Weekly Articles: 3/7/2021

Huzzah! I have been meandering this week in terms of my writing (work + I can get lazy sometimes). Some new topics coming in the future (including: hyperviscous Klebsiella, hemorrhagic fevers, and post-Chikungunya arthritis), whenever I get to them. But for now, just some literature to get through.

This pair of articles are a bit dated (as in earlier in the year) but I cover them now because I have seen ads on the Lancet ID/CID for this new drug. It’s called cabenuva (not capybara) and its a once-monthly injectable drug consisting of cabotegravir and rilpivirine (integrase inhibitor and non-nucleotide reverse transcriptase inhibitor). I’ll go over the 2 articles cited in the ad, which were published in the same edition of NEJM.

Swindells S, Andrade-Villanueva JF, Richmond GJ, Rizzardini G, Baumgarten A, Masiá M, Latiff G, Pokrovsky V, Bredeek F, Smith G, Cahn P, Kim YS, Ford SL, Talarico CL, Patel P, Chounta V, Crauwels H, Parys W, Vanveggel S, Mrus J, Huang J, Harrington CM, Hudson KJ, Margolis DA, Smith KY, Williams PE, Spreen WR. Long-Acting Cabotegravir and Rilpivirine for Maintenance of HIV-1 Suppression. N Engl J Med. 2020 Mar 19;382(12):1112-1123. doi: 10.1056/NEJMoa1904398. Epub 2020 Mar 4. PMID: 32130809.

This was a randomized, multicenter, parallel group, open label trial of HIV patients who have been previously on ART for at least 6 months with viral suppression before enrolling. Exclusion criteria included infection with hepatitis B, virological failure with either integrase inhibitor or NNRTI, mutations, or interruption of ART in the 6 months before enrollment. Patients were enrolled in a 1:1 fashion to long-term injectable or to continue their ART. The scheme is as follows:

Those randomized into the long-term injectable therapy had a lead-in period of PO therapy for 4 weeks (to monitor for side effects) followed by a loading dose of 600mg of cabotegravir and 900mg of rilpivirine injected and then 400/600 every 4 weeks through 52 weeks as maintenance. Primary endpoint was percentage of patients with viral load >50 copies/mL. 618 patients were randomized, with no difference in patients with >50 copies/mL at week 48 in the ITT analysis:

Most of the adverse events here were related to injections, but not different outside of that. 

Orkin C, Arasteh K, Górgolas Hernández-Mora M, Pokrovsky V, Overton ET, Girard PM, Oka S, Walmsley S, Bettacchi C, Brinson C, Philibert P, Lombaard J, St Clair M, Crauwels H, Ford SL, Patel P, Chounta V, D’Amico R, Vanveggel S, Dorey D, Cutrell A, Griffith S, Margolis DA, Williams PE, Parys W, Smith KY, Spreen WR. Long-Acting Cabotegravir and Rilpivirine after Oral Induction for HIV-1 Infection. N Engl J Med. 2020 Mar 19;382(12):1124-1135. doi: 10.1056/NEJMoa1909512. Epub 2020 Mar 4. PMID: 32130806.

The accompanying study was a randomized, multicenter, open-label trial of long-acting injectable therapy in patients with HIV with viral loads of >1000 copies/mL who had not been on any ART prior to enrollment. Patients were first induced with triumeq (abacavir, lamivudine, dolutegravir) for 16 weeks, and those with undetectable viral load then were entered into the study, which randomized patients in a 1:1 fashion to either injectables or continuing their PO therapy (loaded in a similar fashion as above):

Primary endpoint was the percentage of patients with viral load >50 copies/mL at the end of 48 weeks. 566 patients were randomized, with both groups being fairly well balanced. There was no statistical difference in the primary outcome between groups:

And as above, the main side effects related to those in the injection site.

I think this is a fairly exciting time for HIV care. The new integrase inhibitors are fairly potent and have a high barrier to resistance, which has led to people trying two-drug regimens (looking at you, Juuluca) and now, it seems it is possible to do a one injection per month, which patients may prefer. This may be a reasonable first-line therapy in places that have the capabilities to have a dedicated injection clinic, which is one of the downsides of this. Furthermore, missing a dose by a week may have implications in terms of mutations (which wasn’t really seen in these studies), and with scheduling, this may become an issue where you may have to write for backup oral prescriptions for these drugs to keep from achieving mutations in the event someone misses doses. Indeed, patients who were randomized to the injectable regimen had an option of oral bridging if you missed a dose. 

