PROSPECTIVE ASSESSMENT OF THE CLINICAL, RADIOGRAPHIC AND FUNCTIONAL EVOLUTION OF TREATMENT FOR UNSTABLE TROCHANTERIC FRACTURES OF THE FEMUR USING A CEPHALOMEDULLARY NAIL

Trochanteric fractures are extracapsular fractures of the proximal femur involving the area between the greater and lesser trochanters. They are common fractures in the elderly population because of osteoporosis, and are mainly associated with low-energy trauma, such as falling from the individual's own height.

The incidence of fractures of the proximal femur has increased significantly over recent decades and is expected to double over the next 25 years because of increasing life expectancy among the population(1). It has been estimated that nine out of every ten trochanteric fractures occur in individuals over the age of 65 years(2). Around one in every 1000 inhabitants per year, in developed countries, is affected by fractures of the proximal femur(3).

Advanced age and associated comorbidities are responsible for high morbidity and mortality and for the high cost of treatment. In the United States, the expenditure is expected to rise from 8.7 billion in 2009 to 240 billion in 2040. The mortality rate after six months ranges from 12 to 41% and results mainly from clinical complications such as pulmonary thromboembolism and sepsis(4).

The classification systems most used in the literature are AO/ASIF(5) (Figure 1) and the Tronzo system( 6) (Figure 2).

Figure 1 - AO/ASIF classification.

Determination of the fracture pattern in terms of stability is fundamental for assessing the treatment options. Fractures are considered to be unstable in the presence of comminution of the posteromedial cortical bone, reverse obliquity and subtrochanteric extent(2). Unstable fractures are grouped as AO/ASIF 31A2 and 31A3 and Tronzo III, IV and V.

Non-surgical treatment is reserved for patients with comorbidities that put them at unacceptable risk in relation to anesthesia, surgical procedures, or both(2). Through the principle of relative stability, surgical treatment has the aim of achieving functional reduction and stable fixation for pain relief and early return to walking and to the previous state of independence. Although surgical treatment does not change the mortality rate over the first six months, 

Figure 2 - Tronzo classification.

it diminishes the complications resulting from prolonged restriction to bed(2).

Because of the abundant blood supply in the trochanter region, the rates of osteonecrosis and pseudarthrosis are low, thus favoring surgical treatment via internal fixation. Osteosynthesis of trochanteric fractures is the principal surgical treatment method, although replacement using a prosthesis is occasionally indicated(2).

The result from surgical treatment depends on the fracture pattern, bone quality, reduction and fixation method. Among the mechanical complications from the surgical treatment, varus collapse, uncontrolled shortening and fixation failure (which are more common in unstable fractures) can be highlighted(7). A variety of fixation devices have been developed to face up to the difficulties in fixation of trochanteric fractures(8). The implants can be intramedullary or extramedullary. The extramedullary implant most frequently used is the sliding hip screw (DHS). For stable fractures, this type of screw is the implant of choice for treating unstable fractures, according to many authors(4,9-11).

Cephalomedullary systems are biomechanically better for reducing flexor moment, because of better rotational control and better control over varus collapse cervicodiaand shortening, given that their layout is more medial than extramedullary devices are(12). Several studies have reported that osteosynthesis using cephalomedullary devices promotes faster return to walking, shorter duration of surgery and less blood loss(2,13).

The design of proximal femoral nails has evolved and the nails are now in their third generation. The improvements in the design have reduced the occurrences of some complications like intraoperative fractures and fractures below the nail tip (after the operation). The Targon-PF® cephalomedullary nail (Figure 3) presents the differential that the cephalic anti-rotation nail and cephalic jacket of the sliding screw are fixed in the femoral nail itself, thereby avoiding the "Z" effect that occurs with other cephalomedullary nails(14) (Figure 4).

The objective of this study was to prospectively assess the clinical, radiological and functional evolution of osteosynthesis using a cephalomedullary nail (Targon-PF®), in unstable trochanteric fractures of the femur, over a one-year postoperative follow-up. 

Figure 3 - PF-Targon-PF® nail.

Figure 4 - Anteroposterior radiograph
on hip, with "Z" effect.

