Salvage of an Infected Periprosthetic Failed Fracture Fixation in a Nonagenarian

Abstract

Infected nonunions resulting in segmental bone loss are a devastating complication for patients and a difficult problem for surgeons. Adequate soft tissue coverage, return of mobility, fracture stability, and long-term freedom from infection are all important goals of treatment. Although there are numerous published studies that provide some treatment guidelines, there are patients who require unique and individualized solutions. In this report, we present a case in which an antibiotic-impregnated cement spacer was used as a component of the definitive treatment in a geriatric patient with segmental bone loss of the femur secondary to severe infection as a salvage technique to avoid amputation.

Background

Segmental bone loss in the setting of infection, trauma, and tumor remains a challenging problem. A variety of surgical treatments have been described utilizing allograft, vascularized autograft, bone transport, and 2-stage reconstruction incorporating antibiotic-impregnated cement. A systematic review by Struijs et al ¹ evaluated published data on multiple techniques for management of infected long bone nonunions. The studies reviewed included 1- and 2-stage techniques for treatment of infected long bone nonunions. These studies employed different techniques but uniformly reported union rates of 75% to 100% and high rates of persistent infection, thus suggesting that there is no single definitively superior technique. None of these studies included significant numbers of geriatric patients who, because of their age and limited longevity, may require specialized treatment strategies not suitable for younger patients.

Masquelet et al ² popularized a 2-stage technique, similar to that used by reconstructive surgeons in adult patients with bone loss. The first stage consists of debriding the necrotic bone and filling the defect with an acrylic spacer. The cement is removed once the infection is cleared, and cancellous bone graft is placed in the defect. Several studies have presented excellent outcomes with this technique ^{3 –5} ; however, nearly all patients are younger adults, and the results in geriatric patients are unknown.

Although there are many techniques for reconstruction of infected nonunions, there is a paucity of literature guiding treatment of segmental bone loss in the elderly patients. In this population, decreasing the number and rigor of the surgeries and decreasing the length of hospital stay might be as important as achieving a solid bony union for overall quality of life.

Case Report

A female in her mid-90s was referred for treatment of a failed open reduction and internal fixation of her right periprosthetic femur fracture. She had a history of hypertension, spinal stenosis, and atrial fibrillation requiring chronic anticoagulation therapy.

Two months prior to presentation, she fell and sustained a comminuted supracondylar periprosthetic distal femur fracture that was treated with open reduction and internal fixation using a lateral plate with screws and cables. This fracture occurred between an ipsilateral total hip placed in 2003 and a total knee arthroplasty placed in 1999. Both of these were radiographically well fixed and the patient had no pain in either joint prior to her fall. Her immediate postoperative course was complicated by an asymptomatic urinary tract infection and pneumonia, both of which were treated with oral antibiotics. She was discharged to a nursing home. Upon follow-up with her surgeon, several weeks postoperatively, she complained of persistent pain, and radiographs demonstrated failure of fixation (Figure 1). She was then referred to our institution for our further care.

Figure 1.

Radiographs on presentation showing failed fixation of a periprosthetic femur fracture.

At her initial presentation to our facility, she had pain and gross motion at her fracture site. Her incision was well healed without erythema or induration. The decision was made to proceed with hardware removal and revision fixation.

During surgery, a copious amount of foul-smelling material was discovered upon incision of the iliotibial band. A large portion of her vastus lateralis muscle was necrotic. The supracondylar and distal shaft components of the entire femur were necrotic with multiple devascularized bone segments. The hardware was removed, the necrotic bone and muscle were debrided, and the leg was irrigated with saline. Following debridement, there was a 12-cm long defect in the distal femur. Multiple strings of vancomycin-impregnated beads were placed, and the incision was closed over drains. To stabilize her femur, an external fixator was placed on the proximal femur and tibia, spanning the segmental loss (Figure 2). The patient was initially placed on vancomycin. Her cultures grew Clostridium species, nonperfringens, which was sensitive to moxifloxacin and metronidazole. She was placed on moxifloxacin 400 mg by mouth per day and metronidazole 500 mg by mouth every 8 hours.