Esquer Garrigos Z, Jandhyala D, Vijayvargiya P, Castillo Almeida NE, Gurram P, Corsini Campioli CG, Stulak JM, Rizza SA, O’Horo JC, DeSimone DC, Baddour LM, Sohail MR. Management of Bloodstream Infections in Left Ventricular Assist Device Recipients: To Suppress, or Not to Suppress? Open Forum Infect Dis. 2020 Nov 16;8(1):ofaa532. doi: 10.1093/ofid/ofaa532. PMID: 33447628; PMCID: PMC7794653.

This is a retrospective study evaluating the management of bacteremia in LVAD patients and assessing the safety and efficacy of long-acting suppressive therapy. These were defined as follows:

• LVAD-related: source likely either pocket, driveline, or pump
• LVAD-associated: unable to determine that LVAD as the source, but highly suspicious
• non-LVAD: another non-LVAD source 

Primary outcome was occurrence and timing of relapse bacteremia. In this single-center study, 241 patients with LVAD were evaluated, of which 80 had bacteremia totaling 121 distinct episodes. Most were non-LVAD associated (46%), compared to LVAD-related (38%), and LVAD-associated (15.7%). LVAD-related bacteremias presented later than the others and were more persistent as well:

The money image is as follows:

Its a busy figure, so I’ll try to break it down as best as I can. 32 out of 35 patients with LVAD-associated bacteremia were given suppressive therapy, with a relapse rate of 37.5% (rest either died, relapsed before suppressive therapy, or got transplanted). 9 out of 14 of LVAD-related cases got suppressive therapy with none experiencing relapse. 13 of 46 cases of non-LVAD bacteremias got suppressive therapy, with none having relapses. Further, 31 of 33 who did not get suppressive therapy did not suffer relapse. IT should be noted that staph aureus was the most common cause of relapse in those with LVAD-associated infections and LVAD-related infections who got suppressive therapy. There is more detail I wish we could have gotten here, including which drug and duration of therapy. I do not think people would balk at the idea of giving suppressive therapy to those with LVAD-associated staph aureus bacteremia, which recurs in non-LVAD situations anyways. Gram negatives is a bit tricky, but I do see treating with oral regimens for an extended period of time (in comparison of Staph aureus, where I do not blame indefinite suppression, at least until transplant and a bit beyond). Speaking of Staph…

McHale T, Medder J, Geske J, Rupp ME, Van Schooneveld TC. The Effect of Insurance on Appropriate Hospital Discharge Antibiotics for Patients With Staphylococcus aureusBacteremia. Open Forum Infect Dis. 2020 Dec 22;8(1):ofaa568. doi: 10.1093/ofid/ofaa568. PMID: 33511223; PMCID: PMC7817079.

This was one part surprising and one part expected. This is a single-center retrospective study evaluating appropriate Staph Aureus bacteremia (SAB) antibiotic discharge based primarily on insurance (either no insurance, Medicaid, Medicare, or commercial insurance). Primary outcome was receipt of inappropriate antibiotic on discharge. No distinction was made for uncomplicated or complicated (as this would have been difficult to chart review) so a cut-off of 14 days was used. Logistic regression model was used to evaluate the relationship between outcome and exposure. 273 patients were included in the study, with unadjusted models showing that lack of ID consult, lack of nursing service after discharge were both associated with inappropriate antibiotic therapy. Further, increased odds of inappropriate antibiotic therapy was associated with decreasing insurance coverage:

After adjusting for ID consultation and nursing service, here was no association between type of insurance and appropriate antibiotic therapy (OR 1.57, 95% CI 0.98-2.53), though a trend was noticed. Multivariate logistic regression analysis found that lack of ID consultation and lack of nursing services were still associated with inappropriate antibiotic therapy, as opposed to lack of insurance:

Again, ID consultation in Staph Aureus bacteremia leads to appropriate antibiotics (though the definition of appropriate depended on the 14 day threshold, not uncomplicated/complicated). I am a bit surprised that insurance was not an issue, but I think a larger study would suggest that it does, in fact, impact your antibiotic choice. 

