Protein Energy Malnutrition as a Predictor of Early Recurrent Revisions After Debridement Surgery in Patients With Difficult-to-Treat Periprosthetic Infection

1 трАВмАтОЛОгИЯ И ОртОпЕДИЯ рОССИИ / TrAUmAToLogy AND orThopEDICS of rUSSIA Bozhkova S.A., Liventsov V.N., Tikhilov R.M., Romano C.L., Kochish A.Yu., Labutin D.V., Artyukh V.A. Protein Energy Malnutrition as a Predictor of Early Recurrent Revisions After Debridement Surgery in Patients With Difficult-to-Treat Periprosthetic Infection. Травматология и ортопедия России. https://doi.org/10.17816/2311-2905-1717.


Introduction
Chronic prosthetic joint infection (PJI) is one of the serious complications of total joint arthroplasty. The worst treatment results are observed in infections caused by drug-resistant strains of bacteria [1,2]. Lately, the so-called difficult-to-treat (DTT) infection has been defined as an infection caused by difficult-to-eradicate pathogens such as rifampicinresistant staphylococci, ciprofloxacin-resistant gram-negative bacteria and Candida fungi [3]. Multiple attempts to eradicate this infection through debridement surgery result in large bone defects. Resection arthroplasty of the hip coupled with a flap of vastus lateralis muscle filling the bone defect has been shown as an effective way to achieve the control of infection, particularly, in case of the DTT PJI and in some cases restore its function [4,5]. This surgical approach improves the blood supply and the delivery of oxygen, leucocytes and antibiotics to the affected tissue which helps to eradicate pathogens responsible for the chronic infection decreasing the risk of PJI recurrence [6]. трАВмАтОЛОгИЯ И ОртОпЕДИЯ рОССИИ / TrAUmAToLogy AND orThopEDICS of rUSSIA  The imbalance of nutrients or protein-energy malnutrition (PEM) is one of the main causes of increased morbidity and mortality, reduced quality of life, increased frequency and length of hospital stay, and associated health care costs [7,8]. PEM is a proven risk factor for the development of postoperative complications in orthopedics including patients who underwent arthroplasty of large joints as well as spinal surgery [9,10,11]. In addition, PEM not only negatively affects postoperative recovery and wound healing but also increases the risk of early and late surgical site infection [12]. The incidence of PEM in orthopedic patients requiring arthroplasty varies from 8.5% to 50% [13, 14,15]. PEM has been shown both as a risk factor for the onset of prosthetic joint infection and a predictor of the infection recurrence after primary arthroplasty [16]. Surrogate markers of PEM such as total lymphocyte count, serum albumin and transferrin have been extensively used to assess the underlying nutritional status of patients [17,18]. Increased risk of complications after total hip arthroplasty including the recurrence of PJI has been associated with hypoalbuminemia [19]. Low levels of albumin in combination with hemoglobin were also associated with the development of PJI [18].
The number of studies that have addressed the role of PEM in the development of PJI is limited. Most of them are hindered by heterogeneous cohorts without any discrimination between cases of primary and revision arthroplasty, or PJI and trauma-related surgery [15,20]. Given this, we aimed to evaluate the prevalence of PEM in patients DTT PJI of the hip. In addition, the association of preoperative PEM with surgical complications requiring early postoperative revision after resection arthroplasty with a muscle flap or a cement spacer implantation was assessed.

Methods
The data for the study were retrieved from the records of the local PJI register covering the period of 2012-2018. The study was approved by the institutional review board. General informed consent was obtained as a part of the hospital admission.
Of all 132 patients included in the study, 57 were treated with a pedicled muscle flap after the removal of the endoprosthesis. Surgical debridement of the affected tissue was followed by subsequent transplantation of a vastus lateralis muscle flap of the femur retaining the vascular supply. Alternatively, the remaining 75 patients were treated with surgical debridement and installation of an antimicrobial cement spacer in place of the metal components.
The main inclusion criteria were chronic DTT PJI of the hip caused by rifampicin-resistant staphylococci, ciprofloxacin-resistant Gram-negative bacteria and/or Candida fungi. The main laboratory parameters which were taken for the evaluation of the nutritional status included total lymphocyte count, hemoglobin, albumin and total protein. The cut-off values suggesting PEM were defined as follows: total lymphocyte count (TLC) at <1.5×10 9 cells/L, hemoglobin at <12.0 g/dL, albumin at <3.5 g/dL and total protein at <6.5 g/dL. The severity of PEM was defined as the percentage of cases with one, two or three and more of these laboratory values below the cut-off level.
Postoperatively, all patients received intravenous antibiotic therapy followed by oral antibiotics for 6-8 weeks. The proportion of patients requiring revision surgery in the early postoperative period was evaluated regardless of the cause.

Statistic analysis
The data were analyzed using Statistica v 10 (StatSoft, Russia) and GraphPad Prism v 9.0 (GraphPad, USA). Categorical variables are shown as percentages. They were evaluated using Fischer's exact test.
Odds ratios (OR. were calculated with respective confidence intervals (CI) at the confidence level of 95%. The P-value below 0.05 was considered as the indication of statistical significance.

