| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Ann Thorac Surg 2004;77:537-543
© 2004 The Society of Thoracic Surgeons
a Cardiothoracic Surgery Service, Brooke Army Medical Center, Fort Sam Houston, Texas, USA
b Department of Pathology, University of Texas Health Science Center at San Antonio, San Antonio, Texas, USA
c Department of Surgery, University of Texas Health Science Center at San Antonio, San Antonio, Texas, USA
d Department of Orthopedics, University of Texas Health Science Center at San Antonio, San Antonio, Texas, USA
Accepted for publication July 8, 2003.
* Address reprint requests to Dr Cohen, Colonel Medical Corps, US Army, 15638 Dawn Crest, San Antonio, TX, USA 78248-1723
e-mail: david_j_cohen{at}hotmail.com
 |
Abstract |
|---|
 Top Abstract IntroductionMaterial and methodsResultsCommentAcknowledgmentsReferences |
|---|
METHODS: Forty New Zealand white rabbits were divided into four groups: group 1, control (1 ml of normal saline); group 2, Corynebacterium matruchotti 100,000 colonies; group 3, Streptococcus sanguis II 10 colonies; and group 4, C matruchotti 100,000 colonies plus S sanguis II 10 colonies. Animals were inoculated with bacteria through a flexible catheter placed through the aortic valve through a right carotid cut down. Inoculations were repeated every 3 days the first 2 weeks and then twice a week thereafter. At postmortem examination the aortic valves were harvested, embedded in paraffin, and stained with von Kossa stain. They were also examined by scanning and transmission electron micrography.
RESULTS: Group 4 had 93.3% large calcifications (confluent calcium densities that are easily recognized with minimal magnification) and 6.6% small microcalcifications (dustlike microscopic particles requiring a compound microscope to appreciate) of the aortic valves. Group 3 exhibited large calcification in 20% and small in 40% of the aortic valves. Group 1 and group 2 had no evidence of calcification.
CONCLUSIONS: These results suggest that recurrent low-grade endocarditis from calcifying oral bacteria, particularly when occurring with synergistic strains, may be one cause of calcific aortic stenosis.
|
Introduction |
|---|
|
TopAbstractIntroductionMaterial and methodsResultsCommentAcknowledgmentsReferences |
|---|
See independent related article, page 704.
Calcific aortic stenosis is a major public health problem. Calcification of the aortic valve occurs in 55% of the elderly and critical aortic stenosis occurs in 2.9% of the population that reaches 75 years of age [1]. The incidence of congenital bicuspid aortic valve in the general population is 2% [2]. These congenitally bicuspid aortic valves will narrow and calcify more frequently and at an even earlier age than tricuspid aortic valves that develop aortic stenosis [3, 4]. Tissue prosthetic valves have also been noted to calcify early through a process that resembles the natural history of bicuspid aortic valves [5].
Bicuspid aortic valves cause turbulent flow. The turbulence of flow distally is responsible for endothelial damage that facilitates the formation of thrombus [6]. This stage is called nonbacterial thrombotic endocarditis (NBTE). Bacteria can adhere to a damaged, fibrin-coated surface more easily than to an intact surface [7].
Interestingly oral microorganisms are among the many microbes that can mineralize. Of these, Corynebacterium matruchotti has been well studied and a small proteolipid has been isolated from it that can initiate calcification in vitro [8–10]. At least two of the oral streptococci, Streptococcus sanguis II and Streptococcus mitis can also calcify. Mineralization generally occurs in these organisms when they die, which would make identification of them difficult in surgical or post mortem valve specimens. An average person houses a reservoir of calcifying bacteria in the oral cavity. Several studies have shown the participation of oral bacteria in transient bacteremias after dental treatment, tooth brushing, and even chewing [11–13]. These bacteremias are usually subclinical but can be demonstrated by blood cultures taken within 15 minutes of the inciting event [11–13].
We suggest that over an extended period, 30 to 40 years, organisms from transient bacteremias can inoculate an existing area of NBTE. These bacteria could be responsible for further damage and calcification. Most patients with calcific aortic stenosis present clinically with symptoms due to advanced valvular stenosis and have no history of endocarditis. Patients in whom calcific aortic stenosis develops may have experienced subclinical (silent) endocarditis with one or more organisms of low virulence from the oral cavity that are capable of calcifying. Repeated seeding of vegetations on a progressively distorted valve by organisms that are known to induce calcification is a credible scenario that deserves inquiry.
