The Successful Use of Negative-Pressure Wound Therapy in Two Cases of Canine Necrotizing Fasciitis
A 5 mo old female Akita and a 1 yr, 5 mo old male German shorthaired pointer were both evaluated for soft-tissue lesions characterized by rapidly expanding edema, erythema, and pain. Ultrasound was utilized to locate and sample fluid accumulations, and β-hemolytic Streptococcus was isolated from the wounds. Development of systemic symptoms including fever, tachycardia, and tachypnea as well as a lack of response to medical management prompted surgical intervention in both cases. During surgical exploration and debridement, disruption of intermuscular tissue planes was appreciated and necrotizing fasciitis (NF) was suspected. Negative-pressure wound therapy systems utilizing 120 mm Hg of continual negative pressure were applied to wounds for 5 and 4 days for the Akita and German shorthaired pointer, respectively. Resolution of infection was achieved and although the lesions were associated with limbs, amputation was avoided. In both cases, the results of histopathology were consistent with NF. NF is recognized as a rapidly progressive infection associated with high rates of morbidity and mortality. Timely use of negative-pressure wound therapy appears to be a viable management tool to accompany surgical debridement, appropriate antibiotics, and analgesics.
Introduction
Necrotizing fasciitis (NF) is a rare, potentially life-threatening bacterial infection arising in the fascial and subcutaneous tissues.1–5 NF can be associated with trauma, compromised immune states or the origins may be idiopathic.3 β-Hemolytic Streptococcus spp. are the most common cause of NF, representing up to 71% of human cases.1 As with other less virulent soft-tissue infections, swelling and edema is expected: however, NF lesions are more painful, progress more rapidly, and are typically accompanied by systemic symptoms, including fever and tachycardia.1,2 The infection expands quickly along fascial planes between muscle, and histopathology often reveals necrosis of subcutaneous fat, vascular structures, nerves, and deep fascia together with neutrophilic infiltration and thrombosis of adjacent vasculature.1,2 To confirm a diagnosis of NF, surgical identification of eroded fascial planes, isolation of a causative organism, and histopathological evidence is required.1 However, wounds under suspicion of NF should be promptly and aggressively treated.1 To await the results of laboratory testing before performing surgical exploration and debridement would be ill advised because mortality from multiple-organ dysfunction or sepsis has been reported in >70% of human cases where early aggressive management was not instituted.2 Multiple surgical debridements are generally recommended, typically at intervals of 24–48 hr.1
Although case reports describe the successful use of negative-pressure wound therapy (NPWT) in human cases of necrotizing fasciitis, to the authors' knowledge, this is the first report to describe its use in canine NF.5–9
Case Report
Case 1
A 5 mo old female Akita weighing 24.5 kg was presented to her primary care veterinarian with a 2 day history of right hind-limb lameness (day 0). Radiographs demonstrated marked soft-tissue swelling of the right pelvic limb and moderately increased soft-tissue opacity within the stifle. The patient was administered an injection of carprofen and oral tramadol was dispensed (dosages and frequency unknown) while the client awaited a referral appointment. On presentation to New York Veterinary Specialty Center (day 2), the client reported that the swelling of the right hind limb had increased. Examination revealed a nonweight-bearing lameness of the hind limb, marked pitting edema of the right thigh, and enlargement of the right popliteal lymph node. Although palpation of the right thigh elicited signs of discomfort, the stifle joint was stable and nonpainful. The patient was normothermic, no external wounds were visible, and the remainder of the physical examination was unremarkable. Complete blood cell count and serum biochemical profile revealed decreased albumin (24 g/L; reference range, 25–44 g/L) and elevated phosphorus (2.42 mmol/L; reference range, 0.94–2.13 mmol/L). A regional ultrasound of the right hind limb identified a pocket of echogenic fluid cranial to the proximal aspect of the right femur between the vastus medialis and rectus femoris muscles. Adjacent intermuscular soft tissues were diffusely hyperechoic, and mild to moderate regional lymph node enlargement was appreciated. Microscopic evaluation of a sample from the fluid pocket demonstrated numerous degenerate neutrophils and both intra- and extracellular bacterial cocci. A fluid sample was submitted for aerobic culture and sensitivity. The client was advised to soak the right hind limb in a warm magnesium sulfate solution 2–3 times daily. Tramadola was continued (4 mg/kg per os [PO] q 8 hr) and empirical antibiotic therapy with enrofloxacinb (5 mg/kg PO q 24 hr) and metronidazolec (10 mg/kg PO q 12 hr) was initiated pending the aerobic culture results.
