General ID: September 2018

When can patients who have undergone surgery for pyogenic spondylodiscitis be treated with short courses of IV antibiotics? Note the emphasis on IV; this is important later. The authors conducted a 3-year retrospective cohort study including 102 patients who received surgery for pyogenic spondylodiscitis at their university hospital. The primary outcome of interest was recurrence of disease, defined as return of evidence of infection requiring another course of IV antibiotics and/or additional surgery within one year.

The median patient age was 61 and two-thirds were male; the most common comorbidities were diabetes (32%), ESRD (19%), and cirrhosis (17%). The most common findings on culture were S. aureus (40%, with a 60/40 mix of MRSA and MSSA), various gram-negative bacilli (23%), and no organism (25%). A quarter of patients had recurrent disease, and in multivariate analysis the independent predictors of recurrence included positive blood cultures (aOR 2.97; p=0.03) and presence of a paraspinal abscess (aOR 3.36; p=0.02).

The authors then stratified the cohort as being “high risk” for recurrence (i.e. having positive blood cultures and/or a paraspinal abscess) or “low risk” (lacking both risk factors). Forty-two percent of the cohort was classified as high-risk, and their infection recurrence rate was 35%; among the low-risk patients the recurrence rate was 19% (p=0.052 for the comparison). When the authors further stratified the cohort by whether patients had received short-course therapy, defined as <3 weeks of IV antibiotics, or a longer duration of IV therapy, they found that in the high risk group, short course therapy was associated with more recurrent infections (56% vs 22%; p=0.027), but that there was no difference in outcome with short course therapy in the low risk group (16% versus 21%; p=0.461).

I think there are two important caveats to this study. First, patients’ receipt of oral antibiotics before and/or after IV therapy is not reported. So, is this a study showing that less than 3 weeks of antibiotic therapy is as good as longer courses of antibiotics in low-risk patients, or a study showing that switching from IV to PO antibiotic therapy by 3 weeks is as good as switching to PO later? We can’t say. Personally, I’m already comfortable switching from IV to PO antibiotics earlier than 3 weeks – but would still be giving the traditional 6 week total antibiotic duration for vertebral osteomyelitis (or 8+ weeks given risk factors for failure as defined by Park et al; see 26917813). Second, I would consider a 5% absolute difference in the incidence of recurrent disease to be clinically meaningful if the disease in question is spondylodiscitis (spines are important, y’all), so I don’t find the lack of statistical significance between the recurrence rates with short vs long course therapy in a small sample size of low risk patients particularly reassuring. 30223785

Chloroquine is frequently under-dosed in the treatment P. vivax malaria, leading to more recurrent disease. That’s the conclusion of a meta-analysis of 37 studies including 5240 patients published in this month’s Lancet ID. Of the patients who received chloroquine alone, 35% were given less than the optimal 25mg/kg dose, with underdosing being more significantly more common in overweight patients. In a multivariate analysis controlling for patient age, sex burden of parasitemia, and geographic variation in incidence of relapse, each 5mg/kg increase in chloroquine dose between 20mg/kg and 30mg/kg was associated with a reduced rate of recurrent disease (aHR 0.82; p = 0.021), a benefit most pronounced in children (aHR 0.59; p = 0.006). When patients treated with chloroquine and primaquine were included in the analysis, use of primaquine was also highly protective against recurrent disease (aHR 0.1; p <0.0001), to the degree that chloroquine dose was not associated with recurrence rate for the patients also given primaquine. 30033231

Speaking of malaria, Lancet ID also published a double-blind RTC of the NIH’s new malaria vaccine this month, though the results were disappointing. Pfs25H-EPA is a protein-protein conjugate vaccine for P. falciparum that was trialed in 120 patients: 20 patients in a pilot cohort study to establish the vaccine’s safety, and then 100 more patients randomized to receive either 4 doses of Pfs25H-EPA or a control consisting of 3 doses of HBV vaccine followed by one dose of the meningococcal vaccine. Pfs25H-EPA recipients had more treatment-related adverse events than control vaccine recipients (191 vs 126, p = 0.034), only developed fleeting antibody responses with half-lives of 1-2 months, only developed better in vitro serum transmission-reducing activity than the control vaccine recipients after their final dose of Pfs25H-EPA, and did not transmit malaria to feeding mosquitos any less frequently than the control vaccinees. So it goes. 30061051

Prophylactic antibiotics meaningfully reduce exacerbations in patients with stable COPD at the cost of antibiotic resistance. Wang et al performed a meta-analysis of twelve RTCs (pooled n=3683) examining the effect of prophylactic antibiotics on frequency of exacerbations and quality of life in patients with stable COPD. The severity of COPD in the patient population was heterogeneous, though most studies involved patients receiving multiple medications (e.g. inhaled steroids, long-acting B2 agonists, theophylline, etc). The studies used azithromycin, clarithromycin, erythromycin, roxithromycin, moxifloxacin, and doxycycline; in most trials antibiotics were given daily, though in two azithromycin was given thrice weekly, and two other studies used an on-off cycle approach (e.g.; moxifloxacin given for 5 days every 8 weeks). Durations of treatment and followup ranged from 3 to 36 months.

