COVID-19 Group 2

Synopsis: Currently, consensus recommendations, such as those from the Society of Critical Care Medicine, recommend that COVID-19 patients requiring mechanical ventilation should be managed using a lung-protective strategy (ex. lower tidal volume, higher PEEP, early proning, etc.). However, as experience treating COVID-19 grows, Gattinoni et al describe two different “phenotypes” of COVID-19 lung disease, denoted Type L and Type H. The presentation of these two phenotypes are hypothesized to depend on severity of the infection and host response, ventilatory response of the patient to hypoxemia, and time elapsed between onset of the disease and hospitalization.

Type L is an an early disease phenotype, characterized by

1. Low elastance (high or normal compliance)
2. Low ventilation to perfusion ratio (ventilation is NORMAL and hypoxemia is due to vasoplegia and loss of hypoxic vasoconstriction)
3. Low lung weight (or normal lung weight where most ground glass densities are located subpleurally or along lung fissures)
4. Low recruitability (due to near normal lung volumes and low amounts of “non-aerated” tissue).

The authors postulate that Type L develops in the beginning of the disease, when there is only a modest amount of interstitial lung edema, as well as vasoplegia, which accounts for the patient’s hypoxemia. In this phase, the patient compensates by increasing minute ventilation via increased tidal volumes and respiratory rate, driving down PaCO2. Since they are still able to generate high tidal volumes and negative intrathoracic pressure, these patients do not typically present with dyspnea. However, this increase in minute ventilation can facilitate patient self-inflicted lung injury (P-SILI) due to the combination of negative inspiratory intrathoracic pressure and increased lung permeability due to inflammation. Increased edema generated by this process facilitates the transition to the Type H phenotype.

Type H is similar to “classic” ARDS and is characterized by

1. High elastance (low compliance)
2. High right-to-left shunt (secondary to cardiac output perfusing non-aerated areas of lung)
3. High lung weight (due to increased edema)
4. High lung recruitability (which is why recruitment maneuvers and proning are useful)

Given the differences between these two phenotypes of COVID-19 lung disease, the authors propose the following treatments based on phenotype. Treatment of patients with Type L disease include reversing hypoxemia with increased FiO2 (especially if not dyspneic), consideration of non-invasive ventilation options in those who are dyspneic, and measurement of inspiratory pleural pressures, to prevent P-SILI. They also recommend modest increases in PEEP to prevent intrathoracic pressure changes (balanced with fact that high PEEPs might cause hemodynamic changes in the setting of normal compliance associated with Type L patients). The authors recommend early intubation once inspiratory pleural pressures increase. Initial ventilator management should include high tidal volumes and lower PEEP settings, if compliance remains normal. In addition, proning may be less useful, since lung volumes are theoretically normal. Once patients “convert” to Type H, traditional ARDS protocols for mechanical ventilation and adjuvant therapies should be employed.  

Synopsis: The association between SARS CoV-2 (virus causing COVID-19), underlying cardiovascular disease (CVD) and myocardial injury is not known. This single center, case-series analyzed outcomes among COVID-19 patients with underlying CVD and myocardial injury at the Seventh Hospital in Wuhan from Jan 23, 2020 to Feb 23, 2020. The main independent variables under evaluation were CVD and troponin-T (TnT) level. The primary outcome under evaluation was mortality. A total of 187 patients were included in the case-series; mean age was 58.5 (SD 14.66) and 48.7% were male. 66 (35.5%) participants had underlying CVD (HTN, CAD, CMP) and 52 (27.8%) had elevated TnT levels.

Compared with patients with normal TnT, those with higher TnT were significantly older (71.40 vs 53.53), male (65.4% vs 42.2%), had higher rates of comorbidities (HTN, CAD, CMP, diabetes, COPD, CKD) and ACEi/ARB use history (21.1% vs 5.9%). On admission, patients with elevated TnT had significantly higher levels of WBC, neutrophil count, AST, creatinine, other cardiac injury markers (CK-MB, NT-proBNP), inflammatory biomarkers (CRP, procalcitonin, globulin), D-dimer and longer PT. There was no significant difference in therapy received by patients with normal or elevated TnT levels, except for higher rates of glucocorticoid therapy (71.2% vs 51.1%) and mechanical ventilation (59.6% vs 10.4%) in those with elevated TnT.

Almost one-quarter (n = 43) of study participants died. Mortality was significantly higher in those with elevated TnT than in patients with normal TnT (59.6% vs 8.9%, p<.001). Mortality was documented in 7.62% (8 of 105) of patients with normal TnT and no CVD, 13.33% (4 of 30) with normal TnT and underlying CVD, 37.5% (6 of 16) with elevated TnT and no CVD and 69.44% (25 of 36) with elevated TnT and underlying CVD. Patients with higher TnT also developed significantly more frequent complications, including ARDS (57.7% vs 11.9%), malignant arrhythmias (11.5% vs 5.2%), acute coagulopathy (65.8% vs 20.0%) and AKI (36.8% vs 4.7%). Plasma TnT levels demonstrated significantly positive linear correlation with CRP (b=0.530, p<0.001) and NT-proBNP b=0.613, p<.001).