Rhee C, Chiotos K, Cosgrove SE, Heil EL, Kadri SS, Kalil AC, Gilbert DN, Masur H, Septimus EJ, Sweeney DA, Strich JR, Winslow DL, Klompas M. Infectious Diseases Society of America Position Paper: Recommended Revisions to the National Severe Sepsis and Septic Shock Early Management Bundle (SEP-1) Sepsis Quality Measure. Clin Infect Dis. 2021 Feb 16;72(4):541-552. doi: 10.1093/cid/ciaa059. PMID: 32374861.

This is not really an article, but more of IDSAs views on the septic shock early management bundle (SEP-1). This requires hospitals to enforce compliance with the sepsis bundle, which includes getting blood cultures, giving IV fluids, and antibiotics within a certain amount of time. The IDSA reports several concerns with this bundle, including clinicians feeling compelled to give antibiotics to avoid being penalized in syndromes that mimic sepsis (there is a good table in this paper on mimics). One interesting thing they note is that time-to-antibiotics and mortality is more strongly associated with septic shock than just sepsis. They also go through some possible biases in studies supporting early antibiotic therapy, as well as issues that may lead to clinicians not adhering to sepsis-bundle compliance, such as not being able to measure lactate or volume overload states (i.e. CHF or renal failure). IDSA recommends the following:

• Removing sepsis without shock from SEP-1, given that the evidence does not support immediate antibiotics here and there may be benefit from further data gathering in this stage
• Elimination of documentation of possible or suspected septic shock and to modify persistent hypotension to more objective and reproducible criteria
• Do not measure lactate (which I am a fan of)

Things such as hemodynamic monitoring and fluid management are deferred to other societies. I recommend checking out this paper to get the full benefit of it.

S Scott Sutton, Joseph Magagnoli, Tammy Cummings, James Hardin, Association Between the Use of Antibiotics, Antivirals, and Hospitalizations Among Patients With Laboratory-confirmed Influenza, Clinical Infectious Diseases, Volume 72, Issue 4, 15 February 2021, Pages 566–573, https://doi.org/10.1093/cid/ciaa074

This is a retrospective cohort study evaluating antibiotic and antiviral usage in a VA cohort with lab-confirmed influenza and its impact in hospitalization rates. Primary outcome was inpatient hospitalization rates within 30 days of influenza diagnosis, categorized into all-cause or respiratory hospitalization. 12,806 cases were evaluated, with those with more  comorbidities being treated with both antibiotics and antivirals. Most commonly prescribed antibiotics were macrolide, followed by penicillins and quinolones. Those who got antibiotics + antivirals had a lower incidence of hospitalizations that was highest within 5 days:

In those with any type of therapy, antibiotic + antiviral therapy was more protective of hospitalization but this did not reach statistical significance:

The authors note the absolute magnitude of reduced risk of dual therapy to be small. Despite this, the population studied was older and had more comorbidities which may not be generalizable to a general population. Further, risk factors for admission were not evaluated here, which would be helpful to risk-stratify those who would benefit from dual therapy. 

Castillo-Mancilla JR, Cavassini M, Schneider MP, Furrer H, Calmy A, Battegay M, Scanferla G, Bernasconi E, Günthard HF, Glass TR; Swiss HIV Cohort Study. Association of Incomplete Adherence to Antiretroviral Therapy With Cardiovascular Events and Mortality in Virologically Suppressed Persons With HIV: The Swiss HIV Cohort Study. Open Forum Infect Dis. 2021 Jan 21;8(2):ofab032. doi: 10.1093/ofid/ofab032. PMID: 33604408; PMCID: PMC7880264.

