Determination of site of infection in endocarditis

Infective endocarditis by Klebsiella species: a systematic review

This study aimed to systematically analyze all cases of infective endocarditis (IE) by Klebsiella species in the literature. A systematic review of PubMed, Scopus and Cochrane library (through 27^th January 2021) for studies providing epidemiological, clinical, microbiological as well as treatment data and outcomes of IE by Klebsiella species was performed. In this review, a total of 66 studies were included, providing data for 67 patients. A prosthetic valve was present in 16.4%, while the most common causative pathogen was K. pneumoniae followed by K. oxytoca. The aortic valve was the most commonly infected intracardiac site, followed by the mitral valve. The diagnosis was based on transthoracic echocardiography in 46.2%, while the diagnosis was set at autopsy in 9.2% of included patients. Blood cultures were positive in 93.8%. Fever and sepsis were the most frequent clinical presentations, followed by embolic phenomena, paravalvular abscess, and heart failure. Cephalosporins, aminoglycosides, and carbapenems were the most frequently used antimicrobials. Surgical treatment along with antimicrobials was performed in 37.3% of included patients. Clinical cure was noted in 80.3%, while the overall mortality was 19.4%. Infection at the aortic valve was independently associated with mortality by IE. This systematic review gives a comprehensive description of IE by Klebsiella and provides information on epidemiology, clinical manifestations, therapeutic strategies and their outcomes.

Utility and safety of coronary angiography in patients with acute infective endocarditis who required surgery

OBJECTIVES

To assess the benefit/risk ratio to perform a coronary angiography (CA) before surgery for infective endocarditis (IE).

METHODS

We conducted a single-center prospective registry including 272 patients with acute IE intended for surgery and compared patients who underwent a preoperative CA (n = 160) with those who did not (n = 112). A meta-analysis of 3 observational studies was also conducted and included 551 patients: 342 who underwent a CA and 209 who did not.

RESULTS

In our registry, combined bypass surgery (CABG) was performed in 17% of the patients with preoperative CA. At 2 years, the rate of the primary composite end point (all-cause death, new systemic embolism, stroke, new hemodialysis) was similar in the CA (38%) and no-CA (37%) groups. In-hospital and 2-year individual end points were all similar between groups. There were only 2 episodes of systemic embolism after CA and only one possibly related to a vegetation dislodgement. In the meta-analysis, combined CABG was performed in 18% of the patients with preoperative CA. All-cause death was similar in both groups: odds ratio, 0.98 [0.62-1.53], P = .92. Only 5 cases of systemic embolism possibly related to a vegetation dislodgement were reported. New hemodialysis was numerically more frequent in the CA group: odds ratio, 1.68 [0.79-3.58] (18% vs 14%, P = .18).

CONCLUSIONS

In daily practice, two-thirds of the patients with acute IE who required surgery have a preoperative CA leading to a combined CABG in 18% of the patients. Our results suggest that to perform a preoperative CA in this context is not associated with improved prognosis.

Pneumococcal bacteremia and focal infection in young children

We reviewed a consecutive case series of 178 immunocompetent children aged 3-36 months without central venous lines who had blood cultures positive for Streptococcus pneumoniae by either of paired broth and quantitative culture methods. The incidence of accompanying focal infection was significantly greater in patients with > 10 colony-forming units (cfu)/mL than in patients with < or = 10 cfu/mL (30.4% vs 12.9% respectively, p = 0.04). No significant relationships existed between the magnitude of bacteremia and the age, gender, presenting temperature, interval until the blood culture turned positive, total peripheral blood white cell count, absolute neutrophil count, or absolute band count. Overall, the quantitative method detected 59/178 (33.1%) of the isolates, including five isolates (2.8%) that the broth method failed to detect.

