A 39-Year-Old Man With a Skin Infection

DR REYNOLDS: Mr M is a previously healthy 39-year-old man with erythema and swelling of his left index finger. Over a 2-day period, redness, pain, and swelling of the dorsal aspect of the proximal phalanx of his index finger on his left hand appeared, worsened, and spread, and when he was unable to close his fist, he sought care at the local emergency department.

Mr M does not recall any insect, animal, or human bites to the area. He has not had any trauma to his hand. Mr M's only ill contact is his young son, who has recurrent otitis media and has taken a number of different antibiotics recently. Mr M had 1 previous episode of what might have been a skin infection. About 6 months ago, he noticed an expanding area of redness on his thigh; a physician family member treated him with an unknown oral antibiotic and the area resolved.

Mr M's only medical problem is familial hypercholesterolemia, which has been treated with limited medication adherence. He has had episodes of dyshidrotic eczema in the past but denies any currently. Mr M occasionally takes simvastatin, 80 mg/d; he has no known drug allergies.

Mr M is a senior manager for a large company and travels frequently within the United States. Although he was born and educated in Europe, he has not traveled internationally in the past few years. He is married, monogamous, and lives with his wife and toddler son. He exercises regularly by biking, jogging, and swimming.

On physical examination, Mr M appeared fit and in no distress. His temperature was 96.7°F (36°C), his blood pressure was 124/70 mm Hg, his heart rate was 64/min, and his oxygen saturation was 95% on room air. He had a palpable, firm left epitrochlear lymph node but no palpable axillary nodes. There was no heart murmur nor stigmata of endocarditis. His left index finger showed an erythematous, edematous area of approximately 3 × 3 cm between the metacarpal phalangeal (MCP) and proximal interphalangeal (PIP) joints on the dorsal surface. Lymphangitic streaking was noted on the dorsum of the hand and extending to the lateral side of the forearm. Range of motion of the left index finger at the MCP and PIP joints was significantly limited.

Laboratory evaluation revealed a normal white blood cell count and lactate level. A hand radiograph showed soft tissue swelling without a foreign body or bony abnormality. Blood cultures were drawn but the lesion had no drainage to culture.

Mr M was given intravenous cefazolin and vancomycin for 24 hours in the emergency department. Because he did not improve, he was seen by the hand service and underwent an incision and drainage procedure. The wound did not reveal any pus; a wick was placed. He was admitted to the hospital and at the time of his interview was slowly improving with intravenous vancomycin and levofloxacin.

It started probably about 3 days ago, when I had a small—what appeared to be—mosquito bite, and it started to itch. Probably at that point, it hadn't grown in size, but with the itch, I applied some anti-itch cream and some Neosporin. Then probably a day or so later, it started to grow in size and began to throb, especially in the evenings. I took some ibuprofen. And then 2 days ago I felt sufficient pain, and I wasn't seeing any sort of improvement, so I felt that I should go to the ER.

Probably 8 months ago, I had what may have been a skin infection in my thigh where I was perhaps bitten; it started out like a small mosquito bite. I had some itchiness, and then it grew. At the time, my father-in-law prescribed some antibiotics. I don't know if that cured it or if it was already healing. But that particular infection grew to, gosh, 3 times the size of a quarter. I mean, it was very large.

I traditionally have not taken antibiotics. This is probably the third time in my life—I’m almost 40 years old—and it would never be the first course of action. I have one side of my family that, any time I’m sick for whatever reason, they will jump at the opportunity and say, “You’ve got to go get antibiotics.”

I also did notice when I went to the walk-in clinic to get this infection looked at that there were posters everywhere regarding antibiotics and whether they are absolutely necessary. I came away with the impression that there was probably a very loose prescription of antibiotics these days. And I even noticed it with my 1-year-old son, who's been on them ever since birth. And that's why we’ve got a particular concern with his situation, because at some point, they just don't work.

