By Dean L. Winslow, MD, FACP, FIDSA

Professor of Medicine, Division of General Medical Disciplines, Division of Infectious Diseases and Geographic Medicine, Stanford University School of Medicine

Dr. Winslow reports no financial relationships relevant to this field of study.

SYNOPSIS: A before-and-after intervention study compared 170 patients treated with either oxacillin IV or vancomycin IV for six weeks (plus gentamicin IV given during the first five days) to 171 patients who were treated with TMP/SMZ IV plus clindamycin IV for the first week followed by TMP/SMZ PO (without clindamycin) to complete a six-week course. Mortality and hospital length of stay were significantly less in the TMP/SMZ-treated patients.

SOURCE: Tissot-Dupont H, Gouriet F, Oliver L, et al. High-dose trimethoprim-sulfamethoxazole and clindamycin for Staphylococcus aureus endocarditis. Int J Antimicrob Agents 2019, June 8. [Epub ahead of print].

This prospective study was conducted from 2001-2017 with the control group of either oxacillin (for methicillin-susceptible Staphylococcus aureus [MSSA]) or vancomycin (for methicillin-resistant S. aureus [MRSA]) prescribed from 2001-2011 and the trimethoprim sulfamethoxazole (TMP/SMZ) plus initial clindamycin (T&C) regimen given from 2012-2017. The groups were well-matched at baseline, with the T&C patients slightly older than the control patients (64.4 vs. 59.4 years). Underlying diseases (diabetes, hypertension, chronic kidney disease, dialysis, coronary artery disease, etc.) were similar between the groups. Fifty-seven percent of patients in both groups had native valve endocarditis. Twenty-five of the T&C patients and 20% of the control patients had prosthetic valve endocarditis, and 28% of patients in both groups had cardiac devices. Twelve percent of the T&C patients and 11% of the control patients had MRSA.

Although septic failures were not significantly different between the arms in the ITT analysis, the on-treatment analysis showed that six patients in the control group and three patients in the T&C group failed to clear their blood cultures despite the addition of other antibiotics. Although not statistically different, 13 patients in the control group and eight patients in the T&C group experienced a relapse of infection. By ITT analysis, in-hospital death was seen in 18% of control patients and 10% of T&C patients. Although the investigators did not directly compare nephrotoxicity rates, acute kidney injury was seen in nine T&C patients and in only one control patient. Allergic reactions were similar in both groups.


I have always loved TMP/SMZ. Trimethoprim was just one of the many life-saving antimetabolites discovered by Trudy Elion and George Hitchings in their laboratory at Burroughs-Wellcome in North Carolina. TMP/SMZ also is > 90% orally bioavailable and is bactericidal against most sensitive organisms, including S. aureus and Listeria.1 Current U.S. and European guidelines recommend four to six weeks of IV antibiotics to treat S. aureus endocarditis.

This study differs from other trials of TMP/SMZ for S. aureus endocarditis and bacteremia in that investigators used high-dose IV TMP/SMZ (960/4,800 mg daily) and IV clindamycin (1,800 mg/day) for the initial seven days of treatment, followed by monotherapy with oral TMP/SMZ (320/1,600 mg or 2 DS tabs) three times per day to complete a six-week course. They also allowed the addition of IV rifampin (1,800 mg/day) and IV gentamicin (180 mg/day) in both groups if patients had persistently positive blood cultures or if they had a cardiac abscess. By this liberal definition of “persistent bacteremia,” 28% of the T&C patients were noted but the number in the control arm was not mentioned. (All members of the control group also received IV gentamicin during their first seven days of treatment.) The study also seemed a bit unusual in that 67% of the control patients and 52% of the T&C patients underwent surgery at some point in their treatment courses.

Although the results of this study are a bit difficult to interpret, it reassures me that high-dose TMP/SMZ is a reasonable choice for the treatment of S. aureus endocarditis. My friend, Norm Markowitz, published a highly cited study back in the early 90s using standard dose TMP/SMZ that appeared to show that TMP/SMZ may have been inferior to vancomycin for the treatment of MSSA, but not MRSA endocarditis.1 However, many have questioned this interpretation because of the small number of patients studied from his largely injection drug use patient population in Detroit and the low dose of TMP/SMZ used in his trial.

The Tissot-DuPont et al study throws another confounding factor into the mix: the use of clindamycin “to inhibit toxin production” during the first seven days of treatment. Many of the patients in this study were very sick (including 14% in the T&C arm who were in septic shock), so the use of clindamycin may have been reasonable in these cases. Although I usually do not get too concerned if a patient with S. aureus endocarditis does not clear the bacteremia in 48 hours, I suspect the antagonism of the bactericidal effect of TMP/SMZ by clindamycin played a role. Others have shown both in vitro and in vivo antagonism of the bactericidal activity of cell wall active antibiotics by rifampin in patients with S. aureus bacteremia.2 Years ago, we showed that rifampin and other bacteriostatic antibiotics antagonize the bactericidal action of cell wall active antibiotics and TMP/SMZ against Listeria,3 so I suspect clindamycin is doing the same thing.

The authors did not give a lot of detail about the nephrotoxicity encountered in this study (which resulted in dose reduction rather than discontinuation of TMP/SMZ in most cases they listed). My suspicion is that the direct effect of TMP on reducing the tubular secretion of creatinine may have caused some “over-calling” of nephrotoxicity. However, having said that, I am cautious in using TMP/SMZ, particularly in patients with mild underlying renal insufficiency in the setting of diabetes since TMP/SMZ not infrequently causes the serum potassium to rise a bit (possibly related to underlying type IV RTA, which is common in patients with diabetes).

In any case, this is a very interesting large study that eventually may change practice (and spare patients the expense and complications from PICC lines), but it should be replicated in a larger, prospective, randomized, controlled trial.


  1. Winslow DL, Pankey GA. In vitro activities of trimethoprim and sulfamethoxazole against Listeria monocytogenes. Antimicrob Agents Chemother 1982;22:51-54.
  2. Markowitz N, Quinn EL, Saravolatz LD. Trimethoprim-sulfamethoxazole compared with vancomycin for the treatment of Staphylococcus aureus infection. Ann Intern Med 1992;117:390-398.
  3. Riedel DJ, Weekes E, Forrest GN. Addition of rifampin to standard therapy for treatment of native valve infective endocarditis caused by Staphylococcus aureus. Antimicrob Agents Chemother 2008;52:2463-2467.
  4. Winslow DL, Damme J, Dieckman E. Delayed bactericidal activity of beta-lactam antibiotics against Listeria monocytogenes: Antagonism of chloramphenicol and rifampin. Antimicrob Agents Chemother 1983;23:555-558.