General ID: August 2018

Four months of rifampin is as effective as and better tolerated than nine months of isoniazid for the treatment of latent tuberculosis infection. This multinational open-label study randomized 6012 patients with LTBI (defined as a positive TST or IGRA assay with intention to treat by the patient’s clinician) to receive either four months of daily rifampin (max dose 600mg) or nine months of isoniazid (max dose 300mg) with vitamin B6. Patients whose contacts had active TB resistant to either isoniazid or rifampin were excluded from the trial; while HIV infection did not preclude enrollment, only 255 patients with HIV-infection were enrolled. The investigators followed the participants clinically every 1-2 months to ascertain whether they developed active TB during treatment. Treatment completion was measured by pill counts and defined as receipt of at least of 80% of the doses within 12mo (for rifampin) or 18mo (for isoniazid). Participants were followed for up to 28 months after treatment, and >95% of participants completed the full 28mo of follow-up.

More patients in the rifampin arm completed LTBI therapy (79% vs 63%). Adverse events occurred more frequently in the isoniazid arm (6% vs 3%), including adverse events leading to permanent cessation of the trial drug (5% vs 3%), and drug-induced hepatoxicity (1.7% vs 0.3%). During the follow-up period, the primary endpoint of active TB (either microbiologically confirmed or clinically diagnosed) occurred in 8 of 3443 patients treated with rifampin versus 9 of 3416 patients treated with isoniazid (0.1 cases per 100 person-years of follow-up for both, difference less than 0.01 cases per 100 person-years).

During college I interned with my local county public health department in the TB control division. My project that semester was to review the charts of all of the contacts of the preceding year’s active TB cases and determine who had completed their courses of isoniazid and, for those who hadn’t, how far into the course they had gotten. Our LTBI treatment completion rate was mediocre, and we had a lot of loss to follow-up at the 3-4 month mark. Accordingly, I’m a big fan of the new 4-month rifampin and 12-week rifapentine-isoniazid regimens for LTBI, as my experience says that’s about as long as people are willing to take LTBI meds anyway. What do you think - are we ready yet to say that rifampin should be the preferred regimen for LTBI? 30067931

What predicts clinical failure in prosthetic joint infection managed with debridement and implant retention (DAIR)? Two studies published this month address this topic. The first retrospectively evaluated late acute prosthetic joint infections (i.e.; infections with rapid clinical onset that occurred more than 3 months after the arthroplasty) managed with DAIR across 27 medical centers. The authors defined treatment failure as the need for either implant removal or suppressive antibiotic therapy, relapsed or new infection after surgery, or death. Out of 340 patients, the treatment failure rate was 45%. More than half of failures (76/139) occurred in patients with infection due to Staphylococcus aureus. Other significant risk factors for failure of DAIR included fracture as the original indication for prosthesis (OR 5.4), rheumatoid arthritis (OR 5.1), age over 80 (OR 2.6), male gender (OR 2) and CRP over 150mg/L (OR 2); the only protective factor was exchange of the mobile components of the prosthesis during the procedure (OR 0.35). 30092305

OK, so I read that and thought “Wow, implant retention without suppressive antibiotic therapy for a S. aureus infected prosthetic joint? Someone’s braver than I am.” But in fact, in France there is a whole group of people braver than I am, and they retrospectively studied the predictors of treatment success at 2 years of follow-up in S. aureus prosthetic joint infection managed with DAIR across six hospitals. This group defined treatment failure as needing to remove the prosthesis, needing to do subsequent debridements or give additional courses of antibiotics after the first, death, or recurrent infection; importantly, they didn’t consider suppressive antibiotic therapy alone to constitute failure. The authors included 137 patients in the study, and reported successful treatment in 76% of cases, of whom 14/104 received suppressive antibiotics. In multivariate analysis, treatment success was associated with “longer” treatment durations (HR 0.78; they don’t define “longer,” but do state the mean treatment duration was 12.9 weeks) and combination therapy with rifampin (HR 0.08). Smoking was the only risk factor associated with treatment failure (HR 3.6). The authors noted that antibiotic regimens including rifampin for less than 3 weeks were significantly less likely to be successful than those with >3 weeks of rifampin, and calculated that 10 weeks of rifampin probably offered maximal benefit.

So, I suppose if you have a patient under 80 who doesn’t smoke, doesn’t have rheumatoid arthritis, didn’t get their prosthetic joint for a fracture, and who is willing to take 3 months of antibiotics including rifampin, DAIR (ideally with exchange of the mobile components of the joint) without antibiotic suppression might be a reasonable option for S. aureus PJI. 30083889

One dose of cefazolin may not be adequate prophylaxis against surgical site infections for prolonged operations. This study examined the pharmacodynamics of perioperative cefazolin in patients undergoing cardiac surgery and how they (and length of surgery) related to the incidence of subsequent sternal wound infection over the next 30 days. Forty patients were enrolled in the pharmacodynamic study, 70% undergoing CABG and 30% valve replacement or repair. The mean duration of surgery was 4.6 hours. Eight patients developed sternal surgical site infections, and infection was associated with both lower cefazolin concentrations at the time of wound closure and longer duration of surgery (OR 2.9 for each additional hour of surgery). Surgery duration of 346 minutes (about 5.5 hours) was a clinically meaningful threshold for increased surgical risk: sternal wound infection occurred in 60% of patients with surgeries longer than this cutoff versus in 14% of those with shorter surgeries. A similar threshold was found for cefazolin concentration at closure (104mg/L; infection occurred in 30% of those with concentrations below the threshold versus 6% of those at or above it).

Length of surgery and cefazolin concentration at wound closure weren’t closely correlated in multivariate analysis, suggesting that these are two independent risk factors for sternal wound infection. The next step, in my mind, is a randomized controlled trial comparing standard intraoperative cefazolin prophylaxis to enhanced prophylaxis with a second dose given a few hours into surgery to maintain therapeutic levels of cefazolin to the end of the operation. 30150469

Trichomonas infection may be more difficult to eradicate in pregnant women. This retrospective cohort study examined Trichomonas vaginalis infection in pregnant women treated at a single academic medical center over nine years, comparing women treated with single or multiple doses of metronidazole. In total, 542 women were diagnosed with Trichomonas infection over the course of 565 pregnancies; of these, 352 received single-dose metronidazole and 74 received multi-dose regimens of metronidazole. In total, 44% of Trichomonas tests obtained 3 weeks after treatment remained positive, and multi-dose regimens of metronidazole were no more effective than single doses at eradicating infection (40% vs 45% persistent positives, p=0.5). Ninety-eight patients received three or more courses of treatment over the course of a single pregnancy. Women with multiple pregnancies affected by Trichomonas were much more likely to have positive post-treatment tests (aOR 20.1), and obese women were less likely to have positive post-treatment tests (aOR 0.3).

In my mind, the simplest explanation for these findings is that the pregnant women with recurrent infection were either nonadherent to therapy or were being reinfected by their untreated sex partners. This hypothesis is supported by a subgroup analysis of women who had negative post-treatment tests after one versus multiple courses of treatment: the women who needed multiple courses of treatment to clear their infections were more likely to report adverse events associated with the drug (13% vs 1%) and less likely to report their sex partner's treatment (56% vs 87%). However, in three cases patients experienced treatment failure after 5 or more courses of metronidazole, and samples sent to the CDC confirmed that they had acquired metronidazole-resistant Trichomonas. The authors conclude that pregnant women with Trichomonas infection should be routinely tested 3 weeks after treatment to confirm eradication of the infection, and that treatment of sex partners is critical to prevent the cycle of reinfection. 30067546