Weekly Articles 8/8/2021

Not as many articles this week to go over, but nevertheless some interesting ones.

REMAP-CAP Investigators; ACTIV-4a Investigators; ATTACC Investigators. Therapeutic Anticoagulation with Heparin in Critically Ill Patients with Covid-19. N Engl J Med. 2021 Aug 4. doi: 10.1056/NEJMoa2103417. Epub ahead of print. PMID: 34351722.

More on anticoagulation and COVID. This was an international, multiplatform, randomized controlled trial comparing therapeutic anticoagulation vs prophylactic anticoagulation with LMWH or UFH in critically ill patients with COVID. This essentially combined the REMAP-CAP trial, the ATTACC trial and ACTIV-4a trials, and combined these platforms to increase the number of patients and the robustness of the data. Patients who had severe COVID, defined as the receipt of ICU-level respiratory or cardiovascular support, were included. Patients were randomized in a 1:1 fashion to therapeutic anticoagulation or standard of care prophylactic anticoagulation in an open label fashion. Primary outcome was organ support free days, evaluated in an ordinal scale through day 21. Primary analysis involved a Bayesian cumulative logistic regression model to calculate the posterior distribution for the proportional odds ratio for organ support-free days. 1103 patients were enrolled across all platforms, with enrollment ending early on December, 19 2020 due to futility after an interim analysis. Both groups were fairly well balanced and when it came to the primary outcome, the median organ support free days for the therapeutic AC group was 1 compared to 4 in the thromboprophylaxis group (adjusted OR 0.83, 95% CI 0.67 to 1.03), with posterior probability of futility of 99.9% and posterior probability of inferiority of 95%:

The median adjusted OR for survival to hospital discharge was 0.84 (95% CI 0.63 to 1.11). 

Notably, though those in the therapeutic AC had lower rates of thrombotic events, the rates of death was similar between both groups. It is unclear what is going on here, in terms as to why this is the case, but it may be the case that by itself, AC does not really do much in light of absence of anti-inflammatory therapy. In other words, there are other causes of death outside of just clotting. Though the multiplatform design may introduce some bias due to the definitions of severe COVID, I think it may be time to abandon therapeutic anticoagulation in all patients. 

O’Brien MP, Forleo-Neto E, Musser BJ, Isa F, Chan KC, Sarkar N, Bar KJ, Barnabas RV, Barouch DH, Cohen MS, Hurt CB, Burwen DR, Marovich MA, Hou P, Heirman I, Davis JD, Turner KC, Ramesh D, Mahmood A, Hooper AT, Hamilton JD, Kim Y, Purcell LA, Baum A, Kyratsous CA, Krainson J, Perez-Perez R, Mohseni R, Kowal B, DiCioccio AT, Stahl N, Lipsich L, Braunstein N, Herman G, Yancopoulos GD, Weinreich DM; Covid-19 Phase 3 Prevention Trial Team. Subcutaneous REGEN-COV Antibody Combination to Prevent Covid-19. N Engl J Med. 2021 Aug 4. doi: 10.1056/NEJMoa2109682. Epub ahead of print. PMID: 34347950.

This was part of the trial using REGEN-COV, which was a combination of casirivimab and imdevimab. This was a randomized controlled trial evaluating the use of the combination in household members of patients with COVID who had negative RT-PCR. Patients were randomized in a 1:1 fashion to subcutaneous 1200mg of REGEN -COV or placebo. Primary endpoint was proportion of patients with symptomatic RT-PCR-confirmed COVID during the 28 day period. 2067 patients with baseline negative RT-PCR were evaluated, of which 1505 who had negative serology were included. Both groups were fairly well balanced in terms of their baseline characteristics. Symptomatic SARS-COV2 infection occurred in 11/753 patients in the experimental group (1.5%) compared to 59/752 in the placebo group (7.8%), for a relative risk reduction of 81.4% (OR 0.17):

The aggregate total number of weeks with symptoms and the peak viral load were lower in the experimental group:

The relative risk reduction for those over the age of 50 was 78.1%, with 0.8% of those who got REGEN-COV having serious adverse events (compared to 1.1% in the placebo group). While these results are encouraging, the implementation of a program such as this makes it a bit difficult to fathom, outside of cost. Further, it would have to warrant contact tracing, which is another hurdle to climb. I would like to see higher risk patients being evaluated, as I think these may benefit the most overall.

