Interprosthetic femoral fractures: Morbidity and mortality in a retrospective, multicenter study.

INTRODUCTION: Interprosthetic femoral fractures (IFF) are becoming more frequent; however they have not been the subject of many publications and the largest study on this topic includes only 30 cases. The complication rate and clinical outcomes have only been evaluated in small case series. This led us to conduct a retrospective, multicenter, observational study in IFF patients with at least 12 months' follow-up to (1) determine the mortality and morbidity (2) determine the clinical and radiological outcomes and (3) identify elements of the treatment indications.

HYPOTHESIS: The morbidity and mortality rates will be comparable to those in recent studies on this topic.

MATERIALS AND METHODS: The study included 51 patients (49 women, 2 men) with a mean age of 82.8±9.2 years [55-97], a mean Parker score of 4.9±2.4 and a mean Katz score of 4.4±1.4 who had suffered an IFF between 2009 and 2015. According to the SoFCOT modifications of the Vancouver classification, 30 fractures were interprosthetic in the shaft segment where there were no implants (19 double C and 11 type D (corresponding to a type C with less than two diaphysis widths between the extension stems of the hip and knee implants)) while 21 were periprosthetic, with 12 around the THA (11 B1 and 1 B3) and 9 around the TKA (8 B1 and 1 B3). One patient was treated conservatively with an external fixator but died the next day, 2 patients received a new total hip arthroplasty and 47 underwent plate fixation of their fracture (one patient was treated non-operatively because of poor medical condition).

RESULTS: One patient was lost to follow-up, and nine died during the first 6 months. Six early surgical site complications occurred and 13 general ones. Within 1 year of the IFF, there were six mechanical complications, two surgical site infections and two cases of loosening. The mean follow-up was 27.6±17.2 months. The mean time to union was 19.25±8.8 weeks. The mean final Parker score was 3.37±2.6 and the mean Katz score was 2.98±1.8; both were significantly lower than the initial scores. Six patients died between months 12 and 50. The overall mortality at the final review was 31% (16/51) with a median survival of 3.45 years.

DISCUSSION: Our hypothesis was not confirmed because the mortality and morbidity in our study were higher than in other published studies. In the six relevant studies identified, the surgical site infection rate was 12.3%, the major revision rate was 11.6% and the mortality rate was 6.5%. In our study, these values were 24%, 24% and 31%, respectively. These worse results may be explained by the very fragile nature of the studied population and the surgeons not following appropriate technical rules for fracture fixation.

LEVEL OF EVIDENCE: IV, Retrospective study.