Periprosthetic fractures occur in the bones near a prosthetic implant, both during and after surgery. This complication represents one of the main threats for orthopedic surgeons and one of the most challenging complications in modern arthroplasty. As the global population in Western countries ages and total hip replacements become increasingly common, the impact of periprosthetic fractures is becoming increasingly relevant for national health systems.^[1-3]

The incidence of periprosthetic fractures is estimated to occur in a percentage ranging from 0.1% to 4% of all hip prosthesis implants.^[3]

The most important patient-related risk factors include advanced age, female sex, comorbidities^[4], and prosthesis aging^[5].

Periprosthetic fractures should be suspected for all patients with a previously known history of hip arthroplasty who access the ER complaining of inguinal or trochanteric acute pain, hip functional limitation, and lower limb instability.^[6]

The diagnostic approach involves imaging exams, using plain X-rays for primary evaluation to detect fracture lines. CT scans can be useful to assess the fracture pattern, the degree of displacement, and an eventual loss in bone stock. The Vancouver classification considers fracture location, stability between the prosthetic stem and surrounding bone, and the quality of local bone stock to determine the approach of choice.^[7]

Conservative treatment is mainly reserved for fracture lines limited to the trochanters (Vancouver A). In the remaining cases, surgical treatment is largely recommended by modern literature. The surgical approach to fractures surrounding the prosthetic bed (Vancouver B) largely depends on the stability of the stem over the surrounding bone. Prosthetic revision and replacement are recommended in mobilized stems, whereas open reduction and internal fixation (ORIF) are the treatments of choice for unmobilized implants. The same approach is also the first-line approach to fractures that are localized distally, clear to the prosthetic stem (Vancouver C).^[7,8]

The treatment of periprosthetic fractures does not only challenge orthopedic and trauma surgeons but also represents a threat to patients’ functionality and quality of life. Blood loss, prolonged immobilization, and complex surgeries may have a negative impact, especially on fragile patients like the majority of those who develop periprosthetic fractures.^[9,10]

This study aims to evaluate the clinical impact of periprosthetic fractures, focusing on how fractures and their treatment affect patients’ functionality and quality of life.

This single-center retrospective study was conducted according to the ethical standards in the 1964 Declaration of Helsinki and its later amendments. Our study consisted of reviewing all hip periprosthetic fractures that required surgical treatment in our institution between January 2018 and December 2024.

Inclusion criteria were a previous total hip arthroplasty (THA), an endoprosthetic replacement of the femur head and neck, and a radiological diagnosis of periprosthetic fracture corresponding to a class A, B, or C according to the Vancouver Classification (Fig. 1).

Exclusion criteria were an age lower than 18 years, fractures that occurred in bones other than the femur or without previous arthroplasty, and a post-operative follow-up shorter than 6 months.

For each patient, we collected data regarding their age, gender, and whether the first prosthesis was implanted to treat osteoarthritis or due to a femoral neck fracture. The time elapsed between the first arthroplasty and the fracture was measured for all patients. Once the diagnosis of periprosthetic fracture had been established, fractures were cataloged according to the Vancouver classification. The interval between the diagnosis of periprosthetic fracture and its surgical treatment was measured in days. The surgical treatment of choice was chosen considering the AO/OTA principles and evaluating the peculiarity of each single case. The duration of all surgeries was recorded in minutes. For those who required a prosthetic revision, we distinguished between cases that received cemented and cementless implants (Fig. 2).

For each case, we evaluated the number of days of hospitalization after surgery and the number of blood units administered to prevent critical anemia. At the moment of discharge, we also assessed whether cases could walk and bear weight on the treated limb. The postoperative follow-up consisted of routine X-ray evaluations and outpatient visits performed within one, three, and six months after surgery. Further evaluations were performed yearly, with eventual variations depending on the necessities of each single case. The reported walking ability of each case before the periprosthetic fracture was compared to what we could state during the outpatient visit performed within 6 months after surgery.

Each post-operative complication (Grade III or higher according to the Clavien-Dindo Classification)^[11] that did not represent an exclusion criterion has been reported. Statistical analysis was performed using Stata SE 13 (StataCorp LLC). Statistical significance was set at 0.05 for all endpoints.