The client re-presented the patient within 24 hr of discharge (day 3). On examination, the patient was dehydrated, tachycardic, tachypneic, unable to rise, and demonstrated a depressed mentation. Increased swelling of the right hind limb was noted relative to the previous examination, and the dog also demonstrated an increased pain response on palpation of the thigh. The dog was admitted and IV fluid therapy administered. The spectrum of antibiotics was broadened to include ampicillin with sulbactamd (20 mg/kg IV q 8 hr). Oral enrofloxacin and metronidazole were replaced with their IV counterpartse, and morphinef (0.5 mg/kg IV q 6 hr) was initiated.
Following 12 hr of fluid therapy, the wound was surgically explored (day 3) under general anesthesia. Preoperative ultrasonography was employed to locate the greatest accumulation of fluid, and a medial incision was made to approach the right thigh. The sartorius muscle was displaced cranially exposing a large pocket of purulent material between the rectus femoris and vastus medialis muscles. The area was drained, debrided, and flushed thoroughly with sterile saline. Two penrose drainsg, each measuring 2.5 cm, were placed between the muscle bellies, exiting distal to the original incision. The incision was closed routinely and a modified Robert Jones bandage applied. The dog was continued on IV fluids, antibiotics, and analgesics. Following 36 hr of in-patient supportive care (day 4), the patient's mentation had improved and the dog demonstrated a normal appetite. Respiration and heart rate had returned to normal and the dog was discharged with amoxicillin trihydrate/clavulanate potassiumh (13.75 mg/kg PO q 12 hr) and tramadol (4 mg/kg PO q 8 hr) in addition to the previously dispensed enrofloxacin and metronidazole.
Two days later (day 6), the client reported that the dog's condition had deteriorated again. The dog stopped eating and drinking and demonstrated aggression when touched. Physical examination revealed a continued inability to rise, pyrexia (39.6°C), tachycardia, and tachypnea. The right hind limb remained substantially edematous, moderate erythema of the thigh had developed, and palpation of the thigh revealed evidence of extreme discomfort. The right stifle demonstrated crepitus, reduced flexion and extension angles, and mediolateral instability. Culture and sensitivity testing of the previously submitted fluid yielded heavy growth of β-hemolytic streptococci susceptible to both enrofloxacin and amoxicillin trihydrate/clavulanate potassium.
Based on the isolation of β-hemolytic streptococci and the clinical progression, NF was suspected, with potential involvement of the right stifle. Prior to repeat surgical exploration, a complete blood cell count and serum biochemical profile revealed leukocytosis (16.5 × 109/L; reference range, 4–15.5 × 109/L), neutrophilia (13.2 × 109/L; reference range, 2.06–10.6 × 109/L), monocytosis (1.81 × 109/L; reference range, 0–0.84 × 109/L), hypoproteinemia (48 g/L; reference range, 50–74 g/L), and hypoalbuminemia (25 g/L; reference range, 27–44 g/L). The penrose drains were removed under general anesthesia, and a lateral approach was made to the right hind limb from the greater trochanter to the distal thigh (Figure 1A). The fascial plane between the vastus lateralis and biceps femoris muscle was abnormally thickened and easily disrupted on blunt dissection. A direct communication between the lateral and medial aspects of the thigh was appreciated cranial to the femur. Aggressive debridement was performed via curettage, and an incisional biopsy of fascia and adjacent muscle was collected and submitted for histopathology. The region was lavaged with 1 L of sterile saline, and Manuka honeyi (25 g) was applied to the wound bed. The incision was then loosely packed with several pieces of gas sterilized polyester foamj. A fenestrated 12-gauge red rubber catheterk was inserted into a block of foam within the wound bed. Stoma pastel was applied in a ring around the wound, the incision was left open, and an impervious adhesive drapem was applied to the region creating an airtight seal. Ingress and egress portals were established in the right stifle, and the joint was lavaged with 550 mL of sterile saline. A suction pumpn with an in-line collection canister was utilized to apply 120 mm Hg of continual negative pressure to a red rubber catheter, completing the NPWT system (Figure 1B).