Miriam B Michael, Siham M Mahgoub, Reiad Khan, Thomas A Mellman, Constance C Mere, Alem Mehari, Tammey J Naab, Uzoamake Nwagowugwu, Susan Ihaegwara, Celia J Maxwell, Absence of Antibody Reponses and Severe COVID-19 in Patients on Hemodialysis Following mRNA Vaccination, Open Forum Infectious Diseases, Volume 8, Issue 8, August 2021, ofab337, https://doi.org/10.1093/ofid/ofab337

A few COVID related case reports. This details 2 patients with ESRD on dialysis who had received a COVID vaccination at least 3 weeks prior to admission. Both patients presented with acute hypoxic respiratory failure, required transfer to a medical intensive care unit, and subsequently died due to respiratory failure. When evaluating their immunoglobulin G antibody levels against SARS-CoV spike protein, their index was <1.0, which was negative. This highlights the need for an intact immune system for mounting a response and the importance of herd immunity. 

Racette SD, Alexiev BA, Angarone MP, Bhasin A, Lima K, Jennings LJ, Balasubramanian S, Matsuoka AJ. Kikuchi-Fujimoto disease presenting in a patient with SARS-CoV-2: a case report. BMC Infect Dis. 2021 Aug 3;21(1):740. doi: 10.1186/s12879-021-06048-0. PMID: 34344305; PMCID: PMC8329643.

This one is a bit wild. Kikuchi syndrome is a self-limiting disorder with cervical adenopathy and systemic symptoms such as fever and malaise with a distinctive pathology on biopsy (histiocytic necrotizing lymphadenitis). This is a very rare disease. Anyways, this case report highlights a fairly young Indian physician with latent tuberculosis who had COVID-19 3 months prior to presentation (not severe) who came with fever, chills, and neck swelling. He got antibiotics for presumed folliculitis but did not improve so presented to the ED with ongoing fevers and tachycardia. He was found to have lymphopenia, reduced neutrophils (ANC 800), and elevated LDH 289. His work-up was negative and was sent for an excisional biopsy. His pre-biopsy SARS-CoV2 was positive. He got an excisional biopsy of his right cervical node, which showed follicular hyperplasia with expanded area of focal necrosis. CMV and SARS-CoV2 stains of the biopsy sample were negative and the findings were consistent with Kikuchi-Fujimoto disease. He did well with outpatient prednisone. It was presumed this was due to prior SARS-CoV2 infection. 

Samanta P, Clancy CJ, Marini RV, Rivosecchi RM, McCreary EK, Shields RK, Falcione BA, Viehman A, Sacha L, Kwak EJ, Silveira FP, Sanchez PG, Morrell M, Clarke L, Nguyen MH. Isavuconazole Is as Effective as and Better Tolerated Than Voriconazole for Antifungal Prophylaxis in Lung Transplant Recipients. Clin Infect Dis. 2021 Aug 2;73(3):416-426. doi: 10.1093/cid/ciaa652. PMID: 32463873; PMCID: PMC8326536.

This was a single-center, retrospective study evaluating voriconazole vs isavuconazole in lung transplant recipients for preventing invasive fungal infection. IFI was defined per the EORTC/MSG criteria. 300 patients were evaluated, with 144 receiving isavuconazole and the remaining 156 receiving voriconazole. While both groups were relatively similar, those in the isavuconazole group tended to require ECMO more frequently post-transplant, and all patients in the isavuconazole received inhaled amphotericin B compared to only 44% in the voriconazole group. Both received similar types of immunosuppression pre-transplant. At 12 months post-transplant, there was no difference in IFI rates between both groups (7% in the isa group and 8% in the vori group):

Breakthrough IFI occurred in 5 patients per group (total of 10), with multivariate logistic regression finding that IFI occurred when >7 units of RBC were received or when a mold-positive resp culture was found in either donor or recipient:

A lower number of patients in the isavuconazole group discontinued their antifungal prophylaxis and lower still had hepatotoxicity (15% vs 28%) or neurotoxicity (0% vs 13%) when compared to voriconazole. I am not surprised these were the results, though only one patient in the voriconazole group had an infection with mucor. I think isavu is a reasonable option for antifungal prophylaxis, though the single-center design and the fact all of the isavu had adjacent inhaled amphotericin B may limit its generalizability. I would not be surprised if in the next 5 years, most centers transition to isavuconazole, though. 