The initial step in testing this hypothesis is to demonstrate that bacteria, which are capable of calcifying in the oral cavity, can induce calcification when seeded onto an aortic valve with an injured endothelial surface. That is the focus of this paper.
|
Material and methods |
|---|
|
TopAbstractIntroductionMaterial and methodsResultsCommentAcknowledgmentsReferences |
|---|
Preoperatively and on days 2 and 4 after catheter placement the animals were each given 300,000 IU benzathine penicillin subcutaneously as a prophylactic antibiotic. This perioperative antibiotic treatment was to ensure that no surgical infection would develop that could seed the aortic valve or catheter before the subsequent injection of the experimental inoculum. Indwelling catheters were placed as follows. General anesthesia using a combination of xylazine (6.5 mg/kg), acepromazine (1 mg/kg), and ketamine (32 mg/kg) was administered intramuscularly. The neck was clipped free of hair and the area prepared and draped for aseptic surgery. The animals were placed in the left lateral recumbent position. A 3-cm skin incision was made ventral to the external jugular vein. The sternomastoid and cleidomastoid muscles were separated exposing the carotid artery. The carotid artery was elevated from the sheath and ligated distally. A small arteriotomy was performed proximally; the polyethylene catheter was introduced and passed through the aortic valve into the left ventricle. Correct positioning was determined subjectively by feel and initial attempts were confirmed by roentgenogram. Placement of the catheters was designed to cause mild endothelial damage to the valve leaflets. The catheters were secured in position with silk suture to the carotid artery. The catheters were flushed with heparinized saline and the proximal end tunneled to exit the back of the neck. The surgical incision was closed in layers using 3-0 and 4-0 polygalactin 910 (Vicryl). During recovery nalbuphine was used as analgesic. The investigator or a member of the veterinary staff made daily observations.
Animals that survived the initial surgical procedure were randomly divided into four groups: group A (control) received physiologic saline only; group B received 105 C matruchotti in saline; group C received 101 S sanguis II (note: higher doses were frequently lethal before the end of the experiment); and group D received 105 C matruchotti plus 101 S sanguis II. The experiment was carried out in two rounds. In the first there were 10 animals in each group; in the second round we added another five animals to groups A and D because we deemed these two groups to be the most important (see Table 1).
|
Injection
After a 5-day surgical recovery period, the first bacterial suspensions were injected through the implanted carotid artery catheters. Injections were repeated every 3 days for the first 2 weeks and then twice a week (Monday and Thursday) for the remainder of the study. This dosage schedule was selected because in pilot studies it produced calcifications in aortic valves in group D within 50 days yet allowed most animals to survive until the end of the experiment. After each injection the catheter was flushed with heparinized saline in order to maintain its patency. Animals in which the carotid catheter became clotted received subsequent bacterial injections through the lateral ear vein. Animals were kept until they died or were sacrificed at 50 days.
Pathology
All animals underwent limited gross examination after death including gross examination of the thoracic and abdominal organs. The hearts together with the contiguous aortic roots were removed. The aortic valves were dissected and examined grossly, selectively photographed, low-voltage x-ray films taken with a Faxitron unit, tissue fixed in neutral buffered formalin and processed for paraffin embedding. Valve sections, 6 µ to 8 µ, were stained with hematoxylin and eosin for general study by light microscopy. Von Kossa stains were also prepared to test for calcium phosphate and Gram stains were prepared to test for bacteria.
A serial section from the same paraffin block was mounted on an aluminum planchet and an energy dispersive analytical roentgenogram (EDAX) area mapping was performed for calcium in a scanning elecron microscope (SEM). This tissue section was then coated with gold-palladium, examined by SEM, and the morphology was recorded on film. Another tissue section adjacent to that studied by SEM (above) was deparaffinized and processed for transmission electron microscopy (TEM). The resultant epoxy embedment of tissue, taken just a short distance from the periphery of a large calcification, was sectioned normal to the plane of the paraffin section.
Statistics
In this study the independent variable was the bacterial agent (control, C matruchotti S sanguis II, or both). In this paper "large" calcification refers to the relatively large, confluent densities that were easily recognized with minimal magnification, as opposed to the "small" calcifications that represent the dustlike microscopic particles requiring a compound microscope to resolve them. The dependent variable was calcification (0 = none, 1 = small, 2 = large). The null hypothesis is that there will be no difference in findings as a function of agent. The appropriate test is a Kruskal-Wallis one-way analysis of variance (ANOVA) followed by Mann-Whitney rank sum tests corrected for multiple comparisons. Because six comparisons are possible among four agent groups, a Bonferronni corrected significance level of p = 0.5/6 = 0.0083 should be used. The results are shown in Table 1.
|
Results |
|---|
|
TopAbstractIntroductionMaterial and methodsResultsCommentAcknowledgmentsReferences |
|---|
|
Light microscopy
The tissue beneath these large linear vegetations all showed evidence of injury and necrosis consistent with catheter trauma. Defects in traumatized tissue from all groups of animals were filled with fibrinous thrombus. Consistently valves from only two of the experimental groups (C and D) had colonies of or individual bacteria within the vegetations. Large colonies of bacteria were easily recognizable with hematoxylin and eosin stained sections, as were large nodules of calcification. Calcification was confirmed with von Kossa stain and the presence of bacteria with Gram stain. The size ranges and distribution of the bacterial colonies and the calcified nodules overlapped in adjacent sections prepared either with hematoxylin and eosin, Gram, or von Kossa stains (Fig 2).