I don't think I have any medical problems that would make me prone to skin infection. I run, I swim, I cycle. I would say probably I play more than the usual person, and I consider myself to be very fit.

I had read one symptom being a fever or a temperature. I experienced none of those, and if it wasn't for the throbbing pain, I would not have checked into the doctor's at all, because if the redness had maintained—even if it had spread and not hurt, again, I probably wouldn't have thought a whole lot about it.

So, my real question is, what other symptoms do you look for? Because I could have come too late, given the fact that it was spreading.

What are the most common pathogens causing skin infections in otherwise healthy adults, what pathogen might be causing Mr M's infection, and what are the risk factors? Mr M exercises, had another skin infection, recently took an antibiotic, and has a son taking antibiotics. Do any of these put him at increased risk? What has caused the dramatic rise in prevalence of community-acquired methicillin-resistant Staphylococcus aureus (CA-MRSA)? What is the treatment strategy for simple, uncomplicated cellulitis? Is oral or parenteral therapy appropriate for more serious skin infections? Is Mr M at risk of more skin infections? What do you recommend for Mr M?

DR MOELLERING: What are the most common pathogens causing skin infections in otherwise healthy adults and what does Mr M likely have?
The common pyodermas include cellulitis, erysipelas, impetigo, furunculosis, and folliculitis.¹ In addition, since Mr M's infection involves the hand, paronychia should be included in this list. Uncomplicated cellulitis is almost always caused by S aureus or Streptococcus pyogenes (group A streptococci).^{1 - 2} Very rarely, other streptococci or gram-negative organisms have been isolated from these lesions, but this is the exception rather than the rule. Erysipelas, on the other hand, is a relatively specific manifestation of group A streptococcal skin infection and may be distinguished from cellulitis by the fact that the intracutaneous edema in this process produces palpable margins of the lesion and a peau d’orange appearance of the involved skin.^{1 - 2} Impetigo is a superficial dermatologic infection characterized by honey-colored crusting of the overlying skin and is caused by either S pyogenes or S aureus. ^{1 - 2} Furunculosis is invariably caused by S aureus, and folliculitis can be caused by a variety of organisms, including S aureus, Candida species, Pseudomonas aeruginosa, and, rarely, Enterobacteriaceae.^{1 - 2} The etiologic agents of paronychiae include S aureus, S pyogenes, Candida species, and herpes simplex (herpetic whitlow).^{1 - 2} A number of less common organisms associated with occupational or environmental exposure also cause skin infections.² Human bites may involve anaerobic bacteria, which are part of the normal mouth flora, Eikenella corrodens, viridans streptococci, and S aureus.³ Cat bites are associated with Pasteurella multocida infections, and Capnocytophaga canimorsus has been associated with dog bites.³ Significantly, with the exception of S aureus, all of the bite-associated organisms are susceptible to penicillin, and in some cases, especially with P multocida, they are relatively resistant to the oral cephalosporins.⁴ Erysipeloid is an infection due to Erysipelothrix rhusiopatheae and is seen in fish and crab handlers and occasionally in butchers and other persons exposed to raw meat.^{5 - 6} It often involves the fingers, presents as a subacute cellulitis with accompanying pain, and may also produce lymphangitic streaking, as was the case for Mr M.⁶ Bites of seals have been associated with Erysipelothrix and Mycoplasma species infections, while exposure to salt water and fresh water may lead to severe skin and soft tissue infections due to Vibrio vulnificus or other halophilic Vibrio species or Aeromonas or Pseudomonas species.^{6 - 9}

