Femoral neck fracture is a common fracture. Previous studies have shown that good closed reduction and internal fixation using cannulated screws in the treatment of femoral neck fracture is efficacious[34,35]. However, neck shortening in the course of treatment is still worthy of attention. A considerable proportion of patients present this phenomenon on follow-up imaging, which is less concerned. Its incidence in this study was 30.33% (37/122), very similar to that in a previous study [17].The total incidence of nonunion was 12.29%, in accord with previous studies[36,37].
Here, we retrospectively studied data from 122 young patients who had Pauwels type II femoral neck fractures treated using FPTCS and analyzed possible factors influencing neck shortening. Univariate analysis showed that patient age, sex, BMD, BMI, fracture type, and reduction quality were risk factors for neck shortening. Age and BMD were closely related to the occurrence of neck shortening. With increasing age, BMD may decline, and it is clear that bone quality may determine the probability of femoral neck fracture [38].Reduction of the axial anti-compression strength of the femoral neck leads to its shortening. Sex is also an important influencing factor. Compared with men, women have thinner bone cortex and lower bone density. After menopause, estrogen decreases rapidly, which further affect the process of fracture repair[39]. BMI also increases the risk of neck shortening after fracture[40]. Increased BMI may directly increase the axial pressure on the fracture end, leading to neck shortening. Therefore, for patients with these risk factors, a comprehensive preoperative evaluation is essential, although this complication may result from a combination of these factors.
Non-conditional logistic regression analysis showed that fracture type (Garden III, IV) and reduction quality (Grade III, IV) were the main causes of postoperative neck shortening. Higher fracture type, posterior medial cortex comminution, and lower reduction quality could indicate a greater risk of postoperative neck shortening in young patients who had Pauwels type II femoral neck fractures. Therefore, more attention should be paid to these three indicators during clinical observations. We found that fracture type (Garden III, IV) may increase the likelihood of neck shortening, as previously reported[9,18]. Garden type III and IV fractures are unstable and often accompanied by comminution or posterior medial bone cortical defects[41].Comminuted fractures increase bone absorption after surgery and are more likely to produce neck shortening after healing. Destruction of the posterior medial cortex often leads to a lowering of both the quality of reduction and resistance to axial loading, and results in neck shortening, even when complete anatomical reduction is achieved in the operation[42]. Garden's Alignment Index is commonly used to evaluate reduction quality. When the evaluated fracture reduction quality fails to meet the standard of anatomical reduction, the probability of postoperative neck shortening is greatly increased[18]. It is possible that the stress on the fracture’s broken end is not uniform, leading to collapse at the fracture site, with consequent neck shortening.
In this study, a univariate analysis showed that patient age, sex, BMD, BMI, fracture type, and reduction quality were risk factors for femoral neck shortening. Importantly, a multifactor analysis showed that, in the shortening group, the main causes of neck shortening were the fracture type (Garden III, IV), reduction quality (Grade III, IV), and the posterior medial cortex comminution; additionally, secondary causes might be BMD (T≤ −2.5) and BMI (≥28.0 kg/m2).These results indicated that the Harris score in the shortened group was lower than that in the non-shortened group. That finding suggested that neck shortening would result in reduced hip function, which was mainly manifested as claudication and poor abduction strength [43]. A potential mechanism for these manifestations could be that neck shortening might have affected the arm of the abductor muscle, which reduced the ability of the hip joint to maintain a stable gait and pelvic balance [44]. Consequently, patients had to increase the abductor muscle strength to compensate when walking, which resulted in walking claudication. In future, it would be interesting to explore whether there might be some quantitative relationship between the degree of femoral neck shortening and the hip function score.
Although early mobilization following young femoral neck fracture surgery is of superior importance, to date, there is no unified postoperative rehabilitation protocol. In this study, the postoperative weight-bearing program was determined by the physician in charge, and it depended on many factors, including age, sex, BMD, BMI, fracture type, posterior medial cortex comminution, and reduction quality. In general, partial weight bearing activities, with the use of an assistive device, were allowed as soon as patients could tolerate them. In patients with the risk factors mentioned above, partial weight-bearing was delayed until 8 weeks after surgery. During those 8 weeks, the patient performed functional exercises for the affected limb, without bearing weight.
This study has some limitations. First, our results were based on a small number of patients. We still did not accurately predict whether neck shortening was related to the healing rate of the femoral neck fracture. Second, we did not include patients with Pauwels type I and III fractures. Third, because there are few studies on neck shortening and all methods for measuring neck shortening are still in their initial stages, more accurate methods of measurement need to be further explored.
To reduce the effect of postoperative neck shortening on hip function, we have changed our treatment of femoral neck fractures by using fully threaded screws or a femoral neck system (FNS), instead of FPTCS. Currently, a clinical follow up is ongoing to evaluate the degree of postoperative neck shortening and its effects on hip function, when treated with fully threaded screws or FNS. Those findings will be reported in future studies.
In conclusion, Our results support the use of fracture type, the presence of comminution of the posterior medial cortex, and reduction quality as important reference indexes to predict the possibility of neck shortening after internal fixation with FPTCS for Pauwels type II femoral neck fracture in young patients. BMD and BMI may also be risk factors. The results also suggest that fracture type, posterior medial cortex comminution, and reduction quality might be useful for evaluating postoperative neck shortening.