Figure 2.

Intraoperative radiographs following debridement, placement of external fixator, and antibiotic-impregnated methyl methacrylate beads.

The patient was taken for 2 subsequent surgical debridements during this same hospitalization. The surgical team had multiple discussions with the family regarding limb salvage versus amputation, and the deleterious effects that her age would have on the likelihood of limb salvage with the goal of achieving a stable bony union. In the presence of an aggressive infection and a 12-cm bone defect, we felt it unlikely that we could ever achieve a bony union of her femur and that any attempt would require multiple surgical procedures combined with a prolonged period of disability and immobility. In this patient, amputation would also be a life-threatening procedure that would also result in permanent severe disability. Given the proximity of the total hip to her infection, it was felt that a hip disarticulation would likely be necessary. Due to the morbidity and mortality associated with that procedure, we elected to proceed with limb salvage that was tailored to this patient’s age and orthopedic condition.

Our goals were to surgically eradicate the infection and provide enough structural stability to the construct to permit mobilization. We thus addressed the femoral fracture with a long lateral locking plate in conjunction with a large tobramycin impregnated intercalary cement spacer to address the bony defect (Figure 3). 2.4 g of tobramycin was mixed in the 80 mg packet of bone cement. The spacer was sized to approximately equalize leg lengths and achieve inherent soft tissue stability.

Figure 3.

Immediate postoperative radiographs demonstrating the revised plating and the antibiotic-impregnated bone–cement block.

At 2 weeks follow-up, her pain had improved, the incision was healing without drainage, and she was weight-of-leg weight bearing. Radiographs at 6 weeks postoperatively showed no evidence of hardware failure, and given her limited activity demands she was advanced to 50% weight bearing. There was concern that the patient was developing neuropathy from the metronidazole, and she was switched to Augmentin at 2 months postoperatively. Her inflammatory markers normalized, and antibiotics were discontinued at 4 months. The patient continued to do well and was advanced to weight bearing as tolerated with a walker. At her last follow-up at 18 months post-surgery, she had minimal pain in her right leg and her radiographs remained unchanged (Figure 4). She was predominantly in a wheelchair secondary to complications of her spinal stenosis but was able to weight bear as needed for assisted transfers. Her radiographs show maintenance of her fixation and no migration of the cement block. She passed away approximately 2.5years later with no further complications, no clinical signs of infection, and no additional interventions for her right lower extremity.

Figure 4.

Final follow-up radiographs at 18 months.

Discussion

The literature is sparse regarding antibiotic-impregnated cement for definitive fixation in weight-bearing limbs. Many authors have described managing segmental bone loss with temporary antibiotic-impregnated cement spacers with subsequent staged bone grafting. However, given this patient’s age and the magnitude of the bone loss, regenerative bone healing seemed an unrealistic goal. Alternatives providing early, definitive treatment allowing for mobilization were felt to be more appropriate for this elderly woman. Hip disarticulation and transfemoral amputation were considered although the mortality rates for these procedures in a patient her age are dismal, and the transfemoral amputation would have left her at risk for a periprosthetic hip infection. Hip disarticulation accounts for only 0.5% of lower extremity amputations in the United States, and the outcomes reported are variable. ⁶ Zalavras et al ⁷ reported 1-month mortality rates for 15 patients undergoing hip disarticulation. Of the 7 patients, 1 whose surgery was done emergently died within the first month. Researchers at the VA performed a large review of 2375 patients receiving a lower extremity amputation. ⁸ Of the cases, 0.7% were hip disarticulations, and the average age was 63. After adjusting for age, etiological factors, and comorbidities, there was a 13-fold increased risk of in-hospital mortality among patients with hip disarticulation compared to transtibial amputation. There was an almost 7-fold increase in mortality for age greater than 86.