True cellulitis must be distinguished from noninfectious conditions that may mimic it. Patients with severe stasis dermatitis may have overlying erythema that is difficult to distinguish from cellulitis,¹ but generally this occurs bilaterally, which would be exceedingly rare in the case of cellulitis. There may be sufficient erythema overlying the lesions in deep vein thrombophlebitis to raise the possibility of cellulitis. Insect bites with hypersensitivity responses and acute gout (podagra) may also produce cutaneous changes reminiscent of cellulitis.¹ Fixed drug eruptions and contact dermatitis can also sometimes mimic cellulitis. The latter is germane for Mr M because he applied to his finger an antibiotic ointment that is a potential cause of contact dermatitis. Erythema migrans (associated with Lyme disease), when extensive, can sometimes be mistaken for common cellulitis.¹ Finally, the early stages of pyoderma gangrenosa and lesions of carcinoma erysipeloides (particularly inflammatory breast carcinoma) may also be confused with cellulitis.¹

Given the above, what is the etiology of Mr M's cellulitis? The salient features here are that there was no history of trauma or other exposure. The lesions were located on the fingers. The process had an acute to subacute onset with fairly severe pain. There was no suppuration or crepitance. He had marked lymphangitic streaking with local epitrochlear adenopathy and was afebrile. These findings, particularly Mr M's prominent lymphangitic streaking, suggest group A streptococcus as the source. Lymphangitic streaking can be caused by E rhusiopatheae or herpes simplex, as noted earlier, but there is nothing to implicate either of these organisms. Finally, although it is not typical of staphylococcal infections, staphylococcal cellulitis may also be associated with lymphangitic streaks.^{1 - 2} Thus, Mr M should receive antibiotic coverage for both S pyogenes and S aureus.

What are the risk factors for skin infections in otherwise healthy adults? Does Mr M's jogging, biking, and swimming put him at risk? The risk factors for skin infection in otherwise healthy adults include skin trauma and fungal dermatoses that lead to skin fissuring, as well as nasal carriage of S aureus and oropharyngeal or rectal carriage of S pyogenes.^{1 - 2} Nasal carriage of S aureus has been repeatedly demonstrated to be a significant risk factor for subsequent development of staphylococcal infections.^{10 - 12} Humans are the predominant reservoirs of S aureus; 15% to 35% of normal individuals harbor S aureus in their nares or pharynx at a given point.^{13 - 14} About 30% of these patients have prolonged carriage, 50% of the population shows intermittent carriage, and about 20% never carry these organisms.^{13 - 14} Vaginal carriage occurs in approximately 10% of postmenopausal women.¹³ Rectal and perineal carriage has also been described, particularly in patients colonized with MRSA.¹⁵ Staphylococcus aureus is usually spread by direct person-to-person contact, but inanimate objects or contaminated materials (such as whirlpool baths) may help transmit the organisms among individuals such as wrestlers and football players with skin trauma that predisposes them to infection.^{16 - 17} Mupirocin prophylaxis has been shown to prevent S aureus infections due to nasal carriage. A meta-analysis of 10 cohort or randomized controlled studies in 2445 renal dialysis patients demonstrated that mupirocin prophylaxis reduced the risk of S aureus infections by 68% in all patients, 80% in hemodialysis patients, and 63% among peritoneal dialysis patients.¹¹ Staphylococcus aureus bacteremia was also reduced by 78% in hemodialysis patients. Finally, peritonitis and exit-site infections were reduced by 66% and 62%, respectively, among peritoneal dialysis patients. Unfortunately, prolonged use of mupirocin has been associated with the emergence of resistance.¹⁸

There is no evidence that jogging or biking, as Mr M does, regularly predisposes to skin infections (in the absence of trauma). Exposure to improperly chlorinated fresh water has been associated with Aeromonas and Pseudomonas infections.^{8 ,19}

Does Mr M's use of an oral antibiotic 6 months ago make a resistant organism more likely? How much antibiotic resistance exists in the community? What has caused the dramatic rise in prevalence of CA-MRSA?As far as the previous use of antimicrobials is concerned, antimicrobial exposure within the previous 3 to 4 months is potentially considered to be associated with subsequent infection with an organism that is resistant to that antibiotic.²⁰ The exposure 6 months ago may have played some role in selecting a resistant organism in Mr M's case, especially if the infecting organism is MRSA. A study from the United Kingdom demonstrated that the risk of infection with CA-MRSA increased with the number of antibiotic prescriptions administered during the previous year and was highest for quinolones.²¹

