Merge pull request #864 from rando2/jf-edits

Incorporate new data from RECOVERY trial

content/20.pharmaceuticals.md

@@ -149,7 +149,7 @@ The effectiveness of favipiravir for treating patients with COVID-19 is currentl
 An open-label, nonrandomized, before-after controlled study was recently conducted [@doi:10.1016/j.eng.2020.03.007].
 The study included 80 COVID-19 patients (35 treated with favipiravir, 45 control) from the isolation ward of the National Clinical Research Center for Infectious Diseases (The Third People’s Hospital of Shenzhen), Shenzhen, China.
 The patients in the control group were treated with other antivirals, such as lopinavir and ritonavir.
-It should be noted that although the control patients received antivirals, two subsequent large-scale analyses, the WHO Solidarity trial and the RECOVERY trial, identified no effect of lopinavir or of a lopinavir-ritonavir combination, respectively, on the metrics of COVID-19-related mortality that each assessed [@doi:10.1056/NEJMoa2023184; @doi:10/fnx2; @doi:10.1056/NEJMe2034294].
+It should be noted that although the control patients received antivirals, two subsequent large-scale analyses, the WHO Solidarity trial and the Randomized Evaluation of COVID-19 Therapy (RECOVERY) trial, identified no effect of lopinavir or of a lopinavir-ritonavir combination, respectively, on the metrics of COVID-19-related mortality that each assessed [@doi:10.1056/NEJMoa2023184; @doi:10/fnx2; @doi:10.1056/NEJMe2034294].
 Treatment was applied on days 2-14; treatment stopped either when viral clearance was confirmed or at day 14.
 The efficacy of the treatment was measured by, first, the time until viral clearance using Kaplan-Meier survival curves, and, second, the improvement rate of chest computed tomography (CT) scans on day 14 after treatment.
 The study found that favipiravir increased the speed of recovery, measured as viral clearance from the patient by RT-PCR, with patients receiving favipiravir recovering in four days compared to 11 days for patients receiving antivirals such as lopinavir and ritonavir.
@@ -218,7 +218,7 @@ This large-scale, open-label trial enrolled 11,330 adult in-patients at 405 hosp
 Patients were randomized in equal proportions into four experimental conditions and a control condition, corresponding to four candidate treatments for COVID-19 and SOC, respectively; no placebo was administered.
 The 2,750 patients in the remdesivir group were administered 200 mg intravenously on the first day and 100 mg on each subsequent day until day 10 and assessed for in-hospital death (primary endpoint), duration of hospitalization, and progression to mechanical ventilation.
 There were also 2,708 control patients who would have been eligible and able to receive remdesivir were they not assigned to the control group.
-A total of 604 patients among these two cohorts deceased during initial hospitalization, with 301 in the remdesivir group and 303 in the control group.
+A total of 604 patients among these two cohorts died during initial hospitalization, with 301 in the remdesivir group and 303 in the control group.
 The rate ratio of death between these two groups was therefore not significant (0.95, _p_ = 0.50), suggesting that the administration of remdesivir did not affect survival.
 The two secondary analyses similarly did not find any effect of remdesivir.
 Additionally, the authors compared data from their study with data from three other studies of remdesivir (including [@doi:10.1056/NEJMoa2007764]) stratified by supplemental oxygen status.
@@ -457,7 +457,7 @@ All of these factors could potentially influence treatment outcome.
 Furthermore, the authors acknowledge that the effect of the drugs might be different in females and pediatric subjects, since these subjects were not part of the study.
 The reported result that HCQ + AZ is safer than HCQ contradicts the findings of the previous large-scale analysis of twenty years of records that found HCQ + AZ to be more frequently associated with cardiac arrhythmia than HCQ alone [@doi:10.1101/2020.04.08.20054551]; whether this discrepancy is caused by the pathology of COVID-19, is influenced by age or sex, or is a statistical artifact is not presently known.
 