Identification of the bacteria by culture was not performed.
|
|
|
Scanning electron microscopy
The general contour of the section matches that of the aortic valve with the sinus of Valsalva filled with thrombus as shown in Figure 2. There is much background noise but the areas of bacterial colonies (hematoxylin and eosin and Gram stains) and calcification demonstrated by light microscopy stand out as lighter areas in approximately the same arrangement as in the stained tissue sections (Fig 5, a).
The distribution of calcium as revealed by the EDAX area mapping for calcium (Fig 5, b and c), confirms our observations made with the light microscope on von Kossa stained material
|
|
The lifespan for all groups had about the same means and ranges but the appearance of calcifications in the valves differed markedly between groups. Neither controls nor animals that received C matruchotti alone developed any aortic valve calcifications. Animals that received S sanguis II alone developed some calcifications but these were predominantly small and sparse. When the two organisms were injected together however valves from every animal were calcified and in only one animal were the calcifications exclusively small. There were significant differences in findings between control and S sanguis II, between control and both, between C matruchotti and S sanguis II, between C matruchotti and both, and between S sanguis II and both. There was no significant difference between control and C matruchotti.
|
Comment |
|---|
|
TopAbstractIntroductionMaterial and methodsResultsCommentAcknowledgmentsReferences |
|---|
The increased incidence of acute and subacute endocarditis in bicuspid aortic valves has been well known since the mid 1800s. This increase has been explained successfully in multiple experimental studies by the postobstruction (aortic valve stenosis) low pressure sink where bacteria would tend to attach to damaged or altered tissue [6]. In our model it would be the aortic side of the narrowed valve. The increased incidence of calcification in bicuspid aortic valves has been an enigma. Several mechanical theories have been postulated but none of them explains why one bicuspid aortic valve becomes calcified and another one remains free of disease [2]. Kim and associates [15] have shown that the calcification process in aortic valve stenosis may be associated with vesicles. These vesicles may be similar to the matrix vesicles which help to initiate calcification in cartilage and bone. However they may also represent cell or bacterial debris.
The role of endocarditis in aortic valve calcification has only been superficially investigated. Moore [16] reported several cases of aortic valve endocarditis with foci of calcification. Reports have implicated cocci as the cause of Shiley pericardial valve infection, calcification, and stenosis [5, 17]. Bacteria with the capability to calcify are present abundantly in the oral flora and enter the blood stream daily during normal daily trauma to mucosal surfaces [11–13]. Some of these bacteria may attach themselves to the aortic valve leaflets. A fulminant clinical endocarditis may not ensue but a low-grade subclinical chronic recurrent infection may occur. This phenomenon might occur because the number of organisms, which enter the blood stream in a transient oral bacteremia, is small and the bacteria are quickly cleared. Furthermore, organisms such as C matruchotti are indolent and unaggressive. These organisms with the ability to calcify could initiate the calcification process. The ability of bacteria to calcify is related to a bacterial proteolipid that can be distinguished from human tissue proteolipids [18]. Such a theory would explain why some bicuspid aortic valves become calcified and others do not because not all persons have the same oral flora.
We used a modified rabbit model for endocarditis. Animals underwent a series of injections through the catheter previously placed in the left ventricle. This technique was chosen to assure that the aortic valves were exposed to the controlled number of bacteria injected. A second option, which was to inject the inoculum through the marginal ear veins, was less desirable as the reticuloendothelial system of the lungs might have cleared a number of microorganisms or resulted in septic pulmonary embolic infarction. In that case we would have no assurance that the aortic valves were actually exposed to a sufficient number of organisms. In our experiments calcification occurred as early as two weeks after initiation of bacterial injections. Consequently this model can only be thought of as an accelerated model. In humans aortic valve calcification occurs over a period of 40 to 80 years.
That a combination of organisms was necessary to cause a high frequency of large aortic valve calcifications has interesting ramifications. Corynebacterium matruchotti, which reputedly is the oral microorganism that calcifies most readily, causes no calcifications to occur when administered by itself [19, 20]. Only when S sanguis II, even in the tiny amounts used, is also injected does calcification consistently occur. The most likely explanation is that S sanguis II can adhere to the damaged valve [21–23] and C matruchotti likely adheres to the attached streptococci. Filamentous organisms and coccoids often form what is known as a "corncob"; the best known of these is the Actinomyces viscosis and S mutans combination that occurs in dental plaque [24]. Streptococci sanguis II by itself can cause some calcifications although they are less frequent and smaller than the calcifications which result from the combination of organisms.
