JOURNAL OF CLINICAL MICROBIOLOGY, Oct. 2010, p. 3791–3793 Vol. 48, No. 10 0095-1137/10/$12.00 doi:10.1128/JCM.00238-10 Copyright ? 2010, American Society for Microbiology. All Rights Reserved.
Aortic Valve Endocarditis Possibly Caused by a
1,245April Buscher,Linda Li,Xiang Y. Han,and Barbara W. Trautner1,2,3 *
2Section of Infectious Diseases, Department of Internal Medicine, Baylor College of Medicine, Houston, Texas; Houston Center for 1Quality of Care and Utilization Studies, Houston, Texas; Michael E. DeBakey Veterans Affairs Medical Center, Houston, 34Texas; Department of Surgery, Baylor College of Medicine, Houston, Texas; and Department of Laboratory Medicine, 5The University of Texas M. D. Anderson Cancer Center, Houston, Texas
Received 4 February 2010/Returned for modi，cation 3 March 2010/Accepted 28 June 2010
Haematobacter is a newly proposed genus for a group of fastidious Gram-negative aerobic bacilli isolated mostly from blood samples from patients with septicemia. The Haematobacter genus currently includes two species, H. massiliensis and H. missouriensis. We report isolation of a novel Haematobacter-like species from the blood of a 65-year-old man who suffered from probable aortic valve endocarditis. The possible causative role was suggested by the monomicrobial culture and the absence of another causative agent in a patient with probable endocarditis by Duke criteria. This fastidious organism could not be identi;ed by routine biochemical tests. Sequencing analysis of the 16S rRNA gene (1,425 bp) best matched the known Haematobacter species yet was substantially different with a nucleotide similarity of 96.7%. This strain also reduced nitrate to nitrite, unlike known species. This case is likely the ;rst reported case of endocarditis possibly caused by a Haema- tobacter-like organism.
urinary catheter placement revealed large amounts of blood CASE REPORT and 71 red blood cells per high-power microscopy ，eld. HIV A 65-year-old man presented to the emergency room in test results were negative. December 2008 with a 3-week history of worsening shortness Admission chest X-ray showed patchy densities in the right of breath, orthopnea, subjective fevers, chills, night sweats, dry upper and lower lung ，elds as well as the perihilar region of cough, and sore throat. The shortness of breath occurred ini- the left lung consistent with pulmonary edema. An electrocar- tially during exertion, but on the day of presentation, it also diogram showed sinus tachycardia with premature atrial com- occurred at rest. The patient had little prior medical care. He plexes, a right bundle branch block, and left ventricular hyper- had a 50-pack-per-year history of smoking as well as a history trophy. A bedside transthoracic echocardiogram showed of alcohol abuse. moderate to severe aortic insuf，ciency as well as a lesion on the On physical examination, the patient was in moderate respi- aortic valve. ratory distress. He was afebrile with a heart rate of 116/min and Blood cultures using two sets of aerobic and anaerobic bot- blood pressure of 149/88 mm Hg. Chest auscultation revealed tles were obtained on admission. Antimicrobial therapy with diffuse wet rales in the lung ，elds, a III/VI early diastolic vancomycin and piperacillin-tazobactam was initiated 6 h later, murmur heard best at the left upper sternal border with radi- along with a regimen of diuretics, beta-blockers, and statins for ation to the apex, and an S3 gallop. Oral thrush, bilateral his heart failure and hypertension. On hospital day 3, a trans- axillary lymphadenopathy (1 to 2 cm), and bilateral lower ex- esophageal echocardiogram revealed further details of the dis- tremity pitting edema were also noted. eased aortic valve, i.e., a partially ！ail left coronary cusp along Laboratory tests revealed a white blood cell count of 18.9 3with the presence of two small mobile vegetations, one (0.4 by 10/ l with 85% neutrophils, hemoglobin level of 11.4 g/dl, 0.3 cm) on the tip and another (0.2 by 0.1 cm) at the base. The blood urea nitrogen level of 42 mg/dl, creatinine level of 2.4 ejection fraction was slightly impaired (45 to 49%) (normal mg/dl, total bilirubin level of 2.2 IU/ml (normal range, 0.2 to value, 55%). 1.2 IU/ml), and erythrocyte sedimentation rate of 94 mm/h Meanwhile, a Gram-negative bacillus, strain BC14248, grew (normal range, 0 to 20 mm/h). The following laboratory results in one of the two aerobic bottles after an incubation of 5 days were also abnormal: lactate dehydrogenase level of 346 U/ml in the BacT/Alert automated culturing system (bioMe?rieux (normal range, 100 to 190 U/ml), creatinine kinase level of 524 Inc., Durham, NC). The patient then met Duke criteria for U/ml (normal range, 25 to 250 U/ml), creatinine kinase level in probable endocarditis (echocardiographic ，ndings of infective muscle and brain of 15.4 ng/ml (normal range, 0 to 5 ng/ml), endocarditis, positive blood culture not meeting major criteria, and troponin I level of 7.70 ng/ml (normal range, 0.01 to 0.08
fevers, and hematuria) (4). On subculture, the organism grew ng/ml). The levels of transaminases and alkaline phosphatase
were within normal limits. A urinalysis conducted 5 days after best on 5% sheep blood agar, but the appearance of colonies
(1-mm size) required 48-hour incubation at 37?C with 5% CO. 2 The colonies were light gray, round, raised, glistening, mucoid, * Corresponding author. Mailing address: Michael E. DeBakey Vet- and sticky. There was no hemolysis. Similar growth was also erans Affairs Medical Center (102), 2002 Holcombe Boulevard, Hous- seen with nonselective buffered charcoal yeast extract agar. ton, TX 77030. Phone: (713) 794-8610. E-mail: email@example.com. However, the organism grew even more slowly on chocolate Published ahead of print on 7 July 2010.
