Advertisement

A novel treatment strategy using continuous local antibiotic perfusion: A case series study of a refractory infection caused by hypervirulent Klebsiella pneumoniae

Open AccessPublished:December 19, 2020DOI:https://doi.org/10.1016/j.jos.2020.11.010

      Abstract

      Background

      Conventional topical antimicrobial therapy cannot maintain a constant local concentration, resulting in uncontrolled infection and complications. We propose continuous local antibiotic perfusion (CLAP), which can maintain a constant appropriate local antibiotic concentration for a long time with less invasiveness and complications. CLAP is clearly different from traditional treatment because it uses negative pressure to direct the continuously infused antibiotic solution to the center of infection and excrete it outside the body. This study aimed to demonstrate the effectiveness of CLAP by presenting cases in which even refractory bone and soft-tissue infections caused by the hypervirulent Klebsiella pneumoniae (hvKp) could be cured without significant tissue loss and dysfunction.

      Methods

      This study is a case series in which four patients with limb infection due to hvKp were treated by CLAP. hvKp was defined by a positive string test. The therapy included intra-soft-tissue antibiotic perfusion and intramedullary antibiotic perfusion. Gentamicin (60 mg/50 cc) was infused continuously through dual-lumen tubes and bone marrow needles at low-flow rates (2 mL/h). Negative pressure was used to collect the antimicrobial solution and eliminate the dead space.

      Results

      The infection was controlled in all four patients after a mean period of 44.3 days. The mean maximum blood concentration of gentamicin was 1.6 μg/dL, and no cases of renal dysfunction or ototoxicity occurred. After CLAP, wound closure was required in two patients and iliac bone grafting was required in one patient. As sequelae, there were one case of osteoarthritis and one case of higher brain dysfunction due to hypoxia.

      Conclusions

      Our results suggest that intractable hvKp infections can be controlled by CLAP. CLAP may give us the option to directly control local infections with less systemic complications. Therefore, it is considered a valuable treatment for further basic and clinical research, and this research report may be a first step.