METHODS

Between November 2008 and March 2009, 70 patients with trochanteric fractures underwent femoral osteosynthesis. Out of this group, 40 patients underwent osteosynthesis by means of the Targon- -PF® cephalomedullary nail because they presented unstable trochanteric fractures.

The patients included in this study were over 60 years of age, with unstable fractures classified according to the AO/ASIF system as 31A2 or 31A3. They were included only after they authorized their participation through signing a free and informed consent statement.

The exclusion criteria included the presence of femoral fractures with subtrochanteric extent, pathological fractures due to tumor lesions, previous incapacity to walk and associations with other fractures that would interfere with rehabilitation. Three patients were excluded because they were under 60 years old, because their ages could have caused distortions in the functional assessment of walking. Thus, the final sample was composed of 37 patients.

Fourteen men (37.8%) and 23 women (62.2%) were assessed. The right side was affected in 18 cases (49%) and the left side in 19 cases (51%). The patients' mean age was 77.7 years, with a range from 60 to 95 years. The most common trauma mechanism was a fall from the individual's own height, in 36 cases (97.3%), There was one case (2.7%) resulting from trauma due to a projectile from a firearm.

All the fractures were classified using the AO/ ASIF and Tronzo classification systems. According to the AO/ASIF system, 27 patients (73%) presented type 31A2 fractures and 10 (27%) had type 31A3 fractures. According to the Tronzo classification, 23 patients (62%) were in Tronzo III, four (11%) were in Tronzo IV and 10 (27%) patients were in Tronzo V.

Comorbidities were found to be present in 31 patients (83.8%). Cardiovascular diseases were seen in 25 patients (67.5%), diabetes in 11 (29.7%), Parkinson disease in three (8.1%), pulmonary disease in two (5.4%) and Alzheimer's disease in two (5.4%). Other comorbidities present included epilepsy, alcoholism, hypothyroidism and chronic kidney failure. There were no comorbidities in six patients (16.2%). The patients underwent surgical treatment as soon as their clinical conditions allowed this. The mean time elapsed from hospital admission to the date of the surgery was 7.1 days, with a total mean duration of hospital stay of 9.9 days.

All the patients underwent osteosynthesis using a Targon-PF® cephalomedullary nail, after indirect reduction of the fracture on the orthopedic table with the aid of fluoroscopy. The nails used had a distal diameter of 10 or 12 mm, single proximal diameter of 17 mm, mediolateral angle of 6° and cervicodiaphyseal angles of 125°, 130° and 135° between the screws in the neck and the axis of the intramedullary nail. The implant was chosen after preoperative planning, according to the cervicodiaphyseal angle of the proximal extremity of the contralateral femur and the diameter of the diaphyseal medullary region.

Drug prophylaxis for deep vein thrombosis (DVT) was administered during the pre and postoperative periods. The patients also received prophylactic antibiotic therapy at the time of induction of anesthesia and for 48 hours after the operation, consisting of cefazolin. During the immediate postoperative period, standard analgesia was administered and the patients were instructed to gradually start to walk again according to their tolerance level.

Through a prospective cohort evaluation with cross-sectional analyses, the patients who underwent osteosynthesis using Targon-PF® to treat unstable intertrochanteric fractures were assessed one week, two weeks, one month, two months, six months and one years after the operation. At each return visit, clinical, radiological and functional assessments were made on the patients.

The clinical assessment was made by means of analyzing occurrences of clinical complications, including assessment of the mortality rate and its causes.

The radiological assessment included an analysis on the maintenance of the quality of reduction, implant positioning, fracture consolidation and synthesis material failure. The quality of fracture reduction and implant positioning were evaluated by means of radiographic examination during the immediate postoperative period, using anteroposterior (AP) and lateral radiographic views of the proximal femur. The reduction was considered to be ideal when the cervicodiaphyseal angle on the AP radiograph was between 130° and 135°. If the angle was smaller, the case was considered to be one of varus reduction; and if it was larger, valgus reduction(15). Implant positioning was assessed by means of the tip-apex index(12). Fracture consolidation, varus collapse and synthesis material failure were assessed by means of radiographic analysis at subsequent returns.