Staphylococcus aureus resistance is particularly concerning because the organism is virulent and versatile and has repeatedly shown the ability to thwart antimicrobial therapy. The publication of the entire genomic sequence of strains of MRSA provides important insight into the ability of this organism to adapt to changing environmental conditions. Kuroda et al²² determined the entire genome sequences of 2 related MRSA strains, N315 (an MRSA strain isolated in 1982) and Mu50 (a vancomycin-intermediate S aureus strain isolated in 1997). They found that the genome is composed of approximately 2600 open reading frames (genes), many of which appear to have been acquired by lateral transfer from other organisms. This ability to acquire new genes undoubtedly accounts for this organism's remarkable ability to adapt to changing environmental conditions and develop resistance. When penicillin was first discovered, virtually all strains of S aureus were susceptible to the drug, but after only 10 years of use of penicillin at the Boston City Hospital, almost three-quarters of the S aureus causing infections in the hospital had developed penicillin resistance.²³ Initially, these staphylococcal isolates were confined to the hospital, but by 1967, there was no difference in susceptibility in inpatient and outpatient isolates and more than 80% from both settings were resistant to penicillin.²³ Currently, more than 90% of S aureus organisms worldwide are resistant to penicillin.²³ The discovery of the antistaphylococcal penicillins in the late 1950s provided some initial relief from the problems posed by penicillin-resistant staphylococci. Methicillin was first used clinically in 1959, but by 1961 the first strain of MRSA was described. These organisms subsequently caused significant problems in Europe and, in the late 1970s, spread to the United States, where they began to cause significant problems in tertiary care hospitals, especially burn and intensive care units.²⁴

Unlike penicillin resistance in S aureus, which is mediated by the production of β-lactamases that hydrolyze the β-lactam ring of penicillin and inactivate it, methicillin resistance has been shown to be due to the acquisition of genes that encode penicillin binding proteins (specifically, PBP2′ or 2A) that exhibit decreased affinity for β-lactams.²⁵ These altered penicillin-binding proteins are encoded by a mecA gene that is located on a transposable element inserted into the staphylococcal chromosome. The presence of the altered penicillin-binding proteins encoded by mecA enables the organism to complete cell wall synthesis even in the presence of high concentrations of standard penicillins, cephalosporins, or carbapenems. Of the 5 transposable elements encoding the mecA gene, types I, II, and III are found primarily in the hospital-associated strains of S aureus, while types IV and V are characteristic of CA-MRSA.²⁶

The prevalence of MRSA in hospitals in the United States has been steadily increasing and shortly after the turn of the century, the frequency of methicillin resistance among S aureus isolates in US hospitals surpassed 50%.²⁷ Not surprisingly, strains of hospital-acquired MRSA have spread to nursing homes throughout the United States.²⁸ Community-acquired MRSA was first described in the United States in Detroit and Boston in the 1980s among intravenous drug abusers, many of whom were taking oral antibiotics in an attempt to prevent staphylococcal skin infections at their injection sites.^{29 - 31} In the 1990s, outbreaks of staphylococcal disease due to community-acquired S aureus were described among the Aboriginal populations of Australia.³² The first real problems with CA-MRSA in the United States surfaced in the late 1990s with a description of 4 fatal infections among Native American children in Minnesota and North Dakota.³³ In the past decade, CA-MRSA has exploded throughout the United States.³⁴ By 2001, 72% of community-associated S aureus in Houston were resistant to methicillin, a percentage that increased to 76% in 2004.³⁵ In Atlanta, 72% of S aureus within the community were methicillin resistant in 2003,³⁶ and at present, 60% to 80% of community-acquired S aureus in the entire United States is methicillin-resistant.^{34 - 37} A report from the National Naval Medical Center in San Diego clearly documents the striking increase in CA-MRSA that occurred there between 2001 and 2004³⁸ and mirrors what has been described in most of the rest of the country.^{34 - 37} Another recent study of skin infections from emergency departments situated in various US locations also documents the striking increase in CA-MRSA, which accounted for 78% of all S aureus isolates from skin infections in that particular study.³⁷ Although most infections caused by CA-MRSA are non–life-threatening infections of the skin and soft tissues, a recent study documented that between July 2004 and December 2005, a significant number of cases of serious invasive MRSA infections in 9 locations in the United States (1234 of 8987 cases studied) were caused by CA-MRSA.³⁹