      1. Introduction

      To treat bone and soft-tissue infections, orthopedic surgeons may perform debridement of the infected tissue, implant removal, and limb amputation. If the infection subsides after these treatments, they try to restore the dysfunction and reconstruct tissue defects using various techniques, such as microsurgery [
      • Rhomberg M.
      • Frischhut B.
      • Ninkovic M.
      • Schwabegger A.H.
      • Ninkovic M.
      A single-stage operation in the treatment of chronic osteomyelitis of the lower extremity including reconstruction with free vascularized iliac bone graft and free-tissue transfer.
      ] and the Ilizarov technique [
      • Dendrinos G.K.
      • Kontos S.
      • Lyritsis E.
      Use of the Ilizarov technique for treatment of non-union of the tibia associated with infection.
      ]. This strategy has been widely applied but is not ideal.
      In contrast, multiple osteomyelitis and abscesses associated with systemic infections treated with a similar strategy can cause irreversible severe tissue loss and dysfunction. Therefore, in such cases, orthopedic surgeons only perform incision drainage and administer antibiotics, waiting for the infection to settle. To help orthopedic surgeons treat infections and avoid undesirable situations in which they cannot treat the infection, minimally invasive treatment methods that can directly control local infections are needed.
      In the orthopedic field, various methods of local delivery of antimicrobials for control of local infections have been studied. For example, antibacterial powders have been sprayed in the surgical field [
      • Li S.
      • Rong H.
      • Zhang X.
      • Zhang Z.
      • Wang C.
      • Tan R.
      • Wang Y.
      • Zheng T.
      • Zhu T.
      Meta-analysis of topical vancomycin powder for microbial profile in spinal surgical site infections.
      ], antibacterial liquids have been injected locally after surgery [
      • Lawing C.R.
      • Lin F.C.
      • Dahners L.E.
      Local injection of aminoglycosides for prophylaxis against infection in open fractures.
      ], and antibacterial drugs have been locally administered in foreign materials such as bone cement [
      • Wininger D.A.
      • Fass R.J.
      Antibiotic-impregnated cement and beads for orthopedic infections.
      ]. In these methods, the release of the antibacterial agent is temporary and the local concentration of the antibacterial agent is unstable and only slightly above the minimum inhibitory concentration (MIC) [
      • Anagnostakos K.
      • Wilmes P.
      • Schmitt E.
      • Kelm J.
      Elution of gentamicin and vancomycin from polymethylmethacrylate beads and hip spacers in vivo.
      ,
      • Klemm K.
      The use of antibiotic-containing bead chains in the treatment of chronic bone infections.
      ]. Furthermore, if bone cement is inserted, bacteria can attach to its surface after the release of the antimicrobial agent, and can grow in the dead space after the bone cement has been removed [
      • Neut D.
      • Van de Belt H.
      • Stokroos I.
      • van Horn J.R.
      • Van der Mei H.C.
      • Busscher H.J.
      Biomaterial-associated infection of gentamicin-loaded PMMA beads in orthopaedic revision surgery.
      ].
      Here, we developed a method of continuous local antibiotic perfusion (CLAP, Fig. 1A) that can maintain a high local concentration of antibacterial agents at the site of infection for a long time. Previously, a method of injecting antibiotics through a catheter placed in the body against periprosthetic infection has been reported [
      • Whiteside L.A.
      • Roy M.E.
      One-stage revision with catheter infusion of intraarticular antibiotics successfully treats infected THA.
      ], but CLAP is completely different from the previous method in the following four points: (1) Removes only biologically inactive tissue not to increase the dead space, (2) performs dead space control applying negative pressure to all incisions and abscess formation sites, (3) injects appropriate concentration antibacterial drugs continuously and induces them to the infected site by negative pressure, and (4) reduces the systemic side effects of antibacterial drugs by continuously draining them by negative pressure.
      Fig. 1
      Fig. 1(A): Image of continuous local antibiotic perfusion (CLAP). (B, C, D): A series of photographs confirming the state of CLAP using a contrast agent. Antibiotics administered topically to a bone or abscess are induced to the center of the infection by negative pressure and collected via dual-lumen tubes.
      We focused on hypervirulent Klebsiella pneumoniae (hvKp). Owing to its high viscosity, hvKp can be disseminated throughout the body to form abscesses that are difficult to drain and are intractable [
      • Shon A.S.1
      • Bajwa R.P.
      • Russo T.A.
      Hypervirulent (hypermucoviscous) Klebsiella pneumoniae: a new and dangerous breed.
      ,
      • Prokesch B.C.
      • TeKippe M.
      • Kim J.
      • Raj P.
      • TeKippe E.M.
      • Greenberg D.E.
      Primary osteomyelitis caused by hypervirulent Klebsiella pneumoniae.
      ]. Moreover, hvKp easily acquire resistance to antibacterial drugs [
      • Lee C.R.
      • Lee J.H.
      • Park K.S.
      • Jeon J.H.
      • Kim Y.B.
      • Cha C.J.
      • Jeong B.C.
      • Lee S.H.
      Antimicrobial resistance of hypervirulent Klebsiella pneumoniae: epidemiology, hypervirulence-associated determinants, and resistance mechanisms.
      ] and Klebsiella is listed by the World Health Organization as a pathogen that has priority for the development of new drugs [
      • Tacconelli E.
      • Magrini N.
      Global priority list of antibiotic-resistant bacteria to guide research, discovery, and development of new antibiotics.
      ].
      We demonstrate the potential of CLAP, showing that CLAP can control infection even in case series of refractory bone and soft tissue infections of hvKp without significant tissue loss, dysfunction, or systemic complications.

      2. Materials and methods

      This study was a case series including four patients who had bone and soft-tissue infections due to hvKp, such as multiple osteomyelitis, deep abscess of an extremity, and necrotizing fasciitis. hvKp was defined by a positive string test, and no genomic analysis was performed.
      The patients were treated with CLAP at our hospital between January and December 2018. We fully explained to the patients that this treatment was an off-label use of antibacterial drugs. This treatment was approved by the Certified Clinical Research Review Board of our hospital as an off-label use of aminoglycoside antibiotics. The patients and/or their families were informed that the case data would be submitted for publication, and they provided their consent.
      The examination items were age, sex, country of birth, underlying condition, positive culture, site of infection, treatment period, maximum gentamicin blood concentration, additional surgery, and sequelae.

