Indications for removing intravenous catheters causing SAB are the same, regardless of the microorganism's methicillin susceptibility.123,124 Three prospective observational studies observed that the removal of the CVC (including those without complications) was associated with an earlier clinical response and with a lower rate of relapse.31,125,126

Tunneled CVC (i.e. Hickman® catheters) or surgically placed CVC (i.e. Port-a-cath®) may be kept only when they are highly needed and no easy alternative vascular access is available; when there are no signs of infection in the skin, the tunnel or the surgical incision; on the grounds of hemodynamic stability; and when there are no metastatic distant complications of the SAB. In these circumstances a conservative treatment may be attempted by local (antibiotic-lock) and systemic therapy.127,128 However, failure rates higher than 50% with this conservative approach have been reported in various observational studies.127,129–132 Keeping the CVC responsible for SAB during more than 72h has been identified as an independent risk factor of persistent bacteremia and death.133 Among patients undergoing hemodialysis via CVC with catheter-related MRSA bacteremia, not removing the intravenous line is an independent predictor of complications.134Recommendations:

The presence of inflammatory signs at the site of insertion of any intravenous line responsible for SAB forces the prompt removal of the catheter (A-II). Catheters should be also removed if infection is suspected (presence of catheter and no other obvious focus), and the catheter is easily replaceable (A-II).

A conservative approach to CRB caused by S. aureus should be only attempted in exceptional circumstances (i.e.: absolute impossibility of removing the catheter for technical reasons) and taking into account the clinical and baseline characteristics of the patient (B-II). In these cases, the antibiotic lock therapy must be administered in combination with an effective systemic antimicrobial treatment (B-II). Anyway, the persistence of bacteremia beyond the first 72hours of a conservative management will lead to the immediate removal of the catheter (B-II).

Initial assessment of a patient with SAB should include a detailed history investigating the presence of predisposing factors for IE, prosthetic material, penicillin allergy; possible primary focus, and data of severe sepsis or shock.

Patients with CRB by S. aureus are at high risk of presenting with distant hematogenous complications, especially when the catheter cannot be removed and/or if the antibiotic treatment is not appropriate.134 The risk of developing IE may be higher than 25% in some types of SAB, this leading to prolonged therapy.135–137

Recent observational studies, including all cases from a cohort of patents with SAB, have found that the risk of developing IE is strongly associated with the presence of complications of this infection, the persistence of bacteremia >72h after the onset of adequate therapy, the recurrence of bacteremia or the presence of intracardiac devices.138–140 The systemic performance of transesophageal echocardiography (TEE) to all patients presenting with CRB caused by S. aureus is some controversial, due to the lack of prospective studies specifically addressing this question.141

One of the most important predictive parameters of complicated SAB is the persistence of bacteremia after removing the intravenous catheter.142,143 The treatment of persistent bacteremia is discussed in the next chapter. In these circumstances, it is necessary performing TEE in order to rule out IE.142 Since septic thrombophlebitis may give place to a similar clinical picture, the absence of echocardiographic signs of IE makes it necessary to perform an ultrasound study of the central veins or other vascular accesses where the responsible catheter had been placed, and eventually extend the length of therapy.

In recent years, the use of new molecular imaging methods such as 18F-fluoroglucose (18F-FDG) PET-CT has shown higher sensibility than TEE for the detection of an inflammatory valvular or perivalvular process among patients carrying intracardiac devices, with a very high negative predictive value, and it seems reasonable to recommend its performance in patients with suspected intracardiac device-associated infections when TEE has not been diagnostic.144–146 The usefulness of this technique for the systematic evaluation of patients with SAB has not yet been well defined.Recommendations:

A careful evaluation of the patient's symptoms and an exhaustive clinical examination are essential in cases of catheter related SAB in order to rule out possible sources of the infection. The presence of eventual metastatic septic foci must be identified (B-II).

Blood cultures must be taken after 72h of the onset of appropriate antimicrobial therapy in order to rule out complicated bacteremia (A-II).

Systematically performing TEE to all patients with CRB by S. aureus in order to decide the length of therapy remains controversial. The absence of valvular risk (no valvular disease, neither previous nor diagnosed at the moment of SAB) along with a clinical and microbiological response (negative blood cultures) to therapy within the first 72h after the catheter removal and onset of adequate antibiotics are associated with a favorable outcome (absence of complications or relapse) in more than 95% of patients receiving treatment for at least 14 days after negative blood cultures (B-II).

The length of therapy needs to be adapted to the findings of the TEE or central veins ultrasonography, when indicated (A-II).

The role of new imaging molecular techniques for the diagnosis of intracardiac device-associated infections has not been fully elucidated (C-II).

