Category Archives: Infection

Immunotherapies for cancer in the ICU

The first lecture of the day was given by the expert in the area Elie Azoulay.  He talked through immune therapies that are being used in cancer.

Immune therapy “boosts” the immune system and restores its ability to eradicate cancer cells.  There are loads of different types of immune therapy – but Azoulay focussed on the ones that are of relevance to intensivists – Adoptive Cell Transfer, encapsulating “CAR T-cells” and “checkpoint inhibitors”.

Cancer cells normally find ways to act on checkpoints (molecules on T Cells) to avoid being attacked by the immune system.  Checkpoint inhibitors, drugs like pembrolizumab and vivolumab [act on PD-1] or atezolizumab [acts on PD-L1] activate the immune system to get to work on tumours.  BUT the usual safeguards against autoimmunity within the body are also affected.  Other drugs that target CTLA-4 (such as ipilimumab used in melanoma)  act as a type of “off switch” on T Cells.  And they work in solid organ tumours – particularly in combination the oncology trial results are impressive… This paper is an overview of the field from a couple of years ago as quoted below:

There are lots of these drugs – how might you identify if your patient has received one? Well hopefully it will be abundantly clear from their treatment or oncologist – but they are all monclonal antibodies of course so end in -mab.  Heres a list:

The other type of treatment in this category then is Chimeric Antigen Receptor T Cells (CAR-T cells).  These are the patient’s own T Cells, apheresed, stimulated and expanded and then re-infused.

The treatments and trials of note are Tisagenlecuecel in the ELIANA trial for young people with refectory B Cell ALL and JULIET trial for high grade B Cell lymphoma and Axicabtagene cioleucel for relapsed B Cell lymphoma in the ZUMA-I trial.

The reason these second or third line cancer treatment matters for intensive care though is because of the serious adverse event rate.  All this immune system jiggery pokery comes at the cost of upsetting normal function and some 30 to 40% of patients will get some sort of complication:

So what will we need to do on ICU?

The lists of critical care support is quite long as theses patients can get multi organ failure requiring support! They range from ruling out infection (e.g. LP in neurotoxicity – is it CAR-T related or CNS infection/sepsis?) and admiting for close observation/monitoring, good symptom control/IV fluids through to oxygenation and ventilation for acute respiratory failure, vasopressors and shock treatment and even renal and cardiac support and monitoring.  There are specific treatments – steroids are the mainstay but blocking the cytokines responsible for the cytokine storm (for example with IL6 antagonists /  tacilizumab) and other rescue strategies.

Some of the complications are still not fully understood – for example neurotoxicity might be related to the parenchyma effect of CAR-T cells or might be a break down of the blood brain barrier, and earlier onset cytokine storm seems to lead to worse neurotoxicity – prompting some people to think there is a link.  But its still an area of research…

The current reality in many units is that CART therapy is bringing patients to ICU for reversible pathology, and because CAR T therapy is an exciting area, perhaps perhaps it will expand beyond its current remit in cancer to other conditions… So we need to be ready!

My favourite bit of Azoulays talk today was his patient information leaflet – enjoy!


ICM Experimental 1: Mechanisms of Multi-organ Failure

How to protect and resuscitate the mitochondrium: Role of Metformin and Cyclosporin

(Jean-Charles Preiser)

3 post-injury phases: Early Phase attempts to protect cellular integrity at the expense of functionality

  • changes in macro / microcirculation
  • decreased oxygen consumption
  • decreased energy expenditure
  • metabolic shutdown

Hyperlactataemia closely related to severity of sepsis; in established sepsis it is defect of oxygen utilisation, rather than impaired oxygen transport


Mitochondria in critical illness– more than just a powerhouse – also key roles in cell signalling (ROS), calcium homeostasis, regulation of apoptosis. Stressed mitochondria release DAMPs, further increasing inflammation and systemic toxicity


Metformin directly decreases mitochondrial respiration and increases aerobic glycolysis

Several studies have shown improved outcomes associated with metformin use in ITU patients

Cyclosporin A inhibits opening of mitochondrial permeability transition pore (mPTP), thereby preventing cell death

Rabbit model: cardiac arrest –> resuscitation with either 1. Control or 2. Cyclosporin A or 3. Non-immunosuppressive analogue-Inhibitor of mPTP opening –> organs harvested with functional markers recorded


Cyclosporin A (and non-immunosuppressive analogue-Inhibitor of mPTP opening) – protective effect on liver, kidney and heart, but not lung


  1. Cyclosporine before PCI for acute MI (800 pts): randomisation to bolus of Cyclosporine or Placebo –> disappointingly same primary outcome (death any cause / cardiac compromise) in both groups   ??dose ??timing-related
  2. Out-of-hospital cardiac arrest (800pts): randomisation to bolus of cyclosporine vs placebo – No difference in SOFA scores 24 hrs post-admission   ??dose ??timing-related ??inclusion criteria


Late and Recovery phases post-injury: Long stayers >5 days – decreased mitochondrial biogenesis, dysregulated lipid oxidation –> weakness, muscle inflammation, impaired anabolic recovery


Inhibitors of mitochondrial function may prevent later organ failure if given very early post-injury BUT prolonged mitochondrial dysfunction is not desirable


Recommended reading: Feeding mitochondria: Potential role of nutritional components to improve critical illness convalescence


The role of the Glucocorticoid (GC) receptor in circulatory shock

(Sabine Vettorazzi)


GC production and function is required for efficient response to inflammation – abnormal GC production associated with higher sepsis mortality in animal models and humans