      2.1 Treatment protocol

      CLAP was carried out according to the following treatment protocol:
      • (1)
        Detection. In order to estimate the spread of an abscess as closely as possible, contrast-enhanced computed tomography (CT) should be performed. If osteomyelitis is suspected, magnetic resonance imaging (MRI) should be performed to estimate the spread of osteomyelitis in detail.
      • (2)
        Design. First, the spread of the infection is estimated in detail from the imaging findings (Fig. 2A and B). Based on this, the arrangement of the dual-lumen tube (Salem Sump Tube™, Cardinal, unknown) and the position of insertion of the bone marrow needle (Tohoku University bone marrow puncture needle; Senko Medical Instrument Manufacturing Co., Tokyo, Japan) are determined.
        Fig. 2
        Fig. 2(A): Preoperative contrast-enhanced computed tomography (CT) image of the left lower leg. Green: infected site in the soft tissue. Yellow: dual-lumen tube. (B): Preoperative magnetic resonance imaging (MRI) image of both lower legs. Orange: infected site in the bone. Gray: bone marrow needle. (C): Intraoperative photograph at the start of continuous local antibiotic perfusion (CLAP). (D): Photo of both lower limbs after CLAP. Signs of infection have disappeared and wounds such as incisions and drain holes have been closed.
      • (3)
        Defense of the tissue. In surgery, all procedures should be performed in a protective manner to prevent further invasion of the tissue (Fig. 2C). The tissue that is clearly not biologically active is removed, and the bone marrow needles and dual-lumen tubes are placed to be as minimally invasive as possible, based on preoperative examination. To avoid obstruction during treatment, we inserted a dual-lumen tube with the largest possible diameter, depending on the size of the dead space. At this point, if there is a fracture, internal fixation should be performed to obtain enough fixation according to the principles of fracture treatment. Performing osteosynthesis during the first operation stabilizes the bone and protects the surrounding tissue.
      • (4)
        Do thoroughly. For 14 days after surgery, gentamicin (60 mg/50 cc) is continuously administered at a low flow rates (2 mL/h) by a syringe pump through the bone marrow puncture needles and dual-lumen tubes. Of course, there are parts (in the blood and outside the surgical field) where antibiotics do not reach sufficiently with CLAP; therefore, systemic administration of antibiotics based on drug sensitivity is often necessary. If there are remaining signs of infection or wound healing problems, Steps 1–4 are repeated thoroughly until Step 5 is reached.
      • (5)
        Disappearance of signs of infection. The goals of CLAP are wound closure and improvement of clinical findings, such as redness, tenderness, spontaneous pain, exudate, and fever (Fig. 2D). We prioritized clinical findings and C-reactive protein (CRP) and erythrocyte sedimentation rate (ESR), which are markers of inflammation that are generally considered to be proportional to the degree of infection [
        • Lin Z.
        • Vasudevan A.
        • Tambyah P.A.
        Use of erythrocyte sedimentation rate and C-reactive protein to predict osteomyelitis recurrence.
        ], were used only as a reference finding. We comprehensively evaluated wound healing, clinical symptoms, and CRP/ESR, and considered the termination of antibiotic administration.

      2.2 Surgical procedure

      2.2.1 Intra-soft-tissue antibiotic perfusion

      It is used for soft-tissue infections. For spaces in which soft-tissue infection is a concern, the tip of the dual-lumen tube where the antimicrobial is released should be placed deeper and farther into the space. In contrast, the dual-lumen tube discharge holes should be placed over a wide area in the space (Fig. 2A). As a result, the space where infection is a concern is filled with antibacterial drugs, and a negative pressure is applied uniformly and drained sufficiently.

      2.2.2 Intramedullary antibiotic perfusion

      It used for bone infection. A bone marrow needle is inserted into the area with suspected bone infection. If possible, the needle is placed so that the affected area is sandwiched (Fig. 2B). First, a hole is made in the cortical bone with a 2.4 mm Kirschner wire, and a bone marrow needle is inserted using a hammer. Reverse flow of blood from the bone marrow needle is a good guide for correct placement of the needle. Provided that there is no allergy of contrast media, contrast media should be injected to ensure that the antibiotic reaches the site of infection (Fig. 1B, C, D).

      2.2.3 Dead space management

      Obvious necrotic tissue and abscesses are then removed, and the wound is closed as much as possible. Thus, the space is made more airtight and more easily filled with the antibacterial agent. The wound is then covered with negative pressure wound therapy (NPWT, Renasys; Smith & Nephew Medical Ltd., Kingston upon Hull, UK), whether it is open or closed (Fig. 3A). The suction tube of Renasys is branched, and the suction port of the Salem Sump Tube is connected to the tip of the branch (Fig. 3B). This keeps the same negative pressure on the wound surface and inside the wound.
      Fig. 3
      Fig. 3(A): An example of continuous local antibiotic perfusion (CLAP). A soft port of Renasys was attached to the wound. (B): The suction ports of the dual-lumen tubes inserted into the wound were connected to the branch of the suction tube of Renasys.

      2.3 Postoperative management

      The suction pressure of the NPWT is set at a low negative pressure (from −25 to −80 mmHg). The negative pressure is adjusted according to the condition of the wound (open or closed, size of the dead space, and fragility of the soft tissue). The ward staff needs to periodically check that the drain is not clogged by confirming that the gas–liquid interface in the tube has moved after rapid delivery of the drug solution. If the tube continues to be clogged, the antibacterial drug will go around the body and cause complications such as renal dysfunction. Therefore, it is necessary to pay close attention not to clogged drainage. In order to assess the systemic effects of CLAP, the blood levels of gentamicin are measured once a week after surgery. We have not evaluated the amount of fluid discharged because the exudate in the wound was also mixed and could not be accurately evaluated.

