High Potential for Complications After Traumatic Exposure in Patients With a Total Knee Replacement

Objective. The purpose of this study is to describe cases of traumatic exposure during the early postoperative period in patients with a total knee replacement (TKR) and to report the treatments instituted and complications of this adverse event. Methods. A retrospective review of postoperative patients with TKR performed at the Universidade de São Paulo from 2001 to 2017 who subsequently were treated at the emergency room due to trauma to the TKR region accompanied by surgical wound dehiscence in deep planes and implant exposure was conducted. The initial treatment, evolution, and complications of each patient were reported. Results. In 16 years, there were 3224 TKRs performed at the study institution. Among this population, 4 (0.1%) patients had trauma dehiscence of the surgical wound during the immediate postoperative period. All patients were women between the ages of 64 and 88 years with comorbidities (eg, diabetes mellitus and/or hypertension). The mean time between the surgery and trauma was 6.7 ± 6.2 days. All patients underwent surgical cleaning, debridement with polyethylene exchange, and primary closure of the surgical wound and received adjuvant treatment with broad-spectrum antibiotic therapy. Of the 4 patients affected, 3 developed an infection, 3 required new surgery after initial cleaning, 2 lost their prostheses, and 1 lost a limb, requiring a transfemoral amputation. Conclusions. Traumatic dehiscence with implant exposure during the initial postoperative period of TKR is an infrequent event (0.1%) with a high potential for complications (75%), which may lead to loss of the prosthesis and even the limb. Measures should be taken to prevent falls and gather reports from several centers to broaden the knowledge of this rare event, identify prognostic factors, and define the best treatment algorithm.

Population aging and the advancement of surgical techniques have considerably increased the rate of performance of total knee replacement (TKR) procedures for the treatment of osteoarthritis.¹ It is estimated that from the early 2000s to 2025, the number of arthroplasties performed annually will double.² According to Kurtz et al,³ TKR outnumbers total hip replacements by a large amount. Therefore, it would be reasonable to assume the absolute number of complications from this procedure will also increase. 

The main and most cited TKR-related complications in the literature are surgical wound dehiscence or delayed healing, hematoma formation, deep vein thrombosis, residual instability of the prosthesis, stiffness or range of motion loss, aseptic loosening, and postoperative infection.⁴ In a survey of more than 17 000 knee prostheses, Galat et al⁵ found a complication rate of 0.33% related specifically to the surgical wound, including delayed healing, dehiscence or necrosis, formation of subcutaneous hematomas, and persistent surgical wound drainage. However, little discussion exists in the literature regarding surgical wound complications due to traumatic events during the early postoperative period, which can trigger a cascade of new complications and serious consequences.^5,6

To the best of the authors’ knowledge, only 1 study⁶ specifically addresses the topic. Sershon et al⁶ reported an incidence of only 0.05% of traumatic exposure of the prosthesis in a series of almost 28 000 patients; however, frequent complications resulting from this event were demonstrated.⁶ Because this event is uncommon, no details exist regarding the best form of prevention and type of treatment. With this in mind, prevention policies, treatment, and management protocols for such cases must be improved. 

The objectives of the present study were to report cases of traumatic exposure of the TKR implant during the early postoperative period and draw attention to this possible complication, which, although rare, can have serious consequences. The authors hypothesize that cases of traumatic dehiscence will present a high rate of complications such as infection and implant loss.

The present study was submitted to and approved by the institutional ethics committee. This is a retrospective review of postoperative patients with a TKR performed at the Universidade de São Paulo from 2001 to 2017 who were treated in the emergency room for trauma to the TKR region accompanied by surgical wound dehiscence in deep planes and implant exposure. Patients with non-traumatic dehiscence and other postoperative complications were excluded from this study.

The patients included initially were assisted in the emergency department of the institution, and their wounds underwent crude dirt removal, superficial cleaning, and placement of occlusive dressing with subsequent referral to the surgical center for deep cleaning and final closure. Due to the unique condition of each patient, the subsequent procedures were based on the injuries presented. Broad-spectrum antibiotic therapy was introduced, normally teicoplanin and Amikacin (Sagent Pharmaceuticals) or Meropenem (Pfizer), and a tetanus vaccination was recommended.

