1 feb Episode 107 Prognostic parameters Part 1

Mon, 02/01/2021 - 21:48

Dear colleagues,

First some news from the ever-changing vaccine front.

  1. News from China: A Chinese colleague sent me news about the Chinese vaccines from open official sources that he was so kind to translate.

These press releases start in Chinese, followed by English.

  • In the first paper from People’s Daily on Jan 39 by Sinopharm, one of the producers of inactivated SARS-CoV-2, we learn that the phase 3 has included over 60,000 participants (of 125 nationalities!).  There were no safety issues and the efficacy (against disease?) was 79.34 %.  They further claim that it is a “complete vaccine that could protect against  “different strains ofb the world (but unclear whether this includes the UK, SA and BR variants)
  • The second paper CCTV Network of today  just states that China has 16 vaccines in development of which 6 are in phase 3, but unclear which ones.  
  • The third paper, also published today, indicates that Cansino will continue with phase 3 trial of an Adeno 5 vector vaccine.  I’m rather skeptical about the success of this particular vaccine as the published phase ½ trial clearly showed high reactogenicity and incomplete seroconversion.  The problem is that Adeno 5 (in contrast to Ad26) is a common serotype, which has been abandoned in the West since the poor results of the HIV STEP trial.  Nevertheless the Russian Sputnik V also uses Ad(, but only as a booster for Ad26.  

 

  1. A discussion in Nature Briefings on how to adapt the next generation of COVID vaccines to the new variants (South-African and Brazilian).

 

  1. A discussion on the relative importance of SARS-CoV-2 transmission by aerosol and via  contaminated surfaces (fomites): clearly most evidence today is in favor of airborne transmission, while fomites usually have too little virus, which is non-infectious.  Nevertheless hand hygiene remains important….  

 

In the episode 107, I want to discuss recent papers  (mostly just published in Jan 2021) related to prognostic markers.  Not just for their clinical usefulness, but also because some of the biomarkers (such as cytokines or molecular pathways) could give clues to novel therapeutic (anti-inflammatory) treatments, which I will discuss in the next episodes….

Large scale studies and meta-analyses

  1. A very comprehensive large scale study by Rishi Gupta in Lancet Resp Med (Ep 107-1) on the International Severe Acute Respiratory and Emerging Infections Consortium Coronavirus Clinical Characterization Consortium (ISARIC4C) score, evaluated on over 70,000 hospitalized COVID patients in the UK.  A similar score for mortality was previously published. They took relatively “simple” parameters:
    • Clinical: age, sex, respiratory rate, breathing room air or receiving extra oxygen; coma score.
    • RX infiltrates: yes or no
    • Lab: presence of nosocomial infections; oxygen saturation; C-reactive protein; lymphocyte count; urea concentration.  

In slide 2 (Fig 1), the relation between the considered parameters with risk on deterioration is shown.

In slide 3 (Fig 4) the relation between mortality and deterioration for several ages in the total cohort is shown and an example of data from 10 individuals how they predict deterioration and mortality.

This tool can be used freely

  • ISARIC4C data access application form see https://isaric4c.net/sample_ access
  • 4C Mortality and 4C Deterioration risk calculators see https://isaric4c. net/risk  

 

  1. A very nice and detailed meta-analysis of lab markers, indicative of severity and mortality, mainly based on early Chinese studies from colleagues at  UCLouvain (Ep 107-2)

Slide 4 shows a long list of markers, but slide 5 (Table 3) indicates which are associated with mortality: elevated IL-6 (interleukin-6, macrophage activation); lymphopenia, raised biluribin (liver dysfunction) and LDH (lactate dehydrogenase, marker of tissue damage). 

Slide 5 provides the parameters that are most discriminative between non-severe and severe disease: elevated procalcitonin and CRP (C-reactive protein, marker of inflammation); lymphopenia and thrombocytopenia; elevated D-dimer (activation of coagulation system) ; elevated LDH and liver transaminases; creatinemia (kidney dysfunction) and creatine-kinase (CK = damage to muscles).