Quantitative blood cultures for diagnosis and management of catheter-related sepsis in pediatric hematology and oncology patients

Paired quantitative blood cultures collected simultaneously via catheter and peripheral vein in Isolator 1.5 ml tubes, were performed in 50 febrile hematology children. Samples were taken to diagnose catheter-related sepsis (CRS) without catheter removal and to monitor the therapeutic efficiency of antimicrobials administered through the infected device by infusion and/or by the antibiotic lock technique (ALT). In 7 children (14%) the colony counts from catheter blood samples were 30-fold higher than the colony counts from peripheral samples, suggesting CRS; in 7 other patients (14%), identical colony counts in both samples suggested sepsis was not catheter-related. One patient (2%) had septicemia caused by E. coli found in the urinary tract; only the peripheral blood cultures were positive. In 6 patients (12%), the Isolator system was not effective for diagnosing bacteremia or CRS; in 29 patients (58%) the febrile episode was not microbiologically documented. All episodes of CRS were cured whatever the treatment was: infusion or ALT.

Quantitative aspects of septicemia

For years, quantitative blood cultures found only limited use as aids in the diagnosis and management of septic patients because the available methods were cumbersome, labor intensive, and practical only for relatively small volumes of blood. The development and subsequent commercial availability of lysis-centrifugation direct plating methods for blood cultures have addressed many of the shortcomings of the older methods. The lysis-centrifugation method has demonstrated good performance relative to broth-based blood culture methods. As a result, quantitative blood cultures have found widespread use in clinical microbiology laboratories. Most episodes of clinical significant bacteremia in adults are characterized by low numbers of bacteria per milliliter of blood. In children, the magnitude of bacteremia is generally much higher, with the highest numbers of bacteria found in the blood of septic neonates. The magnitude of bacteremia correlates with the severity of disease in children and with mortality rates in adults, but other factors play more important roles in determining the patient's outcome. Serial quantitative blood cultures have been used to monitor the in vivo efficacy of antibiotic therapy in patients with slowly resolving sepsis, such as disseminated Mycobacterium avium-M. intracellulare complex infections. Quantitative blood culture methods were used in early studies of bacterial endocarditis, and the results significantly contributed to our understanding of the pathophysiology of this disease. Comparison of paired quantitative blood cultures obtained from a peripheral vein and the central venous catheter has been used to help identify patients with catheter-related sepsis and is the only method that does not require removal of the catheter to establish the diagnosis. Quantitation of bacteria in the blood can also help distinguish contaminated from truly positive blood cultures; however, no quantitative criteria can invariably differentiate contamination from bacteremia.

Limited usefulness of quantitative culture of blood drawn through the device for diagnosis of intravascular-device-related bacteremia

The use of a differential quantitative blood culture technique (Isolator) to diagnose intravascular-device-related bacteremia (IDRB) was studied prospectively. During septic episodes in 44 patients, blood was obtained simultaneously through the suspected infected device and from a peripheral venipuncture. The blood samples were processed by the Isolator technique, which enables easy quantification of microorganisms. The cannula was removed, and its tip was cultured semiquantitatively. Of the 52 cannulas studied, 15 were the cause of IDRB, but only 7 of these showed a significantly higher bacterial count in blood obtained through the device compared with peripheral blood. The bacterial count was higher in blood drawn through the device than in peripheral blood in four of six cases that did not fulfill the definition of IDRB. Some blood cultures obtained through the device were positive despite negative cultures of peripheral blood and cannula tips (six cannulas). Quantitative blood cultures were not useful in diagnosing IDRB in this study.

Evaluation of ceftriaxone and other antibiotics against Escherichia coli, Pseudomonas aeruginosa, and Streptococcus pneumoniae under in vitro conditions simulating those of serious infections

In pursuit of an in vitro system capable of reliably predicting the activities of antibiotics in serious infections and in infections occurring in immunocompromised hosts, we evaluated the abilities of four drugs to achieve virtually complete killing of bacterial cells growing in human body fluids in amounts which are very high and close to those likely to be present in serious infections; drug concentrations varied with time as they vary in human bronchial secretions or blood or urine (dynamic concentrations). The rationale for such a test was (i) to set up in vitro conditions as close as possible to those the antibiotics encounter in serious infections and (ii) to hold the drugs capable of almost completely killing the bacteria used in the assay to be highly active in vitro and likely to be the most efficacious in the treatment of serious infections. Among the antibiotics used, ceftriaxone proved to be highly active under conditions simulating pulmonary infections and septicemias caused by Streptococcus pneumoniae (bacteria grown in bronchoalveolar fluid or blood; antibiotic concentrations varying with time as in human bronchial secretions or blood) and under conditions simulating blood and urinary infections caused by Escherichia coli (bacteria grown in human blood or urine; antibiotic concentrations varying as in the various fluids). Gentamicin (not tested against pneumococci) appeared to be highly active only under conditions simulating urinary infections caused by E. coli; aztreonam (not tested against pneumococci) and ampicillin (tested only against pneumococci) did not appear to be highly active under any of the test conditions. Only the combination of gentamicin plus either ceftriaxone or aztreonam appeared to be highly active under conditions simulating serious septicemias and urinary infections caused by Psudomonas aeruginosa.

