cases of shortening occurred. Nine lost an average of 50° of knee motion. No device failures occurred.

Flexible nailing in children

Cramer et al [24] conducted a prospective trial of 53 femur fractures treated with Ender nails in children aged 5-14 years (average, 8.5 years). Isolated injuries averaged only a 3-day stay in the hospital. Bridging callus was observed in an average of 3 weeks, while the time to healing averaged 12 weeks. On average, full weightbearing was achieved at 30 days.

Forty-nine patients had no angulations after treatment. Four reported angulations were less than 15° in any plane and less than 10° rotation. No leg length discrepancy exceeded 2 cm. An average of 7 mm of overgrowth occurred. Additional results included five complications, four butterfly fragments displaced on rod insertion, one comminution at the fracture site with passage of rods, three cases of inflammation at the insertion site, one hematoma at the removal site, and two broken locking screws after healing.

Flynn reported on 50 cases treated with a titanium elastic nail for pediatric femur fractures with excellent results. [25] Thirty-eight had excellent results, including good alignment, leg lengths within 1 cm, no wound problems or irritation at fracture site, no malrotation, and full ROM at the knee. Some had a knee effusion that resolved when the nails were removed. Nine patients experienced irritation at the insertion site. One nonunion was treated with successful repeat nailing. One fracture occurred with 20° varus angulation; remodeling reduced this to 7°. One repeat fracture occurred.

Frei et al retrospectively reviewed medical records and plain X-ray images of 22 children (between 2–15 years old) treated with elastic stabile intramedullary nailing for unstable femoral shaft fractures. A complication rate of 13.6% was reported. Two children experienced retrotorsion of the femoral neck; diminished anteversion of the femoral neck occurred in one child. [26]

Rigid intramedullary nailing in children

Momberger et al retrospectively reviewed 50 patients aged 10-16 years who were treated with statically locked intramedullary nailing with a greater trochanter starting point. [27] No infection, nonunion, osseous necrosis, or nerve palsies occurred. Full weightbearing occurred at an average of 5.7 weeks.

Complications included one intraoperative fracture, one delayed union, one DVT, and patellofemoral pain in one patient. Of these patients, 25 underwent radiographic analysis, which revealed an average of only 1 mm in limb-length discrepancy, and none had more than 11 mm discrepancy. The average increase in articulotrochanteric distance compared with the contralateral side was 4.5 cm.

A systematic review study analyzed retrospective data on rigid, locked, intramedullary nail insertion sites in pediatric femur fracture patients and whether the entry site of the nail affected the risk of avascular necrosis (AVN). The results found that the AVN rate for the piriform fossa insertion site was 2% and the rate for the tip of the greater trochanter entry site was 1.4%. No cases of AVN were reported for the lateral trochanter insertion site, indicating that this site is associated with the lowest risk of AVN. [28]

Spica casting in children

Infante et al conducted a study comparing immediate spica casting in three different weight classes from 10 to 100 lb for closed, isolated femur fractures in children. A total of 175 fractures were treated with one and one half spica casts. Union was achieved in all fractures by 8 weeks. All patients were discharged from the hospital within 24 hours.

Infante et al recommend from their findings and past literature that in patients 10-80 lb with close isolated fractures of the femur, spica casting should still be used as a criterion standard. [29] Group 3 could not be included in these recommendations because of the low number of patients in this group.

Table. Spica Casting Results (Open Table in a new window)


Average Shortening Before Casting, cm

Average Shortening After Casting, cm

Average Time of Casting Needed, wk

Average AP Varus/Valgus Before

Average AP Varus/Valgus After

1 (10-49 lb)






2 (50-80 lb)






3 (81-100 lb)






Plate fixation in children

Kregor et al reported 15 fractures treated with plate fixation. [30] All healed in an average of 8 weeks. No infections were reported, and 14 of 15 healed with anatomic alignment. Overgrowth averaged 9 mm. No restrictions in activities were reported at 26 months. Ward also reported on 25 fractures treated with plate fixation. Healing occurred in 23 of 25 in an average time of 11 weeks. [31]

External fixation in children

Blasier et al reported on 139 fractures treated with external fixation. [32] All healed in an average of 11.4 weeks. Eighteen were followed up at 2 years, and 15 patients had overgrowth averaging 8.7 mm, three had shortening averaging 7.7 mm, and none required treatment. Pin tract infection was common; six required intravenous antibiotics for pin tract infection treatment. Two repeat fractures occurred.

Miner and Carrol reported on 37 fractures treated in children aged 4-14 years. [33] Healing occurred in 36 of 37. Minimal leg-length discrepancies and angulations were reported. Pin tract infection was reported in 72.7%; more alarming was the 21.6% rate of repeat fracture. Patients with bilateral femur fractures seemed to be at the greatest risk of repeat fracture.