Introduction

Distal femoral fractures account for about 3–6% of all femoral fractures [1,2,3]. Previously published reviews on distal femoral nonunion have provided valuable insights into the epidemiology, risk factors, and treatment of this condition [4,5,6]. The distal femoral nonunions, although rare, negatively impact the quality of life (QOL) and pose considerable treatment challenges [7, 8]. Patients with open and comminuted fractures and infections are most prone to develop nonunion [9, 10]. Open reduction and internal fixation (ORIF) combined with cancellous autografting is the most successful treatment option, with a union rate of 97.4% and an average time to heal of 7.8 months [11, 12].

Several non-modifiable and modifiable risk factors for non-union have been identified, including open fractures, fractures with bone loss, and those associated with osteoporosis, diabetes, obesity, and malnutrition [13, 14]. In addition, certain patient-related factors like smoking, advanced age, and obesity are also considered to affect fracture healing negatively [14, 15]. Importantly, inadequate fracture fixation is primarily responsible for non-union and implant failure [16]. The rigidity score has been advocated to predict the risk of nonunion [17, 18]. This score was significantly associated with the risk of non-union [18]. Still, no significant association was detected between non-union and postoperative weight-bearing status or other previously identified risk factors [17]. Karam et al. [19] reported an overall similar non-union rate of 17.8% in both locked plating of closed periprosthetic and non-periprosthetic distal femoral and comminution, which was the most crucial predictor of non-union.

Although nonunions are a well-known complication of fractures, evidence on this condition is limited. The present investigation systematically updates current evidence, reviews existing modalities, innovations, and related outcomes, and discusses future perspectives on the management of nonunion of the distal femur.

Methods

Eligibility Criteria

All the clinical studies investigating treatment options for nonunion distal femur fractures were accessed. According to the author’s language capabilities, English, German, Italian, French, and Spanish articles were eligible. Only studies with levels I to V of evidence, according to the Oxford Centre of Evidence-Based Medicine [20], were considered. Reviews, opinions, letters, and editorials were not considered, as were animals, in vitro, biomechanics, computational, and cadaveric investigations. Missing quantitative data under the outcomes of interests warranted the exclusion of the study.

Search Strategy

This study was conducted according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses: the 2020 PRISMA statement [21]. The following PICO algorithm was established for the database search:

  • P (Problem): nonunion distal femur fractures;

  • I (Intervention): management;

  • C (Comparison): different modalities of fracture healing;

  • O (Outcomes): union.

In May 2024, PubMed, Web of Science, Google Scholar, and Embase were accessed without time constraints. The Medical Subject Headings (MeSH) used for the database search are reported in the Appendix. No additional filters were used in the database search.

Selection and Data Collection

Two authors (FM and LS) independently performed the database search. All the resulting titles were screened by hand and, if suitable, the abstract was accessed. The full text of the abstracts which matched the topic was accessed. If the full text was not accessible or available, the article was not considered for inclusion. A cross reference of the bibliography of the full-text articles was also performed for inclusion. Disagreements were debated and mutually solved by the authors. In case of further disagreements, a third senior author (RV) decided.

Data Items

Two authors (FM and LS) independently performed data extraction. The following data were extracted: author and year, name of the journal and study design, level of evidence, length of the follow-up, number of included patients, mean age, percentage of women, type of initial and definitive treatment, time to definitive treatment, number of unions with or without major complications and main findings. Data were extracted in Microsoft Office Excel version 16.72 (Microsoft Corporation, Redmond, USA).

Assessment of the Risk of Bias and Quality of the Recommendations

The risk of bias was evaluated following the guidelines outlined in the Cochrane Handbook for Systematic Reviews of Interventions [22]. Nonrandomised controlled trials (non-RCTs) were assessed using the Risk of Bias in Nonrandomised Studies of Interventions (ROBINS-I) tool [23]. Seven domains of potential bias in non-RCTs were assessed. Possible cofounding and the nature of patient selection before the start of the comparative intervention are assessed by two domains. A further domain is used to determine bias in the classification during the intervention. The final four domains assess the methodological quality after the intervention comparison has been implemented and relate to deviations from previously intended interventions, missing data, erroneous measurement of outcomes, and bias in the selection of reported outcomes. The figure of the ROBINS-I was elaborated using the Robvis Software (Risk-of-bias VISualization, Riskofbias.info, Bristol, UK).

Synthesis Methods

The main author (FM) performed the statistical analyses following the recommendations of the Cochrane Handbook for Systematic Reviews of Interventions [22]. The IBM SPSS software (version 25) was used. For dichotomic data, the number of events and observations were reported. For continuous data, the arithmetic mean and standard deviation were used.