Detection of central venous catheter-associated sepsis

To determine whether contamination of the catheter was the cause of pyrexis in patients receiving total parenteral nutrition through central venous catheters, semiquantitative culture of blood drawn through the line kept in place was performed. Eighty-three catheters from 75 patients were studied. The predictive value of a negative culture using this method was greater than 97%.

Clinical correlations of serial quantitative blood cultures determined by lysis-centrifugation in patients with persistent septicemia

The potential clinical value of colony counts determined by the lysis-centrifugation blood culture method was studied by reviewing the records of eight patients with persistent septicemia in whom colony counts were available on at least 3 days. Colony counts of the five patients who survived decreased steadily as the patients improved. One of the three patients who died had counts repeatedly below 1.0 CFU/ml while she was clinically stable and higher counts when her condition deteriorated. Two patients died despite decreasing colony counts. One was improving and died unexpectedly of an unrelated cause; the other died of candidiasis, but declining serial arabinitol/creatinine ratios suggested a partial response to therapy. In addition, septicemia related to infected intravenous catheters was documented by demonstrating large differences in colony counts determined simultaneously from two different sites in two patients and by demonstrating a precipitous drop in CFU per milliliter after removal of the infected catheter in one patient. Routine availability of colony counts appears to be an important advantage of the lysis-centrifugation method.

Quantitative blood cultures in the evaluation of septicemia in children with Broviac catheters

We applied quantitative methods of analysis to all blood cultures drawn during the course of treatment in 28 children with Broviac catheters in a central vein. Thirty febrile episodes in 14 of these patients were evaluated. Samples of blood obtained from a peripheral vein and through the central catheter were cultured quantitatively on agar plates and nonquantitatively in standard broth media. Catheters were judged to be a source of septicemia nine times in seven children. In all nine positive catheter samples, the concentration of pathogens was 10 times as great as that observed in the peripheral venous sample. The blood drawn through the Broviac catheter contained greater than or equal to 2000 colony-forming units per milliliter in six cases. Quantitative cultures in two patients with septicemia not attributable to the catheter yielded low colony counts in the catheter sample. Cultures of blood samples drawn through the catheter when a child was well were not helpful in predicting subsequent septicemia. The technique of inoculating blood directly onto agar plates is easily performed and superior to standard broth cultures, because it detected pathogens within 16 hours and identified infections with multiple organisms.

Methods of quantitative microbiological analyses that support the diagnosis, treatment, and prognosis of human infection

The microorganisms responsible for the production of an infection may be considered to be in two classes: classical microbes and host-defined microbes. Classical microbes are those pathogens which fulfill the Koch-Henle postulates, and their isolation from a host indicates infection. They are not normally part of the body's normal flora, although they may be acquired by the host and enter into a passive relationship known as the carrier state. Examples of this type of microbe are Bacillus anthracis (anthrax), Yersinia pestis (plague), and Bordetella pertussis (whooping cough). Pathogens that require specific hosts have largely replaced the classical pathogen as a cause of infection in hospitalized patients. Especially in recent years, with the advent of new modes of anticancer treatment and the general ability of the medical community to extent a patient's life span by chemotherapy and innovative surgery, the contribution to morbidity and mortality by microbes has substantially increased. These host-specific pathogens are largely part of the body's normal flora. It is incumbent upon the clinical microbiologist to be able to distinguish the patient's normal microbial load, an increased load due to physiological factors, but not representing infection, and a significant change from normal which should be considered infection. The ability to distinguish infection from noninfection is one of the prime responsibilities of the clinical microbiology laboratory and has contributed to the development of the infectious disease subspecialty of internal medicine. This article will examine a critical question: Is there a relationship between the numbers of microorganisms isolated from a specimen and the production of infection, and, if so, does this relationship vary for the different anatomical sites of the body?