-Finally, findings from the Randomized Evaluation of COVID-19 Therapy (RECOVERY) trial were released on October 8, 2020.
+Finally, findings from the RECOVERY trial were released on October 8, 2020.
 This study used a randomized, open-label design to study the effects of HCQ compared to SOC at 176 hospitals in the United Kingdom [@doi:10.1056/NEJMoa2022926].
 This large study enrolled 11,197 hospitalized patients whose physicians believed it would not harm them to participate.
 Patients were randomized into either the control group or one of the treatment arms, with twice as many patients enrolled in the control group as any treatment group.
@@ -597,8 +597,8 @@ Often used to treat conditions such as RA [@doi:10.1101/cshperspect.a016295], TC
 It has also been approved to treat CRS caused by CAR-T treatments [@url:https://dailymed.nlm.nih.gov/dailymed/drugInfo.cfm?setid=2e5365ff-cb2a-4b16-b2c7-e35c6bf2de13].<!-- To Do: add manual ref-->
 While secretion of IL-6 can be associated with chronic conditions, it is a key player in the innate immune response and is secreted by macrophages in response to the detection of pathogen-associated molecular patterns and damage-associated molecular patterns [@doi:10.1101/cshperspect.a016295].
 An analysis of 191 in-patients at two Wuhan hospitals revealed that blood concentrations of IL-6 differed between patients who did and did not recover from COVID-19.
-Patients who ultimately deceased had higher IL-6 levels at admission than those who recovered [@doi:10/ggnxb3].
-Additionally, IL-6 levels remained higher throughout the course of hospitalization in the patients who ultimately deceased [@doi:10/ggnxb3].
+Patients who ultimately died had higher IL-6 levels at admission than those who recovered [@doi:10/ggnxb3].
+Additionally, IL-6 levels remained higher throughout the course of hospitalization in the patients who ultimately died [@doi:10/ggnxb3].
 This finding provided some early evidence that COVID-19 deaths may be induced by the hyperactive immune response, often referred to as CRS or cytokine storm syndrome (CSS), as IL-6 plays a key role in this response [@doi:10.1128/mmbr.05015-11].
 In this context, the observation of elevated IL-6 in patients who died may reflect an over-production of proinflammatory interleukins, suggesting that TCZ could potentially palliate some of the most severe symptoms of COVID-19 associated with increased cytokine production.
 
@@ -624,7 +624,7 @@ Therefore, no conclusions can be drawn about its efficacy for the treatment of C
 However, early interest in TCZ as a possible treatment for COVID-19 emerged from a very small retrospective study in China that examined 20 patients with severe symptoms in early February 2020 and reported rapid improvement in symptoms following treatment with TCZ [@doi:10.1073/pnas.2005615117].
 Subsequently, a number of retrospective studies have been conducted in several countries.
 Many studies use a retrospective, observational design, where they compare outcomes for COVID-19 patients who received TCZ to those who did not over a set period of time.
-For example, one of the largest retrospective, observational analyses released to date [@doi:10/d2pk], consisting of 1351 patients admitted to several care centers in Italy, compared the rates at which patients who received TCZ deceased or progressed to invasive medical ventilation over a 14-day period compared to patients receiving only SOC.
+For example, one of the largest retrospective, observational analyses released to date [@doi:10/d2pk], consisting of 1351 patients admitted to several care centers in Italy, compared the rates at which patients who received TCZ died or progressed to invasive medical ventilation over a 14-day period compared to patients receiving only SOC.
 Under this definition, SOC could include other drugs such as HCQ, azithromycin, lopinavir-ritonavir or darunavir-cobicistat, or heparin.
 While this study was not randomized, a subset of patients who were eligible to receive TCZ were unable to obtain it due to shortages; however, these groups were not directly compared in the analysis.
 After adjusting for variables such as age, sex, and SOFA (sequential organ failure assessment) score, they found that patients treated with TCZ were less likely to progress to invasive medical ventilation and/or death (adjusted HR = 0.61, CI 0.40-0.92, _p_ = 0.020); analysis of death and ventilation separately suggests that this effect may have been driven by differences in the death rate (20% of control versus 7% of TCZ-treated patients).
@@ -657,7 +657,6 @@ Taken together, these findings suggest that future clinical trials of TCZ may wa
 However, these studies should be approached with caution, not only because of the small number of patients enrolled and the retrospective design, but also because they performed a large number of statistical tests and did not account for multiple hypothesis testing.
 In general, caution must be exercised when interpreting subgroup analyses after a primary combined endpoint analysis. <!--Need to be careful about overinterpreting subgroup analyses. Combined endpoints (of 2 events, like death OR ICU admission) are also problematic, though they are often used to increase power.-->
 These last findings highlight the need to search for a balance between impairing a harmful immune response, such as the one generated during CRS/CSS, and preventing the worsening of the clinical picture of the patients by potential new viral infections.
-
 Though data about TCZ for COVID-19 is still only just emerging, some meta-analyses and systematic reviews have investigated the available data.
 One meta-analysis [@doi:10.2139/ssrn.3642653] evaluated 19 studies published or released as preprints prior to July 1, 2020 and found that the overall trends were supportive of the frequent conclusion that TCZ does improve survivorship, with a significant HR of 0.41 (_p_ < 0.001).
 This trend improved when they excluded studies that administered a steroid alongside TCZ, with a significant HR of 0.04 (_p_ < 0.001).
@@ -667,7 +666,16 @@ Although these estimates are similar, it is important to note that they are draw
 A different systematic review of studies investigating TCZ treatment for COVID-19 analyzed 31 studies that had been published or released as pre-prints and reported that none carried a low risk of bias [@doi:10.1016/j.pulmoe.2020.07.003].
 Therefore, the present evidence is not likely to be sufficient for conclusions about the efficacy of TCZ.
 