Results
General characteristics of the cohort are depicted in Figure 1. Patients treated with a muscle flap had a longer median duration of PJI and a lower percentage of at least two episodes of infection recurrence before the hospital admission as compared to patients managed using two-stage revision with a cement spacer.
Hypoalbuminemia and decreased hemoglobin were the most common markers of malnutrition in the cohort (Fig. 2).
Whereas the overall fraction of patients with one abnormal malnutrition marker was comparable to those without PEM in the muscle flap group but higher in the cement spacer group (Fig. 3).
Association of the advanced malnutrition (at least three abnormal markers) and post-operative complications requiring surgical revision was only statistically evident in the cement spacer group (Fig. 4, 5).

Discussion
Despite the worse functional outcomes compared to the use of an antimicrobial spacer, resection arthroplasty of the hip joint using a vastus lateralis muscle flap is an effective procedure to achieve PJI control [4]. The cohort of patients in the present study has previously been assessed to compare the efficacy of resection arthroplasty with a muscle flap and a cement spacer implantation for treatment of DTT PJI [5]. In all patients with the muscle flap and cement spacer, the infection remission was achieved in 96.5% (n = 55) and 45.3% (n = 34) of cases, respectively. Late recurrence of PJI was only in 2.5% (n = 2) of cases with the muscle flap vs 26.7% (n = 20) of cases with a cement spacer. Thus, resection arthroplasty with a vastus lateralis muscle flap proved to be a strategy of choice for management of recurrent DTT PJI. In contrast, for this    The average preoperative prevalence of PEM in patients undergoing arthroplasty depending on the heterogeneity of the cohort has been widely reported at 8.5%, 12.3%, 53% and 80% [13,15,15,20,21]. At the time of hospital admission, most of our patients (>70%) had laboratory signs of PEM. It is well known that prolonged systemic inflammatory response due to the persistent infection eventually leads to the loss of muscle mass and function, which could itself be the manifestation of PEM [7,8]. For instance, cases of revision arthroplasty due to septic complications had a higher rate of hypoalbuminemia as compared to aseptic cases [10]. Given this, it is feasible to assume that long-term chronic DTT PJI with multiple prior failed revisions could have been the primary reason for the impaired preoperative nutritional status of our patients. This would also explain the rather high prevalence of PEM in our study compared to previous reports. It is noteworthy to mention that the already abnormal nutritional status of patients at the time of hospital admission could have also been further aggravated by the surgery.
Several laboratory values have been considered as surrogate markers for the early diagnosis of PEM. For instance, albumin and hemoglobin have been shown as independent factors associated with PJI occurrence after primary elective THA [18]. Whereas albumin had the highest specificity and a positive predictive value compared to all other markers such as total lymphocyte count and transferrin. Other studies have also confirmed the association of hypoalbuminemia with complications of arthroplasty from surgical site infection to pneumonia, cardiovascular pathology, and urinary tract infection [9,21,22,23]. Like the above reports, our findings showed low albumin and hemoglobin as the most prevalent (>50%) abnormal laboratory values.
Our previous study showed that in the early postoperative period, the only indication for the secondary revision in the spacer group (28%, n = 21) was the recurrence of the infection. Alternatively, main indications (35%, n = 20) for post-operative revision in the muscle flap group were necrosis of the flap (8.8%, n = 5), culture-negative hematomas (10.5%, n = 6), and recurrence of the infection including wound dehiscence or prolonged wound drainage for more than 7 days (15.7%, n = 9) [5]. PEM was found to be associated with the prolonged surgical site discharge and subsequent deep infection which required additional treatment after elective total joint arthroplasty [16]. In our cohort, a decrease of 3 or more laboratory markers of PEM increased the risk of postoperative wound revision by 6 times in patients with a cement spacer (OR 6.11, CI95% 1.06-35.35, P<0.04) and 2 times in patients with a muscle flap (OR 2.0, CI95% 0.47-8.56, P = 0.35) (see Fig. 4, 5).
Due to the retrospective nature of our study, some medical records such as BMI index and a complete list of concomitant pathology were missing for a part of the cohort and thus were omitted from the analysis.
Enteral nutrition is a process of providing the patients with necessary supplements has been long regarded as a method of choice for PEM correction in general medicine and hip replacement in geriatric patients [24]. Unfortunately, due to the limited number of studies addressing potential benefits of PEM correction in orthopedic patients the exact guidelines for preventing early post-operative complications including PJI recurrence via nutritional supplementation before and after arthroplasty are yet to be established. In a clinical study of 162 cases of primary knee arthroplasty, it has been shown that patients with constant perioperative nutritional management had a lower number of postoperative complications such as wound drainage, hematomas, superficial and deep infection as well as the reduced number of albumin transfusions and a shorter hospital stay [25]. This suggests that implementing the protocols for hypoproteinemia correction before any major orthopedic surgery such as primary and revision arthroplasty including patients with PJI could improve its outcome.
Overall, both previous reports and our data show that malnutrition is associated with the occurrence of surgical site infection and relapse of PJI after revision arthroplasty. In particular, the risk of postoperative revision was significantly associated with multiple pathological markers of PEM in two-stage revision with a cement spacer but not with a muscle flap surgery. Considering the mounting evidence that malnutrition is a modifiable risk factor for PJI, more studies are required to demonstrate the efficacy of preoperative PEM correction.

Authors' contributions
Svetlana A. Bozhkova -research concept and design, data statistical processing, manuscript writing and editing.
Vitaly N. Liventsov -data collection and analysis, manuscript writing.