There are also prosthetic arthroplasty options that might be useful in this situation. Total femoral replacement could have addressed the intercalary nonunion and provided the potential for immediate weight bearing. ^{9 –11} One of the authors (HAC) has used this option in similar patients. However, to mitigate the risks of persistent infection, this would have been best performed as a staged procedure to follow placement of an antibiotic spacer and given this patient’s age and the nature of her infection, she would have remained at high risk of a recurrence with few options for salvage of her limb. In addition, even with preservation of the abductor mechanism via a trochanteric slide (“flip”) osteotomy, this is an extensive surgery requiring femoral resection and revision knee replacement, and functional recovery would have been unlikely in this particular patient. Finally, 1 additional option for this patient would be to use a prosthetic distal femur and knee with compressive osseointergration technology (Compress implant, Biomet Inc, Warsaw, Indiana). ¹² As with the total femoral replacement, this would be a much more expensive option than our intercalary polymethylmethacrylate spacer and would also be best delayed until after a complex hardware removal and antibiotic spacer placement. Recovery after these 2 procedures would be difficult and there would be a real risk of recurrent periprosthetic infection as described previously.

There are several case reports in the arthroplasty literature of using antibiotic spacers as definitive treatment following infected total joints. Two-stage reconstruction of infected arthroplasty is now the standard of care, and a variety of techniques using articulating and nonarticulating antibiotic spacers have been described. Scharfenberger et al ¹³ reported on 28 patients treated with a preformed antibiotic impregnated femoral implant with a cement head that contains a metal rod as a core. Of the 28 patients, 18 did not have the second stage of the reconstruction, either because of comorbidities or refusal to undergo additional surgery. At a 2-year follow-up, they were reported as “functioning well” although no specific mention of their weight bearing and ambulatory state was made. Other small series have demonstrated similar results. In our own personal experience, articulating cement spacers rapidly lead to progressive bony erosion and pain once weight bearing commences.

Special consideration must be given to the amount and type of antibiotics used in the cement, especially in elderly patients. The Food and Drug Administration has approved the use of “low dose” commercially premade spacers which contain 1 g tobramycin/40-g packet cement. ¹⁴ The safety profile for systemic levels of tobramycin eluted from cement is much higher. Up to 3.6 g per 1 40 g packet of cement of tobramycin can be given before serum levels rise higher than 3 mg/L. ¹⁵ By comparison, patients receiving intravenous tobramycin peak dosing levels are 5 to 8 mcg/mL. Moxifloxacin has not been shown to need age-based dosing, ¹⁶ and metronidazole clearance is not affected by decreased renal function. We recommend the help of infectious disease specialists, however, to balance these potent drugs.

For the patient presented in this report, a difficult decision was made to implant a long intercalary periprosthetic femoral spacer stabilized in part by a locking plate, with the preoperative goal that this be the definitive treatment. The infection, the bone loss, and the patient’s advanced age dictated what we consider an individualized option that would not be expected to provide acceptable long-term results in a typical, younger and healthier patient. We must emphasize that this was a salvage procedure that should not be used routinely for patients with nonunions and bone loss. Our solution likely offered the best chance of restoration of at least partial mobility immediately, thus improving quality of life and reducing the risks of serious morbidity and mortality in this very elderly patient.

Conclusion

Little is known about the natural history of antibiotic-impregnated cement retained in the body. Although biomechanical studies show little difference in the mechanical properties of the cement once the antibiotics have been added, ¹⁰ with weight bearing in an active patient, the construct would likely eventually fail due to degradation of the cement or failure at the bone–cement junction. In this case, our patient was able to return to an acceptable quality of life, with minimal pain, weight bearing as needed for assisted transfers, and no evidence of continued infection. Although this technique would not be recommended for younger, more active patients, it may be an acceptable salvage alternative in patients for whom amputation or segmental grafting are not reasonable options.