Results

Study Selection

The literature search resulted in 179 articles. All search results were extracted and checked for relevance. Of these, 84 studies were discarded because they were duplicates. Following the defined inclusion criteria, the abstracts of 95 articles were reviewed, and 47 studies were excluded because they did not match the eligibility criteria. The reasons that led to exclusion were: study design (N = 28), low level of evidence (N = 2), not evaluating the un-union of distal femur fracture (N = 9), and language limitations (N = 8). An additional 13 articles were excluded because they did not offer quantitative data on the outcomes of interest. In conclusion, 35 investigations were included in the present analysis. The results of the literature search are shown in Fig. 1.

Fig. 1
figure 1

PRISMA flow chart of the literature search

Full size image

Methodological Quality Assessment

The ROBINS-I was applied to investigate the risk of bias in the included studies. Given the small number of patients studied in the included investigations, the risk of bias based on confounding and the selection of participants was rated in four studies as serious and in four other studies considered critical because they reported data from only one or two patients. The remaining studies were rated as having a low to moderate risk of bias in these domains. The protocol of intervention was well reported in most studies, and no significant deviation from the interventions was identified, leading to an overall low to moderate risk of bias in the classification of interventions and deviation from intended interventions. Data were adequately reported in the most included studies, and the measurement of the outcomes was equivalent among the groups. Given the lack of randomisation of the investigators and patients, the measurement of the outcomes was evaluated with a moderate risk of bias in all of the studies. When described, the reported results corresponded to the planned protocol in most included studies. However, the exact details of a protocol are often missing, leading to a moderate risk of bias in most of the studies. Given the overall poor methodological quality in the included studies, the overall risk of bias was predominantly moderate. The assessments of the methodological quality are given in Fig. 2.

Fig. 2
figure 2

The ROBINS-I of non-randomised control trials

Full size image

Study Characteristics and Results of Individual Studies

Data from 832 patients were collected. Of them, 34.2% (239 of 698 patients) reported an open fracture, and 24.5% (78 of 319 patients) reported infection at the fracture site. The mean length of the follow-up was 28.9 ± 13.2 months. The mean age of the patients was 53.8 ± 14.7 years. The generalities, patient characteristics and main results of the included studies are reported in Table 1.

Table 1 Generalities, patient characteristics, and main results of the included studies
Full size table

Result Syntheses

At a mean of 21.7 ± 20.9 months, 84.5% (703 of 832) of patients reached complete union without major complications, and 3.8% (32 of 832) reached complete union with major complications. On the other hand, 8.7% (72 of 832) patients showed signs of persistent non-union.

Discussion

The present systematic review revealed that with surgical management of the distal femoral nonunion 84.5% (703 of 832) of patients reached complete union without major complications, and 3.8% (32 of 832) reached complete union with major complications at a mean of 21.7 ± 20.9 months. But, 8.7% (72 of 832) patients showed signs of persistent non-union.

Distal femoral nonunions are challenging to manage, and patient risk factors should be critically evaluated when treating acute distal femoral fractures, since several modifiable risks, such as intake of alcohol and steroids, smoking, and diabetic control, might reduce the risk of nonunion [4, 10, 24]. In addition, optimal fracture alignment and compression are also essential when treating acute fractures [25]. Infections are important risk factors, especially in open fractures or fractures that fail to heal [26, 27]. There is good evidence to guide decision-making and successful union with conventional trauma fixation techniques in the majority of patients. Biological augmentation of fracture healing is not necessary in most cases but needs to be considered in higher-risk cases with a mobile atrophic pattern.

The definition of nonunion is heterogeneous and is related to the time elapsed from the injury to the diagnosis [28]. Reported time intervals ranged from 3 to 12 months [28]. A cut-off of 6 months was most commonly used to define nonunion [28]. Few studies reported a cut-off of 12 months and 9 months [28]. More recently, the US Food and Drug Administration (US FDA) proposed a cut-off of 9 months without evidence of progression to healing over the previous 3 months [29, 30]. The diagnosis of nonunion is often achieved through imaging using plain radiography [31]. However, consensus on the appropriate criteria still needs to be improved [32, 33]. The absence of bridging callus on at least three of four cortices on radiographs is a common assessment criterion, as is the movement of the fracture during radiographic stress testing, lack of bony/cortical continuity, or lack of bridging bone trabeculae [34]. Clinical criteria used to diagnose nonunion are also heterogeneous and not commonly used; among them are pain over the fracture site, pain during weight bearing and natural mobility at the fracture site [35, 36].