One hundred and twenty periprosthetic fractures were diagnosed and treated surgically in our institution during the examined period. Ninety-eight of these fractures involved the hip region, 94 were localized in the proximal femur, and four were in the periacetabular region. The 94 proximal femur periprosthetic fractures represented the cohort of our study. Our cohort included 61 females (65%) and 33 males (35%), with a mean age of 83.4 (56-97) years at surgery.

All the hip periprosthetic fractures were classified according to the Vancouver system: only four were classified as type A (4.3%), 36 were type B1 (38.3%), 30 were type B2 (31.9%), five were type B3 (5.3%), and 19 were type C (20.2%).

Most of them (69%; 65 cases) had had their primary hip replacement due to osteoarthritis, whereas only 34% of our cases (29) had suffered from a femoral neck fracture. Seventy-six patients (70%) had a non-cemented prosthesis, while 18 had cemented implants. On average, patients developed periprosthetic fractures nine years after their first implant. The mean interval between radiological diagnosis of periprosthetic fracture and surgical intervention was 2.8 days (0-15).

Different approaches have been used depending on the fracture pattern, bone stock quality, and the prosthetic stem’s eventual mobilization. Open reduction and internal fixation with plates and screws were sufficient in 58 patients. In two cases, a bone allograft was used to restore the bone continuity before plate fixation. In two cases, the good bone stock and the peripheral fracture pattern allowed the revision of the mobilized stem without fixation. Revision of the femoral stem and fracture fixation were associated in 31 patients. In one case, the proximal femur was replaced by a megaprosthetic implant due to a massive lack of periprosthetic bone and severe periprosthetic loosening.

An analysis of variance (ANOVA) showed a significant difference in the duration of surgery between the Vancouver classification groups (p=0.013), suggesting that the fracture class impacts the duration of surgical interventions. A Tukey Post Hoc test was conducted to compare the groups further. The Post Hoc test results showed a statistically significant difference between the B1 and B2 groups (SE=12.897, p=0.006), suggesting that fractures classified as B2 required a significantly longer surgery compared to those classified as B1. None of the other pairs of groups showed significant differences. Another ANOVA analysis was conducted to assess the relationship between the duration of surgery and the type of surgical procedure performed. Our data suggest that the type of surgery had a statistically significant influence on the operating times (F=8.786, p<0.0001). The post-hoc test, conducted using the Tukey Post-Hoc test for multiple comparisons, revealed that the most significant difference could be found between those who had ORIF alone and a combination of ORIF and prosthetic revision surgery (p=0.001).

No significant differences were found when relating the Vancouver classification to days between fracture and surgery (p=0.438), time between implantation and fracture (p=0.811), and postoperative blood transfusion (p=0.315).

Patients’ mean hospitalization was 10.6 days (5-27). During these days, 78 cases (83%) experienced post-operative anemia and required transfusion with at least one packed blood cell unit. The remaining sixteen cases (17%) did not suffer from anemia. A Pearson correlation test highlighted a statistically significant positive linear correlation between the preoperative hospitalization time and surgical times (r=0.426; p=0.001). A similar test also reported a statistically positive linear correlation between surgical times and post-operative days of hospitalization (r=0.378; p=0.001). The linear correlation between the duration in days of the preoperative and the post-operative hospitalization was even more remarked and statistically significant (r=0.681; p=0.001).

Eleven cases (11.7%) experienced at least one complication directly correlated with the fracture or the surgical intervention, including infections (2), dislocations (6), or further fractures (3). Eight cases (9%) died within six months after the fracture due to conditions not directly correlated with the periprosthetic fracture. Our casuistry was summarized in Table 1.

At their latest follow-up, 61 patients (65%) returned to the level of walking autonomy they had before the fracture. In contrast, 35% (33) reported a decline in their walking ability compared to the preoperative condition. Among the 61 patients who maintained a postoperative ambulation equal to the preoperative one, only 6 (9.8%) developed complications (including deaths). In contrast, 13 (39.4%) of the 33 patients whose ambulation worsened had postoperative complications. A chi-square test revealed that patients with complications were more likely to report worsened gait after surgery compared to those who did not develop complications (p=0.010).

A summary of our results is reported in Table 1.

Periprosthetic fractures of the proximal femur are a topic of increasing interest in modern trauma surgery. Their incidence is expected to increase for the years to come in Western countries, as a result of the gradual aging of the general population and the increasing prevalence of people that underwent total hip replacements.^[12,13] Our study population reflected this tendency, being primarily composed of elderly patients, with a mean age of 83.4 years.