The treatment of choice for CRB caused by MSSA will be a β-lactamase resistant isoxazolic penicillin, such as intravenous cloxacillin at a dose of 2g/4h.147 In the case of intolerance or allergy to β-lactams, vancomycin may be used. However, this antibiotic is less effective than β-lactams for MSSA bacteremia, the incidence of relapse and persisting bacteremia being higher.88,91,92,148,149

Some recent studies (but not all) have observed a worse prognosis in cases of MSSA bacteremia when vancomycin MIC is ≥1.5mg/L (measured by E-test), regardless of the antibiotic chosen as treatment.8,9,150 Although there are no definitive information in this field., alternative antimicrobial regimes could be considered, such as daptomycin, alone or in combination with β-lactams or fosfomycin, especially in the case that blood cultures remain positive and/or clinical improvement is not evident after catheter removal (C-III).

In the case of MRSA, susceptibility to vancomycin must be considered in the prognosis of bacteremia,65,81 along with the clinical response. In patients with no fever and clinical stability after 24–48h of antibiotic treatment and removal of the catheter, therapy with glycopeptides could be continued regardless of the isolate's MIC to these antibiotics (as long as it is within the susceptibility range). Although there are no controlled studies, in patients with not so favorable clinical response and isolates with low susceptibility to vancomycin (MIC≥1.5mg/L measured by E-test) it seems reasonable to treat with alternative antibiotics.103,151

In this regard and with the available data, daptomycin and linezolid are the two possible candidates to be considered. Daptomycin is a rapidly bactericidal antibiotic against MRSA. In a large, randomized trial including patients with SAB, daptomycin showed a similar efficacy as cloxacillin and vancomycin for bacteremia caused by MSSA and MRSA, respectively.94 In the subgroup of patients with non-complicated SAB, good results were obtained with the currently recommended dose of 6mg/kg/d. Also, the rate of renal toxicity was lower in patients treated with daptomycin as compared with those treated with vancomycin.113

Linezolid is a bacteriostatic antibiotic with activity against MRSA, a 100% bioavailability and good diffusion to tissues (including poorly vascularized areas). Its efficacy has been proven in the setting of pneumonia, bacteremia and severe skin and soft tissue infections. In a meta-analysis of 5 randomized controlled trials in patients with bacteremia caused by MRSA, linezolid was non-inferior to vancomycin.152 A more recent meta-analysis found that linezolid had a higher efficacy against Gram-positive microorganisms, as compared with glycopeptides or β-lactams153; however, poor information on the specific pathogens or sites of infection was provided, thus strong conclusions on the efficacy of linezolid for SAB could not be drawn. In a posterior randomized clinical trial, linezolid proved to have a similar efficacy as comparators for CRB by S. aureus, including MRSA.154 No data are available for treatment with linezolid in patients with SAB due to strains with low sensitivity to vancomycin.Recommendations:

• •

  The treatment of choice of an episode of CRB caused by MSSA is cloxacillin (B-I).

• •

  Alternatively, patients may be treated with daptomycin (A-I) or a glycopeptide (B-II).

• •

  The best antimicrobial treatment in episodes caused by a strain of MSSA with low susceptibility to vancomycin (MIC≥1.5mg/L measured by E-test) has not been elucidated. This panel suggests to use a combination of cloxacillin and daptomycin when blood cultures remain positive and/or clinical improvement is not evident after catheter removal (C-III).

• •

  In the case of CRB caused by MRSA, vancomycin is the treatment of choice (B-II). It may be continued in stable patients with negative blood cultures after 72h of treatment, regardless of the susceptibility of vancomycin (C-III).

• •

  Alternatively, patients may be treated with daptomycin (A-I).

• •

  Linezolid should be only used in patients who cannot take the previous agents (B-II).

The identification of patients with SAB that are evolving favorably – thus may be classified as having non-complicated bacteremia – may be helpful for deciding not to extend the therapy beyond 14 days. This will consider the patient's baseline features, the clinical evolution and the microbiological parameters.

The absence of diabetes, intravascular devices (such as pacemakers or vascular prosthesis), or any immunosuppressant condition (such as neutropenia, or being under therapy with corticoids or other immunosuppressant drugs) is associated with a good prognosis. The removal of the catheter and the absence of complications such as IE, septic thrombophlebitis or any metastatic foci are also associated with a good outcome. Early defervescence and negative blood cultures within the first 72h are also associated with a non-complicated evolution and, all together, may support a treatment not longer than 14 days.31,80,88,134,135,142,155,156Recommendations:

An episode of CRB caused by S. aureus may be considered as non-complicated in the basis of several characteristics of the host (such as absence of diabetes, immunosuppressant conditions and intravascular devices), of the clinical presentation, and of the clinical and microbiological evolution (clearance of bacteremia in less than 3 days of treatment).

The length of therapy for episodes of non-complicated CRB caused by S. aureus has not been well defined by controlled studies. A meta-analysis showed that, after 10–14 days of systemic antimicrobial therapy, the rate of relapse was only 6.1%.157 In more recent series including well-selected cases of non-complicated CRB, the rate of recurrence has been higher in patients treated for less than 14 days, as compared with patients treated for a longer period.155 There is no available information regarding length of therapy in CRB caused by MRSA. However, according to various controlled trials,94,158 it seems reasonable to treat these cases for a similar period of time, as long as the bacteremia is non-complicated and the catheter has been removed. Sequential oral therapy with drugs such as linezolid could be considered in clinically stable patients, with no metastatic complications and with negative blood cultures after the onset of treatment and the removal of the intravenous line.154Recommendations:

Systemic antibiotics in cases of non-complicated CRB caused by S. aureus must be administered for a period not shorter than 14 days (A-II). In patients with favorable clinical and microbiological evolution, sequential oral antibiotics may be considered (A-II).

