In prosthetic valve endocarditis, would you ever use a cephalosporin with an aminoglycoside for “synergy”? Or would cephalosporin alone be sufficient in the setting of a pan-sensitive Strep.?
Adding gentamicin for 2 weeks for PVE caused by pan-sensitive viridans group strep or strep bovis is a great area of controversy. As you know, the IDSA guidelines state that the treatment in such a case can be with OR without gentamicin for 2 weeks (Table 6), as do the UK guidelines. The European guidelines do not recommend such addition. IDSA does not provide the strength of recommendation for adding gentamicin, nor references, and combined with the lack of this recommendation in other guidelines implies paucity in the evidence to support this practice.
Gentamicin added to any beta-lactam exhibits synergy for strep in any situation, and many studies that have observed synergy, or at least additive effects in vitro. Mechanistically, when the cell-wall agent inhibits further synthesis of the peptidoglycan layer, halting it’s growth, the osmotic stability of the bacteria against the serum is disrupted. Once this occurs, the bacteria dies essentially by being exposed to an environment that is too hypertonic for it to function. However, the internal mechanism of the bacteria such as protein synthesis and nuclear division may not halt immediately. This is where the gentamicin comes in. Although otherwise impenetrable to amingolycosides, the now broken cell wall allows gentamicin to enter and exhibit activity against 30S ribosomal subunit, leading to a misread genetic code and inhibited translocation. As you can see, the two antibiotics truly work together, leading to shorter time to kill and faster sterilization as shown in laboratory simulations.
The question is how relevant is this clinically? Streptococcus that is fully susceptible to beta-lactams literally melts away, and the concern for sterilization of prosthetic material is addressed with extending the duration of therapy to 6 weeks rather than with additional antimicrobials. Unfortunately the two strategies have not been compared head-to-head in a robust way. In my literature search, I found only one study that is particularly relevant. This RCT of 61 patients with both native and PV IE that compared the efficacy of monotherapy with ceftriaxone 2g IV once daily with combination therapy with ceftriaxone 2g IV once daily and 3 mg of gentamicin/kg IV once daily for the initial 2 weeks as therapy for endocarditis due to penicillin-susceptible streptococci. Clinical cure was observed for in 96.2% of monotherapy recipients and 96% of combination therapy recipients and no patient had evidence of relapse. 27.5% required cardiac surgery after initiation of treatment, including five monotherapy recipients and nine combination therapy recipients. The study concluded that the regimes are deemed equivalent and the gentamicin addition may be redundant. However, this is a small trial and more than half of the patients had native valves. [Clinical Infectious Diseases. 27(6):1470-4, 1998 Dec]
In addition, a Japanese study looked at 31 patients who had definite or possible infectious endocarditis due to penicillin sensitive streptococcus. Of these patients, 27 were treated with ceftriaxone plus gentamycin combination therapy and four with ceftriaxone monotherapy and all patients achieved clinical cure. This study wanted to make point however that ceftriaxone is an alternative beta-lactam to penicillin and not so much to show that gentamicin may not be necessary. [J Infect Chemother. 16(3):186-92, 2010 Jun]
As you can see, the studies are not exceptionally rigorous and good evidence is lacking. Here are some practical tips for managing patients:
1. When streptococcus is identified, most patients are started on vancomycin empirically in case the bug is highly resistant (MIC over 2). Although resistance is considered significant for MICs greater than 0.125, usually a beta-lactam can still be used with gentamicin, up to an MIC of 2, where vancomycin is usually preferred.
2. Once you have your susceptibility, switch your patients to combo therapy if the MICs are between 0.125 and 2 or if the strep is a nutritionally variant type (e.g. granulicatella), regardless of the MIC. These are harder to kill and should be treated with combination therapy. Gentamicin here is used for 2 weeks.
3. If the strain is actually fully susceptible, I would treat with gentamicin as a combination while the patient is being investigated. I would do this even in a patient with a native valve, but only if surgery and prosthetic valve placement appears likely. These next 48-72 hours can reveal a great deal about the severity of infection, the damage done and the patient’s clinical status. In addition, a few days of gentamicin is not likely to cause much toxicity.
4. I would use 3mg/kg once daily, and not divided q8h, as toxicity is lower with this regimen.
5. I would continue gentamicin for 2 weeks in a case of fully susceptible strain if the damage to the current valve was so bad that it needed to be completely replaced, there was very large vegetations present (large bacterial burden), bacteremia is persistent despite 48-72 hours of therapy, there are large pulmonary emboli or the patient is not improving much clinically. If the patient is doing well, cleared the bacteremia and had only minor valve damage or no damage at all, I would stop the gentamicin at this point.
6. I would ensure that the renal function does not deteriorate, and have a low threshold for stopping gentamicin if it does. I would perform therapeutic drug monitoring if I commit to 2 weeks of therapy, checking the trough level before the 4th dose and wanting it to be undetectable, then twice more (at one week and mid second week).