The patients were followed up for at least 2 years for early detection of subsequent complications, such as surgical site infection or early prosthesis loosening, and received specific treatment according to the evolution of each case. The evolution of each patient and any complications were recorded. 

In the last 16 years, there were 3224 TKRs performed at the study institution. Among this population, 4 (0.1%) patients were treated in the emergency room of the hospital for traumatic dehiscence of the surgical wound during the first 2 weeks of the postoperative period. All patients were women between the ages of 64 and 88 years, and their comorbidities included 2 with diabetes mellitus and arterial hypertension, 1 with only diabetes,  and 1 with only hypertension. All patients had adequate glycemic control and had undergone arthroplasty via a longitudinal approach. The deep planes were sutured with absorbable multifilament thread, preventing future granuloma formation that could irritate local tissues, and the skin was sutured with nonabsorbable monofilament thread. The mean time between the surgery and trauma was 6.7 ± 6.3 days. Two falls occurred in the patients’ homes, 1 occurred in the hospital during hospitalization, and 1 occurred in the street in front of the hospital when the patient was discharged.

All patients underwent surgical cleaning, debridement with exhaustive washing of the surgical site with saline solution, polyethylene exchange, collection of culture material for an antibiogram, and primary closure of the surgical wound. In addition, all patients received adjuvant treatment with broad-spectrum antibiotic therapy until the culture results were obtained, according to the treatment protocol determined by the Commission on Hospital Infection Control of the study institution. The mean time between trauma and surgical debridement was 3.7 ± 0.9 hours. 

The Table shows the characteristics of each patient studied.

Case 1
A 77-year-old woman slipped and fell during hospitalization when trying to get out of bed without nursing assistance. After a good initial outcome, with the exception of the range of motion, which was 0° to 80°, the patient presented with an infection 7 months after traumatic implant exposure, with the development of joint effusion, pain, and fistula in the anterior region of the right knee. She required a 2-stage TKR exchange and presented growth of Staphylococcus epidermidis, Enterococcus faecalis, and coagulase-negative staphylococci at the time of arthroplasty removal. After the second stage of the revision, the patient progressed satisfactorily with her infection under control and a range of motion of 0° to 100° (Figure 1).

Case 2
An 88-year-old woman slipped and fell on wet asphalt in front of the hospital. The patient presented at the time of trauma with loss of the patellar component, irreparable damage to the extensor mechanism, and medial collateral ligament (MCL) insufficiency of the right knee; she had to undergo secondary surgical reconstruction of the extensor mechanism and MCL with an allograft after 2 months. Growth of S epidermidis was observed in cultures collected during the second surgery, and antibiotic therapy was instituted for 6 months. This patient did not develop clinical signs of postoperative infection and presented with good knee stability, good healing of the extensor mechanism, and a range of motion of 0° to 90° (Figure 2).

Case 3
A 64-year-old woman slipped and fell in the middle of the night on her way to the bathroom in her home. The patient developed an acute infection after the first debridement with persistent wound drainage, pain, and constant hyperemia on the distal edge of the surgical wound. Although all culture results were negative, the patient developed an uncontrollable infection even after multiple surgical cleanings; the prosthesis with spacer placement was removed and broad-spectrum antibiotic therapy was initiated. The patient had wound dehiscence with skin loss, requiring gastrocnemius flap rotation for cutaneous coverage. Due to poor progression, persistent pain, and difficult infection control, a transfemoral limb amputation was performed at 8 months after the initial fall.

Case 4
A 78-year-old woman, who was wearing socks, slipped and fell on the waxed wooden floor of her home. Of the 4 patients presented, this patient was the only one with a favorable outcome, with no need for new interventions after initial debridement and negative bacterial cultures, thus allowing the primary prosthesis to be kept, with a range of motion of 0° to 110° (Figure 3). 

The main finding of this study is the low rate of traumatic dehiscence of the surgical wound after TKR (0.1%) and high rate of complications resulting from this event. Of the 4 patients reported herein, 3 required new surgeries after initial cleaning, 2 lost their prostheses, and 1 lost a limb, requiring a transfemoral amputation due to uncontrollable infection.

