James S. Tan

Antibiotic prophylaxis has been recommended to prevent infective endocarditis in patients with valvular heart disease undergoing dental procedure. Erythromycin and clindamycin have been recommended as alternative agents in those who are allergic to penicillin. We report a penicillin-allergic patient who developed infective endocarditis despite clindamycin prophylaxis. Viridans-group Streptococcus that was later identified as Streptococcus sanguinis resistant to clindamycin was isolated from her blood. Initial therapy with clindamycin did not improve her symptoms. Her symptoms responded to linezolid therapy. She underwent mitral valve replacement because of deteriorating cardiac function. Postoperatively, she was treated successfully with linezolid and levofloxacin intravenously for 4 weeks. She remained well on her 30-month follow-up. Viridans-group Streptococcus, as a group, has been shown to be increasingly resistant to various antimicrobial agents including clindamycin and erythromycin. Although the percentage of resistant viridans-group Streptococcus strains is still low in most areas, local resistance should be monitored. If this trend should continue, the value of these antibiotics as prophylactic agent should be reevaluated.

Superficial fungal infections of the foot (tinea pedis and onychomycosis) are common among elderly patients. Although most authorities believe that patients with diabetes mellitus have an increased predisposition to dermatophytic infections, some controversies still remain. Because these infections disrupt the skin integrity and provide an avenue for bacterial superinfection, elderly diabetic patients with dermatophytic infection should be promptly treated with an antifungal agent. For most dermatophytic infections of the foot, topical agents are usually effective and less expensive than oral agents. Laboratory diagnosis of fungal infection prior to institution of therapy is recommended. Proper technique for obtaining the specimen is important to ensure a higher chance of isolating the infecting fungus. Commonly used anti-dermatophytic agents that are also active against the yeasts include the imidazoles, the allylamines-benzylamines and the hydroxypyridones, which are also effective against most of the moulds. Oral therapy for tinea pedis, although not well studied, should be limited to patients with more extensive infections, such as vesicobullous and moccasin type, resistant infections or chronic infections. In addition, oral agents should also be considered in diabetic and immunosuppressed patients. On the other hand, treatment of onychomycosis of the foot usually requires systemic therapy. Griseofulvin is the least effective agent when compared with the newer agents. Terbinafine, itraconazole and fluconazole have been shown to have acceptable cure rates. More recently, topical treatment of the nail with 8% ciclopirox nail lacquer, bifonazole with urea and amorolfine have been reported to be successful. Over the past decade, fungal foot infections of the skin and nail are more effectively treated with the introduction of numerous topical and oral agents.

We describe 4 patients infected with levofloxacin-resistant pneumococci after therapy for community-acquired pneumonia (CAP). The 4 patients had 15 episodes of CAP; Streptococcus pneumoniae was isolated from blood or sputum samples obtained during 14 of the episodes. The underlying medical condition was Bruton agammaglobulinemia in 3 patients and chronic lymphoid leukemia in the other. The initial episode of CAP in each patient was due to a levofloxacin-susceptible strain. One of 4 reinfections and 5 of 6 relapses were due to levofloxacin-resistant strains. All of these strains had amino acid substitutions in the quinolone-resistance–determining region of the genes parC and gyrA. The time between episodes of pneumonia varied from 1 to 4 months. In immunocompromised patients with suspected or proven pneumococcal infection, it may be prudent not to use fluoroquinolone monotherapy empirically when the patient has a history of fluoroquinolone therapy in at least the past 4 months.

Sir—We agree with Khatib and Fakih [1] that the diagnosis of Legionella coinfection should be made on firm grounds when infection is suspected to be due to multiple agents. We used the stricter criteria recommended by Marston et al. [2], such as a 4-fold increase in the serologic titer or positive results of urinary antigen tests for detection of Legionella pneumophila serogroup 1 (Lp1). In their letter, Khatib and Fakih [1] stated that they questioned the specificity of serologic tests because of cross-reactions with other organisms. They quoted a review article by Waterer et al. [3] that questioned the specificity of serologic tests for Legionella species because of false-positive test results for infections due to several bacteria. Four papers were referenced by Waterer et al. [3]: Tsai and Fraser [4], Boswell et al. [5], Musso and Raoult [6], and Johnson et al. [7].

We describe 6 patients who had bacteremic community-acquired pneumonia and unsuspected Legionella pneumophila coinfection. We reviewed case records o

We describe a patient with community-acquired pneumonia due to Legionella pneumophila serogroup 6. This patient was found to have bronchoalveolar car