      3. Results

      Four patients with hvKp were treated with CLAP. In all four cases, the infection was controlled after a mean duration of CLAP of 44·3 days (range: 30–61 days) without significant tissue loss. The average age of the patients was 53 years old (range: 46–72 years old). One patient was a male and three were females. Patient 2 was from the Philippines, and the other three patients were from Japan and had no history of travel outside Japan. Patient 1 was in contact with Vietnamese people at work. Two patients had diabetes, and one patient was using an immunosuppressant for systemic lupus erythematosus (SLE). String test–positive K. pneumoniae was detected in all patients. In all patients except Patient 1, blood culture at the time of first treatment was positive for K. pneumoniae. Cultures of abscesses or bone marrow fluid were positive for K. pneumoniae in all patients. The average duration of antibiotic treatment, including systemic administration such as infusion and oral administration, was 60.0 days (range: 35–70 days). The average number of CLAP sites was 3.2 (range: 1–5). The mean maximum blood concentration of gentamicin was 1.6 μg/dL (range: 0.8–1.9 μg/dL), and there were no cases of renal dysfunction or ototoxicity. After CLAP was completed, wound closure was required in two patients and iliac bone grafting was required in one patient. As sequelae, there were one case of osteoarthritis after pyogenic arthritis and one case of higher brain dysfunction due to hypoxia during sepsis. All cases are being followed up, but at least one year after the end of treatment, the progress has been favorable without relapse of the infection. The data for each case are shown in Table 1.
      Table 1Details of each case.
      CaseSexAge (years)Country of birthUnderlying conditionsPositive cultureSite of infectionTreatment period (days)Maximum gentamicin blood concentration (μg/dL)Additional surgerySequelae
      1Male47JapanNoneFemoral abscess, lower leg abscessLiver, lungs, femur, lower leg611.9Iliac bone graft
      2Female47PhilippinesDiabetes mellitusBlood, stool, talar, tibiaSepsis, femur, tibia, fibula, talar, ankle joint431.6Ankle osteoarthritis
      3Female72JapanDiabetes mellitus, obstructive arteriosclerosis, chronic renal failureBlood, lower leg abscessSeptic shock, superior mesenteric artery, lower leg430.8Wound closure
      4Female46JapanSLEBlood, buttocks abscessSeptic shock (cardiopulmonary arrest), buttocks abscess301.9Wound closureHigher brain dysfunction
      Mean5344.31.6
      SLE: systemic lupus erythematosus.
      Bold emphasize septic infection.

      3.1 Case presentation

      3.1.1 Case 1

      A 47-year-old man with no history of underlying disease visited the outpatient clinic with left femoral pain, redness, and swelling without an apparent cause. During the previous consultation, CT findings showed multiple abscesses in the left femur, left lower leg, liver, and lungs. Therefore, the patient started antimicrobial treatment with meropenem after incision and drainage with the previous physician. K. pneumoniae was detected by a bacterial culture test for abscesses (Blood culture was not performed.) The previous physician did not perform the string test. One week after the start of treatment, a pathological fracture occurred in the left femur when the patient tried to flex his left hip joint by himself (Fig. 4A and B). He was transferred to our hospital for infection control and fracture management. The abscess was punctured for bacterial culture, and K. pneumoniae was detected again. Because there were multiple systemic abscesses, it was suspected that the pathogen was hvKp. Therefore, a string test was performed, with a positive result. There were no intracranial and intraocular lesions. When the fracture was exposed during the first operation, there was a viscous abscess filling the bone and medullary cavity, and the periosteum had melted (Fig. 4C). CLAP was performed, and at the same time the femoral fracture was fixed with an intramedullary nail (Fig. 5A). Bone resorption occurred because of the initial infection; however, after CLAP, a new periosteal bone was formed, and bone fusion was partially obtained at the location of the fracture (Fig. 5B). Two months after the initiation of CLAP, the antibiotic treatment was changed from intravenous injection to oral administration, and the patient was discharged from the hospital. Two months later, the ESR became normal, and antibiotic treatment was completed. A free iliac bone graft was performed 12 months after the first surgery because a bone defect was left in a part of the pathological fracture (Fig. 5C). After bone grafting, bone fusion was obtained all around (Fig. 5D). The patient was considered to be free from relapse if they completed 2.5 years without any antibiotics, could walk with a full load, and had returned to the same job as before. Table 2 shows the transition of C-reactive protein/erythrocyte sedimentation rate. Erythrocyte sedimentation rate remains negative after antibiotic treatment.
      Fig. 4
      Fig. 4(A, B): X-ray and contrast-enhanced computed tomography (CT) of a left femoral trochanteric fracture. Osteomyelitis over the entire length of the femur and multiple abscesses spreading around the femur were observed. (C): Intraoperative photograph of the first operation (fixed with an intramedullary nail). When only the skin was incised, an abscess around the bone erupted, exposing the periosteum-melted femur.
      Fig. 5
      Fig. 5(A, B, C, D): Radiographs of the left femur at 2.5, 4.5, 12, and 19 months after the first operation. (E): Photo of left thigh 19 months after first surgery. There are no signs of relapse of the infection, leaving only small scars, such as incisions and drain holes.
      Table 2Changes in C-reactive protein (CRP) and erythrocyte sedimentation rate (ESR) in case 1.
      Postoperative days (week)−101238123648120
      Events during treatmentTransfer to our hospitalfirst operationend of CLAPdischarge from hospital Change to oral antibioticsend of oral antibioticsbefore bone graftingfinal follow up
      CRP (mg/dL)11.8710.318.276.493.969.472.140.330.140.09
      ESR (mm)11094734931811