Myocardial injury was associated with significantly higher mortality among COVID-19 patients; worse in presence of underlying CVD as compared to absence of CVD. Mortality was significantly less in patients with underlying CVD and no myocardial injury. There are multiple confounding variables that should be considered when interpreting these findings including age, health status (comorbidities) and use of ACEi/ARB. Limitations include selection bias (only the sickest got hospitalized), small sample size and retrospective methodology with case-series analysis, thus an inability to make causal inferences about the relationship between underlying CVD and myocardial injury and mortality among patients with COVID-19. Though additional research is necessary to better understand this association, it is still critical to closely follow and replace electrolytes to avoid arrhythmias and cardiac dysfunction in patients with COVID-19. 

Synopsis: With the advent and rapid progression of SARS-COV-2 over the last few months have alerted medical professionals for extra-pulmonary manifestations of the disease. While respiratory complications remain the paramount cause of mortality in this heterogeneous group of patients, clinicians should be alert to other organ systems where the novel coronavirus can have a detrimental effect on. One area where this is apparent is increased presence of severe systemic clotting problems that these patients face. Initial autopsies in China have shown large clot burden not only in the lungs, but in kidneys, heart, and the liver.

Zhang and associates from Peking Union Medical College Hospital in the Sino–French New City Branch of Tongji Hospital in Wuhan have presented their experience with three SARS-COV-2 confirmed patients. Average age in their case series was 68 with all of their patients having 2+ co-morbidities including diabetes, hypertension, stroke, coronary artery disease. One of the patients on physical examination had bilateral upper and lower extremity ischemia. Subsequent dedicated brain imaging showed bilateral extensive cerebellar, cortical, and brainstem infarcts in all the patients. When the authors analyzed coagulation markers in these patients, they observed that all three had elevated antiphospholipid antibodies which included Anticardiolipin IgA, anti–β2 glycoprotein I, IgA, and IgG. Authors concluded that these antibodies can also arise transiently in

The presence of these antibodies may rarely lead to thrombotic events that are difficult to differentiate from other causes of multifocal thrombosis in critically patients, such as disseminated intravascular coagulation, heparininduced thrombocytopenia, and thrombotic microangiopathy. However, despite these observations one should be cautious to drawing any conclusions as antiphospholipid antibodies are well known to be transiently positive at the time of acute infectious illness. Additionally, antiphospholipid antibody titers and lab assay used were not reported in the article presented.

Synopsis: Convalescent plasma has shown promise in treating other viral infections (Ebola, H1N1, H5N1, SARS, MERS),raising the possibility that it may impact COVID-19. This uncontrolled case series of five critically ill patients (severe ARDS requiring mechanical ventilation and persistent high viral loads) in Shenzen, China from January 20 to March 25, 2020 monitored outcomes after transfusion with convalescent plasma.

Donors had to be asymptomatic for at least 10 days before donation. Recipients were 10 to 22 days into hospitalization before transfusion and transfusion occurred the same day as donation. As early as post-transfusion day (PTD) 1 for one patient and by PTD 12 for all five patients, viral load was no longer detectable and there was sustained elevation in virus-specific antibodies. All patients demonstrated improvement in PaO2/FiO2. Three patients were extubated at 2, 9, and 9 days post transfusion and were able to be discharged; two patients remained intubated at the conclusion of the study period. One of the intubated patients had been on ECMO and was able to be decannulated PTD 5. Clinical and laboratory values showed promising trends (improved fever, SOFA score, CRP, procalcitonin, and IL-6 levels).

Limitations include small sample size and non-representative demographics (all non-smokers, fewcomorbidities) and possible confounding of results by continued antiviral therapy after transfusion. Additionally, the two patients who remained intubated at the conclusion of the study had superimposed bacterial pneumonias, suggesting that patients with isolated respiratory sequelae of a purely viral infection may benefit most from this treatment. Though a limited case series, this study supports the examination of convalescent plasma in a clinical trial setting. 

Synopsis: The authors reviewed the literature regarding tracheostomy during the 2003 SARS outbreak and summarized findings from three case series, two case reports, and their institution's experience and plans for COVID-19. Their summary notes:

• Create a dedicated experienced team for performing trachs to reduce variability and improve efficiency and safety.
• Plan well prior to entering the room as it is hard to communicate with PPE/PAPR on.
• Most trachs were open tracheostomies performed bedside in a negative pressure ICU room. Pros: minimizes transport and connections on and off ventilator. Cons: limited space, suboptimal positioning, bringing surgical instruments bedside.
• They do not recommend perc trachs as it increases the amount of aerosolization (bronch, serial dilations) and more frequent connection/disconnection to vent when patients require high vent settings.
• If trach occurs in OR, it should be in a negative pressure OR with well-defined route to and from OR.
• Minimize aerosolization with paralysis, hold ventilation prior to tracheotomy, minimize suctioning. If suctioning, use a viral filter.
• Standard PPE (N95, goggles, cap, gown, gloves) was the minimum for performing tracheostomy. Some centers required shoe covers, face shields, and/or PAPR.
• Don/doff should be supervised by infection control nursing staff.
• There were no cases of SARS transmission to healthcare workers performing tracheostomies. 