In patients with no apparent source of SAB an exhaustive anamnesis should be made in order to rule out potential origins of the infection. The presence of permanent devices or foreign bodies, such as catheters, pacemakers, valve prosthesis or orthopedic prostheses must be specifically addressed. Also, patients must be carefully searched for unnoticed skin lesions, or symptoms suggesting distant infectious metastasis (which may occur in up to 30% of cases), such as back pain (indicating vertebral osteomyelitis or epidural abscess), low back pain (in the case of renal or psoas abscesses) and prolonged fever or sweating (suggesting IE). Adequate complementary tests will be performed according to these findings.

The frequency of IE among patients presenting with SAB ranges between 10% and 30%, the higher rate being associated with community-acquired primary SAB or in patients with intracardiac medical devices.142,159,160 Therefore, these patients should undergo an echocardiography. Initially it may be a transthoracic echocardiography (TTE), but if no vegetations are found, the performance of a TEE is recommended.136,161Recommendations:

A careful evaluation of the patient's symptoms and an exhaustive clinical examination are essential in cases of primary SAB in order to rule out possible sources of the infection (C-I). A reliable echocardiographic test should be performed in carriers of intracardiac devices and in cases of community-acquired SAB (A-II).

The basis for the choice of the antimicrobial regime for patients with primary SAB is not different from that of CRB caused by S. aureus (see above). The majority of studies have analyzed heterogeneous samples of patients, including cases with no identifiable source of SAB.8,9,65,81,88,91,92,94,103,113,147–153 Therefore, recommendations on treatment under this circumstance may be the same.

The absence of an identifiable origin of SAB is a risk factor for complication, especially in community-acquired cases. The length of therapy depends on whether the bacteremia is complicated or not. For non-complicated cases, a minimum of 14 days of therapy is recommended in order to avoid relapse of the infection. However, not all authors agree with this opinion, some of them suggesting that community acquired SAB, or cases with no identifiable source should be always considered as complicated.155,162,163

Patients carrying intravascular devices (either valvular or vascular prostheses) should receive a longer treatment (4–6 weeks). Patients with previous valvular heart disease where a TEE (performed after 5–7 days of the onset of bacteremia) has found no vegetations may be treated for 14 days.Recommendations:

Recommendations for the specific definitive antimicrobial treatment for primary SAB do not differ from those of CRB by S. aureus (B-II). The duration of antibiotics should be no shorter than 14 days (B-II). In patients carrying intravascular prostheses, the length of therapy will depend on the findings of the complementary tests performed to discard a secondary involvement of these devices (C-I).

The persistence of positive blood cultures after 72h of appropriate antimicrobial treatment indicates complicated SAB, leading to the repetition of new blood cultures for assessing the clearance of bacteremia. There are no solid recommendations on the frequency for repeating these cultures. However, the persistence of bacteremia beyond 7 days of adequate therapy is an important landmark. When this happens, treatment failure must be considered, as well as the need for changing the antibiotics and double check for the presence of non-drained infectious foci.

The optimal management of complicated SAB includes the administration of appropriate antimicrobial therapy and the identification and drainage of infectious foci, either primary or secondary. The intravenous line catheter is one of the most frequent origins of SAB.

Apart from the intravenous catheter, and given the high frequency of subclinical venous thrombosis in the setting of CRB, some authors support the systematic performance of vein ultrasounds.164 However, this may detect many abnormalities which may be difficult to interpret. A more pragmatic approach is to perform such ultrasound test only in cases of persistent bacteremia. In such cases, the finding of thrombosis is diagnostic of septic thrombophlebitis, leading to a minimum of 4 weeks of antimicrobial treatment.

A frequency of 30–40% of hematogenous seeding of a foreign body in the course of SAB has been reported.165,166 Therefore, patients carrying a device or prosthesis must follow a careful evaluation. The eventual infection of the prosthesis will need to be managed accordingly, the odds of therapeutic failure being higher. In the case of prosthetic valve endocarditis, early surgery is associated with a better prognosis.167

The performance of an 18F-FDG-PET/CT in the context of complicated SAB with neither evident source nor IE may help to find the origin of the infection and/or distant septic metastasis.168

The risk of having IE in the setting of SAB is very high, especially among carriers of prosthetic valves (5–20% of all episodes).142,169 Some studies advocate for the performance of TEE to all patients with SAB (either complicated or not), arguing that it is cost-benefit in terms of length of therapy,170 while some groups prefer a more conservative approach, and indicate TEE in the case of persistent bacteremia or in carriers of intracardiac devices.138Recommendations:

Blood cultures must be repeated every 72hours in order to monitor the microbiological response to antibiotic therapy (A-II). Make it sure that an intravenous catheter left in place is not the origin of the persistent bacteremia (A-II). When a foreign-body (i.e. prosthetic joints or prosthetic valves) becomes infected, the indication of surgery for debridement and/or removing the device must be considered (A-II). It is necessary to perform an echocardiography to all patients with complicated SAB. In patients carrying an intracardiac device or in those with persistent bacteremia, a TEE is preferable (A-II).