 

  1. The study by Oscar Peiro on admission predictors of mortality (Ep 107-3) is a retrospective single center study on 196 consecutive patients, with expected risk factors for mortality (Table 1 in slide 6).  They studied 4 biomarkers: cardiac troponin I (cTnI), D-dimer, C-reactive protein (CRP)) and lactate dehydrogenase (LDH) and calculated for each the best discriminatory cut-off point for survival/mortality, based on optimal sensitivity and specificity: cTnI (21 ng/L); D-dimer (1112 ng/mL); CRP (10 mg/dL) and LDH (334 U/L). Survival and ROC curves (Slide 7) suggest that cTNI and D-dimers are the best predictors; but CRP and LDH are also very good.

 

  1. The study by Maestre-Muñiz  (Ep 107-4) is similar but in a more rural setting:  over 400 COVID patients with 92 % presenting a comorbidity and COPD as an important one (22 %).  Table 1 (slide 9-10) shows the clinical predictors, with an expected influence of age, cardiovascular disease and COPD. The effect of age, COPD and respiratory insufficiency are further illustrated on slide 11.  The discriminatory lab parameters are shown in Table 2 (slide 12).  Clearly, most significant is the low partial arterial pressure of oxygen (PaO2), thrombocytopenia (lymphopenia only tendency); increased CRP, urea, creatinine and LDH (but not liver enzymes, IL-6 and D-dimers)

 

Cardiac and endothelial markers

 

  1. Xingjuan Shi summarizes a lot of mainly Chinese studies on predictors of cardiac complications (Ep 107-5). 
  • Table 1 confirms that disease severity and mortality has been strongly associated with diabetes, hypertension and cardiovascular disease in most studies (slide 13).
  • Table 2  (slide 14)shows that cardiac injury, arythmias, heart failure and mortality are associated with various cardiac biomarkers, such as
    • NT BNP = N-terminal B-type natriuretic peptide = biomarker of cardiac stress
    • CK-MB = creatine kinase-myocard band= biomarker of myocardial injury
    • hs-Tn1 = high-sensitivity troponin I = biomarker of myocardial injury,

But also more general markers, such as Ferritin (= marker for iron, but also increases with infection); LDH = lactate dehydrogenase = marker for tissue damage; D-dimer = activation of coagulation system

 

  1. In the prospective series of 123 patients in Tuebingen (Ep 107-6), mortality was clearly associated with impaired left ventricular function, residual volume and tricuspid function (slide 15).  Very remarkably, in this series, clinical parameters, such as age, male gender, obesity and various cardio-vascular risk factors ( hypertension, smoking, diabetes…) were not associated with mortality, but many of the expected biomarkers were elevated in the non-survivors: leukocytosis, increased CRP, procalcitonin, liver enzyme AST, Creatine-kinase (CK) as well as the typical cardial markers troponin-1 and NT-pro-BNP.

 

  1. A prospective Italian study (Ep 107-7) takes this one step further and calculates the predictive power of elevated troponin and BNP towards all-cause mortality: the odds ratio of combined hsTNI/BNP increase is 18! (slide 17) In multivariate analysis, this “cardiac marker”  outperformes kidney dysfunction (lowered eGFR) and lymphopenia.  Desaturation was a weak predictor in this series and D-dimer as well as ferritin had no predictive value! (Slide 18).

 

  1. Finally, Vassiliou (Ep 107-8) investigate in a small series of 38 critically ill Greek patients, of whom 10 died, marker of endotheliopathy as possible predictors.  They found the soluble E- selectin, P-selectin; Angiopoetin-2 and soluble ICAM-1 were all clearly associated with mortality , whereas others (Ang-2, VEGF, VE-cadherin and von Willebrand factor) were not. (see slide 19 and 20).  This result is remarkable, as many expected markers of inflammation etc were not significantly different in this small series (see Table 1).    

 

Best wishes,

Guido