Synopsis: Work from the team that performed the preclinical development of remdesivir has now published that an orally bioavailable antiviral drug shows promise in the fight against COVID-19 in animal models. This team led by Ralph Baric at UNC Chapel Hill and Mark Denison at Vanderbilt University with the support of an NIH grant from the University of Alabama Birmingham expanded upon previous findings that a drug out of the Emory Institute of Drug Development (EIDD) was able to block RNA replication in a broad spectrum of coronaviruses. The drug, 𝛽-DN4-hydroxycytidine (NHC, EIDD-1931), is a nucleoside analog that is selectively incorporates into viral RNA in place of cytidine or uridine, resulting in increased rates of lethal viral mutagenesis without inducing mutation in host cell RNA. In this new paper published in Science Translational Medicine, the authors first found that EIDD-1931 had antiviral activity against SARS-CoV-2 in an established human lung epithelial cell line as well as primary human airway epithelium cultures derived from patients undergoing surgery at UNC. High-fidelity deep sequencing approaches (Primer ID NGS) demonstrated that EIDD-1931 was effective against remdesivir-resistant infections in a manner consistent with viral mutagenesis. Given remdesivir works via a different mechanism (chain terminator of RNA synthesis), this data hints at a role for combined therapy in the future.

The authors then tested the prophylactic and therapeutic utility of an orally bioavailable prodrug form of NHC, EIDD-2801. Because there are no mouse models available that can recapitulate SARS-CoV2 pathogenesis, the authors used mice infected with mouse-adapted SARS-CoV and mice genetically modified to be susceptible to MERSCoV. These mice were treated with EIDD-2801 either two hours before infection exposure or 12, 24, or 48 hours after exposure and continued the drug BID through day 6 post infection. Pretreatment with EIDD-2801 prevented the weight loss and lung damage seen in control animals not treated with the drug. Further, EIDD-2801 also treated the infection if administered within 12 hours of exposure to the virus - including a reducing viral load, weight loss, lung damage, and defects in pulmonary function. Per the authors, symptoms appear at 1-2 days post infection in mice which correlates to day 7-10 post initiation of symptoms in humans. Regarding mechanism, viral genomic RNA by qRTPCR decreased in an initiation-time dependent manner in the lungs of EIDD-2801 treated mice, but the virus particles that remained were seemingly non-infectious as viral titers were reduced to the limit of detection by plaque assay. The authors then again used deep sequencing to find that EIDD-2801 induced mutagenesis in the targeted viral RNA but not host cell RNA. Interestingly, mutations included frequent sequence changes to stop codons consistent with their proposed mechanism of error catastrophe. Limitations included the previously noted lack of animal models of SARSCoV2 and that the 20-29 week old mice used here did not recapitulate the age-related increase in pathogenesis seen in humans.

In conclusion, this data suggests that NHC (EIDD-1931) and its orally bioavailable prodrug (EIDD-2801) are able to prevent viral replication and reduce lung damage due to SARS-CoV2. Although the prodrug was not explicitly tested against SARS-CoV2 in mice, NHC’s utility against SARS-CoV2 in human lung cells combined with promising data for the prodrug against related viruses in vivo proposes an exciting role for EIDD-2801 as a pill in the fight against COVID-19. In terms of timing, this drug would likely be similar to Oseltamivir (Tamiflu) for influenza in that the protective effect is maximized if administered closer to the onset of symptoms, however there is likely utility in severe cases where viral replication is extended. Commentary on the study can be found in AAAS’s EurekAlert! where one of the lead authors, UNC Assistant Professor Timothy Sheahan, noted, “With three novel human coronaviruses emerging in the past 20 years, it is likely that we will continue to see more. EIDD-2801 holds promise to not only treat COVID-19 patients today, but to treat new coronaviruses that may emerge in the future." This data supports that EIDD-2801 should be rapidly evaluated in non-human primates including a newly described primate model for SARS-CoV-2. Clinical trials to test the efficacy of the drug in humans are reportedly expected to begin later this spring.

Links: 

• Paper: https://stm.sciencemag.org/content/early/2020/04/03/scitranslmed.abb5883
• “preclinical development of remdesivir” - https://jvi.asm.org/content/93/24/e01348-19.long
• “previous findings” - https://jvi.asm.org/content/93/24/e01348-19.long
• “Eureka alert” - https://www.eurekalert.org/pub_releases/2020-04/uonc-ana040320.php
• “newly described” - https://www.biorxiv.org/content/10.1101/2020.03.21.001628v1