Successful Treatment of Necrotizing Fasciitis in the Hind Limb of a Great Dane

Introduction

Necrotizing fasciitis (NF) is a rapidly progressive and potentially life-threatening bacterial infection of the subcutaneous (SC) tissues and fascial planes, characterized by massive soft-tissue necrosis exempting the muscle and skin.^1–3 The bacteria responsible for the majority of veterinary cases are beta-hemolytic, Lancefield group G Streptococcus sp. with genetic characteristics typical of Streptococcus canis.^2,4–9 Canine cases with NF can be presented with clinical signs typical of soft-tissue bacterial infection, including fever, depression, and lethargy, as well as localized soft-tissue swelling, heat, edema, and pain of the affected area.^1–3 Advanced imaging such as ultrasound, fistulogram, and magnetic resonance imaging or computed tomography can be used to identify the extensive exudate progressing along fascial planes that is characteristically present in NF cases but not readily apparent on physical examination.^2,10 Beta-hemolytic streptococcal infection is the cause of both NF and streptococcal toxic shock syndrome (STSS); therefore, early recognition and treatment of infection are essential.^2–8

Once a diagnosis of NF is suspected, immediate and complete surgical debridement is indicated to stop progression of the disease and prevent STSS. Additionally, local tissue ischemia and poor perfusion make attempts at medical management alone ineffective.² Definitive diagnosis of NF is achieved by fluid culture and histopathology.² Postoperatively, cases are supported aggressively with intravenous (IV) fluids, broad-spectrum antibiotics, and specific treatments for concurrent vasculitis, cellulitis, and biochemical abnormalities that may be present because of severe systemic inflammation and possible organ failure (e.g., hypoproteinemia, electrolyte abnormalities, hypoglycemia). Persistence or progression of infection may require further surgical intervention. Amputation may be necessary in severe or recurrent cases.¹

The following case report describes the diagnosis and treatment of NF in the hind limb of a Great Dane puppy. Reports of NF in the veterinary literature are few, and details of postoperative and supportive treatments contributing to improved outcome are uncommon.^1,2,4–8 This is the first report to describe a combination of broad-spectrum antibiotic therapy, local amikacin infusions, pentoxifylline administration, and physical rehabilitation in the successful restoration of near-normal limb function following initial surgical intervention.

Case Report

A 5-month-old, 34.6-kg, intact female Great Dane was evaluated by the referring veterinarian for signs consistent with food bloat and acute onset of nonweight-bearing, right hind-limb lameness. No history of trauma was present. Abdominal radiographs showed a food-distended stomach. Radiographs and orthopedic examination of the pelvis and right hind limb were unremarkable. No abnormalities were present on complete blood count (CBC) and serum biochemical profile.

The puppy was treated with gastric lavage and transferred to an emergency hospital for continued care. The emergency clinician repeated the orthopedic examination. The right stifle had mild effusion, regional soft-tissue swelling, and pain on extension; however, no cranial drawer (in flexion or extension) or cranial tibial thrust could be elicited. The right popliteal lymph node was subjectively enlarged (approximately 2 cm wide × 2.5 cm long × 1.5 cm deep) and firm on palpation compared to the contralateral limb. The dog was given hydromorphone^a (0.1 mg/kg IV) for analgesia. A right lateral abdominal radiograph, obtained to evaluate gastric dilatation or volvulus, was unremarkable. Right stifle radiographs showed joint effusion and mild irregularity of the distal femoral condyle. The puppy was discharged from the emergency clinic that evening. Discharge instructions included activity restriction and administration of tramadol^b (4 mg/kg orally [PO] q 8 to 24 hours) for suspected right partial cranial cruciate ligament rupture.