Butt JH, Fosbøl EL, Verhamme P, Gerds TA, Iversen K, Bundgaard H, Bruun NE, Larsen AR, Petersen A, Andersen PS, Skov RL, Gislason GH, Torp-Pedersen C, Køber L, Olesen JB. Dabigatran and the Risk of Staphylococcus aureus Bacteremia: A Nationwide Cohort Study. Clin Infect Dis. 2021 Aug 2;73(3):480-486. doi: 10.1093/cid/ciaa661. PMID: 32478836.

I have written a lot about staphylococcus aureus bacteremia. The issue with this bug is that it can create coagulases that form a staphylothrombin complex. Most anticoagulants do not affect this cascade, with the exception of dabigatran, which is a direct oral thrombin inhibitor. This study was a retrospective cohort study to study the incidence of SAB in patients with atrial fibrillation, comparing the use of dabigatran with anti-Xa inhibitors such as apixaban or rivaroxaban. Primary outcome was SAB, with patients who had IE being those who were in the hospital for 14 days (in Denmark, all those who had IE had to stay for 14 days). Only those who had a first time prescription were evaluated.  112,537 patients were evaluated. All groups were otherwise fairly similar, though the dabigatran group was a few years younger and tended to have more hypertension. By Cox regression analysis, treatment with dabigatran is associated with a lower incidence of bacteremia when compared to anti-Xa inhibitors (adjusted IRR 0.76, 95% CI 0.63-0.93). Dabigatran, but not apixaban, was associated with lower SAB incidence compared to rivaroxaban (aIRR 0.68, 95% CI 0.55-0.84 for dabigatran; aIRR 0.87, 95% CI 0.73-1.10):

Notably, when it came to IE, there was no association with lower rates of IE for the dabigatran group. The authors also performed multiple sensitivity analysis that included only patients between certain time periods, those who were anticoagulant-naive, those who had DOACs irrespective of the indication, and those who did not obtain edoxaban, and the results pretty much remained the same. In a way, I am not surprised that the rates of IE were not different, as the clot must have been fairly formed for vegetation to occur, as for the AC to not have any effect. As to why this is, I am unsure. Perhaps certain strains of Staph aureus are hell-bent on making coagulase. Is this practice changing? Not exactly, unless someone has an underlying condition that requires AC, but I do not see this being an add-on to someone with staph aureus bacteremia (I can see how this would cause more problems). 

Duval X, Le Moing V, Tubiana S, Esposito-Farèse M, Ilic-Habensus E, Leclercq F, Bourdon A, Goehringer F, Selton-Suty C, Chevalier E, Boutoille D, Piriou N, Le Tourneau T, Chirouze C, Seronde MF, Morel O, Piroth L, Eicher JC, Humbert O, Revest M, Thébault E, Devillers A, Delahaye F, Boibieux A, Grégoire B, Hoen B, Laouenan C, Iung B, Rouzet F; AEPEI-TEPvENDO study group. Impact of Systematic Whole-body 18F-Fluorodeoxyglucose PET/CT on the Management of Patients Suspected of Infective Endocarditis: The Prospective Multicenter TEPvENDO Study. Clin Infect Dis. 2021 Aug 2;73(3):393-403. doi: 10.1093/cid/ciaa666. PMID: 32488236.

PET-SCANs for IE. This is a multicenter prospective study to assess the diagnostic performance of F-FDG-PET/CT for patients with suspected native valve or prosthetic valve endocarditis. PET-CT was used for the valve itself, reading perivalvular uptake, as well as for the whole body. Two definitions of IE was used:

1. The modified Duke criteria (prior to scanning)
2. Using PET-CT. Any valvular uptake when it came to valves were considered major criteria. Emboli or aneurysms anywhere else were minor criteria. 