      3.1.2 Case 4

      A 45-year-old woman treated for SLE visited our hospital's outpatient department with back pain and developed ventricular fibrillation. Lifesaving treatment was performed by the medical emergency team, and extracorporeal membrane oxygenation (ECMO) was performed. Hyperkalemia associated with septic shock was thought to be the cause of the ventricular fibrillation. Contrast-enhanced CT was performed to find the cause of sepsis, and revealed an abscess in the left buttock (Fig. 6A). An emergency operation was performed to control the sepsis source. The surgery revealed a high-viscosity abscess filling the space from the subcutaneous tissue to the muscles (Fig. 6B). We judged excision of the infected tissues would cause death because of the widely spread abscess and her poor general condition. Therefore, CLAP was performed as minimally invasive source control of the infection. After incision to the shallowest abscess, dual-lumen tubes were inserted at the site of abscess spread on contrast-enhanced CT. The abscess was washed using a dual-lumen tube and the wound was closed and covered with NPWT. After surgery, her general condition was stabilized using source control with CLAP while maintaining circulatory dynamics with ECMO (Fig. 6C). ECMO was terminated after 7 days, and the patient was discharged from the intensive care unit after 30 days. Since then, no worsening of her general condition and no relapse of local infection signs has been observed. Although wound healing was delayed because of the underlying disease, the wound was closed after three months after washing and ointment application of bucladesine sodium (Fig. 6D). Although the patient's higher brain dysfunction due to her initial poor general condition remained, there was no recurrence of infection after 1 year. CRP/ESR did not become negative due to repeated aspiration pneumonia and urinary tract infection.
      Fig. 6
      Fig. 6(A): Computed tomography (CT) at the onset of septic shock. An abscess containing air following the sciatic hiatus was observed. (B): Intraoperative photograph of the first operation. (C): Postoperative photograph in the intensive care unit. (D): Photograph of the left femur 3 months after the first operation.