The functional assessment was made through analysis on Harris scores, progression of walking and pain scale evolution. At the return visit one year after the operation, a final functional assessment was made using the Harris score (a functional assessment scale with a maximum score of 100 points that includes evaluations on four categories: pain, mobility, daily activities and range of motion). Scores lower than 70 are considered to be poor; between 70 and 80, reasonable; 80 to 90, good; and 90 to 100, excellent (Annex 1)(16). During the one-year follow-up, progression of walking and pain scale evolution were assessed. Walking ability was assessed by dividing the patients into five groups: non-walking, walking with the aid of a stick, with crutches and with a walking frame and walking unaided. Before suffering the fracture, 16.2% of the patients were already using a walking frame, 34.3% were using a stick and 59.5% were walking unaided. For the pain evaluation, a visual analogue scale graded from 0 to 10 was used. This scale was directly proportional to the pain level reported by the patient: 0 - free from pain; 1 to 3 - mild pain; 4 to 6 - moderate pain; and 7 to 10 - severe pain.

The statistical analysis was performed using the SPSS software. Student's t test was used for normally distributed independent variables. Differences were considered to be statistically significant when the rejection level for the nullity hypothesis (P) was 0.05 (significance level of 95%).

RESULTS

The clinical assessment over the one-year period showed that 28 patients (75.7%) did not have any postoperative clinical complications, while the other nine patients (24.3%) presented complications. The mortality rate over the evaluation period was 13.5% (five patients). Of these, four patients (80%) died due to sepsis resulting from bronchopneumonia and one patient (20%) died due to an episode of upper digestive tract hemorrhage. The other clinical complications comprised two cases (5.4%) of deep vein thrombosis, one case (2.7%) of heel ulcer and one case (2.7%) of acute arterial obstruction (Table 1).

From the radiographic evaluation, the mean cervicodiaphyseal angle immediately after the operation was 132.5°, with a standard deviation of 9.8. The reduction after the operation was considered ideal in 31 patients (83.6%) (Figure 5), while there were three cases of varus reduction (8.2%) and three cases of valgus reduction (8.2%) (Table 2). The mean tip-apex index was 22.8 (standard deviation = 8.3), and 27 patients (73%) had an index = 25 mm and the other ten

Annex 1 - Harris Hip Score

Table 1 - Postoperative clinical complications.

Figure 5 - Anteroposterior radiograph on hip, with ideal reduction.

Table 2 - Cervicodiaphyseal angle measurement.

Table 3 - Tip-apex index measurement.

Figure 6 - Anteroposterior and lateral radiographs on right hip
with cut-out.

The functional assessment using the Harris score one year after the operation showed that the mean was 69.2 with a standard deviation of 9.3. In 16% of the patients, the result presented was excellent, 19% good, 28% reasonable and 38% poor. The Harris scale showed a statistically significant inverse relationship with age (p = 0.023), with a coefficient of -0.0402, i.e. younger ages were associated with higher Harris scores (Figure 7). The Harris score was not influenced by the fracture classification according to the AO/ ASIF system (Table 4).

The evaluation of the progression of walking showed that the quality of walking gradually improved over the assessment period. Despite the instructions to gradually start walking early on, as tolerated, 35 patients (94.6%) did not do any active walking during the first week, but only bore weight on the limb.

Figure 7 - Harris score versus age.

Annex 2 - Clinical, functional and radiographic assessment on the patients who underwent osteosynthesis using a cephalomedullary nail.

Table 4 - Harris scale for AO/ASIF classification.

At the assessment two months after the operation, it was observed that only three patients (8.6%) were not walking actively. Six months after the operation, only 3.1% were still not walking, while 50% were walking with the aid of stick, 34.4% with a walking frame and 12.5% without using any aid (Figure 8). Before the fracture, 16.2% of the patients were already using a walking frame, 34.3% were using a stick and 59.5% were walking unaided.

The evaluation on the evolution of pain by means of the visual analogue scale graduated from 0 to 10 showed that there was a progressive diminution over the assessment period, with a mean of 5.19 in the first week and 2.25 one year after the surgery (Figure 9).

Figure 8 - Time versus progressing of walking.

Figure 9 - Pain scale versus time.

