Factors influencing results and complications in proximal periprosthetic femoral fractures: a retrospective study at 1- to 8-year follow-up.

BACKGROUND AND AIM: Proximal periprosthetic femoral fractures (PPFFs) are gradually increasing and surgical management is often associated to high risk of complications, due to elderly population and associated comorbidities. We retrospectively assessed 79 patients at 1- to 8-years follow-up, focusing on factors that may have affected results and complications.
METHODS: Seventy-nine PPFFs were classified according to Vancouver classification in: 4 type B1, 22 type B2, 29 type B3, 24 type C. The average age of the patients was 81.5 year-old (61-95). 37 PPFFs (12 type B2 and 25 type B3) were treated with uncemented tapered fluted revision stems (Link MP Recostruction Stem, Waldermar Link GmbH) and 42 patients with open reduction and internal fixation (ORIF). Patients were clinically and radiographically assessed at a mean 5-year follow-up (1-8 years). Elixhauser Comorbidity index was analysed.
RESULTS: All PPFFs, except two, healed within 10 months. Harris Hip score at follow-up was 65.6 in ORIF group and 79.8 in revision hip group (p<0.05). 35.7% and 32.4% had blood transfusions in ORIF and revision hip group respectively (p> 0.05). 5 patients in ORIF group and 2 patients in revision hip group died within 12 months from surgery (p<0.05).
CONCLUSIONS: Both ORIF and hip revision arthroplasty with modular uncemented tapered fluted stems are valid procedures for Vancouver B2 and B3 PPFFs. Restricted weight-bearing, high number of comorbidities with elevated Elixhauser Comorbidity Index and ORIF are associated to high 12-month mortality.

Introduction

Proximal periprosthetic femoral fractures (PPFFs) are the second reason for hospital 30-days readmission in patients with total hip replacement (THR) after infections (1) and the third most common reason for reoperation after aseptic loosening and dislocation in patients who had THR (2). Because the demand of THR is estimated to grow by 174% by 2030, as reported by Kurtz et al. (3), it is possible that the incidence of PPFFs will increase as well. The risk of having a PPFF is about 1.7% after primary THR with a cumulative risk of 3.5% at 20 years (4) and their incidence is three times higher after revision arthroplasty than primary THR (5,6). Furthermore, most of the times, PPFFs occurred in elderly patients for low-energy trauma (7) and aging people have higher risk of falls (8). So the increasing number of THR in the elderly may contribute to raise their incidence.

PPFFs can be classified according to the Vancouver system proposed by Duncan and Masri, based on fracture site, stability of the stem and femoral bone quality (9). Type A involves lesser or greater trochanter. B fractures involve the bone surrounding the femoral stem and are subclassified in B1 (stable stem), in B2 (unstable stem but good bone stock) and B3 (unstable stem and poor quality bone stock). Fractures distal to the tip of the stem are classified as type C. Surgical management of PPFFs can include both osteosynthesis and revision hip arthroplasty, depending on fracture characteristics.

Besides being technically demanding, PPFFs are also very expensive problems with a 90-days readmission rate that can reach 27.3% (10). In fact, their surgical treatments are usually high-cost with long hospital stay, need of medical support and blood transfusions (11). The perioperative complications and mortality are particularly high as well: these patients are generally fragile people with other systemic comorbidities and limited mobilization, that need long physiotherapy and, most of the times, failed to return to their pre-injury functional level, requiring previously unnecessary assistance, with further social costs (6). The average total cost for a PPFF was estimated at US$ 30,979 (12).

In the present study we retrospectively analysed at an average 5-years follow-up the series of PPFFs admitted in our Department, focusing on the factors that may have influenced their outcomes.

Material and methods

89 patients with PPFFs were admitted in our Department from January 2013 to March 2020. 10 patients were treated conservatively, thus excluded from the present study. The remaining 79 consecutive patients were retrospectively analysed at an average 5-year follow-up (range 1-8). A proper informed consent form has been signed and the study was conducted in according with the Declaration of Helsinki.

The mean age of the patients was 81.5 years (range 61-95). 49 were female (62%) and 30 were male. The PPFFs were classified according to Vancouver classification (9): 4 cases were type B1, 22 were type B2, 29 were type B3 and 25 were type C (Graph. 1). PPFFs occurred in 8 cemented primary THR (10.1%) and in 71 uncemented stem THR, and in 7 anatomical stems and in 72 straight stems.

Graph 1.

Patients’ characteristics and fractures classification according to Vancouver system.

