Recurrent Bioprosthetic Valve Endocarditis in Intravenous Drug Users

Background

Serratia marcescens is a gram-negative organism that belongs to the Klebsiella-Enterobacter - Serratia group; however, it has multiple strains [1]. Over the last 3 decades, S. marcescens has been reported as a cause of nosocomial infections [2]. The common reservoirs of this microorganism are the respiratory tract, urinary tract, digestive tract, and artificial nails [3,4].

Infective endocarditis (infective endocarditis) caused by S. marcescens is rare and is reported to cause only 0.14% of infective endocarditis cases [5]. S. marcescens infective endocarditis can have a high fatality rate, at approximately 63% [6]. It has been reported in previous case reports that the eradication of S. marcescens is extremely difficult due to highly resistant strains, despite using appropriate antimicrobial agents [7,8].

Among the Enterobacteriaceae group, it is rare that any of the microorganisms in this group is resistant to carbapenems through production of beta-lactamases, but some strains of Serratia have been found to be resistant to carbapenems by 2 mechanisms either through beta-lactamase production or diminished outer membrane permeability and AmpC cephalosporinase production [2]. In the 2 cases presented here, meropenem was used as part of the treatment regimen and thus could have led to recurrence.

We present 2 cases of recurrent S. marcescens right-sided endocarditis in 2 patients with active intravenous (i.v.) drug use. Both patients presented with S. marcescens infective endocarditis within the previous 6 months before being diagnosed again with bioprosthetic valve infective endocarditis due to S. marcescens. We highlight the importance of recurrence and the treatment approach in these 2 cases, along with reported cases in literature.

Case Reports

CASE 1:

A 30-year-old woman with a body mass index of 25 had a significant medical history of active i.v. heroin use for a 5-year period and tricuspid valve replacement due to S. marcescens endocarditis 2 months prior to this incident presentation. The patient presented to the Emergency Department (ED) with fever, chills, shortness of breath, and chest pain. Laboratory test results revealed a WBC count of 20×109/L and a C-reactive protein level of 20 mg/L. She was later admitted to the Intensive Care Unit because of septic shock, as her clinical condition worsened and her systolic blood pressure dropped to 85 mmHg. She also showed signs of confusion and mental disorientation. The patient did not respond to i.v. fluids and required vasopressors. Owing to the recent medical history of infective endocarditis and the patient’s symptoms, blood cultures were taken. A trans-esophageal echocardiogram showed a bio-prosthetic valve in the tricuspid position with thickened leaflets and 1.7×1-cm mobile vegetation and infection (Figure 1).

The patient underwent redo sternotomy, an explant of an old tricuspid valve bioprosthesis, debridement of tricuspid valve annulus, and valve replacement.

The initial antibiotic regimen consisted of gentamicin 3 mg/kg/day divided twice per day and ciprofloxacin 200 mg i.v. twice daily. The patient’s blood cultures again came back positive for S. marcescens. The patient’s antibiotic regimen was switched to meropenem 2 g every 8 h and i.v. vancomycin 2 g every 24 h after the culture results were available. Treatment with i.v. antibiotics was continued for 6 weeks.

CASE 2:

In another very similar case, a 30-year-old female patient with a body mass index of 22 was admitted to the hospital for tricuspid bioprosthetic valve S. marcescens endocarditis after tricuspid valve replacement 5 months prior to her incident presentation. The patient remained healthy during the prior 5-month period after her tricuspid valve replacement with a bio-prosthesis, with no signs or symptoms of infection. However, the patient continued i.v. drug use during this period after the surgery. She had been a cocaine and heroin i.v. drug user for over 3 years. She presented to the ED with pleuritic chest pain and night sweats. On physical examination, her temperature was 38.9°C, and she had edema of the lower extremities and a grade 3/6 systolic murmur without any signs of radiation. Laboratory test results revealed a WBC count of 15×109/L and an erythrocyte sedimentation rate of 40 mm/h. Transesophageal echocardiography (TEE) was performed, and the size of the vegetation on the valve was found to be 0.68×0.87 cm (Figure 2).