Citation: Journal of the American Animal Hospital Association 51, 1; 10.5326/JAAHA-MS-6033
The dog remained hospitalized with the NPWT system in place. Enrofloxacin was discontinued, and supportive care included IV fluid support and continuation of morphine, ampicillin with sulbactam, metronidazole, as well as the initiation of fentanylo (50 µg/hr transdermally).
Although the patient remained unable to rise, since the placement of the NPWT system, substantial improvement in mentation was noted. The heart rate, respiratory rate, and temperature returned to normal. Following 48 hr of continual suction, the open cell foam was replaced under general anesthesia (day 8). Exuberant fresh granulation tissue had developed within the wound bed. Additional physical debridement at that time did not appear necessary. An additional 25 g of Manuka honey was instilled into the wound, and the NPWT system was reapplied.
By day 11, the patient was voluntarily eating and drinking and, with sling assistance, was making attempts to ambulate. The patient was sedated, and the NPWT apparatus was removed from the right lateral thigh. The entire wound bed consisted of apparently healthy granulation tissue with no evidence of further necrosis or exudate (Figure 1C). Two 2.5 cm Penrose drains were temporarily placed into the resulting defect, and the skin edges were debrided and reapposed. The dog was discharged with oral amoxicillin trihydrate/clavulanate potassium, metronidazole, and tramadol (dosages as previously described).
By day 11, the intraoperative tissue sample had been evaluated microscopically and the findings were consistent with NF. The muscle, fibrous, and adipose tissues were diffusely infiltrated with neutrophils (Figure 2). Lesser numbers of lymphoid cells and macrophages were also identified. Considerable necrosis and fibrin exudation was present throughout the sample.
Citation: Journal of the American Animal Hospital Association 51, 1; 10.5326/JAAHA-MS-6033
By day 22, the incision was completely healed. The swelling of the right hind limb had resolved, and palpation of the thigh at that time elicited evidence of only mild discomfort. Antibiotic therapy was continued for a total of 2 mo. Supportive joint therapy included two intra-articular hyaluronic acidp injections (days 36 and 112), two right stifle shockwave sessionsq (1000 E6 shocks, 20 mm trode; days 93 and 107), and a physical therapy regiment utilizing an underwater treadmill (days 100, 112, 119, 147, and 161). By day 161, substantial improvement in mediolateral joint stability was appreciated, flexion and extension angles of the right stifle were seen to be 50 and 145°, respectively, and the patient bore moderate weight on the right hind limb during ambulation. At the 1 yr follow-up, the client reported the dog was using the leg well without the need of oral analgesics, and examination revealed only low-grade residual lameness associated with the right hind limb.
