Incidence of complications and functional outcomes following segmental femoral shaft fractures: a critical review of the literature

Femoral shaft fractures have an annual incidence ranging from 10 to 21 per 100,000 patients and are usually sustained either after high energy trauma in the young or low energy mechanism in the elderly with osteoporosis [1, 2]. Segmental femoral shaft fractures (SFSF) are defined as fractures occurring at two or more different levels of the femoral shaft [3] and are usually caused by high energy injuries, such as road traffic accidents (RTA), falls from height or heavy crush injuries [4].

Modern stabilisation techniques of SFSF include open reduction and internal fixation (ORIF) with plates and screws, external fixation, and intramedullary nail (IMN) [5], which remains the mainstay of treatment [4, 6]. However, SFSF have unique characteristics that make them challenging to manage effectively, as they are more prone to shortening, rotational deformity, malalignment, and additionally the compromised blood supply of the segments combined with the associated significant soft tissue insult puts them at higher risk of non-union [7‐9]. To our knowledge, there has been no recent literature overview on SFSF, and therefore, the purpose of this study is to evaluate the current treatments and outcomes of SFSF in adults.

As a basis for this narrative review, a systematic search of the literature was conducted to assess the available evidence regarding SFSF in August 2023 using the 2009 Preferred Reporting Items for Systematic Review and Meta-analysis (PRISMA) guidelines [10].

All studies identified in the English language published onwards from 1995 via the following electronic databases were searched: Scopus, Embase (via Scopus) and MEDLINE (via PubMed). Specific research strings were formulated for each database using the following keywords and/or MeSH terms: 1) (Segment*)AND ( (femur)OR (femoral))AND (fractur*) and 2) Segment*AND fem* AND fractur*.

Inclusion criteria were studies presenting cases with SFSF reporting patient demographics, treatment methods and outcomes. Irrelevant studies, for example studies including Non-SFSF, segmental bone defects, studies including proximal and distal femoral fractures such as ipsilateral femoral neck and shaft fractures, femoral neck fractures, intercondylar fractures, intracapsular fractures were excluded. In addition, studies that do not report outcomes of treatment separately were also excluded. Moreover, biomechanical studies, in vitro studies, animal studies, review articles, non-English language literature, editorials, comments, opinions, letters to the editor, published abstracts, errata (unless they provide original data) were also excluded. The reference lists of the eligible studies and relevant review articles were cross-checked to identify additional relevant studies.

Two researchers (A.V and S.K) independently reviewed all studies (title, abstract and full text) that met inclusion criteria and extracted the relevant data. Any disagreements were resolved by the supervisor (G.C) through discussion.

The PRISMA flow diagram is shown in (Fig. 1). The search yielded 3,393 results and after duplicates were removed 2,647 articles were eligible for screening exclusion based on Title/Abstract. The full text of fifty-four references were assessed for eligibility. Thirty two studies were excluded at this stage: 2 duplicate studies [11, 12], 11 studies that reported mixed outcomes and there were no separate outcomes specific to SFSF [13‐24], 7 studies that were in foreign language [23, 25‐30], 7 studies that reported on SFSF with ipsilateral femoral neck fractures [20, 22, 24, 31‐34] and 5 studies that were abstracts only [35‐39]. Overall, 22 studies (18 retrospective case series [4, 7, 40‐55]), 1 prospective control group study [56] and 3 case reports [57‐59] with a total of 313 patients (314 fractures) met the inclusion criteria and formed the basis of this review.

Non-SFSF are relatively infrequent injuries with an incidence of about 10/10,000 patients [65‐67]. Only 2% of these are open fractures [68]. In general, these fracture types demonstrate an age and gender related bimodal distribution, as they result from high energy trauma in young male patients or fall from standing in elderly females [68, 69]. Associated injuries are common and often necessitate DCO [70]. Several treatment options exist with IMN being the gold standard [4, 6]. In this narrative review the authors are dealing with segmental femoral shaft fractures and not segmental femoral fractures which is a more generic term and may include proximal or distal femoral fractures.