This is a prospective multicenter cohort study of people living with HIV on ART in Switzerland. Patients with >6 months of viral suppression and no history of cardiovascular disease at baseline were enrolled. Primary endpoint was any cardiovascular event as a composite outcome. ART adherence was determined based on a self-reported questionnaire that contained 2 westions. Over 6971 patients were included. Multivariate analysis found that missing >1 ART dose in the previous 4 weeks was associated with increased CVD events in univariate analysis (HR 1.11, 95% CI 0.76-1.61) and multivariate analysis had a HR of 1.35 (95% CI 0.9-2.03). With imputations of missing data, results remained non-statistically significant:

When it came to non-CVD-related mortality, missing >1 dose of ART was associated with a HR for non-CVD related mortality of 1.46 (95% CI 1.02-2.09) in univariate analysis and 1.44 (95% Ci 1.00-2.07) in multivariate analysis. In a model assessing >2 doses missed, HR was 2.21 (95% CI 1.38-3.56) in univariate analysis and 2.21 (95% 1.37-3.57) in multivariate analysis. The idea behind higher CVD-related mortality and lack of ART adherence relates to low-level HIV replication and thus, enhanced inflammation and immune activation. Despite this hypothesis, this study did not reveal higher CVD-related mortality but found higher overall mortality, which I assume amounts to something (don’t miss your ART, kids). 

Rhoads JLW, Willson TM, Sutton JD, Spivak ES, Samore MH, Stevens VW. Epidemiology, Disposition, and Treatment of Ambulatory Veterans With Skin and Soft Tissue Infections. Clin Infect Dis. 2021 Feb 16;72(4):675-681. doi: 10.1093/cid/ciaa133. PMID: 32047886.

This is a multicenter retrospective study of VA patients diagnosed with SSTI in outpatient clinics and EDs over a 13 year period. Over 1.5 million patients were evaluated, and one of the notable findings was a decrease in the incidence of SSTI from 4.58 per 1000 patient years in 2005 to 3.27 per 1000 patient-years in 2018:

Only 5.8% of these cases were admitted, mostly from the ED. antibiotics were more likely to be prescribed by ED physicians than primary care physicians (90% vs 61%), and the ED was more likely to use anti-MRSA therapy as well (51.4% vs 35.1%):

The pattern of treatment was similar for purulent and non-purulent SSTI, with the nonpurulent being more likely to receive beta-lactam monotherapy:

This doesn’t surprise me much, at least when it comes to the ED giving more MRSA therapy than primary care. The study does note a low percentage of abscesses being drained in both settings (8.1% in the ED and 2.6% in primary care), which may not help especially in purulent cellulitis cases. 

Alcorn K, Gerrard J, Cochrane T, Graham R, Jennison A, Irwin PJ, Barbosa AD. First Report of Candidatus Mycoplasma haemohominis Infection in Australia Causing Persistent Fever in an Animal Carer. Clin Infect Dis. 2021 Feb 16;72(4):634-640. doi: 10.1093/cid/ciaa089. PMID: 32006025.

This one is both fascinating and, at the same time, infuriating. This is a case report describing an Australian woman who presented with a 4 day history of fevers and myalgia, with a preceding history of 4 months of fatigue and weight loss. She is a wildlife reserve assistant, with not so interesting travel history. She presented with fever, neutropenia, t hrombocytopenia, elevated LDH and CT abdomen/pelvis revealed splenomegaly. She got pip-tazo without improvement and subsequently doxycycline, which improved her pancytopenia and fevers. She presented multiple times with similar symptoms as she self-discontinued her doxycycline, leading to a spleen biopsy and subsequently, splenectomy. She eventually gets prednisone (thinking it was an inflammatory disease), but she worsened with severe fever, weakness, abdominal pain, severe anemia, elevated LDH. She had a repeat blood film which revealed the following:

16s PCR ultimately revealed Mycoplasma haemohominis, which was the second case reported in humans. As seen above, this organism can be seen as small, ring-like inclusions in red blood cells. In this patient, it was present in 2% of erythrocytes. This is a well-known organism in the veterinary world, and has been found in mammals such as rats, bats, dogs, monkeys, reptiles, and cats (where it causes fever and hemolytic anemia). 

Chen S, Pawelec G, Trompet S, Goldeck D, Mortensen LH, Slagboom PE, Christensen K, Gussekloo J, Kearney P, Buckley BM, Ford I, Jukema JW, Westendorp RGJ, Maier AB. Associations of Cytomegalovirus Infection With All-Cause and Cardiovascular Mortality in Multiple Observational Cohort Studies of Older Adults. J Infect Dis. 2021 Feb 3;223(2):238-246. doi: 10.1093/infdis/jiaa480. PMID: 32909605; PMCID: PMC7857154.