DISCUSSION

Trochanteric fractures are correlated with high morbidity and mortality rates. The most prevalent and most important clinical complications are deep vein thrombosis, pulmonary thromboembolism, superficial infection and deep infection(4).

The general incidence of mortality after trochanteric fractures is described in the current literature as 6 to 11% within the first month and 14 to 36% within the first year(16). Mortality typically results from cardiopulmonary, thromboembolic and septic complications. In our sample, the mortality rate was 13.5% (five patients) over the one-year period. The cause of death was sepsis due to bronchopneumonia in four cases (80%), while one case (20%) was due to upper digestive tract hemorrhage.

Deep vein thrombosis (DVT) is a substantial cause of morbidity and mortality during the follow-up of cases of fracture of the proximal femur(4). Symptomatic DVT occurs in around 2%, and 85% of these cases are diagnosed during the first five weeks after the fracture. Factors such as prolonged immobility, coagulopathy and delayed surgical treatment favor occurrences of DVT(17). In the present study, antithrombotic drug prophylaxis was instituted in all cases, and two patients (5.4%) were diagnosed with DVT. No cases of pulmonary thromboembolism were found among our sample.

No cases of superficial or deep infection were identified in the present study. The other clinical complications were one case (2.7%) of acute arterial obstruction and one case (2.7%) of pressure ulcer on the heel.

According to Schipper et al(8), the most common postoperative complication seen in radiological evaluations is cut-out. Its incidence has been described in the literature as 0.7 to 10.6%(2). In the present study, there was only one case of cut-out (2.7%). Inadequate reduction, especially with varus presentation, incorrect implant positioning and advanced osteoporosis are the factors that determine its occurrence.

Implant positioning should follow the concepts introduced by Baumgaertner et al(12), in which the distance between the tip of the sliding screw and the center of the femoral head should not be more than 25 mm, as a sum of the AP and lateral-view radiographs (tip-apex index < 25 mm), which facilitates telescoping of the dynamic system of the implant and reduces the risk of cut-out(15). Also according to Baumgaertner et al(12), occurrences of cut-out increase considerably when the tip-apex index is greater than 25 mm, independent of the quality of the fracture reduction. The tip-apex index has been described for osteosynthesis using a DHS. Use of this index for assessing the adequacy of positioning of cephalomedullary nails remains controversial, especially with regard to nails difficulty in positioning the sliding screw at the center of the femoral head in AP-view radiographs. Thus, there is a greater tendency towards positioning the sliding screw at a lower location on AP radiographs, especially in patients with a short femoral head and neck.

Despite this proviso, this index has been used by some authors for assessing whether implants have been positioned correctly(15). In our study, we found a mean tip-apex index of 22.8 (standard deviation of 8.3). There were 27 patients (73%) with a tip-apex index = 25 mm and 10 (27%) with an index > 25 mm. Even with the presence of the anti-rotation nail in the Targon-PF, we were able to respect the principles recommended by Baumgartner, for most of our patients. The patient in this study who presented cut-out had advanced osteoporosis and inadequate implant positioning (tip-apex index of 40 mm), as risk factors.

Werner-Tutschku et al(18) reported that the main reason for occurrences of cut-out is an initial unsatisfactory reduction, generally with varus presentation. As well as favoring occurrences of cut-out, varus alignment may result in Trendelenburg gait. In our sample, the mean cervicodiaphyseal angle was 132.5°, with a standard deviation of 9.3. The postoperative reduction was considered to be ideal in 31 patients (83.6%), and there were three cases with varus reduction (8.2%) (Table 2). According to Werner-Tutschku et al, the incidence of varus consolidation is 11.2%(18). The angular deviation did not favor cut-out and did not statistically influence the quality of gait in this sample (p = 0.442).

In a meta-analysis, Kaplan et al(2) presented a mean time taken to achieve consolidation of four months, independent of the device used. On the other hand, Bride et al10 reported that consolidation occurred after an average of six months. According to Crawford et al(19), the consolidation rate found among patients treated with a cephalomedullary nail was 89%(19). In the present study, consolidation was observed in all the patients after six months, except for one case that presented cut-out one month after the operation, who then underwent resection arthroplasty. Consequently, consolidation could not be assessed. The abundant blood supply in the trochanteric region, associated with stable synthesis, favored this result.