Some studies published decades ago, including intravenous drug users, observed a 1-day reduction in the duration of bacteremia with combined therapy. However, the addition of an aminoglycoside in the first days of therapy had no impact on mortality, and was significantly associated with renal toxicity.111,113,173

The good intracellular and tissue diffusion of rifampin could potentially lead to a faster control of complicated bacteremia. However, there is no clinical evidence supporting that the combination with rifampin is more active than β-lactam monotherapy.122 Notwithstanding, an ongoing randomized study will try to confirm whether the addition of rifampin to the standard treatment of SAB is able to decrease the mortality.174Recommendations:

The treatment of choice for complicated bacteremia caused by MSSA is cloxacillin, either 2g every 4h, or administered in continuous infusion (A-I). Combined therapy is recommended in the following scenarios: (1) persistence of fever; lack of improvement of signs and symptoms (B-III); (2) microbiological failure detected by the positivity of subsequent blood cultures, especially in episodes by an isolate with vancomycin MIC≥1.5mg/L (measured by E-test). The possible options for combined therapy are: (1) cloxacillin 2g/4h iv+daptomycin 10mg/kg/d iv; (2) cloxacillin 2g/4h iv+fosfomycin 2g/6h iv (A-III). The length of therapy in complicated bacteremia is variable, ranging between 4 and 6 weeks according to the clinical evolution and the source of infection. The length of combined therapy is not established, but it seems reasonable to maintain at least it until blood cultures became negative.

Vancomycin MIC is a key parameter for the efficacy of this antibiotic in MRSA complicated bacteremia, the ratio AUC/MIC over 400 being associated with optimal antimicrobial activity. When the MIC value is ≥1.5mg/L, this PK/PD goal may not be easily achieved with the standard dose of vancomycin (15mg/kg/12h), this probably leading to treatment failure, especially in complicated bacteremia.65,177 In order to achieve this AUC/MIC ratio, doses of vancomycin may be increased so the trough levels are over 15mg/L. However, this has been associated with higher renal toxicity, especially in long treatments.178 Therefore, this strategy is no longer recommended.

Importantly, it has been observed that the vancomycin MIC creeping is associated with a higher likelihood of treatment failure, regardless of the method for determining the MIC value and also the antibiotic treatment administered.43 While this has been questioned by a recent meta-analysis,7 this would suggest that MRSA isolates with low susceptibility to vancomycin may associate other unknown virulence parameters which may lead to treatment failure.179

As previously discussed, daptomycin possesses a faster bactericidal effect and less toxicity than vancomycin. The efficacy is similar,94 or even superior when vancomycin MIC is higher than 1.5mg/L (measured by E-test).103,151,180 Doses of daptomycin have been increased since its approval (from 6 to 8–10mg/kg/d) in order to avoid the emergence of resistance,181 and to reduce the mortality, although the latter point has not been clearly proven.6,182

Several in vitro studies have observed synergy with the combination of daptomycin plus β-lactams against MRSA,183,184 including VISA isolates. In addition, the combination may avoid the emergence of resistance185 and also increase the activity of β-lactams by a mechanism which is independent from the gene mecA (seesaw effect) in isolates with low susceptibility to daptomycin.186

Clinical experience with this combination is scarce. In a study including 7 consecutive patients with persistent MRSA bacteremia who had been previously treated with daptomycin or vancomycin, the combination of daptomycin (8–10mg/kg/d) plus oxacillin or nafcillin (2g/4h) led to a rapid clearance of bacteremia in all seven cases. In vitro studies performed on 3 isolates from this study showed an enhanced membrane daptomycin binding and a higher bactericidal activity.117 In patients registered in the Cubicin® Outcomes Registry and Experience (CORE, 2005–09), with mild or medium renal impairment and treated with daptomycin for SAB (mainly caused by MRSA), a trend toward a better outcome was observed among patients that were also treated with β-lactams.172 Thus, this combination may be more effective than monotherapy for complicated bacteremia caused by MRSA. However, more studies supporting this hypothesis are necessary, and also defining which β-lactams are synergistic and what doses should be used.

Fosfomycin is a bactericidal antibiotic with activity in the early steps of the cell wall peptidoglycan synthesis. Its unique mechanism of action makes cross-resistance with other antibiotics very rare, most isolates of MRSA being susceptible. However, fosfomycin must be administered in combination with a second drug in order to avoid the rapid development of resistance.