Why bacteria "stick" better to bicuspid valves than to tricuspid valves is not precisely known. It is known that bacteria do not adhere to functionally intact endothelial surfaces and that some sort of injury and thrombus formation is necessary to facilitate adhesion [22]. General dogma suggests that increased turbulence surrounds bicuspid valves in comparison with tricuspid valves. This turbulence could cause injury to the valve surface. While this is a likely explanation there is little direct evidence to support it. In the rabbit endocarditis model that we used valve injury was necessary for endocarditis to occur. Even if injury and thrombus formation is not the predisposing event in humans, recent work [6] has shown that endothelial cells change their characteristics with lateral flow, and flow is certainly altered in the vicinity of bicuspid valves.
Most of the calcification occurred within vegetations. Few calcifications were in solid tissue and when present in solid tissue, they tended to be in nodules of new tissue. Some calcium deposits were nodular and the same size range as ubiquitous colonies of bacteria.
In recent studies of nonrheumatic aortic valve stenosis, Chlamydia pneumoniae was identified as being present by immunohistochemical analysis, electron microscopy, or polymerase chain reaction in 53% of patients with advanced aortic valve stenosis [25, 26]. As in our hypothesis there is a suggestion in this human study that recurrent infection had occurred on these valves even though it was not possible to culture organisms from the specimens.
Finally it is by no means certain that all (or for that matter, any) of the valve calcification that occurs is due to calcification of the microorganisms themselves. Aortic valves can exhibit cartilaginous metaplasia, cartilage formation, and even bone formation. The valve is composed of a mesenchymally derived connective tissue and mesenchymal cells can occasionally be stimulated to redifferentiate into cells, which synthesize and mineralize a matrix. The possibility certainly exists that aortic valve cells themselves are responsible for most of the mineralization. Alternately a small amount of mineralization in the bacteria might trigger the redifferentiation process in the valve cells. As an example of this kind of process, mineralization of epiphyseal cartilage growth plate triggers its remodeling and bone formation.
The hypothesis that minor trauma to the valve endothelium serves as a site for recurrent infection with indolent calcifying oral bacteria is consistent with clinical observations. It would explain why calcific aortic stenosis is more common and occurs earlier in patients with a bicuspid aortic valve rather than a tricuspid aortic valve as the turbulent blood flow associated with bicuspid valves would more commonly cause the minor inciting injury necessary. The hypothesis is also consistent with the fact that not all patients with a bicuspid or tricuspid aortic valve get calcific aortic stenosis since not all individuals have the requisite type of bacteria or synergistic combination of bacteria in their oral flora.
It is likely that there are several mechanisms that result in aortic valvular calcification. Eventually an advanced "end-stage" morphology with scarring and calcification can develop, which masks the etiology. Although this study does not prove that oral microorganisms cause aortic valve calcification in humans it clearly demonstrates that the possibility exists. The study of human aortic valves and their association with the calcifiable bacteria needs to be investigated. If this theory proves to be true, then some cases of calcific aortic stenosis may be preventable by identifying high-risk patients and initiating an early treatment plan designed to change the patient's typical oral bacterial flora.
|
Acknowledgments |
|---|
|
TopAbstractIntroductionMaterial and methodsResultsCommentAcknowledgmentsReferences |
|---|
|
References |
|---|
|
TopAbstractIntroductionMaterial and methodsResultsCommentAcknowledgmentsReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  M. Rattazzi, L. Iop, E. Faggin, E. Bertacco, G. Zoppellaro, I. Baesso, M. Puato, G. Torregrossa, G. P. Fadini, C. Agostini, et al. Clones of Interstitial Cells From Bovine Aortic Valve Exhibit Different Calcifying Potential When Exposed to Endotoxin and Phosphate Arterioscler. Thromb. Vasc. Biol., December 1, 2008; 28(12): 2165 - 2172. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. A. Bratos-Perez, P. L. Sanchez, S. Garcia de Cruz, E. Villacorta, I. F. Palacios, J. M. Fernandez-Fernandez, S. Di Stefano, A. Orduna-Domingo, Y. Carrascal, P. Mota, et al. Association between self-replicating calcifying nanoparticles and aortic stenosis: a possible link to valve calcification Eur. Heart J., February 1, 2008; 29(3): 371 - 376. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| ANN THORAC SURG | ASIAN CARDIOVASC THORAC ANN | EUR J CARDIOTHORAC SURG |
| J THORAC CARDIOVASC SURG | ICVTS | ALL CTSNet JOURNALS |
