Streptococcus bovis Clone Causing Two Episodes of Endocarditis 8 Years Apart

Abstract

A patient had endocarditis caused by Streptococcus bovis twice 8 years apart. According to pulsed-field gel electrophoresis (PFGE) the two isolates were identical. Seven unrelated blood isolates of S. bovis yielded unique PFGE patterns. Considering this heterogeneous population structure our findings demonstrate the long-term persistence of an S. bovis clone in a patient with recurrent endocarditis.

Streptococcus bovis can be found in the lower intestine in 10 to 16% of healthy people (3). Colonization is more frequent (up to 56%) in patients with malignant or nonmalignant neoplasias of the gastrointestinal tract (3). These patients have an increased risk of bacteremia and endocarditis caused by S. bovis (3, 4). The basis for this association is unclear. It has been suggested that carcinoma of the colon may promote gastrointestinal colonization by S. bovis (1). Here, we report on a patient who had the same clone of S. bovis isolated from his blood on two occasions 8 years apart. The second episode of bacteremia was related to clinically documented endocarditis, and endocarditis was likely during the first episode.

A 69-year-old male developed nightly chills in May 1989. A systolic murmur was noted, and a transthoracic echocardiogram showed a thickened posterior leaflet of the mitral valve with a prolapse of both leaflets but no vegetations. S. bovis was isolated from cultures of his blood. The strain was intermediately sensitive to penicillin (MIC, 0.125 μg/ml), sensitive to trimethoprim-sulfamethoxazole, and resistant to clindamycin, erythromycin, and tetracycline. A test for beta-lactamase production was negative. The patient recovered under treatment with amoxicillin-clavulanic acid. A screen for occult blood in the patient’s stool was negative, and no further examination of the gastrointestinal tract was performed. After 8.5 years, in December 1997, the patient experienced a new episode of fatigue, weight loss, and dyspnea. A strain of S. bovis was again isolated from blood cultures, with a resistance pattern identical to the isolate from 1989. An echocardiogram showed a severely altered posterior leaflet of the mitral valve with a vegetation. The patient was successfully treated with penicillin, followed by ceftriaxone and gentamicin. Endoscopies revealed a 3-cm tumor in the fundus of the stomach (histology inconclusive) and a tubulovillous adenoma in the colon. The patient refused any further workup of his gastrointestinal lesions.

Pulsed-field gel electrophoresis (PFGE) was done on the two isolates from our patient and seven additional S. bovis isolates that had been cultured from the blood of seven patients seen at our institution during 1997. The isolates had been stored at −70°C. PFGE was performed on the isolates as previously described (2). Briefly, whole genomic DNA was restricted with SmaI, and fragments were separated in a CHEF DRIII unit (Bio-Rad, Glattbrugg, Switzerland) under the following conditions: with 0.5× Tris-borate-EDTA running buffer, at 6 V/cm, 14°C, and a 120°C angle, with a 1.2- to 54-s ramped switch time for 18 h. Gels were stained with bromide, and banding patterns were compared with the naked eye. PFGE revealed that the two isolates from the patient showed identical patterns. All other S. bovis strains had unique patterns without obvious similarities (Fig. 1).

FIG. 1.

PFGE of S. bovis strains isolated in 1989 (lane 1) and 1997 (lane 2) from the patient with recurrent invasive infection. Lanes 3 to 9 are blood isolates from seven patients with S. bovis bacteremia recovered in the same institution in 1997. M, molecular weight markers.

The repeated isolation of an identical clone of S. bovis many years apart documented in this patient must be interpreted against the background of the high genetic diversity of clinical isolates of S. bovis, as evidenced by the PFGE analysis of several contemporary clinical blood isolates. Before molecular typing methods became available, the time interval between two episodes of an infection due to the same organism was an important criterion to differentiate a relapse of the infection from reinfection with a new strain. Furthermore, sensitivities to antibiotics were taken into account, but both criteria were neither sensitive nor specific. The advent of molecular typing techniques has substantially improved the capability to differentiate relapse from reinfection. The isolation of genetically identical or closely similar strains during two infectious episodes may be taken as evidence for persistence of the pathogen, if the observed clone occurs infrequently (<5%) in the general population, i.e., if there is a low probability of becoming infected with the same clone twice. Considering the high genetic heterogeneity of epidemiologically unrelated isolates of S. bovis documented here, it appears highly unlikely that our patient acquired the same clone of S. bovis twice many years apart. We can not reliably determine, however, whether our patient had persistent, asymptomatic infection of his heart valve for 8.5 years or had a persistent gastrointestinal colonization which led to an infection of the mitral valve on two separate occasions. It is not known whether long-term latent infections of heart valves can occur, but given the lack of effective host defenses against cardiac vegetations, this is not very likely, and it therefore appears more plausible that our patient had two independent episodes of endocarditis resulting from chronic gastrointestinal colonization by the same clone.

Such long-term mucosal colonization with an identical clone has not been documented previously. Little is known about the transmission kinetics and population genetics of S. bovis, and studies using molecular tools are needed to characterize the biology of bacterial colonization of the gastrointestinal tract in more detail. Such studies should help to define whether the long-term persistence of a single clone in this patient is typical only for S. bovis or can occur with other mucosal colonizers as well. Another important issue raised by the present case is whether certain clones of S. bovis have features favoring a persistent colonization of the gastrointestinal tract or bloodstream invasion.