41 patients (3 type B1, 10 type B2, 4 type B3, 24 type C) were treated with open reduction and internal fixation (ORIF) with LCP (Locking Compression Plate) and cables (Fig. 1) and one case classified as B1 with a greater trochanter plate (Cable-Ready plate, Zimmer, Inc, Warsaw, IN, USA) (Fig. 2).

Figure 1.

Type B3 fracture in a cemented stem in a 94 year-old woman (a,b). The fracture has been treated with a LCP plate, although the partial impairment of the cement mantle (c).

Figure 2.

Seven-year radiographic control in a B1 fracture in a 85 year-old women treated with Cable-Ready plate.

37 patients (12 type B2, 25 type B3) underwent to revision arthroplasty with a cementless Link MP Reconstruction stem (Waldermar Link GmbH, Hamburg, Germany) via direct lateral approach. It is a tapered fluted titanium modular stem with distal fixation. The distal part of the implant has different diameters and lengths, with a 2-degree taper angle with longitudinal ribs that increases rotational stability, and a 3-degree bow, that follows the normal anterior bow of the femoral shaft. The proximal part has different neck-shaft angles and different offsets and allows increasing the leg length given to 1 cm modular blocks. Changing of the version is also permitted after distal segment implantation (Fig. 3).

Figure 3.

Cementless Link MP Reconstruction stem (Waldermar Link GmbH, Hamburg, Germany).

The post-operative weight-bearing was restricted for 6 weeks in all patients who underwent to ORIF and was full in 30/37 (81.1%) patients who had hip revision arthroplasty.

The patients underwent to regular post-operative clinical and radiographic controls at one month, three months, 6 months and yearly. Medical records and radiographs were retrospectively analyzed considering the pre-operative and the last functional and radiographic control. On the radiographs were considered signs of implant loosening according to Gruen’s zones (13), fracture healing, and hardware breakage.

Need of blood transfusion and complications occurred 12-months within index surgery were also recorded. Type of stem, femoral bone stock and quality were considered on the pre-operative radiographs. Functional abilities were assessed at follow-up using HHS score, focusing on the return to the pre-injury functional level. The number of comorbidities considered in the Elixhauser Comorbidities Index (14) were also recorded, dividing the patients in four groups based on this number: 0, 1-2, 3-4, 5+ comorbidities. The Elixhauser Comorbidities Index was also calculated using van Walraven algorithm (14).

A comparison between patients who had hip revision and ORIF was made too, analysing the aforementioned variables.

A two-tailed Student t-test was used for statistical analysis, considering the difference as statistically significant if the p-value was inferior to 0.05.

Results

77 out of 79 (97.5%) PPFF were healed within 10 months. One patient died a couple of weeks after surgery, because of cardiovascular complications, so it has been considered as lost at follow-up. The other patient, who was a B3 treated with ORIF, needed a further revision at 14 months for non-union, using a double plating technique. Neither radiolucency lines nor subsidence were identified around the femoral stem; no hardware breakage occurred (Figs. 4-5).

Figure 4.

A periprosthetic fracture-dislocation occurred 7 days after THR in a 78 year-old man (a,b). Radiographic control at 5-year follow-up treated with an uncemented tapered fluted stem showed fracture healing without stem subsidence (c,d).

Figure 5.

A B2 fracture occurred in a 86 year-old woman with an anatomic stem (a,b). Radiographic control at 8 years showed fracture healing without stem subsidence or osteolysis.

The mean HHS was 79.8 (46-100) in the revision arthroplasty group and 65.6 (51-97) in the ORIF group. This difference was statistically significant (p<0.05). Particularly, 4 patients who had revision stem and 7 patients who had ORIF had very low functional ability (HHS< 70) and were not able to walk at the follow-up. All the 7 patients who had ORIF, and one patient who had revision (he previously had a contralateral thigh amputation) had severe disability before the injury, so surgery did not affect their clinical conditions.

In the revision group 13 patients had no pain, 20 patients had mild pain, 4 patients had moderate pain; in the ORIF group 6 patients had no pain, 29 patients had mild pain, 7 patients had moderate pain.

As regards complications, 5 patients in ORIF group (11.9%) and 2 patients in revision hip group (5.4%) died within 12 months from surgery (p<0.05). One of the 5 patients died two weeks after ORIF for congestive heart failure. In the revision hip group we also had one myocardial infarction, three hip dislocations (one revised) and two early infections treated with surgical debridement. In the ORIF group we had one early infection treated with surgical debridement and one case of non-union (Tab. 1).

Tabella 1.