      4. Discussion

      In this study, we used CLAP to treat intractable abscesses and osteomyelitis caused by hvKp. The infection was controlled in all patients, and they were cured without significant tissue loss and dysfunction.
      There is a history of researchers trying to treat infections by locally administering antibacterial agents to the site of infection [
      • Li S.
      • Rong H.
      • Zhang X.
      • Zhang Z.
      • Wang C.
      • Tan R.
      • Wang Y.
      • Zheng T.
      • Zhu T.
      Meta-analysis of topical vancomycin powder for microbial profile in spinal surgical site infections.
      ,
      • Lawing C.R.
      • Lin F.C.
      • Dahners L.E.
      Local injection of aminoglycosides for prophylaxis against infection in open fractures.
      ,
      • Wininger D.A.
      • Fass R.J.
      Antibiotic-impregnated cement and beads for orthopedic infections.
      ]. In contrast, a method of injecting antibiotics once a day from a catheter placed in the body has also been reported [
      • Whiteside L.A.
      • Roy M.E.
      One-stage revision with catheter infusion of intraarticular antibiotics successfully treats infected THA.
      ]. However, these methods cannot achieve constant local concentrations: too high a concentration can cause complications like renal dysfunction [
      • James A.
      • Larson T.
      Acute renal failure after high-dose antibiotic bone cement: case report and review of the literature.
      ,
      • Aeng E.S.
      • Shalansky K.F.
      • Lau T.T.
      • Zalunardo N.
      • Li G.
      • Bowie W.R.
      • Duncan C.P.
      Acute kidney injury with tobramycin-impregnated bone cement spacers in prosthetic joint infections.
      ] and too low a concentration can induce drug-resistant bacteria [
      • Anagnostakos K.
      • Hitzler P.
      • Pape D.
      • Kohn D.
      • Kelm J.
      Persistence of bacterial growth on antibiotic-loaded beads: is it actually a problem?.
      ,
      • Drlica K.
      The mutant selection window and antimicrobial resistance.
      ]. Therefore, in order to control infections safely and reliably, it is important to keep the local concentration of the antibacterial drug at a constant appropriate level. In our opinion, conventional treatment lacked the stability of topical antibiotics, as well as a mechanism for collecting administered antibiotics.
      Minimum biofilm eradication concentration (MBEC) is usually measured on biofilms made in 24 h [
      • Lebeaux D.
      • Chauhan A.
      • Rendueles O.
      • Beloin C.
      From in vitro to in vivo models of bacterial biofilm-related infections.
      ,
      • Ceri H.
      • Olson M.E.
      • Stremick C.
      • Read R.R.
      • Morck D.
      • Buret A.
      The Calgary Biofilm Device: new technology for rapid determination of antibiotic susceptibilities of bacterial biofilms.
      ], but it is also known that the more mature the biofilm, the higher the MBEC [
      • Singla S.
      • Harjai K.
      • Chhibber S.
      Susceptibility of different phases of biofilm of Klebsiella pneumoniae to three different antibiotics.
      ,
      • Høiby N.
      • Henneberg K.Å.
      • Wang H.
      • Stavnsbjerg C.
      • Bjarnsholt T.
      • Ciofu O.
      • Johansen U.R.
      • Sams T.
      Formation of Pseudomonas aeruginosa inhibition zone during tobramycin disk diffusion is due to transition from planktonic to biofilm mode of growth.
      ]. Conversely, it has been reported that the longer a biofilm is exposed to an antimicrobial, the more its MBEC decreases [
      • Castaneda P.
      • McLaren A.
      • Tavaziva G.
      • Overstreet D.
      Biofilm antimicrobial susceptibility increases with antimicrobial exposure time.
      ]. In CLAP, an appropriate local antimicrobial concentration was maintained over a long period, so it was thought that even attached bacteria that had formed biofilms could be eradicated.
      A 5D approach has been proposed for the treatment of orthopedic infections [
      • Winkler H.
      Treatment of chronic orthopaedic infection.
      ]. This approach has the following components: (1) detection, (2) debridement, (3) destruction, (4) dead space management, and (5) decontamination. As shown below, CLAP can satisfy all the 5D with minimal invasion: (1) identify the location of the infection using contrast-enhanced CT and MRI, (2) remove the tissue with apparently lost biological activity, (3) eradicate the abscesses and biofilms with constant appropriate concentration local antibiotics, (4) eliminate the dead space by applying continuous negative pressure within the wound, and (5) administer antibiotics for the floating bacteria. Thus, CLAP might be one of ideal therapy for orthopedic infection.
      In this study, we focused on cases of intractable bone and soft-tissue infections caused by hvKp. The characteristics of infection by hvKp are its viscosity and systemic dissemination. hvKp infections have a mortality rate of 3%–42% and 55% in community-acquired pneumonia with bacteremia [
      • Shon A.S.1
      • Bajwa R.P.
      • Russo T.A.
      Hypervirulent (hypermucoviscous) Klebsiella pneumoniae: a new and dangerous breed.
      ]. Necrotizing fasciitis due to hvKp is reported to cause 47% of deaths by septic shock and has a very high probability of limb amputation [
      • Cheng N.C.
      • Yu Y.C.
      • Tai H.C.
      • Hsueh P.R.
      • Chang S.C.
      • Lai S.Y.
      • Yi W.C.
      • Fang C.T.
      Recent trend of necrotizing fasciitis in Taiwan: focus on monomicrobial Klebsiella pneumoniae necrotizing fasciitis.
      ]. To date, there have been no published reports detailing the surgical treatment of limb infections with hvKp [
      • Russo T.A.
      • Marr C.M.
      Hypervirulent Klebsiella pneumoniae.
      ]. Liver abscesses from hvKp are said to be refractory to treatment because the drain can easily become blocked because of its high viscosity [
      • Tan Y.M.
      • Chung A.Y.
      • Chow P.K.
      • Cheow P.C.
      • Wong W.K.
      • Ooi L.L.
      • Soo K.C.
      An appraisal of surgical and percutaneous drainage for pyogenic liver abscesses larger than 5 cm.
      ]. In CLAP, clogging of the drainage tube is prevented by draining the pus while continuing the flow of the antibiotics. Thus, CLAP can even overcome intractable bone and soft-tissue infections by hvKp.
      The main limitation of the present study is that it is a series of only four cases and there were no controls to show the effectiveness of CLAP statistically. However, the study fully demonstrates the effectiveness of CLAP in four cases of very rare infections by hvKp, in the sense that the therapy not only saved lives but also preserved the functional limb. We are currently testing a variety of cases, including post-traumatic infections, periarticular joint infections, and post-spine surgery infections, and will report on large case series and collaborative research with other institutions in the future.
      The second limitation is the definition of hvKp infection. In this study, hvKp infection was defined by a positive string test. However, it has been reported that viscosity and virulence do not necessarily match, and many studies have been conducted to search for virulence factors by genomic analysis [
      • Catalán-Nájera J.C.
      • Garza-Ramos U.
      • Barrios-Camacho H.
      Hypervirulence and hypermucoviscosity: two different but complementary Klebsiella spp. phenotypes?.
      ]. No genomic analysis was performed in this study, and therefore the four patients may not have been infected by the same pathogen. However, it was reasonable to describe the pathogen as “hypervirulent” K. pneumoniae, because these patients developed multiple abscesses and osteomyelitis due to bloodstream infection with K. pneumoniae and were resistant to conventional therapy.