In SFSF, the authors found that the mean age was 36.2 years (range, 18–74 years) and the majority were males (4.8 ratio). These findings show that the male preponderance is higher, and the age is lower in SFSF compared to Non-SFSF where the mean age was 68 years old and male to female ratio was 1:2 respectively [67]. Additionally, the most common mechanism in SFSF was by far an RTA in 76% of cases, with only 15% of cases being a fall, compared to 48% and 37% in the non-SFF counterpart respectively [1]. It can be concluded that SFSF are more frequently the result of high-energy injuries, and occur mostly in young males, compared to the bimodal distribution in Non-SFSF.

In this study the most common associated injuries include the chest (27%), lower leg (24%) and abdomen (20%). In a review of 26,357 non-SFSF fractures, the respective percentages were 18.9%, 20.5% and 6.2% [71]. It can be concluded that SFSF have significantly higher incidence of associated injuries, which would also be expected by the higher energy involved. However, SFSF associated injuries were reported in detail in 75% of cases and in another 33% of studies, associated injuries were not reported separately but collectively as “polytraumas”, which may or may not account for the discrepancy.

As far as classification, like in Non-SFSF, most SFSF studies (67%) report the AO/OTA classification [4, 7, 12, 41, 43, 44, 46, 48‐53, 55]. Two studies [42, 56] report the Winquist and Hansen Classification [61], whereas one study [40] used the Seinsheimer classification [63]. In 1 study [54] the authors used a less known classification system, which divided the femoral shaft in 5 zones and described the fracture segments accordingly. Of note, one study used the AO/OTA classification for each segment separately [44]. Interestingly, only 4 fractures out of 314 had more than one segment [12, 55, 59], whereas the rest had one segment. Regarding open fractures according to Gustilo-Anderson classification, the findings in SFSF indicate a 16% of open fractures, which contrasts with Non-SFSF of 9% only [1], and the authors speculate that this may be because of the higher energy required to produce a SFSF.

Associated injuries often occur in conjunction with femoral shaft fracture, in both Non-SFF and SFF. Recent multicentre studies show that the most commonly associated orthopaedic injuries for Non-SFSF are lower leg fractures (20%), chest injuries (19%) and head injuries (14%) [68, 70, 72]. In this study, the majority of associated injuries included chest trauma (27%), followed by lower leg fractures (24%) and abdominal trauma (20%). Head injury incidences were comparable among SFSF and Non-SFSF. It can be concluded that SFSF present with higher rates of associated injuries compared to Non-SFSF, probably due to the more frequent polytrauma/higher energy mechanism. The authors tried to assess whether the treatment approach changes in patients with SFSF according to the nature and presence of associated injuries, however no specific pattern was identified.

Antegrade reamed IMN remains the gold standard procedure for Non-SFSF [68].This study found that 72% of fractures were treated with IMN [7, 40, 42‐44, 46, 49, 50, 52‐55, 57‐59], 22% ORIF with plate [12, 41, 45, 48, 51, 56], and in 6% both IMN and ORIF were simultaneously performed [4, 44]. This combo approach was only reported by only two studies: Liu et al.[4] used it in a relatively young population (mean age 38 years, range 27–52). The authors used a plate first in order to maintain reduction and prevent rotational instability prior to reaming and inserting an antegrade IMN. Although the plate served the aforementioned purpose, it was arbitrarily left in situ, although they acknowledge that it could have also been removed [4]. Ferreira et al. [44] decided to use this combo approach in two middle aged patients whose segmental injury involved a relatively more distal segment.

