Weekly Articles 3/14/2021

Richard G Wunderink, Antoine Roquilly, Martin Croce, Daniel Rodriguez Gonzalez, Satoshi Fujimi, Joan R Butterton, Natasha Broyde, Myra W Popejoy, Jason Y Kim, Carisa De Anda, A Phase 3, Randomized, Double-Blind Study Comparing Tedizolid Phosphate and Linezolid for Treatment of Ventilated Gram-Positive Hospital-Acquired or Ventilator-Associated Bacterial Pneumonia, Clinical Infectious Diseases, 2021;, ciab032, https://doi.org/10.1093/cid/ciab032

This is a double-blind, multicenter, phase 3, randomized non-inferiority trial evaluating tedizolid and linezolid in hospital acquired and ventilator pneumonia. Patients were randomized in a 1:1 fashion to either a 7d course of tedizolid or a 10 day course of linezolid across 122 study sites. Pneumonia was diagnosed based on radiographic and clinical criteria with cultures obtained via ET aspirate, BAL, mini-BAL, or purulent sputum. Primary outcomes were 28-day all caused mortality and investigator-assessed clinical response at the test of cure visit (7-14 days after last infusion or time of failure) in the ITT population. 726 patients were enrolled, with baseline characteristics being well matched in both groups. In the ITT population, tedizolid was non-inferior to linezolid in 28 day all-cause mortality, though this was not the case for the investigator assessed clinical cure:

Linezolid also outperformed tedizolid in terms of clinical cure by pathogen in the ITT population, including MSSA and MRSA, which was statistically significant:

A post-hoc logistic regression model was used to assess factors that predict investigator assessed clinical response, including APACHE II score, geographic region, and baseline pathogen and no factors were found to be predictors of clinical response. I often find it difficult to interpret these types of studies, comparing one antibiotic to another one within the same class, as they tend to perform equally well. In this one, at least they both performed well enough in one of the primary outcomes, but when it comes to looking at the microorganisms cultured and those who achieved cure, tedizolid did not perform as expected. Indeed, many times these drugs are used with the idea of covering MRSA or VRE. While enterococcus was not really looked at, the performance of tedizolid when it comes to staph aureus certainly gives one pause. 

Carmo TA, Ferreira IB, Menezes RC, Telles GP, Otero ML, Arriaga MB, Fukutani KF, Neto LP, Agareno S, Filgueiras Filho NM, Andrade BB, Akrami KM. Derivation and Validation of a Novel Severity Scoring System for Pneumonia at Intensive Care Unit Admission. Clin Infect Dis. 2021 Mar 15;72(6):942-949. doi: 10.1093/cid/ciaa183. PMID: 32146482.

This is an observational, single center cohort study that aimed to create a new prediction score for patients with pneumonia admitted to the ICU. 200 patients were enrolled in the initial cohort followed by 362 patients in the external validation cohort. This score, which is known as the Pneumonia Shock score, was then compared to SAPS-3, qSOFA for ICU-specific scores and the CURB-65 and CRB-65. Primary outcome was ICU mortality. Stepwise multivariate logistic regression model yielded 8 variables associated with mortality:

2 points were given to age and vasopressor use, with the rest receiving 1 point, for a range of 0 to 10 points. In patients with score <2, mortality was 9.3% while in those whose score was >3, mortality was >26%. In the discovery cohort, the AUC was 0.80:

The Pneumonia shock outperformed the ICU and pneumonia scores when it came to predicting in-patient mortality:

I am always very wary about prediction scores. A lot of the pneumonia ones are very good for risk stratifying and assessing mortality in the ED. This one is nice in that it allows you to assess mortality risk in the ICU, but this is a single center study, so it is definitely not ready to be used in the clinical setting. 

Sabur NF, Brar MS, Wu L, Brode SK. Low-dose amikacin in the treatment of Multidrug-resistant Tuberculosis (MDR-TB). BMC Infect Dis. 2021 Mar 10;21(1):254. doi: 10.1186/s12879-021-05947-6. PMID: 33691624; PMCID: PMC7945371.