The clinical experience with the combination daptomycin-fosfomycin is limited. Notwithstanding, in vitro studies have proven synergy with this combination.118 In addition, avoidance or delay of daptomycin-resistance emergence has been observed.185 Various experimental models have confirmed the synergy of the combination, one of the first being a model of experimental enterococcal endocarditis.187 More recently, a model of experimental foreign-body infection caused by MRSA has also proven the activity of the combination.188

Clinical experience with daptomycin plus fosfomycin is still anecdotal. The most important experience included three patients with left-side endocarditis (one by MSSA on a prosthetic valve and two by MRSA) treated with daptomycin (10mg/kg/d) plus fosfomycin (2g/6h); a surgical management was considered, but outcome was favorable for the three patients with only medical treatment.118 Another case of IE caused by daptomycin-resistant MRSA was treated with daptomycin (12mg/kg/d) plus fosfomycin (6g/6h), also with good results.189 It must be stressed that the intravenous administration of fosfomycin includes a high sodium concentration, therefore caution is recommended in patients with liver cirrhosis or heart failure.

As with the combination with β-lactams, daptomycin plus fosfomycin seems to be promising in the management of MRSA complicated bacteremia, but again further prospective studies proving the superiority of this treatment are necessary.

The mechanism of synergy of this combination is not fully understood. The early inhibition of the cell wall synthesis due to fosfomycin's activity may produce impairment in PBP2a. The wall synthesis would depend again on the activity of PBP2, and therefore the bacteria would become susceptible to imipenem. Various in vitro, experimental studies and clinical experience support the efficacy of the combination of fosfomycin and β-lactams,190–192 and synergy has been observed for the combination with imipenem.193 However, clinical experience is still very scarce. In a Spanish multicentre study, fosfomycin-imipenem was administered to 16 patients as salvage therapy. Clearance of bacteremia was observed within 72h in all patients.194 This supports the use of this combination as salvage therapy for patients with complicated bacteremia caused by MRSA.Recommendations:

The best treatment for complicated MRSA bacteremia has not been elucidated. The treatment with vancomycin is associated with a high rate of treatment failure, especially in the following situations: (1) if vancomycin MIC≥1.5mg/L (measured by E-test) (A-II); (2) if the patient has renal impairment or is at risk of renal toxicity (A-II).

Doses of 6mg/kg/d of daptomycin have been associated with treatment failure and emergence of resistance. Daptomycin at doses of 10mg/kg/d is the treatment of choice for MRSA complicated bacteremia (A-III).

Patients with persistent bacteremia or severe sepsis or shock in the setting of treatment with high doses of daptomycin may benefit from combined therapy. The options are: (1) daptomycin (10mg/kg/d)+fosfomycin (2g/6h) (A-III); (2) daptomycin (10mg/kg/d)+cloxacillin 2g/4h (A-III); imipenem (1g/6h) plus fosfomycin (2g/6h) (A-III).

The administration of high doses of fosfomycin may lead to sodium overload and hypokalemia (1g of fosfomycin-disodium carries 13.5mEq [330mg] of Na). The duration of treatment for complicated bacteremia is variable, ranging from 4 to 6 weeks, depending on the clinical evolution and the source of the infection.

S. aureus is the most frequent etiology of IE worldwide, both in native valve endocarditis and in early prosthetic valve endocarditis, as well as in the infection of intracardiac devices.213 In the study by Murdoch and cols.213 including 2781 cases diagnosed from 2000 to 2005 in several countries from the five continents, S. aureus was the cause in 31%, it being methicillin-resistant in 27%. More recent studies report increasing rates of staphylococcal IE,214 despite the decrease of cases in intravenous drug users (IVDUs). The increasing rate is probably in relation with more prevalent health-care associated cases,215 implantation of intracardiac devices, patients in hemodialysis, diabetes mellitus or MRSA skin colonization.80

In spite of the increasing incidence of IE caused by S. aureus, therapeutical innovations and the indication for surgery have barely improved the prognosis of this infection over the last 25 years, current mortality rates being over 20% and around 40% for MSSA and MRSA IE, respectively.80,216 Surgery is performed in 25 to 35% of cases, with no differences between MSSA and MRSA episodes.80,216 This is observed in many old and recent studies, where S. aureus has proven to be an independent risk factor for mortality in left side IE.80,213 Therefore, providing appropriate antimicrobial treatment during the first 24–48h is key in the approach to patients with suspected IE caused by S. aureus, until definitive identification of the etiology and its antibiotic susceptibility profile are available. Empirical treatment recommendations for IE are summarized in Table 2.Recommendations:

The empirical antimicrobial treatment of an episode of complicated bacteremia or IE should include S. aureus whenever there are reasonable doubts on its potential role as etiology, given its high and increasing incidence and severity.

Therefore, active antibiotics against S. aureus should be included in the empirical treatment in the following cases: suspicion of community-acquired IE (either in IVDUs or not); suspicion of acute IE or presenting with severe sepis (B-II); and early IE, associated to pacemakers or defibrillators (B-II), or in nosocomial cases or in health-care associated cases (B-II).

To date, the clinical evidence suggests that the appropriateness of the empirical antibiotic treatment has an important impact in the prognosis of complicated SAB.6 In this context, the choice of the most appropriate empirical antibiotics may be guided by three fundamental principles: (i) patients with a clinical presentation strongly suggesting IE and having particular epidemiological and/or clinical characteristics making S. aureus a probable etiology, the results of blood cultures not being available at the time of starting the antibiotics; (ii) patients with a clinical presentation strongly suggesting IE and blood cultures being positive for Staphylococcus spp, the species and the susceptibility to methicillin still unknown; (iii) SAB in a patient with risk factors for IE, the susceptibility to methicillin still unknown.