Genetic studies of community-acquired S aureus isolates from the United States have shown that all of the isolates thus far contain genes that encode Panton-Valentine leukocidin (PVL).⁴⁰ The PVL genes encode a toxin that damages cell membranes and destroys white blood cells. Previous studies among methicillin-susceptible S aureus containing these particular genes had suggested that there was increased mortality in patients with pneumonia due to PVL-containing strains compared with those infected with strains without PVL.⁴¹ Nonetheless, it is not clear at the present time whether the increased virulence that appears to be characteristic of a number of the community-acquired isolates is due to the PVL itself or to the production of other toxins or other factors in these organisms. Indeed, 2 recent studies have come to differing conclusions concerning the importance of PVL in the pathogenesis of CA-MRSA infections.^{42 - 43} By pulsed-field gel electrophoresis, a number of different types of CA-MRSA have been described, including USA300, USA400, USA1000, and USA1100.⁴⁴ All of these organisms contain the PVL genes, and many of them contain genes for other exotoxins as well. Interestingly, one of these clones (USA300-0114) proliferated and is now almost entirely responsible for the widespread dissemination of these organisms throughout the United States.⁴⁵ Although the genetic composition of this organism has been determined and a number of unique genes have been found in it, the exact basis for its achieving preeminence and for the incredibly rapid spread of this organism throughout the United States is still not clear.

Another interesting observation is that rates of nasal colonization with CA-MRSA appear to be much lower than for methicillin-susceptible S aureus. Kuehnert et al⁴⁶ found that the prevalence of nasal colonization by methicillin-susceptible S aureus in a National Health and Nutrition Examination Survey was 27% to 37% among a variety of populations studied in the United States, while the colonization rate for MRSA was 0.5% to 2.2%. This study was conducted in 2001-2002 and, thus, may have preceded the current explosion of CA-MRSA infections occurring in the United States. Nonetheless, it raises the possibility that nasal colonization may play less of a role in the spread of this organism than it does in methicillin-susceptible S aureus or that CA-MRSA may be more virulent (likely to cause infection upon colonization) than methicillin-susceptible S aureus.

Outbreaks of infections due to CA-MRSA in the United States were first reported among Native American children in Minnesota, Nebraska, and North Dakota,³³ among men who have sex with men,³⁴ and among prison inmates,³³ followed by reports of infections in contact sport participants (fencers, wrestlers, football players).^{40 ,44} Factors such as crowding, person-to-person contact, and low socioeconomic status (which are also relevant for methicillin-susceptible S aureus infections) appear to explain the increase in disease prevalence in many of the affected populations.⁴⁴ Although Mr M does not have any of these numerous risk factors, given the dramatic rise in prevalence of CA-MRSA, most individuals in the United States could be considered at risk. Unlike the hospital-acquired strains, CA-MRSA is more likely to cause skin and soft tissue infections, such as Mr M’s, and somewhat less likely to cause respiratory and urinary tract infections.⁴⁷ They have also been associated with a number of “new staphylococcal syndromes,”^{34 ,40} including necrotizing skin infections that are initially misdiagnosed as “spider bites.” The infections cause a good deal of purulence, and abscesses are not uncommon. Community-acquired MRSA is also known to cause pyomyositis, necrotizing fasciitis, and septic thrombophlebitis of the extremities. A “pelvic syndrome” has been described among young children^{40 ,48} that includes septic arthritis of the hips, pelvic osteomyelitis, pelvic abscesses, and septic retroperitoneal thrombophlebitis. Waterhouse-Friderichsen syndrome has also been seen for the first time with CA-MRSA.⁴⁹ Finally, these organisms are capable of causing rapidly progressive necrotizing pneumonia.⁵⁰ Despite receiving appropriate antimicrobial therapy, some patients with CA-MRSA pneumonia have had very severe diseases with lung necrosis, empyema, and, in some cases, rapid progression to death within 48 to 72 hours.⁵¹ In a recent series of 10 cases of severe CA-MRSA pneumonias in children, 6 died.⁵²