3792 CASE REPORTS J. CLIN. MICROBIOL.
original Gram-negative bacillus, was found in multiple blood cultures. To identify the fastidious Gram-negative rod, we ampli，ed and sequenced its 16S rRNA gene using the MicroSeq 500 system (Applied Biosystems, Inc., Foster City, CA) and two more sets of 16S primers, primers 5 -TGCCAGCAGCCGCGGTAATAC and 5 -CGCTCGTTGCGGGACTTAACC and primers 5 -GCA CAAGCGGTGGAGCATGTG and 5 -AGGAGGTGATCCA ACCGCA (3). Nearly the full length (1,425 bp) of the gene was obtained (GenBank accession no. GU396991), and BLAST searches showed the best match with an uncultured bacterium (6) with 98.9% identity (1,402 of 1,417 bp without gap sites). The second best matches were with Haematobacter massiliensis (GenBank accession no. AF452106) for 96.7% identity (1,354/ 1,400 bp), with Haematobacter missouriensis (GenBank acces-
sion no. DQ342315) for 96.8% identity (1,342/1,387 bp), and
with Haematobacter genomospecies (GenBank accession no.
DQ342319) for 96.5% (1,339/1,387 bp). Yet, these matches all
contained four gap sites. Other close matches were with two Rhodobacter species (GenBank accession no. DQ342322 and CP000661) for identities of 96.5% to 95.0%. Therefore, this or- ganism probably represents a novel Haematobacter-like species.
FIG. 1. Gram stain of the organism showing serpentine rod forms.
Discussion. De，nitive or probable infective endocarditis (IE) by modi，ed Duke criteria consists of either a pathological agar or plain Trypticase soy agar and did not grow at all on cardiac specimen or ful，llment of a range of clinical criteria MacConkey agar. With Gram staining, the organism was seen (4). This patient ful，lled the clinical criteria for probable IE.
He satis，ed one major criterion by having oscillating mobile as short to long serpentine rods (Fig. 1).
Biochemically, the organism was positive for catalase, oxi- vegetations attached to the aortic valve and a new valvular dase, and urease. It reduced nitrate to nitrite and produced regurgitation. He also met three minor criteria in that he had
subjective fevers at home prior to admission, a positive blood HS by the lead acetate method. However, it did not produce 2indole or hydrolyze gelatin or esculin. The API 20NE system culture of an organism not known to cause endocarditis, and (bioMe?rieux) yielded a biocode of 1041044 at 48 h, and the hematuria (albeit after urinary catheter placement). The pa- Vitek GNI card result was 40504004140, but neither code tient’s presentation of new congestive heart failure was also gave con，dent identi，cation. The strain was also sent to the impressive, along with IE-associated laboratory abnormalities, Centers for Disease Control and Prevention (CDC) for further including anemia, marked leukocytosis, and an elevated eryth- work-up, but it did not match with any known organism. The rocyte sedimentation rate (5). He showed a right bundle organism was tested for susceptibility to antimicrobial agents branch block on his admission electrocardiogram as well. The
possible causative role of the Haematobacter-like species for by Etest (bioMe?rieux) using standard Mueller-Hinton agar and
a 40-hour reading; it was pansensitive to ceftazidime, cefepime, this patient’s endocarditis was suggested by the monomicrobial cipro！oxacin, amoxicillin-clavulanate, piperacillin-tazobactam, culture at the time of his initial presentation along with asso- amikacin, imipenem, and trimethoprim-sulfamethoxazole with ciated signs and symptoms. Although Enterococcus grew from
his blood cultures 3 months later, this was possibly because his MICs of 0.5 g/ml for all these antimicrobial agents. Subse-
quently, the patient’s antibiotic was changed to ceftriaxone for initial endocarditis with the Haematobacter-like organism had
damaged his heart valves and created a predisposition for ease of administration.
endocarditis. During the patient’s hospitalization, an axial computed to-
The genus Haematobacter was proposed in 2007 by a team at mography of the neck and thorax showed evidence of irregular
the CDC based on studies of 13 strains isolated in France (1 soft-tissue lesions involving the anterior portion of the right
strain) and the United States (12 strains) (2). The French vocal cord; these lesions were suspected to be laryngeal carci-
strain was initially named Rhodobacter massiliensis based noma. He also had a spiculated nodule in the right upper lobe
of the lung. However, the patient refused further work-up. He mainly on the close match of its 16S rRNA gene with a few
known Rhodobacter species (up to 96%) (1). However, this improved clinically on treatment in that his acute heart failure
strain differed from the Rhodobacter species by the apparent symptoms, leukocytosis, renal failure, and hematuria resolved.