In the setting of IE, the clinical features suggesting S. aureus as the etiology include acute clinical presentation, with severe sepsis or septic shock, the presence of major embolic events (i.e., ictus), a new regurgitating murmur and, in the case of right side IE, the presence of pulmonary emboli.80,216 In the setting of SAB, the features associated with a higher likelihood of having IE are: community-acquired primary SAB; SAB of unknown source; hospital-acquired SAB; persistent fever and bacteremia; the presence of acute skin lesions; previous valve diseases or the presence of prosthetic valves; IVDUs; and previous episodes of IE.142,217 In a study including 736 patients with hospital-acquired SAB (mainly CRB episodes), independent risk factors for IE were the presence of a prosthetic valve or other intracardiac devices, and prolonged bacteremia.138 In the absence of persistent bacteremia, intracardiac devices, hemodialysis or osteomyelitis, the likelihood of IE is very low (negative predictive value near 100%), therefore TEE would not be necessary.138Recommendations:

In the setting of SAB, it is recommended considering the diagnosis of IE until it has been ruled out by complementary tests (namely TEE) in the following scenarios: community-acquired episodes (B-II); IVDUs (B-II); presence of skin lesions suggesting hematogenous seeding (B-II); and nosocomial bacteremia in the presence of prosthetic valves or intracardiac devices (B-II).

Risk factors for MRSA have been previously discussed. To date, there are few papers addressing the likelihood of MRSA as etiology of staphylococcal IE. In a recent study including 72 episodes of IE caused by S. aureus, 22% were caused by MRSA, the risk factors being: nosocomial origin, surgical procedures within the previous 6 months, surgical wound infection and the presence of an intravenous catheter.218-222 In the prospective International Collaboration on Endocarditis (ICE) study, including 424 episodes of IE caused by S. aureus, 33% of cases were due to MRSA and were significantly associated with diabetes, chronic immunosuppressant therapy, cancer, acquisition in a health-care environment, presence of intravascular devices, and recent invasive procedures.80

MSSA is more frequent in community acquired SAB with no apparent source (>90%)80 and in right side native valve IE among IVDUs. However, except for rare cases, the clinical presentation is rarely able to predict whether the etiology of IE will be MSSA or MRSA. Persistent bacteremia is more frequently caused by MRSA.80,218Recommendations:

The empirical antimicrobial treatment for IE should include activity against MRSA in any of the following instances: nosocomial cases (B-II), previous nasal or skin colonization by MRSA (B-II); patients from nursing-homes (B-II) or in hemodialysis (B-II), surgical procedure within the 6 months preceding the bacteremia (B-II), or the presence of certain baseline conditions (diabetes, cancer, immunosuppressant therapy) (B-II).

From a practical point of view, the previous considerations may be simplified according to the place of acquisition of the infection. For community-acquired cases, empirical treatment should focus MSSA, the episode frequently involving a native valve. However, if the infection has been acquired in the hospital (or in a health-care related environment), empirical treatment should also include MRSA, and prosthetic valves (fundamentally early prosthetic valve IE) or intracardiac devices may be involved.

As previously discussed, the treatment of choice for IE caused by MSSA are anti-staphylococcal penicillins, such as cloxacillin or nafcillin.85,112,226 Cefazolin's activity may be impaired in high-inoculum infection such as staphylococcal IE, so caution in this setting is required.85 Notwithstanding, efficacy was similar in a case–control study including 49 patients treated with cefazolin and 84 with nafcillin, tolerance being higher for cefazolin.149 In addition to these β-lactam antibiotics, some experts support the addition of daptomycin in severely ill patients, in accordance with the excellent results reported in small case series and in experimental studies.117,183Recommendations:

When community-acquired IE caused by S. aureus is suspected, the treatment of choice is cloxacillin (B-II).

In critically ill patients, or in patients with severe sepsis or septic shock, many experts recommend adding daptomycin to the treatment with cloxacillin (C-III).

Patients allergic to β-lactams may be treated with cefazolin (if no previous anaphylaxis has been reported) (B-II), or with the combination of daptomycin plus fosfomycin (C-III).

Current guidelines recommend adding gentamicin to cloxacillin for the treatment of native valve IE caused by MSSA.112,226 As previously discussed, there is scarce clinical evidence proving a better activity of this combination, while some studies have proved that this increases the risk of renal toxicity at the end of the episode, even when the aminoglycoside is administered at low doses.113Recommendation:

The addition of gentamicin in the empirical treatment of community-acquired IE caused by S. aureus during the first 3–5 days is not recommended (D-I).