Compared with health care–associated MRSA, which is often resistant to multiple antimicrobial agents, the community-associated strains are relatively susceptible.⁴⁷ They are universally resistant to the antistaphylococcal penicillins and currently available cephalosporins, but most have been susceptible to vancomycin, trimethoprim/sulfamethoxazole, tetracyclines, rifampin, gentamicin, and ciprofloxacin.⁴⁷ They have shown variable susceptibility to clindamycin, and a significant number are resistant to erythromycin.⁴⁷ Recent studies show some disturbing trends among the most commonly isolated clone, USA300.^{53 - 54} In a San Francisco study,⁵³ one-quarter (of 188 strains tested) had become resistant to tetracycline and 63% to ciprofloxacin, again demonstrating the rapid emergence of resistance to fluoroquinolones when these agents are used alone to treat staphylococcal infections. In a study in Boston, levofloxacin resistance occurred in 58 of 73 (79%) of USA300-0114 clones tested.⁵⁴

What is the treatment strategy for simple, uncomplicated cellulitis? Is there any utility in obtaining blood cultures in such infections? In simple, uncomplicated cellulitis without fever or other constitutional symptoms, blood cultures are almost invariably negative,⁵⁵ but they may be useful in patients with more severe disease.⁵⁵ Mr M has a more serious infection, with rapid progression and lymphangitic streaking; thus, blood cultures were appropriate in this case despite his lack of fever. Even more important are cultures of the material from the surgical drainage. Because of the increasing prevalence of CA-MRSA, wound cultures should be obtained whenever possible.

Several years ago, the therapeutic options for oral therapy of uncomplicated cellulitis consisted of an antistaphylococcal penicillin or an oral cephalosporin (cephalexin or cefadroxil), possibly with penicillin for better coverage against group A streptococci.⁵⁶ Now, however, with the emergence of CA-MRSA, treatment choices are more difficult (Table). Linezolid is an effective but relatively expensive choice for treating such infections because it covers all staphylococci and streptococci effectively.⁶⁶ However, it may not be necessary to use such a “big gun” for uncomplicated cases or for cases in which the areas of purulence have undergone appropriate surgical drainage.⁶⁷ In these cases, the options include trimethoprim-sulfamethoxazole, minocycline or doxycycline, or clindamycin.⁴⁰ Trimethoprim-sulfamethoxazole is not optimally effective against S pyogenes, and if that is a likely pathogen (as it could be in Mr M's case), alternative therapy should be used or a β-lactam antibiotic added. Trimethoprim-sulfamethoxazole has not been evaluated definitively in a randomized trial for the treatment of uncomplicated staphylococcal cellulitis. An older, randomized, nonblinded trial demonstrated that trimethoprim-sulfamethoxazole was slightly less effective than vancomycin for bacteremia and endocarditis due to methicillin-susceptible S aureus, but appeared relatively equivalent for treatment of MRSA infections.⁶⁸ Trimethoprim-sulfamethoxazole has some evidence of effectiveness against S aureus, including its bactericidal activity against S aureus in vitro,⁶⁹ and a recent retrospective study suggesting that trimethoprim-sulfamethoxazole was effective for such infections.⁵⁷ However, a small trial demonstrated that trimethoprim-sulfamethoxazole was effective in only 11 of 14 patients treated.⁵⁸