Her initial antibiotic regimen consisted of ciprofloxacin 200 mg i.v. twice daily until the blood cultures were reported. The patient’s antibiotic regimen was switched to i.v. meropenem 2 g every 8 h and i.v. vancomycin 2 g every 24 h for 4 weeks based on cultures, and she was eventually transferred to a tertiary cardiovascular surgery center. No signs of valvular failure, no changes in ECG or abscess, or any indications for surgery were seen in the patient; therefore, a second surgery was not needed. The patient had a follow-up after 2 months. Blood cultures were negative, and TEE showed no vegetations on the prosthetic tricuspid valve.

Discussion

S. marcescens infections are associated with i.v. drug abuse and hospitalizations [9]. Cohen et al reported 19 cases of endocarditis caused by S. marcescens in 1980, of which 74% of the patients were i.v. drug abusers and had a prosthetic valve [10]. In other words, the risk for S. marcescens infective endocarditis increases in nosocomial settings and in patients with comorbidities. A previous study on 16 patients infected with Serratia showed that most of patients had chronic debilitating diseases, as well as mutual predisposing factors, including urinary catheterization and mechanical ventilation. Also, patients who have been exposed to broad-spectrum antibiotics or corticosteroids, in addition to prosthetic valves and grafts, were at an increased risk of this infection [1,11].

S. marcescens seems to have a preference for the left side of the heart [10]. Unlike previous reports, both of our patients had recurrent right-sided endocarditis. In our case series, meropenem was chosen based on similar other case treatment protocols in the literature, and vancomycin was used, as the cultures were susceptible to it. The cultures showed resistance to beta-lactams, which explains that the strain causing the infection produces beta-lactamase.

It is possible in our first case that the first infection was not properly treated, as blood cultures were negative after the first infection, but valve cultures were not repeated after treatment had ended, as the patient had a bioprosthetic valve. We believe that the insufficient treatment period of this patient’s previous infection and continued i.v. drug use had contributed the most to the relapse of infection. The patient underwent a TEE after 1 month of finishing the treatment, and her valves showed no signs of vegetation or infection.

Also, recurrent tricuspid valve endocarditis, despite valve replacement for the same infection, warrants investigations to rule out possible distant seeding that is unique to S. marcescens infections. Seeding, as a mechanism of high bacterial resistance, can explain prosthetic valve endocarditis. This also explains that valve replacement for S. marcescens endocarditis in active i.v. drug users should be considered, owing to the poor success rates of eradicating the bacteria once valve replacement is pursued, if performed before completely treating the infection.

A previous study on 19 cases of endocarditis caused by S. marcescens in i.v. drug users that took place between 1969 and 1974 revealed that 14% of endocarditis cases in drug abusers were caused by Serratia, and most patients were treated with an aminoglycoside alone or an aminoglycoside with a beta-lactam or chloramphenicol. Most of the cases’ outcomes were sepsis and death: 13 out of 19 patients died, and 12 out of the 13 died from sepsis [12].

Another report of S. marcescens infective endocarditis in a patient with drug abuse and untreated hepatitis C had a sensitive strain to multiple classes of antibiotics, including fluoroquinolones, beta-lactams, and aminoglycosides. The case had a good outcome and resulted in negative blood and urine cultures of Serratia [6]. It is important to note that Serratia-caused endocarditis is a rare infection. The Infectious Diseases Society of America has still not defined how these cases should be treated [13]; however, endocarditis is mainly treated by a combination of an aminoglycoside or a fluoroquinolone with a beta-lactam for a 6-week period.

Cases of drug abuse-associated left-sided infective endocarditis have a higher fatality rate, of 65.9%, following treatment, compared with a fatality rate of 6.5% for right-sided infective endocarditis after treatment. It has also been reported that surgical treatment has a slightly increased mortality for right-sided infective endocarditis [10]. However, it has been recommended that patients with left-sided S. marcescens endocarditis should be treated with antibiotics along with valve replacement to achieve better outcomes [8].