In this setting, the treatment should include activity against MRSA. If information on blood cultures yielding S. aureus is available, the goal is to provide effective antibiotic treatment for both MSSA and MRSA, taking into account the considerations previously discussed. If no blood cultures have been taken or are not available (i.e. fever and new valve prosthetic dehiscence in the post-operative period), antibiotics with activity against Gram-negative microorganisms should also be included in the treatment. As previously mentioned, experimental studies on the combination of daptomycin and β-lactams (cloxacillin, nafcillin, ampicillin, carbapenems, ceftriaxone or piperacillin/tazobactam) also observed that synergy occurred with a wide range of these drugs, and also with fosfomycin.118,183Recommendations:

In the setting of health-care related IE caused by S. aureus, monotherapy with vancomycin nor daptomycin are not recommended (D-II).

In this context, daptomycin in combination with cloxacillin is recommended (B-II). For patients allergic to β-lactams, cloxacillin may be substituted by fosfomycin (C-III).

Faced a suspected case of IE but no available blood cultures, the use of daptomycin in combination with a β-lactam with activity against nosocomial Gram-negative microorganisms is recommended (C-III).

The studies proving a higher efficacy of cloxacillin, nafcillin and cefazolin over other β-lactams have already been referred.85,149 Clinical evidence of daptomycin's efficacy for IE caused by MSSA is also scarce, results being non-conclusive.94 However, we know that patients with left valve IE caused by MSSA with vancomycin MIC≥1.5mg/L (measured by E-test) have higher mortality, this leading to the search for alternative treatments.171 In IVDUs with right valve IE, with no pulmonary emboli and with negative blood cultures after 72h of treatment, a two-week course of cloxacillin may be enough, gentamicin adding no benefits.114 The standard length of therapy recommended for left valve non-complicated IE is 4 weeks, or 6 weeks in the case of pulmonary emboli or periannular complications.7,112,226Recommendations:

For native valve left side IE caused by MSSA, cloxacillin for 4–6 weeks is recommended (B-II), and two weeks in non-complicated right valve IE among IVDUs (A-I). Daptomycin may be added to cloxacillin in the case of persistent bacteremia detected by the positivity of subsequent blood cultures, especially in episodes caused by an isolate with vancomycin MIC≥1.5mg/L (measured by E-test) (C-III). Systemic combination with gentamicin is not recommended (D-II). In patients allergic to β-lactams, the combination of daptomycin to fosfomycin is recommended (C-III).

In addition to previous considerations, in prosthetic valve IE caused by MSSA, the length of therapy will be at least 6 weeks, due to the difficulties for eradicating the biofilm on the surface of the prosthesis.112,226-234 Also, according with a recent study on IE caused by Enterococcus faecalis, gentamicin may be added to the treatment on a once-daily dose during the first two weeks.112,235Recommendations:

Cloxacillin is recommended in prosthetic IE caused by MSSA (C-II), in association with rifampin after the first 5 days of treatment (C-III), and gentamicin in a once-daily dose during the first two weeks of therapy.

In the case of allergy to β-lactams, the same combination of antibiotics may be used, with the substitution of cloxacillin by daptomycin (C-III).

Some available evidence has proven a bad prognosis for vancomycin-treated complicated bacteremia caused by MRSA, when vancomycin MIC is higher than 1.5mg/, although a recent meta-analysis has failed to prove so.7 Also, some studies have shown a higher efficacy of daptomycin as compared with vancomycin in complicated bacteremia caused by MRSA when vancomycin MIC is higher than 1.5mg/L.103,151 Therefore, daptomycin will be the treatment of choice for these cases. Also, there is scarce but excellent clinical results of the synergistic combination of daptomycin plus β-lactams or fosfomycin against MRSA isolates causing bloodstream infection and IE. These results have been also observed in experimental studies.117,118,183 Overall, this evidence supports the priorization of these combinations in the setting of native valve IE caused by MRSA.Recommendations:

Daptomycin plus cloxacillin is recommended in native valve IE caused by MRSA when vancomycin MIC is≥1.5mg/L (B-II).

The same treatment may be administered when vancomycin MIC is <1.5mg/L (measured by E-test), or vancomycin at doses providing trough levels of 15–20mg/L (B-II).

In patients allergic to β-lactams, the combination of daptomycin plus fosfomycin is recommended (B-II), or the use of vancomycin at doses providing trough levels of 15–20mg/L (B-II).

Neither the addition of rifampin (D-III) or gentamicin (D-III) to the treatment are recommended.

The need for avoiding vancomycin when its MIC is high places daptomycin in the first line. Also, there is no available evidence supporting the use of alternative agents with good activity against biofilm different from rifampin or gentamicin, so its use is still recommended.Recommendations:

In prosthetic valve IE caused by MRSA with vancomycin MIC≥1.5mg/L (measured by E-test), the use of daptomycin, in combination with rifampin after 5 days of treatment, and gentamicin in one single daily dose during the first two weeks of therapy is recommended (C-III). Daptomycin plus fosfomycin could be used alternatively (C-III).

In the case of MIC<1.5mg/L (using E-test), the same combination may be used (C-III), or vancomycin combined with rifampin after 5 days of treatment, plus gentamicin in a one single daily dose during the first two weeks of treatment (B-II).