Clindamycin has been used successfully in a number of skin and soft tissue infections due to S aureus.³⁴ However, some organisms are resistant to erythromycin and susceptible to clindamycin on routine laboratory testing. Such organisms may be resistant to erythromycin via efflux or the presence of genes encoding an inducible methylase (macrolide-lincosamide-streptogramin B [MLS_B]) that creates resistance to macrolides, lincosamides, and streptogramins.⁶⁰ In the inducible strains, the production of the methylase is induced by erythromycin but not by clindamycin; hence, the organisms appear susceptible to clindamycin in vitro. However, treatment with clindamycin creates the possibility of mutations, which lead to constitutive production of MLS_B and resistance to clindamycin, resulting in therapeutic failures,⁶⁰ as documented in anecdotal cases.^{70 - 71} This mechanism of resistance can be detected in the clinical microbiology laboratory by using a 2-disk method known as the D-test and should always be performed before clindamycin is used to treat an infection due to a macrolide-resistant organism.⁶⁰ In areas where resistance is low, clindamycin remains a useful drug for these infections.^{34 ,72} The other mechanism of erythromycin resistance is efflux, and clindamycin will work since it is not subject to the same efflux mechanisms as the macrolides. Long-acting tetracyclines (doxycycline or minocycline) may be effective for MRSA infections, but more data are necessary to define their role against CA-MRSA.⁵⁹ One recent small trial in patients from an area with a high prevalence of CA-MRSA found that doxycycline was effective in all 20 patients treated.⁵⁸

Is oral or parenteral therapy appropriate for more serious skin infections? Although the need for parenteral therapy for Mr M is not clear-cut, he probably was treated with intravenous antibiotics because of the location of the infection on his hand and because of the presence of lymphangitic streaking. Indications for initial parenteral therapy in patients with cellulitis include rapidly spreading lesions, a prominent systemic response (chills, temperature >100.5°F), and/or presence of concomitant disease or conditions predisposing to immunosuppression (neutropenia, immunosuppressive therapy, asplenia, preexisting edema, cirrhosis or renal failure). ² A number of parenteral agents can be used in this setting. Until recently, antistaphylococcal penicillins such as oxacillin or nafcillin were considered drugs of choice, but if MRSA prevalence is greater than 30%, vancomycin (or teicoplanin outside of the United States) is the agent of choice.⁶⁵ Linezolid is a reasonable alternative to vancomycin and covers both staphylococci (including MRSA) and streptococci.⁶⁶ Although quinupristin-dalfopristin is likely to be effective in this setting, it is not widely used at present because it can cause thrombophlebitis and an arthralgia/myalgia syndrome.⁷³ Daptomycin is a reasonable alternative to vancomycin.⁷⁴ Tigecycline has also been studied in this setting, but its spectrum of activity is considerably broader than necessary for uncomplicated infections of the skin and skin structures.⁶⁴ A number of investigational agents are completing phase 3 trials, including dalbavancin,⁷⁵ oritavancin,⁷⁶ telavancin,⁷⁷ ceftobiprole,⁷⁸ and ceftaroline.⁷⁹ Of these, dalbavancin has particular potential for CA-MRSA infections. This drug has an exceedingly long serum half-life with the potential to treat such infections with a single dose on an outpatient basis.⁷⁵ Currently, however, the drug has not been marketed in the United States or elsewhere.