Three days later, the puppy was presented again to the emergency clinic for rapid progression of right hind-limb lameness, severe lethargy, and anorexia. The puppy had not responded to treatment with prescribed tramadol or the carprofen^c (2.2 mg/kg PO q 12 hours) prescribed by another hospital during the 3-day period between being discharged and returning to the emergency clinic. Physical examination revealed mental depression, mild tachycardia (128 beats per minute), mild tachypnea (respiratory rate 40 breaths per minute) with normal respiratory effort, hyperthermia (39.5°C), and dry mucous membranes. The puppy was thin (body condition score 3/9) and had lost 5 pounds since her first presentation 3 days earlier. The right hind limb was nonweight-bearing and was dragged during ambulation. Severe, circumferential, soft-tissue swelling and pitting edema extended from the level of the right coxofemoral joint to the right hind digits. The limb was diffusely hot to the touch, and the proximal limb was markedly painful on palpation. The stifle could not be flexed, and the other joints of this limb had severely decreased range of motion. Neurological deficits in the right hind limb included loss of conscious proprioception and markedly diminished withdrawal and hyporeflexive hind-limb spinal reflexes. Peripheral pain sensation was intact. Digital rectal examination was unremarkable.

Radiographic evaluation of the right hind limb revealed increased stifle effusion and diffuse, soft-tissue swelling of the limb as compared to the previous examination 3 days earlier. The bones of the right hind limb were within normal limits. The in-hospital serum biochemical profile was unremarkable, while the CBC demonstrated a monocytosis (4644/μL, reference range 150 to 1350/μL) with slightly toxic-appearing neutrophils on slide evaluation; otherwise, the CBC was within normal limits. Urinalysis showed a urine specific gravity of 1.016 and an inactive sediment. A sample was submitted for culture. Fine-needle aspiration of SC soft-tissue swelling in the popliteal region yielded white-yellow, opaque fluid. Fluid cytology (performed in-house) was consistent with a septic, purulent exudate containing numerous, highly degenerate neutrophils and many cocci in chains. A sample was submitted for aerobic and anaerobic culture. Therapy was initiated with ampicillin trihydrate^d (22 mg/kg SC), buprenorphine^e (0.01 mg/kg intra-muscularly [IM]), and isotonic crystalloids (lactated Ringer’s solution [LRS],^f 90 to 120 mL/kg IV q 24 hours).

Ultrasound evaluation of the right hind limb identified SC and intermuscular pockets of echogenic fluid in the lateral and caudomedial femoral region, extending distal to the hock. Doppler-flow evaluation demonstrated good arterial blood flow throughout the right hind limb. After obtaining initial diagnostic results, immediate surgical exploration and debridement of the limb was recommended to the owners because of the presumptive diagnosis of severe soft-tissue infection.

Within 6 hours of the puppy’s second presentation to the hospital, the right hind limb was explored via a lateral approach to the femur. Purulent, murky fluid exuded when the fascia lata was incised. The superficial and deep muscular fasciae appeared necrotic and were easily dissected from surrounding soft-tissue structures using blunt technique. During surgical exploration, fluid tracts extended along intermuscular fascial planes cranial and medial to the femur, as well as along superficial fascial planes distal and lateral to the hock and the stifle joint. A second incision was made on the lateral aspect of the distal tibia to explore the distal extent of the fistulous tract and continue debridement. No foreign material was identified. Grossly necrotic soft tissues were debrided, and the surgery site was copiously lavaged with sterile saline. Two 10-French, closed-suction drains with 100-mL reservoirs^g were placed to achieve drainage of the entire affected area. Drains were secured with 2–0 nylon^h in purse-string and friction suture patterns. The remaining SC tissues and skin were closed without defect.

A bandage was not placed on the limb to allow for visual monitoring, cold therapy, and passive range-of-motion exercises. A cold compress was applied to the affected leg for 10 minutes q 4 hours. After the compress was removed, the leg was massaged using mild to moderate pressure, and the joints were slowly cycled through gentle flexion and extension 10 to 15 times.¹¹ Massage and range-of-motion exercises of the affected limb were well tolerated in the immediate postoperative period.

Postoperative therapy included a constant-rate infusion of fentanylⁱ (4 to 5 μg/kg per hour IV), ampicillin^j (22 mg/kg IV q 8 hours), amikacin^k (15 mg/kg IV q 24 hours), and maintenance crystalloids supplemented with 20 mEq/L potassium chloride^l (120 mL/kg IV q 24 hours). Drain fluid production was quantified q 4 hours, and fluid cytology was performed daily.