Skin-related complications after arthroplasty are known to increase procedure-related complications. Vince et al⁷ and Vince and Abdeen⁸ reviewed the complications resulting from healing problems and recommended early intervention in necessary cases. Galat et al⁵ found surgical wound-related complications led to a greater number of surgeries and infections than cases that evolve with good healing, thus initial healing should be a key factor taken into account for the success of the arthroplasty. Galat et al⁵ also reported the presence of diabetes was related to worse surgical wound evolution. Accordingly, in the present series, the only patient who did not require a new intervention was also the only patient who did not have diabetes.

Regarding surgical wound exposure and problems related to a traumatic event, the literature on the subject is very scarce. Sershon et al⁶ reported a rate of only 0.05% of cases with this type of complication, with an unsuccessful rate of 43% for the initial therapy. Their case series⁶ included 14 patients, with only 11 cases of TKR, which demonstrates the rarity of this event. To the best of the authors’ knowledge, there is not another study that has addressed this patient population. It is important to emphasize these numbers because rarely will any center have a large case series of such a rare event. However, it is interesting to continue searching for these cases so that a compilation of studies can form the foundation for guiding treatment and identifying prognostic factors for success. Due to the need for new surgeries, the failure rate of the initial therapy was 75% in the present series.

In non-traumatic cases of surgical wound dehiscence, well-established factors exist in the literature that can minimize these events, including less traumatic techniques, oxygen supplementation, or negative pressure wound therapy.^9-11 However, in traumatic dehiscence, in addition to being potentially more serious due to the associated trauma, which can cause more diverse and heterogeneous injuries, and given the small cohort of patients and the low incidence of this complication, it remains difficult to make firm recommendations as to the optimal management.

For the prevention of a traumatic dehiscence, the only effective measure would be to avoid falls during the postoperative period. Although they appear to be infrequent events, the rates of postoperative falls in TKR patients vary from 12% to 38% according to di Laura Frattura et al¹²; in the first year following a TKR procedure, Chan et al¹³ reported the rate as 17.2%. Preventive measures should be implemented, including the preference for adductor canal blocks, performed in the middle third of the thigh, rather than a femoral nerve block for preservation of the quadricep strength,¹⁴ as well as attention to the patient’s environment, who are often elderly individuals. Even though all patients in this small series were women, the authors could not make conclusions that could eventually favor this gender for this type of complication.

The limitations of the present study include its small sample size, but the authors believe even with a small number of patients, this study is important due to the small number of similar cases reported in the literature. With a compilation of several studies, it would be possible in the future to perform an evaluation of all patients to obtain more robust data. The retrospective data analysis also may be considered a limitation, although data collection was performed prospectively. It is possible that other patients may have experienced a fall event with prosthesis exposure and sought treatment at another service; although this possibility seems unlikely.

Traumatic dehiscence with implant exposure during the initial postoperative period after TKR is an infrequent event (0.1%) with a high potential for complications (75%), which can lead to loss of the prosthesis and even the limb. Measures should be taken to prevent falls, and reports should be collected from several centers to expand the number of cases of this rare event in the search for prognostic factors and the optimal treatment algorithm. 

Authors: Camilo Partezani Helito, MD, PhD^1,2; Marcel Faraco Sobrado, MD^1,2; Marco Aurelio Cotegipe Negrelli, MD¹; José Ricardo Pécora, MD, PhD¹; Riccardo Gomes Gobbi, MD, PhD¹; and Fabio Janson Angelini, MD, PhD¹

Affiliations: ¹Grupo de Joelho, Instituto de Ortopedia e Traumatologia, Hospital das Clínicas HCFMUSP, Faculdade de Medicina, Universidade de São Paulo, São Paulo, SP, Brazil; and ²Hospital Sírio Libanês, São Paulo, Brazil

Correspondence: Marcel Faraco Sobrado, MD, Grupo de Joelho, instituto de Ortopedia e Traumatologia do Hospital das clínicas da Faculdade de Medicina da Universidade de São Paulo, rua Ovidio pires de campos 333, São Paulo, São Paulo 01246000 BRAZIL; marcelfs@gmail.com

Disclosure: The authors disclose no financial or other conflicts of interest.