April 2019: General ID

DAAs for HCV continue to be magic: safety and efficacy of simeprevir-sofosbuvir in a large French cohort with HCV genotypes 1-4.  I don’t often review articles about HCV. With apologies to all the HCV researchers and specialists out there, I find the topic incredibly tedious.  In 2019, we have several pan-genotypic regimens, all of which achieve very high rates of cure in just about everybody with minimal side effects.  It seems the only barriers to the total eradication of HCV are societal: namely, diagnosing our patients and getting them access to these new medications.  But anyway, sometimes I think I ought to check back in with the HCV world and make sure nothing big has changed.  So here we are.

The authors enrolled patients into a large observational cohort, ANRS CO22 HEPATHER, which consisted of 9432 French denizens with genotype 1-4 HCV monoinfection who received DAAs between 2013 and 2018.  This study examined the outcomes of those who received sofosbuvir/simeprevir with (n=63) or without (n=536) ribavirin, for either 12 or 24 weeks.  Of these individuals, 71% had experienced treatment failure with an interferon-based regimen, and 56% had cirrhosis – i.e. this was a “difficult to treat” population.  Nonetheless, the overall SVR12 rate was 93% and did not vary by genotype, treatment duration, or the use of ribavirin.  Severity of liver disease correlated with the rate of severe side effects, but not with treatment outcome.  Premature treatment discontinuations were rare (3%), particularly discontinuations due to adverse events (1.5%), and half of these patients managed to achieve SV12 anyway.  The most common adverse events were asthenia (experienced in 16-38%), headache (4-22%), and pruritis (6-28%); each of these was more common in the groups who received 24 weeks of treatment. Two patients died at week 12 from sudden cardiac arrest, which was potentially attributable to the study drugs (sofosbuvir is associated with bradyarrhythmias).

The funny thing is, this study is already obsolete, as “sim/sof” has been supplantanted by newer pangenotypic regimens (e.g. sofosbuvir/ledipasvir, daclatasvir/voxilaprevir, and grazopevir/elbasvir) offering similarly high SVR rates that are equally or better tolerated and offer shorter courses of therapy.  So, I stand by what I told my fellowship PD when asked why I didn’t choose our HCV clinic elective: in a few years, everyone with HCV who can get into clinic will have been cured. 30940090

DAA treatment of HCV infection reduces the incidence of diabetes.  Since we’re talking about HCV this month, we might as well do it twice.  This retrospective study examined the ERCHIVES database for patients with HCV infection who did not have diabetes, HIV, or HBV coinfection and who were treated with either an interferon-based (n=4764) or DAA-based (n=21,279) regimen.  The authors matched these cases 1:1 to age/race/sex/propensity-score matched controls who were not treated for their HCV infections.

The authors found that incidence of diabetes was 20.6 per 1000 person years in the untreated controls versus 19.8 in the interferon-treated group and 9.9 in the DAA-treated group (p<0.0001).  Diabetes incidence was lower for those who achieved SVR (13.3 vs 19.2; p<0.0001), and patients with more advanced fibrosis/cirrhosis saw greater benefit in diabetes risk reduction.  In multivariate analysis, both DAA treatment (HR 0.53; 95% CI 0.46-0.63) and SVR (HR 0.81; 95% CI 0.7-0.93) were associated with reduced incidence of diabetes.  In contrast, interferon-treated and untreated individuals were no different in their rates of diabetes-free survival. One more reason everybody with HCV ought to receive DAAs. 30977808 

Native joint septic arthritis outcomes are better for small versus large joints.  This retrospective study examined outcomes of adult patients with native joint septic arthritis (i.e. excluding PJI) seen at a single New Zealand hospital over a five-year period.  The authors stratified the cohort into patients with small joint arthritis (hereafter, SJA, which included septic arthritis of the small joints of the hands and feet as well as septic arthritis of the clavicle; n=250) and large joint arthritis (hereafter, LJA, defined as septic arthritis involving any other joint; n=302).  They further extracted demographic and clinical data about the patients and their treatment outcomes.  Treatment failure was defined as a composite endpoint including death, readmission for ongoing, relapsed, or recurrent infection, amputation or excision arthroplasty, and arthrodesis.

The incidence of LJA was greater than SJA (13 vs 8 per 100,000 person-years) and was markedly different by patient ethnicity, with higher rates among the Maori and Pacific-islanders than Europeans (30 and 43 vs 17 per 100,000 person-years); this association may be confounded by socioeconomic status, which also correlated with septic arthritis incidence.  The authors further found that older age was associated with increased incidence of LJA (at a rate of an additional 5 episodes/100,000 per 10-year increase in age), but not SJA.  Common comorbidities included tobacco use (35%), osteoarthritis (29%), diabetes (24%), and gout (15%).

The most common sites of septic arthritis included the knee (21%), interphalangeals of the hand (20%), metacarpals (18%), and the genohumeral joint (12%); polyarticular infection occurred in 7% of cases.  Synovial WBC counts, available in 195 cases, were <50,000 in 46%, and initial CRP was normal in 10%.  Eighty percent of cases yielded a microbiologic diagnosis and the most common organism was S. aureus (53%, of which 13% were MRSA), though the microbiology of LJA and SJA differed, with more SJA being associated with animal and human bite wounds and due to streptococci, Eikenella, anaerobes, or mixed infections.