The puppy developed a moist, frequent cough with harsh lung sounds 15 hours postoperatively. Diagnostic evaluation was not performed because of significant financial constraints. Nebulization, coupage, and pentoxifylline^m (10 mg/kg PO q 8 hours) were initiated for treatment of suspected streptococcal bronchopneumonia and because of concerns for developing STSS. The frequency of coughing markedly decreased in the following 24 hours. By 48 hours postoperatively, the cough had resolved, lung sounds were normal in all fields, and the puppy was eupneic.

No bacteria were cultured from the cystocentesis urine sample. A beta-hemolytic Streptococcus sp. was identified on initial fluid culture results 2 days after surgery. Clinical signs, surgical findings, and culture results were consistent with the suspected diagnosis of NF. To confirm this diagnosis, an incisional biopsy from the fascia overlaying the right vastus lateralis muscle was obtained via a 1.5-cm, lateral femoral incision under medetomidineⁿ anesthesia (420 μg/m² IV). Histopathology showed foci of soft-tissue necrosis along the fascial layer with mild hemorrhage and inflamed granulation tissue. No infectious organisms were identified on examined tissue sections. These findings were consistent with a diagnosis of NF. Clindamycin^o (13 mg/kg PO q 12 hours) was initiated based on current treatment recommendations for canine NF.² Therapy with passive range-of-motion exercises, massage, and limb compresses was continued as previously described, except that warm compresses were used instead of cold to improve regional blood flow and promote tissue healing.

On day 3 postoperatively, the right hind limb remained severely swollen and nonweight-bearing, but the puppy’s attitude and appetite were markedly improved. Discomfort was no longer shown during physical therapy. In addition, the hock and stifle had slight improvement in range of motion. Aspiration of a SC swelling noted in the right hock region yielded fluid that, on cytology, contained numerous chains of cocci. The pocket was drained by needle aspiration and was then infused with amikacin^k (250 mg diluted 1:1 with saline). This treatment was repeated q 24 hours for 2 more days in addition to treatment with systemic antibiotics. The puppy was maintained on IV crystalloids (LRS,^f 60 to 90 mL/kg q 24 hours) for the duration of amikacin therapy. Due to the owner’s financial limitations, daily urinalysis to monitor for nephrotoxicity was not performed.

Serial girth measurements of the right hind limb showed diminished swelling of the thigh and stifle on day 4 postoperatively. Drain fluid production decreased to 0.13 mL/kg per hour (a 61% reduction from postoperative fluid production on day 1). Ultrasound examination of the limb was performed, and none of the extensive SC or intermuscular fluid pockets identified in the initial diagnostic ultrasound was present. Only the small fluid pocket in the right hock region remained. Fluid cytology from the right hock region showed a marked decrease in the number of cocci present.

On day 5 postoperatively, the puppy was bright, eating well, and minimally weight-bearing on the affected limb. The fentanyl constant-rate infusion had been tapered on days 2 and 3 postoperatively, then was discontinued on the third postoperative day. Pain was well controlled with tramadol^b (3 mg/kg PO q 12 hours). Bacteria were absent from fluid obtained from the right hock region, and only minor swelling was palpable. While fluid production from the limb had decreased to 0.1 mL/kg per hour (a 70% decrease from day 1 production), drain fluid cytology revealed numerous rods and chains of cocci. Fluid samples from the drain were submitted for anaerobic and aerobic bacterial culture. While awaiting culture results, amikacin^k (250 mg diluted 1:1 with saline) was infused retrograde into each drain q 24 hours for 2 days, and IV amikacin^k was discontinued. Since bacterial growth was present despite treatment with antibiotics recommended for NF, amoxicillin^p (22 mg/kg PO q 12 hours) was discontinued, and cefpodoxime^q (8.5 mg/kg PO q 24 hours) was started. Clindamycin^o was continued at 13 mg/kg PO q 12 hours.

Drain fluid culture results available on day 7 postoperatively identified Serratia marcescens, which was susceptible to amikacin and first-generation cephalosporins. Drains were removed on day 8 postoperatively, when cytological evaluation confirmed moderate polymorphonuclear inflammation and absence of bacteria. Cultures were negative for anaerobes, fungi, and mycoplasma.