lack of red pigmentation and several biochemical reactions, He was discharged from the hospital and went home on ceftri-
and further characterizations of similar CDC strains led to axone to complete 6 weeks of therapy for this probable native
revision to the novel genus Haematobacter (2). Organisms aortic valve endocarditis. After this treatment, he was clinically
within Haematobacter are fastidious, nonfermentative, aerobic, stable until 3 months later, when he was admitted for endo-
carditis again. However, this time only Enterococcus, not the and short to long serpentine Gram-negative rods. They are
VOL. 48, 2010 CASE REPORTS 3793
positive for catalase, oxidase, and urease but negative for ni- 1,389 bp respectively (3, 6). The biologic relevance of a shorter
16S rRNA gene remains to be determined. trate reduction or indole production. Approximately half of
In summary, our case is likely the ，rst reported case of the strains of Haematobacter also grow on MacConkey agar.
endocarditis possibly caused by a Haematobacter-like organ- Due to the limited number of positive biochemical reactions,
ism. This entity should be added to the vocabulary of clinical de，nitive identi，cation requires 16S rRNA gene sequencing at
microbiologists. present. The genus currently includes two described species, H.
Nucleotide sequence accession number. The 16S rRNA gene massiliensis (n 7) and H. missouriensis (n 5), and one
sequence of the Haematobacter-like species strain BC14248 unnamed genomospecies (n 1), all separated mainly by
has been deposited in GenBank under accession number 70% hybridization of genomic DNA. These organisms have GU396991. similar biochemical features, and the 16S rRNA gene se- quences of H. massiliensis and H. missouriensis differ by only 2
We thank the microbiology laboratories at the Houston Michael nucleotides (of 1,389 bp). DeBakey VA Medical Center, the CDC Special Pathogens Branch, Our strain (BC14248) exhibits the general features of and the University of Texas M. D. Anderson Cancer Center for bio- Haematobacter in its fastidiousness, Gram stain morphology, chemical studies on this organism. and limited biochemical reactions. However, this strain is also This work was supported in part by a Veterans Affairs Research and Development grant (B4623) (to B.W.T.) and by the Houston HSR&D mucoid and glistening on culture plates, similar to the French Centers of Excellence (HFP90-020). We have no con！ict of interest. strain, whereas other U.S. Haematobacter strains have not been
reported to have these features. The main difference of our REFERENCES isolate lies in the 16S rRNA gene sequence; the 3.4% diver- 1. Greub, G., and D. Raoult. 2003. Rhodobacter massiliensis sp. nov., a new gence including four gap sites would satisfy new species de，- amoebae-resistant species isolated from the nose of a patient. Res. Microbiol. 154:631–635. nition in view of the general acceptance that a 2 to 3% diver- 2. Helsel, L. O., D. Hollis, A. G. Steigerwalt, R. E. Morey, J. Jordan, T. Aye, J. gence connotes a separate bacterial species. Proposal of a new Radosevic, D. Jannat-Khah, D. Thiry, D. R. Lonsway, J. B. Patel, M. I. Haematobacter species can be considered in the future if other Daneshvar, and P. N. Levett. 2007. Identi，cation of “Haematobacter,” a new genus of aerobic Gram-negative rods isolated from clinical specimens, and similar strains are isolated. reclassi，cation of Rhodobacter massiliensis as “Haematobacter massiliensis The MicroSeq 500 system and our additional 16S primers comb. nov.” J. Clin. Microbiol. 45:1238–1243. 3. LaSala, P. R., J. Segal, F. S. Han, J. J. Tarrand, and X. Y. Han. 2007. First would normally amplify and determine a 1,500-bp segment of reported infections caused by three newly described genera in the family the 16S rRNA gene. The 1,425-bp sequence from this Haema- Xanthomonadaceae. J. Clin. Microbiol. 45:641–644. tobacter-like species was shorter than expected. By comparing 4. Li, J. S., D. J. Sexton, N. Mick, R. Nettles, V. G. Fowler, Jr., T. Ryan, T. Bashore, and G. R. Corey. 2000. Proposed modi，cations to the Duke criteria the sequence with the full-length gene of Escherichia coli for the diagnosis of infective endocarditis. Clin. Infect. Dis. 30:633–638. (1,541 bp; J01859), numerous deletions in the Haematobacter 5. Mylonakis, E., and S. B. Calderwood. 2001. Infective endocarditis in adults. N. Engl. J. Med. 345:1318–1330. gene were noted, which would account for the full length being 6. Rintala, H., M. Pitkaranta, M. Toivola, L. Paulin, and A. Nevalainen. 2008. only 1,467 bp. The available lengths from the H. massilensis Diversity and seasonal dynamics of bacterial community in indoor environ- and H. missouriensis genes were also shorter at 1,414 bp and ment. BMC Microbiol. 8:56.