Several strategies are used for the management of distal femoral nonunion, with each treatment option having merits and demerits [37,38,39,40]. Dynamisation is occasionally attempted as a less invasive method to achieve union in femoral diaphyseal fractures with an axial stable fracture configuration [41, 42]. The main surgical technique used is the revision ORIF, which involves osteosynthesis with freshening of bone ends and bone grafting [43, 44]. It offers a high chance of achieving union and allows for restoring the joint alignment [43, 44]. Another minimally invasive technique involves retrograde intramedullary nailing, which is more suitable for fractures where the original surgery used plates and screws [45, 46]. Being less invasive, it reduces soft tissue disruption. It may potentially provide faster healing, but it is not suitable for all non-union types, and there may be the possibility of malalignment if nail placement is inaccurate. A combined plate and nail fixation construct is also used, using a locking plate and an intramedullary nail for enhanced stability [47, 48]. Ilizarov external fixator is another surgical option used for treating distal femoral nonunion, and it can provide high union and low complication rates in experienced hands [49, 50]. Although autologous bone grafting is usually considered an essential part of any surgical fixation technique to promote bone growth and improve the chances of healing of nonunion, it requires an additional surgical site for graft harvesting, increasing the potential for pain and discomfort [51, 52]. It is also beneficial for complex non-unions or those with significant bone loss [52]. However, being more invasive than single procedures, there may be an increased risk of infection and hardware complications. In resistant and challenging distal femoral nonunion in older people with associated bone loss, osteoporosis, and secondary knee osteoarthritis, mega prosthesis has been successfully used as a one-stage salvage procedure [53,54,55].

Saxena et al. [47] reported the outcomes of 10 cases of non-union of distal femur treated by fixation with nail plate construct and bone grafting, achieving union in 10.3 months on an average duration.

The only previously systematic review on this topic was done by Ebraheim et al. [4], who assessed 19 studies and found that the most common initial treatment was ORIF with plating, and the most common definitive treatment was with fixed angle plating combined with cancellous autografting, with a union rate of 97.4% and average union time of 7.8 months.

In a retrospective study of 31 cases of distal femoral non-union, which were treated by anatomical lateral locking plates and autogenous bone grafting, stable union was achieved in all cases, with a significant improvement in the functional scores [43]. The authors concluded that optimal stability, good compression at the non-union site, maintenance of axial alignment, bone end freshening, adequate use of cortico-cancellous bone autograft, respecting the biology, and the correct choice of the implant (including the size) are essential to achieve union at the fracture site [43].

In a retrospective study of 18 cases of atrophic distal femoral non-union that were managed using a combination of J-shaped bone graft and double plating union was achieved in all cases at an average of 22.1 ± 5.5 months (range 14–34 months) postoperatively [56]. The average healing time was 6.72 ± 2.80 months [56].

A recently published bibliometric study (1990–2023) found a growing interest in fracture non-union globally through published studies [57].

Distal femoral fractures are complex injuries, and even with advancements in treatment, non-union remains a potential complication. While there has been progress in understanding and treating this condition, there are still gaps in the current research. Studies have not yet established a clear consensus on the most optimal surgical technique for nonunions, and guidelines for managing nonunion, especially of the distal femur, are missing. More research is needed on long-term patient outcomes after treatment for non-union. This can help assess the effectiveness of different approaches and identify areas for improvement. More research is needed to better understand the influence of patient characteristics, such as age, comorbidities, and bone quality, on the risk of developing non-union and treatment success. Further investigations into these areas can lead to more standardised protocols, improved success rates, and improved outcomes. Strategies to improve the outcome of the surgical management of nonunion treatment are evolving.

Logistic regression predictive models can help develop and identify patients at high risk of non-union before surgery. This would allow for early intervention and potentially a reduction in non-union rates [14]. Specific growth factors, like bone morphogenetic proteins (BMP), are used locally at the fracture site to stimulate bone formation, enhance healing, and potentially reduce the need for extensive grafting procedures [58]. However, its use is still limited, as it is a relatively new approach with ongoing research and high cost compared to traditional methods [59, 60]. Genetic influence has been proposed, especially polymorphisms in the genes ANXA3, BMP2, CALY, CYR61, FGFR1, IL1β, NOG, NOS2, PDGF gene, and TACR1 are susceptible to nonunion [61, 62]. Hence, gene therapy may help stimulate bone growth and improve healing rates in non-unions [63]. Refinement in minimally invasive surgical (MIS) techniques for treating non-unions could lead to faster recovery times and less surgical burden on patients [64, 65]. In this context, there is also interest in robotic-assisted surgery (RAS), which might improve precision and accuracy during revision surgery for non-unions [66, 67]. Developing bioactive implants that promote bone healing and enhance the fusion rate at the fracture site could be crucial [68]. Using biodegradable implants would also eliminate the need for a second surgery to remove hardware. Moreover, advanced imaging techniques like Magnetic Resonance Imaging (MRI) may better assess bone healing progress in patients [24, 69]. However, the presence of metallic implants may generate artefacts. Developing biosensors to monitor the bone healing process in real-time may thus allow for adjustments to treatment plans as needed [70, 71]. Focusing on these areas, researchers are developing more effective strategies for distal femoral nonunion, leading to improved patient outcomes and a faster return to function.

Conclusion

This systematic review of distal femoral nonunion based on 35 studies and 832 cases revealed that with appropriate surgical treatment 84.5% of patients reached complete union without major complications, and 3.8% (32 of 832) reached complete union with major complications at a mean of 21.7 ± 20.9 months. But, 8.7% (72 of 832) patients showed signs of persistent non-union.