GC binds to Glucocorticoid receptor (GR) –> translocates to nucleus

  • GR monomer –> Represses inflammation
  • GR dimer –> Induces inflammation


GR Dimerisation-deficient mouse model (GRdim) created, to clearly discriminate effects of GR monomers from GR dimers (which are completely absent)

Mice subjected to endotoxin-mediated shock: GRdim mice died faster than wild-type mice

  • more severe lactate acidosis
  • higher inotropic requirement for haemodynamic stability
  • impaired lung compliance
  • increased osteopontin levels (with decreased levels of anti-inflammatory IL-10)


GR Dimer is important for survival during LPS induced toxic shock, at least during the first 6 hours of shock in GRdim mice (only observed for 6 hours with intensive care support)


Why and how the heart fails

(Alain Rudiger)


Four shock states –

  • shock –> cardiac dysfunction, MOF
  • combination of shock states commonly occur


Cardiac dysfunction during Inflammatory Shock:

  • Elevated BNP and Troponin
  • Arrhythmia
  • Systolic and diastolic dysfunction


Is this myocardial injury the reason for cardiac dysfunction?

Little evidence of myocyte necrosis in pts dying of sepsis + impaired contractility is Reversible = functional impairment rather than structural damage


Adaptive myocardial depression when myocardium is at risk

  • reprogramming of genes
  • activation of fetal genes (survival program)
  • decreased energy expenditure to keep cells viable


Complex process of gene activation / down regulation in multiple cell-types at different times


Rat model 6 hours post-onset of faecal peritonitis:  500+ genes up/down regulated

  • –> downstream: signalling cascades e.g. blunted adrenergic cascade / blunted calcium transport
  • –> further downstream: affects electromechanical coupling and decreases myocardial contractility


Close interactions between shock, cardiac dysfunction and MOF

The heart fails as a result of multiple factors:

  • Insufficient preload
  • Excessive RV afterload
  • Arrhythmia, dyssynchrony
  • Diastolic dysfunction and impaired contractility

ICM Experimental 1: Concepts of Critical Illness


Inflammatory balance and imbalance

(Peter Pickkers)


Originally thought that pro-inflammatory phase occurs first, followed by anti-inflammatory phase

Sepsis-induced immunosuppression: from cellular dysfunctions to immunotherapy


Initial inflammatory response NOT predictive of which patients become immunoparalysed later


Broad defects in the energy metabolism of leukocytes underlie immunoparalysis in sepsis.

Acute phase of infection- Warburg effect is seen in immune cells

This is absent in immunoparalysis, but can be reversed by IFN-gamma –> pharmacological intervention possibilities

Mounting overwhelming pro-inflammatory and anti-inflammatory responses early on = poorer outcome (refractory shock, immunoparalysis, late secondary infections and death)


To Attack or Tolerate: Novel concepts in host response

(Michael Bauer)


Innate immune system has memory, related to Warburg metabolism

Tiny doses of beta-glucan / LPS induce state of trained immunity; higher doses can cause immunoparalysis

Central role of Haem / Fe:

  • in the presence of free haem, a benign response turns into profound infection independent of pathogen dose
  • Hemopexin / haem oxygenase binds / removes extracellular free haem, uncontrolled sepsis can be controlled


Potential to manipulate haem response in addition to antimicrobials and anti-inflammatory drugs


Host Stress and Strain

(Mervyn Singer)

Critical illness strain on body analogous to that seen in Hooke’s Law of elasticity

  • within reversible zone –> acute illness and organ dysfunction –> recovery
  • beyond a ‘point of no return’, enters irreversible zone –> death


Less resilient pt: ‘point of no return’ shifts to left à tolerate less strain

Hardy pt: ‘point of no return’ shifts to right à more resistant to death


Intervention strategy to target specific phase of illness



Cardiac Troponin T (NOT as a marker of cardiac event) as an independent predictor of in-hospital mortality in emergency dept pts with suspected infection (de Groot et al., Emerg Med J 2014; 31:882-8)


Pathophysiological responses shouldn’t be simply viewed as disturbances of physiology, but as factors that accommodate the body to the insult

Is ‘shutting down’ an evolutionary mechanism?

–> Permissive hypercapnoea, hypoxaemia, hypotension, oliguria, anaemia etc could be beneficially targeted to help organs reach recovery phase, e.g.

  • ? avoid stressors (catecholamines, pain, severe cardioresp upset)
  • ? avoid arterial hyperoxia or even normoxia
  • ? induce hibernation or suspended animation
  • ? vagal stimulation of anti-inflammatory cholinergic pathway
  • ? beta-blockade +/- alpha-2-agonism
  • ? chill out psychologically
  • Acceptable patient-specific ranges need to be identified; timing, dosing and duration of intervention are Critical

Further reading:

The stress response and critical illness: A review (Cuesta et al., Crit Care Med 2012; 40:3283-9)


Stressing the obvious? An allostatic look at critical illness (Brame et al., Crit Care Med 2010; 38[Suppl.]: S600-7)


Changing the paradigm: Personalised antibiotics dosing for critically ill


(Prof Jeffery Lipman)

Interesting and thought provoking talk about how we might be inappropriately dosing antibiotics in our critically ill patients.

pK changes in ICU patients (also changes during the course of their ICU admission). Hence dosing is complex and studies confirm that it is very variable across ICUs.


IMG_0084 IMG_0087 IMG_0088



pD properties of antibacterials and hence kill characteristics different across the various classes of antibiotics.

IMG_0091 IMG_0093 IMG_0095



Proposed solutions:



The problem is that trials which had looked at continuous vs intermittent dosing of antibiotics e.g. BLINGs and BLISS, have focussed on clinical cure rather than mortality. Watch out for BLING3