Case 2
A 1 yr, 5 mo old male German shorthaired pointer weighing 23.5 kg was presented for a right forelimb lameness. A distal humeral fracture had been repaired several months previously utilizing intramedullary pinning and cerclage wires. Physical examination revealed a nonweight-bearing lameness of the right forelimb with palpable instability of the distal humerus. Radiographs taken under general anesthesia revealed a malunion of the distal humerus and the elbow was found to be ankylosed at 130° of extension relative to the weight-bearing angle of 160° in the contralateral limb. Surgical revision involved removal of the previously placed implants, use of an autogenous corticocancellous graft, and application of a lateral 3.5 mm locking plater and a type 1a transarticular lateral external fixator.s One month following the revision, the patient was bearing moderate weight on the right forelimb; however, 6 wk following the revision surgery, the client re-presented the dog due to acute onset of pain and swelling associated with the cranioventral chest and right axillary region (day 0). The client reported that the dog was anorexic and was no longer using the right forelimb. On physical examination, the dog was pyrexic (40°C), dehydrated, and demonstrated a depressed mentation. Pitting edema and erythema was identified extending from the right axillary region to the ventrolateral thorax, and the dog was nonweight bearing on the right forelimb. Palpation of the swelling elicited signs of substantial discomfort. A complete blood cell count revealed a granulocytosis (25.5 × 109/L; reference range, 3.5–12 109/L), monocytosis (2.3 109/L; reference range, 0.3–1.5 109/L), and anemia (hematocrit was 0.29; reference range, 0.37–0.55). Serum biochemical profile was unremarkable. Ultrasonography demonstrated an area of hypoechogenicity in the axillary region surrounded by hyperechoic tissues. A fluid aspirate from the area was submitted for cytology and aerobic culture and sensitivity. The dog was admitted, and supportive care included IV fluid therapy, ampicillin and sublactam (30 mg/kg IV q 8 hr), tramadol (4 mg/kg PO q 8 hr), and warm compression of the axillary region q 6 hr.
After 24 hr of in-patient supportive care, the dog had tachycardia, tachypnea, and an inability to rise from lateral recumbency. The edema and erythema expanded to include the entire ventral thoracic area, the abdominal region, and the right forelimb. Substantial pain was evident on palpation of the edematous areas. Microscopic evaluation of the fluid sample collected on day 1 revealed subacute neutrophilic and septic inflammation. Based on the rapid progression of clinical signs, necrotizing fasciitis was suspected.
While under general anesthesia, the lateral external fixator was removed and a skin incision was made caudal to the right front limb. Substantial purulent discharge was appreciated in the axillary area and the fascial planes between the serratus dorsalis, serratus ventralis, external abdominal oblique, and scalenus muscles were thickened and disrupted. Regional soft tissues and muscle were debrided, and the area flushed thoroughly with sterile saline. A soft-tissue sample was collected and submitted for histopathology. Manuka honey (50 g) was applied to region, and a NPWT system as described in case 1 was placed with 120 mm Hg of continual negative pressure applied to the wound bed. In-hospital supportive care included clindamycint (11 mg/kg IV q 12 hr), famotidineu (0.5 mg/kg IV q 12 hr), meropenemv (12 mg/kg IV q 8 hr), morphine (0.5 mg/kg q 8 hr intramuscularly), fresh-frozen plasma (1 unit IV), fentanyl (75 µg/hr transdermally), and supportive fluid therapy. By day 2, swelling had progressed to include the inguinal region, and a low albumin (17 g/L; reference range, 23–44 g/L) promoted initiation of a hetastachw constant rate infusion (20 mg/kg/day IV).
By day 3, the fever, tachycardia, and tachypnea had resolved and the patient regained the ability to ambulate with assistance. The patient was sedated and the NPWT was system removed. Healthy granulation tissue was observed lining the wound bed, and no additional debridement appeared necessary. Manuka honey (25 g) was instilled into the region prior to reapplication of the NPWT system. The aerobic culture yielded β-hemolytic sterptococcal growth that was susceptible to clindamycin and amoxicillin trihydrate/clavulanate potassium. The histopathology results were consistent with NF in that the tissue architecture was disrupted secondary to infiltration with neutrophils and lesser numbers of lymphoid cells and macrophages.
By day 5, the patient's mentation had substantially improved and regional edema had resolved. The patient was once again sedated, and the NPWT system was removed. The wound bed consisted of healthy granulation tissue. A Jackson-Pratt suction drainx was placed, and the cranial half of the wound was closed. Excessive skin tension precluded direct closure of the caudal portion of the wound, which was left open to heal by second intention. The dog was discharged on day 9 with a clindamycin (11 mg/kg PO q 12 hr) and amoxicillin tridhydrate/clavulanate potassium (13.75 mg/kg PO q 12 hr) for a total of 6 weeks of antibiotic therapy. The Jackson-Pratt drain was removed on day 11, and the wound healed uneventfully. Examination on day 40 revealed fair weight bearing on the right forelimb. Radiographs demonstrated healing of the distal humeral fracture and although ankylosis of the joint persisted, the revision surgery had resulted in an improved weight-bearing angle.
