In December 1998, a 57-year-old Italian man with a severe mitral insufficiency of rheumatic origin underwent a nonbleeding dental procedure without antimicrobial prophylaxis. One month later, he was hospitalized at the University Hospitals of Geneva with fever (38°C–39°C), rigors, and weight loss of 4 kg in 3 weeks. There was no history of prior endocarditis, drug abuse, traumatism, or concomitant disease. At admission he was hemodynamically compensated and without fever. Over the cardiac apex, a grade 5/6 proto-mesosystolic murmur was audible. Transesophageal echocardiography showed vegetation on the mitral valve without abscesses. An ophthalmologic examination showed an embolus near the right macula. The urinary sediment exhibited microhematuria. In all blood-culture bottles, a gram-positive Streptococcus sp. was grown. Thus, infective endocarditis was diagnosed by the presence of 2 major and 3 minor criteria according to modified Duke criteria (5).

An antimicrobial drug treatment with intravenous (i.v.) penicillin G 6 × 4 million units/day and gentamicin 3× 1 mg/kg/day for 3 weeks was initiated. After an excellent clinical course and normalization of inflammation markers, treatment was switched to i.v. ceftriaxone 1× 2 g/day for another 3 weeks to enable outpatient therapy (6). No secondary abscesses occurred. Several infected teeth were extracted during treatment. Because of the severity of the preexisting mitral regurgitation, an elective replacement with a mechanical prosthesis was performed in March 1999. The patient was considered cured. Three years later he died of cerebral hematoma attributed to oral anticoagulation. Autopsy did not show any sequelae of former infection.

All 6 blood-culture bottles (3 BACTEC aerobic plus/F and 3 BACTEC anaerobic lytic/F; Becton Dickinson, Sparks, MD, USA.) were positive with gram-positive cocci in chains. The isolate grew as transparent α-hemolytic colonies (0.5- to 1-mm diameter) on sheep blood Columbia agar after an incubation of 24 h at 35°C in 4% CO₂-enriched atmosphere. The biochemical identification system API 20 Strep (bioMérieux, Lyon, France) was used to attempt the identification. The numerical profile was 4241450, indicating S. sanguinis (95% similarity). Agglutination with Lancefield antisera (Streptokit, bioMérieux) was negative for groups A, B, C, D, F, and G. Antimicrobial MICs were determined with Etest (AB Biodisk, Solna, Sweden). Results were interpreted according the available Clinical Laboratory Standard Institute (CLSI, formerly NCCLS) criteria (7) (Table).

The genetic sequence of the 16S rRNA was determined by a capillary sequence analyzer (ABI 3130 XL DNA Analyzer, Applied Biosystems (Foster City, CA, USA) and compared with the nucleotide sequences in GenBank. Of >1,000 bp in the 16S rRNA sequence, only 2 differed from the previously published S. sinensis HKU4 (AF432856) (1), a sequence identical to that of HKU5 (AF432855) and HKU6 (AF432857) (4). Results were positive for identifying the superoxid dismutase (sodA) (primer from reference 8) and RNA polymerase β-subunit (rpoB) (9) housekeeping gene sequences. Concerning sodA, for >404 bp there was 97.3% homology with the positive S. sinensis strain AY386220 (9) and only 2 bp differences in >286 bp (99.3% homology) with the S. sinensis strain EF451825. Concerning rpoB, 485 bp were identical over 516 nt sequences (94% homology) of the S. sinensis strain AF199923 (9).

We report a case outside Hong Kong of infective endocarditis caused by a strain of S. sinensis. Information on this novel pathogen has been published for 3 Chinese patients (2–4). Ours is the fourth published case worldwide. Because the original S. sinensis (1) is most closely related to our strain, we believe that our isolate belongs to that species. The close relationship of the sequence of our strain to the referential sodA and rpoB gene sequences also identifies our Geneva strain as S. sinensis. We have submitted the sodA and rpoB sequencing results to GenBank (accession nos. EF585234 and EF591041, respectively).

A 16S rRNA sequence with only 3 base differences from HKU4 had been detected by a German group in the aortic valve of another patient (GenBank accession no. AY049738, unpub. data), and other sequences have been reported by a French group (EF371928, unpub. data) and a British group (AY386220, unpub. data). We do not know more about these cases. Our Geneva strain differed only in 2 nt bases from the HKU4 strain (on positions 43 and 48) and only in 1 base (position 66) from the German sequence, whereas it is identical with the French sequence. This might suggest the emergence of a European strain of S. sinensis. From a clinical point of view, all reported patients (1,4) had an underlying chronic rheumatic heart disease of the mitral valve as the major risk factor. Only l patient had a preceding tooth extraction (4), like our patient with teeth abscesses. No other infections besides endocarditis caused by S. sinensis have been described thus far in humans or in animals. All previously reported patients were successfully treated by i.v. penicillin G or ampicillin for 4 weeks (combined therapy with gentamicin during the first 2 weeks) without necessity for surgical intervention. In light of other streptococcal endocarditic recommendations (10,11) and reported experiences with S. sinensis (4), our 6 weeks of antimicrobial drug treatment was probably excessive. The elective mitral valve replacement was indicated because of the severity of the preexisting regurgitation.

In conclusion, our case outside Hong Kong confirms that S. sinensis causes endocarditis throughout the world. Like other viridans streptococci, S. sinensis might be part of the human oral flora. The real number of cases is probably underestimated because commercial kits misidentify S. sinensis as S. intermedius or S. anginosus (1). With adequate sequencing technology, further reports may indicate the real prevalence of this emerging pathogen.