      5. Conclusion

      We have successfully used CLAP to cure the refractory infection associated with hvKp. It was suggested that CLAP may have potential that directly controls local infection without significant tissue loss, dysfunction, and systemic complication. Therefore, it is a valuable treatment and the first step for further basic and clinical research.

      Funding

      This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

      Declaration of competing interest

      The authors declare no conflicts of interest associated with this manuscript.

      Acknowledgements

      We are deeply grateful to the orthopedic surgeons (especially the hand surgery team at Chiba University), emergency physicians, and ward staff of each hospital who willing accepted and supported this new treatment.
      We would like to thank the people at Smith and Nephew Corporation in charge of these hospitals for a detailed explanation of how to use Renasys and the prompt delivery of the machine.
      The authors would like to thank Enago (www.enago.jp) for the English language review.

      References

        • Rhomberg M.
        • Frischhut B.
        • Ninkovic M.
        • Schwabegger A.H.
        • Ninkovic M.
        A single-stage operation in the treatment of chronic osteomyelitis of the lower extremity including reconstruction with free vascularized iliac bone graft and free-tissue transfer.
        Plast Reconstr Surg. 2003 Jun; 111 (discussion 2362-3): 2353-2361
        • Dendrinos G.K.
        • Kontos S.
        • Lyritsis E.
        Use of the Ilizarov technique for treatment of non-union of the tibia associated with infection.
        J Bone Joint Surg Am. 1995 Jun; 77: 835-846
        • Li S.
        • Rong H.
        • Zhang X.
        • Zhang Z.
        • Wang C.
        • Tan R.
        • Wang Y.
        • Zheng T.
        • Zhu T.
        Meta-analysis of topical vancomycin powder for microbial profile in spinal surgical site infections.
        Eur Spine J. 2019 Dec; 28: 2972-2980
        • Lawing C.R.
        • Lin F.C.
        • Dahners L.E.
        Local injection of aminoglycosides for prophylaxis against infection in open fractures.
        J Bone Joint Surg Am. 2015 Nov 18; 97: 1844-1851
        • Wininger D.A.
        • Fass R.J.
        Antibiotic-impregnated cement and beads for orthopedic infections.
        Antimicrob Agents Chemother. 1996 Dec; 40: 2675-2679
        • Anagnostakos K.
        • Wilmes P.
        • Schmitt E.
        • Kelm J.
        Elution of gentamicin and vancomycin from polymethylmethacrylate beads and hip spacers in vivo.
        Acta Orthop. 2009 Apr; 80: 193-197
        • Klemm K.
        The use of antibiotic-containing bead chains in the treatment of chronic bone infections.
        Clin Microbiol Infect. 2001 Jan; 7: 28-31
        • Neut D.
        • Van de Belt H.
        • Stokroos I.
        • van Horn J.R.
        • Van der Mei H.C.
        • Busscher H.J.
        Biomaterial-associated infection of gentamicin-loaded PMMA beads in orthopaedic revision surgery.
        J Antimicrob Chemother. 2001 Jun; 47: 885-891
        • Whiteside L.A.
        • Roy M.E.
        One-stage revision with catheter infusion of intraarticular antibiotics successfully treats infected THA.
        Clin Orthop Relat Res. 2017 Feb; 475: 419-429
        • Shon A.S.1
        • Bajwa R.P.
        • Russo T.A.
        Hypervirulent (hypermucoviscous) Klebsiella pneumoniae: a new and dangerous breed.
        Virulence. 2013 Feb 15; 4: 107-118
        • Prokesch B.C.
        • TeKippe M.
        • Kim J.
        • Raj P.
        • TeKippe E.M.
        • Greenberg D.E.
        Primary osteomyelitis caused by hypervirulent Klebsiella pneumoniae.
        Lancet Infect Dis. 2016 Sep; 16: e190-e195
        • Lee C.R.
        • Lee J.H.
        • Park K.S.
        • Jeon J.H.
        • Kim Y.B.
        • Cha C.J.
        • Jeong B.C.
        • Lee S.H.