Overall, antibiotic prophylaxis reduced the risk of any COPD exacerbation (RR 0.82), with no difference between the continuous and intermittent antibiotic strategies. The most effective antibiotics for exacerbation prevention were erythromycin and azithromycin, with NNTs of 4 and 7, respectively; moxifloxacin had no statistically significant effect. Antibiotics also reduced the frequency of exacerbations (RR 0.69), though in subgroup analysis this benefit was only observed with macrolides. Gains in quality of life (measured by a standardized questionnaire in pulmonary research, the SGRQ) were also observed, but only in the studies of prophylaxis given for greater than 6 months. Of note, antibiotic prophylaxis was not associated with reductions in hospitalization or all-cause mortality.

But no good deed goes unpunished: antibiotic prophylaxis was associated with a significant increase in carriage of antibiotic resistant organisms (OR 4.49 with 95% CI 2.48-8.12 across four studies reporting this data). Longer-term use of antibiotics and continuous versus intermittent prophylaxis did not result in statistically significant differences in acquisition of resistant organisms.

To summarize, macrolide prophylaxis (particularly azithromycin and erythromycin) decreases the risk and frequency of COPD exacerbations with relatively low NNTs, though it does not affect hospitalization or mortality; the benefit of prophylaxis with other agents is less clear. These benefits come at the cost of increased antibiotic resistance, suggesting that macrolide prophylaxis should be restricted to COPD patients with the highest risk of exacerbations, and that other antimicrobials should be used when patients receiving prophylaxis develop bacterial pneumonia. 30189002

It’s not just Zika: symptomatic Dengue infection is also associated with congenital neurologic malformations. The authors used a national database to review live births from Brazilian women with Dengue infection during pregnancy between 2006 and 2012 (well before Zika was recognized in Brazil). The outcome of interest was the presence of the ICD-10 code for congenital CNS malformation in the child. Out of 16 million recorded live births, the incidence of neurologic malformation was 0.08% (13,600 cases). Dengue infection during pregnancy was nonsignificantly associated with a 50% increase in risk of neurologic congenital malformation (95% CI 0.97-2.27), and significantly associated with ICD-10 codes for malformation of the spinal cord (OR 5.4) and ‘other congenital malformations of brain’ (OR 4.5). Yes, this is retrospective epidemiologic data and there are definite limitations to using ICD coding as a surrogate for the presence of disease, but it makes sense that Dengue, which can infect the CNS and is also a flavivirus, might produce some of the same sequelae as Zika. This is bad news for those of us in the southern US, as Dengue is already endemic in the South Texas and will probably spread north as the climate warms. 30124410

And speaking of flaviviruses, there’s a nice review of St. Louis Encephalitis Virus disease in the US in this month’s issue of AJTMH. The authors reviewed national surveillance data from 2003-2017, which encompassed 193 cases of human infection with a median 10 cases per year and no reported outbreaks in the past 15 years. SLE occurred most commonly Arkansas, Arizona, and Mississipi; the median patient age was 57, the most common presentations were encephalitis (60%), meningitis (13%), and nonspecific febrile illness (18%), and the case fatality rate was 6%. 30182919

Waning immunity to the mumps vaccine may be driving outbreaks in the US. This group from the Viral Disease and Immunization Services divisions of the CDC reviewed the epidemiology of mumps outbreaks (>20 cases in a cluster) in the US from 2010-2015. They identified 23 outbreaks with 20-500 cases per outbreak, involving 1791 cases in total (42% of all mumps disease reported to the CDC during that time). Outbreaks lasted for a median 3 months and all involved some sort of crowded close-contact encounter (e.g. jails, schools, sports leagues, and bars); the most common setting was a university campus. The mean case age was 23 years and 70% of the outbreaks primary affected young adults aged 18-24. Only 68% of cases had received the MMR vaccine and only 62% had received 2 or more doses.

The authors point out that the mumps component of MMR is actually the least immunogenic, conferring only 88% protection after 2 doses, which may be inadequate to maintain herd immunity in densely populated settings even with good vaccine coverage. This problem may actually have been exacerbated by the near-elimination of mumps in the US, resulting in a lack of natural boosting of vaccine immunity by exposure to sick contacts. A third dose of vaccine is already given in outbreak settings (though the evidence to support this, reviewed by the authors, is only modest). Perhaps it is time to make a third dose universal for college freshmen, just as we have with the meningococcal vaccine. 30204850