The mean antibiotic treatment durations were 25 days for SJA and 40 days for LJA.  Among the cases of SJA, treatment for two weeks was not associated with more treatment failure than 3-6 weeks of treatment (18% vs 14%; p=0.28).  On the other hand, treatment of LJA with two weeks of therapy resulted in a higher rate of treatment failure (50% vs 22% for longer courses of treatment; p=0.02 [my calculation]).  Patients received definitive therapy with oral antibiotics in 38% of cases after a mean 8 days of inpatient IV treatment; oral antibiotics were not associated with treatment outcome overall, but did yield more treatment failures in patients with LJA (25% vs 11%; p=0.04).  Unfortunately, they don’t describe the oral antibiotics or doses used.  Surgery was performed in 83% of episodes, and approach (open vs arthroscopic) had no bearing on the treatment outcome. 

The overall rate of treatment failure was 17%, and included a 5% rate of mortality.  In a multivariate regression model, risk factors for treatment failure included LJA (OR 1.7), the presence of any intra-articular prosthetic material (OR 3.4), age (OR 1.14 per 5-year increment), and number of surgical procedures (OR 1.4 per procedure).

These are important because they suggest the adequacy of short course therapy (i.e. 2 weeks) and oral antibiotics for small joint septic arthritis, reinforcing the findings of a recent randomized controlled trial on the same topic (30992295).  Unlike that study, which was underpowered to detect differences between hand and large joint infections, here there is a clearer signal here suggesting that small and large joint arthritis are therapeutically distinct entities.  In this cohort, large joint septic arthritis had poorer outcomes and specifically poorer outcomes with short course and oral therapies.  Now, should we believe that in light of OVIVA (and noting that neither antibiotic duration nor route made it into the final multivarial regression model)?  Maybe not.  But, on the other hand, there’s so little data on this topic that I don’t have any other data on this question, much less higher quality data, at which to point. 30941403

When it comes to duration of S. aureus, every additional day of bacteremia portends a poorer outcome. This multicenter, prospective observation study examined adults with S. aureus bacteremia and examined their risk factors and clinical outcomes stratified by the bacteremia’s duration.  The authors classified bacteremia as short (1-2 days), intermediate (3-6 days), or prolonged (1 week or more).

Of 884 patients (mean age 57, 70% male), 63% had a short, 28% intermediate, and 9% prolonged bacteremia.  Prolonged bacteremia was most often due to methicillin-resistant S. aureus (p<0.001) and associated with a 1d delay in ID consultation (p=0.005).  Overall antibiotic choice was not associated with bacteremia duration, and in the subset of patients with MSSA, initial therapy with a beta-lactam (versus switch from vancomycin) was not associated with a reduction in the proportion of intermediate to prolonged bacteremias.  Interestingly, however, switch to definitive therapy with daptomycin took longer in the patients with prolonged vs intermediate vs short bacteremia (4 vs 3 vs 1 days; p=0.0017).  On the other hand, if you exclude patient in the short group, those who were switched to daptomycin vs continued on vancomycin had similar durations of bacteremia (median 6 vs 5 days; p=0.98).

Time to source control was longer in the intermediate and prolonged vs short bacteremia groups (3.5 vs 3 vs 1 days; p<0.0001).  Longer durations of bacteremia were associated with more metastatic complications (12% vs 23% vs 33%), longer length of stay (9 vs 12 vs 24 days), and higher rates of 30-day mortality (5% vs 11% vs 22%).  Among those who had a source control procedure, delays of >48hr were associated with longer LOS (median 14 vs 9 days; p<0.0001) and higher mortality (10% vs 4%; p=0.052). Every day of bacteremia was associated with a increased risk of death (RR 1.16; p<0.0001), with the greatest accumulation of risk after 3 days.

To summarize, increased duration of S. aureus bacteremia is associated with greater length of stay and mortality, and 72 hours is the point that best differentiates bacteremic patients who will do well vs poorly. 

While the study was not adequately powered to compare the ability of different antimicrobial regimens to rapidly sterilize the bloodstream of S. aureus, this paper does make me think about other studies addressing that question.  Specifically, I know that the investigators of the recent CAMERA2 trial were put off from combination therapy given the results of their trial showing no composite endpoint benefit – and in fact an increase in mortality – with combination therapy for MRSA bacteremia.  But on the other hand, that mortality finding may have been confounded by the high rates of AKI observed in the trial due to their frequent usage of vancomycin plus an antistaphylococcal penicillin, and they did find clear reductions in bacteremia duration with the combination therapy.  So I’m left wondering: might CAMERA2 have been a “positive” trial if they’d chosen another combination, like vancomycin and cefazolin, or even daptomycin and ceftaroline? 30949675

 Having an active enterovirus infection prevents children from responding to the oral poliovirus vaccine.  This makes sense – viruses can prevent infected cells from being infected by other viruses by triggering innate immune responses (e.g. stimulating interferons and the genes they trigger, priming the cell and adjacent cells into an antiviral state) and by shutting down cellular processes that haven’t been coopted for viral replication.  It’s also potentially clinically important, since the parts of the world still using the oral polio vaccine (i.e. mostly low and middle-income countries) happen to be the same places where sanitation infrastructure is often more limited and the burden of enteric viruses is highest.  This study used stool PCR to assess the prevalence of enteropathogens in children 14 days before and at the time of vaccination with the monovalent type 3 oral polio vaccine (OPV).  The study was conducted in India and enrolled 704 infants aged 6 to 11 months.

A total 56 serotypes of non-polio enteroviruses were detected in the day-of-vaccination samples.  Infection with a non-polio enterovirus was associated with reduced rates of OPV seroconversion (ORs 0.57-0.69 based on enterovirus type).  Recently acquired infections (e.g. those detected at day of vaccination) had a greater interfering effect than persistent infections (e.g. those detected at both time points).  In a subset of 120 infants who also received 16s rRNA sequencing of the stools, bacterial taxa did not differ significantly between OPV responders and nonresponders, although bacterial diversity was greater in the nonresponders.  So, I wonder if including a dose of IPV – or an additional dose of OPV – might be valuable in   30247561