Gold standard was the Duke criteria. The authors calculated the net reclassification index, which estimates whether the addition of PET-CT either upgraded or downgraded patients within the modified Duke criteria. 140 patients were evaluated (70 prosthetic, 70 native). At 6 months, 68% had definitive IE, and 19% had possible IE by the modified Duke criteria. More patients with prosthetic valves had abnormal uptake in PET-CT (67% vs 24%), of which 14% in the PV group was attributed to the valve itself. PET-CT perivalvular uptake was considered a criterion for IE in 26 of the patients. Further, whole-body extracardiac uptake was found in 49% of patients. Overall, PET-CT added at least one duke criteria in 31% of patients, of which 23 were major and 21 were minor. This led to a modification in 21 patients of which 18 (a total of 13% within the cohort) were upgraded. Absolute NRI was 12.1%, 20% in prosthetic valve and 4.3% in native valve). 

37 patients had their management modified due to the PET-CT findings, 22 for antibiotics and 7 for surgery. Further 40% of patients benefited from PET-CT, largely due to the non contributing baseline echocardiography. The expense of PET-CT may be a huge limiting factor, though I can see some benefit of the application for native valve endocarditis. Prosthetic valve, given the underlying possible abnormal uptake due to the prosthetic valve itself, I am unsure of. 

Snyder RE, Cooksey GS, Kramer V, Jain S, Vugia DJ. West Nile Virus-Associated Hospitalizations, California, 2004-2017. Clin Infect Dis. 2021 Aug 2;73(3):441-447. doi: 10.1093/cid/ciaa749. PMID: 32525967.

This is more of an epidemiological study. This is a study evaluatiating the WNV-associated hospitalizations in California from 2004 through 2017. They included only residents of California. 3109 residents were hospitalized, all of which were initial hospitalizations (any one who had a code for WNV had their records evaluated to look at only their initial hospitalizations, rather than a subsequent admission where they just carried on the diagnosis). 94% of these admissions occurred between June and November, with a peak at around Aug/Sept:

Overall, the majority of patients were male and older than 60:

77% had West Nile non-neuroinvasive disease, with 21.4% having ICD codes for meningitis, seizures, paralysis, coma, or GBS. Roughly 80% had underlying conditions such as hypertension, cardiovascular disease, or diabetes. 47% were discharged to either rehab, SNF, or supportive home care and 8.8% died during their hospitalization. Multivariate analysis found that cardiovascular disease, CKD, male gender, immunosuppression, increasing age, and diabetes were associated with in-hospital death: 

Julia L Marcus, Kenneth Levine, Whitney C Sewell, Patricia Solleveld, Kenneth H Mayer, Douglas S Krakower, Switching From Tenofovir Disoproxil Fumarate to Tenofovir Alafenamide for Human Immunodeficiency Virus Preexposure Prophylaxis at a Boston Community Health Center, Open Forum Infectious Diseases, Volume 8, Issue 8, August 2021, ofab372, https://doi.org/10.1093/ofid/ofab372

This was a retrospective cohort study from a community health center evaluating the frequency and predictors of switching from Truvada (TDF/FTC) to Descovy (TAF/FTC) for PrEP after Descovy availability. Only patients assigned male at birth were evaluated, with primary outcome being TAF/FTC switch at the first 12 months of availability. They evaluated whether or not the switch was warranted by looking at those who had CrCl <60 or reduced bone density as well as cardiovascular risk factors. 2892 patients were included. 11% had hypertension, 0.6% had reduced bone density. In the first 12 months of Descovy availability, 343 (11.9%) switched from Truvada to Descovy. Of those who switched. 7% were indicated, 79% were unnecessary and 14% were potentially harmful. Risk of switching to Descovy was higher in Hispanics (HR 1.4, 95% CI 1-1.8), those who had medicare (HR 2.0, 95% CI 1.2-3.4), those whose CrCL <60 (HR 5.2, 95% CI 3.5-7.9), those with hypertension (HR 1.6, 95% CI 1.2-2.1), or diabetes (1.8, 95% CI 1.1-2.8). High total cholesterol, high LD were associated with reduced incidence of switching. I think the switch from TDF to TAF in terms of reduced nephrotoxicity and bone health are well known. I think it is also interesting they are looking at the metabolic implications of this, as TAF may adversely affect those as well. 