As shown in Table 1, a selection of 11 case reports of infective endocarditis caused by S. marcescens has been summarized from the literature and includes multiple factors: valve involved, source of infection, definitive antibiotic treatment based on culture sensitivity, and outcome of treatment. A case of bacterial pneumonia resulted in endocarditis causing mild thickening in the mitral valve; the strain was sensitive to sodium colistimethate and cephalothin, and despite using the appropriate treatment, the patient’s condition deteriorated, leading to sepsis and death [14].

The second case shows a patient who underwent multiple teeth extractions, leading to endocarditis involving the mitral valve. She was treated with a combination of 3 antibiotics, chloramphenicol, streptomycin, and penicillin, and the outcome was the same as the previous case, sepsis and death. The third case involved treatment with 4 different antibiotics for a case of bilateral vegetation endocarditis in the mitral and tricuspid valves. Although the strain was sensitive to all of the antimicrobial agents used, the patient’s condition deteriorated and resulted in death. The last case showed successful treatment of endocarditis caused by a strain of S. marcescens that was susceptible to ciprofloxacin; a 6-week antibiotic regimen resulted in negative cultures and no relapse for more than 1 year following treatment.

Eight out of the 12 cases mentioned below resulted in patient death. Half of the cases had vegetations on the mitral valve only, and 1 of them involved both the tricuspid and mitral valves, while 2 cases involved the aortic valve. First-generation cephalosporins were used in 3 cases; penicillin and chloramphenicol were the most used antibiotics. Aminoglycosides were also commonly used. In our 2 cases, the choice of antibiotic treatment was made according to sensitivity and was consistent with previous reported studies.

Another report of a case of mitral valve infective endocarditis caused by teeth extraction was treated with penicillin, streptomycin, and chloramphenicol; however, the outcome was septicemia and death [15]. Another case of mitral valve endocarditis was treated only with meropenem and resulted in death due to brain herniation [16]. Another report described treatment of a case of S. marcescens mitral valve endocarditis with ceftazidime and gentamicin that resulted in negative cultures, but unfortunately led to death due to uncontrolled heart failure.

Recurrent S. marcescens infective endocarditis has been rarely described in the literature. Poyet et al described a 63-year-old patient who presented with 5-time recurrent infective endocarditis due to different types of streptococci [17].

Although very rare, treatment of recurrent S. marcescens endocarditis can be challenging especially in the setting of i.v. drug abuse (Table 1).

It can be concluded that most cases with negative cultures have included cephalosporins and gentamicin in the treatment regimen; therefore, these agents have mostly promising outcomes in the management of infective endocarditis caused by S. marcescens. We also believe that incomplete treatment and ongoing i.v. drug abuse can be major causes for increased recurrence and relapse rates.

It has been reported in the literature that prosthetic valve infective endocarditis is mostly caused by staphylococci, which makes vancomycin an essential agent in the treatment of these patients, as was seen in the cases in this report [18]. Intravenous drug use is one of the common causes of recurrent infective endocarditis, as seen in a case of a man with 3 infections within 7 months; social support, compliance with treatment, and abstaining from i.v. drug use are essential for avoiding recurrence and for successful recovery [19].

Conclusions

S. marcescens is a rare cause of infective endocarditis that is commonly found in i.v. drug users. Newer generations of cephalosporins and gentamicin show promising outcomes in some cases. Recurrent S. marcescens infective endocarditis can be challenging to treat and can be attributed to improper treatment of the first infection, as well as continued i.v. drug abuse, and can require different treatment measures, including the control of infection source.

This case series reinforces that patients who previously had infective endocarditis are at higher risk of recurrence, leading to higher morbidity and mortality; thus, they should be followed up more frequently, especially when it comes to highly resistant organisms, such as S. marcescens.