This was an observational cohort that evaluated a low dose amikacin (8mg/kg) compared to standard dose (15mg/kg) for MDR TB. This is based on the idea that aminoglycosides have concentration dependent activity, meaning the higher the concentration = greater killing. This was a study of 49 patients in a referral center in Canada for amikacin for TB. Mean duration of amikacin was 139 days, and starting dose was ~8.9mg/kg. A total of 12% of patients had sensorineural hearing loss and 23% had nephrotoxicity:

84% had successful treatment outcome and only 2% died during therapy:

Rates of ototoxicity ranges from 18% to up to 37%, so it seems a lower dose amikacin may be a reasonable option going forward.

Nickel K, Halfpenny NJA, Snedecor SJ, Punekar YS. Comparative efficacy, safety and durability of dolutegravir relative to common core agents in treatment-naïve patients infected with HIV-1: an update on a systematic review and network meta-analysis. BMC Infect Dis. 2021 Feb 26;21(1):222. doi: 10.1186/s12879-021-05850-0. PMID: 33637050; PMCID: PMC7908737.

This is one of those meta-analysis that would have been easier to understand if there had been a better image than this:

However, I found it interesting and is open access, so I encourage you to read it if you are interested in HIV-care. Ok, this is a meta-analysis that evaluated DTG against other guideline-recommended HIV therapies in treatment-naive patients. RCTs that compared treatments were included, and analyses were adjusted by NRTI backbone combinations. Third-drugs were compared to one another (raltegravir, bictegravir, dolutegravir, darunavir/ritonavir, Kaletra, atazanavir, elvitegravir, efavirenz):

’

20 studies were included. At week 96, a higher proportion of patients on DTG achieved viral suppression compared to ritonavir-boosted therapies, EFV, and EVG/c, and higher but not statistically significant compared to RPV, RAL and BIC. Ideally, we would have a better image, so here is one from the 2019 meta-analysis to illustrate my point (this 2019 one goes through week 48 rather than week 96;i’ll update it once BMC ID updates their crappy and meaningless image):

Both RAL and BIC were superior to the ritonavir-boosted regimens. DTG was also superior in those who had a VL >100k when compared to ritonavir-based regimens and raltegravir, but equal when compared to BIC, EFV, and EVG/c. DTG also achieved the greatest increase in CD4 compared to EFV, DRV/r, and BIC. DTG also had lower rates of discontinuation at week 96 compared to ritonavir-based regimens, EFV, RAL, and EVG/c. So this would continue to suggest that integrase inhibitors seem to be better than other third drugs for HIV when it comes to viral suppression and CD4 count improvements, though always be wary of fatty liver and weight gain.

Anesi JA, Lautenbach E, Tamma PD, Thom KA, Blumberg EA, Alby K, Bilker WB, Werzen A, Tolomeo P, Omorogbe J, Pineles L, Han JH. Risk Factors for Extended-Spectrum β-lactamase-Producing Enterobacterales Bloodstream Infection Among Solid-Organ Transplant Recipients. Clin Infect Dis. 2021 Mar 15;72(6):953-960. doi: 10.1093/cid/ciaa190. PMID: 32149327.

This is a multicenter, retrospective study that sought to evaluate the risk factors for ESBL-Enterobacteriales bacteremia in solid organ transplant patients. Case patients were compared to control patients with gram negative bacteremia that were not ESBL-producing organisms. 988 bacteremic episodes were evaluated, with most common organisms isolated being Kleb pneumo and E. coli. Multivariate analysis found that prior isolation of ESBL-organisms in the previous year, steroid-containing chronic immunosuppressive regimen, acute rejection treated with steroids in the past 3 months, exposure to third-gen cephalosporins and echinocandins in the prior 6 months, and exposure to TMP-SMX in the prior 6 months were associated with increased risk of ESBL-bacteremia:

Most of these are not surprising. The results remained fairly equivalent when adjusted for year of diagnosis. In sub-group analysis, unique risk factors for each organ transplant emerged including exposure to voriconazole/posaconazole in the liver transplant group, exposure to daptomycin or SSTI with ESBL organisms in the kidney transplant group, African American/Black race in the heart transplant, and need for re-operation within 4 weeks of transplant and exposure to cefepime in the prior 6 months:

Given the association of TMP-SMX, a multivariate analysis found that prophylaxis was significantly associated with ESBL-bacteremia (aOR 2.10, 95% CI 1.03-4.28) compared to just exposure (aOR 1.98, 95% CI 0.97-4.01). Most of these risk factors are not terribly surprising; i.e the more antibiotics you use and the longer you’re in the hospital, the higher your risk of ESBL. TMP-SMX exposure and echinocandins are a surprising risk factor, though this could relate to microbiome disruption, as TMP-SMX can also have a broad spectrum of activity. Basiliximab, at least as induction in heart transplant patients, was associated with a protective effect.

Kassis C, Durkin M, Holbrook E, Myers R, Wheat L. Advances in Diagnosis of Progressive Pulmonary and Disseminated Coccidioidomycosis. Clin Infect Dis. 2021 Mar 15;72(6):968-975. doi: 10.1093/cid/ciaa188. PMID: 32108231.

This is a retrospective study that evaluated the sensitivity and specificity of Coccidiodes antigen EIA in combination with antibody detection. 158 cases were evaluated along with 487 control cases. Immunodiffusion was most frequently positive in cases, followed by complement fixation and then antigone detection:

The combination of antibody and antigen yielded a sensitivity of 93%, specificity of 96.1%, PPV 88.5% and NPV 95.1%, compared to antibody alone which yielded a sensitivity of 84.2% and specificity of 96.5%. Not surprising, the sensitivity and specificity of antibody testing was higher in proven cases (87.2% sensitivity in proven vs 76.1% in probable cases) as well as for antigen testing (55.4% vs 58.7% in proven and probable, respectively). The combined antigen/antibody sensitivity was 91.6%, compared to 87.2% for antibody alone in proven cases, while the combination testing sensitivity was 95% in the probable cases compared to 76% for antibody alone. Sensitivity was also higher for antibody detection in disseminated cases (92%) compared to pulmonary cases (80%), which was similar for antigen detection (79% in disseminated vs 42% in pulmonary). Antibody and antigen testing was also influenced by immunosuppression, with those who were immunocompetent having higher sensitivities for both. 

There are a lot of numbers in this article (and I hate numbers), but the crux of it seems to be that the combination of antigen and antibody testing yields a higher sensitivity in patients with high pre-test probability of cocci. Further, based on this, seems that immunodiffusion has higher sensitivities than complement fixation. 

COVID Articles

Simonovich VA, Burgos Pratx LD, Scibona P, Beruto MV, Vallone MG, Vázquez C, Savoy N, Giunta DH, Pérez LG, Sánchez MDL, Gamarnik AV, Ojeda DS, Santoro DM, Camino PJ, Antelo S, Rainero K, Vidiella GP, Miyazaki EA, Cornistein W, Trabadelo OA, Ross FM, Spotti M, Funtowicz G, Scordo WE, Losso MH, Ferniot I, Pardo PE, Rodriguez E, Rucci P, Pasquali J, Fuentes NA, Esperatti M, Speroni GA, Nannini EC, Matteaccio A, Michelangelo HG, Follmann D, Lane HC, Belloso WH; PlasmAr Study Group. A Randomized Trial of Convalescent Plasma in Covid-19 Severe Pneumonia. N Engl J Med. 2021 Feb 18;384(7):619-629. doi: 10.1056/NEJMoa2031304. Epub 2020 Nov 24. PMID: 33232588; PMCID: PMC7722692.

This is a double-blind, placebo-controlled, multicenter trial in Argentina that assigned patients in a 2:1 fashion to convalescent plasma and placebo in patients with severe COVID pneumonia. Patients needing mechanical ventilation and having multiorgan failure were excluded. Minimum antibody titer was 1:400. Primary outcome was clinical status at 30 days after intervention as represented by the WHO ordinal scale. 334  patients were enrolled. Patients in the convalescent plasma were more likely to have COPD and had higher ferritin levels, but were otherwise fairly well balanced. There was no statistical difference between convalescent plasma and placebo when it came to the distribution of clinical outcomes in the ordinal scale:

This held true after adjustment for COPD, sex, and tobacco history (OR 0.92, 95% CI 0.59 to 1.42). 