As described in the Non-SFSF literature, ORIF does not provide better outcomes compared to IMN, ORIF can be used in diaphyseal fractures with further proximal or distal extension, where IMN may be contraindicated or infeasible [66, 73‐75]. In the results for SFSF reported herein, the majority (85%) were treated with antegrade nails, and 99% of all nails were reamed. Although, in a recent meta-analysis, reamed or unreamed technique for femoral shaft fracture IMN remains controversial [28], reamed IMN is generally preferred due to shorter union time and lower rates of non-union as it was not found to increase blood loss and ARDS rates [28, 73]. However, the unreamed technique has the advantage of reducing operative time and therefore may still be indicated for patients with comorbidities, pathological femoral fractures, or severely injured patients [76, 77]. In 28% of SFSF open reduction was used to restore length, alignment, and rotation. Apart from these reasons, open reduction in SFSF can be used to stabilize the segmental part during reaming and therefore prevent devitalization of the fragment. This can be achieved by applying pressure downwards using a Hohmann retractor, by applying a Schanz pin to manipulate the fragment or by using a blocking screw [49]. In the Non-SFSF literature there were cases where open reduction rates were lower (33%), compared to SFSF [78]. However, in a recent study with subtrochanteric fractures with diaphyseal extension open reduction rates were significantly higher (48%) [79]. The reason for this could be that passing the guide wire becomes easier especially in more complex fracture patterns and there was no difference between union rates in closed and open reduction method [78, 80].

As far as rehabilitation, the majority of patients were advised initial non-weightbearing and progress to full weightbearing based on the amount of callus present. In 29% of SFSF partial weightbearing was initiated after 6 weeks post-operatively, and in 48% advancement was based on callus formation, on a case-by-case basis. On the other hand, in Non-SFSF treated with IMN, partial weightbearing is initiated the first two weeks and is followed by full weightbearing after 4 weeks according to the callus presence [81, 82].

In SFSF the functional outcomes reported are generally good in 70% of cases, compared to the literature for the Non-SFSF (93%) [73], whereas in 19% are excellent. In addition, in SFSF mean time to union is 20 weeks, which is significantly longer in comparison with Non-SFSF which is 14 weeks [5, 73, 74, 83, 84]. The authors speculate that because SFSF are more often the result of high-energy injury compared to the Non-SFSF as alluded to earlier, this may cause a higher disruption in fracture biology due to the nature of the injury and therefore compromised vascular supply and slower healing times.

The most common complications reported in SFSF include delayed union (5%) followed by non-union (4%), knee stiffness (2%), and rotational malalignment (2%). These occurred less frequently in Non-SFSF in which delayed union and non-union were 2.2% and 2.1% respectively [73]. Of interest, SFSF studies do not report the amount of shortening except for 1 study [12] which would be interesting to know as an outcome of treatment of SFSF.

This study has some limitations. The level of evidence is low, as most studies are retrospective case series and case reports. The mean follow-up was 1.45 years (range: 0.75–3 years), which is short and even absent in some studies. None of the studies report comorbidities, American society of anesthesiology grade (ASA) or injury severity score (ISS) during admission of the cases included. In addition, in 33% of cases [52, 54] were described as “polytraumas” which is a general term, and the associated injuries were not all recorded in detail. Only 7 studies report DCO [12, 43, 44, 49, 56, 58, 59], while in the remaining ones, even the newer ones, it is not reported whether or how it was done. Except for 2 studies [12, 56], all studies reported descriptive outcomes and do not employ specific outcome instruments. Furthermore, none of the included studies report mortality rates. Only in 2 studies mean operative time is reported [12, 47]. In addition, time to union is not reported in detail in many studies or the results were mixed with non-segmental fracture types and the exact union time could not be estimated [7, 40, 42, 44, 50, 51, 55, 57, 58]. There were also studies where the outcome of the patients was partially reported [40‐43, 57].

SFF are high-energy complex injuries which are challenging to manage. Their mean age is significantly younger than Non-SFF, with a higher male preponderance and a higher associated injury incidence. They are open in about a quarter of cases, which is significantly higher than Non-SFF. Like Non-SFF, IMN is the mainstay of treatment, but SFF have about 50% longer time to union and about double the rates of nonunion compared to Non-SFF.

Based on these limitations, future research should aim in eliminating the aforementioned limitations: higher quality studies should include a detailed description of the patient group, including comorbidities, ASA, ISS, associated trauma injuries, DCO versus early total care, specific outcome instruments, mortality rates, follow-up time and time to union. In addition, studies comparing SFSF vs NSFSF groups are needed to directly assess the differences and provide sturdier insights, which are currently lacking.