As previously mentioned, there is very scarce evidence supporting the use of linezolid or ceftaroline for IE caused by MRSA.200,224,225 In a recently published multicentric study, the combination of imipenem plus fosfomycin was used as salvage therapy for MRSA bloodstream infection (12 patients with IE, 2 with vascular graft infection and 2 with complicated bacteremia). Cure rate was 69%, and only 1 in 5 deaths was associated with MRSA infection. All patients had negative blood cultures within the first 72h after starting this combination, which showed to be safe, except for a cirrhotic patient with ascites secondary to fosfomycin-related sodium overload.194

One of daptomycin's drawbacks in the treatment of IE is the emergence of resistance. To avoid this, the combination of imipenem plus fosfomycin, or the use of linezolid seems appropriate. More recently, the combination of daptomycin plus ceftaroline has shown good activity in vitro,223,236 and excellent but scarce clinical results against daptomycin-resistant MRSA isolates.237 In specific clinical scenarios, daptomycin plus linezolid could be a good choice for IE caused by MRSA presenting with pneumonia or meningitis. In the latter, daptomycin plus ceftaroline might also have a better efficacy than either monotherapy.Recommendations:

In patients with IE caused by MRSA presenting clinical failure with previous recommended schedules, the administration of daptomycin plus fosfomycin may be used (B-II). Fosfomycin plus imipenem could be also used (C-II). If this cannot be done, either because of allergy or a high risk of sodium overload, ceftaroline, either alone (B-II) or combined with daptomycin (C-II), or linezolid, alone (C-II) or associated with daptomycin (C-III), may be valid alternatives.

Persistent bacteremia by MRSA is recognized in international guidelines as the only microbiological criteria indicating per se the need for surgery.238 Persistent positive blood cultures must lead to eradication of primary (valves) or metastatic infectious foci in order to ameliorate the prognosis. In the European guidelines, persistent bacteremia is defined as positive blood cultures after 10 days of appropriate antimicrobial treatment.226 However, it has been recently proven that the prognosis of left side IE worsens if blood cultures still yield microorganisms 48–72h after the onset of treatment, mortality rate being twice as high.239 Therefore, a case with persistent bacteremia (positive blood cultures in the third day of appropriate treatment) in the setting of IE caused by MRSA should be considered for surgical management, even if there are no other features indicating so.Recommendations:

Patients suffering from IE caused by S. aureus share the same indications for surgery as other cases due to other microorganisms, with the exception of prolonged MRSA bacteremia. Therefore, international guidelines may be followed (A-II), but if blood cultures after 72h from the onset of appropriate treatment still yield MRSA, complementary tests should be performed in order to rule out metastatic foci, and the cardiac surgeons should contacted (B-II).

There are few data regarding this question, and none specific for IE caused by S. aureus. Classically, the guidelines have recommended an arbitrary period of 4–6 weeks after surgery in order to avoid relapse, even if valve cultures were negative.240 Alternatively, a complete course of prosthetic valve IE was recommended for cases of operated native valve IE, subtracting the days of antibiotic treatment received before the surgery.112 However, a study involving 358 cases of IE submitted to valve surgery proved that, in cases with negative valve cultures, two weeks of treatment after surgery, or completing the initial scheduled treatment, were sufficient.241 The most appropriate length of therapy in patients with positive valve cultures is uncertain, current guidelines recommending to complete a whole treatment after the surgery.112,226Recommendations:

In patients with native or prosthetic valve IE caused by S. aureus undergoing valve replacement and cultures being negative, it is recommended to administer two more weeks of therapy or simply finish the initially scheduled treatment (B-II).

In patients with positive valve cultures after surgery, it is recommended to restart the treatment of IE (i.e., ≥4 weeks for native valve IE, and ≥6 weeks for prosthetic IE) (C-III).

Adherence to quality-of-care indicators in the management of SAB is frequently lower than desired.31,34,246 General interventions performed in any patients with bacteremia usually include early reporting of preliminary results of blood cultures (i.e., results of Gram stain), management advice including antimicrobial therapy, and control of the source of infection.242

Apart from those, interventions specifically addressed at improving the management of SAB have been based on infectious disease specialist (IDS) consultation, either only when solicited or routinely offered to physicians in charge of the patients.31,33,36,243,245,246 These studies included comparative parallel cohorts31,33,244–246 or a pre-post desing,36,243 but they all showed improvement after IDS consultation and even lower mortality, too.

However, the recommendations provided by IDS were not structured. A recent study investigated the impact of a protocolized intervention based on IDS active (non-solicited) consultation comprising 6 quality-of-care indicators (see above).34 The intervention consisted on an combined activity performed by clinical microbiologists and IDS who provided specific recommendations in a structured form with a bundle of measures; the form was included in the patients’ chart so the physician in charge could consult it, and it was updated 3 days per week; this activity had been previously informed to all hospital departments, and was associated with improved adherence to quality-of-care indicators and reduced mortality.Recommendations:

Active, unsolicited IDS consultation for management and follow-up should be provided to physicians in charge of all patients with SAB (BII). The specialized recommendations to physicians in charge of the patients with SAB should be provided in a structured manner so all quality-of-care indicators of the management are considered (BII).