This is a meta-analysis of 5 observational cohorts evaluating the relationship between CMV serology positivity and cardiovascular mortality and all-cause mortality. 10,122 patients were evaluated. Cox regression was applied with different models for adjustment of confounders being applied (model 1: age/sex, model 2: age/sex, education, BMI, smiking status, and co-morbidities, model 3: all of the above + CRP). Overall, CMV serostatus was not associated with all-cause mortality when all CMV IgG titers were compared to CMV seronegativity. This was the same case for cardiovascular mortality. Though when titers were divided into 4 quartiles (1 being lowest, 4 being highest), there was an association with CMV seropositivity and all-cause mortality in model 1, but not the other models:

When comparing relative to quartile 1, there was no association between higher titers and all-cause and cardiovascular mortality:

A few things to note; this was a mostly white population (the cohorts were from Denmark). Further, I was unable to find the thresholds for the quartiles (unless I missed it, in which case, silly me) so I cannot comment on this. There was a trend towards increased mortality, as you can see above, however how this changes with a different population remains to be seen. I do get the rationale, though. CMV is not an infection you clear but rather it remains latent, so it stands to reason it may lead to higher inflammation and thus, higher cardiovascular events. But this was not seen in this paper. 

COVID related articles:

Bielecki M, Züst R, Siegrist D, Meyerhofer D, Crameri GAG, Stanga Z, Stettbacher A, Buehrer TW, Deuel JW. Social Distancing Alters the Clinical Course of COVID-19 in Young Adults: A Comparative Cohort Study. Clin Infect Dis. 2021 Feb 16;72(4):598-603. doi: 10.1093/cid/ciaa889. PMID: 32594121; PMCID: PMC7337655.

This is a prospective study of soldiers in a Swiss Army Base that evaluated the efficacy of social distancing. 508 soldiers, divided in 3 companies, were evaluated. Those in the second and third company had no social distancing (stationed in the same barracks, shared kitchen/communal areas), while company 1 was stationed 3km away from the other 2. So no contact between company one and 2 and 3. They followed an outbreak that began in company 3, with subsequent outbreak in those companies. 9 days after the first case, strict social distancing and hygiene measures were enacted. As noted below, company 1 had no outbreaks:

On day 35, 363 asymptomatic soldiers were sampled for RT-PCR, serological testing, and systematic questionnaire. Of the 88 tested in company 1, there was evidence of past or active exposure in 13 soldiers (15%), compared to 64% in company 2 and 59% in company 3.

Dong YM, Sun J, Li YX, Chen Q, Liu QQ, Sun Z, Pang R, Chen F, Xu BY, Manyande A, Clark TG, Li JP, Orhan IE, Tian YK, Wang T, Wu W, Ye DW. Development and Validation of a Nomogram for Assessing Survival in Patients With COVID-19 Pneumonia. Clin Infect Dis. 2021 Feb 16;72(4):652-660. doi: 10.1093/cid/ciaa963. PMID: 32649738; PMCID: PMC7454485.

This is a retrospective study of 628 patients with COVID-19 that aimed to develop a prognostic tool to evaluate the incidence of severe COVID-19 infection. Prognostic factors were screened out using LASSO regression and then analyzed in a multivariate Cox proportional hazard model to identify factors associated with survival. These were then used in an in-hospital survival prediction model. 390 patients were allocated in the training cohort while were placed in the 238 in the validation cohort. Hypertension, neutrophil-to-lymphocyte ratio (NLR), and NT-pro BNP were independent factors associated with in-hospital survival:

A nomogram was then established to predict 14 and 21-day survival, which is visualized as below, using a hypothetical patient with a history of HTN, an NLR of 4, and NT-proBNP of 1000:

• HTN = 0 points
• NLR = 36
• NT-pro BNP = 22

This yields a total of 58, and thus a 75% survival at day 14 and 62% at day 21. In other words, the more points you get, the higher the chances of you living (its confusing, since its backwards from the usual scoring system. In other words, very low lymphs, presence of hypertension, and high BNP = death):

The AUC for the 14-day nomogram was 0.922 in the training cohort and 9.22 in the validation cohort, while it was 0.923 in the 21-day training cohort and 0.881 in the validation cohort. Here is a table with HR, which I think is helpful:

I mean, I do not know if this changes how people approach prognosis, but it is a simpler one to use if someone hits the door and it negates the use of things such as D-dimer, ferritin, CRP, etc. It would still need external validation as this was done in one hospital in Wuhan, China.