Results and complications in the ORIF group and in the revision arthroplasty group. * identifies statistically significant results.

	ORIF group	Revision group
Death	5 (11.9%)	2 (5.4%)	p<0.05*
Major complications	1 early infection1 non union	2 early infections3 dislocations1 myocardial infarction
Transfusions	15 (35.7%)	12 (32.4%)	p>0.05
HHS	65.6	79.8	p<0.05*
Pain: - no pain - mild - moderate	6 (14.3%)29 (69%)7 (16.7%)	13 (35.1%)20 (54.1%)4 (10.8%)	p<0.05*
Elixhauser Comorbidity Index	19.9	21.1	p>0.05
Return to pre-injury functional level	35 (83.3%)	29 (78.3%)	p>0.05

12 patients in the revision group (32.4%) and 15 patients in the ORIF group (35.7%) needed post-operative blood transfusions (p>0.05).

The Elixhauser Comorbidity Index was 19.9 in the ORIF group and 21.13 in the revision group (p<0.05). The allocation of the patients based on number of Elixhauser comorbidities is described in Graphic 2 with relative post-operative complications.

Graph 2.

Allocation of the patients based on number of Elixhauser comorbidities.

Conclusions

PPFFs, although rare, are becoming more frequent, due to the increasing number of primary THR and to the ageing of the population. This kind of injury usually occurs for low energy trauma in elderly people with associated comorbidities (4,5). These fractures have poor clinical outcomes and, most of the times, the involved fragile patients do not return to their pre-injury functional level, with long in-bed period favouring other systemic complications, like lower respiratory tract infections, cardiovascular failure, bedsores or further limited mobilization (15,16). Female gender is 40% more likely to have a PPFF (6). 62% of our series was female.

PPFFs are also described to be more common in uncemented stems. In a meta-analysis Bissias et al. described cemented implants as a protective factor against PPFF, reducing the risk by 41% (6). In our series only 10.1% of the stems involved in a PPFF were cemented, suggesting that also in our patients the majority of PPFFs occurred in cementless prostheses.

In the present study we aimed to analyse clinical and radiological results at an average 5-years follow-up in Vancouver type B and C fractures using ORIF or revision hip arthroplasty with an uncemented modular tapered fluted stem. We have also tried to compare the two different treatments for B2 and B3 fractures.

Patsiogiannis et al. in their review emphasized that pre-operative evaluation of fracture site, characteristic of cement mantle and osteolytic areas surrounding the stem should influence the surgeon choice (17). In fact, good results can be also achieved with ORIF even in B2 and B3 fractures, as reported in the present study and from other Authors (18).

A comparison among results about managing Vancouver B2 and B3 fractures is quite difficult, because of multiple provided technical solutions and various surgeon preferences, that might influence the results (18).

In the present study we had fair clinical results at an average 5-years follow-up after surgical management of PPFFs. The HHS was significantly higher in the revision group (79.8 vs 65.6), with a greater percentage of patients without pain (35.1% vs 14.3%), but with a slightly higher return to their pre-injury activity level in the ORIF group (83.3% vs 78.3). In our opinion extremely low pre-injury functional level might explain the latter. In fact, surgery did not affect at all the functional ability in 7 patients in the ORIF group and in one patient in the revision group (previously thigh-amputated), and it affected only in minor part the functional abilities of the remaining patients.

As regards systemic complications, in our study they occurred more frequently in Vancouver B3 and C fractures, with 5 deaths within 12 months after surgery in the ORIF group and 2 deaths and one acute myocardial infarction in the revision group. The mortality rate was double in the ORIF group in comparison to revision group (11.9% vs 5.4%). It is well-known that PPFFs have high mortality rate, reaching 60% at 5-years in males, over 75 years, with an American Society of Anesthesiologists (ASA) grade ≥ 3 (19). In a match-paired study based on the New Zealand Registry on 232 hip revisions for PPFFs the 6-month mortality rate was 7.3% in comparison to 0.9% in 232 hip revisions for aseptic loosening, with lower functional outcomes in the first group (20).

Since PPFFs are severe injuries, involving elderly patients that usually assume antiplatelets or anticoagulants drugs, it also important to consider the transfusion rate. More than one third of our patients has been transfused in both groups without statistically difference in the two groups. In the literature the transfusion rate after PPFFs surgery is also significantly high (11,21).