This is another nail in the coffin of convalescent plasma, unfortunately, though this was a study on more severe COVID so it may bias the results in that way. 

Kalil AC, Patterson TF, Mehta AK, Tomashek KM, Wolfe CR, Ghazaryan V, Marconi VC, Ruiz-Palacios GM, Hsieh L, Kline S, Tapson V, Iovine NM, Jain MK, Sweeney DA, El Sahly HM, Branche AR, Regalado Pineda J, Lye DC, Sandkovsky U, Luetkemeyer AF, Cohen SH, Finberg RW, Jackson PEH, Taiwo B, Paules CI, Arguinchona H, Erdmann N, Ahuja N, Frank M, Oh MD, Kim ES, Tan SY, Mularski RA, Nielsen H, Ponce PO, Taylor BS, Larson L, Rouphael NG, Saklawi Y, Cantos VD, Ko ER, Engemann JJ, Amin AN, Watanabe M, Billings J, Elie MC, Davey RT, Burgess TH, Ferreira J, Green M, Makowski M, Cardoso A, de Bono S, Bonnett T, Proschan M, Deye GA, Dempsey W, Nayak SU, Dodd LE, Beigel JH; ACTT-2 Study Group Members. Baricitinib plus Remdesivir for Hospitalized Adults with Covid-19. N Engl J Med. 2021 Mar 4;384(9):795-807. doi: 10.1056/NEJMoa2031994. Epub 2020 Dec 11. PMID: 33306283; PMCID: PMC7745180.

This is the second part of the remdesivir trial, ACTT-2. This study evaluated remdesivir and baricitinib, a PO JAK 1 and 2 inhibitor that inhibits several cytokines involved in COVID-19 (including IL-2, IL-6, IL-10, IFN-gamma). This was a double-blind, placebo-controlled trial in 67 sites worldwide, and patients were randomized in a 1:1 fashion to receive either remdesivir by itself or remdesivir with baricitinib. Remdesivir was loaded in 200mg on day one followed by 100mg daily from days 2-10. Baricitinib was given as a 4mg dose for 14 days. Primary outcome was time to recovery at day 28 as defined by an eight-category ordinal scale. 1033 patients underwent randomization, with both groups being fairly well balanced. 21% of patients required non-invasive ventilation or HFNC, while 11% required either mechanical ventilation or ECMO. Recovery ratio was better in the combination group (1.16, 95% CI 1.01 to 1.32) with a median recovery of one day faster in the combination group (median 7d vs median 8d). The highest difference was seen in those who at baseline required noninvasive ventilation, with a median time to recovery in the combination group being 10 days compared to 18 days in the monotherapy group (recovery rate 1.51, 95% CI 1.1-2.08). The recovery rate was not statistically significant in those without oxygen, with supplemental oxygen, and those in mechanical ventilation.

Ok, busy table. This may look a bit better:

Mortality at day 28 was 5.1% in the combination group vs 7.8% in the monotherapy group (HR 0.65, 95% 0.39 to 1.09). There was no significant difference in adverse events. Given the broad immunosuppressive actions of this JAK inhibitor, I should not be surprised this actually worked, though there is some concern for possible thrombosis with such drugs which may prevent its broad use (on top of cost). There is another trial with this drug ongoing (NCT04421027) but given that steroids worked about the same and are cheaper and more well known, I am not sure how this will catch on. 

Thelen JM, Buenen AGN, van Apeldoorn M, Wertheim HF, Hermans MHA, Wever PC. Community-acquired bacteraemia in COVID-19 in comparison to influenza A and influenza B: a retrospective cohort study. BMC Infect Dis. 2021 Feb 22;21(1):199. doi: 10.1186/s12879-021-05902-5. PMID: 33618663; PMCID: PMC7897875.