Dawson P, Rabold EM, Laws RL, Conners EE, Gharpure R, Yin S, Buono SA, Dasu T, Bhattacharyya S, Westergaard RP, Pray IW, Ye D, Nabity SA, Tate JE, Kirking HL. Loss of Taste and Smell as Distinguishing Symptoms of Coronavirus Disease 2019. Clin Infect Dis. 2021 Feb 16;72(4):682-685. doi: 10.1093/cid/ciaa799. PMID: 32562541; PMCID: PMC7337666.

This is a prospective study from the Wisconsin Department of Health services that identified households of a lab-confirmed COVID-19 index case with the goal of describing symptom profiles. 90 participants from 26 households, including 64 household members were evaluated. 25% of the household members (16/64) tested positive for COVID. The most common symptoms in the 42 individuals that tested positive were as follows:

• Cough in 81%
• Headache in 76%
• Fever in 64%
• Loss of taste/smell in 62%
• Nasal congestion in 62%

There was no difference in age, sex, race/ethnicity, or co-morbidites between those who reported loss of taste/smell and those who didn’t. In none of the patients was it the only symptom. In the remaining 64 household members, loss of taste/smell was reported in 12 cases, of which 10 tested positive. PPV was 83% (95% CI 55-95%), which was higher than for fever and cough:

So that is cool, it may be that loss of taste/smell is a far more common symptom than I thought, though in the absence of other symptoms it may not mean much. 

Ng OT, Marimuthu K, Koh V, Pang J, Linn KZ, Sun J, De Wang L, Chia WN, Tiu C, Chan M, Ling LM, Vasoo S, Abdad MY, Chia PY, Lee TH, Lin RJ, Sadarangani SP, Chen MI, Said Z, Kurupatham L, Pung R, Wang LF, Cook AR, Leo YS, Lee VJ. SARS-CoV-2 seroprevalence and transmission risk factors among high-risk close contacts: a retrospective cohort study. Lancet Infect Dis. 2021 Mar;21(3):333-343. doi: 10.1016/S1473-3099(20)30833-1. Epub 2020 Nov 2. PMID: 33152271; PMCID: PMC7831879.

Here is a paper you won’t see come out of the United States, since I do not think anyone here would subject themselves to such scrutiny. This is one out of Singapore, and it evaluated the prevalence of SARS-CoV2 and risk factors in exposed individuals in SIngapore. They analyzed close contacts of confirmed cases over a 4 month period, including household contacts (those who share the same address), high risk contacts (contact for at least 30min), and low-risk contacts (contact 10-30 min). Household and high risk contacts were placed under legally enforced quarantine for 14 days, with public health officials assessing contacts via phone 3 times per day (yep). Those who were symptomatic were transferred to the hospital, and those who were diangosed were admitted until they had 2 negative nasopharyngeal PCR. All household and work/social contacts were given a risk factor questionnaire as well as serology testing. 

1114 PCR confirmed cases and 13026 close contacts were evaluated. Secondary attack rate was higher for household contacts (5.9%) compared to work or social contacts (1.3%). Using Bayesian modelling estimates along with serology and symptom questionnaires and adjusting for testing rates among the subjects, the overall secondary infection rate was 11% for household contacts, 4% for social contacts, and 5% for work contacts:

They also found that around 60% of cases were likely missed. The authors also did univariate analysis and multivariate analysis for risk factors for COVID-acquisition for household contacts, which included contact with more than one case, spouse/partner of the case, being handed an object touched by someone with COVID, sharing a meal with a COVID case, sharing a bedroom, being in the same car, and speaking with someone with COVID, even less than 30 min (I am unclear about the mask here). 

For work/social contacts, it was very similar, though having direct physical contact with a case and sharing a meal were not associated with increased risk. Further, being spoken to by a COVID patient for <30min did have statistical significance, but barely.

Again, this is very interesting and it is nice to have some good data from a large cohort to assess risk factors. But this is fairly intense monitoring and I do not see something like that happening in the states.