        Antimicrobial resistance of hypervirulent Klebsiella pneumoniae: epidemiology, hypervirulence-associated determinants, and resistance mechanisms.
        Front Cell Infect Microbiol. 2017 Nov 21; 7: 483
        • Tacconelli E.
        • Magrini N.
        Global priority list of antibiotic-resistant bacteria to guide research, discovery, and development of new antibiotics.
        World Health Organization, 2017
        • Lin Z.
        • Vasudevan A.
        • Tambyah P.A.
        Use of erythrocyte sedimentation rate and C-reactive protein to predict osteomyelitis recurrence.
        J Orthop Surg. 2016 Apr; 24: 77-83
        • James A.
        • Larson T.
        Acute renal failure after high-dose antibiotic bone cement: case report and review of the literature.
        Ren Fail. 2015 Jul; 37: 1061-1066
        • Aeng E.S.
        • Shalansky K.F.
        • Lau T.T.
        • Zalunardo N.
        • Li G.
        • Bowie W.R.
        • Duncan C.P.
        Acute kidney injury with tobramycin-impregnated bone cement spacers in prosthetic joint infections.
        Ann Pharmacother. 2015 Nov; 49: 1207-1213
        • Anagnostakos K.
        • Hitzler P.
        • Pape D.
        • Kohn D.
        • Kelm J.
        Persistence of bacterial growth on antibiotic-loaded beads: is it actually a problem?.
        Acta Orthop. 2008 Apr; 79: 302-307
        • Drlica K.
        The mutant selection window and antimicrobial resistance.
        J Antimicrob Chemother. 2003 Jul; 52: 11-17
        • Lebeaux D.
        • Chauhan A.
        • Rendueles O.
        • Beloin C.
        From in vitro to in vivo models of bacterial biofilm-related infections.
        Pathogens. 2013 May 13; 2: 288-356
        • Ceri H.
        • Olson M.E.
        • Stremick C.
        • Read R.R.
        • Morck D.
        • Buret A.
        The Calgary Biofilm Device: new technology for rapid determination of antibiotic susceptibilities of bacterial biofilms.
        J Clin Microbiol. 1999 Jun; 37: 1771-1776
        • Singla S.
        • Harjai K.
        • Chhibber S.
        Susceptibility of different phases of biofilm of Klebsiella pneumoniae to three different antibiotics.
        J Antibiot (Tokyo). 2013 Feb; 66: 61-66
        • Høiby N.
        • Henneberg K.Å.
        • Wang H.
        • Stavnsbjerg C.
        • Bjarnsholt T.
        • Ciofu O.
        • Johansen U.R.
        • Sams T.
        Formation of Pseudomonas aeruginosa inhibition zone during tobramycin disk diffusion is due to transition from planktonic to biofilm mode of growth.
        Int J Antimicrob Agents. 2019 May; 53: 564-573
        • Castaneda P.
        • McLaren A.
        • Tavaziva G.
        • Overstreet D.
        Biofilm antimicrobial susceptibility increases with antimicrobial exposure time.
        Clin Orthop Relat Res. 2016 Jul; 474: 1659-1664
        • Winkler H.
        Treatment of chronic orthopaedic infection.
        EFORT Open Rev. 2017 May 11; 2: 110-116
        • Cheng N.C.
        • Yu Y.C.
        • Tai H.C.
        • Hsueh P.R.
        • Chang S.C.
        • Lai S.Y.
        • Yi W.C.
        • Fang C.T.
        Recent trend of necrotizing fasciitis in Taiwan: focus on monomicrobial Klebsiella pneumoniae necrotizing fasciitis.
        Clin Infect Dis. 2012 Oct; 55: 930-939
        • Russo T.A.
        • Marr C.M.
        Hypervirulent Klebsiella pneumoniae.
        Clin Microbiol Rev. 2019 May 15; 32: e00001-e19
        • Tan Y.M.
        • Chung A.Y.
        • Chow P.K.
        • Cheow P.C.
        • Wong W.K.
        • Ooi L.L.
        • Soo K.C.
        An appraisal of surgical and percutaneous drainage for pyogenic liver abscesses larger than 5 cm.
        Ann Surg. 2005 Mar; 241: 485-490
        • Catalán-Nájera J.C.
        • Garza-Ramos U.
        • Barrios-Camacho H.
        Hypervirulence and hypermucoviscosity: two different but complementary Klebsiella spp. phenotypes?.
        Virulence. 2017 Oct 3; 8: 1111-1123