Mr M has improved after drainage and 3 days of antibiotics. What course of antibiotics do you recommend now? Assuming that Mr M has improved sufficiently, he could be discharged home taking oral antimicrobial therapy. His infection was most likely caused by group A streptococci or S aureus, and unless the postoperative cultures demonstrate otherwise, this infection could be due to a CA-MRSA strain. In addition, since his infection involves his hand, antimicrobial coverage should be adequate to cover those possibilities, and, although it is expensive, oral linezolid would be a reasonable choice. Trimethoprim-sulfamethoxazole does not cover group A streptococci well and, given the lymphangitic streaking, should not be used. Clindamycin is also problematic because no organisms have been cultured to test for genes that encode resistance. Minocycline or doxycycline have only limited data available on efficacy against CA-MRSA.⁵⁹ Finally, fluoroquinolones do not represent first-line therapeutic agents for such infections, given the significant resistance to these compounds among the USA300 strain, as described in reports from San Francisco⁵³ and Boston.⁵⁴ Rapid emergence of resistance to fluoroquinolones also has been documented in hospital-acquired strains of MRSA.⁸⁰

Now that he has had 2 skin infections, does Mr M need to worry that he will have more and, if so, how can he avoid them? Does he need an evaluation for underlying immune deficiency? Because the initial episode Mr M described was probably a very mild infection (if that), he does not fit the description of a patient with recurrent infections. Most of the time, patients with recurrent infections have nasal colonization with recurrent staphylococcal furunculosis or compromised extremities with lymphedema secondary to cancer therapy or venous harvesting for coronary artery bypass grafting.² They often experience recurrent group A streptococcal infections as well. Mr M has none of these conditions. The immunological diseases associated with staphylococcal infections include chronic granulomatous disease, Chédiak-Higashi syndrome, and Job syndrome,⁸¹ but all are congenital abnormalities and Mr M would have experienced infectious problems at a much younger age. For these reasons, Mr M does not need an evaluation for underlying immune deficiency.

DR REYNOLDS: Mr M was discharged from Beth Israel Deaconess Medical Center after he received 72 hours of intravenous antibiotics. Just before his discharge, his wound culture (obtained in the operating room) grew MRSA. The species was sensitive to gentamicin, levofloxacin, rifampin, tetracycline, trimethoprim-sulfamethoxazole, and vancomycin.

QUESTION: For a patient in whom you diagnose CA-MRSA soft tissue infection, what is the role of testing for treating a nasal carriage state?

DR MOELLERING: Because the nasal carriage rate of CA-MRSA is so low in studies carried out to date, there are no recommendations for routine nasal cultures in this setting. If studies under way begin to show higher prevalence of nasal carriage or demonstrate rectal or perineal carriage, then such cultures might have clinical relevance.

QUESTION: On one level you're suggesting that anyone who has anything red in their skin should run to the hospital, reach out their arm, and say, “Give me an IV so that my lung doesn't look like a liver.” On another level, we can't hospitalize everybody who has cellulitis. We can do a lot at home now. Do we have to hospitalize everyone who comes in with cellulitis?

DR MOELLERING: No, certainly that is not my take-home message. Although CA-MRSA is capable of causing these very serious infections, most of the infections that they cause are much less severe, and some will respond to surgical drainage without antimicrobial therapy.

A recent study describes patients with infections due to CA-MRSA seen in a number of emergency departments across the United States.³⁷ In it, the authors could not determine the effect of antimicrobial therapy. That is because the patients were preselected to have purulent collections. When these lesions are treated with appropriate surgical drainage, the role of antibiotics is less important.^{67 ,72} But some of these infections produce rapidly spreading cellulitis, necrotizing fasciitis, or other serious complications. Those are the ones in which effective antimicrobial therapy is mandatory.¹

QUESTION: Would it be possible to use trimethoprim-sulfamethoxazole and penicillin together for cellulitis?

DR MOELLERING: Absolutely—a perfectly reasonable thing to do. Here you have 2 relatively inexpensive agents that would cover both S aureus and S pyogenes. The only minor drawback is the need to administer the penicillin 4 times per day for optimal coverage. It is an alternative regimen that could have been considered in Mr M's case.