CASE #1: Cellulitis
Cellulitis is an infection of the subcutaneous fat that is so closely related to erysipelas, an infection of the dermis, that some authorities use the terms interchangeably. Unlike erysipelas, the borders of cellulitis are not raised or sharply demarcated. Cellulitis is marked by warmth, erythema, edema, and tenderness. Erythematous streaks may radiate from the affected tissue. The skin may be smooth or have the texture of orange peel. Occasionally ulcers or bullae are seen.
Most cellulitis is caused by Streptococcus pyogenes or Staphylococcus aureus. Other bacteria, such as non-group A streptococci (i.e., groups B, C, and G) and the gram-negative microorganisms Escherichia coli, Pseudomonas, Proteus, Serratia, Enterobacter, Citrobacter, and marine Vibrios, can also cause cellulitis. IV drug abusers who “skin pop” (inject liquid drugs into the fat rather than blood vessels) have been reported to suffer from cellulitis due to Enterococcus faecalis and Enterobacteriaceae organisms that otherwise only very rarely cause cellulitis, as well as Bacteroides and Clostridium species. Rare causes of cellulitis include atypical mycobacteria and deep fungal pathogens, the most common being cryptococci.
The estimated incidence of cellulitis is two to three cases per 100 people per year (24.6/1,000 person-years). Cellulitis affects men and women equally in most studies, although a higher incidence among males has recently been reported. Most cases involve the leg. Facial cellulitis and perianal cellulitis are most common in children. Breast cellulitis is most often caused by streptococcus and occurs primarily in women with breast cancer or in those who have undergone lumpectomy with lymph node dissection for stage 1 or 2 breast cancer.
No test is particularly sensitive or specific for cellulitis. A WBC count >10,000/µL with fever is present in up to 42% of patients. Needle aspiration of the advancing edge of cellulitis may be subjected to Gram's stain and cultured. Culture is diagnostic in approximately 30% of patients. Blood cultures have a higher positive yield in facial and periorbital cellulitis. Ultrasonography can diagnose an underlying abscess and help guide needle aspiration. CT or MRI may rule out underlying fasciitis or osteomyelitis. Radiographs reveal gas in the tissues, helping to establish diagnoses of fasciitis or gangrene, which require immediate surgical intervention.
The differential diagnoses for leg cellulitis include stasis dermatitis, allergic contact dermatitis, necrobiosis lipoidica diabeticorum, necrotizing fasciitis, toxic epidermal necrolysis, and gas gangrene, all of which are similarly painful. Facial cellulitis must be distinguished from Stevens-Johnson syndrome, angioedema, rosacea, allergic contact dermatitis, and lupus. Perianal cellulitis can be confused with diaper dermatitis but will not respond to topical antifungal agents or barrier creams. Both facial and perianal cellulitis tend to be unilateral. Impetigo results in alteration of the epidermis but not the dermis or fat and is not typically painful.
With the increasing prevalence of methicillin-resistant S. aureus (MRSA), choosing the correct antibiotic for empiric treatment is more important than ever. In healthy patients, mild cellulitis may be treated on an outpatient basis with oral antibiotics. I use trimethoprim-sulfamethoxazole 160/800 mg (Bactrim DS, Septra DS) b.i.d., which does not cover streptococcus but does cover many cases of community-acquired MRSA, with cefdinir (Omnicef) 300 mg b.i.d. (to cover streptococcus). Avoid using quinolones, as rapidly emerging resistance can extend to other classes of antibiotics.
IV antibiotics should be given for more extensive cellulitis; cellulitis involving burns or ulcers, fever, and elevated WBCs; and cellulitis in the context of immunosuppression (e.g., diabetes). At the very least, patients should receive amoxicillin-clavulanic acid (Augmentin) or piperacillin-tazobactam (Zosyn). Whether the IV antibiotic should always be vancomycin is arguable; a first-generation cephalosporin (e.g., cefazolin [Kefzol, Ancef]) administered with vancomycin can increase the potency of the regimen. Empiric vancomycin is necessary when the patient has impaired circulation or immune dysfunction (e.g., secondary to diabetes).
Gram-negative coverage is not generally needed but can be added if testing finds a gram-negative pathogen or if the patient fails to improve. Abscesses must be drained for antibiotic treatment to be effective. Underlying osteomyelitis requires an extended treatment regimen that can last six weeks.
Our patient's fever and WBC count normalized within four days of starting empiric vancomycin and cefepime. After one week of trimethoprim-sulfamethoxazole 160/800 mg b.i.d. with cefdinir 300 mg b.i.d., the cellulitis had abated altogether. The causative pathogen was not identified.