As elders are often complex patients to manage, the treatment of periprosthetic fractures frequently goes beyond the surgical intervention alone.^[9] A correct diagnosis represents the first milestone to direct patients towards the most correct therapeutic pathway. Although the timing between the trauma and its surgical treatment is now largely considered as a prognostic factor for primary hip fractures, the impact of treatment delay on the outcomes of proximal femur periprosthetic fractures is still controversial.^[14-16]

A study by Boddice et al.^[16] highlighted that delays in the surgical management of periprosthetic fractures may not increase mortality or perioperative complications, but they do contribute to an extension of hospital stays. Our study showed a significant correlation between the waiting time before surgery (r=0.681) and the duration of postoperative recovery in our hospital. This data confirms the importance of establishing early diagnosis and limiting surgical delays.

In our experience, patients with uncemented implants represented the larger share of those who had been diagnosed with periprosthetic fractures. Although this difference could depend on the local trends in first implant surgery and the age of visited cases, our data suggest that proximal femurs that received cementification could be less likely to develop postoperative fractures.

The Vancouver classification represents a reliable and well-established instrument to determine the best treatment of choice for periprosthetic fractures.^[3,7,8] Although effective for planning and as a preoperative guideline, the Vancouver classification should not be considered a prognostic factor. In our experience, the Vancouver classes showed no direct correlation with the duration of preoperative and postoperative hospitalizations, postoperative blood transfusions, complications, and postoperative walking ability. However, the Vancouver classification could indirectly influence the outcome of patients’ treatments. Determining the type of surgery, the classification reasonably influences surgical times.^[17] In particular, in our population surgical procedures for B1 were significantly shorter than those for B2 fractures. In our experience, the duration of surgical procedures also influenced the postoperative intercourse, as suggested by increased postoperative hospitalizations for those who had longer surgeries. The type of surgery also intuitively influences the type and the incidence of intraoperative and postoperative complications. In commonly old and complex patients like those who had hip replacements and developed periprosthetic fractures, complications are not only a threat to the surgical success but also to the patients’ quality of life and even their survival.

In our population, 65% of patients regained their pre-fracture autonomy, while 35 % showed a decline in walking ability. The onset of new walking limitations was significantly most common in patients who developed postoperative complications, testifying their adverse impact on the prognosis of old and fragile patients.^[18]

Pavlović et al.^[19] also reported that only 63% of their patients with periprosthetic fractures could return to walk, with or without assistance, within one year after surgery. These results confirm how periprosthetic hip fractures can limit patients’ walking ability and reduce their autonomy. In turn, reduced autonomy may have an adverse social impact, leading to a worse quality of life for patients and their loved ones. It also negatively impacts family budgets and national welfare costs.^[20,21]

Periprosthetic fractures also threaten patients’ survival in a mid-to-long-term scenario. Lamb et al.^[22] reported a one-year mortality of 4841 patients from 35 cohorts included in their literature review. Our data confirm this risk. Nine percent of our cases died within six months after their periprosthetic fractures, although no death was directly correlated with fracture or surgery-induced complications. Although death rates of patients with periprosthetic hip fractures seem to be slightly lower compared to those who have fractures on their native hips^[22-24], this eventuality is far from being remote, and patients should be carefully made aware of it by trauma surgeons for ethical and legal reasons.

We acknowledge the limitations of our study. Its retrospective nature limited the availability of some preoperative data and did not allow the complete standardization of methods and post-operative data collection. Furthermore, the size of our cohort impaired the significance of some statistical analyses.

Beyond these limitations, our study provides evidence of the complex treatment of proximal femur periprosthetic fractures. Despite its pivotal role in establishing the treatment of choice for each patient, the Vancouver classification did not emerge as a prognostic factor. On the contrary, the surgical delay was directly correlated with post-operative hospitalization days, and the onset of complications had a significant impact on functionality and autonomy, leaving their mark on patients’ lives after the fracture.

The Vancouver classification is a useful tool for surgical planning, but does not represent a prognostic factor. Periprosthetic hip fractures can reduce patients’ walking ability, especially in case of complications, and even put at risk the survival of the most fragile patients.