This is a retrospective study comparing patients with influenza A during the 2015/2016 season or 2016/2017 season, influenza B during the 2017/2018 season, and COVID-19 in terms of rates of bacteremia. 1324 patients were included, and overall were fairly well balanced, and after correcting for  contaminants, the occurrence of relevant bacteremia was statistically lower in those with COVID-19:

Furthermore, COVID patients with bacteremia were older (median 83yo vs 70yo p-value 0.03), with overall mortality being higher for COVID patients than for those with influenza. In those with relevant bacteremia, the mortality rates between the groups are not statistically different:

This is a 2-center study, and notably, cultures were not always obtained prior to antibiotic use, which may affect the results. Either way, this would seem to support the practice of not routinely using antibiotics in COVID patients, though the severity of these patients are not described which would be helpful to risk stratification in terms of who would benefit from antibiotics. 

Gokhale Y, Mehta R, Kulkarni U, Karnik N, Gokhale S, Sundar U, Chavan S, Kor A, Thakur S, Trivedi T, Kumar N, Baveja S, Wadal A, Kolte S, Deolankar A, Pednekar S, Kalekar L, Padiyar R, Londhe C, Darole P, Pol S, Gokhe SB, Padwal N, Pandey D, Yadav D, Joshi A, Badgujar H, Trivedi M, Shah P, Bhavsar P. Tocilizumab improves survival in severe COVID-19 pneumonia with persistent hypoxia: a retrospective cohort study with follow-up from Mumbai, India. BMC Infect Dis. 2021 Mar 5;21(1):241. doi: 10.1186/s12879-021-05912-3. PMID: 33673818; PMCID: PMC7934984.

This is a single-center retrospective study evaluating a single dose of 400mg tocilizumab in COVID patients with persistent hypoxia, bilateral pulmonary infiltrates, and elevated inflammatory markers. These patients were compared to a historical group that only required 2 of the above inclusion criteria (hypoxia/infiltrates). Primary outcome was death or recovery. 269 patients were included, of which 118 were historical controls. Patients in the tocilizumab group tended to have more co-morbidities, have lower O2 saturation rates, higher RR, and higher rates of non-invasive and mechanical ventilation:

Mortality was 53% in the tocilizumab group compared to 63% in the historical control group, with median survival being longer in the tocilizumab group (18d vs 9d, p=0.007):

On multivariate analysis, only tocilizumab use and oxygen saturation were associated with improved mortality:

Given the single center nature of the study as well as its reliance on historical controls, I’d be wary about its generalizability. Id take a look at:

Rosas IO, Bräu N, Waters M, Go RC, Hunter BD, Bhagani S, Skiest D, Aziz MS, Cooper N, Douglas IS, Savic S, Youngstein T, Del Sorbo L, Cubillo Gracian A, De La Zerda DJ, Ustianowski A, Bao M, Dimonaco S, Graham E, Matharu B, Spotswood H, Tsai L, Malhotra A. Tocilizumab in Hospitalized Patients with Severe Covid-19 Pneumonia. N Engl J Med. 2021 Feb 25. doi: 10.1056/NEJMoa2028700. Epub ahead of print. PMID: 33631066.

Known as COVACTA, a phase 3, international, randomized, double-blind, placebo-controlled trial evaluating tocilizumab to placebo in patients with severe COVID pneumonia. Inclusion criteria included O2 saturation <93% or P/F ratio <300. Patients were randomized in a 2:1 fashion to tocilizumab 8mg/kg or max of 800mg single infusion or placebo, with the option of repeating dose at 8-24hrs if there was no improvement. Primary endpoint was clinical status at day 28 based on a 7 day ordinal scale. 452 patients were randomized, and 438 were included in the mITT population. Groups were fairly well balanced, although more patients in the placebo group received steroids (29% vs 19%). Clinical status at day 28 for the tocilizumab group was a median of 1 compared to 2 in the placebo group (difference -1, 95% CI -2.5 to 0):

Moreover, there was no significant differences in mortality between groups, although tocilizumab tended to favor days to improvement by 2 categories in the ordinal scale, and they had lower ICU stay days, lower incidence of mechanical ventilation, and lower incidence of clinical failure in those not in mechanical ventilation:

There were also no significant differences in adverse events between groups. ‘Tis unfortunate for tocilizumab, but while IL-6 is one of the major drivers of severity in these patients, it is not the only one so I am not surprised that targeting that one interleukin would not give favorable outcomes. There were trends towards better outcomes in terms of ICU length of stay, and decreased incidence of going to the ICU or mechanical intubation, but this study was not powered to look at these, so those results would be speculative, at this juncture. One thing to note is the data was analyzed by the sponsor, so take that as you will. 

REMAP-CAP Investigators, Gordon AC, Mouncey PR, Al-Beidh F, Rowan KM, Nichol AD, Arabi YM, Annane D, Beane A, van Bentum-Puijk W, Berry LR, Bhimani Z, Bonten MJM, Bradbury CA, Brunkhorst FM, Buzgau A, Cheng AC, Detry MA, Duffy EJ, Estcourt LJ, Fitzgerald M, Goossens H, Haniffa R, Higgins AM, Hills TE, Horvat CM, Lamontagne F, Lawler PR, Leavis HL, Linstrum KM, Litton E, Lorenzi E, Marshall JC, Mayr FB, McAuley DF, McGlothlin A, McGuinness SP, McVerry BJ, Montgomery SK, Morpeth SC, Murthy S, Orr K, Parke RL, Parker JC, Patanwala AE, Pettilä V, Rademaker E, Santos MS, Saunders CT, Seymour CW, Shankar-Hari M, Sligl WI, Turgeon AF, Turner AM, van de Veerdonk FL, Zarychanski R, Green C, Lewis RJ, Angus DC, McArthur CJ, Berry S, Webb SA, Derde LPG. Interleukin-6 Receptor Antagonists in Critically Ill Patients with Covid-19. N Engl J Med. 2021 Feb 25. doi: 10.1056/NEJMoa2100433. Epub ahead of print. PMID: 33631065.

More Tocilizumab. This is part of the REMAP-CAP trial, which is an international, adaptive platform trial that evaluated several treatment strategies for severe pneumonia. There were several domains included within this trial, including immunosuppression, antiviral, etc, and this one looked at the immunomodulatory domain of the trial. Patients with severe COVID pneumonia requiring ICU care were eligible to be enrolled. The domain included 5 interventions; tocilizumab, sarilumab (IL-6 antagonist), anakinra (IL-1 antagonist), interferon beta-1a, and control. Tocilizumab was given at a dose of 8mg/kg up to 800mg once with possible re-dosing 12-24hrs later, while sarilumab was given at a dose of 400mg once only. Primary outcome was the number of respiratory and cardiovascular organ support-free days at day 21 based on an ordinal score. Statistical analysis was done using a Bayesian cumulative logistic model, which calculates posterior probability distributions of organ-support free days. 895 patients were randomized (366 to tocilizumab, 48 to sarilumab, 412 to control, 69 to other arms. Patients were fairly well balanced, though those in the Tocilizumab group had a higher APACHE II score (13 vs 10 compared to sarilumab), higher CRP, and shorter time from ICU admission to enrollment. The median number of organ support free days was 10 in the tocilizumab group, 11 in the sarilumab group, and 0 in the control group (aOR 1.64, 95% CI 1.25 to 2.14 for tocilizumab, 1.76, 95% CI 1.17 to 2.91 for sarilumab compared to control):

This yielded a posterior probabilities of superiority of >99.9% for tocilizumab and 99.5% for sarilumab. Pooled mortality was lower in the IL-6 group (27% vs 36%): 

I’ll be honest, some of this statistical wizardry sometimes goes over my head, but one of the keys here was the relatively early enrollment of patients within 24 hours of organ support, which may have allowed for a better result compared to other trials. Further, this was a VERY sick population of patients, so perhaps this led to the seen outcome. I am still having issues making sense of these results, so I’ll leave people smarter than me to make a recommendation, but I still do not see myself reaching out for tocilizumab when we have the ‘Roids.