-Additionally, there are possible risks associated with the administration of TCZ for COVID-19.
+On February 11, 2021, a preprint describing the first randomized control trial of TCZ was released as part of the RECOVERY trial [@doi:10.1101/2021.02.11.21249258].
+Of the 21,550 patients enrolled in the RECOVERY trial at the time, 4,116 adults hospitalized with COVID-19 across the 131 sites in the United Kingdom were assigned to the arm of the trial evaluating the effect of TCZ.
+Among them, 2,022 were randomized to receive TCZ and 2,094 were randomized to SOC, with 79% of patients in each group available for analysis at the time that the initial report was released.
+The primary outcome measured was 28-day mortality, and TCZ was found to reduce 28-day mortality from 33% of patients receiving SOC alone to 29% of those receiving TCZ, corresponding to a rate ratio of 0.86 (95% CI 0.77-0.96; _p_ = 0.007).
+TCZ was also significantly associated with the probability of hospital discharge within 28 days for living patients, which was 47% in the SOC group and 54% in the TCZ group (rate ratio 1.22, 95% CI 1.12-1.34, _p_ < 0.0001).
+A potential statistical interaction between TCZ and corticosteroids was observed, with the combination providing greater mortality benefits than TCZ alone, but the authors note that caution is advisable in light of the number of statistical tests conducted.
+Combining the RECOVERY trial data with data from seven smaller randomized control trials indicates that TCZ is associated with a 13% reduction in 28-day mortality (rate ratio 0.87, 95% CI 0.79-0.96, _p_ = 0·005) [@doi:10.1101/2021.02.11.21249258].
+While this initial report did not include the full results expected from the RECOVERY trial, this large-scale, randomized controlled trial therefore offers strong evidence that TCZ may offer benefits for COVID-19 patients, even at this initial stage of analysis.
+
+There are possible risks associated with the administration of TCZ for COVID-19.
 TCZ has been used for over a decade to treat RA [@doi:10.1177/1759720X18798462], and a recent study found the drug to be safe for pregnant and breastfeeding women [@doi:10.1093/rheumatology/kez100].
 However, TCZ may increase the risk of developing infections [@doi:10.1177/1759720X18798462], and RA patients with chronic hepatitis B infections had a high risk of hepatitis B virus reactivation when TCZ was administered in combination with other RA drugs [@doi:10.1111/1756-185X.13010].
 As a result, TCZ is contraindicated in patients with active infections such as tuberculosis [@doi:10.1186/s12967-020-02339-3].
@@ -677,7 +685,8 @@ Reactivation of hepatitis B and herpes simplex virus 1 was also reported in a sm
 A July 2020 case report described negative outcomes of two COVID-19 patients after receiving TCZ, including one death; however, both patients were intubated and had entered septic shock prior to receiving TCZ [@doi:10.1016/j.chest.2020.04.024], likely indicating a severe level of cytokine production.
 Additionally, D-dimer and sIL2R levels were reported by one study to increase in patients treated with TCZ, which raised concerns because of the potential association between elevated D-dimer levels and thrombosis and between sIL2R and diseases where T-cell regulation is compromised [@doi:10.1016/j.chest.2020.06.006].
 An increased risk of bacterial infection was also identified in a systematic review of the literature, based on the unadjusted estimates reported [@doi:10.2139/ssrn.3642653].
-TCZ administration to COVID-19 patients is not without risks, may introduce additional risk of developing secondary infections, and should be approached especially cautiously for patients who have latent viral infections.
+In the RECOVERY trial, however, only three out of 2,022 participants in the group receiving TCZ developed adverse reactions determined to be associated with the intervention, and no excess deaths were reported [@doi:10.1101/2021.02.11.21249258].
+TCZ administration to COVID-19 patients is not without risks and may introduce additional risk of developing secondary infections; however, while caution may be prudent when treating patients who have latent viral infections, the results of the RECOVERY trial indicate that adverse reactions to TCZ are very rare among COVID-19 patients broadly.
 
 In summary, approximately 33% of hospitalized COVID-19 patients develop ARDS [@doi:10.1186/s13054-020-03240-7], which is caused by an excessive early response of the immune system which can be a component of CRS/CSS [@doi:10.1016/j.chest.2020.06.006; @doi:10.1186/s12967-020-02339-3].
 This overwhelming inflammation is triggered by IL-6.
@@ -727,7 +736,7 @@ It was found that the different antibody isolates did not target the same epitop
 It was also demonstrated that binding affinity of the antibodies does not reflect their capability to compete with ACE2 binding.
 Furthermore, no conclusions about correlations between the severity of disease and the ability to produce nAbs can be drawn at this point.
 Rather, higher nAb titers were more frequently found in patients with severe disease.
-Correspondingly, higher levels of anti-spike IgG were observed in patients that deceased from infection compared to patient that recovered [@doi:10.1172/jci.insight.123158].
+Correspondingly, higher levels of anti-spike IgG were observed in patients that died from infection compared to patient that recovered [@doi:10.1172/jci.insight.123158].
 
 Results from the SARS and MERS epidemics thus provide valuable lessons for the design of nAbs for the current outbreak.
 The findings for SARS-CoV-1 and MERS-CoV can aid in identifying which structures constitute suitable targets for nAbs, despite the fact that the RBD appears to be distinct between the three coronavirus species.