The puppy was discharged 9 days after surgery, with owner instructions to continue tramadol^b (3 mg/kg PO q 12 hours), clindamycin^o (13 mg/kg PO q 12 hours), and cefpodoxime^q (8.5 mg/kg PO q 24 hours). Pentoxifylline^m was discontinued when fluid cytology (performed in-house) was negative for bacteria. Pentoxifylline^m had been administered for 7 days. At the time of discharge, the puppy was playful, moderately weight-bearing on the affected limb, and placing the limb normally during ambulation.

An examination was performed 40 days after discharge. The puppy’s right hind limb was normal in size, had normal flexion and extension of all joints, and was near normal in function. A persistent, mild, intermittent lameness was present (i.e., the puppy was not 100% weight-bearing 100% of the time). The puppy had gained weight (current weight was 43.7 kg), and the owner reported she was doing well at home.

Discussion

The few reports of canine NF in the veterinary literature focus primarily on the pathophysiology, clinical signs, and initial treatment of this disease. The majority of dogs reported with NF are presented with a limb (less commonly the neck or thorax) affected by severe, rapidly progressing cellulitis, fever, depression, and a variable history of mild trauma.^1–7 Increasing awareness of this disease has been the first step in facilitating early recognition and expeditious treatment, since physical examination findings are not pathognomonic.^1,2 In cases of suspected NF, the diagnosis is supported by easy finger fascial dissection and the presence of copious, malodorous exudate at the time of surgical debridement and lavage. Fluid culture and histopathology of tissue from the leading edge of the region affected are required for definitive diagnosis.¹

In the present case, history and physical examination findings were similar to those of previously reported NF cases with no known history of trauma. Early recognition of suspected NF was facilitated by ultrasound evaluation of the affected limb, which revealed intermuscular fluid pocketing, prompting aspiration and cytology. Reported cases are still too small in number to identify significant risk factors; however, this case makes Great Danes the most commonly affected breed (3/13).^5,6 Other breeds affected by NF include Irish wolfhound, border collie, Doberman pincher, cocker spaniel, and German shorthaired pointer (n=2 of each breed).

Once NF was suspected because of the progression of clinical signs and results of diagnostic testing, the affected limb was immediately treated with surgical debridement, lavage, and closed-suction drains. In humans, reported survival rates are 70% if surgery is performed within 24 hours of diagnosis of NF infection and only 35% if surgery is ≥24 hours after diagnosis.¹ The closed-suction drains allowed for daily fluid cytology. Previously reported canine cases of NF have been treated with open drainage or amputation because of the extent or location of the necrotic tissue.^1,2,5,9 The location of NF in this case allowed primary closure and the use of closed-suction drains.

Whether the administration of carprofen contributed to disease progression is unknown. Nonsteroidal antiinflammatory drugs (NSAIDs) have been purported to inhibit neutrophil activity and thus enhance the rapid progression of NF.¹² One study evaluating the virulence of the Streptococcus sp., isolated from the first reported cases of canine NF, found that worsening infection could not be attributed to increased bacterial virulence.⁶ These findings give possible support to the theory that NSAID administration may exacerbate streptococcal infection in dogs with NF.¹³ Other concerns with NSAIDs are their potential to mask initial presenting signs, causing a delay of early diagnosis and treatment and increasing the risk of gastrointestinal ulceration.² In this case, the puppy received an NSAID for 3 days, which may have delayed diagnosis and masked the intense pain commonly reported in cases of NF.^2–7

Recommended broad-spectrum antibiotic therapy for treatment of NF in dogs is a combination of a penicillin, an aminoglycoside, and clindamycin.² The use of fluoroquinolones as part of broad-spectrum antibiotic therapy is controversial,¹ as enrofloxacin^r use may have contributed to worsening NF in previous reports.⁴ Specifically, one study indicates that fluoroquinolones may induce activity of bacteriophages that enhance the virulence of Streptococcus canis.^2,14 Once the recommended antibiotic therapy is initiated, adjustments can be made based on results of bacterial culture and sensitivity. This is important, as secondary pathogens were identified in this case (Serratia marcescens), and secondary pathogens have been identified in other NF cases.^1,4 Additionally, in veterinary medicine, Streptococcus sp. is not the only pathogen associated with NF.⁹ Escherichia coli and beta-hemolytic streptococci have been associated with NF in veterinary medicine.