One complication from using cephalomedullary nails is fractures of the femoral diaphysis below the tip of the implant(7). Lack of experience on the part of the surgeon and the inadequate design of first and second- -generation nails have favored this complication(14). Nails of the latest generation have a radius of curvature that fits the anatomical shape of the femur better. In contrast with older reports, recent studies have reported significant decreases in this complication(2). The rate of diaphyseal fracture below the nail tip ranges in the literature from zero to 2.1%(2). In the present study, no cases of fractures distal to the implant were found.

With regard to functional evaluation, the Harris score is the scale used by the majority of authors for postoperative functional assessment of trochanteric fractures(16). Schipper et al(8) found a mean score of 66.80 (standard deviation = 17.94) with a proximal femoral nail of PFN® type, and 69.50 (standard deviation = 16.00) with a nail of Gamma Nail® type, after one year. The mean value found in the present study was 69.3 (standard deviation = 9.3) with the Targon-PF®; 16% of the patients presented excellent results, 19% good, 28% reasonable and 38% poor. The functional result according to the Harris score was considered to be poor or moderate for the majority of the patients who underwent osteosynthesis with a cephalomedullary nail. The great difficulty in functional assessment of fractures of the proximal femurs using the Harris score is that it is impossible to make an assessment before the surgery and thus, no comparisons from before to after the operation can be made, as used in elective surgery. Given the advanced age of these patients, with preexisting limitations, they would possibly present scores that are already compromised. One important indicator is that the Harris score showed a statistically significant correlation with age (Figure 7), and was not influenced by the Tronzo and AO classifications, and not even by the quality of the reduction and positioning of the implants. Thus, the energy of the trauma and the quality of the surgery, evaluated radiographically, did not influence the functional result, but the patient's age significantly influenced the functional result, such that older patients had worse results (coefficient: -0.0402). The previous functional level of these more elderly patients was probably already worse than that of the younger patients, which influenced these results. Therefore, from our point of view, a new scale for use in functional assessments on hip fractures needs to be developed, since the Harris score is more indicated for elective surgery such as hip arthroplasty to treat arthrosis, because of its comparative possibilities. With a new scale, it would be possible to better assess the influence of the trauma energy and the quality of the surgery in functional assessments, after removing the influence of the patient's previous state from the assessment. We emphasize that even though the mean Harris score of 69.3 found in the present study is low, it is similar to values found by other authors(16). For a better functional assessment on the patients, we analyzed another two important parameters separately: recovery of walking ability and evolution of the pain scale.

According to Pajarinen et al(13), patients who underwent osteosynthesis with a cephalomedullary nail, in unstable trochanteric fractures, presented a significantly faster return to their previous level of walking(13). Herrera et al(9) reported on a study involving 250 patients treated with the PFN® and Gamma nail® cephalomedullary nails, in which around 50% of the patients had recovered their previous walking capacity, one year after the surgery(4). In the present study, we assessed the recovery of walking ability over the course of time. We divided the subjects into five groups: no walking; walking with a frame; walking with crutches; walking with a stick; and walking unaided (Figure 8). The greatest evolution in the quality of walking occurred over the first two months after the operation, such that only 8.6% of the patients were still not walking at that time. One year after the operation, only 3.1% of the patients were still not walking, while 50% were walking with the aid of a stick, 34.4% with a walking frame and 12.5% without any aid. Thus, one year after the operation, 40.6% of the patients presented same walking capacity as presented previously.

An assessment of pain was made by Nuber et al(20), through a six-month follow-up with successive scores, comparing patients who received a DHS with those who received an intramedullary nail. The pain scores were considerably greater in the group treated with intramedullary nails. In the present study, the patients were assessed using a visual analogue scale for pain and showed progressive diminution of pain at the return visits over the one-year period, as presented in Figure 9. The evolution of the pain scale was not influenced by the trauma energy, age, reduction parameters or implant positioning parameters.

CONCLUSION

Osteosynthesis using a cephalomedullary nail Targon®-PF, used in unstable trochanteric fractures, resulted in low rates of clinical complications, excellent stabilization, few mechanical complications and adequate functional results. The authors consider that this is an appropriate technique for treating unstable trochanteric fractures of the femur.