Li L, Hsu SH, Wang C, Li B, Sun L, Shi J, Ren Y, Wang J, Zhang X, Liu J. Characteristics of viral pneumonia in non-HIV immunocompromised and immunocompetent patients: a retrospective cohort study. BMC Infect Dis. 2021 Aug 6;21(1):767. doi: 10.1186/s12879-021-06437-5. PMID: 34362320; PMCID: PMC8343364.

This study evaluated the etiology of viral pneumonia in non-HIV immunocompromised and immunocompetent individuals using sputum and BAL samples. The authors evaluated patients who were admitted with CAP and they defined immunocompromise as those who had SOT, HSCT, BMT, were undergoing chemotherapy and had neutropenia, had an autoimmune disorder requiring steroids for at least >10mg/d for three weeks, had prior splenectomy or cirrhosis. 860 patients were included, of which 370 were immunocompromised. While the tables are quite busy, there were a few trends:

• During flu season, the most common viruses in the immunocompromised group were CMV (15%), influenza A (17%), and RSV (13%). In non-flu season, CMV accounted 43% of the cases
• Co-infections in the immunocompromised cohort was common, with 55% having other pathogens, the most common being PJP (22%), nosocomial pathogens (36%) and aspergillus (14%).
• Immunocompetent patients had high rates of influenza A (38%) and RSV (15%) during the flu season. During non-flu season, no virus dominated and it was a mix of rhinovirus, RSV, parainfluenza, or adenovirus.
• Co-infection was not very common in the immunocompetent (40%). 
• Patients with CMV tended to have CTD (43%), some sort of GN or nephrotic syndrome (22%), while those with influenza and other non-influenza tended to have ILD. 
• Surprisingly, those who had CTD or some sort of nephrotic/nephritis syndrome had higher mortalities (30% and 39%) compared to the rest of the cohort.

Again, the tables are interesting but quite busy, but those who are immunocompromised tended to have more CMV during non-flu season and during flu-season it accounted for a not insignificant proportion of cases. Notably, immunocompetent individuals tended to shed virus for a shorter period of time:

Tsai YH, Huang TY, Chen JL, Hsiao CT, Kuo LT, Huang KC. Bacteriology and mortality of necrotizing fasciitis in a tertiary coastal hospital with comparing risk indicators of methicillin-resistant Staphylococcus aureus and Vibrio vulnificus infections: a prospective study. BMC Infect Dis. 2021 Aug 9;21(1):771. doi: 10.1186/s12879-021-06518-5. PMID: 34372768.

This prospective study evaluated the association between LRINEC sores with MRSA and V. vulnificus necrotizing fasciitis and to determine the characteristics associated with survival. 184 patients were included, of which 20 died (mortality rate of 10.9%). V. vulnificus was the most common pathogen (22%) followed by staph aureus (17%), Beta-hemolytic strep (6.5%) and aeromonas (6%). Those who died were more likely to have low systolic blood pressure (50% vs 12.2%) and have bacteremia on presentation (65% vs 24%). Comparing V. vulnificus and MRSA, those with hepatic dysfunction, seawater or seafood contact, and SBP <90 were more likely to be infected with V. vulnificus (63% vs 31% for hepatic dysfunction, 93% vs 12.5% for seawater contact, and 22.5% vs 0 for SBP). When looking at the individual components of the LRNIEC score (CRP, WBC, Hbg, Sodium, Cr, glucose), more patients in the MRSA group had CRP >150, though the rest components did have no statistical difference. Notably, a higher proportion of the patients with MRSA had scores >6.

Platelet  counts <150k were more prevalent in the V. vulnificus cohort (26 vs 5). Overall, I think this is an interesting read but doesn’t necessarily add much. The numbers were a bit too low to look for any trends, and it solidifies that while the LRINEC score is useful for prognosticating, it may not be terribly useful for diagnostic purposes (sounds like something to write an in-depth post about).