The epigenetic signature in muscle of critically ill patients: relation with muscle weakness and impact of metabolic interventions

Major advances in intensive care medicine have allowed great improvements in the outcome of critically ill patients over the last decades, with nowadays a large proportion of patients surviving the acute insult. However, many of these patients report decreased quality of life after ICU discharge, due to persistence of complications acquired during the ICU stay. As a result, the focus of clinicians and researchers has shifted more towards therapies that may prevent the development or reduce the burden of critical illness-related morbidity. Artificial nutrition is one of the cornerstones of such so-called supportive therapies, which are specifically aimed at enhancing recovery of patients while reducing the risk of ICU-related complications. In the last 10 years, several large RCTs have tested various modalities of the use of artificial nutrition. Strikingly, however, when combining all available evidence, early enhanced feeding was shown not to be beneficial for ICU patients, and was even harmful in some of the trials. Indeed, early enhanced nutrition by supplementing insufficient enteral nutrition with parenteral nutrition (Early-PN) was shown to delay recovery and to increase the risk of acquiring complications like muscle weakness in ICU, as compared with accepting an early macronutrient deficit by not initiating supplemental parenteral nutrition during the first week in ICU (Late-PN). This is an important finding, as ICU-acquired weakness hampers the recovery of critical illness, and also does not recover swiftly and fully after the critical illness itself has resolved. Thereby, ICU-acquired weakness affects mortality and quality of life up to years after discharge from the ICU. Therapies targeting ICU-acquired weakness are currently still scarce and insufficient, and are limited to preventive measures, such as not initiating parenteral nutrition during the first week in ICU. From a mechanistic viewpoint, the increased risk of muscle weakness with Early-PN as compared with Late-PN has been explained by the inability of Early-PN to suppress catabolism and its suppressive effect on autophagy, which is a cellular quality control mechanism that is crucial for recovery of organ failure in critical illness. Nutrition-induced suppression of ketogenesis may also have played a role, as exogenous administration of ketones in mice was shown to protect against sepsis-induced muscle weakness.

The general aim of this doctoral thesis was to gain further mechanistic insight into the observed clinical benefits of accepting an early macronutrient deficit (Late-PN) in critically ill patients as compared with early full feeding (Early-PN), with a focus on ICU-acquired weakness, and to explore novel nutritional strategies for ICU patients, which in the future may aid in preventing the development of ICU-acquired weakness.

In a first part, we demonstrated that Late-PN suppressed the growth hormone axis in critical illness as compared with Early-PN. Serum concentrations of growth hormone and IGF-I were lower and those of IGFBP1 were higher when accepting an early macronutrient deficit, whereas IGFBP3 was not affected. Lower growth hormone concentrations with this intervention were explained by a decrease in its non-pulsatile secretion, whereas pulsatile secretion was unaffected. The further suppression of the growth hormone axis statistically appeared to be an undesirable side effect of Late-PN and can thus be considered a “price to pay” for the benefits of Late-PN as compared with Early-PN, as these benefits were attenuated but not annihilated by its effect on the growth hormone axis. Whether the growth hormone axis recovers after critical illness, whether changes in the growth hormone axis may play a role in the persistence of ICU-acquired weakness and whether pharmacological reactivation of the growth hormone axis in the context of Late-PN may further augment the observed clinical benefits thereof remains to be investigated.

In a second part, we demonstrated that DNA methylation is altered in muscle of critically ill patients as compared with volunteers who had never been critically ill but had similar comorbidities as the patients. DNA methylation alterations occurred at 565 individual CpG sites that were associated with 400 unique genes, many of which were identified as highly relevant for muscle structure, muscle function and/or muscle weakness. We also identified two hypomethylated regions in the promotor regions of the HIC1 and NADK2 genes, which play important roles in muscle regeneration and postsynaptic acetylcholine receptors, and in mitochondrial processes, respectively. These findings suggest that aberrant DNA methylation alterations in muscle of critically ill patients may contribute to the development and the persistence of ICU-acquired weakness, which requires further investigation. Importantly, if this would prove to be the case, therapies targeting aberrant DNA methylation may have great potential in the prevention and/or treatment of ICU-acquired weakness.

In a third part, we completed the first step in the design of a fasting-mimicking diet for the ICU. Indeed, we were able to demonstrate that prolonged critically ill patients develop a metabolic fasting response after a 12-hour nutrient interruption, as illustrated by significant increases in serum concentrations of bilirubin and the ketone body beta-hydroxybutyrate, and decreases in insulin requirements to maintain normoglycemia and in serum IGF-I. The documented fasting response is an important finding, as the ability of showing such a response during critical illness has long been debated, and it allows progression to a next step in designing a fasting-mimicking diet. This intervention, consisting of alternating blocks of 12 hours of feeding and 12 hours of fasting, based on our results, will be tested for its ability to enhance autophagy and ketogenesis. As the present study demonstrated that autophagic markers in blood samples were largely unaffected by fasting in patients and matched healthy subjects, and thus may not be a good surrogate of autophagy at the level of other tissues, autophagy will be studied directly at the level of muscle biopsies. If this follow-up trial would be successful, the ultimate goal of this project is to test the impact of the designed fasting-mimicking diet on clinical outcomes including ICU-acquired weakness in a large multicenter randomized trial.

In a fourth part, we focused on GDF15, which is a cellular stress marker that has been shown to induce aversive responses to nutrition when there is low nutritional need or when macronutrients cannot be metabolized, thus possibly indicating whether a patient is ready for enhanced artificial nutrition. Serum GDF15 concentrations were elevated throughout critical illness, a response which was not affected by randomization to Early-PN versus Late-PN, but was more pronounced in ICU non-survivors than survivors. Higher GDF15 concentrations upon ICU admission were independently associated with worse outcome of critical illness, including a higher risk of ICU-acquired muscle weakness. Higher concentrations of GDF15 were significantly but weakly associated with gastrointestinal intolerance, and GDF15 concentrations upon ICU admission did not appear to discriminate patients at low versus high risk of Early-PN-related complications. Thus, from our findings, the potential of GDF15 as a “ready-to-feed indicator” for the individual patient appeared limited. However, the dramatically elevated GDF15 levels during critical illness may explain the low appetite and gastrointestinal intolerance that are often seen in critically ill patients. Future research should focus on investigating the relation between GDF15 and gastro-intestinal intolerance, and on finding other biomarkers that can be used as ready-to-feed indicator.

In conclusion, this PhD project provided valuable insight with regard to the further optimization of nutritional therapies to improve outcome of critically ill patients, with a particular focus on prevention of ICU-acquired weakness. The research presented in this thesis also raised numerous important new questions for further investigations, with perspectives for further improvement of the care and outcome of critically ill patients.