Infective endocarditis in heroin addicts: epidemiological observations and some unusual cases

The total number of cases of heroin-induced endocarditis occurring over a four-year period were reviewed in order to explain an increase in the number of cases in the last year studied (1975). Brown heroin was noted to be used more frequently by addicts during the period of increased incidence. Cultures of "street samples" of brown and white heroin as well as cocaine were obtained in order to elucidate a possible relationship between the increased use of brown heroin and the increased number of endocarditis cases. Despite frequent contamination of both white and brown heroin, none of the common endocarditis-causing pathogens were isolated from the samples. Staphylococcus aureus, the most common etiological agent, frequently resulted in tricuspid endocarditis. That the accepted criteria for tricuspid endocarditis may be present without actual cardiac valve involvement is demonstrated by a most unusual case of hepatic vasculature infection.

Role of cardiac catheterization in infective endocarditis

The benefits and hazards of catheterization and angiography were evaluated in 19 patients with acute aortic and/or mitral endocarditis and heart failure. In 14 patients (74%), the bedside diagnosis of valvular insufficiency and heart failure was proved correct. In three patients with both aortic and mitral valve disease, angiography (without hemodynamic measurements) was necessary to clarify the diagnosis. Angiography detected four aortic aneurysmal erosions that were unsuspected clinically, but missed three others. After angiography, heart failure worsened in two patients with severe progressive aortic insufficiency and one died. Thus, catheterization-angiography was of greatest value if more than one left-sided valve lesion was present, if extravalvular diseases mimicked heart failure, or if extravalvular infection was present. Patients with isolated, clear-cut mitral insufficiency usually do not need these diagnostic procedures, and they are probably contraindicated in patients with severe aortic regurgitation with rapidly progressing heart failure.

Experience with valved conduits for repair of congenital cardiac lesions

Dacron valved conduits or aortic allografts were placed between the right heart and the pulmonary artery for repair of various complex congenital cardiac anomalies in 56 patients (aged 15 days to 33 years; median, 11 years). Forty-four patients had a total of 56 previous palliative procedures, which contributed to postoperative morbidity and mortality. Six patients had a total of seven episodes of early or late sepsis involving the conduit. One patient, treated for early sepsis, again developed infection in the Hancock graft 1 year postoperatively and died. Three other patients, 2 with calcified allografts, developed infections 4 months to 7 years following repair and required graft replacement. Hemodynamic data 1 month to 5 years (mean, 1.6 years) following repair revealed mild to moderate obstruction (less than 45 mm Hg gradient) at the Hancock conduit valve ring in 13 of 19 patients, while 5 had large pressure gradients (greater than 75 mm Hg). All aortic allografts had severe obstruction and calcification necessitating graft replacement. It is anticipated that improved technique and appropriate timing of palliative and corrective operations will substantially reduce or eliminate these problems.

Forward angiography in the identification of vegetations in tricuspid endocarditis

A patient with staphylococcal endocarditis of unknown valvular location and resistant to antibiotic therapy was studied in order to localize the site of infection prior to cardiac surger. The injection of contrast material into the right atrium visualized tricuspid vegetations which were confirmed at surgery. In such situations, forward angiographic studies constitute a safe, simple, and potentially diagnostic procedure which avoids the hazards of advancing a catheter across an infected valve.

Endocarditis due to accidental penetrating foreign bodies

A 15 year old boy had an eight month history of recurrent fever, malaise and poor appetite. Chest roentgenogram revealed a foreign object overlying the right ventricle. Multiple blood cultures grew Enterobacter cloacae. The patients condition improved and blood cultures became negative following gentamicin and carbenicillin therapy. E. cloacae was isolated from the foreign body (a finishing nail) at surgery. Antimicrobial therapy was continued for a total of 30 days, and the patient made an uneventful recovery.