Discussion
NPWT positively alters the microenvironment of a wound, providing several justifications for its use in NF cases.9 In vitro studies demonstrated negative pressure stimulates molecule synthesis and cell proliferation through changes in gene expression.10 Morykwas et al. (1997) demonstrated that NPWT increased the rate of granulation tissue formation in vivo relative to an untreated control group.10 Although negative pressure can decrease blood flow in directly adjacent tissues, blood flow in surrounding tissues several centimeters from a wound margin reached levels 5 times that of basal blood flow.11,12 Increased blood flow is likely due to the direct effects of the pressure gradient, stimulation of neovascularization, and reduction in edema, which would otherwise compress and constrict microvasculature.10 Improving local perfusion will increase O2 concentration, improve access of systemically administered antibiotics, and enhance the ability of humoral and cell-mediated defense mechanisms to penetrate the wound bed. The constant flux of exudate and wound debris away from the site of infection will also directly decrease bacterial load. Morykwas et al. (1997) demonstrated that bacterial load in porcine wounds decreased from 108 to 105 organisms/g of tissue within 5 days with NPWT relative to 11 days in control animals.10
Medicinal grade Manuka honey was applied directly to the wound bed in both cases described herein. When used topically, Manuka honey acts as a local antimicrobial due to its low pH, hyperosmolar characteristics, ability to form hydrogen peroxide, and its high concentration of the antibacterial compound methylglyoxal.13,14 Maddocks et al. (2012) demonstrated topically applied Manuka honey reduced bacterial load within Streptococcus pyogenes biofilms.15
Ultrasonography was used in both cases described herein to locate and sample the lesion. Unlike radiography, ultrasonography provides a means of evaluating intermuscular soft tissues. Ultrasonographic evidence of fasciitis in human cases includes fascia >4 mm in width, distortion of the fascia, and fluid accumulation within the fascial planes.16 Identification of hyperechoic intermuscular tissues surrounding the fluid pocket was appreciated in case 1. Those findings may have represented early evidence of NF.
Oftentimes, management of NF precedes the results of culture and sensitivity testing, and antimicrobial therapy such as penicillins or clindamycin are instituted based on the presumption of streptococcal involvement. Although enrofloxacin was briefly utilized in case 1, fluoroquinolones have been implicated in inducing activity of bacteriophages that enhance the virulence of Streptococcus canis regardless of in vitro susceptibility.1 Enrofloxacin has also been shown to induce microscopic cartilage abnormalities when administered for 30 days to dogs during their growth phase. Therefore, fluoroquinolones should be avoided either when NF is suspected or when alternative antibiotic choices exist in juveniles.
Conclusion
The results of these two cases encourage the use of NPWT as a component of therapy when treating canine NF. The clinical status of both patients was declining prior to application of the NPWT devices. Following application, both infections resolved rapidly and further surgical intervention was not necessary. When utilized in conjunction with traditional local and systemic wound management techniques, it is the authors' opinion that this modality has the potential to decrease morbidity and mortality rates in cases of canine NF. Additional controlled studies are necessary to further assess this treatment option in cases of NF.
A, B: Intraoperative appearance of a necrotizing fasciitis wound of a 5 mo old female Akita prior to surgical debribement (A), following application of a negative-pressure wound therapy (NPWT) system (B), and after 5 days of NPWT. C: Notice the cobblestone appearance of the healthy granulation tissue (asterix), which develops as the tissue grows between pores of the open cell foam within the NPWT system. Stoma paste remains adhered to the surrounding skin (black arrows).
Photomicrographs of sections of tissue from the right thigh of the same dog in Figure 1 demonstrating necrosis, fibrin, and suppurative inflammation. Hematoxylin and eosin staining, original magnification ×200; bar = 50 μm.
Contributor Notes