Of the cases reported in the veterinary literature, the dog of this report was the first with NF to receive local antibiotic infusions in addition to systemic antibiotic therapy. This may be important in the treatment of NF, because continued bacterial production of exotoxin and proteinases causes poor tissue perfusion and tissue necrosis.² The end result is that systemic antibiotics may fail to reach effective concentrations in affected tissues. However, the amount of amikacin that would be absorbed systemically from local administration in dogs with NF is unknown. This is a potential cause for concern with concurrent systemic amikacin administration. Optimally, antibiotics such as amikacin with a short half-life and risk of toxicity should achieve high local concentrations while minimizing systemic absorption.¹⁵ Several methods to achieve these goals are under investigation in human medicine, including local infusion and implantable antibiotics.^16–18 Although antibiotic-impregnated materials (such as polymethylmethacrylate beads) and SC, IV, and IM injections of antibiotics are commonly used in canine and feline medicine, antibiotic infusion into affected areas is an uncommon and rarely studied treatment modality. Though not used in this case, isolated limb perfusion may have been another method of achieving effective antibiotic concentrations locally while limiting systemic side effects. We believe that antibiotic delivery by local infusion may have achieved effective antibiotic concentration in the affected tissues, because the puppy in this case report demonstrated an apparent rapid response to local infusion with amikacin.

This is the first reported use of pentoxifylline in a case of NF. Pentoxifylline has been used in both human and veterinary medicine to improve microcirculation, to enhance tissue healing, and to decrease endotoxic effects of inflammatory mediators.^19,20 Adverse hematological effects observed in humans have not been demonstrated in dogs; however, high levels of pentoxifylline for long-term administration in septic shock cases are reportedly harmful.^19,21,22 Pentoxifylline therapy was initiated in this case when the puppy developed pulmonary signs suggestive of progressive streptococcal infection, raising concern for development of STSS secondary to NF. The reported mortality rate for canine STSS associated with bronchopneumonia is higher than that for NF alone.^4–7 Pentoxifylline administration (10 mg/kg PO q 8 hours for 7 days) may have benefited this puppy, although more research is necessary to clarify the effect of this medication on STSS associated with NF.

Daily physical rehabilitation of the right hind limb was an integral part of the treatment plan aimed at maintaining joint health and returning limb function. Cold therapy was initially used to decrease inflammation for the first 48 hours after surgery. Subsequently, warm compresses were applied frequently to increase regional blood flow and promote tissue healing. Massage and passive range-of-motion exercises were implemented to improve circulation and lymphatic flow, promote joint health, and maintain limb function in a nonweight-bearing, diffusely infected, poorly perfused limb. Reported effects of massage in both small animals and humans include increased local circulation and tissue oxygenation, reduction of edema, breakdown of scar tissue formation, and improved limb range of motion and mobility.²³ Literature on NF in human and veterinary medicine has focused primarily on diagnosis and treatment of the infection; however, interest has increased in improving functional outcome with the addition of physical rehabilitation. A recent study in human medicine reported that 30% of patients surviving NF have mild to severe functional limitation, with extremity involvement being significantly predictive of greater limitation.²⁴

Conclusion

A 5-month-old, female Great Dane was diagnosed with NF of the right hind limb based on physical examination, ultra-sonographic evaluation, fluid cytology and culture, and histopathology. Rapid diagnosis and immediate surgical debridement resulted in a favorable outcome. Success in this case was apparently achieved by broad-spectrum, systemic antibiotic therapy; active closed-suction drainage; daily fluid cytology; local amikacin infusions; pentoxifylline administration; and maintenance of limb function via massage and passive range-of-motion activities.

Acknowledgments

We extend special thanks to Barry Kipperman, DVM, Diplomate ACVIM, for case consultation and to River May, DVM, for cytological evaluation.