In this study we tried to focus our attention on factors that may influence the functional outcome or promote the onset of complications. In our opinion the number of comorbidities described in the Elixhauser Comorbidities Index can be one of these elements. It is proved that this index correlates with in-hospital mortality and readmission and is usually used in research for creating models to control the severity of illness, giving a weight to each comorbidity (14,22). It can also be applied in orthopaedics surgery, for example, for predicting adverse events and length of stay after shoulder arthroplasty (23) or to foresee 30-day mortality risks after femur fracture (24). In the present study we listed the Elixhauser comorbidities of each patient, dividing the subjects into four groups, according to the number of pathologies. 6 patients died within 12 months after surgery in the groups with more than 3 comorbidities (12.8%), instead only one patient died in the group with one or two comorbidities (4.3%). Moreover, in the first group one patient had a post-operative myocardial infarction. The Elixhauser Comorbidities Index was higher in the group with more than three comorbidities, predicting higher mortality rate, according to the literature (19). Varady at al. showed that Elixhauser Comorbities Index is more accurate than the ASA score for predicting 1-year mortality after hip fractures (25) and no other studies in the literature used this index for PPFFs.

Number of comorbidities also affects the length of stay; in fact patients with more than 5 comorbidities are more likely to have longer length of stay than those with 0 to 2 comorbidities with a higher total cost (12).

Restricted weight-bearing after ORIF may also have affect the outcomes of PPFFs in the present study. We allowed early full weight-bearing in 81.1% of the revision hip arthroplasty and restricted weight-bearing for 6 weeks to all patients who underwent to ORIF, because we did not trust in femoral bone quality. In fact, in the literature ORIF in PPFFs had a failure rate from 18% to 33% (26-28). Other studies reported similar restrictions and suggested this factor as a negative factor for poor outcomes in B1 and B2 fractures (7,29). Albeit not complication free, revision with modular uncemented tapered fluted stems allows avoiding weight-bearing restrictions that may influence the post-operative functional recovery in elderly patients and our study showed that these stems could be a valid option in Vancouver B2 and B3 PPFFs. In fact, stems with distal fixation permit to by-pass proximal femoral comminution or bone loss and to avoid structural bone grafts (30), achieving axial and rotational stability and high rate of fracture union (31,32). Even if some Authors showed low complication rate for PPFF osteosynthesis (29), ORIF can sometimes be challenging with higher risk of non-unions and hardware loosening in comparison to revision arthroplasty, with a functional level that might be very poor (20). Park et al. in 27 B2 fractures treated by ORIF with plates and cables had just one refracture, with optimal radiographic results but they stated that mobility of those patients was so low that they were not able to assess them in their clinic (29). Moreta et al. showed marked functional deterioration and high complication rate, despite good radiographic results, as well (33).

In our study we had 100% non union within 10 months in the revision group and 97.6% in the ORIF group with one case of non-union that needed further revision with double plating technique. These results are consistent with the literature (32,34). Abdel et al. had one stem subsidence that appeared well-fixed at the final follow-up and 11% of dislocations using tapered fluted uncemented stems (31). Dislocation was our main orthopaedic complication in the revision group (3 cases - 8.1%).

In addition, modularity of these revision stems improve the versatility of the implant, making easier to modify the neck version in a second time without removing the distal part of fixed stem (during the same surgery or in case of next revision) or to increase the limb length or femoral offset if necessary. Removal of uncemented stems most of the times is easier to perform in comparison to cemented stems and there is no risk of cement interposition in the fracture, impairing the fracture healing.

The present study has some limits. It has a retrospective design and a mid-term follow-up. However the follow-up was enough long to prove fracture healing and functional recovery.

Then we did not considered if they usually assumed antiplatelet or anticoagulant therapy, that may be a confounding factor, affecting the percentage of blood transfusion in the two groups (ORIF versus revision arthroplasty).

We thought we would be interesting to consider other factors that could have influenced the outcomes, like body mass index or bone mineral density or blood tests like post-operative serum albumin.

PPFFs are severe complications of primary and revision THR, usually associated to poor clinical outcomes. Both osteosynthesis and hip revision arthroplasty with modular uncemented tapered fluted stems can be considered valid procedures for Vancouver B2 and B3 PPFFs. Even if not complication-free, revision stems with distal fixation avoid restricted weight-bearing, that together with low bone quality, high number of comorbidities with elevated Elixhauser Comorbidity Index can be factors affecting the post-operative clinical outcomes and increasing 12-month mortality risk. However, regardless these conclusions, patient-specific considerations should be done in management these injuries: conservative treatment of PPFFs involving a loose stem should be considered in patients with lower life expectancy, several chronic diseases and poor functional demand.